Handbook of Hemorheology and Hemodynamics

HANDBOOK OF HEMORHEOLOGY AND HEMODYNAMICS

Biomedical and Health Research Volume 69 Recently published in this series: Vol. 68. Vol. 67. Vol. 66. Vol. 65. Vol. 64.

Vol. 63. Vol. 62.

Vol. 61. Vol. 60. Vol. 59. Vol. 58. Vol. 57. Vol. 56. Vol. 55. Vol. 54. Vol. 53. Vol. 52. Vol. 51. Vol. 50. Vol. 49. Vol. 48. Vol. 47. Vol. 46.

J.-F. Stoltz (Ed.), Mechanobiology: Cartilage and Chondrocyte – Volume 4 R.J. Schwartzman, Differential Diagnosis in Neurology H. Strasser (Ed.), Traditional Rating of Noise Versus Physiological Costs of Sound Exposures to the Hearing T. Silverstone, Eating Disorders and Obesity: How Drugs Can Help S. Eberhardt, C. Stoklossa and J.-M. Graf von der Schulenberg (Eds.), EUROMET 2004: The Influence of Economic Evaluation Studies on Health Care Decision-Making – A European Survey M. Parveen and S. Kumar (Eds.), Recent Trends in the Acetylcholinesterase System I.G. Farreras, C. Hannaway and V.A. Harden (Eds.), Mind, Brain, Body, and Behavior – Foundations of Neuroscience and Behavioral Research at the National Institutes of Health J.-F. Stoltz (Ed.), Mechanobiology: Cartilage and Chondrocyte – Volume 3 J.-M. Graf von der Schulenburg and M. Blanke (Eds.), Rationing of Medical Services in Europe: An Empirical Study – A European Survey M. Wolman and R. Manor, Doctors’ Errors and Mistakes of Medicine: Must Health Care Deteriorate? S. Holm and M. Jonas (Eds.), Engaging the World: The Use of Empirical Research in Bioethics and the Regulation of Biotechnology A. Nosikov and C. Gudex (Eds.), EUROHIS: Developing Common Instruments for Health Surveys P. Chauvin and the Europromed Working Group (Eds.), Prevention and Health Promotion for the Excluded and the Destitute in Europe J. Matsoukas and T. Mavromoustakos (Eds.), Drug Discovery and Design: Medical Aspects I.M. Shapiro, B.D. Boyan and H.C. Anderson (Eds.), The Growth Plate C. Huttin (Ed.), Patient Charges and Decision Making Behaviours of Consumers and Physicians J.-F. Stoltz (Ed.), Mechanobiology: Cartilage and Chondrocyte, Vol. 2 G. Lebeer (Ed.), Ethical Function in Hospital Ethics Committees R. Busse, M. Wismar and P.C. Berman (Eds.), The European Union and Health Services T. Reilly (Ed.), Musculoskeletal Disorders in Health-Related Occupations H. ten Have and R. Janssens (Eds.), Palliative Care in Europe – Concepts and Policies H. Aldskogius and J. Fraher (Eds.), Glial Interfaces in the Nervous System – Role in Repair and Plasticity I. Philp (Ed.), Family Care of Older People in Europe

ISSN 0929-6743

Handbook of Hemorheology and Hemodynamics

Edited by

Oguz K. Baskurt, M.D., Ph.D. Professor and Chairman, Department of Physiology, Akdeniz University Faculty of Medicine, Antalya, Turkey

Max R. Hardeman, Ph.D. Clinical Biochemist, Laboratory for Clinical Hemorheology, Department of Physiology, Academic Medical Center, Amsterdam, The Netherlands

Michael W. Rampling, Ph.D. Honorary Senior Lecturer, Department of Bioengineering, Imperial College, South Kensington, London, UK

and

Herbert J. Meiselman, Sc.D. Professor and Vice-Chairman, Department of Physiology and Biophysics, University of Southern California, Keck School of Medicine, Los Angeles, CA, USA

Amsterdam • Berlin • Oxford • Tokyo • Washington, DC

© 2007 The authors. All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without prior written permission from the publisher. ISBN 978-1-58603-771-0 Library of Congress Control Number: 2007931478 Publisher IOS Press Nieuwe Hemweg 6B 1013 BG Amsterdam Netherlands fax: +31 20 687 0019 e-mail: [email protected] Distributor in the UK and Ireland Gazelle Books Services Ltd. White Cross Mills Hightown Lancaster LA1 4XS United Kingdom fax: +44 1524 63232 e-mail: [email protected]

Distributor in the USA and Canada IOS Press, Inc. 4502 Rachael Manor Drive Fairfax, VA 22032 USA fax: +1 703 323 3668 e-mail: [email protected]

LEGAL NOTICE The publisher is not responsible for the use which might be made of the following information. PRINTED IN THE NETHERLANDS

Handbook of Hemorheology and Hemodynamics O.K. Baskurt et al. (Eds.) IOS Press, 2007 © 2007 The authors. All rights reserved.

v

Foreword The appearance of this Handbook is a timely event, as it is 20 years since the publication of the “Handbook of Bioengineering” [1] that dealt mainly with basic aspects of hemodynamics and hemorheology. Also, in the 1980s and 1990s a number of books were published that focused on clinical aspects of blood rheology [2–5]. In selecting topics for the present handbook the editors have attempted to provide a general overview of both basic science and clinical hemorheology and hemodynamics. Hemorheology and hemodynamics are closely related, the former dealing with all aspects of the flow and interactions of the individual blood cells mostly studied in vitro, the latter with the in vivo relationships among vessel architecture, driving pressure, flow rate and shear stress. The linkage between the in vitro and in vivo research described in the book will be of interest to both basic science and clinical investigators. With respect to hemorheology, the new book successfully updates developments and advances in the flow properties of human blood cells (microrheology). Furthermore, in the chapters on cell mechanics, these flow properties are related to events occurring at the level of the bonds between the interacting corpuscles (platelets and white cells as well as red cells), and between the corpuscles and the vessel wall (molecular rheology). A welcome feature of the handbook is that it includes a chapter on comparative hemorheology, showing that the rheological properties of red cells vary widely among the animal species, thus shedding light on the process of adaptation to a specific environment or lifestyle, and a chapter on neonatal and fetal blood rheology showing the considerable adaptation processes in play at birth and in infancy and childhood. Also dealt with in some depth are the effects of diseases on the mechanical and adhesive properties of red cells and the underlying molecular mechanisms, particularly those found in malaria. A related subject, the damage sustained by red cells due to flow-induced mechanical trauma, is also presented. With respect to hemodynamics, it is evident in the chapters of section III of the handbook that the field has advanced significantly in the last 30 years, particularly with respect to our understanding of microcirculatory blood flow using novel experimental techniques, the latter being the subject of a separate chapter. The handbook closes with chapters on clinical states associated with abnormal blood rheology, including a chapter on the yet controversial subject of rheological therapy. The editors of the handbook have each been active in the fields of bio- and hemorheology for many years, and have published extensively. They have successfully achieved their objective to publish a well-written and well-edited handbook that will be valuable for researchers and students in the field. Shu Chien, MD, PhD Harry L. Goldsmith, PhD [1] R. Skalak and S. Chien, Eds., Handbook of Bioengineering, McGraw-Hill, New York, 1987. [2] S. Chien, J. Dormandy, E. Ernst and A. Matrai, Eds., Clinical Hemorheology, Martinus Nijhoff Publ., Dordrecht, 1987.

vi

[3] G.D.O. Lowe, Ed. Clinical Blood Rheology, CRC Press, Boca Raton, FL, 1988. [4] A.M. Ehrly, Therapeutic Hemorheology, Springer-Verlag, New York, 1991. [5] J.F. Stoltz, M. Singh and P. Riha, Hemorheology in Practice, IOS Press, Amsterdam, 1999.

vii

Preface The fields of hemorheology and hemodynamics are active and expanding areas of research, yet no combined reviews or “handbooks” have been published within the past 20 years. It was thus felt appropriate to attempt such a task. An outlined proposal was submitted to IOS Press and, after cordial telephone and email exchanges, was approved for publication. In planning for this book, the editors realized that it would be impossible to cover in detail the entire field of hemorheology and hemodynamics, and hence that it would be necessary to limit its scope. It was therefore decided to primarily focus on the macro-and micro rheological behavior of blood and its formed elements, on interactions between the formed elements and blood vessel walls, and on the microvascular aspects of hemodynamics; areas such as cardiac hemodynamics and theory for pulsatile flow in large vessels were omitted. Since many aspects of hemorheology and hemodynamics can be affected by disease or a wide variety of clinical states, these areas were deemed relevant as were the hyperviscosity syndromes and therapy for disturbed blood rheology. In addition, discussions of methods in hemorheology and hemodynamics were included to provide a practical framework for studies in these areas. In the search for authors needed to prepare each section, the global nature of the fields was recognized: Australia, Austria, France, Germany, Hungary, Italy, Netherlands, Singapore, South Korea, Turkey, United Kingdom and the United States of America are represented. We sincerely thank every contributor for writing their section, for allowing editorial corrections/modifications, and for accepting “helpful” criticism without threatening the editors with hostile actions. This book would not have been possible without their contributions. We also wish to thank Ms. Rosalinda B. Wenby for her valuable editorial assistance. We hope that we have been successful in reaching our objectives for this book, and that it will be of value to researchers and clinical scientists engaged in hemorheology and hemodynamic studies. In particular, we trust that the book will serve to foster greater cooperative efforts between these fields. It is notable that the field of hemorheology was, in large part, prompted by direct observations of RBC aggregation, “blood sludging”, white cell effects, and disturbed flow in the human retinal microcirculation. These observations have lead to in vitro studies of blood viscosity, cell rheology and aggregation, and blood flow in small tubes and in more complex geometries; studies aimed at understanding relations between in vitro and in vivo hemorheologic and hemodynamic phenomenon are also ongoing. Further collaborative efforts will be of mutual value, and will hopefully lead to improved health for normal individuals and for those with various clinical conditions. O.K. Baskurt M.R. Hardeman M.W. Rampling H.J. Meiselman

This page intentionally left blank

ix

Contents Foreword Shu Chien and Harry L. Goldsmith Preface O.K. Baskurt, M.R. Hardeman, M.W. Rampling and H.J. Meiselman

v vii

I. History of Hemorheology I. History of Hemorheology Michael W. Rampling

3

II. Hemorheology 1. Basic Aspects of Hemorheology Giles R. Cokelet and Herbert J. Meiselman

21

2. Compositional Properties of Blood Michael W. Rampling

34

3. Blood Rheology a. Macro- and Micro-Rheological Properties of Blood Giles R. Cokelet and Herbert J. Meiselman b. Viscoelasticity of Human Blood George B. Thurston and Nancy M. Henderson 4. Cell Mechanics a. Mechanical and Adhesive Properties of Healthy and Diseased Red Blood Cells Brian M. Cooke and Chwee T. Lim

45 72

91

b. Red Blood Cell Aggregation Björn Neu and Herbert J. Meiselman

114

c. Mechanical Properties of Leukocytes and Their Effects on the Circulation Roger Tran-Son-Tay and Gerard B. Nash

137

d. Adhesion of Circulating Leukocytes and Platelets to the Vessel Wall Susan L. Cranmer and Gerard B. Nash

153

5. Pathophysiology of Blood Rheology a. Mechanisms of Blood Rheology Alterations Oguz K. Baskurt

170

b. Hemorheology of the Fetus and Neonate Otwin Linderkamp

191

c. Mechanical Trauma to Blood Marina V. Kameneva and James F. Antaki

206

x

d. Hemorheological Considerations in Stored Blood Transfusion James P. Isbister

228

6. Methods in Hemorheology Max R. Hardeman, Peter.T. Goedhart and Sehyun Shin

242

7. Comparative Hemorheology Ursula Windberger and Oguz K. Baskurt

267

III. Hemodynamics 1. Basic Principles of Hemodynamics Timothy W. Secomb and Axel R. Pries

289

2. Blood Rheology Aspects of the Microcirculation Herbert H. Lipowsky

307

3. In Vivo Hemorheology Oguz K. Baskurt and Herbert J. Meiselman

322

4. Endothelium and Hemorheology Tommaso Gori and Sandro Forconi

339

5. Methods in Hemodynamics Sehyun Shin, Hideyuki Niimi, Max R. Hardeman and Peter.T. Goedhart

351

IV. Clinical Aspects of Hemorheology 1. Hyperviscosity: Clinical Disorders James P. Isbister

371

2. Clinical Significance of Hemorheological Alterations Kalman Toth, Gabor Kesmarky and Tamas Alexy

392

3. Treatment in Clinical Hemorheology: A Current Overview Michel R. Boisseau, Katalin Koltai, Zsolt Pecsvarady and Kalman Toth

433

A Note on the Editors

445

Subject Index

447

Author Index

455

I. History of Hemorheology


Handbook of Hemorheology and Hemodynamics O.K. Baskurt et al. (Eds.) IOS Press, 2007 © 2007 IOS Press. All rights reserved.

3

History of Hemorheology Michael W. RAMPLING1 School of Medicine, Imperial College, South Kensington, London SW7 2AZ, UK.

Introduction If the rate of appearance of publications in the field can be taken as a criterion, hemorheology can be considered as coming of age in the fairly recent past - perhaps forty or so years ago. This relative lateness is due largely to the previous lack of measuring equipment with the required sophistication; a particular problem being the complex nature of blood viscosity and the need for adequate viscometers capable of measuring it. Nevertheless the ease of availability of blood, its dramatic color and its obvious connection to well being have made it a subject of study since ancient times. What is more, many of those ancient studies were of physical properties of blood that have direct hemorheological relevance. So it could be said that hemorheology is one of the oldest of clinical research areas.

1. Ancient history It quickly becomes clear to anyone who has left a tube of blood undisturbed for a significant period that it will separate out into different phases. The very fact that, in these circumstances, movement of blood components is taking place makes it a hemorheological phenomenon, and the phenomenon was well known to the ancients, and probably the first hemorheological phenomenon to be studied. They found that such settled blood generally formed four layers, so the idea developed that blood was composed of four fluids. The top layer, the first of the fluids, became known as the yellow bile or cholera, and is now known to be serum, i.e. the fluid that separates from the blood clot. The next layer was called the mucus or phlegma. This is now referred to as the buffy coat of leukocytes, platelets and fibrin that settles on top of the third layer called the blood or sanguinis. The third layer is actually the packed red cells. Finally, at the bottom is the last layer, i.e. the black bile or melancholia; this is really the red cells that have failed to oxygenate after the blood collection [1]. This separation, and the fact that the rate of separation of the layers is increased in association with disease, was known to the ancients Greeks and they developed it into a diagnostic tool dependent on the relative proportions of the four layers. It would have been a small step for the ancients to have moved from the ideas presented above to an interest in the rate at which the blood layers settle out. Strangely, it was many centuries before proper study of this took place. The person to give it comprehensive investigation and to allow its development into a proper 1

Corresponding author: School of Medicine, Imperial College, South Kensington, London SW7 2AZ, UK; E mail: [email protected]

4

M.W. Rampling / History of Hemorheology

quantitative diagnostic tool was Robin Fåhraeus in the early twentieth century. It became known as the Erythrocyte Sedimentation Rate Test, and we shall return to it later in this chapter.

2. Middle Ages Following on from the idea of the humours in the blood, and their disturbance in disease, it was not a great conceptual jump to the idea that the physician might do something about the imbalance with the aim of alleviating the pathology. The method for doing that was blood letting, which was to become a common therapeutic tool. The bleeding could be induced grossly by lancing a major vessel or, perhaps more subtly, by leaches. As might be expected the efficacy of the treatment was not high and, indeed, often resulted in death of patients already weakened by the disease from which they were suffering. Nevertheless the concept in principle is very close to that of a hemorheological technique that was developed in the twentieth century, i.e. hemodilution, to reduce blood viscosity and assist blood flow [2].

3. Age of Reason The discovery by William Harvey, in the early seventeenth century, of the circulation of the blood was one of the greatest physiological discoveries of all time (Figure 1). Prior to Harvey’s insight, it had been believed that blood ebbed and flowed in the veins and arteries spreading “vital spirit” to all the tissues, and the massive flow from the heart was considered to be involved in the replenishment of the blood that was consumed in the process.[3]. Harvey’s opus magnus, usually referred to as “De Motu Cordis”, was published in 1628 [4]. His insight was based primarily on the observations of the one-way valves in the veins and in and around the heart, the different pressures in the veins and arteries and the effects of ligations on blood flow. They enabled him to hypothesize the idea of blood circulating from the heart through the arteries and returning to the heart via the veins. This was a huge physiological insight and is obviously of great hemorheological significance as, for the first time, the importance of flowing blood became exposed and so the rheological properties of blood became relevant. Harvey’s discovery stands as one of the pivotal points in the history of hemorheology. It is also interesting that, after the discovery, the raison d’etre of blood letting gradually changed from being to readjust the humours to being, rather, to thin the blood and so make blood flow easier. Harvey’s insight can be seen to have been especially acute as he did not have access to microscopy and so, though he could see the arteries and the veins, he could not see the microcirculatory vessels. However, his evidence for a complete circulatory system was so compelling that he assumed that there were “anastomosis of these vessels or pores in the flesh and solid parts permeable to the blood” [5]. It was left to Malpighi in 1660 to be the first to see the microcirculation [6] and to prove microscopically the connection of the network of small vessels between the arterial and venous sides of the circulation.


5

Figure 1. William Harvey (National Portrait Gallery, London)

Figure 2. Antoni van Leeuvenhoek (Whipple Museum of the History of Science, University of Cambridge)

A major problem at this time was that the true nature of blood was still very poorly understood. It was largely looked on as a simple liquid, but that view began to change after Malpighi had seen the red cells, even though he mistook them for fat globules. It was left to Anthoni van Leeuwenhoek (Figure 2) to give the first accurate description of the red cell in 1674 [7]. He was able to see them flowing in the microcirculation and

6


so was the first to confirm Harvey’s postulate about the circulation. He observed the deformation of red cells as they negotiated the capillaries, and the extent of their deformability seems to have amazed him, as he commented in one of his letters that they could elongate “up to three times their original dimensions without break-up”, but he realized that such deformability was necessary for them to negotiate the minute blood capillaries. He also observed red cell aggregation and noticed that it increased during pregnancy and during infections, and postulated that red cells can lose their deformability in certain circumstances and thus cause disease. Perhaps most astonishing of all was his estimate of the diameter of the red cell which he put at 1/3000inch – or 8.5μm [8]. He was clearly one of the founders of the field of hemorheology, and a scientific giant. From now on it was realized that blood had a particulate nature, i.e. it was composed in part of “corpuscles”, and to be complex and an interesting fluid to study. For example, red cell aggregation was investigated by a number to physicians, John Hunter in particular. In the mid 1700’s he reported [9] that “in all inflammatory dispositions …. the red globules become less uniformly dispersed, and their attraction to one another becomes stronger”. So not only did he see the aggregation but he realized that the degree of aggregation increased in association with many pathologies. This is, of course, an area of particular interest to hemorheologists today. However still the white blood cells had not been seen, and their discovery fell to one of Hunter’s students, William Hewson, in 1770 [10]. He also investigated the lymphatics and explained the relationship between the lymph glands and the lymphocytes.

4. Jean Leonard Marie Poiseuille After Harvey’s discovery of the circulation of the blood, physicians began to consider what were the factors that determined the flow of blood in and to different organs. Some invoked nervous influences, some the inherent properties of the walls of the blood vessels and some the flow properties of the blood. A particular advocate of the last, after many microscopic investigations of blood flow in various animals, was Jean Leonard Marie Poiseuille (Figure 3). In order to get clearer views of the factors determining flow in the circulation he simplified the problem and investigated flow in glass tubes. He found that blood was too difficult to use, or to get consistent results from, so ended by performing most of his studies on simpler liquids such as water and alcohol. He published the results of these painstaking studies in the 1840’s [e.g. 11]. Subsequently, Hagenbach took Poiseuille’s data and derived the relationship between the factors determining the rate of flow of a simple liquid through cylindrical tubes that is usually referred to as Poiseuille’s Law [12]. It is summarized in the equation:F/t = (Spr4)/(8Kl) where F/t is the fluid flow rate, p is the driving pressure across the tube, r and l are respectively the radius and length of the tube and K is the viscosity of the fluid. It is of interest that Hagen independently found the same law [13], so the law is sometimes given the name Hagen-Poiseuille. We now know that this equation only applies strictly to Newtonian liquids, but the principles are applicable even to non-Newtonian liquids such as blood. It is also interesting that the difficulties that Poiseuille had encountered in using blood in his ex vivo studies were partly due to its non-Newtonianism, and this in turn means that, though he did not realize it, he was probably the first to react to the


7

non-Newtonian character of blood. Poiseuille’s breakthrough investigations may be considered the real beginning of the science of hemorheology.

Figure 3. Jean Leonard Marie Poiseuille as portrayed on the Poiseuille Medal. The Medal is awarded by the International Society of Biorheology for services to Biorheology. (Prof Shu Chien)

5. Robin Fåhraeus and John Harkness Earlier in this chapter mention was made of the settling of blood and its relevance to the Ancients in showing up the four humours of the blood. There was a renewed interest in the phenomenon once red cell aggregation had been seen under the microscope and it became realized that there was an association between increased aggregation and the rate of settling. Hewson for example showed that defibrinated blood aggregated less and its settling rate was reduced compared to whole blood. It was also known that the blood even of healthy horses aggregated to a remarkable degree and the settling rate was very high. However, a clear view of what was happening here awaited the appearance of Robin Fåhraeus [1] (Figure 4). His interest in the area would appear to have been stimulated by his observations, while working at the Karolinska Sjukhuset in Stockholm in 1917, on the blood from pregnant women, which was already known to settle more quickly than that from healthy non-pregnant women [14]. His publications started soon after [1] and it was his detailed study of the factors affecting the settling of the blood and its variation in sickness and in health that led to the development of the Erythrocyte Sedimentation Rate (ESR) Test as a valuable quantitative diagnostic tool. As old as this test is, and as simple as it is, it is still one of the most common diagnostic tests done world wide. What is more, newer

8


methodologies for measuring the ESR are still being developed and papers on the topic have never stopped being published.

Figure 4. Shu Chien before a portrait of Robin Fåhraeus. Professor Chien was the first recipient of the Fåhraeus Medal that is awarded by the European Society for Clinical Haemorheology for services to Clinical Hemorheology. (Prof Shu Chien)

Fåhraeus made many other contributions to hemorheology, but an area of especial importance resulted from his studies of blood flow in very small diameter glass tubes. These seem to have been stimulated by observations of in vivo flow in microvessels. One observation that particularly impressed him was the flow of blood through the microvessels of the mesentery of the horse. Here, the red cells do not separate at all but flow as a continuous string held together by the forces of aggregation that produce rouleaux, constraining them to flow along the centre of the vessel. This was very different from the granular nature of the flow of human blood that he observed in the nail bed capillaries.[1]. In his subsequent studies he was able to show that the apparent viscosity of blood decreases as the diameter of the tube through which it flows decreases until the diameter became significantly smaller that the red cell when the apparent viscosity starts to rise again, sharply. This is now known as the FåhraeusLindqvist Effect [15]. It is due to the finite size of the cellular compartment of the blood and it becomes significant only when the diameter of the tube through which the blood is flowing is of a similar order of size to that of the cells. In particular it is caused by the “cell reduced “layer near the wall of the tube. The ESR Test that Fåhraeus had done so much to pioneer has the great advantage of simplicity and cheapness – and this is why it is still so widely used, especially in the


9

third world. Nevertheless, it has significant problems, e.g. the necessity for hematocrit correction and the frequency of anomalous results. However, it had been known for a considerable period that the elevated ESR is associated with increase in the degree of rouleaux formation in the blood, and that this in turn was associated with elevated levels of large plasma proteins, especially fibrinogen [1]. So, because the main determinants of plasma viscosity are the large plasma proteins, it was to be expected that there should be some association between plasma viscosity and pathology. Many studies had been made of the relationship of plasma viscosity to individual diseases, but one of the first papers comprehensively to investigate its potential as an alternative to the ESR as a non-specific diagnostic tool to follow the progress of pathology was produced in 1946 by Harkness et al [16]. However the adoption of this new concept in clinical practice was hampered by the lack of a decent plasma viscometer for the clinical setting. This was finally solved when Harkness described such a capillary viscometer in 1963 [17]. It was to go into commercial production and there was a period, especially in England, when plasma viscosity was regularly used as an alternative to the ESR. However, with time more specific biochemical tests for the large proteins that influence plasma viscosity (e.g. fibrinogen) have superseded it and the commercial production of the Harkness Viscometer has ceased. Now, in the developed world at least, plasma viscosity is only used for rare clinical cases, e.g. macroglobulinemia.

6. The Other Pioneers of the Earlier Twentieth Century As was discussed above, Poiseuille was frustrated in his studies of fluid flow in cylindrical tubes. He had wanted to investigate blood but found it too difficult and so produced his ground-breaking research results using simple liquids such as water and alcohol. Nevertheless work on blood did continue, and in the early part of the twentieth century the dependency of the apparent viscosity of blood on the diameter of the tube through which it flows and on its rate of flow was found, as was the fact that it possessed shear-thinning properties [15]. What had been discovered was that blood possessed non-Newtonian characteristics, and that was a property that was going to bedevil the field for a considerable time to come. The problem was that the measurement techniques that were then available disallowed real comparisons of viscometric data to be made between laboratories. In view of these complexities a remarkably brave and forward-thinking study was performed by Whittaker and Winton and published in 1933 [18]. They perfused the isolated hind leg of a dog with blood of various hematocrits, and at various driving pressures and compared the flow rates with those of physiological saline to allow the apparent in vivo viscosity of flowing blood to be calculated. They were able to show that the apparent viscosity was considerably less than that measured on a sample of the same blood by an Oswald-type viscometer. Their data also indicated that there was a logarithmic relationship between viscosity and hematocrit and that, at the driving pressures and flow rates used, the viscosity was almost constant, i.e. the blood behaved as a Newtonian liquid in this in vivo experiment. Although it is now possible to criticize aspects of the investigation, it was nevertheless a major study of in vivo hemorheology. It is a great shame that such studies have been few and far between in the history of hemorheology. Only in the recent past have significant studies in this direction started to appear in anything like adequate numbers. We shall return to this area later.

10


The major development that stimulated the interest in hemorheology that was to burgeon in the second half of the twentieth century was that of viscometers capable of measuring non-Newtonian viscosity in a standardized manner that was reproducible in laboratories across the world. These were originally developed for use on complex liquids of commercial interest, e.g. paints, molten chocolate etc. One of the most notable was the Weissenberg Rheogoniometer [19]. AL Copley, amongst others, made a huge contribution to hemorheology by modifying the instrument to allow its use on blood [20]. For the first time, data from separate laboratories became truly capable of comparison. There followed an inundation of physicists and engineers studying the basic viscometric properties of blood – great names like GW Scott Blair, RL Swank, RE Wells, EW Merrill, but also RL Whitmore who brought much of this work together in a hugely influential book in 1968 [21]. Leopold Dintenfass was another of these early ‘hemo-viscometrists’. He developed his own advanced viscometer, and with it made a huge range of seminal studies in the area. However, perhaps his greatest contribution to the field was his first book [22]. This was much more clinically based than Whitmore’s had been and covered not only basic rheology but also investigations of hemorheological changes associated with a vast range of clinical conditions – the 66 pages of references bear testimony to that fact! A.L. Copley has already been mentioned as one of the investigators who helped bring modern viscometry into hemorheology and he did considerable work of merit afterwards. Even before that, since the early 1940’s, he had been very active in the field, using a rolling ball viscometer of his own design, and had made some of the earliest studies of blood viscosity, yield stress, thixotropy etc [23]. However, perhaps his major contribution to the subject of hemorheology was his tireless promotion of it in everyway possible. He was involved in the establishment of the two journals that promote all things hemorheological, i.e. Biorheology, first published in 1962, and Clinical Hemorheology, first published in 1980. He was also heavily involved in the organizing of the First International Conference on Hemorheology which was held in Reykjavik in 1966. This led ultimately to the establishment of the International Society of Biorheology and its first International Congress was held in Lyon in 1972. These Congresses have since recurred every 3 years. Eventually it was realized that there was a need for a predominantly clinical forum for hemorheologists and so the biennial European Conference on Clinical Hemorheology was born in Nancy in 1979. The final stage in the maturation of the subject was the founding of the International Society of Clinical Hemorheology in 1993 at the First International Conference on Clinical Hemorheology held in Vienna. Such events, together with the arrival of relatively easy to operate commercial viscometers capable of dealing with the vagaries of blood viscosity e.g. from the Wells Brookfield and Contraves companies, were responsible for sucking considerable numbers of clinicians into hemorheological research. Thus the glory days of clinical hemorheology began in the late 1960’s and ran through to the late 1980’s. It was a time of rapid accumulation of research data, facilitated by considerable drug company interest and, therefore, funding. At that time there were considerable numbers of investigators in the US, but they tended to concentrate on the more theoretical aspects of the subject, and groups were active elsewhere, especially in Japan, but the centre of clinical hemorheology was Europe. There were very active groups in Britain and Italy but the majority of the groups were in France and Germany. The number in France was sufficient that for a time they actually had 3 separate societies of clinical hemorheology! The result of all of this activity was the discovery of viscometric abnormalities in a vast array of clinical conditions [24].


11

7. The Glory Days of Hemorheology In the early days viscosity of blood was the main area of study. However, eventually workers in the field began to be interested in the mechanical properties of the red cell, and its deformability in particular. This was triggered by the realization that blood viscosity could be affected by stiffened cells [25], and that the rate of gaseous exchange by the erythrocyte depends on its flexibility and the stirring of its internal hemoglobin as it negotiates the capillaries [26]. The interest was also stimulated by the considerable study of sickle cell anemia at the time – this being a condition where erythrocytes have difficulty negotiating the microcirculation because they become stiffened when deoxygenated [27]. Sickle cell anemia is, of course, an extreme case where reduced red cell deformability becomes a hemodynamic embarrassment; nevertheless it was considered that less extreme conditions could still have circulatory consequences. The result was an interest in the study of erythrocyte deformability in its own right. To begin with, the methodologies for measuring red cell deformability were complex and really only found in high powered research laboratories. Such techniques were based on glass micropipettes, of similar dimensions to the red cell, into which whole red cells were sucked and the aspiration pressure used as an index of the cellular deformability. Alternatively, using even smaller pipettes, part of the red cell membrane was sucked into the pipette and from its radius of curvature and the aspiration pressure an estimate of the detailed mechanical properties of the membrane could be made [28]. These methods were too complex for the routine clinical laboratory. However, simpler methodologies gradually began to be developed, but the most common was based on the use of micropore filters, which were produced by the Nucleopore Company. These filters had pores of a similar size to red cells or small blood capillaries, and simple filtration methods gave indices of erythrocyte deformability [29]. However there was still a problem; the methods were generally manufactured in each laboratory and interlaboratory comparison not adequate. An important breakthrough took place in 1985 when Dormandy et al described the St George’s Filtrometer [30]. This simplified the methodology and became a commercially available product, which appeared subsequently in many laboratories and expanded the rate of presentation of results, in a conforming way. Nevertheless there were still many problems, not least the facts that the presence of white cells tended to make the data inconclusive and the anticoagulant used could affect results. Dormandy was largely responsible for organising frequent workshops, in London, to try to standardize methodology. Ultimately this all led to standardization being imposed from outside in the shape of the International Committee for Standardization in Haematology [31]. Interestingly the influence of the white cell on microfiltration studies was at first seen only as an inconvenience and various methods, both practical and theoretical were developed to try to remove their influence on micropore filtration measurements of erythrocyte deformability. However, eventually the realization that the leukocytes could affect micropore flow reawakened interest in their influence on flow in the microcirculation in vivo. This was remarkable since, as early as the 1930’s, reports were appearing showing that the white cells are stiffer than the red and are capable of plugging microvessel flow [32]. The effects of the leukocytes on micropore flow sparked a burst of activity looking at the mechanics of the white cell using both micropipette and micropore technology. It is now clear that the white cells have deformabilities some three orders of magnitude less that the red and so, even though they are present in healthy blood only in concentrations of about 1 to every 1000 red

12


cells, nevertheless they are roughly equally influential on microvascular flow to the red mass [33, 34]. Once the white cell had become an object of particular interest to hemorheologists, factors of rheological significance, other than its deformability, began to be appear, in particular its ability to adhere to the wall of blood vessels. This is a continuing active area of research [35]. Later it was discovered that the red cell too could under certain circumstance adhere to the walls of blood vessels. An example of this is found in patients suffering from Diabetes mellitus. This is an area especially studied by JeanLuc Wautier, who has shown that the accumulation of RAGE (receptors for advanced glycation end-products) in the membranes of the red cells is probably responsible. Such adhesion can be expected to influence hemodynamics in a deleterious way [36]. One of the factors that pushed the development of clinical hemorheology in the period between the 60’s to 80’s was the commercial development of a number of hemorheologically based pharmaceutical products. Names such as dextran, pentoxyfylline (Trental) and Troxerutine immediately spring to mind. These products were all reputed to have beneficial hemorheological effects of one sort or another. Appropriate preparations of dextran were used to induce hemodilution and improve blood fluidity, and were also thought to reduce red cell aggregation [24]. Pentoxyphylline was reputed to improve red cell [37] and white cell mechanics [38] and Troxerutine to reduce red cell aggregation [39]. There were also more physical techniques appearing in the field with potential hemorheological effects, such as hemodilution, that needed investigation [2, 40. 41]. All of this brought considerable funding into the field to support, at least in part, its research groups, finance its scientific meetings and to fund clinical trials. It had a huge catalytic effect on activity in the area. It is unfortunate that much of this support has now evaporated. In the early days of the modern developments in hemorheology the studies were devoted entirely to the adult, but it was inevitable that the neonate would eventually also be studied and this has shown that fascinating differences exist between the neonate and the adult [42, 43]. Furthermore, once in utero blood sampling techniques had been perfected, the hemorheology of the fetus became the object of study and showed even more extraordinary differences from the adult [44], and has influenced attitudes to in utero blood transfusion. Furthermore, as studies of clinical conditions unique to the neonate expand so it is increasingly being realized that a number of them have hemorheological associations, e.g. neonatal hyperviscosity syndrome [45]. It is interesting, when one looks at science from an historical point of view, to see how topics come into and out of fashion. The study of rouleaux formation was high on the agenda in the 70’s but then interest waned. However, there has been a considerable resurgence of interest in the fairly recent past [46]. This seems to have been triggered by the advent of a new hypothesis to explain the cause of this unusually weak form of cellular aggregation. Previously, the cross-bridging hypothesis had been generally accepted but recently a very different one based on the depletion layer idea has been proposed [47]. In an effort to decide between the hypotheses investigators have been using enzymes to alter the surface properties of the red cell and then to study the effects on the aggregation phenomenon [48, 49]. The conflict between the hypotheses is still on-going. However, the interest in rouleaux formation has spawned a new concept; that of coating poly(ethylene glycol), and related compounds, onto the red cell surface to alter its rouleaux forming potential in a regulated manner [50]. This has two important potential uses, first as a means of altering aggregation tendencies by varying only one parameter in hemodynamics studies thus making their analysis easier and


13

potentially allowing the hemodynamics influence of rouleaux formation to be determined once and for all, second it offers possible therapeutic benefit. Professor Baskurt’s group in Antalya, and their collaborators, have taken full advantage of the former use. They have investigated the effects of red cell aggregation on in vivo hemodynamics, where the aggregation level is altered by coating the cells with poloxamer copolymers, thus the only hemorheological parameter that they vary in their studies is the inherent aggregability of the red cells [51, 52, 53, 54]. Another area on which some work had been done in the past but which had fallen out of favor is that of athletic performance and hemorheology. In the past few years papers in this area have reappeared in earnest. The group from which most recent papers have come is that of Brun [55, 56, 57], but other groups are increasingly getting involved [58, 59]. A unique event in hemorheology took place in 1985 when Dintenfass et al [60] reported on what is believed to have been the first and, so far, the only hemorheological experiment to take place on a space flight. The study was devoted to the question of whether red cell aggregation is affected by zero gravity conditions. The flight took place on a Space Shuttle in early 1985 and the general conclusion seems to have been that the lack of gravity had little, if any, effect on the morphology of the aggregates. An aspect of modern hemorheology has been the steady introduction of newer and more sophisticated methodologies. A particularly good example is the use of optical trapping to study the forces and mechanism involved in red cell aggregation formation [61]. This technology allows the aggregating red cells to be individually manipulated during the aggregation process and the minute forces involved estimated so that the process can be followed in detail. One of the newer instruments on the market, also based on laser technology, is the LORCA (Laser-assisted Optical Rotational Cell Analyser), which allows for rapid automated analysis of red cell deformability and aggregation [62]. Yet another development has been the use of ultrasound backscattering to estimate red cell aggregation [63] It is always good when new groups become involved in hemorheology. So the recent interest from veterinary research groups is to be welcomed. This interest seems to have been triggered by the realization that the significance of hemorheology to human health could extend to other mammals, especially those of economic and environmental importance, e.g. race horses and endangered species in zoos [64]. However hemorheological studies on animals other than the human are not new. In the early 1970’s Shu Chien’s (Figure 4) group reported on a comprehensive study of five different mammalian and four different non-mammalian species [65] and in 1992 Johnn et al published a comprehensive hemorheological comparison of 31 different mammalian species [66]. Furthermore, increasingly workers are using animal models to study in vivo influences of hemorheological changes and the effects of therapeutic regimen [67, 68]. One of the most common of these animal models is the rat for reasons of cheapness, size and relatively similar hemorheology to that of the human adult [69, 70] A continuing problem with our field in the view of the author is that relatively few studies have attempted directly to relate hemorheology to hemodynamics in vivo or in vitro. Too often hemorheological changes ‘found in the viscometer’ are assumed, on the basis of logic, to be of hemodynamic significance. However, what is needed is hemodynamic experiments that prove these hypotheses. Some studies in this direction have, of course, been done. A classic, which has been discussed above, is that of Whittaker and Winton [20] reported as long ago as 1933. Others have artificially

14


altered hemorheological characteristics and studied their effects on hemodynamics both in vitro [71, 72] and in vivo [73, 74, 75, 76, 77, 78]. However this is a very complex area and even today the data is still confusing. For example it is still not clear what influence rouleaux formation has on in vivo hemodynamics. However, the work coming out of Professor Baskurt’s laboratory in Antalya, referred to above [51, 52, 53, 54], is beginning to clarify the situation. There is a crying need for more research groups to become involved in in vivo studies. So the history of hemorheology is a very long one, with a great surge in work in the field between the 1960’s and 1980’s. Since then the areas of particular interest have changed but clinical hemorheology is still very active. An indication of the change and wellness of the subject can be gleaned from looking at the in house journal Clinical Hemorheology. In 1985 only one volume was printed and it contained 11 papers from the UK, 8 from Germany, 7 from France, 4 from the USA, Italy and Austria and 1 each from Japan, Denmark and Greece, i.e. 41 papers in total. This shows the major influence of Western Europe, followed by the US and some contribution from Japan. In 2003 two volumes of the journal were published, but just to take figures from volume 2: there were 6 papers from Germany, 4 from France, 3 from China, 2 each from the USA, Turkey and Hungary and 1 each from Argentina, Egypt, Japan, Switzerland, Portugal, Italy and Poland, i.e. 26 papers. This indicates two things; first, work in the field continues at a similar rate to that in the 80’s and second that it is no longer centered in Western Europe but rather has widened to include most regions of the world. This is also bringing with it newer attitudes and interests, for example the study of hemorheological effects of traditional Chinese pharmaceuticals [79, 80]. So clinical hemorheology is still very much alive and kicking and long may it continue to be so.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]

R. Fahraeus, The suspension stability of the blood, Physiol. Rev. 9 (1929), 241-274. H. Schmid-Schonbein and H Reiger,. Isovolaemic haemodilution. In Clinical Aspects of Blood Viscosity and Cell Deformability. G.D.O. Lowe, J.C. Barbenel and C.D. Forbes, Eds., Springer-Verlag, Berlin 1981 pp 211-226. E.T. McMullen, Anatomy of a physiological discovery: William Harvey and the circulation of the blood, J. Roy. Soc. Med. 88 (1995), 5491-498. W. Harvey, Exercitarti anatomica de motu cordis et sanguinis animalibus. G. Fitzer, Ed., Frankfortam-Main, 1628. W. Harvey, The Circulation of the Blood. Dent & Sons and E.P. Dutton, London, 1923. M. Malpighius and I.I. Epistel, About the lungs, In Classics in Cardiology, Vol 1, FA Willius and TE Keys, Eds., Henry Schumann Dover Public, New York, 1941, pp 89-97. A. van Leeuwenhoek, Microscopical observations concerning blood, milk, bones, the brain, spittle and cuticula, Philosoph. Trans. Roy. Soc. London 9 (1674), 121-128. A. Schierbeek, The Collected Letters from Antoni van Leeuvenhoek. Swets and Zeitlicher, Amsterdam in press. J.F. Palmer, The Works of John Hunter, Longman, London, 1837 A.L. Copley, The history of clinical haemorheology, Clin. Hemorheol. 5 (1985), 765-812. J.L.M. Poiseuille, Recherches experimentales sur la mouvement des liquides dans les tubes de trespetits diametres, Acad. Roy. Sci. Paris 9 (1846), 433-543. E. Hagenback, Uber die bestimmung der zahigkeit einer flusskeit durch den ausfluss der rohren. Ann. Phys. Chem. Poggendorf. 109 (1860), 385-426. G. Hagen, Uber die bewegung des wassers in engen cylindrischen rohren, Ann. Phys. Chem. Poggendorf. 46 (1839), 425-442. A.L. Copley and G.W. Scott Blair, Obituary: Robin Fahraeus 1888-1968, Biorheology 6 (1970), 153154.


15

[15] R. Fahraeus and T. Lindqvist, The viscosity of blood in narrow capillary tubes, Am. J. Physiol. 96 (1931), 562-568. [16] J. Harkness, J. Houston and R.B. Whittington, Plasma viscosity: a clinical test, Br. Med. J. 1 (1946), 268-276. [17] J. Harkness, A new instrument for measuring plasma viscosity, Lancet 2 (1963), 280-281. [18] S.R.F. Whittaker and F.R. Winton, The apparent viscosity of blood flowing in the isolated hindlimb of the dog, and its variation with corpuscular concentration, J. Physiol. 78 (1933), 239-369. [19] G.W. Scott Blair, An Introduction to Biorheology, Chapt 2, Elsevier, Amsterdam, 1974,. pp11-20. [20] R.G. King and A.L. Copley,. Modifications to the Weissenberg Rheogoniometer for haemorheological and other biorheological studies, Biorheology 7 (1970), 1-4. [21] R.L Whitmore, Rheology of the Circulation, Pergamon Press, Oxford, 1968. [22] L. Dintenfass, Blood Microrheology, Butterworth, London, 1971. [23] A.L. Copley, L.C. Krchma LC and M.E. Whitney, Humoral rheology 1 Viscosity studies and anomalous flow properties of human blood systems with heparin and other anticoagulants, J. Gen. Physiol. 26 (1942), 49-64. [24] G.D.O. Lowe, Ed., Clinical Blood Rheology, Vol 2, CRC Press, Boca Raton, 1988. [25] G.D.O. Lowe and J.C. Barbenel, Plasma and blood viscosity, In: Clinical Blood Rheology Vol 1 Chapt 2 G.D.O. Lowe, Ed., CRC Press, Boca Raton 1988 pp11-24 [26] J.A. Sirs. Erythrocyte function and aldosterone, Biorheology 8 (1971), 1-10. [27] J. Stuart and M.W. Kenny, Sickle-cell disease and vascular occlusion. In: Clinical Aspects of Blood Viscosity and Cell Deformability. G.D.O. Lowe, J.C. Barbenel and C.D. Forbes, Eds., Springer-Verlag, Berlin, 1981 pp109-122. [28] M.Paulitscke and G.B. Nash. Micropipette methods for analysing blood cell rheology and their application to clinical research, Clin. Hemorheol. 13 (1993), 407-434. [29] H.L. Reid, A.J. Barnes, P.J. Lock, J.A. Dormandy and T.L. Dormandy, A simple method for measuring erythrocyte deformability, J. Clin. Path. 29 (1976), 855-858. [30] J. Dormandy, P. Flute, A. Matrai, L. Bogar, J. Mikita, G.D.O. Lowe, J. Anderson, S. Chien, E Schmalzer and A. Herschenfeld, The new St George’s Filtrometer, Clin. Hemorheol. 5 (1985), 975-983. [31] International Committee for Standardisation in Haematology, Guidelines for Measurement of Blood Viscosity and Erythrocyte Deformability, Clin. Hemorheol. 6 (1986), 439-453. [32] J.C. Sandison, Contraction of blood vessels and observations on the circulation in the transparent chamber in the rabbit’s ear, Anat. Rec. 54 (1932), 105-127. [33] U. Bragge, R. Skalak and R. Attefors, Granulocyte rheology, Adv Microcirc 7 (1977), 29-48. [34] G.W. Schmid-Schonbein, K-L. P. Sung, H. Tozeren, R. Skalak and S. Chien, Passive mechanical properties of human leukocytes, Biophys. J. 36 (1981), 253-356. [35] J-L Wautietr, G.W. Schmid-Schonbein and G.B. Nash, Measurement of leukocyte rheology in vascular disease: clinical rationale and methodology, Clin. Hemaorheol. 21 (1999), 7-24. [36] M.-P. Waurtier, T. Khodabandehlou, C. Le Devehat and P.-L. Wautier, Modulation of RAGE expression influences the adhesion of red cells from diabetic patients, Clin. Hemorheol. 35 (2006), 379386. [37] S. Saki, A.Y. Kikuchi. A.N. Ohshima and M. Hori, Effects of pentoxifylline and prostaglandin E1 on cold blood perfusion studied by red blood cell deformability, Angiology 38(1987), 514-519. [38] A Matrai and E. Ernst, Pentoxifylline improves white cell rheology in claudicants, Clin. Hemorheol. 5 (1985), 483-492. [39] AGlacet-Bernard, G. Coscas, A. Chabanel, M. Samama and F. Lelong. Agregation erythocytaire et occlusion veineuse retinienne. Etude preliminaire a l’evaluation d’un traitement correcteur rheologique (Troxerutine), In : Hemorheologie et Agregation Erythrocytaire, Vol 4, J-F Stoltz, Ed., EMI, Cachan, 1992, pp. 222-227. [40] S. Montefusco, A. Gnasso, N. Scarpato, P.Rubba, G. Nappi, C. Cortese, G. Pandolfi and A. Postiglione, Hemorheological effects of LDL-apheresis in Familial Hypercholesterolemia, Clin. Hemorheol. 9 (1989), 81-88. [41] B. Walzl, A. Haas, M. Walzl, J. Faulborn, G.E. Sochor, H. Eckhardt, J. Bergloff and H. Lechner, First experiences with heparin-induced extracorporeal LDL precipitation (HELP) in oculary microcirculatory disturbances, Clin. Hemorheol. 14 (1994), 45-52. [42] O. Linderkamp, P.Y.K. Wu and H..J Meiselman, Deformability of density separated red blood cells in normal newborn infants and adults, Pediatr. Res. 16 (1982), 964-968. [43] M.A. Anwar, M.W. Rampling, S. Bignall, and R.P.A. Rivers, The variation with gestational age of the rheological properties of the blood of the new-born, Br. J. Haematol. 86 (1994), 163-168. [44] R. Welch, M.W. Rampling, M.A. Anwar, D.G. Talbert and C.H. Rodeck, Changes in haemorheology with fetal intravascular transfusion, Am. J. Obstet. Gynecol. 170 (1994), 726-732.

16


[45] J.H. Drew, R.L. Guaran, M. Cichello and J.B. Hobbs. Neonatal whole blood hyperviscosity: The important factor influencing later neurologic function is the viscosity and not the polycythaemia, Clin. Hemorheol. 17 (1997), 67-72. [46] M.W. Rampling, H.J. Meiselman, B. Neu and O.K. Baskurt, Influence of cell-specific factors on red cell aggregation, Biorheology 41 (2004), 79-90. [47] H..J. Meiselman, Red blood cell role in RBC aggregation – 1963 -1993 and beyond, Clin. Hemorheol. 13 (1993), 575-592. [48] Z. Wen, W. Yao, L. Xie, Z. Yan, K. Chen, W. Ka and D. Sun, Influence of neuraminidase on the characteristics of microrheology of red blood cells, Clin. Hemorheol. 23 (2000), 51-58. [49] N.Z. Ertan, M.W. Rampling, M.J. Pearson, C.P. Winlove, P.M. Gribbon and D. Denison, Time dependent effects of trypsin treatment of human red cells, Clin. Hemorheol. 13 (1993), 473-480. [50] J.K. Armstrong, H.J. Meiselman, R.B. Wenby, R.B. and T.C. Fisher, Modulation of red blood cell aggregation and blood viscosity by the covalent attachment of pluronic copolymers, Biorheology 38 (2001), 239-247. [51] O. Yalcin, M. Uyuklu, J.K. Armstrong, H.J. Meiselman and O.K. Baskurt, Graded alterations of RBC aggregation influence in vivo blood flow resistance. Am. J. Physiol. 287 (2004), H2644-H2650. [52] O.K. Baskurt, O. Yalcin, S. Ozdem, J.K. Armstrong and H.J. Meiselman, Modulation of endothelial nitric oxide synthase expression by red blood cell aggregation, Am. J. Physiol. 286 (2004), H222H229. [53] O.Yalcin, H.J. Meiselman, J.K. Armstrong and O.K. Baskurt, Effect of enhanced red blood cell aggregation on blood flow resistance in an isolated-perfused guinea pig heart preparation, Biorheology 42 (2005) 511-520. [54] O. Yalcin, F. Aydin, P. Ulker, M. Uyuklu, F. Gungor, J.K. Armstrong, H.J. Meiselman and O.K. Baskurt, Effects of red blood cell aggregation on myocardial haematocrit gradient using two approaches to increase aggregation, Am. J. Physiol. 290 (2006) H765-H771. [55] E. Varlet-Marie, A. Gaudard, J-F Monnier, J-P Micallef, J. Mercier, F. Bressolle and J-F Brun, Reduction of red blood cell aggregability during submaximal exercise: relationship with fibrinogen levels, Clin. Hemorheol. 28 (2003), 139-150. [56] E. Varlet-Marie, A. Gaudard, J. Mercier, F. Bressolle and J-F Brun, Is the feeling of heavy legs in overtrained athletes related to impaired haemorheology? Clin. Hemoreheol. 28 (2003). 151-160. [57] S. Khaled, J-F Brun, J-P Micallef, L. Bardet, L. Cassanas, G. Monnier and J.F. Orseetti. Serum zinc and blood rheology in sportsmen (football players), Clin. Hemorheol. 17(1997), 47-58. [58] A.V. Muravyov, S.V. Draygin, N.N. Eremin and A.A. Muravyov, The microrheological behaviour of young and old red blood cells in athletes, Clin. Hemorheol. 26 (2002), 183-188. [59] A. Temiz. O. Yalcin, H. Resmi and O.K. Baskurt, Can white blood cell activation be one of the major factors that affect hemorheological parameters during and after exercise, Clin. Hemorheol. 26 (2002), 189-194. [60] L. Dintenfass, P.D. Osman and H. Jedrzejczyk, First haemorheological experiment on NASA space shuttle ‘Discovery’ STS 51-C: aggregation of red cells, Clin. Hemorheol. 5 (1985) 917-936. [61] P.J.H. Bronkhorst, J. Grimbergen, G.J. Brakenhoff, R.M. Heethaar and J.J. Sixma, The mechanism of red cell aggregation investigated by means of direct manipulation using multiple optical trapping,. Br. J. Haematol. 96 (1997), 256-258. [62] R. Hardeman, P.T. Goedhart, J.G.G. Dobbe and K.P.LLetterina, Rotational Cell Analyzer L.O.R.C.A.) I A new instrument of laser-assisted optical measurement of various structural hemorheological parameters, Clin. Hemorheol. 14 (1994), 605-618. [63] M. Boynard, J.C. Lelievre and G. Guillet, Aggregation of red blood cells studied by ultrasound backscattering, Biorheology, 24 (1987) 451-461. [64] U. Wendberger, R. Plasenzotti and Th. Voracek, The fluidity of the blood in African elephants (Loxodonta africana), Clin. Hemorehol. 33 (2005), 321-326 [65] S. Chien, S. Usami, R.J. Dellenback and C.A. Bryant, Comparative hemorheology - hemorheological implications of species differences in blood rheology, Biorheology 8 (1971), 35-58. [66] H. Johnn, C. Phipps, S. Gascoyne, C. Hawkey and M.W. Rampling, A comparison of the viscometric properties of the blood of a wide range of mammals, Clin. Hemorheol. 12 (1992), 639-647. [67] H. Baumler, B. Neu, R. Mitlohner, R. Georgieva, H.J. Meiselman and H. Kiesewetter, Electrophoretic and aggregation behavior of bovine, horse and human red blood cells in plasma and in polymer solutions, Biorheology, 38 (2001), 39-51. [68] O.K. Baskurt and H.J. Meiselman, Susceptibility of equine erythrocytes to oxidant–induced rheological alterations, Am. J. Vet. Res. 10 (1999), 1301-1306. [69] X. Nie, Z-Y. Wen, Z-Y Yan, L. Huang , D. Sun and B. Cheng, Effects of morphine on rheological properties of rat red blood cells, Clin. Hemorheol. 22 (2000), 189-196.


17

[70] A. Temiz, O.K. Baskurt, C. Pekcetin, F. Kandemir and A. Gure, Leukocyte activation, oxidant stress and red cell properties after acute, exhausting exercise in rats, Clin. Hemorheol. 22(2000), 253-260. [71] W. Reinke, W. Gaehtgens and P. C. Johnson, Blood viscosity in small tubes: effect of shear rate, aggregation, and sedimentation, Am. J. Physiol. 253 (1987), H540-H547. [72] A.A. Palmer and W. Jedrzejczyk,. The influence of rouleaux on the resistance to flow through capillary channels at various shear rates, Biorheology 12 (1975), 265-270. [73] S. Simchon, K-M Jan and S. Chien, Influence of reduced red cell deformability on regional blood flow, Am. J. Physiol. 253 (1987), H898- H 903. [74] M. Cabel, H.J. Meiselman, A.S. Popel and P.C. Johnson, Contribution of red cell aggregation to venous vascular resistance in skeletal muscle, Am. J. Physiol. 272 (1997), H1020-H1032. [75] R. Greene, J.M.B. Hughes, I.D. Iliff and G.F. Pineo, Red cell flexibility and pressure-flow relations in isolated lungs, J. Appl. Physiol. 34 (1973), 169-175. [76] G.A. Pantely, L.J. Swenson, C.H. Tamblyn, G.Y.F. Seaman, C.G. Anselone, W.B. Johnson and I.D. Bristow, Increased vascular resistance due to a reduction in red cell deformability in the isolated hind limb of swine, Microvasc. Res. 35 (1988,) 86-100. [77] M.P. Doyle, W.R. Galey and B.R. Walker, Reduced erythrocyte deformability alters pulmonary haemodynamics, J. Appl. Physiol. 67 (1989), 2593-2599. [78] H.H. Lipowsky and J.C. Firrell, Microvascular hemodynamics during systemic hemodilution and hemoconcentration, Am. J. Physiol. 250 (1986), H908-H922. [79] Yu Zhui, Ou-Yang Jing-Ping, Liu Yongming, Wei Lei, Tu Shuzheng, Yang Hailu, Zheng Hanqiao and Yan Xiaohomg, Experimental study of the antiatherogenesis effect of Chinese medicine angelica and its mechanisms, Clin. Hemorheol. 22 (2000), 305-310. [80] F. Liao and B.Li, Inhibition of shear–induced platelet aggregation by Chinese herbal medicines, Clin. Hemorheol. 17 (1997), 315-3180.


II. Hemorheology



21

Basic Aspects of Hemorheology Giles R. COKELETa,1 and Herbert J. MEISELMANb Department of Chemical and Biological Engineering, Montana State University, Bozeman, MT 59717, USA and bDepartment of Physiology and Biophysics, Keck School of Medicine, University of Southern California, Los Angeles, CA, 90033, USA. a

Introduction In analyzing blood flows, one is generally interested in how the blood responds to forces (e.g., pressure gradients, shear stresses). The general fluid mechanical procedure used to predict how a fluid flows in response to forces involves three steps: (1) Consideration of all the forces being exerted on an infinitesimally small volume of fluid. This is done by use of the physical principle known as the conservation of momentum, and results in equations which relate the forces to velocity gradients. (2) Introduction of rheological (“constitutive”) equations which are specific to the fluid being analyzed. These equations indicate how the fluid responds to forces, and relate the forces to the resulting velocity gradients. The rheological equations contain fluid specific characteristics (e.g., apparent viscosity as a function of shear rate). (3) Substitution of the rheological equations into the conservation of momentum equations, and integration of the resultant differential equations to obtain macroscopic relationships, such as between flow rates and pressure gradients. In this chapter, this procedure will be illustrated for specific applications to blood.

1. The Continuum Model All fluids consist of particles (molecules, cells, etc.). Analyzing a fluid flow by keeping track of each and every particle in the fluid is an impossible task. To reduce the analytical work to a reasonable level, we use a model of the fluid which ignores the particulate nature of the fluid and replaces the real fluid with a model fluid whose properties vary smoothly and continuously in space. This is the continuum model. Of course, we must be aware of when this model does not adequately represent the real fluid. The limits of applicability of the continuum model are determined empirically. One type of limit is determined by measuring a property of the fluid (e.g., density), using smaller and smaller fluid sample volumes, until it is determined that the fluid properties no longer are independent of sample size. Alternatively, one can determine the macroscopic characteristics of the flow (e.g., pressure drop versus flow 1 Corresponding author: Department of Chemical and Biological Engineering, Montana State University, Bozeman, MT, USA; E mail: [email protected]

22

G.R. Cokelet and H.J. Meiselman / Basic Aspects of Hemorheology

rate in a tube) in smaller and smaller flow channels, until the measured flow properties do not obey the predicted relationship derived by the fluid mechanical procedure described in the Introduction. The limit of the applicability of the continuum model for suspensions is generally taken to be 20-30 times the particle major dimension; this limit for blood is discussed in Chapter II.3.a.

2. Rheological Concepts Since the rheological equations for the response of a specific fluid to forces involves fluid velocities, a coordinate system must be defined so that we can measure distances and velocities. Here, we consider coordinate systems that do not move relative to the local earth’s surface. Three types of coordinate systems are useful in looking at various fluid flows: (1) a rectangular system (2) a cylindrical system (3) a spherical system. Which coordinate system is used for a particular flow situation depends on the flow boundary shape and the directions of the non-zero velocity components. Thus, for laminar, steady flow between parallel plates where only one velocity component is non-zero, one uses the rectangular system. For flows in cylindrical tubes or between concentric cylinders viscometers, one uses the cylindrical coordinate system, while for flow in a cone-and-plate viscometer, one uses the spherical system. After selecting an appropriate coordinate system, one then proceeds with the fluid mechanical procedure described in the Introduction. This will be illustrated for an analysis in the rectangular coordinate system. To obtain the force-response relationships, one considers an infinitesimally small volume of the fluid, whose shape is defined by the chosen coordinate system and where the faces of the small volume must all be perpendicular to the coordinate directions. In the rectangular coordinate system, such an elemental volume would be a cube or rectangular parallelepiped, whose edges are oriented parallel to the coordinate axes, and which have lengths of įx in the x-direction, įy, and įz in the y and z directions. The forces acting on this elemental volume of fluid are then considered. In Figure 1, a force acting on the front face of the elemental volume is shown; it is not generally perpendicular to the face on which it is acting. This force, which is a vector, is broken into three components that are aligned with the coordinate system axes as shown. Forces Fx and Fz act in the plane of the face, and when divided by the area of the face are called shear stresses. Fy is perpendicular to the face, and when divided by the area of the face is known as a normal stress, or pressure. On the top face, another vector force per face area has been decomposed into its three stress components. Each stress is identified by two subscripts, the first of which indicates the face that the stress is operating on, with the face identified by the coordinate system axis perpendicular to the face, and the second identifies the stress direction. Thus, Ĳzx is the stress operating on the face which is perpendicular to the z-direction, and which operates in the x-direction.


23

Figure 1. An elemental volume of fluid, subjected to some forces. The bold force vector acting on the front face is shown decomposed into its three components, Fx, Fy and Fz, acting on that face.

In general, each face of the elemental volume has three stresses acting on it, and so the elemental fluid volume has a total of 18 stresses on it; there is also a gravitational or body force, but its effect is generally considered negligible when studying blood flow. If we imagine that the elemental volume shrinks to a point, so that we can evaluate the stresses at a point, we will have nine different stresses: three normal stresses with one for each coordinate direction, and six shear stresses, two in each plane perpendicular to a coordinate direction. By a mathematical argument, shear stresses which differ only in the order of their subscripts (e.g., Ĳxy and Ĳyx) are equal. Consequently, at any point in the continuum fluid, only three normal and three shear stresses are independent and need to be specified. Having defined the forces acting at a point in the fluid in terms of stresses, we now have to define the fluid’s response to these stresses. Any motion of an infinitesimal fluid volume can be decomposed into one or more basic motions: (1) Pure translation (i.e., movement relative to the coordinate system, without rotation, shape change or volume change) (2) Pure rotation (3) Angular distortion (i.e., a shape change without a volume change) (4) Volumetric distortion. Of these motions, only two are caused by stresses: a volumetric distortion caused by normal stresses, and an angular distortion arising from shear stresses. Since blood is considered to be an incompressible fluid, only the shear stresses need to be considered. One means of measuring the rates of angular distortion is to measure the rates of changes of the angles between adjacent faces of the infinitesimal fluid volume. This is illustrated, for the two-dimensional case, in Figure 2, where a measure of the rate of angular distortion could be the rate of change of the angle Ȗ. Since Ȗ = 1800 – Į – ȕ, the time rate of change of Ȗ is given by

wJ wt

wD wE wt wt

24


where “t” denotes time. We now relate these angular changes to the appropriate velocity components.

Figure 2. Angular distortion of a 2-dimensional fluid element due to shear stresses.

In Figure 2, the lower left corner of the volume moves with velocity components ux and uy. However, the lower right corner may move with slightly different velocities:

ux

wux dx and uy wuy dx . wx wx

Thus, in the time increment when the volume moves from the lower to the upper position, the volume edge originally parallel to the x-axis will become oriented in a new direction, at an angle dĮ relative to the coordinate axis x. The relative displacement, dy, in the y-direction is

(uy

wuy dx ) (uy ) wx

wuy dx . wx

For small angles, the tangent of the angle is equal to the angle in radians, so

dD

dy dx

wuy dx wx dx

By similar reasoning dE .

§ wu · ¨ ¸ © wx ¹

wux dJ . Therefore, wy dt

(

. wux wuy ) { J ; wy wx

where J is defined as the shear rate in the x – y plane. Shrinking the elemental volume to a point, there are three possible angular distortions, one in each coordinate plane through the point. For a rectangular coordinate system, these are

25

G.R. Cokelet and H.J. Meiselman / Basic Aspects of Hemorheology .

Jxy .

Jxz .

Jyz

wux wuy wy wx

(1a)

wux wuz wz wx

(1b)

wuy wuz . wz wy

(1c)

For cylindrical coordinates, the shear rates are .

Jrx .

Jr\

.

Jx\

wur wux wx wr

1 wur wu\ u\ r w\ r wr 1 wux wu\ . r w\ wx

(2a)

(2b)

(2c)

For spherical coordinates .

JrI

. r\

J

.

JI\

uI ) 1 wur r r wr r wI w(

u\ w( ) 1 wur r r r sin I w\ wr

sin I r

w(

u\ ) 1 wuI sin I wI r sin I w\

(3a)

(3b)

(3c)

In many situations, where the flow is steady, laminar flow of an incompressible fluid, these equations are greatly simplified because there is only one non-zero velocity component: ux in flow between parallel plates and in cylindrical tubes, u\ in a concentric cylinder viscometer, and uI in a cone-and-plate viscometer. Thus, for these simple flows:

26


.

Jxy

Jrx

dux dy

dux dr

. u\ Jr\ du\

dr

.

JrI

r

uI ) r r dr

parallel plates

(4a)

cylindrical tube

(4b)

concentric cylinders viscometer

(4c)

cone-and-plate viscometer

(4d)

d(

All the other shear rates for these simple flows are zero. Now that we have measures of the rates of deformation (i.e., shear rates) at a point in any fluid, we must now introduce the unique flow (i.e., rheological) properties of the fluid into our analysis in order to incorporate the shear stresses which cause the deformations. These flow properties are expressed in the constitutive or rheological fluid equations.

3. Constitutive Equations The rheological properties of fluids are measured in viscometers specially designed so that only one velocity component exists and does not vary in the direction of flow. In addition, except for tube viscometers, the fluid space in the viscometer is small enough so that all of the fluid is subjected to approximately the same shear rate. General details of viscometer operation can be found in Van Wazer, Lyons, et al. [1], and special details of determining blood’s rheological properties can be found in Cokelet [2] and Meiselman and Cokelet [3]. The rheological properties of a fluid may be divided into two classes: timeindependent and time-dependent. Since the time-dependent behavior of blood is discussed in Chapter II.3.b. and because blood’s rheological properties are timeindependent in almost all in-vivo and in-vitro flows, we restrict the discussion here to time-independent rheological properties. The relationship between the shear stress and the shear rate must be determined experimentally. To be most useful, this relationship must be expressed as a mathematical equation. Some such equations are the result of curve-fitting experimental data, while others are based on a model of the fluid. Some useful empirical relationships are given here:

W

.

K J

for Newtonian fluids, where K is the viscosity and a constant at all shear rates of interest.

27

G.R. Cokelet and H.J. Meiselman / Basic Aspects of Hemorheology .

W

a J b

W

aJ

for W t b . Here “a” and “b” are constants, with “b” being the minimum shear stress needed to cause flow, known as the yield stress. This is the Bingham fluid model.

n

where “a” and “n” are constants. This is the power law fluid model

It is common to define an “apparent viscosity” by the relationshipKa

{

W . Except J

for Newtonian fluids, it is a function of shear rate. Three equations, based on models of suspensions, are: Here Ko is the suspending fluid Newtonian viscosity, “k” is a constant dependent on particle shape, and H is the volume fraction of the suspension occupied by particles. This equation is applicable for suspensions having a low volume fraction of particles. Einstein showed that k=2.5 for spherical particles.

Ka Ko(1 kH )

W 1/ 2

aJ

1/ 2

b

1/ 2

where “a” and “b” are constants. It is known as the Casson [4] equation; at very low shear rates, b is the yield shear stress. However, for blood, the experimental data can not be fit over all shear rates with only one set of constants “a” and “b”, whereas fairly good fit is possible by applying the equation over several shear rate ranges and thereby obtaining several sets of constants.

Ka Ko (1 0.5kH )2

where

k

k o k f J r1 / 2 1 J r1 / 2

and

Jr {

J Jc

with

k 0, kf , and Jc being constants. This equation, known as the Quemada [5-7] equation, accurately fits blood data over a very wide range of shear rates. The use of the Casson equation to represent blood’s rheological properties is discussed further in Chapter II.3.a. 4. Applications Given the angular distortion or shear rate equations (1a – 4d) and a suitable constitutive equation for a specific fluid, it is now possible to derive an equation using easily measured macroscopic properties of a viscometric fluid flow (e.g., pressure drop versus flow rate in tube viscometers) to determine the coefficients in the constitutive equation.

28


This will be illustrated for the Newtonian fluid model that includes the Einstein type equation above. 4.1. Tube Viscometer For steady, uniform, laminar flow of an incompressible fluid in a cylindrical tube, the appropriate shear rate equation is equation (4b) above:

Jrx

wux wr

(5)

And the constitutive equation for a Newtonian fluid is:

Wrx

KJrx

(6)

Substitution of equation (5) into equation (6) gives:

Wrx

K

wux wr

(7)

Inspection of this equation indicates: 1) for conditions at the centerline where r equals 0, Wrx 0 based on symmetry of the velocity profile about r=0; 2) at the wall, Wrx Wo , which is not zero because, with ux=0 at r=r0, there must be a velocity gradient there. Consequently, we see that Wrx is a function of r, with the relationship between Wrx and r obtained by the application of the principle of the conservation of momentum.

Figure 3. A small fluid shell for the case of flow though a cylindrical tube.

Figure 3 shows a very small “sleeve” of fluid. From equation (7) we see that there is a non-zero shear stress wherever the velocity gradient is non-zero. Consequently, there is a shear stress, Wrx , on the inner cylindrical surface of the elemental volume, and also one on the outer surface, Wrx

wWrx dr , which may differ from Wrx because the wr

velocity gradient may be different on each of the two cylindrical surfaces. Likewise,

29


the pressure at axial positions x and x+dx may be different. Because the flow is uniform, (i.e., the velocity does not change in the flow direction), the momentum added to the elemental volume by flow into the volume at axial position x is balanced by the flow out of the volume at the axial position x+dx. Consequently, the x-component of the momentum equation is:

2SrdrP 2Srdr ( P

wP dx) 2SrdxWrx (2S (r dr )dx)(Wrx dWrx dr ) dr wx

0.

Simplification reduces this equation to

wP wx

dWrx dr

(8)

w 2P If this equation is differentiated with respect to x, we see that 0 and wx 2 wP a constant. Use of the r-component of the momentum equation shows therefore wx wP dP which is a constant that P does not vary in the r direction, and thus wx dx everywhere. If equation (8) is now integrated, we obtain:

dP r dx

Wrx kI

(9)

where kI is a constant of integration. By symmetry of the velocity profile, r=0, so kI=0. Combining equations (7) and (9) yields:

dP r dx

K

Wrx

0 at

dux dr

Separation of variables and integration gives:

dP r 2 dx 2

Kux kI .

At r=r0, ux=0, so the velocity profile is given by:

dP r 0 2 r 2 ( ) 2 dx 2

Kux

which is the well-known parabolic velocity profile. To obtain the total flow rate, Q, one must integrate the equation from r=0 to r=r0: 4

Q

³ 2Sr (dr )ux

dP Sro dx 8K

(10)

30


Equation 10 is, of course, the Hagen-Poiseuille Equation. The Hagen-Poiseuille Equation allows us to determine if experimental data for pressure gradient and corresponding flow rate are fit by the Newtonian fluid flow

dP 'P , which is equal to , versus Q must be linear with dx 'x 8K . Note that to test if the continuum the slope of the line fitting the data equal to r04

equation. That is, a plot of

model was appropriate for fluid flow in a tube of radius r0, one must obtain data with several tube sizes to see if the calculated viscosity is a constant and independent of tube radius. It not, the continuum model has failed for this fluid flowing in the tested tube; this matter is discussed further for the case of blood in Chapter II.3.a. One could, in a manner similar to the above, derive equations analogous to equation (10) for the other constitutive equations. Fortunately, this is not necessary, because the Mooney-Rabinowitz-Weissenberg equation [8, 9] is generally valid for fluids whose rheological properties are not time dependent:

J0

2U (3

d ln U ) d ln W 0

where U = the average velocity divided by the diameter and

J0

= the shear rate at the

tube wall; W o is obtained from equation (9). Thus, by plotting ln U versus ln W 0 , one can obtain the wall shear rate for the corresponding wall shear stress. Note that all of the above equations are only valid for laminar fluid flow. The determinant for laminar tube flow for a Newtonian fluid is embodied in the Reynolds number

Re {

2ro u

K

d 2100 , where u is the average fluid velocity; Re values > 2100

indicate non-laminar flow. 4.2. Concentric Cylinders Viscometer For steady, uniform, laminar flow of an incompressible Newtonian fluid in a concentric cylinders viscometer, equation (9) and the Newtonian fluid constitutive equation give:

Wr\

K (

wu\ u\ ) wr r

u\ ) r Kr wr w(

(11)

With this type of viscometer, one usually measures: 1) the rotational rate of one of the cylindrical surfaces while the other remains stationary; 2) the total torque exerted by the fluid on one of the cylindrical surfaces. The total torque, T, for the case where the torque is measured on the rotating inner cylindrical surface is: T 2SriLiWi where the subscript “i” refers to the inner surface, and Li is the length of the torque measuring surface, usually taken as the fluid gap length. Thus, using equation (11),

31


T

§ u\ · ¨ w( ) ¸ 2Sri 2 Li (Kri¨ r ¸ ) ¨ wr ¸ ¸ ¨ ¹i ©

§ u\ · ¨ w( ) ¸ 2SKri 3 Li¨ r ¸ ¨ wr ¸ ¸ ¨ ¹i ©

If the gap width is very small, it can be assumed, within instrumental precision, that the derivative in this equation is a constant across the gap and equal to with ro being the outer surface radius. In this case, T terms of the angular rotational speed,

T

2SKLi (u\ )i

(u\ / r ) i , ri ro

ri 2 , or, in ( ri r o )

Zi :

2SKLiri 3Zi ( ri r o )

Thus, one can use the measured torque and rotational speed to calculate viscosity. It is necessary to test to see whether the viscometer gap is small enough so that the small gap approximation is valid; this testing is achieved by using various size gaps. If not, one must construct the viscometer with a smaller gap, or utilize equation (11) and the conservation of momentum equation to determine the velocity profile in the gap. Fortunately, Krieger and Elrod [10] developed a method, somewhat analogous to the Mooney-Rabinowitz-Weissenberg equation, for obtaining the shear rate at the rotating surface of a concentric cylinder viscometer, even for non-Newtonian fluids and non-small gaps. For a rotating inner surface, the Krieger-Elrod Equation is

Ji

º (ln s ) 4 d 4Zi d ln Zi (ln s ) 2 d 2Zi 1 ln ...» s « 2 4 ln s ¬ 45Zi (d ln Wi ) 3Zi (d ln Wi ) d ln Wi ¼

Zi ª

where s = ro/ri. Solution of this equation requires graphical determination of the derivatives. For blood, a precision of 1% or better requires three terms of this series for shear rates of 10 s-1 or less, whereas shear rates above 10 s–1 require only two terms. The shear stress on the inside surface can be calculated from the measured torque from the equation:

Wi

T (2Sri 2 Li )

The criteria for laminar fluid flow is again embodied in a Reynolds number for a Newtonian fluid, Re {

ri Z (ro ri ) U

K

.

When the inner surface rotates, Taylor

ro , resulting in incorrect shear stresses. For ro ri either surface rotating, turbulence will occur if Re t 50,000 . vortices will form if Re t 41.3

32


4.3. Cone-and-Plate Viscometer This viscometer employs: 1) a cone with an angle measured from the cone axis to the cone surface of slightly less than 90o; 2) a flat plate. The cone is situated so that its axis is perpendicular to the flat plate and its apex is just slightly separated from the flat plate; in some situations, a small portion of the apex is removed to allow proper separation with the “virtual” apex at the plate surface. Usually, the cone is rotated at selectable constant rates and the resultant torque is measured on the cone; the fluid is in the gap between the cone and the plate. For steady, laminar uniform fluid flow, application of the momentum equation results in:

1 dW TZ W TZ dT

2 cot T

Note that it is assumed that uT

rf (T ) , where f (T ) is only a function of T ;

this assumed velocity satisfies all the boundary conditions. Integration and use of the boundary condition on the flat plate surface gives

W T\

(W T\ ) S ( 2

1 ) sin 2 T

From this equation, it can be seen that the shear stress, as a function of the angle T , varies by 0.12% if the angle in the fluid (i.e., between the cone surface and the flat plate), T L , is 2.0o. Since in most cases the liquid angle is less than 2.0o, one may assume that the shear stress is constant. Thus, the measured torque T is

T

2 ³ 2Sr W T\ dr

where the integration is from r=0 to r=R, the radius of the cone. Rearrangement gives

W T\

3T . 2SR 3

The corresponding shear rate, with the use of .

J T\

Z TL

, where

Z

T L d 2 o , becomes (from equation 4d):

is the rotational rate of the cone.

If the rotational velocity is too high, the fluid flow will have vortices. The criterion for the onset of this condition is again a Reynolds number, defined as Re

R 2ZU

P

. For

a 4o liquid angle, secondary flows are present if Re is greater than 1100; the criterion for onset of vortices formation depends on the fluid angle [11].


33

References J.R. Van Wazer, J.W. Lyons, K.Y. Kim and R.E. Colwell, Viscosity and flow measurement, John Wiley, New York, 1963. [2] G.R. Cokelet, The rheology of human blood, In: Biomechanics: Its foundations and objectives, Y.C. Fung, N. Perrone and M. Anliker, Eds. Prentice Hall, Englewood Cliffs, NJ, 1972. [3] H.J. Meiselman and G.R. Cokelet, Blood Rheology: Instrumentation and Techniques, In: Flow: Its measurement and control in science and industry, R.D. Dowdell, Ed., Instrument Society of America, Pittsburgh, PA, 1974, Vol. 1, Part 3. [4] N. Casson, A flow equation for pigment-oil suspensions of printing ink type, In: Rheology of disperse systems, C. C. Mills, Ed., Pergamon, New York, 1959, pp 84-102. [5] D. Quemada, Rheology of concentrated suspensions and minimum energy dissipation principle, I. Viscosity-concentration relationship, Rheol. Acta. 16 (1977), 82-94. [6] D. Quemada, Rheology of concentrated dispersed systems. II. A model for non-Newtonian shear viscosity in steady flows, Rheol. Acta. 17 (1978), 632-642. [7] D. Quemada, Rheology of concentrated disperse systems. III. General features of the proposed nonNewtonian model. Comparison with experimental data, Rheol. Acta, 17 (1978), 643-653. [8] B. Rabinowitsch, Ueber die viskositaet und elastizitaet von solen, A. Physik. Chemie, A145 (1929), 126. [9] M. Mooney, Explicit formulas for slip and fluidity, J. Rheol. 2 (1931), 210-221. [10] I.M. Krieger and H. Elrod, Direct determination of the flow curves of non-Newtonian Fluids. II. Shearing rate in the concentric cylinder viscometer. J. Appl. Physiol. 24 (1953),134-144. [11] M.E. Fewell and J. D. Hellums, The secondary flow of Newtonian fluids in cone-and-plate viscometers. J. Rheol. 21(1977), 535-565. [1]

See Also: H.A. Barnes, J.F. Hutton and K. Walters, An Introduction to Rheology, Elsevier Science, Amsterdam, 1989. R.G. Larson, The Structure and Rheology of Complex Fluids, Oxford University Press, Oxford, 1998. C.W. Macosko, Rheology: Principals, Measurements and Applications, Wiley-VCH, New York, 1994. A.Y. Malkin and A.I. Isayev, Rheology: Concepts, Methods and Applications, ChemTec Publishing, Norwich, NY, 2005. T.G. Mezger, The Rheology Handbook, Vincent Verlag, Hanover, 2002. F.A. Morrison, Understanding Rheology, Oxford University Press, Oxford, 2001.

34


Compositional Properties of Blood Michael W. RAMPLING1 School of Medicine, Imperial College, South Kensington, London SW7 2AZ, UK

Introduction The function of blood is to feed all the tissues of the body with vital materials and to remove waste. To do this in the human adult it has to traverse the complicated vascular network, which varies in diameter from some 3 cm down to about 5 μm. Furthermore, the blood must circulate above a limiting rate if it is to do its work effectively enough to keep the organism healthy. This rate of circulation is determined by the driving pressure generated by the heart, by the geometrical resistance offered by the vasculature and by the flow properties of the blood. These flow properties are the concern of the hemorheologist and they are dependent on the composition of the blood and the properties of its constituents; hence, knowledge of them is vital to any understanding of hemorheology. This chapter gives an overview of the composition of normal adult human blood and some indication of the ways in which it can be altered in diseased states. There is also discussion of the normal changes in blood composition that take place as the fetus develops through to the neonatal period. Finally, there is a brief review of the variations that occur in other mammals, emphasizing the similarities and the great differences that exist compared with the human adult.

1. Human Adult In the healthy adult human, a little less than half the blood volume is made up of the cellular compartment and the rest is plasma. The plasma is a mixture of many metabolites, proteins and lipoproteins suspended in a complex salt solution. The cellular compartment is also complex, consisting mainly of erythrocytes, or red blood cells, but also of white cells, or leukocytes, and platelets, or thrombocytes. The red cells are essentially all the same in that each is little more than a membranous sack containing a very concentrated hemoglobin solution. Their prime role is to facilitate the transport of oxygen and carbon dioxide around the circulation. The white cells on the other hand are made up of a number of distinct varieties, and they all have complicated interiors, containing organelles, a nucleus, fibres etc suspended in a viscous cytoplasm. They are involved in defense against infection. The platelets are essentially all similar, are much smaller than the other blood cells, 1

Corresponding author: School of Medicine, Imperial College, South Kensington, London SW7 2AZ, UK; E mail: [email protected]

M.W. Rampling / Compositional Properties of Blood

35

are anucleate and have relatively complex contents consisting of vacuoles and fibres suspended in their viscous cytoplasm. They have a role to play in hemostasis. 1.1. Plasma The liquid phase in which the cellular components of blood are suspended is the plasma. It is a complex solution of materials ranging in size from a few tens to millions of Daltons. These solutes make up some 8 to 9% of plasma by weight, the rest being water. 1.1.1. Ions The smallest of the solutes are the ions of the dissolved and dissociated inorganic salts, with molecular weights of a few tens of Daltons and they make up some 1% of the plasma by weight. There are many different ions present in blood, e.g. Na+, Ca++, K+, Cl-, HCO3- and PO4--. However, by far the most concentrated of the cations in Na+ (derived largely from dissociated NaCl) and so it is the most potent from an osmotic point of view. For this reason amongst others, it is necessary that its concentration be tightly controlled. There are potent physiological mechanisms to ensure this, so that in the healthy human the concentration is maintained in the normal range of 135 to 145mM [1]. From a hemorheological standpoint this is very important, because if its concentration rises or falls outside this range, red cells shrink or swell, with this volume change having a profound influence on their mechanical properties and, hence, on their effect on blood viscosity [2]. The other particularly important ion from a hemorheological point of view is the anion HCO3-. It has a normal range of 24 to 30mM [1]. Its importance lies in it being one of the factors involved in controlling blood pH and in maintaining it in the very narrow normal range of 7.35 to 7.45 which is vital to normal bodily function [1]. Again, if the pH deviates far from the normal range, it has deleterious effects on the mechanical properties of red cells and hence on their viscometric effects [3]. 1.1.2. Metabolic Molecules The metabolic molecules generally have molecular weights of a few hundreds of Daltons (e.g., glucose, urea and amino acids) and make up about 1% of the plasma by weight. Their concentrations are generally somewhat less well controlled than the salts mentioned above so, for example, the normal ranges for glucose and urea are 0.7 to 1.0g/l and 80 to 250mg/l (urea nitrogen) respectively [1]. Generally, they have relatively minor hemorheological effect and so no further discussion will be made of them here. 1.1.3. Proteins The plasma proteins are all very large, having molecular weights ranging from a few tens of thousands to millions of Daltons, and make up about 7% of plasma by weight. They are generally subdivided according to their electrophoretic mobility into albumin, Į, ȕ and Ȗ groups. However, this approach masks the wide variety of different proteins that exist in the plasma (e.g., there are carriers of lipids, metals and other factors, many immunoglobulins, clotting factors and fibrinogen). They are necessary to the carriage of many vital materials, to defense against infection, to

36


hemostasis, etc. [1], and are hemorheologically important for two reasons. First, because of their relatively high concentration, their large size and often asymmetrical shapes, they have a large effect on the viscosity of plasma; the viscosity of whole blood is dependent in part on the viscosity of the plasma. The normal range for the viscosity of human plasma is 1.25r0.10 mPa.s at 37oC, while the viscosity of water at the same temperature is 0.69 mPa.s; the difference between the viscosity of water and plasma is due almost entirely to the plasma proteins [4]. In health the concentrations of these proteins are reasonably well controlled but they can vary enormously in diseased states and correspondingly cause plasma viscosity to vary widely [5]. The second reason that they are of hemorheological significance is that some of them cause red cells to stick loosely together in characteristic face-to-face aggregates, like piles of coins, called rouleaux. From a hemorheological point of view rouleaux formation is important because it causes the viscosity of blood to be very dependent on the shear rate to which it is exposed [4]. Thus the viscosity of normal blood is high at low shear rate, but steadily falls as the shear rate increases and the shear forces increasingly disperse the rouleaux (see Chapter II.3.a). This phenomenon is known as shear thinning. There is no doubt that the protein with the greatest effect in this respect is fibrinogen, and while its normal range in health is usually taken as 3.5r0.4 g/l [6], it can rise substantially in association with a wide variety of diseases and hence greatly affect the shear thinning properties of blood [4]. Other large plasma proteins also have rouleaugenic effects (e.g., D2-macroglobulin, IgG and IgM) but they are not as potent as fibrinogen [7]. 1.2. Formed Elements 1.2.1. Erythrocytes In the past the relative volume of the blood that was made up of cells was measured centrifugally. The blood sample was spun in a uniform cylindrical tube, with the ratio of the length of the cell column to that of the cells plus plasma termed the hematocrit. Of course, strictly speaking, it was not a true hematocrit (i.e., volume fraction of only red cells) as the spun cell column would contain leukocytes and platelets as well as red cells. Nowadays, red cell parameters are commonly determined electronically and the hematocrit is obtained by multiplying the measured red cell count and the mean cell volume. This gives a “true” hematocrit (i.e., not contaminated by other cells). The convention now is to call the electronic value the hematocrit and the centrifugal one the packed cell volume (PCV). However in practice, for normal blood, it makes little difference as the red cells vastly outnumber the leukocytes and the platelets are much smaller and less numerous than the red cells. The normal range for the hematocrit differs between men and women and is 40 to 50% and 36 to 46% respectively [8]. By comparison, the leukocytes and platelets together comprise only about 1%. This high concentration of red cells is the main reason that they are hemorheologically of enormous importance. The other reasons are the various physical properties of the red cells. Firstly, they have an unusual shape since they are biconcave discs, about 8.5 Pm in diameter and about 2.4 Pm thick, and so have the capacity to align with the direction of flow. Secondly, they are extremely flexible and so will deform and extend under shear forces. Thirdly, they have the tendency to adhere together loosely, as rouleaux, under the influence of plasma proteins (especially fibrinogen) as


37

mentioned above. Fourthly, they contain a hemoglobin solution of high concentration (normal range 32 to 36 g/dl [8]) which has an effect on the speed with which they can deform under shear forces. These properties act together to give blood a viscosity very substantially in excess of that of plasma alone, and to endow it with prominent shear thinning properties. The relationship between blood viscosity and hematocrit is approximately of the form: log (viscosity) = k1 + k2(hematocrit) where k1 and k2 are shear rate dependent. Typical values for the viscosity of healthy blood at shear rates of 0.277 s-1 and 128.5 s-1 are 39r4 and 4.3r0.2 mPa.s respectively for females and 48r6 and 4.7r0.2 mPa.s respectively for males [6]. The gender differences are, of course, due to the lower hematocrit that is normal for females. The above arguments should make it clear that the red cell mass has a huge influence on the viscosity of normal whole blood, and hence on blood flow in large vessels. However, blood must also pass through the microcirculation where vessels may be as small as a few Pm in diameter. Here the cells are of a size similar to the vessels, so the concept of blood viscosity is hardly relevant because the blood cannot be approximated to a uniform solution or suspension. When such flow is observed under a microscope, a bolus of plasma followed by one or two deformed cells followed by a further bolus of plasma, etc. is seen in microvessels. Under these circumstances, the factors determining resistance to flow are the viscosity of the plasma, the deformability and concentration of the red cells as well as other cells (see below), and the forces involved in holding the red cell rouleaux together. Of course another factor is the size of the red cells. The mean red cell volume (MCV) for healthy adults is 83 to 101 fl and is gender independent [8]. However, this range can be misleading, as it is the average MCV across the healthy adult population. The reason that this is misleading is that, within an individual, the variation in size of their red cell population is far greater than this range; this is because young cells released to the circulation from the bone marrow have a large volume, as much as 120 fl, but subsequently progressively lose membrane, contents and size and deformability as they age. Finally, they are destroyed, after about 120 days in the circulation, when their volume may be as little as 60 fl. A recent area of study involving the red cell concerns cell-specific effects on erythrocyte aggregation. It is now clear that individual red cells vary in their response to rouleaugenic molecules. So, for example, young red cells, newly released from the bone marrow, respond much less than do older cells. Furthermore, red cells from different healthy individuals differ in their rouleaugenic response to molecules such as fibrinogen. The reasons, which are presumably determined by the cell membrane, are not yet clear, but the consequences are hemorheologically significant [9]. 1.2.2. Leukocytes As implied above, the leukocytes have very little role to play in determining the viscosity of normal whole blood because their volume concentration is so much smaller than that of the red cells. However, they do play a major role as a determinant of flow in the microcirculation. There are two reasons for this. First, their internal contents are a complex array of nucleus, organelles, fibres, cytoplasm,

38


etc, which results in them having viscoelastic characteristics (i.e., viscosity, elasticity) much greater than those associated with the red cells. Second, they generally have somewhat larger volumes than the red cells. There are several varieties of leukocytes (i.e., monocytes, granulocytes and lymphocytes), all of which are larger than the red cell: the largest (the monocyte) has a mean volume of about 230fl, while for the smallest (the lymphocyte) it is about 120fl. It is now thought that, in a healthy individual, the white cell population confers an overall resistance to microcirculatory flow similar to that of the whole red cell mass. It is interesting to note the number concentration of the leukocytes is normally 4-8 x109 per liter while that for the red cells is about 5x1012 per liter (i.e., some 3-orders of magnitude difference), thus implying the individual leukocyte has a flow resistance, in small vessel flow, some 3-orders of magnitude greater than that of the red cell; the latter point has been proven by micropipette studies [10]. 1.2.3. Platelets From a hemorheological point of view the platelets are of little importance, even though they have relatively complex internal contents with considerable viscosity. The reason for this finding is that they are much smaller than either the erythrocytes or leukocytes, having diameters of the order of 2-3 Pm; their overall volume in normal blood is even less than that associated with the leukocytes. The consequence is that they neither influence whole blood viscosity nor microvascular resistance to any significant degree. Their primary role is to participate in the hemostatic mechanism, and as such they do have a major part to play in what is essentially the solidification of blood as it clots [1].

2. Human Juvenile It has long been known that the blood parameters change as the child ages from birth through to adulthood. In the relatively recent past, in utero blood extraction techniques have been developed and have allowed the substantial hematological changes that take place during fetal development to be evaluated. 2.1. The Fetus Data have now been published on hematological values in the fetus from 18 to 35 weeks of gestation [11]. They show that, over this period, there are progressive increases in hematocrit from 32% to 48% and in plasma protein concentration from 25 g/l to 43 g/l. The most remarkable change over the same time period is the increase in fibrinogen concentration; at 18 weeks the average value is 0.4 g/l but in many of the samples studied no clottable material (i.e., fibrinogen) was extractable at all, while at 35 weeks it had risen to about 2 g/l. One inevitable hemorheological consequence of these changes in the concentrations of fibrinogen and other plasma proteins is that the plasma viscosity rises from 0.9 to 1.05 mPa.s over the same period. Another is that the viscosity of the blood rises very substantially. Thus at a high shear rate (128.5 s-1) it increases, on average, from 2.5 to 3.7 mPa.s, while at a low shear rate (0.277s-1) it rises from about 3 to 14 mPa.s. What is clear from these data is that at 18 weeks there is virtually no shear thinning associated with the blood


39

of the fetus (i.e. it has almost Newtonian characteristics); this flow behavior is due largely to the extremely low levels of fibrinogen and other rouleaugenic proteins in the blood. There are also more subtle changes with potentially hemorheological consequences. One is the large size of the red cells of the very early fetus. At week 24 the MCV is 135 fl and, although it decreases with gestational age, even at term it is still about 119 fl and thus larger than that of the adult [12]. This, of course, has potential consequences to the resistance to flow in the microcirculation. There are also subtle differences in the erythrocyte membrane composition compared to that of the adult, with these differences affecting the mechanical properties of the membrane [13] and the innate ability of the fetal erythrocytes to form rouleaux. The hemoglobin at this stage is predominately HbF, rather than the adult form HbA1, but its intracellular concentration (i.e., MCHC) changes little from 24 weeks of gestation to term, being about 31g/dl and therefore a little less than that of the adult [12]. Another significant difference in the fetus is that the fibrinogen molecules are more strongly sialinated than in the adult [14], and although the level of sialination steadily decreases towards term, it is still higher at term than that of the adult. This increased sialination hinders its ability to induce rouleaux formation, and thus fetal blood overall exhibits less rouleaux formation than the adult, the blood shows much less shear thinning, and the overall viscosity is low compared with the adult. This raises a potentially important question to those involved with in utero transfusion: is it reasonable to transfuse adult blood which has such different hemorheological properties from that of the fetus? 2.2. At Birth A comprehensive study has been made of the hematological/hemorheological characteristics of the blood of the pre-term babies, as early as 24 weeks, and of the term baby [15]. As with the fetal data discussed above, the results show a steady increase in most hematological parameters with gestational age. Thus, hematocrit rises on average from 45 to 50 % with large variations between individuals, plasma protein concentration changes from 44 to 55 g/l, with increases in IgG, D2macroglobulin and fibrinogen from 2.1 to 2.6 g/l. Comparing these values with those obtained in utero at a corresponding gestational age, it is clear that the ex utero values are higher; indicating a degree of hemoconcentration takes place during birth. The result of the aforementioned changes is that the viscosities of both plasma and blood, at both high and low shear rates, increase over the same period: the shear thinning properties become more pronounced, with the magnitude of the changes being greater than found in the in utero study. Low shear rate (0.277 s-1) blood viscosity increases 3-fold from 12 to 40 mPa.s while the high shear rate (128.5 s-1) value increases by 20% from 3.7 to 4.5 mPa.s and plasma viscosity rises by 20% from 0.96 to 1.15 mPa.s. The alteration in the high shear rate viscosity can be explained almost entirely by changes in hematocrit and plasma viscosity over the period, whereas this explanation is not valid at low shear rate; here there are effects due to altered rouleaux formation secondary to cell-specific factors [9] and to the alteration in fibrinogen sialination that accompanies changes in gestational period.

40


2.3. Childhood At birth the red cells are still relatively large compared to the adult and still contain predominantly HbF. Over the following 6 to 12 months the hemoglobin of the red cell population is progressively replaced by HbA and the red cells acquire a similar size to those of adults. There are no hematological differences between males and females in pre-pubescent children, but post puberty the red cell parameters slowly diverge as the male red cell count increases beyond that of the female to give the normal differences found in adult humans described above.

3. Changes of Hemorheological Significance Associated With Pathology Because of space limitations only a review of the general principles will be given here. 3.1. Plasma Proteins In healthy blood it is the plasma protein concentration, especially of the larger proteins, that is mainly responsible for the elevation of plasma viscosity above that of water, so conditions associated with dramatic changes in plasma protein composition can be expected to change the hemorheological properties of blood. Two obvious examples are Waldenstrom’s macroglobulinaemia and multiple myeloma. In these conditions the concentration of large macroglobulins increases massively and has a very large effect on plasma viscosity, often with up to a 20-fold increase in the viscometric effect of these proteins. Furthermore, many of the proteins in these conditions are abnormal and highly rouleaugenic. The consequent large increase in the strength and size of rouleaux leads to markedly increased low shear rate blood viscosity. These effects have deleterious effects on blood flow in all vessels of the circulation, especially those in the microcirculation [5]. As discussed above, the plasma protein with the greatest rouleaugenic effect is fibrinogen, and hence the relevance of plasma fibrinogen concentration increases in association with many clinical conditions. Examples are diabetes mellitus, hypertension, pregnancy, post surgical trauma, infection, etc. This means that the strength of rouleaux formation and the shear thinning of blood is frequently elevated in association with these conditions [16]. 3.2. Erythrocytes. It should by now be clear that the red cell is the most prominent hematological factor influencing hemorheology. Hence, diseases characterized by alterations in this area are generally associated with hemorheological abnormalities. The most obvious is anemia due to any cause, where the lowering of the whole blood hemoglobin concentration inevitably reduces the viscous effect of the red cell mass. The result is that, though anemia itself can have serious physiological effects, it does not bring hemorheological embarrassment. What makes anemia more of a hemorheological problem is that it is often complicated by associated changes in the individual red cells. For example, in iron deficiency anemia the cells become smaller than normal,


41

while in folic acid- or vitamin B12-deficiency anemia the cells increase in size. In thalassemia, the cells are again often smaller than normal and they are also less deformable. Such changes may have negative influences on microvascular flow. However, most significant of all in this respect is sickle cell anemia where the cells sickle in low oxygen conditions, become quite rigid and have very substantial problems negotiating the microcirculation. In addition, the repeated sickle-unsickle cycles during flow through the body eventually leads to their membranes becoming damaged and irreversibly rigid. As a consequence, the resistance to blood flow in large vessels is lower than normal due to the low hematocrit, while in the microcirculation it can be considerably higher, even leading to stasis . At the other end of the scale is the problem of increased red cell concentration. There are various pathologies that lead to this; examples are respiratory disease where the resultant hypoxia stimulates red cell production and polycythemia consequent to malignant changes in the hemopoietic tissue [17]. Inevitably, they lead to substantial increases in blood viscosity. Such changes can also occur in the neonate, in that excessive production of red cells can result from intrauterine hypoxia. However, the effects can also result from passive in utero events such as transfusion between mother and foetus, between twin and twin, or as a result of delayed clamping of the cord [18]. All of these phenomena cause the blood viscosity to increase considerably and in extreme cases can be very deleterious to blood flow. 3.3. Leukocytes As mentioned above, leukocytes in the healthy individual are not of sufficient concentration to affect whole blood viscosity, whereas they do have a significant resistive effect at the microcirculatory level due to their relative stiffness and enhanced viscoelastic characteristics compared to the erythrocyte. Nevertheless, there are conditions in which the leukocyte count can become extremely high: the most prominent are the classic leukemias. It is not unusual for the leukocyte count to increase by an order of magnitude or even more. Under these circumstances they are relevant to the viscosity of the whole blood and, because of their large volume concentration, relative inflexibility, irregular shape, and viscoelasticity, they can cause the viscosity to rise enormously and so affect large vessel flow. However, equally important is that now their resistive effect at the microcirculatory level far exceeds that of the red cell mass, especially as the deformability of leukemic leukocytes is less even than that of normal leukocytes [19]: all of these factors have very deleterious effects on tissue perfusion with obvious consequences [20].

4. Other Mammals Until fairly recently the study of hemorheology has concentrated on the human, and other mammals have largely been ignored. The situation is now beginning to change as investigators look to animal models for hemorheological studies not possible in the human, as veterinary workers begin to realize that the hemorheological effects in pathology found in the human may also exist in mammals of economic importance, and as investigators realize that the study of mammalian hemorheology may offer new insights into human hemorheology, especially in pathological states. Thus

42


fascinating data are beginning to emerge showing that some hematological/ hemorheological parameters seem to be held remarkably constant across the whole class of mammals, while others vary widely. 4.1. Plasma Proteins Comparative data on variations across the mammalian kingdom of the concentrations of the different plasma proteins are difficult to locate, but data from Johnn, et al. [21] indicate that they are likely to vary significantly between mammals. In their study, 31 different mammalian types were investigated with results indicating that the plasma viscosity varied widely: at one extreme were the New Zealand rabbit and the cheetah with values of 1.1 mPa.s and 1.2 mPa.s respectively, and at the other the chimpanzee and goat each with a value of 1.6 mPa.s. These variations were only partly explained by variations in the concentrations of fibrinogen, because while the rabbit had one of the lowest fibrinogens at 2.4 g/l, that of the cheetah was relatively high at 3.8 g/l. Similarly, that of the goat was high at 5.4 g/l while that of the chimpanzee was fairly average at 3.0 g/l. It is notable that these fibrinogen levels are all overshadowed by the level in the black buck which was recorded as 7.0 g/l. Plasma protein data on nine different domestic mammals from Windberger, et al. [22] are useful since they show that total plasma protein concentrations varied from as low as 43 g/l in the mouse to as high as 72 g/l in the cat and in cattle. However, they found only a partial correlation between plasma protein concentration and plasma viscosity. 4.2. Erythrocytes The most important finding here is the remarkable constancy of the hematocrit and whole blood hemoglobin across mammals. In the study by Johnn, et al. [21] mentioned above, they found the lowest hematocrit in the goat at 31% and the highest in the wallaby at 53%. However, these are extremes and when the values for the 31 mammals studied were averaged the mean and standard deviation were 40r5%, indicating the narrowness of spread across the class as a whole. Whole blood hemoglobin shows an even narrower range judging by data from Gascoyne and Hawkey [23] who show a mean and SD of 14.8r2.1 g/l. This seems remarkable considering the large variation in size, heart rate, life style, etc. between the animals. On the other hand, they found enormous differences in MCV, with the goat having the smallest values at 22 fl and the Asian elephant had 6-fold greater value of 120 fl! Windberger, et al. [24] have shown that for the African elephant the MCV is even larger at 138 fl. In spite of this, the mean cell hemoglobin concentration is remarkably conserved. Gascoyne and Hawkey [23] found a coefficient of variation of only 8.5% (34.6r3.0, mean±SD). In view of the relative constancy of hematocrit and whole blood hemoglobin concentration, this narrow range of hemoglobin means that the red cell count varies by a similar order of magnitude to that of the MCV across the mammals. There is no obvious connection between the MCV and the size, heart rate, life styles, etc. of the mammals studied. Johnn, et al. [21] measured the high and low shear rate (128.5 and 0.222 s –1 respectively) blood viscosity characteristics of the 31 mammals they investigated, and found that high shear rate values varied by some 2-fold, with this variation


43

almost totally explained by the variations in hematocrits and plasma viscosities between the mammals. On the other hand, the low shear rate values varied by some 18-fold and this could only be partly explained by the hematocrit and plasma viscosity differences. The other factor of great relevance here is the variation in the inherent rouleaugenicity of the red cells from the various mammals. Thus, for example, in sheep and cattle, rouleaux formation is non-existent, while in the horse and elephant it is very increased such that if it occurred in a human it would be considered pathological. The reasons for the differences may be partly due to different levels of fibrinogen in the plasma of these animals, but it is now becoming increasingly clear that there must be cell specific factors involved as well [9]. That is, in the non-rouleaux forming animals such as sheep, some peculiarity of the membrane structure inhibits rouleaux formation. This has been shown by Kaibara [25] who found that bovine red cells would not form rouleaux in their own plasma, but if they were treated with trypsin to remove significant surface protein they would then form intense rouleaux. Furthermore, in animals where rouleaux formation is massive such as the horse, it is to be expected that membrane factors will be found to be facilitate aggregate formation. Further study in this area is clearly warranted. 4.3. Breed Variation The data discussed above are clear in showing large differences of some hemorheologically important variables between mammals. What is also of interest is to what extent these variable values are conserved across breeds within an individual mammal species. Such studies are rare; indeed, the only one that the author has found is a study of nine different breeds of dog [26]. Very substantial differences in whole blood viscosity were found in this study, so high shear rate (128.5 s-1) values varied by 50% between the highest and the lowest, while the low shear rate (0.277 s-1) values varied by 140%. Of the breeds, it was the more athletic, that undergo brief periods of intense anaerobic exercise, which had the highest values (e.g., greyhounds and deerhounds), while the golden retriever and Springer spaniel were at the other end of the range. However, the differences could be explained entirely by the different hematocrits between the dogs, with such other potential factors such as cellular parameters and plasma composition not being significantly different between the breeds. In this context, an interesting study was done by Anwar and Rampling [27] who compared the hematological/ hemorheological parameters of five different ethnic groups of humans. The only significant difference found between the groups was that Caucasians had significantly lower concentrations of plasma proteins and hence lowered plasma viscosity.

References [1] [2] [3] [4]

S.I. Fox, Heart and circulation, Chapt. 13, In: Human Physiology, Times Mirror, Dubuque, 1996, pp. 342-385. H.J. Meiselman, E.W. Merrill, E.R. Gilliland, G.A. Pelletier and E.W. Saltzman, Influence of plasma osmolarity on the rheology of human blood, J. Appl. Physiol. 22 (1967), 772-781. P.W. Rand, W.H. Austin, E. Lacombe and N. Barker, pH and blood viscosity, J. Appl. Physiol. 25 (1968), 550-559. G.D.O. Lowe and J.C. Barbenel, Plasma and blood viscosity, In: Clinical Blood Rheology, Vol I, G.D.O. Lowe, Ed., CRC Press, Boca Raton, (1988), pp 11- 44.

44

[5] [6]

[7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]

[22]

[23] [24] [25] [26] [27]


T. Somer, Rheology of paraproteinaemias and plasma hyperviscosity Syndrome, Bailliere’s Clin. Haematol. 1 (1987), 695-723. M.W. Rampling, J.R. Brown, R.J. Robinson, M.D. Feher, S. Cholerton and P.S. Sever, The short term effects of abstention from tobacco by cigarette smokers on blood viscosity and related parameters, Clin. Hemorheol. 11 (1991) 441-446 M.W. Rampling, Red cell aggregation and yield stress, In: Clinical Blood Rheology, Vol I, G.D.O. Lowe, Ed., CRC Press, Boca Raton, (1988), 45-64. J.V. Dacie and S.M. Lewis, Practical Haematology, Churchill Livingstone, London (1995) pp. 1217. M.W. Rampling, H.J. Meiselman, B. Neu and O.K. Baskurt, Influence of cell-specific factors on red cell aggregation, Biorheology, 41 (2004), 91-112. S. Chien, White blood cell rheology, In: Clinical Blood Rheology, Vol I, G.D.O. Lowe, Ed., CRC Press, Boca Raton, (1988), 87-109. C.R. Welch, M.W. Rampling, M.A. Anwar, D.G. Talbert and C.H. Rodeck, Gestational reference ranges for fetal haemorheological parameters, Clin. Hemorheol. 14(1994), 93-103. D.G. Nathan and F.A. Oskai, Hematology of Infancy and Childhood, Vol 1, WB Saunders Co, Philadelphia (1990) p 30. O. Linderkamp, G.B. Nash, P.Y.K. Wu & H.J. Meiselman, Deformability and intrinsic material properties of neonatal red blood cells, Blood 67(1986), 1244-1250. J.L. Francis and D.J. Armstrong, Sialic acid and enzymic desialination of cord blood fibrinogen, Haemostasis 11 (1982), 223-228. M.A. Anwar, M.W. Rampling, S. Bignal and R.P.A.Rivers, The variation with gestational age of the properties of the blood of the new-born, Brit. J. Haematol. 86 (1994), 163-168. T. Somer and H.J. Meiselman, Disorders of blood viscosity, Ann. Med. 25 (1993), 31-39. T.C. Pearson and G. Wetherley–Mein, The course and complications of ideopathic erythrocytosis, Clin. Lab. Haematol. 1 (1979), 189-196. E.A. Letsky, Fetal and neonatal transfusion, Brit. Med. J., 300(1990), 862-866 M.A. Lichtman and G.A. Kearney, The filterability of normal and leukemic leukocytes, Blood Cells 2 (1976), 491-501. M.W. Rampling, Hyperviscosity as a complication in a variety of disorders, Semin. Thrombos. Haemostas. 29 (2003), 459-465. H. Johnn, C. Phipps, S. Gascoyne, C. Hawkey and M.W. Rampling, A comparison of the viscometric properties of the blood from a wide range of mammals, Clin. Hemorheol. 12 (1992), 639-647. U. Windberger, A Bartolovitsch, R. Plasenzotti, K.J. Korak and G. Heinze, Whole blood viscosity, plasma viscosity and erythrocyte aggregation in nine mammalian species: reference values and comparison data, Exp. Physiol. 88 (2003), 431-440. S.G. Gascoyne and C. Hawkey, Patterns of variation in vertebrate haematology, Clin. Hemorheol. 12 (1992), 627-637. U. Windberger, R. Plasenzotti and Th. Voracek, The fluidity of blood in African elephants (Loxodonta africana), Clin. Hemorheol. 33 (2005), 321-326. M. Kaibara, Rheological behaviors of bovine blood forming artificial rouleaux, Biorheology 20 (1983), 583-592. A.R. Bodey and M.W.Rampling, A comparative study of the haemorheology of various breeds of dog, Clin. Hemorheol. 18 (1998), 291-298. M. Anwar and M.W. Rampling, Comparative haemorheology of five healthy, ethnically diverse groups: results of a pilot study, Clin. Hemorheol. 14 (1994), 697-707.

45


Macro- and Micro-Rheological Properties of Blood b,

Giles R. COKELETa and Herbert J. MEISELMAN 1 Department of Chemical and Biological Engineering, Montana State University, Bozeman, MT 59717, USA and bDepartment of Physiology and Biophysics, Keck School of Medicine, University of Southern California, Los Angeles, CA, 90033, USA. a

Introduction In order to quantitatively understand the conditions of blood flow through various in vitro and in vivo geometries, the flow properties of blood must be experimentally determined. In this chapter, we initially consider the rheological behavior of blood under conditions where the blood is treated as a homogenous fluid and thus where the formed elements (e.g., red blood cells, white blood cells) are tacitly ignored. This approach is then modified in order to consider flows where the blood cell characteristic dimensions approach those of the geometries in which the flow takes place. The former approach yields the macro-rheological properties of blood while the latter yields micro-rheological characteristics; in general, data obtained in geometries of 200 μm or less are considered micro-rheological. It is of interest to note that the study of blood rheology dates, at least, to the work of Poiseuille who attempted to derive an equation for blood flow in tubes. However, due to experimental difficulties associated with blood coagulation he was unsuccessful with these attempts, and thereafter turned to simpler fluids such as water and oil to develop his well-known equation [1].

1. Special Problems in Measuring Blood Rheology Prior to discussing specific rheological properties, it is of value to discuss some difficulties and potential artifacts associated with rheological measurements on blood and related fluids. Unless otherwise stated, it is the authors’ belief that all data presented herein are unaffected by such experimental problems. There are three properties of blood, and related fluids, which can lead to difficulties in measuring rheological properties: (1) The plasma proteins are surfactants and form a protein layer or film at fluid-air interfaces; (2) The erythrocytes are denser than plasma; (3) At low shear rates, the erythrocytes aggregate. The surfactant layer formed by plasma proteins at the fluid-air interface has mechanical strength and is thus a semi-rigid film. Consequently, the pressure drop 1 Corresponding author: Department of Physiology and Biophysics, Keck School of Medicine, University of Southern California, Los Angeles, CA, 90033, USA; E mail: [email protected]

46

G.R. Cokelet and H.J. Meiselman / Macro- and Micro-Rheological Properties of Blood

across such interfaces, as in a capillary viscometer, is not zero. Likewise, the surfactant layer at an interface can transmit significant torque to or cause drag on the torquemeasuring element in rotational viscometers. In general, this effect is a significant problem only for plasma, serum and RBC-plasma or RBC-serum suspensions at very low hematocrits; at normal or high hematocrits, the additional torque due to the film is only a small percent of the total. An unfortunate result of this additional torque is the artifact of non-Newtonian behavior for plasma or serum (Figure 1); data obtained at high shear rates and thus high shear stresses yields the true Newtonian viscosity. 7

PLASMA VISCOSITY (mPa.s)

6

DONOR A DONOR B

5 4 3 2 1 0

0

500

1000

1500

2000

-1

SHEAR RATE (SEC )

Figure 1. Viscosity-shear rate data for human plasma which shows protein film artifacts. These results were obtained in a cone-plate viscometer in which the film added extra drag at the radius of the cone. The apparent non-Newtonian behavior is the result of this added drag; dashed lines indicate viscosity values obtained by extrapolation of viscosity to infinite shear rate (i.e., to the reciprocal of shear rate = zero).

Procedures to minimize or eliminate the influence of the protein layer depend on the type of viscometer used for the measurements. In the case of capillary viscometers, one can use fluid filled pressure transducers that connect to the fluid reservoirs below the fluid-air interfaces, or can obtain pressure drop and flow rate data using two capillaries of the same diameter but significantly different lengths. In the latter case, one subtracts the lengths and pressure drops for the two capillaries, at a selected flow rate, to obtain a pressure drop free of the effects of the surfactant interfaces. In the case of rotational viscometers, one useful approach is to insert a “guard” ring through the interface, thereby isolating the torque-sensing element from the film-associated drag. Note, however, that the guard ring must rotate at the same angular velocity as the torque-measuring element of the viscometer, and is thus not useful for cone-plate viscometers in which the cone both rotates and senses torque. Alternatively, one can use two bobs that are of different lengths or two cones that are of different radii; in this method, a subtraction technique will eliminate the surfactant effects if the interface is the same for the two bobs or cones.


47

Red cell sedimentation is slow for individual cells, since the density difference between the cell (1.096 gm/ml) and the plasma (1.025 gm/ml) is small. Consequently, the sedimentation rate of an individual erythrocyte in plasma is only about 0.13 μm/minute or 0.08 mm/hr. If the suspending medium is serum and thus does not promote RBC aggregation, and if the RBC volume fraction H (i.e., hematocrit) is 15% or greater, the presence of neighboring red cells causes the settling to be hindered and hence slower. For hindered settling free of wall effects, the sedimentation rate (U, μm/min) is given by the equation where H is hematocrit [2]: U = 0.13 (1 – H)4 40 50 uM ID 100 uM ID 500 uM ID

HEMATOCRIT INCREASE (%)

35 30 25 20 15 10 5 0

0

5

10

15

20

25

30

DISTANCE SETTLED (uM)

Figure 2. Calculated increase of hematocrit of the settled RBC phase for non-flowing blood in horizontal tubes. Horizontal axis is the distance RBC settle from the top inside wall of the tube. Minimal effects are noted for large tubes whereas large increases can occur in smaller tubes.

RBC aggregation, as seen in plasma at stasis or slow flow, greatly increases the sedimentation rate since the settling rate of particles is dependent on the square of the particle size. A typical, steady, hindered settling rate for H=40% normal human blood is about 200 μm/min or about 10 mm/hr. Usually, sedimentation in viscometers is not a significant problem if the wall shear rate is above about several inverse seconds. However, at lower shear rates, the effect of erythrocyte sedimentation must be considered, especially for horizontal tube and cone-and-plate viscometers. Figure 2 presents calculated percent increases of hematocrit versus settled distance in 50, 100 and 500 μm tubes. It is obvious that minor settling has almost no effect in large tubes whereas even a 10 μm settling (i.e., about one RBC diameter) markedly elevates hematocrit in smaller tubes. Erythrocyte aggregation also causes another effect termed red cell syneresis. This phenomenon is the inward movement of RBC as they aggregate, leaving layers of cellpoor plasma at the walls of viscometers of all types [3]. Although syneresis is usually considered to be a slow process (e.g., separation of fluid from a gel), the effects of

48


syneresis are seen very soon after starting the viscometer. This has led to conflicting data for blood’s rheological properties for shear rates below about 2 s-1. For example, consider the events that occur when using a concentric cylinder viscometer filled with well-mixed and well-stirred blood: upon starting the viscometer at a rotational speed corresponding to a steady shear rate of about 2 s-1 or less, the torque-time record of the viscometer first rises from zero to a peak, then decays in an exponential manner with time, finally reaching an essentially steady value after approximately 30 minutes. This time-dependent torque response is due to the startup of the fluid motion at very short times, syneresis of the erythrocytes away from the viscometer surfaces at all times, and finally, at very long times, sedimentation effects. Note that essentially all investigators agree that using the final steady-state value of torque is incorrect: it is the result of fully completed syneresis plus a variable degree of sedimentation: its use results in shear stress and viscosity values that are too low. 1000

RELATIVE VISCOSITY

PEAK TORQUE EXTRAPOLATE

100

10 0.001

0.01

0.1

1

10

-1

SHEAR RATE (SEC )

Figure 3. Viscosity-shear rate results for normal human blood at a hematocrit of 40%. Two different methods were used for assigning the correct torque at low shear rates. For shear rates below 1 s-1 the use of peak torque underestimates viscosity, with the difference becoming greater with decreasing shear. Peak torque values from [4, 5] and extrapolated to zero time from [6, 7].

Two different methods have been proposed for assigning the correct torque at shear rates below about 2 s-1: Chien’s approach uses the peak torque value to calculate the apparent viscosity of blood [4, 5], while Merrill and co-workers extrapolate the exponentially decaying torque data back to zero time and use that torque value to calculate apparent viscosity [6, 7]. The rationale for the latter procedure is that during the time of the exponential torque decay, only the syneresis effect is a significant cause for the decreasing torque behavior; extrapolation back to zero time assumes that the syneresis effect is the same in nature even in the earlier, pre-peak time period. Figure 3 shows a comparison of the results of the two methods for a normal blood at a 40% hematocrit, where it can be seen that the two methods diverge at shear rates less than 1 s-1; results are usually coincident at higher shear.


49

2. Macro-Rheological Behavior of Normal Human Blood Literature reports describing the macrorheological behavior of human blood are numerous [e.g., 8-16] and may be consulted for additional details; reviews of the development of hemorheology [17] and of the molecular biology aspects of hemorheology [18] also exist. In general, it is agreed that normal human blood or reconstituted RBC-plasma suspensions are non-Newtonian fluids without measurable normal stresses. In addition, for the usual time scales encountered in macrorheology studies, rheological properties are history and time independent (i.e., no evidence of thixotropy or rheopexy). That is, although RBC disaggregation and RBC deformation occur during establishing viscometric flow and do require small but finite periods of time, their time scales are relatively short and thus do not affect steady shear results. 50

SALINE PLASMA

VISCOSITY (mPa.s)

40

30

20 RIGID CELLS 10

0 0.1

1

10

100

1000

10

4

-1

SHEAR RATE (SEC )

Figure 4. Typical viscosity-shear rate results for three RBC suspensions all at 40% hematocrit: 1) RBC in plasma; 2) RBC in saline; 3) chemically-fixed rigid RBC in saline. RBC aggregation increases viscosity at lower shear rates for the RBC in plasma suspension, with cell deformation occurring at higher shear rates. RBC-plasma and RBC-saline data are not coincident at high shear due to different suspending phase viscosity.

Figure 4 presents typical viscosity-shear rate results for three RBC suspensions, all at a hematocrit of 40%: 1) RBC in plasma; 2) RBC in isotonic buffer; 3) RBC made extremely rigid by fixation in glutaraldehyde and suspended in isotonic buffer. The observed hematocrit of the rigid cell suspension was 66% which, due to poor packing of rigid cells, corresponds to a true hematocrit of 40% [19]. Dealing first with the RBC-plasma suspension, the feature most notable is the strong effect of shear rate on viscosity and hence the marked non-Newtonian flow behavior. Over the shear rates shown (0.5 to 1,500 s-1), there is an 11-fold decrease in blood viscosity, with essentially Newtonian behavior at or above 1,000 s-1; the exact shear rate necessary for Newtonian flow is hematocrit dependent, with higher shear needed at higher hematocrits. In contrast, the RBC-buffer suspension is much less shear dependent, with only a 2-fold

50


decrease over the range of shear rate and Newtonian flow expected prior to 1,000 s-1. Since normal RBC in plasma undergo reversible aggregation (i.e., rouleaux formation, see chapter II.4.b) while those in buffer do not, RBC aggregation is the primary determinate of low shear blood viscosity. With increasing shear forces, the RBC aggregates in the plasma suspension become progressively dispersed and the viscosities approach each other. Note that since plasma has a higher viscosity than buffer, the two curves in Figure 4 will not coincide at higher shear but would do so if suspending viscosities were equalized [9, 11]. The data for the rigid, non-deformable RBC suspension in Figure 4 demonstrates the role of RBC deformability as a determinant of blood rheology. Since unlike normal cells, rigid RBC are unable to deform in response to shear forces, increases of viscosity at high shear are ascribed to decreases of RBC deformability. Dintenfass [20], and subsequently other investigators, have utilized this behavior at higher shear (i.e., > 200 s-1 or sufficient to disperse RBC aggregates) to assess RBC rigidity via a calculated “Tk” parameter: Tk = [(Kr0.4 – 1) / Kr0.4 ] / H where Kr is relative viscosity (i.e., suspension viscosity divided by suspending medium viscosity) and H is hematocrit. In overview, normal human blood behaves as a shear thinning non-Newtonian fluid, with this behavior generally ascribed to two processes [14, 15]: 1) RBC aggregates formed at stasis or low shear are dispersed by increasing fluid shear forces, with complete dispersion achieved in the range of 80-120 s-1; 2) RBC undergo deformation and alignment with flow at medium to high shear rates. Thus, as shown in Figure 4, increases of viscosity at low shear usually implies enhanced RBC aggregation while increases at high shear are ascribed to decreased RBC deformability. 2.1. Hematocrit Effects Not surprisingly, the volume fraction of particles in a suspension affects its rheological behavior. In the case of normal blood, this volume fraction is represented by the volume fraction of RBC (i.e., hematocrit); unless aggregated, platelets are too few in number and volume to affect viscosity, and white cells only influence viscosity when their volume fraction is abnormally elevated [21]. Figure 5A presents blood viscosity data for hematocrits of 20, 30, 40 and 50% over a 200-fold range of shear rate (i.e., 0.1 to 20 s-1), and Figure 5B presents viscosity at selected shear rates for the four hematocrits. It is obvious from these figures that hematocrit affects both viscosity values and the degree of non-Newtonian behavior. At the lowest shear rate shown, a hematocrit change from 20 to 50% increases viscosity by about 9-fold whereas at the highest shear the same hematocrit change causes a 3-fold increase. Deviation from Newtonian behavior is also hematocrit dependent: at 20% hematocrit there is a 6-fold decrease in viscosity over 0.1 to 20 s-1 whereas at 50% the decrease is about 20-fold.

G.R. Cokelet and H.J. Meiselman / Macro- and Micro-Rheological Properties of Blood 120

20% 30% 40% 50%

VISCOSITY (mPa.s)

100

80

60

40

20

0

0.1

1

10

100

-1

SHEAR RATE (SEC )

Figure 5A. Effect of hematocrit on viscosity-shear rate data for normal human RBC suspended in autologous plasma. Note that hematocrit affects both viscosity and the degree of non-Newtonian behavior.

120 -1

0.11 SEC 0.51 SEC -1

100

VISCOSITY (mPa.s)

-1

2.0 SEC 20 SEC -1

80

60

40

20

0 0

10

20

30

40

50

60

HEMATOCRIT (%)

Figure 5B. Hematocrit-viscosity data at four different shear rates for normal human RBC suspended in plasma. The effects of hematocrit changes are greater at lower rates of shear. Data are from Figure 5A.

51

52


2.2. Yield Shear Stress The reversible RBC aggregation observed in normal blood is similar to that for particles found in many other suspensions (e.g., clay), thus prompting Merrill and coworkers to consider the existence of a yield shear stress for human blood [22, 23]. These investigators utilized a data-fitting model developed by Casson [24] for pigmentoil suspensions used as printing inks: W1/2 = aJ1/2 + Wy1/2 where W is shear stress, J is shear rate and Wy is yield shear stress. Figure 6A presents a so-called “Casson Plot” for the experimental data shown in Figure 5A. It is clear that for each hematocrit the data can be fitted by linear regression (r>0.99), and extrapolation to zero shear rate yields the yield shear stress. As expected [22], the yield stress is strongly hematocrit dependent, with the relationship well-fitted by a power law equation (Figure 6B). At constant hematocrit, the yield stress increases with the strength of RBC aggregation [23], and has been used by some investigators as an index to red cell aggregation in disease and for cells in various polymer solutions [26, 27].

1/2

SQUARE ROOT SHEAR STRESS (mPa )

14 20% 30% 40% 50%

12 10 8 6 4 2 0

0

1

2

3

4

5

SQUARE ROOT SHEAR RATE (SEC -1/2)

Figure 6A. Shear rate-shear stress data for normal human RBC in plasma. Note that the square root of shear rate and shear stress are on linear axes (“Casson” plot). Straight lines obtained via linear regression (r>0.99) with their intercepts being the square root of yield shear stress. Data are from Figure 5A.

Using the Casson Equation to determine yield stress requires shear rate-shear stress data at very low rates of shear (e.g., 20 s-1 and lower) in order to obtain a linear fit in the region near zero shear. Application of the equation to data obtained at higher shear rates is possible, but extrapolation to zero shear will result in incorrect values of yield stress. Although the data shown in Figures 5 and 6 were obtained with a rotational


53

viscometer, direct determinations of yield stress in a tube viscometer (i.e., residual pressure drop at zero flow) are consistent with the results presented herein [27]. Note that both the Casson extrapolation method and the tube viscometer data were obtained going from low shear to zero shear; measurements made starting from zero shear are fraught with artifacts, including RBC settling or syneresis in tubes (Figure 2). 10

YIELD STRESS (mPa)

8

6

4

2

0

15

20

25

30

35

40

45

50

55

HEMATOCRIT (%)

Figure 6B. Yield shear stress results obtained from Figure 6A plotted versus hematocrit. Curved line through points obtained using power law regression (exponent = 3.1, r = 0.999) in agreement with [22].

2.3. Cellular Determinants of Low Shear Viscosity Low shear viscometry is a technique commonly used to assess RBC aggregation [9, 12], and its use follows from the classical description of the factors that affect the low and high shear rate rheological behavior of RBC suspensions [9]: at constant hematocrit and temperature, low shear blood viscosity is primarily determined by RBC aggregation while high shear viscosity is dependent on RBC deformability (Figure 4). It is thus tacitly assumed that comparisons of blood viscosity at low shear will reflect relative degrees of red cell aggregation. Experimental findings now exist that challenge the unique contribution of aggregation to low shear viscosity. One approach is the use of chemical agents that produce a dose-dependent, constant volume change of RBC shape from the normal biconcave morphology either to a crenated, spiky, spherical form or to a mono-concave form [28]. Coincident with such dose-related shape changes are progressive decreases of RBC deformability: rather than causing complete cell rigidity (e.g., Figure 4), the cells are less deformable at low stress levels yet deform normally at higher stress levels [29]. Figure 7 presents rheologic data for crenated RBC, suspended in buffer, at increasing levels of the shape-altering agent (DNP, di-nitrophenol) and hence at greater extents of morphology changes. Microscopic examination of these suspensions indicated absolutely no evidence of RBC aggregation, yet there are marked, dose-

54


dependent increases of low shear viscosity: 5 mM DNP causes a 2-fold increase at 37.5 s-1 and a 15% increase at 150 s-1, whereas at 1,500 s-1 the data are identical with control. Based solely on the usual interpretation of such data [9, 11], DNP would be considered to be an agent causing aggregation that, with sufficient shear stress, is dispersed: in fact, no aggregation occurs and increasing shear serves to deform the cells, eventually resulting in a normal deformed appearance [29]. Similar observations have been reported for RBC treated with hydrogen peroxide or very low levels of glutaraldehyde, or heated to 48 qC: either unaltered or very slight changes of aggregation in plasma yet markedly increased low shear viscosity [30]. 35 CONTROL 1 mM DNP 2 mM DNP 5 mM DNP

VISCOSITY (mPa.s)

30 25 20 15 10 5 0

10

100 -1 SHEAR RATE (SEC )

1000

Figure 7. Viscosity-shear rate data for human RBC re-suspended at 60% hematocrit in saline containing increasing concentrations of the shape-altering chemical di-nitrophenol (DNP). RBC do not aggregate in DNP-saline solutions but deformability at low shear stresses is reduced; normal deformation occurs at high shear stress [29].

2.4. Constitutive Equations The relationships between shear stress and shear rate for blood must be determined experimentally and expressed as mathematical equations, usually referred to as constitutive equations. Given the complex macro-rheological behavior of blood described above, it is not surprising that a single equation fails to completely describe the effects of various rheological variables (e.g., hematocrit, shear rate). Thus, several approaches to defining these equations exist, with some the result of curve-fitting experimental data and others based on a particular rheological model. Chapter II.1 (Basic Aspects of Rheology) presents constitutive equations of two types. One type results from using empirical relationships and includes Newtonian fluids, the Bingham fluid model and the power-law fluid model. All of these use the ratio of shear stress to shear rate to define an “apparent viscosity” for blood that, with the exception of Newtonian fluids, is a function of shear rate. The other type results from the use of models of suspensions and includes an Einstein-model approach based


55

on the volume fraction of the suspension occupied by particles (e.g., hematocrit), the Casson Equation detailed above in section 2.2, and an equation developed by Quemada. While the Einstein model is only useful for suspensions having low volume fractions of particles and the Casson Equation is really only applicable at very low shear rates, the one by Quemada accurately fits blood data over a very wide range of shear. Several other approaches to constitutive equations for blood have been published (e.g., [31-36]), a comparison of constitutive equations has been presented by Easthope and Brooks [37], and a microstructure-based constitutive equation, developed by Owens and co-workers, has been applied to the pulsatile flow of blood [38].

3. Micro-Rheological Properties of Human Blood As the characteristic dimension of a flow channel approaches the size of the particles in a suspension, one should expect that the simple continuum model of the suspension will fail to be applicable. Often, this limit of the applicability of the continuum model begins to manifest itself at characteristic channel dimensions that are about 30 times the particle diameter: in the case of blood with a characteristic RBC dimension of 8 μm, an apparent failure occurs at about 300 microns. This was demonstrated by Fåhraeus and Lindqvist [39], who found that the apparent viscosity of blood was a function of tube diameter, for diameters of 300 microns and less, when they flowed constant-hematocrit blood from a well-stirred reservoir through a tube. Subsequent studies by others showed that the hematocrit of the blood flowing from the tube is equal to that of the blood in the feed reservoir, as long as the tube diameter is 20 microns or greater [40, 41]. This means that, at steady state, there is no accumulation or loss of erythrocytes from the blood as it flows through the tube. However, Fåhraeus earlier had shown that the average hematocrit of the blood in the tube was less than that of the blood in the feed reservoir [42]. The practical consequence of these findings is that the local hematocrit across the cross section of the tube is not uniform, with the hematocrit highest near the center of the tube where the velocity is highest. All of these factors are discussed in more detail in the material below. 3.1. The Fåhraeus Effect The finding that, for blood flowing steadily in tubes with diameters of less than 300 microns, the average hematocrit of the blood in the tube is less than the hematocrit of the blood in the reservoir feeding the tube is known as the Fåhraeus Effect. This effect is generated in the concentration entrance length of the tube, in which erythrocytes move towards the central region of the tube as they flow downstream. This entrance length is estimated to be about the distance that the blood travels in a quarter of a second for blood where red blood cell aggregation is negligible and the vessel diameter is greater than about 20 microns. In a survey of all the data available at the time, Goldsmith et al. [43] compiled the information about the Fåhraeus Effect, and their results are shown in Figure 8.

56


D, μm Figure 8. Relative tube hematocrit as HT/HD (i.e., tube /discharge hematocrit ratio) versus tube diameter D in microns. The shaded area contains all the data in the literature as of 1988. The data points are the original data of Fåhraeus [39]. All flow rates are sufficiently high to prevent red cell aggregation, and the feed hematocrits were 40 - 45%. HT is the average hematocrit of the blood in the tube, and HD is the hematocrit of the blood leaving the vessel which, at steady state, also equals the hematocrit of the blood actually entering the tube. Reproduced with permission from [43].

In capillaries of diameters below about five microns, the red cells flow down the tube in single file and the erythrocyte tends to fill the lumen [44]. As a consequence, HT/HD approaches unity as the capillary diameter approaches 2.7 microns, the minimum diameter tube in which a human erythrocyte can fit without breaking its membrane. A detailed graph of the literature data, for three feed reservoir hematocrits, is presented in Pries et al. [45]. Erythrocytes are not the only blood cells that show the Fåhraeus Effect. White cells and platelets also have such an effect, but it is qualitatively different. For white blood cells, the radial distribution depends on the blood flow rate: at high flow rates they tend to be in the central regions of the flow, whereas at low flow rates, when erythrocytes can aggregate, the white cells are pushed to the wall [46, 47]. For the platelets, they are always found in higher concentrations near the tube wall, and so, at low flow rates, have their concentration in the tube as high as 1.3 - 1.4 times their concentration in the blood leaving the tube, and about a 1.1 times greater at high blood flow rates [43]. The general qualitative rule for blood flow in tubes is that the largest particles are in the central region of the flow, and the smallest particles are near the wall. At high shear rates (i.e., flow rates), the red cells and white cells prefer to be in the faster flowing regions of the tube, but at low flow rates, where red cell aggregation occurs, the red cell aggregates are the biggest particles in the flow, and so they prefer to be in the central region, displacing the white cells towards the perimeter of the flow. For details on the mechanisms of particle radial migration, see Goldsmith [48, 49].


57

3.2. The Fåhraeus -Lindqvist Effect The finding that for small tubes with diameters below about 300 microns and for faster flow rates which do not allow appreciable erythrocyte aggregation, the effective viscosity of the blood depends on tube diameter is known as the Fåhraeus-Lindqvist Effect. The effective viscosity is determined from the Hagen-Poiseuille equation using the measured pressure drop, the flow rate, and tube dimensions. In Figure 9 the relative effective viscosity of the blood (i.e., effective viscosity divided by the plasma viscosity) for such relatively fast flows is indicated by the shaded area, which includes all literature data. The filled squares are the original data of Fåhraeus and Lindqvist [39]. Also shown are data for slow flows, where red cell aggregation can occur.

Figure 9. The Fåhraeus-Lindqvist Effect. The shaded area contains all the published data for normal human blood flowing at high rates, with the filled squares being the original data of Fåhraeus and Lindqvist [39]. Top upward arrow indicates asymmetric syneresis and bottom downward arrow indicates axisymmetric syneresis; vertical dashed line indicates critical diameter for RBC entry (see text for other curves). Reproduced with permission from [43].

In Figure 9 the top curve from Reinke et al. [50] indicates the relative effective viscosity of blood for very slow flow in horizontal tubes where red cell aggregation and sedimentation occur; average wall shear rates were in the range of 1.6 to 4.0 s-1. At steady state, one anticipates that red cells have settled on the bottom of the horizontal tube, causing a high hematocrit there and consequently a very low flow rate, while above this more concentrated hematocrit region, the flow is faster because the

58


hematocrit will be lower than in the settled region. The average hematocrit in the tube will thus be higher than that of the blood flowing form the tube. The bottom two curves in this figure are for blood flowing slowly through vertically oriented tubes, with curve P indicating the minimum relative effective viscosity for slow flow in tubes, and curve D similarly indicating the minimum slow flow relative effective viscosity for blood to which dextran 250 kDa has been added to the plasma to enhance erythrocyte aggregation. With the vertical tube orientation, syneresis makes the red cells move into the core of the flow, leaving an appreciable plasma wall region [51], thereby resulting in greatly reduced resistance to flow. When dextran 250 is added to the plasma, the erythrocyte aggregation is enhanced, and the red cell aggregation is “tighter”, making the plasma wall layer even larger and reducing the flow resistance even further. A detailed graph of the literature data for the Fåhraeus -Lindqvist Effect is in Pries et al. [45]. Considering the Fåhraeus and Fåhraeus-Lindqvist Effects together, one asks whether or not the Fåhraeus-Lindqvist Effect is the result of the Fåhraeus Effect alone. This question has been approached in two ways: (1) In the first, a model of the blood flow is constructed with a wall layer of plasma and a core containing the red cells and plasma. The velocity profile calculated from this model is not parabolic due to the wall layer, but the average velocity of the blood must equal the measured flow rate, and the average hematocrit in the tube must match the measured Fåhraeus Effect. Using the measured Fåhraeus-Lindqvist Effect and the Fåhraeus Effect, the width of the plasma layer can be calculated from this model. Typically, this wall layer thickness is about 0.5 - 3 microns. In reality, the wall layer thickness varies in the axial direction [51], and so this thickness must be considered to be an average width. Accurate measurements of this wall layer thickness requires high speed optical recording of the flow. From such recordings, this average wall layer thickness is found to be comparable to that calculated with the model. This implies that the continuum model of blood can be used to predict these flows. (2) The other technique uses the Mooney-Rabinowitz-Weissenberg (MRW) method (see Basic Aspects of Rheology, chapter II.1) to analyze the flow data from small tubes. This method requires that all data, regardless of the tube diameter, must fall on one curve when plotted as the wall shear stress versus (8U/D), where U is the average flow velocity and D is the diameter of the tube. Consequently, if one so plots the data for flow in a small tube where the Fåhraeus and Fåhraeus-Lindqvist Effects occur, and compares those data with data for blood flow in a large tube with the same average hematocrit in the larger tube as in the small tube, all the data should fall on one curve. This approach was used to analyze data obtained for blood flow in a 29 micron diameter tube and the corresponding data from flow in an 811 micron tube, and the results are reproduced in Figure 10.


59

W w (dynes/cm2)

100

10

1

0.1 1

100

10 U

1000

(sec-1)

Figure 10. Use of the MRW method to analyze blood flow data of 29 and 811 micron diameter tubes, with the average hematocrit of the blood in the tubes being equal. Reproduced with permission from [52].

In Figure 10, the solid curves represent the data from the 811 micron diameter tube, and the points represent the data obtained from the 29 micron tube, with the average hematocrits in the two tubes being equal. The fact that the two data sets for a given tube hematocrit define the same curve indicates that the MRW method was satisfied, and that the effective viscosity of the blood in the smaller tube can be predicted from the macroscopic rheological properties of the blood, if one uses the correct, measured average tube hematocrit. Some evidence indicates that this is true even for tubes with diameters down to about 20 microns [53]. While this makes it easy to predict the pressure drop – flow rate relationship for blood flow in tubes greater than 20 microns in diameter using macroscopic rheological data, the fact that the MRW method is satisfied is just a coincidence. In reality, there is a non-uniform red cell distribution in such blood flows due to the wall layer of plasma, and this violates one of the conditions imposed in the derivation of the MRW method. Nevertheless, from a practical viewpoint, this finding also supports the concept that one can predict blood flow behavior in small tubes or vessels from the macroscopic rheological properties of the blood.

60


The famous Whittaker-Winton paper [54] reported that the in-vivo effective viscosity of blood, as measured in the hindlimb of the dog, is about half that found with a glass capillary viscometer. As these authors suggested, this difference in effective viscosities is due to the Fåhraeus-Lindqvist Effect, as can be demonstrated with a very simple model of the hindlimb vascular network which utilizes the Fåhraeus-Lindqvist Effect. 3.3. Flow in Networks of Vessels The average hematocrits of blood flowing in in-vivo networks of small vessels in the microcirculation have been measured by optical densitometry. While these measurements have an inherent uncertainty in them because the method is calibrated with blood in circular artificial tubes and the cross sections of in-vivo vessels are not circular, they nevertheless are a good indication of the hematocrits in such vessels. Such measurements have been made by Lipowsky et al. [55] and Kanzow et al. [56], as well as by others. For vessels which are smaller than about 30 microns in diameter, one can determine vessel average hematocrits from video records of the flow using tracer fluorescent red blood cells [57]. Figure 11 shows some data for in-vivo vessel average hematocrits as a function of vessel diameter.

Figure 11: The ratio of the vessel average hematocrit to the systemic hematocrit versus vessel apparent diameter, for arterial, capillary and venous vessels in the rat mesenteric microcirculation. The curves are for the Fåhraeus Effect for two blood conditions. Adapted from Kanzow et al. [56].


61

There are a number of characteristics of Figure 11 that are striking. First, the data in the larger vessels are not as scattered as the data for the vessels where the diameter approaches that of capillaries. For the capillaries, the distribution of observed hematocrits is very much dispersed. The solid curve on the graph is the Fåhraeus Effect as determined for suspensions of red cells in saline at room temperature, and the dashed curve represents the Fåhraeus Effect for red blood cells suspended in plasma at a temperature of 37°C. Consequently, the second outstanding feature of this graph is that the Fåhraeus Effect can not alone explain the very low hematocrits seen in many of the capillaries and neighboring vessels. In some of the vessels, the hematocrit ratio can exceed unity, as shown by Lipowsky et al. [58] and by Desjardins and Duling [59]. In the later paper, the data are reported only for vessels in which the flow was very fast, and all those data are clustered about a hematocrit ratio of unity regardless of the vessel diameter. These characteristics of the vessel hematocrit distribution must be explained by some mechanism which has not yet been considered here. This mechanism is basically due to the Fåhraeus Effect, and the way red blood cells are distributed at an arterialtype bifurcation from the inflow vessel into each of the two downstream vessels. This is illustrated in Figure 12.

Figure 12: Blood flow through a series of small diameter arterial-type bifurcations, showing an idealized hematocrit distribution at various axial positions. Reproduced with permission from [53].

If the blood coming into the left inflow vessel has traveled a long distance in that vessel, the hematocrit profile across the vessel cross section will be axisymmetric and show a wall layer of plasma and a core of erythrocytes in plasma. This hematocrit distribution is represented by a “top hat” type of distribution, where no erythrocytes are in the wall layer and the hematocrit in the core is constant, regardless of radial position. As the flow approaches the first bifurcation the flow is laminar since the flow has a very low Reynolds number (i.e., inertial forces are extremely small compared to the viscous forces). The dashed line is a separation surface: all the fluid above this curve flows ahead into the continuing downstream vessel and all the fluid below this surface flows into the smaller downstream vessel which is pointed downward. It has been shown that for many flow distributions into the downstream vessels this separation surface is relatively planar [60, 61]. The flow into the smaller vessel contains the wall layer next to it, as well as some of the core region of the inflow vessel. Consequently, the average vessel hematocrit in this vessel will be lower than the average hematocrit in the inflow vessel, and its value will depend on the fraction of the inflow that goes into it. The blood flowing in the continuing main vessel will have to move towards the lower surface of the vessel, as indicated by the streamlines shown in the figure. Because the Reynolds number is so low, the red cells will follow the streamlines, so

62


that the hematocrit profile downstream from the first bifurcation will no longer be axisymmetric, and the wall layer at the lower surface of the vessel will be extremely small. At the second bifurcation, the process repeats itself, except that because of the non-axisymmetric hematocrit profile in the upstream vessel, the hematocrit in the smaller downstream vessel is higher than the average hematocrit in the inflow vessel. Of course, what happens at the second bifurcation depends on the orientation of the second bifurcation relative to that of the first bifurcation. If the second bifurcation had the smaller side branch headed upward, the hematocrit in the side vessel would have been very small. At the third bifurcation, with an almost equal flow into the two downstream vessels, one downstream vessel will have a very low hematocrit and the other a very high hematocrit. Therefore, it is possible for the average relative vessel hematocrit, as well as the relative hematocrit of the blood flow from a vessel, to be greater than unity and as low as zero. This is why the earlier cited experimental data have such a wide range of values.

1.0

RBC Flux fraction

0.8

0.6

0.4

HD= 20% Diameter= 20 μm *

0.2

0.0 0.0

0.2

0.4 0.6 Flow fraction

0.8

1.0

Figure 13: In-vitro data for blood flow through a bifurcation with all vessel diameters equal to 20 microns, with the inflowing blood average hematocrit of 20%, and with the inflowing vessel very long so that the hematocrit profile in the entrance to the bifurcation is axisymmetric. Reproduced with permission from [63].


63

Experimental data have been obtained which describe the characteristics of such flows through bifurcations. An example of in-vitro data is shown in Figure 13. This figure is a plot of the fraction of the red cells flowing into the bifurcation that go into one of the downstream vessels, F*, as a function of the fraction of the blood volumetric flow going into the same vessel, Q*. The asterisks indicate data when the downstream vessel examined is directed downward as was the first bifurcation in Figure 13, and the triangles represent data for the flow directions shown in the lower right diagram in the figure. If the fraction of red cells going into a downstream vessel were equal to the fraction of the blood volume going into that vessel, the data would have fallen on the dashed diagonal line. The fact that, within the precision of the measurements, the data all define one curve representing F* as a function of Q* regardless of the flow directions is to be expected, because of the low Reynolds number of the flow and the fact that all the vessels were the same size. As shown by the non-zero intercept of the curve on the F* axis, there is a minimum flow rate into the downstream vessel below which no red cells flow into the branch; the downstream vessel then only receives plasma from the upstream vessel wall layer.

1.0

RBC Flux fraction

0.8

0.6

0.4

Ȝ 0.08, 0.16, 0.40 0.40, 0.80 1.00

Ȝ=0.04

0.2

Ȝ=1.00 0.0 0.0

0.2


0.8

1.0

Figure 14: F* versus Q* for bifurcations with all vessels the same size and axisymmetric inflows. O is the ratio of the red cell maximum diameter to the vessel diameter. The dotted lines represent in vitro data for human blood in tubes of 20, 50 and 100 microns diameter [62]. The solid lines are for model studies using flexible disks suspended in silicone oil [63], and the dashed line is for data from rabbit ear capillaries [64].

64


Data for single bifurcations, with axisymmetric inflow, are depicted in Figure 14. As the lines in the figure show, the fraction of the incoming red cells which go into a branch depends strongly on the branch diameter, and the minimum flow fraction required to get any red cells in the branch is also strongly dependent on the O value. Up to this point, data from only single bifurcations with axisymmetric inflows have been described. Enden and Popel [61], in their theoretical study, found that the shape of the dividing surface separating the branch flow from the continuing flow was affected by the value of O . As the branch vessel diameter decreased, the separation surface changed from an almost flat surface when the vessel diameters were about equal to a semi-circular cylindrical surface, budging away from the branch, when the diameter ratio was 0.2 (Q*=0.2). This indicates, at this volume flow split, that as the branch diameter decreases, the average hematocrit of the blood flowing into the branch increases since the amount of core blood from the inflow vessel going into the branch increases and the amount of plasma wall layer decreases.

1.0

RBC Flux fraction

0.8

0.6

0.4

0.2

0.0 0.0

0.2


0.8

1.0

Figure 15: The F* vs. Q* plot for two successive bifurcations, with axisymmetric flow into the first bifurcations. The curve represents the data for the first bifurcation. The asterisks represent data at the second bifurcation when Q* at the first bifurcation is 0.50; the diamonds are the data for the second bifurcation when Q* at the first bifurcation is 0.30. Reproduced with permission from [62].


65

In vivo blood flow networks consist of multiple series of bifurcations. Since it takes about 10 seconds of flow time after a bifurcation to reestablish an axisymmetric hematocrit profile in a vessel, it is unlikely that a non-axisymmetric profile is altered very much between successive bifurcations. Consequently, the F* versus Q* plot for the second and subsequent bifurcations in a series of bifurcations will depend on what happens at the previous bifurcations. This is illustrated with in-vitro data in Figure 15.

Figure 16: Blood flow through an arterial bifurcation in the rat mesentery. In the upper picture, the flow is occurring in its unperturbed state, whereas in the lower picture, the flow in the right hand vessel of the bifurcation has been reduced by partial occlusion of the vessel downstream from the picture. The arrows indicate the flow directions. Reproduced with permission from [66].

66


From these data, it appears that as the fraction of the blood flow which goes into the branch vessel of the first bifurcation increases, the data for the branch of the second bifurcation shift downward. This occurs because with the alternating orientation of the branches, as the blood flow fraction increases at the first bifurcation the hematocrit profile between the vessels is shifted downward in the figure, creating a larger plasma layer at the upper wall of the continuing vessel before the second bifurcation. Of course, the orientation of the second bifurcation relative to the first bifurcation will have a significant effect on red cell distribution at the second bifurcation. This bifurcation effect also occurs in-vivo. An example of this was published in Pries et al. [65] and is shown in Figure 16. This figure shows blood flow through a rat mesenteric arterial-type bifurcation, first in its self-regulating state, and then with flow in the right hand branch retarded. In the latter picture, one can clearly see the non-axisymmetric hematocrit distribution in the branch, with a plasma layer at the upper surface, due to skimming off of the plasma layer that is in the inflow vessel. By adjusting the occlusion extent on the side branch, Pries et al. [61] were able to obtain an F* - Q* plot for this bifurcation. This is shown in Figure 17.

1.0

RBC Flux fraction

0.8

0.6

0.4

0.2

0.0 0.0

0.2


0.8

1.0

Figure 17: The F* - Q* for the bifurcation shown in Figure 16. Reproduced with permission from [66].


67

In this figure, the open circles represent data from the right hand branch, and the filled circles are for the continuing vessel. These two sets of data are not independent, since the sum of the two F*s must equal one, as does the sum of the two Q*s. There is some uncertainty in these data, because of the methods of instrument calibration and the assumption that the vessels are circular in cross section [62], but they are nevertheless qualitatively correct. It is clear that this bifurcation red cell distribution effect greatly affects red cell distribution (and oxygen delivery) in the microcirculation. 3.4. Flow in Microvascular Networks of Vessels A considerable amount of in-vivo data on blood flow in the microcirculation has been gathered (e.g., [55, 68]). Other than data on flow direction and pressure drop via micropipette, these data are only qualitative, because the instruments used to measure the data are calibrated using axisymmetric blood flow in circular cross section tubes, whereas in-vivo vessels are not circular in cross section and have varying dimensions and cross-sectional shapes in the axial direction. While in-vitro data are generally accurate, they are not directly applicable to in-vivo blood flow since they generally obtained using blood flow in circular cross section tubes whose diameter does not vary in the axial direction. In spite of these limitations, both in-vivo and in-vitro data have been used to model blood flow in microcirculatory networks of vessels. A very early physical model of the microcirculation was made by A. Fick [69], who constructed his model from glass tubes and rubber connectors; the model consisted of seven levels of vessels in a tree-like trifurcating system with arterial, capillary and venous vessels. The vessels at each level were of different diameters with the smallest vessels being model capillaries. Pressures were measured with manometers at each level of the model. His conclusion from his experimental data was that the major pressure loss occurs in the post-capillary vessels. This erroneous conclusion arises, at least in part, because all his vessels except “capillaries” had turbulent flow, unlike in in-vivo microcirculatory blood flow which is always laminar. While models such as Fick’s are overly simplistic, they can be used to estimate average flow behavior in large networks of vessels. For example, one can model mathematically the Whittaker and Winton experiment [54], in which suspensions of red blood cells in serum were pumped through the isolated hindlimb of the dog, and overall pressure drop and flow rate data were measured. The conclusion from this study was that the effective viscosity of the suspensions, as calculated with the data and Poiseuille’s equation, was about half that of the viscosity of the blood as measured in a tube viscometer. A very simple mathematical model, tree-like in nature, of this experiment can be constructed with the vessel diameter, length and number data for a generic vascular network from the main artery branch to the main venous branch contained in Johnson [70]. From the average hematocrit data for various in-vivo vessels in Lipowsky et al. [55], one can assign to each level of vessel in the model network a vessel hematocrit. Using an appropriate high shear rate apparent viscosity for each vessel, one can then calculate the overall pressure drop for a given overall flow rate in the network. In a separate calculation, one may assume that the hematocrit in each vessel is equal to the systemic hematocrit, and calculate an overall pressure drop at the same flow rate as used in the first calculation. The ratio of the two pressure drops from these calculations is about 0.59, while Whittaker and Winton measured a ratio of 0.52. This simple model calculation strongly supports the idea that the

68


Whittaker-Winton Effect is due to the lower average hematocrits found in the smallest vessels of the microcirculation. Although all the necessary data for red cell distribution at bifurcations and pressure drop-flow rate information for non-axisymmetric flow in micro vessels are not available, several investigators have nevertheless constructed mathematical models of blood flow in microvascular networks utilizing the information that is available. A number of different approaches have been used. Pries, et al. [71] used data for network topology and topography, distribution of red cells at bifurcations, effective viscosity as a function of vessel hematocrit, etc., based on in-vivo data, whereas Kiani et al. [72] used the same in-vivo network topology and topography data as Pries et al., but in-vitro data for red cell distribution at bifurcations and the Fåhraeus and Fåhraeus-Lindqvist Effects. There is a significant difference in the method for assigning vessel hematocrit to each vessel: the Pries et al. method calculates discharge hematocrits at all nodes in the network at an instant in time, then assigns to each vessel an average hematocrit based on its discharge hematocrit and the Fåhraeus Effect. The effective viscosity for each vessel is obtained from the Fåhraeus-Lindqvist Effect. The Kiani et al. method keeps track of “slugs” of blood as they travel through the network so that at any instant, the hematocrit varies along a vessel length. The latter method volume averages the hematocrit in a vessel, and based upon this hematocrit an effective viscosity to that vessel. As a consequence of these methods for assigning vessel hematocrits, the Pries et al. method results in steady flows in all vessels whereas the Kiani et al. method shows stationary oscillatory flow in many vessels in a network [72, 73]. Regardless of the details of the computational model, comparisons of predicted vessel parameters to corresponding in-vivo measured parameters are not satisfactory: comparisons of individual vessel hematocrits are very poor and, while somewhat better, comparisons of flow velocities are also not very good [53]. These difficulties arise from two sources: the inaccuracies of the in-vivo data and the inapplicability of the invitro data for the reasons mentioned above. A review of the foundations of modeling the microcirculatory blood flow is given by Popel [74], and in-vivo data are available for several microvascular networks and two rat mesenteric networks on the internet at www.physiology.arizona.edu/people/secomb/network.html. Note, however, that in spite of the limitations of current models, they do duplicate many average in-vivo microcirculatory characteristics [75, 76].

References [1] [2] [3] [4] [5] [6] [7] [8]

S.P. Salvatore and R. Skalak, The history of Poiseuille’s law. Am. Rev. Fluid Mech. 25 (1993), 1-19. R.F. Probstein, Physicochemical Hydrodynamics: An Introduction, Buttersworth, Boston, 1989, pp. 139-135. G.R. Cokelet, J.R. Brown, S.O. Codd and J.D. Seymour, Magnetic resonance microscopy determined velocity and hematocrit distributions in a Couette viscometer. Biorheology 42 (2005), 385-399. S. Usami, S. Chien and M.I. Gregerson, Viscometric characteristics of blood of the elephant, man, dog sheep and goat. Am. J. Physiol. 217(1969), 884-890. S. Chien, Shear dependence of effective cell volume as a determinant of blood viscosity. Science 168 (1970), 977-978. G.R. Cokelet, E.W. Merrill, E.R. Gilliland, H. Shin, A. Britten and R.E. Wells, The rheology of human blood-measurement near and at zero shear rate. Trans. Soc. Rheology 7 (1963), 303-317. G.R. Cokelet, Rheology of Human Blood, ScD Thesis, Massachusetts Institute of Technology, Cambridge, MA, 1963. G.R. Cokelet, The rheology of human blood, In: Biomechanics: It’s Foundations and Objectives, Y.C. Fung, N. Perrone and M. Anliker, Eds., Prentice Hall, New Jersey, 1972, pp. 63-103.

G.R. Cokelet and H.J. Meiselman / Macro- and Micro-Rheological Properties of Blood [9] [10]

[11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39]

69

S. Chien, Biophysical behavior of red cells in suspensions, in: The Red Blood Cell, D.M. Surgenor, Ed., Academic Press, New York, 1975, pp. 1031-1133. J.A. Dormandy, Measurement of whole blood viscosity, In: Clinical Aspect s of Blood Viscosity and Cell Deformability, G.D.O. Lowe, J.C. Barbanel and C.D. Forbes, Eds, Springer-Verlag, New York, 1981, pp. 67-78. G.R. Cokelet, The rheology and tube flow of blood, In: Handbook of Bioengineering, R. Skalak and S. Chien, Eds., McGraw-Hill, New York, 1987, pp. 14.1-14.17. G.D.O. Lowe and J.C. Barbanel, Plasma and blood viscosity, In: Clinical Blood Rheology, G.D.O. Lowe, Ed., CRC Press, Boca Raton, 1988, pp. 11-44. T. Somer and H.J. Meiselman, Disorders of blood viscosity, Ann. Med. 25 (1993), 31-39. J.F. Stoltz, M.Singh and P. Riha, Blood viscosity and models, J.F. Stoltz, M. Singh and P. Riha, Eds., IOS Press, Amsterdam, 1999, pp. 15-26. O.K. Baskurt and H.J. Meiselman, Blood rheology and hemodynamics, Sem. Thromb. Hemost. 29 (2003), 435-458. O.K. Baskurt, O. Yalcin and H.J. Meiselman, Hemorheology and vascular control mechanisms. Clin. Hemorheol. Microcirc. 30 (2004), 169-178. S. Usami, Development of hemorheology, Clin. Hemorheol. Microcirc. 23 (2000), 77-83. W.H. Reinhart, Molecular biology and self-regulatory mechanisms of blood viscosity, Biorheology 38 (2001), 203-212. W.D. Corry and H.J. Meiselman, Modification of erythrocyte physicochemical properties by millimolar concentrations of glutaraldehyde, Blood Cells, 4 (1978), 465-480. L. Dintenfass, Hyperviscosity in Hypertension, Pergamon Press, Sydney, 1981, 250pp. M.A. Lichtman, J. Heal and J.M. Rowe, Hyperleukocytic leukemia: rheological and clinical features and management, Baillieres Clin. Hematol. 1 (1987), 725-746. E.W. Merrill, E.R. Gilliland, G. Cokelet, H. Shin, A. Britten and R.E. Wells, Rheology of human blood near and at zero flow, Biophys. J. 3 (1963), 199-213. E.W. Merrill, G.C. Cokelet, A. Britten and R.E. Wells, Non-Newtonian rheology of human blood, Circ. Res. 13 (1963), 48-55. N. Casson, A flow equation for pigment-oil suspension of the printing ink type, In: Rheology of Disperse Systems, C.C. Mills, Ed., Pergamon Press, New York, 1959, pp 84-104. H.J. Meiselman, E.W. Merrill, E. Salzman, E.R. Gilliland and G.A. Pelletier, The effect of dextran on the rheology of human blood, J. Appl. Physiol. 22 (1967), 480-486. B.K. Lee, T. Alexy, R.B. Wenby and H.J. Meiselman, Red blood cell aggregation quantitated via Myrenne Aggregometer and yield shear stress, Biorheology (in press, 2007). E.W. Merrill, A.M. Benis, E.R. Gilliland, T.K. Sherman and E.W. Salzman, Pressure-flow relations of human blood in hollow filters at low flow rates. J. Appl. Physiol. 20 (1965), 954-967. B. Deuticke, Transformation and restoration of biconcave shape of human erythrocytes induced by amphiphilic agents, Biochim. Biphys. Acta 163 (1968), 494-502. H.J. Meiselman, Morphological determinants of red cell deformability, Scand. J. Clin. Lab. Invest. 41 (1981), 27-34. O.K. Baskurt and H.J. Meiselman, Cellular determinants of low shear blood viscosity, Biorheology 34 (1997), 235-247. F.J. Walburn and D.J Schneck, A constitutive equation for whole human blood, Biorheology 13 (1976), 201-218. C.M Rodkiewicz, P. Sinha and J.S. Kennedy, On the application of a constitutive equation for whole human blood, Trans. ASME 112 (1990), 198-206. X.Y. Luo and Z.B Kuang, A study on the constitutive equation of blood, J. Biomechanics 25 (1992), 929-934. J.B. Zhang and Z.B Kuang, Study on blood constitutive parameters in different blood constitutive equations, J. Biomech. 33 (2000), 355-360. P. Neofytou, Comparison of blood rheological models for physiological flow simulation, Biorheology 41 (2004), 693-714. M. Anand and K.R. Rajagopal, A shear-thinning viscoelastic model for describing the flow of blood, Int. J. Cardiovasc. Med. Sci. 4 (2004), 59-68. P.L. Easthope and D.E Brooks, A comparison of rheological constitutive functions for whole human blood, Biorheology 17 (1980), 235-247. J. Fang and R.G. Owens, Numerical simulations of pulsatile blood flow using a new constitutive equation, Biorheology 43 (2006), 637-660. R. Fåhraeus and T. Lindqvist, The viscosity of blood in narrow capillary tubes, Am. J. Physiol. 96 (1931) 562-568.

70


[40] G.R. Cokelet, Macroscopic rheology and tube flow of human blood. In: Microcirculation, J. Grayson and W. Zingg, Eds., Plenum Press, New York, 1976 ,pp 9-31. [41] A.R. Pries, K.H. Albrecht and P. Gaehtgens, Model studies of phase separation at a capillary orifice, Biorheology 18 (1981), 355-367. [42] R. Fåhraeus , The suspension stability of the blood, Physiol. Rev. IX (1929), 241-274. [43] H.L. Goldsmith, G.R. Cokelet and P. Gaehtgens, Robin Fåhraeus : evolution of his concepts in cardiovascular physiology, Am. J. Physiol. 257 (1989), H1005-H1015. [44] P. Gaehtgens, C. Duehrssen and K.H. Albrecht, Motion, deformation and interactions of blood cells and plasma during flow through narrow capillary tubes, Blood Cells 6 (1980), 799-812. [45] A.R. Pries, D. Neuhaus and P. Gaehtgens, Blood viscosity in tube flow: dependence on diameter and hematocrit, Am. J. Physiol. 263 (1992), H1770-H1778. [46] U. Nobis, A.R. Pries, G.R. Cokelet and P. Gaehtgens, Radial distribution of white cells during blood flow in small tubes, Microvasc. Res. 29 (1985), 295-304. [47] H.L. Goldsmith and S. Spain, Margination of leukocytes in blood flow through small tubes, Microvasc. Res. 27 (1984), 204-222. [48] H.L. Goldsmith, Red cell motions and wall interactions in tube flow, Fed. Proceed. 30 (1971),15781588. [49] H.J. Goldsmith, The microrheology of human erythrocyte suspensions, Appl. Mech., Proceed. of 13th Intern. Congress of Theoret. and App. Mech. E. Becker and G.K. Mikhailov, Eds. Springer-Verlag, Berlin 1973, pp 85-103. [50] W. Reinke, P. Gaehtgens and P.C. Johnson, Blood viscosity in small tubes: effect of shear rate, aggregation and sedimentation, Am. J. Physiol. 253 (1987), H540-H547. [51] G.R. Cokelet and H.L. Goldsmith, Decreased hydrodynamic resistance in the two-phase flow of blood through small vertical tubes at low rates, Circ. Res. 68 (1991), 1-17. [52] J.H. Barbee and G.R. Cokelet, Prediction of blood flow in tubes with diameters as small as 29 microns, Microvasc. Res. 3 (1971), 17-21. [53] G.R. Cokelet, Viscometric, in-vitro and in-vivo blood viscosity relationships: how are they related?, Biorheology 36 (1999), 343-358. [54] S.R.F. Whittaker and F.R. Winton, The apparent viscosity of blood flowing in the isolated hindlimb of the dog, and its variation with corpuscular concentration, J. Physiol. Lond. 78 (1933), 339-369. [55] H.H. Lipowsky, S. Usami and S. Chien, In-vivo measurements of “apparent viscosity” and microvessel hematocrit in the mesentery of the cat, Microvasc. Res. 19 (1980), 297-319. [56] G. Kanzow, A.R. Pries and P. Gaehtgens, Flow dependent hematocrit distribution in microvessel networks, Biblthca anat. 20 (1981), 149-152. [57] I.H. Sarelius and D.R. Duling, Direct measurement of microvessel hematocrit, red cell flux, velocity and transit time, Am. J. Physiol. 243 (1982), H1018-H1026. [58] H.H. Lipowsky, L.E. Cram, W. Justice and M.J. Eppihimer, Effect of erythrocyte deformability on invivo red cell transit time and hematocrit and their correlation with in-vitro filterability, Microvasc. Res. 46 (1993), 43-64. [59] C. Desjardins and B.R. Duling, Microvessel hematocrit: measurement and implications for capillary oxygen transport, Am. J. Physiol. 252 (1987), H494-H503. [60] F. W. Rong and R.T. Carr, Dye studies in flow through branching tubes, Microvasc. Res. 39 (1990), 186-202. [61] E. Enden and A.S. Popel, A numerical study of the shape of the surface separating flow into branches in microvascular bifurcations, Trans. A. S. M .E. 114 (1992), 398-405. [62] G.R. Cokelet, Blood Flow through Arterial Microvascular Bifurcations. In: Microvascular Networks: Experimental and Theoretical Studies. A.S. Popel and P.C. Johnson, Eds., Karger, Basel 1985, pp 155167. [63] B.M. Fenton, R.T. Carr and G.R. Cokelet, Nonuniform red cell distribution in 20 to 100 micron bifurcations, Microvasc. Res. 29 (1985), 103-126. [64] S. Chien, C.D. Tvetenstrand, M.A.F. Epstein and G.W. Schmid-Schhoenbein, Model studies on distributions of white and red blood cells at microvascular bifurcations, Am. J. Physiol. 248 (1985), H568-H576. [65] G.W. Schmid-Schoenbein, R. Skalak, S. Usami and S. Chien, Cell distribution in capillary networks, Microvasc. Res. 19 (1980), 18-44. [66] A.R. Pries, K. Ley, M. Claassen and P. Gaehtgens, Red cell distribution at microvascular bifurcations, Microvasc. Res. 38 (1989), 81-101. [67] G.R. Cokelet and I.H. Sarelius, Perceived vessel lumen and cell-blood ratio: impact on in-vivo blood flow rate determination, Am. J. Physiol. 262 (1992), H1156-H1193. [68] A.R. Pries, T.W. Secomb and P. Gaehtgens, Biophysical aspects of blood flow in the microvasculature, Cardiovasc. Res. 32 (1996), 654-667.


71

[69] A. Fick, Ueber den Druck in den Blutkapillaren, Pfluegers Archiv. 42(1888), 482-488. [70] P.C. Johnson, Ed., Peripheral Circulation, Wiley, New York 1978, pp 3. [71] A.R. Pries, T.W. Secomb, P. Gaehtgens and J.F. Gross, Blood flow in microvascular networks, experiments and simulations, Circ. Res. 67 (1990), 826-834. [72] M.F. Kiani, A.R. Pries, L.L. Hsu, I.R. Sarelius and G.R. Cokelet, Fluctuations in microvascular blood flow parameters caused by hemodynamic mechanisms, Am. J. Physiol. 266 (1994), H1822-H1828. [73] R.T. Carr, J.B. Geddes and F. Wu, Oscillations in a simple microvascular network. Ann. Biomed. Eng. 33 (2005), 764-771. [74] A.S. Popel, Network Models of Peripheral Circulation. In: Handbook of Bioengineering. R. Skalak and S. Chien, Eds., McGraw-Hill, New York 1987 chapter 20. [75] A.R. Pries, T.W. Secomb and P. Gaehtgens, Structure and hemodynamics of microvascular networks: heterogeneity and correlations, Am. J. Physiol. 269 (1995), H1713-H1722. [76] A.R. Pries, T.W. Secomb and P. Gaehtgens, Biophysical aspects of blood flow in the microcirculation, Cardiovasc. Res. 32 (1996), 654-67.

72


Viscoelasticity of Human Blood George B. THURSTONa,1 and Nancy M. HENDERSONb a University of Texas at Austin, USA b Vilastic Scientific, Inc., USA

1. Introduction and Overview of Blood Viscoelasticity 1.1. Time-Varying Flow of Blood in Vivo The most obvious feature of the circulation is the pulse. Pulsatile flow can be analyzed as containing steady plus harmonic components. The heartbeat of 60-120 beats per minute gives a fundamental Fourier component of 1 to 2 Hz. Because the flow is time varying, pressure-flow relations are a function of both the shear viscosity and the shear elasticity of the blood. The viscoelasticity of blood has a direct effect on the propagation of the pulse throughout the arterial system [1]. Blood flow in vivo covers a wide range of shear rates and varied vascular geometry (smooth wall of uniform diameter, tapered vessels, bifurcations, side branches, stenoses). 1.2. Red Cell Concentration and Deformation Normal human blood contains a high concentration of red blood cells (RBC), which are elastic elements. The maximum theoretical volume concentration of red cells without squeezing and deforming is 58%. Because normal cell concentration is in the range from 30 to 60%, flow cannot occur through the varied geometry of the circulation without elastic cell deformation and orientation and hence storage of elastic energy in the cells. Blood flows only because the RBC are deformable and can be reoriented to slide on the low viscosity plasma. The elastic deformability of cells means that energy can be stored in and recovered from cell deformation. The elastic energy is measurable when flow changes with time. Oscillatory flow is particularly useful for measuring this energy and characterizing viscoelastic properties of blood.

2. Measurement Methods The viscoelastic effects in whole blood are evident from several experimental observations. These effects have been seen in both optical and mechanical measurements. Observations are made in time-varying flow where stored elastic energy is recovered. The first identification of blood viscoelasticity was done using oscillatory flow in a cylindrical tube [2, 3] and in an oscillating Couette geometry [4]. The oscillating cone-plate geometry was used both for measurements and visual 1

Corresponding Author: 1000 Madrone Road, Austin, TX 78746 USA; E mail: [email protected]

G.B. Thurston and N.M. Henderson / Viscoelasticity of Human Blood

73

observations of flow-induced cell organization [5, 6]. Stress response to sudden initiation and cessation of flow shows an elastic response [7, 8, 9]. The elastic effects are evident from observations of light reflection from and transmission through blood following sudden cessation of flow [10, 11] and during sinusoidal oscillatory flow [12]. The accuracy and reproducibility of rheological measurements can be affected by the tendency of blood cells to sediment, and by the sample confinement size and geometry. Four confinement geometries have been used to measure the viscoelasticity of blood: cone-plate, plate-plate, concentric cylinder (Couette) and cylindrical tube. The geometry of rotational instruments, such as the cone-plate and plate-plate, can accelerate cell sedimentation during very low shear rotation [13], producing a twophase fluid in the shear plane. A vertically oriented cylindrical tube, with large reservoirs at each end, and the Couette geometry are relatively insensitive to sedimentation during measurements. Because blood viscoelasticity is dependent on the confinement size [14, 15], maintaining a uniform and fixed geometry is important for reproducibility. Therefore, a fixed uniform geometry such as a cylindrical tube is preferable to the variable confinement size of the cone-plate and the user-established gaps of plate-plate and cone-plate geometries. Hence, for the data presented herein, oscillatory flow in a vertically oriented cylindrical tube was utilized (Vilastic-3 Viscoelasticity Analyzer, Vilastic Scientific, Inc., Austin, TX USA). Details of the principle of measurement are given later (see Section 5).

3. Parameters for Shear Flow and Viscoelasticity Consider the action of a steady shearing force on a small cubical volume of blood of height H. The force produces shear stress and shifts the shape of the blood volume to a parallelogram (Figure 1). The displacement, D, of the cube has two components: elastic deformation, E, and slippage, S. The elastic deformation is accompanied by storage of elastic energy within the structure of the blood, while the slippage is associated with a continuous input of viscous energy. When the force is removed, the deformed cube undergoes a partial restoration of shape due to the recovered elastic energy, but the cube remains deformed due to energy loss from slippage. In steady flow, the displacement component continues to increase. Measurements of non-time-varying force and velocity provide no information regarding elastic energy storage. In any time-varying flow, however, the elastic energy component varies with time and may be either increasing or decreasing. Figure 1 can be used to define the parameters: shear stress (F/A), shear rate (V/H), and shear strain (D/H).

Figure 1. Diagram of a small cubical volume of height, H, in shear. The displacement, D, due to deformation is composed of two parts: an elastic part, E, and a slippage part, S. With constant force F, E remains constant but S continues to increase. When F is removed, E diminishes to zero and S remains.

74


Figure 2. Sinusoidal deformation of a cubical volume of fluid. The sinusoidal time varying shear rate and shear stress differ in phase by the angle, I, as shown.

Sinusoidal time varying flow provides a basis for differentiation of the elastic and viscous properties of blood. The time dependences are Shear Rate = J m cos(Z t )

(1)

Shear Stress = W m cos(Z t I)

(2)

S Shear Strain = J m cos(Z t ) 2

(3)

and also,

where Z is the radian frequency (2·S·frequency). Figure 2 shows the cubical volume of blood in oscillatory shear and the time dependence of shear rate and shear stress. For a purely viscous fluid, the phase angle difference (I) between the shear rate and shear stress is 0°. If the fluid is purely elastic then the angle is 90°. For a viscoelastic fluid, such as blood, the angle is between 0° and 90°. The magnitude and phase relation between the stress, strain and shear rate are described by complex numbers. For example Shear Stress = Real Part of (W m e i (Zt I) ) = Real Part of (W * eiZt ) The complex coefficient of viscosity (K*) is given by K*

W* J *

W m e iI J m

(4)

Expanding equation 5 yields K*

W m (cos(I) i sin(I)) J m

K*

Kc iKcc

c W'm iWcm J m

(6)

where K’ is the viscosity and K” is the elasticity. Shear stress relations are W*

c Wcm iWcm

(8)

75


c is the elastic stress. Similarly, the complex where Wcm is the viscous stress and Wcm modulus of rigidity, G*, can be determined from the complex ratio of the shear stress, W*, to the shear strain, J*. G*

c Wc Wcm i m Jm Jm

W* J*

G c i G cc

(9)

where G’ is the storage modulus and G” is the loss modulus. The complex coefficient of viscosity, K*, is related to the complex rigidity by G* i Z K * ; G ' Z Kcc ; G cc

Z Kc

(10)

The three basic parameters: stress, strain and shear rate, are often measured as rms (root mean square) values as opposed to maximum values. The maximum values can be replaced by the product of the rms value times 2 . When utilizing rms values the subscript “m” is removed from the above equations. With sinusoidal oscillatory flow, viscous energy is dissipated while elastic energy is alternately stored and recovered. The energy dissipated per unit volume in each cycle is proportional to the viscosity Ed

§ 2S· Kc J 2 ¨ ¸ © Z ¹

(11)

while the maximum energy per unit volume stored and recovered twice during the cycle is proportional to the elasticity, Es

§1· Kcc J 2 ¨ ¸ © Z¹

(12)

We can show the viscoelastic action for each measurement by an apparent relaxation time that can be obtained by assuming the material behaves as a single Maxwell element. The single element Maxwell model consists of a spring (P) and dashpot (K) in series. For sinusoidal stress and strain the viscosity and elasticity are Kc

Kcc

K 1 Z Tr 2 K Z Tr 1 Z Tr 2

(13)

(14)

where Tr = (K/P) is the relaxation time. From the ratio of Equations. (13) and (14), the Kcc tan(I) . The relaxation time can also be apparent relaxation time is Tr c Z ZK obtained from the stored and dissipated energies indicated in of Equations. (11) and (12), Tr

Kcc Z Kc

Es Ed f

(15)

76


where f is the frequency. Therefore, the relaxation time is a direct indicator of the energy ratio, which depends on how energy is stored and dissipated by the microstructure of the blood. Because of the complexity of the microstructure of blood, there are many modes where elastic energy can be stored and recovered and viscous energy dissipated. The result is a very wide spectrum of relaxation times. This spectrum is sensitive to details of how the red cells are organized and interact, so it is sensitive to the rate of flow, vessel size, etc. The rheology of blood must be studied over a wide range of timings, and one way to do this is to examine its viscoelastic properties for a wide range of frequencies. The Maxwell relaxation time obtained using a single Maxwell element (Equations 13-15) is ultimately determined by the complete relaxation spectrum for many Maxwell elements in parallel. Each has its own characteristic spring (constant Pi) and dashpot (constant Ki) and hence its own relaxation time, Ti= (Ki /Pi) [16, 17]. When the shear rate (and shear stress) is increased, a point will be reached where the Ki and Pi diminish in equal proportions, affecting the longest relaxation times first. Thus, the relaxation spectrum becomes modified. With increasing shear rate further degradation of the longest relaxation process will occur along with degradation of the contribution from shorter relaxation processes. The contribution of the shortest relaxation processes will diminish at the higher shear rates and stresses.

4. Moduli for Blood 4.1. Viscoelasticity Profile

A profile for the viscoelasticity of blood is obtained from measurements near the pulse rate (2 Hz) while progressively increasing the rate of flow in a 1 mm inside diameter (ID) cylindrical tube. For this tube size and frequency, the velocity across the diameter in the cylindrical tube is parabolic (see section 5.). These measurements show the viscosity and elasticity and consequently, how the cellular arrangements are affected by the rate of flow. Figure 3 shows how the viscosity and elasticity change with shear rate at the wall (rms values) over a range comparable with flows in vivo. The 44% hematocrit blood was measured at 22 °C. At the lowest shear rates, near 1 s-1, the blood is nearly in its resting state; in that the rate of flow and internal stresses are small and do not materially disrupt the quiescent state of cell aggregates. With increasing shear rate, the large aggregates begin to break-up into smaller collections of cells. At shear rates above 100 s-1, the disaggregated cells form a layered-like structure where layers of stretched and aligned cells slide on separating plasma layers [18]. The disruption of aggregates and formation of cell layers induced here by oscillatory flow can also occur with steady flow and pulsatile flow [19, 20]. The levels of the viscous and elastic components of the shear stress (Equation 8) for the data in Figure 3 are shown plotted versus the rms value of the shear strain in Figure 4. The dependence of the elastic stress on strain shows that the blood is capable of storing elastic energy with increasing strain. The dashed line adjacent to the elastic stress marks a linear relation between stress and strain. The progressively decreasing ability of the elastic stress to follow this linear relationship is indicative of the degradation of cell aggregates. The maximum elastic shear stress occurs near unit strain. This maximum represents an elastic yield stress, which is the maximum stress the


77

structure can accommodate without major disruption. At strains above the yield value, the aggregates undergo a catastrophic break-up, which is shown by the decreasing ability to store elastic energy. The viscoelasticity profile and elastic yield stress are affected by several factors such as aggregation tendency, cell deformability, hematocrit and plasma composition. [21, 22, 23]. The Maxwell relaxation times for the data in Figure 3 are shown in Figure 5. At low shear rates, the Maxwell relaxation time is nearly constant at 0.027 seconds and diminishes with increasing shear rate. The constant value at low shear rates is indicative of the relaxation characteristics of the aggregates. As the aggregate size diminishes with shear rate, the relaxation time decreases. At high shear rates, the relaxation time reflects the elastic character of the stretched and aligned cells. 100 Viscosity

Viscosity and Elasticity (mPa s)

Elasticity

10

1

0.1

0.01 1

10

100

1000

Shear Rate (1/sec) Figure 3. Viscoelasticity Profile. The viscosity and elasticity for a range of shear rates measured at 2 Hz, 22 °C in a 0.1 cm ID cylindrical tube with length of 6 cm. The blood sample is from a normal donor with a hematocrit of 44%.

4.2. Steady Flow vs. Oscillatory Flow Behavior

Profiles for frequencies from 0.1 Hz to 2 Hz, along with the profile for steady flow (zero frequency), are shown in Figure 6. The difference between steady flow viscosity and oscillatory viscoelasticity is that in steady flow the strain is progressively increasing without limit, whereas in oscillatory flow the strain is limited and reversing periodically. Consequently, the elastic energy stored in steady flow is not recovered and remains hidden. All profiles (steady and oscillatory) tend to merge at high shear rates. This behavior is consistent with the truncation of relaxation processes with increasing shear as postulated in the generalized Maxwell model for blood [17]. The viscoelasticity profile changes with the frequency of the oscillation because of the relationship between the speed of deformation and the characteristic time of the blood structure, the Maxwell relaxation time (Equation 15). At low shear rates, where blood is in its quiescent states, the value of viscosity decreases by 16% and the elasticity decreases by 34% when changing the frequencies from 1 to 2 Hz (Figure 6).


Figure 4. Viscous and Elastic Stress. Measurements were made at 2 Hz at 22°C in a 0.1 cm ID cylindrical tube with length of 6 cm. The blood sample is from a normal donor with a hematocrit of 44%. Results are for viscosity and elasticity data versus shear rate presented in Figure 3.

0.1 Maxwell Relaxation Time

Maxwell Relax. Time (s)

78

0.01

0.001

0.0001 1

10

100

1000

Shear Rate (1/sec) Figure 5. Maxwell relaxation times for 44% hematocrit blood measured in a 0.1 cm ID tube with length of 6 cm. Measurements were made at 2 Hz, 22 °C. Results are for viscosity and elasticity data versus shear rate presented in Figure 3.

79


Figure 6. Viscoelasticity profile for frequencies from 0.1 Hz to 2 Hz, along with the profile under steady flow (zero frequency). Measurements were made at 22 °C. Steady flow measurements were performed with a concentric cylinder Couette viscometer (Contraves LS30) with a gap of 0.5 mm. Oscillatory flow measurements were made with a 0.1 cm ID cylindrical tube with length of 6 cm.


100 Viscosity (Normal) Elasticity (Normal) Viscosity (Hardened) Elasticity (Hardened)

10

1

0.1

0.01 1

10

100

1000

Shear Rate (1/sec) Figure 7. Viscoelastic properties of normal blood (45% hematocrit) and a suspension of heattreated RBC in plasma (45% hematocrit) measured at 2 Hz. Measurements were made in a 0.1 cm ID cylindrical tube with length of 6 cm. The temperature was 22°C.

80


4.3. Deformability

For blood to flow in the circulatory system, the red blood cells must be deformable. With decreased deformability, the pressure required to maintain flow will necessarily increase. In the extreme case, as in sickle cell disease, microvessels may become plugged [24]. The elasticity obtained via a viscoelasticity profile is a direct measure of deformability because it relates directly to the amount of energy storage due to the elastic deformation of the red blood cells (Equation 12). At high shear rates, a decrease in deformability causes an increase in elasticity and Maxwell relaxation times. Deformability of red blood cells can be decreased by treatment with acetaldehyde [23] or heat [25]. Figure 7 compares the viscoelastic profile for normal blood with a suspension of heat-treated RBC in plasma. To prepare the heated treated RBC sample, cells washed with phosphate buffered saline were subjected to a 48 °C bath for 9 minutes and then 22 °C for 1 minute. The cells were re-suspended in autologous plasma [26]. At low shear rates, the viscosity and elasticity are slightly elevated above the normal values. Because of the weaker aggregates in the hardened sample, the elasticity begins to diminish at lower shear rates than in the normal blood sample. This may be due to a decrease in cell-to-cell contact area. At high shear rates, the elevated elasticity of the heated cells can be associated with decreased ability of the cells to conform into layers. As opposed to heat treatment, intense hardening of cells by acetaldehyde causes cells to resist stable layer formation because of their limited ability to conform to neighboring cells and to release trapped plasma. The result is a strong dilatancy, where the decreasing elasticity trend of normal blood at high shear rates is reversed [23]. 4.4. Aggregation

The tendency of red cells in blood to aggregate is affected by the concentration of plasma proteins. In particular, fibrinogen concentration plays a significant role for red cell aggregation in plasma suspensions [27]. The tendency of red blood cells to aggregate can be also be enhanced by the addition of high molecular weight dextrans [28] such as dextran 150 (Mw=150,000) [29] and dextran T100 (Mw=100,000) [30]. Dilution of plasma will lower the concentration of all plasma components, including fibrinogen, resulting in a reduction in aggregation tendency. The effects of this dilution are seen in Figure 8, where normal blood is compared with blood containing plasma diluted by 75% with isotonic saline. At low shear rates, the dilution diminishes the viscosity by 37% and the elasticity by 60%, most likely due to the reduction of aggregate size in the diluted plasma sample. At high shear rates where cells are disaggregated, the values of elasticity are nearly equal and the reduction in viscosity of the diluted plasma sample is due to the overall reduction in plasma viscosity. 4.5. Temperature

While normal body temperature is approximately 37 °C, rheological evaluations of blood are often done near "room" temperatures between 20 and 25 °C. The effects on viscoelasticity are significant and therefore, when comparing samples, care must be taken to match temperatures. Figure 9 shows the viscoelasticity profile for temperatures of 15, 22 and 37 °C. It is seen that the character of the viscosity and

81


elasticity is the same but values are shifted with temperature. The functional form is K=D10-ET. At unit strain (shear rate = 12.56 s-1) for a normal blood at 2 Hz and the temperature in the range of 9 and 42 °C, the equations for the blood sample in Figure 9 are [31]

K c 21.47 10 0.01264T

[mPa.s]

(16)

K cc 4.213 10 0.01022T

[mPa.s]

(17)

[mPa.s]

(18)

K plasma

3.362 10 0.01017T

where T is temperature in °C. While the coefficient, D, is specific to the data in Figure 9, the exponent, E, can be assumed the same for other normal blood samples at this strain. Therefore, it can be used to project normal data acquired at 2 Hz and a strain of 1 (shear rate = 12.56 s-1) to a reference temperature, To KTo KT 10 ETo T

(19)

where K(T) is the viscosity or elasticity at the temperature (T), To is the target temperature and E is experimentally determined at a specific shear rate.


100

10

1

0.1 Viscosity (Normal Plasma) Elasticity (Normal Plasma) Viscosity (Diluted Plasma) Elasticity (Diluted Plasma)

0.01 0.1

1

10

100

1000

Shear Rate (1/sec) Figure 8. Viscosity and elasticity versus shear rate for 44% hematocrit cells in normal plasma and for the same cells in plasma diluted by 75% with isotonic saline. Measurements were made at 2 Hz and 22°C in 0.1 cm ID cylindrical tube with length of 6 cm.

82



100

10

1

Viscosity (15°C) Viscosity (22°C) Viscosity (37°C) Elasticity (15°C) Elasticity (22°C) Elasticity (37°C)

0.1

0.01 0.1

1

10

100

1000

Shear Rate (1/sec) Figure 9. Viscoelasticity profile for blood at three temperatures. The frequency of measurement was 2 Hz and the hematocrit was 47%. Measurements were made in a 0.1 cm ID cylindrical tube with length of 6 cm.

4.6. Hematocrit

The viscoelasticity of blood is strongly dependent on hematocrit (i.e., volume fraction red cells). The normal range of hematocrits varies with age and gender and can range from 30% to 60% [32]. Figure 10 shows viscoelasticity profiles for blood with hematocrits of 35%, 45%, 55% and 65%. These samples were derived from a single donor by separating and re-suspending red blood cells in plasma at the desired hematocrit. Measurements were made at 2 Hz and 22 °C. While the character of the curves remains essentially the same for all suspensions, their magnitudes differ significantly with hematocrit. For Figure 10, the functional dependence of viscoelasticity on hematocrit at a shear rate 10 s-1 is given by

K c 1.9736 exp(3.2361 H )

[mPa.s]

(20)

K cc 0.15344 exp(5.240 H )

[mPa.s]

(21)

where H is the hematocrit expressed as a fraction [31]. Similar functional expressions have been determined from statistical analysis of large sample of normals [33, 22]. Because in any study population the hematocrits will vary, researchers have sought ways to account for this variability and its effects on the measured viscoelasticity. One approach is to shift the viscoelasticity data to those values expected if the hematocrit were a standard value such as 45%. This approach requires the determination of shifting functions based on correlations between hematocrit and viscoelasticity of a statistical sampling of normals. These shifting functions are valid only within the


83

normal range of hematocrits [34, 35]. Some have found that the relationship between hematocrit and viscoelasticity for normal blood differs from that of pathological blood [34] and choose to analyze data for pathological blood at the native hematocrit. An additional method is to create a reference set of viscoelasticity data for various hematocrits by the serial dilution of a normal blood sample with plasma and to compare these references with the measured values [36].


100

10

1 Visc. (65%H) Visc. (55%H) Visc. (45%H) Visc. (35%H) Elas. (65%H) Elas. (55%H) Elas. (45%H) Elas. (35%H)

0.1

0.01 0.1

1

10

100

1000

Shear Rate (1/sec) Figure 10. The viscoelasticity profile of blood with hematocrits of 35%, 45%, 55% and 65%. Measurements were made at 2 Hz and 22°C in a 0.1 cm ID cylindrical tube with length of 6 cm.

4.7. Thixotropy

Blood is thixotropic because the flow-induced microstructure requires time to respond to a change in flow. This has been observed by monitoring the effects of time on viscosity in steady flow [37] and on viscoelasticity in oscillatory flow [17]. Light transmission through blood during start-up of flow shows the time for flow-induced microstructure to reach equilibrium [11]. Light transmission following cessation of flow reveals the time for the microstructure to recover and is attributed to recovery of aggregates [10]. The time to reach equilibrium after flow initiation is much shorter than the time for the microstructure to recover after the cessation of flow. The time required for the microstructure to change increases with the magnitude of the stepped increase of flow. At higher shear rates, structural changes occur more rapidly than at lower shear rates. When producing a viscoelasticity profile with increasing steps in shear rate, the flow condition must be maintained long enough to achieve equilibrium before acquiring a data point. For viscoelasticity profiles as presented herein, the blood is subjected to 5 seconds of shear for each data point before data acquisition.

84


4.8. Intrinsic Viscosity

The intrinsic viscosity of blood is a number descriptive of the contribution to the viscosity of a single cell in plasma without interaction with other cells. Such values are determined from measurements of the viscosity while sequentially diluting the cell concentration and extrapolating the viscosity to zero concentration. The intrinsic viscosity is defined as

>Kc@

ª Kc K o º Lim « » c v o0 ¬ K o c v ¼

(22)

where K’ is the viscosity of the solution (blood), Ko is the solvent (plasma) viscosity and cv is the volume concentration of cells. The intrinsic elasticity is outside the range of available measurement techniques [38]. However, from Einstein's viscosity relation for a suspension of rigid spheres, the intrinsic viscosity is 2.5 [39] and for other shapes it is greater than 2.5 [40]. Taylor has shown that the intrinsic viscosity of deformable droplets is below 2.5 [41]. The values for red blood cells in native plasma are below 2 as would be true for deformable objects, while hardened cells are above 3 [38]. Therefore, intrinsic viscosity is an indicator of cell deformability. 4.9. Small Tubes and Boundary Layers

Cell concentration must diminish at the walls of a confining tube because of a boundary layer created by geometric exclusion of cells. It has been shown for cylindrical brass tubes that the thickness of this layer is approximately four cell diameters. With larger tubes of diameters > 1 mm (i.e., > 125 cell diameters), this boundary layer has minimal effect on the measured viscoelasticity [15]. In contrast, for small tubes, the effects become enhanced showing a decrease in viscosity with tube diameter [26]. This is similar to the Fahraeus-Lindqvist effect in steady flow [14]. 4.10. Other Topics Regarding Blood Viscoelasticity

The effects of geometry and plasma environment on the viscoelasticity of blood have been investigated. The influence of flow through tortuous paths such as in porous media provides information about how the viscoelasticity is affected by geometric confinement, which is similar to the tortuosity of arterioles [26]. Additives routinely used in clinical situations such as dextran and ringers also influence blood viscoelasticity [31]. The long-term storage [23] and freezing of blood [42] will alter the viscoelasticity profile as well.

5. Oscillatory Tube Flow with a Viscoelastic Fluid

There are several parameters useful for describing oscillatory flow in cylindrical tubes having the nominal size of small arteries. These parameters start with the pressure gradient in the tube (pressure drop per unit length) and volume flow through it. From these the impedance components (resistance and reactance) can be determined. In the case where the filling fluid is homogeneous and viscoelastic, hydrodynamic theory can

85


be used to relate the flow to the shear stress, shear strain and shear rate at the tube wall. These last parameters then relate to the viscoelasticity of the fluid. This sequence of relations works as well in reverse for calculating pressure-to-flow values. In the case of pulsatile flow, the Fourier components that describe the pulse are additive and can be treated separately by assuming linear relationships. 1000 P' |P"|

P' and |P"| (Pa)

100

10

1 Negative Values

0.1 1.00E-10

1.00E-09

Positive Values

1.00E-08

1.00E-07

3

Volume Flow (m /s) Figure 11. The pressure components of blood (44% H) in a 0.1 cm ID cylindrical tube with length of 6 cm measured at 2 Hz and 22 °C.

To illustrate these relations, consider oscillatory flow in a cylindrical tube with radius of 0.05 cm and length of 6 cm. The tube is filled with normal human blood with 44% hematocrit at room temperature (22 °C). The frequency of oscillation is 2 Hertz. Figure 11 shows the measured pressure-to-flow relation where P' is the component in phase with the volume flow (U) and P" is the component in quadrature with the flow. The negative values are phased 90° behind the flow and positive values are 90° ahead of the flow. For low flow rates the negative values of P” show that the elastic effects of the fluid dominate while at high flow rates the positive values of P” show that the inertial effects due to blood density dominate. Near a volume flow of 3.0E-9 m3/s the elastic effects are equal in size to the inertial effects and pressure is in phase with the flow [43, 44]. This condition is called viscoelastic resonance and occurs when 6 sin I

Y

where I

(23)

cc arctan §¨ K c ·¸ and Y is a dimensionless parameter © K¹

Y

a

UZ K*

(24)

where a is the radius of the tube, U is the density of blood and Z is the radian frequency. The impedance to flow in the tube (Z) is obtained from the ratio of the pressure to flow

86


Pc Pcc i U U

Z

R iX

(25)

where R is the resistance and X is the reactance. These values are plotted in Figure 12. Oscillatory tube flow theory for a viscoelastic fluid serves to relate the impedance to the viscoelastic properties of the fluid and shearing conditions. The density of the fluid is also required. The dimensionless parameter Y serves to separate the fluid motion into two types. If Y<1 then the velocity profile is parabolic (as true for Poiseuille steady flow). At Y values approximately above 5, the velocity profile begins to show wave motion emanating from the tube wall and damping as the wave moves toward the center of the tube [43]. As Y goes to approximately 20, the shear waves are greatly attenuated very close to the tube wall and the fluid movement in the center of the tube is like a piston. The case of Y<1 covers moderate sized and smaller arteries and all of the following applies to this small Y condition.

Resistance (R)

3

Resistance and |Reactance| (Pa s/m )

1.E+11 |Reactance| (|X|)

1.E+10

1.E+09

1.E+08 Negative Values

1.E+07 1.E-10

1.E-09

Positive Values

1.E-08

1.E-07

3

Volume Flow (m /s) Figure 12. The resistance and reactance of blood (44% H) in a 0.1 cm i. d. cylindrical tube with length of 6 cm measured at 2 Hz and 22 °C.

The resistance and reactance are related to the viscosity Kc and elasticity Kcc by R X

8KcL

(26)

Sa 4 4UZL 3Sa

2

8KccL Sa 4

(27)

where a is the radius of the cylindrical tube and L is the length of the tube. From these the viscoelastic coefficients can be calculated for a known impedance or the impedance can be calculated for known viscoelasticity. The data in Figure 3 show the viscosity

87


and elasticity values used to derive the impedance values in Figure 12. These data are plotted versus rms value of the shear rate at the tube wall. The magnitude of the shear rate at the wall is obtained from J

§ 4U · UZa 2 ¨¨ 3 ¸¸ 1 i 24K* © Sa ¹

(28)

The calculations for viscous and elastic stress are given by Equation (8) and the magnitude of the shear strain is J J . The stress-strain relationships of the data Z presented in Figure 12 are matched with those shown in Figure 4. Knowing the viscoelastic properties and flow conditions, the Maxwell relaxation time can also be calculated with Equation (15). The Maxwell relaxation times for the data presented in Figure 12 are matched with those shown in Figure 5.

6. Clinical Applications

Correlations between pathological conditions and blood viscoelasticity have been documented. Values of viscosity and elasticity have been correlated in several conditions: stroke [35, 45], cardiovascular disease [46, 35], peripheral vascular disease [35], arteriosclerosis [47, 48], infarction [47], apoplexy [49], diabetes [50], hyperthyroidism [51] and sickle cell anemia [52]. In addition to correlations with pathology, viscoelasticity data has been used to monitor the therapeutic aspects of erythrocytapheresis for sickle cell anemia [36], pentoxifylline (Trental£) for heart disease [46, 53], procaine hydrochloride [54], leeches [55], ginko biloba for diabetes [56] and for stroke [57]. The effects of exercise [58], hydration [59], age and oral contraceptives [60] on blood viscoelasticity have also been studied. Pumps are used to circulate blood during cardiopulmonary bypass and the effects of these devices on blood viscoelasticity have been investigated. Blood viscoelasticity has been found to be sensitive to sublethal damage caused by pumps [61]. The changes in blood viscoelasticity have been used to analyze effects of pump types (pulsatile versus non-pulsatile) and the use of hypothermia during bypass [62, 63]. Characterization of the viscoelasticity of blood from clinical model animals [64] also has relevance to blood pumps [61].

References [1] [2] [3] [4] [5]

G.B. Thurston, Effects of viscoelasticity of blood on wave propagation in the circulation, J. Biomech. 9 (1976), 13-20. G.B. Thurston, The viscoelasticity of blood and plasma during coagulation in circular tubes, Proceedings of the Sixth Conference of European Society for Microcirculation, Aalborg. S., Karger Basel, Aalborg, Denmark 1971, pp 12-15. G.B. Thurston, Viscoelasticity of human blood, Biophys. J. 12 (1972),1205-1217. Lessner, A., Zahavi, F., Silberberg, A., Frei, E. H. and Dreyfus, F., The viscoelastic properties of whole blood, in: Theoretical and Clinical Hemorheology, H. Hartert and A. L. Copley, Eds., Springer-Verlag, Berlin, 1971, pp. 194-205. A.L. Copley, R.G. King, S. Chien, S. Usami, R. Skalak and C.R. Huang, Microscopic observations of viscoelasticity of human blood in steady and oscillatory flow, Biorheology 12 (1975), 257-263.

88 [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36]

G.B. Thurston and N.M. Henderson / Viscoelasticity of Human Blood S. Chien, S., R.J. King, R. Skalak, S. Usami and A.L. Copley, Viscoelastic properties of human blood and red cell suspensions, Biorheology 12 (1975), 341-346. M. Lucius and J.F. Stoltz, Importance de l’agregation erythrocytaire sure les proprietes viscoelastiques et thixotropes du sang, In: Hemorheologie et Aggregation Erythrocytaire, J. F. Stolz, Ed., Editions Medicales Internationales, Paris, 1986, pp. 47-56. J.F. Stoltz and M. Lucius, Viscoelasticity and thixotropy of human blood, Biorheology 18 (1981), 453473. D.E. McMillan, J.S. Strigberger and N.G. Utterback, Rapidly recovered transient flow resistance: A newly discovered property of blood, Am. J. Phys. 256 (1987), H919-H926. H. Schmid-Schonbein, K.A. Kline, L. Heinich, E. Volger and T. Fischer, Microrheology and light transmission of blood. III. The velocity of of red cell aggregate formation, Pflugers Arch. 354 (1975), 299-317. A. Gaspar-Rosas and G.B. Thurston, Erythrocyte aggregate rheology by transmitted and reflected light, Biorheology 25 (1988), 71-487. G.B. Thurston, Light transmission through blood in oscillatory flow, Biorheology 27 (1990), 685-700. H. Schmid-Schonbein, P. Gaehtgens and H. Hirsch, On the shear rate dependence of red cell aggregation in vitro, J. Clin. Invest. 47 (1968), 1447-1453. R. Fåhraeus and T. Lindqvist, The viscosity of blood in narrow capillary tubes, Am. J. Physiol. 99 (1931), 563-568. G.B. Thurston, The viscosity and viscoelasticity of blood in small diameter tubes, Microvas. Res. 11 (1976), 133-146. G.B. Thurston, Shear rate dependence of the viscoelasticity of polymer solutions. I. Theoretical model, J. Non-Newtonian Fluid Mech. 9 (1981), 57-68. G.B. Thurston, Rheological parameters for viscosity, viscoelasticity and thixotropy of blood, Biorheology 16 (1979), 149-162. G.B. Thurston, Plasma release-cell layering theory for blood flow, Biorheology 26 (1989), 199-214. G.B. Thurston, Elastic effects in pulsatile blood flow, Microvasc. Res. 9 (1975), 145-157. G.B. Thurston, Non-Newtonian viscosity of human blood: Flow-induced changes in microstructure, Biorheology 31 (1994), 179-192. U. Kasser, P. Heimburg and E. Walitza, Viscoelasticity of whole blood and its dependence on laboratory parameters, Clin. Hemorheol. Micro. 9 (1989), 307-312. G.B. Thurston, Effects of hematocrit on blood viscoelasticity and in establishing normal values, Biorheology 15 (1978), 239-249. G.B. Thurston, Erythrocyte rigidity as a factor in blood rheology: Viscoelastic dilatancy, J. Rheol. 23 (1979), 703-719. P.P. Klug and L. Lessin, Microvascular blood flow of sickled erythrocytes, Blood Cells 3 (1977), 263272. G.B. Nash and H.J. Meiselman, Alteration of red cell membrane viscoelasticity by heat treatment: effect on cell deformability and suspension viscosity, Biorheology 22 (1985), 73-84. G.B. Thurston and N.M. Henderson, Effect of flow geometry on viscoelasticity, Biorheolgy 43 (2006), 1-18. S. Chien, S. Usami, R.J. Dellenback and M.I. Gregersen, Shear-dependent interaction of plasma proteins with erythrocytes in blood rheology, Amer. J. Phys. 219 (1970), 143-153. D.E. Brooks, R.G. Greig and J. Janzen, Mechanisms of erythrocyte aggregation In: Erythrocyte Mechanics and Blood Flow, G. R. Cokelet, H. J. Meiselman and D. E. Brooks, Eds., Alan R. Liss, New York, 1980, pp. 141-148. G.B. Thurston, Blood viscoelasticity and relaxation processes: Influence of aggregation tendency, In: Hemo-Rheology and Diseases, J. F. Stoltz and P. Drouin, Eds., Doin Editeurs, Paris, 1980, pp. 51-66. V. Ribitsch, Determination of red blood cell suspensions rheological properties in oscillatory flow, Clin. Hemorheol. 9 (1989), 313-317. G.B. Thurston, Viscoelastic properties of blood and blood analogs, In: Advances in Hemodynamics and Hemorheology, T. Howe, Ed., JAI Press, New York, 1996, pp. 1-30. E. Braunwald, Ed., Harrison’s Principles of Internal Medicine, McGraw-Hill, New York (2001). U. Kasser and P. Heimburg, Quality control in blood viscoelastometry, Clin. Hemorheol. 8 (1988), 93103. E. Walitza, I. Anadere, H. Chmiel and S. Witte, Evaluation of viscoelasticity measurement of human blood, Biorheology 25 (1988), 209-217. H. Chmiel, I. Anadere and E. Walitza, The determination of blood viscoelasticity in clinical hemorheology, Clin. Hemorheol. 10 (1990), 363-374. G.B. Thurston, N. Henderson and M. Jeng, Effects of erythrocytapheresis transfusion on the viscoelasticity of sickle cell blood, Clin. Hemorheol. 30 (2004), 61-75.


89

[37] C.R. Huang, M. Siskovic, R.W. Robertson, W. Fabisiak, E.J. Smitherberg. and A.L. Copley, Quantitative characterization of thixotropy of whole human blood, Biorheology 12 (1975), 279-282. [38] R.B. More and G.B. Thurston, Intrinsic viscoelasticity of blood cell suspensions: effects of erythrocyte deformability, Biorheology 24 (1987), 297-309. [39] A. Einstein, A new determination of molecular dimensions, Ann. Physik 19 (1906), 289. [40] G.B. Jeffery, The motion of ellipsoidal particles immersed in a viscous fluid, Proc. Royal Soc. A102, 123 (1923). [41] Taylor, G., The viscosity of a fluid containing small drops of another fluid, Proc. Royal Soc. A138 (1932), 41. [42] T.P. Kidwell, M.S. Mandell, K.R. Diller and G.B. Thurston, Freeze induced alterations in blood rheology, In: 1975 Advances in Bioengineering: Presented at the ASME Winter Annual Meeting, Nov 30-Dec 5, 1975, American Society of Mechanical Engineers, New York, 1977, pp.58-60. [43] G.B. Thurston, Theory of oscillation of a viscoelastic fluid in a circular tube, J. Acoust. Soc. Amer. 32 (1960), 210-213. [44] G.B. Thurston, The effects of frequency of oscillatory flow on the impedance of rigid, blood-filled tubes, Biorheology 13 (1976), 191-199. [45] P. Koletringer, W. Langsteger, P. Lind, O. Eber and F. Reisecker, Morning increase in blood viscoelasticity of patients with ischemic stroke, Stroke 21 (1990), 826-827. [46] S. Witte and I. Anader, Modifications of viscoelastic properties during cardiovascular diseases, Clin. Hemorheol. 9 (1989), 831-837. [47] I. Anadere, H. Chmiel and P. Heimburg, Viscoelastic parameters of blood in patients with myocardial infarction compared to normal donors, Rheol. Acta 21 (1982), 611-613. [48] K.M. Hell, A. Balzereit, U. Diebold and H.D. Bruhn, Importance of blood viscoelasticity in arteriosclerosis, Angiology 40 (1989), 539-546. [49] H. Chmiel, I. Anadere, E. Walitza and S. Witte, The measurement of density and its significance in blood rheology, Biorheology 20 (1983), 685-696. [50] S. Ikemoto, T. Tanaka, J. Yamamoto, K. Kuchiba, M. Akiyama, T. Maeda, T. Yokose and Y. Isogai, Blood viscoelasticity, in clinical Medicine, In: Hemorheologie et Aggregation Erythrocytaire, J. F. Stolz, M. Donner and A. L. Copley, Eds., Editions Medicales Internationales: Chachan, 1991, pp. 189195. [51] P. Koletringer, W. Langsteger, P. Lind and O. Eber, Erythrocyte aggregation in hyperthyroid patients: Measurement under fluid conditions as elasticity of blood at low shear-rates, In: Hemorheologie et Aggregation Erythrocytaire, J. F. Stolz, M. Donner and A. L. Copley, Eds., Editions Medicales Internationales: Chachan, 1991, pp. 261-264. [52] W.J. Drasler, C.M. Smith II and K.H. Keller, Viscoelastic properties of oxygenated sickle erythrocyte membrane, Biorheology 26 (1989), 935-949. [53] W. Berman, N. Berman, D. Pathak and S.C. Wood, Effects of pentoxifylline (Trental) on blood flow, viscosity and oxygen transport in young adults and inoperable cyanotic congenital heart disease, Pediatr. Cardiol. 15 (1994), 66-70. [54] S.O. Sowemimo-Coker, G. Yardin and H.J. Meiselman, Effect of procaine hydrochloride on aggregation behavior and suspension viscoelasticity of human red blood cells, Biorheology 26 (1989), 951-972. [55] H. Chmiel, I. Anadere and K. Moser, Hemorheological changes under blood leaching, Clin. Hemorheol. 9 (1989), 569-576. [56] S.Y. Huang, C. Jeng, S.C. Kao, J.J. Yu and D.Z. Liu, Improved haemorrheological properties by Ginko biloba extract (Egb 761) in type 2 diabetes mellitus complicated with retinopathy, Clin. Nutr. 23 (2004), 615-621. [57] I. Anadere, H. Chmiel and S. Witte, Hemorheological findings in patients with completed stroke and the influence of ginkgo biloba extract, Clin. Hemorheol. 5 (1985), 411-420. [58] M.R. Fredde and S.C. Wood, Rheological characteristics of horse blood: significance during exercise, Resp. Physiol. 94 (1993), 323-335. [59] G.A. Vlastos, C.C. Tangney and R.S. Rosenson, Effects of hydration on blood rheology, Clin. Hemorheol. 28 (2003), 41-49. [60] H. Kolleger, W. Oder, K. Zeiler, C. Baumgartner, C. Lind, B. Oder, C. Sagmeister and L. Deecke, Viscoelasticity of whole blood as a function of age, gender, cigarette smoking and intake of oral contraceptives, Clin. Hemorheol. 10 (1990), 499-503. [61] P.J. Marascalco, S.P. Ritchie, A. Snyder and M.V. Kameneva, Development of standard tests to examine viscoelastic properties of blood of experimental animals for pediatric mechanical support device evaluation, ASAIO J. 52 (2006), 567-574.

90


[62] A. Undar, N.M. Henderson, G.B. Thurston, T. Masai, E.A. Beyer, O.H. Frazier and C.D. Fraser, The effects of pulsatile versus nonpulsatile perfusion on blood viscoelasticity before and after deep hypothermic circulatory arrest in neonatal piglet model, Artif. Organs 23 (1999), 717-721. [63] A. Undar, Effect of hypothermic cardiopulmonary bypass on blood viscoelasticity in pediatric cardiac patients, ASAIO J. 51 (2005), 522-524. [64] U. Windberger, V. Ribitsch, K.L. Resch and U. Losert, The viscoelasticity of blood and plasma in pig, horse, dog, ox and sheep, J. Exp. Anim. Sci. 36 (1994), 89-95.


91

Mechanical and Adhesive Properties of Healthy and Diseased Red Blood Cells Brian M COOKE a,1 and Chwee T LIM b Department of Microbiology, Monash University, Victoria, Australia b Department of Mechanical Engineering & Division of Bioengineering, National University of Singapore, Singapore a

Introduction For red blood cells (RBC) to survive in the harsh hemodynamic environment of the circulation in vivo, they must remain non-adhesive and maintain a set of unique mechanical properties, particularly remarkable deformability and extreme membrane stability. In healthy humans, RBC survive for about 120 days in the circulation, whereas in certain pathological conditions where membrane stability, cellular deformability or adhesiveness are compromised, the lifespan of the RBC can be dramatically reduced or its function severely compromised, often with severe clinical manifestations. Of a number of disorders affecting the mechanical and adhesive properties of human RBC, homozygous sickle cell disease and malaria are arguably the most important and certainly, in the case of malaria, the most studied. In this chapter, we review the structure-function relationships that determine the mechanical and adhesive properties of RBC and describe some techniques and methods, old and new, for quantifying these important rheological properties. In particular, we concentrate on RBC infected with malaria parasites as a specific example of how recent research on this human pathogen has not only advanced our knowledge of this important human disease and opened up new possible avenues for therapy, but has also increased our understanding of RBC structure-function relationships at the molecular level and the mechanisms that regulate and maintain their unique rheological properties.

1. Structure and Function of Red Blood Cells In simplistic terms, RBC are often considered to be no more than simple biological containers of hemoglobin that effectively transport and exchange oxygen and carbon dioxide throughout the tissues and organs of the body. In fact, they are probably the best understood of all eukaryotic cells, particularly in the composition and physical nature of the membrane skeleton and its relationship to the overall rheological properties of the cell [1-6]. Each of the approximately five million red cells present in each microliter of human blood must traverse the microcirculation more than 1000 times each day during their average lifetime of 120 days. During each passage, the 1

Corresponding Author: Molecular and Cellular Rheology Laboratory, Department of Microbiology, Monash University, Victoria 3800, Australia; E mail: [email protected]

92

B.M. Cooke and C.T. Lim / Mechanical and Adhesive Properties of Healthy and Diseased RBC

RBC, with an average diameter of about eight Pm, must survive exposure to harsh extrinsic shear forces and pass through capillaries and venules of the microcirculation or fenestrations in the spleen with diameters or openings as small as one-quarter of the diameter of the RBC. This ability is possible because RBC are highly deformable, nonadhesive structures that can undergo rapid and reversible shape changes repeatedly when exposed to hemodynamic shear. Normally, RBC can deform to form ellipses and align linearly in large blood vessels under arterial shear flow [7]. Further, they can bend and fold producing ‘slipper’ and other forms [8] which enable the cell to pass through biological apertures as small as 3 μm in diameter. At the cellular level, the general principles of cell deformability and classical methods for its measurement have been well reviewed in the past [9-12]. Essentially, there are three major factors that determine the deformability of normal RBC; (i) the relatively low viscosity of the cytoplasm, (ii) a high surface area to cell volume ratio, and (iii) a viscoelastic membrane skeleton that is able to freely rotate around the cytoplasm (‘tank-treading’) to reduce hydraulic resistance and facilitate transluminal passage [8, 13, 14]. Since for normal, healthy adult RBC, the cytoplasm behaves as an ideal Newtonian fluid with relatively low viscosity, the majority of their unique rheological properties are attributed to the mechanical properties of the RBC membrane skeleton (Table 1). At the molecular level, possibly the best understood of these factors is the composition and organization of the membrane skeleton and the mechanisms by which it combines structural resilience with extreme pliability in health. When the structure of the membrane skeleton becomes perturbed, this alteration is frequently reflected in changes to the material properties of the whole cell which, in turn, can have dramatic effects on RBC function.

2. Molecular Determinants of the Material Properties of Red Blood Cells The normal RBC is composed of a plasma membrane overlying a well-defined twodimensional network of structural proteins termed the membrane skeleton. Its contents are essentially a solution of hemoglobin in water. 2.1. The Plasma Membrane The composition of the plasma membrane is typical of most mammalian cell membranes comprising an asymmetric bilayer of amphiphillic phospholipids with glycolipids and cholesterol intercalated in between. Predominantly uncharged species such as phosphatidylcholine and sphingomyelin predominate in the outer leaflet while anionic species such as phosphatidylserine and phosphatidylethanolamine prevail in the inner leaflet. Under normal physiological conditions, this asymmetry is maintained by the action of ATP-dependent translocases (i.e., flippase and floppase) and inactivation of the ATP-independent phospolipid scramblases (e.g., scramblase 1), which oppose the action of the translocases. Changes in the equilibrium distribution of phospholipids in the RBC membrane do occur under conditions where the translocase-scramblase balance appears to become perturbed. This is most notably demonstrated in sickle, thalassemic and diabetic RBC which abnormally expose phosphatidylserine in their outer lipid leaflet. The surface area of the membrane remains relatively constant, in excess of that needed for a sphere of equal volume, throughout the in vivo lifetime of the RBC; lipid can be lost in the form of microvesicles that bud off the RBC during


93

prolonged storage, ATP depletion or in certain genetic conditions that result in uncoupling of the membrane from the underlying skeleton (see Section 3). Table 1. Rheological properties of normal adult RBC Parameter

Typical Value

Mean cell volume

90 fl

Membrane surface area

~140 μm2

There is about 40% excess surface area over that required to enclose the cell volume in a sphere. Allows changes in shape without need to expand the membrane

Membrane surface viscosity

0.7 μPa.s.m

This value is much greater than the surface viscosity of a simple lipid bilayer. Approximate conversion to a bulk viscosity can be made by dividing by the membrane thickness; the value is then also much greater than the cytoplasmic viscosity

Cytoplasmic viscosity

6.4 mPa.s

Highly dependent on mean cell hemoglobin concentration

Membrane shear elastic modulus

6.0 μN/m

The gradient of a stress/strain curve when the cell undergoes linear stretching; property attributed to the membrane skeleton; confers the remarkable elastic behaviour on RBC

Bending elastic modulus

1.6 x10-19 N.m

Determines e.g., resistance to folding in absence of elongation

Elastic modulus for area compressibility*

300-600 mN/m

High value reflects great resistance of membrane to area dilation; only approx. 2% dilation can occur before RBC rupture

Time for shape recovery

0.12 s

Determined by ratio of membrane surface viscosity to shear elastic modulus

0.31 s

Determined by ratio of cytoplasmic viscosity to bending elastic modulus

(after extensional deformation) Time for shape recovery (after bending deformation)

Notes

* The relative magnitudes of the elastic moduli mean that deformation of RBC in the circulation typically occurs at constant membrane area (no dilation), by a combination of folding (bending) and elongation (shear), with the shear elastic modulus making the greater contribution to resistance to deformation.

2.2. Membrane Proteins The lipid bilayer of RBC is punctuated by a number of proteins. While some of these remain peripheral and do not penetrate into the core of the bilayer, others termed integral membrane proteins traverse the bilayer and interact directly with proteins of the underlying membrane skeleton. These integral membrane proteins, predominantly band 3 and four of the five glycophorins (A-D), anchor the plasma membrane to the membrane skeleton and thus play a major role in determining and maintaining the structural integrity of the RBC. Integral membrane proteins, particularly glycophorin A, are also highly decorated with an abundance of sialic acid residues that confer a layer of high negative charge on the surface of the RBC and tend to counter adhesion of the cells to other circulating cells and to the vascular endothelium. Alterations to surface

94


charge topography do appear to have deleterious effects on the circulatory properties of RBC. In sickle cells, for example, clustering of negative charges on the RBC surface has been proposed as one of a number of mechanisms which mediate their adhesion to vascular endothelial cells [15]. 2.3. The Membrane Skeleton The network of proteins that constitutes the membrane skeleton of the RBC forms the structural support over which the plasma membrane is draped. Twelve major membrane skeleton proteins can be identified by their molecular mass after separation by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and staining with Coomassie Brilliant Blue, including spectrin, actin, ankyrin, proteins 4.1, 4.2, and 4.9, p55, and adducin [16]. A highly ordered arrangement of tetramers of D and E spectrin, their interconnection at the ternary complex, with actin and protein 4.1, and the bonds to the overlying RBC membrane via band 3 and glycophorin C (GPC) provide the structural integrity of the RBC and underpin its ability to repeatedly deform in the microcirculation [5, 17]. The stability of the membrane skeleton is influenced not only by the primary sequence of its component proteins but also by the levels of protein phosphorylation [18, 19]. In addition to these interactions, spectrin and protein 4.1 engage in a number of other protein-protein interactions that affect cellular mechanical properties. Spectrin binds to adducin and ankyrin as well as a number of kinases [20]. Protein 4.1 binds to membrane proteins such as band 3, GPC, and p55, a palmitoylated membrane protein, and cytoplasmic calcium regulating proteins such as calmodulin. These interactions in turn influence both the elasticity and mechanical stability of the erythrocyte membrane [21]. More recently, development of gene knockout mice that do not express particular RBC membrane skeleton proteins has added considerable tools for more precise and sophisticated structure-function studies [22]. Furthermore, the recent arrival of more sophisticated mass spectrometry (MS) based protein identification techniques, such as quadrupole time of flight and Fourier transform MS (QTOF-MS), has identified more than 300 proteins at the RBC membrane skeleton [23]. This clearly increases the complexity of this important biological interface and sets a formidable challenge to determine the precise function of these proteins in RBC. 2.4. The RBC Cytoplasm In the circulation, mature RBC are devoid of all internal organelles and contain only a solution of hemoglobin at a concentration of approximately 350 g/L. As RBC age in vivo, mean cell hemoglobin concentration increases due to a reduction in cell volume and its viscosity increases sharply. This reduces whole cell deformability and, in particular, increases the time constant for shape recovery. Additionally, concentrationdependent binding of hemoglobin to the RBC membrane skeleton further decreases the rate at which the cell can deform [24]. In certain pathological conditions such as sickle cell anemia or hemoglobin C disease, the presence of abnormal hemoglobin and increased mean cell hemoglobin concentration can have a significant effect on reducing RBC deformability, often with severe clinical consequences [25]. The presence of abnormal cytoplasmic inclusions, typically Heinz bodies [26] or malaria parasites, also drastically reduces both membrane and whole cell deformability.


95

3. Pathophysiology and Molecular Basis of Genetic Disorders Associated with Altered Mechanical and Adhesive Properties of Red Blood Cells Albeit clearly incomplete, our understanding of the relationship of the protein network of the membrane skeleton to the rheological properties of the intact RBC has been advanced by many years of study of pathological states such as inherited disorders of RBC including the hemoglobinopathies, the thalassemias and hereditary sphero- and ovalocytosis [3-6, 27-30]. 3.1. Hereditary Spherocytosis Hereditary spherocytosis (HS) is the most common of all inherited disorders of RBC membranes. The disease manifests as a hemolytic anemia of varying severity due to the premature destruction of poorly deformable, osmotically fragile, spherical shaped RBC. The principal cellular abnormality is associated with well defined genetic defects in a number of major RBC proteins including spectrin, band 3, protein 4.2 or, most commonly, ankyrin that result in uncoupling of the RBC membrane from the underlying membrane skeleton [31]. Budding of lipid membrane micro-vesicles leads to loss of membrane surface area resulting in a spherical, doughnut shaped RBC with decreased deformability and increased osmotic fragility. 3.2. Hereditary Elliptocytosis Hereditary elliptocytosis (HE), characterized by the presence of elliptical or ovalshaped RBC in the circulation, results from genetic defects in membrane skeletal proteins that, in contrast to HS, maintain the structural integrity of the RBC skeleton, predominantly Į/ȕ spectrin and ȕ-spectrin/protein 4.1 interactions. Defects in GPC can also manifest as HE but this is most likely secondary to a corresponding deficiency in protein 4.1. Disruption to the structural integrity and mechanical stability of elliptocytic RBC leads to their fragmentation under hemodynamic stress and consequential hemolytic anemia of varying severity [4]. 3.3. South East Asian (Melanesian) Ovalocytosis In contrast to HE RBC described above, individuals with South East Asian Ovalocytosis (SAO) have quite unique elliptical/ovalocytic RBC that are hyper-stable, rather than unstable [27]. SAO RBC are very rigid and show decreased osmotic fragility, increased thermal stability and increased resistance to shape change. The molecular defect is defined by a short 27 base pair deletion in the gene encoding the most abundant of the integral RBC membrane proteins, band 3 [32]. The deletion results in major structural alterations of the protein at the boundary of the cytoplasmic and membrane domains, resulting in dramatic effects on its function including increased binding to ankyrin, increased level of phosphorylation, reduced lateral and rotational mobility and reduced ability to transport anions [33, 34]. Despite the fact that the shear elastic modulus of SAO RBC is at least three times that of normal RBC, the consequent remarkable decrease in dynamic deformability does not seem to reduce their survival in vivo or prevent them from circulating: only mild or unapparent intravascular hemolysis or anemia is seen in these individuals. Interestingly, in the absence of any adhesive abnormality for SAO RBC described to date, this finding

96


suggests that an increase in membrane rigidity per se is not a major rheological factor that determines the survival time of RBC in the circulation. 3.4. Hemoglobinopathies In these inherited conditions, pathology results directly from changes in either hemoglobin structure (i.e., structural hemoglobinopathies) or concentration (e.g., thalassemias). Collectively, these conditions are too numerous and diverse to discuss extensively in this current chapter; the pathobiology of these conditions has been detailed previously [35, 36]. Of those that act through changes both in the mechanical and adhesive properties of RBC, the best example is undoubtedly homozygous sickle cell anemia (SCA). SCA is a devastating disease with protean clinical manifestations, most of which result from physical trapping of grossly mechanically-impaired sickled RBC in the microvasculature leading to vaso-occlusive crises which are frequently life threatening. In sickle cells, the normal adult hemoglobin (HbA) is replaced by abnormal sickle hemoglobin (HbS), which can be inherited in either a heterozygous or homozygous state. In heterozygotes termed sickle-cell trait, HbAS RBC contain approximately equal proportions of HbA and HbS. Except under extreme conditions of low oxygen tension or oxidative stress, HbAS individuals remain largely clinically asymptomatic. In contrast, homozygotes inherit two copies of the HbS gene (i.e., HbSS) and bear the full brunt of this condition. Unlike HbA, HbS polymerises into long rigid rods upon deoxygenation, leading to profound changes in cell morphology (i.e., the classical ‘sickle’ shape). During repeated cycles of deoxygenation and reoxygenation, sickle cells become progressively dehydrated and finally become irreversibly sickled. Dehydration increases intracellular hemoglobin concentration which in turn dramatically increases the internal viscosity of the RBC, thereby further exacerbating a general loss of cell deformability; increased intracellular hemoglobin also markedly shortens the delay time prior to polymerisation. The overall reduction in static and dynamic cell deformability is also due in part to a marked increase in the shear elastic modulus of the RBC membrane itself [37]. Although modulus increase is also likely mediated through a bulk effect of increased hemoglobin concentration, which is perhaps inordinately increased adjacent to the membrane, the precise molecular mechanism for this still remains unclear [38]. In addition to reduced deformability, HbSS-containing RBC are also abnormally adhesive. The mechanism by which the presence of sickle hemoglobin induces this adhesive propensity remains incompletely understood. However, sickle RBC can bind, at least in vitro, to endothelial cell expressed receptors such as DvE3 integrin, Gp1b, CD36, VCAM-1 and P-selectin [15, 39, 40]. Ligands on the surface of the RBC which may mediate these interactions have also been identified, including CD36, D4E1 integrin, sulphated glycolipids, CD44 or aggregated surface charge [15]. Some of these interactions also seem to require plasma factors such as von Willebrand Factor, fibronectin or thrombospondin as bridging molecules. Additionally, membrane mechanical abnormalities, resulting in the exposure of an otherwise cryptic epitope on the band 3 protein which can bind to endothelial cells, have been proposed [41]. Although abnormally adhesive, the degree of alteration appears to occur to a much lower extent than in RBC parasitized by P. falciparum malaria parasites. Comparison of adhesion of normal (HbAA), sickle (HbSS) and P. falciparum-infected RBC to vascular endothelial cells under flow conditions showed that the relative levels of adhesion was in the ratio of 1:3:1000 [42]. Thus, it seems likely that the direct physical


97

mechanical trapping of sickle cells in the small diameter vessels of the microcirculation, consequent upon their abnormal mechanical properties, is the primary event in the pathogenesis of sickle cell anemia with altered cellular adhesive properties playing, if at all, an accessory role.

4. Malaria – A Model RBC Rheological Disease Malaria is caused by hemoprotozoan parasites of the genus Plasmodium. Despite intense research and public health efforts, malaria remains the most serious and widespread parasitic disease of humans. About 500 million people become infected with malaria every year and up to 3 million die as a result of the infection [43]. Four species of Plasmodium infect humans but P. falciparum and P. vivax together cause the vast majority of the disease worldwide. Infections caused by P. vivax contribute most to the debilitating morbidity associated with malaria infection [44], predominantly through acute, recurrent episodes of fever and a chronic anemia. Almost all malaria related deaths, however, are due to P. falciparum because of its apparently unique ability to cause severe clinical syndromes such as cerebral malaria and multi-organ failure that, even if treated early, are frequently fatal. The enhanced pathogenicity of P. falciparum is related to two major parasiteinduced alterations to RBC that are of interest to rheologists: loss of RBC deformability and increased adhesiveness of infected RBC for multiple cellular targets including other RBC, vascular endothelial cells, platelets, dendritic cells and syncytiotrophoblasts of the placenta (see [45, 46] for recent reviews). Ironically, these alterations are not only the key to the unique pathogenicity of P. falciparum but are also essential for the survival of the parasite in the host circulation. In the blood, P. falciparum parasites undergo repeating cycles of asexual replication inside RBC, each cycle taking 48 hours to complete. Over the first half of the cycle, young ‘ring forms’ of the parasite develop into mature, pigmented ‘trophozoites’. During the second half of the cycle, trophozoites develop further into mature schizonts that eventually lyse the RBC releasing 8-32 daughter ‘merozoites’ into the circulation which rapidly re-invade new RBC. RBC containing ring stage parasites appear to show little rheological impairment and continue to circulate. In fact, detection of ring-infected RBC in peripheral blood smears still remains the gold standard test for the diagnosis of malaria in infected individuals. In contrast, RBC containing trophozoites or schizonts are notably absent from the peripheral circulation because these mature forms accumulate in the microvasculature of most if not all organs of the body [47]. These sequestered parasitized RBC (PRBC) can perturb or completely obstruct blood flow in small diameter vessels of the microcirculation and are the most likely cause of the severe vaso-occlusive syndromes, such as cerebral malaria, that accompany the disease [48]. Acquisition of such abnormal circulatory behavior for RBC is paralleled by the production of a number of stage-specific parasite-encoded proteins that associate with the RBC cytoskeleton, either by direct interaction with cytoskeletal proteins or by transient or permanent insertion into the RBC membrane. Production of these proteins is associated with morphological alterations to the RBC including the appearance of electron-dense ‘knob-like’ structures that protrude from the RBC membrane (Figure 1).

98


Figure 1. Morphological and molecular alteration of the RBC membrane induced by P. falciparum malaria parasites. A) Schematic representation of the interaction of exported malaria proteins with proteins of the RBC membrane skeleton. An area of the RBC membrane forming a single knob structure is represented; B) Transmission electron micrograph of a RBC infected with a P. falciparum malaria parasite (P). Electron dense knobs at the membrane skeleton are indicated by black arrow heads. Appearance of the membrane skeleton in normal uninfected RBC is shown for comparison (RBC); C) The surface of a RBC infected with a P. falciparum parasite (P) imaged by Atomic Force Microscopy (AFM) showing the presence of knobs on the RBC membrane; D) A portion of the RBC membrane of the RBC in panel C; E) High magnification of a single knob structure on the RBC membrane imaged by AFM. Image in panel B kindly provided by Professor David Ferguson, University of Oxford, UK. Abbreviations for Figure 1A: K, knob-associated histidine rich protein; PfEMP1, P. falciparum erythrocyte membrane protein 1; EMP3, P. falciparum erythrocyte membrane protein 3; 332, P. falciparum 332 protein; FEST, falciparum exported serine/threonine kinase; MESA, mature parasite-infected erythrocyte surface antigen; A, ankyrin; Gp, glycophorin; 4.1, protein 4.1.


99

The majority of these exported proteins are synthesized during the later stages of the parasite’s life cycle, co-incident with the time of maximum alteration in cellular mechanical and adhesive properties. A comprehensive description of this intriguing set of exported parasite proteins is beyond the scope of this chapter and has been reviewed extensively elsewhere [49-51]; some examples are detailed in Table 2. At least 10 proteins that are exported by the parasite to the RBC membrane skeleton have been identified and characterized to a greater or lesser extent. At least two of these, KAHRP and PfEMP3, bind directly to spectrin and reduce the flexibility of the cytoskeletal network [52]. More recently, an informed bioinformatic analysis of the complete and annotated genome of P. falciparum, which became available in 2002 [53], has revealed that almost 400 proteins are predicted to be exported into the RBC; at least 70 are likely to interact with the membrane skeleton and play a role in RBC remodeling [54]. Table 2. Exported malaria proteins and their interactions with the RBC membrane skeleton. Protein

Interactions with the RBC membrane skeleton and functional consequences

References

RESA

155 kDa phosphoprotein; present in RBC infected with young ring stage parasites; binds to spectrin; little or no measurable effect on membrane rigidity; increases the thermal stability of infected RBC and may protect RBC against fragmentation during episodes of malarial fever.

[55-59]

MESA

250300 kDa phosphoprotein; binds to the 30 kDa domain of RBC protein 4.1; competes with RBC p55 protein for binding to protein 4.1 and is likely to modulate p55 function; precise function remains unknown.

[60-64]

KAHRP

80109 kDa histidine-rich protein; essential for the formation of RBC membrane knobs and adhesion of infected RBC to the vascular endothelium under flow; may be involved in trafficking PfEMP1 to the surface of infected RBC; binds to Į-spectrin, actin and ankyrin; interacts with the cytoplasmic tail of PfEMP1 and likely anchors PfEMP1 into the membrane of parasite-infected RBC; contributes approximately 50% of the total increase in membrane rigidity in infected RBC.

[52, 65-72]

PfEMP3

315 kDa protein located at the membrane skeleton of infected RBC; possibly concentrated at knobs; not essential for cytoadherence of infected RBC, most likely binds to spectrin, contributes approximately 15% of the total increase in membrane rigidity in infected RBC; precise function remains unknown.

[73-75]

Pf332

Giant protein (c. 750 kDa); associated with the membrane skeleton in RBC infected with mature parasites; may be exposed on the infected RBC surface; precise function remains unknown.

[76-78]

FEST

210 kDa serine-threonine kinase associated with the membrane skeleton; possibly responsible for phosphorylation of MESA, RESA or other parasite or RBC proteins.

[79]

PfEMP1

265285 kDa, highly antigenically variable protein, spans the RBC membrane and binds to the RBC skeleton via a conserved cytoplasmic domain that binds to spectrin, actin and KAHRP; extracellular domain exposed on the RBC surface mediates adherence of infected RBC to vascular endothelial cells, other RBC (normal and malaria-infected), platelets and placental syncytiotrophoblasts; clusters at knobs; different variants of PfEMP1 bind to different receptors and mediate quantitatively and qualitatively different forms of adhesion.

[80-84]

The knob structures at the membrane skeleton of parasitized RBC (PRBC) are intriguing. They are essential for adhesion of PRBC to vascular endothelium [69] and for process of sequestration that causes many of the frequently fatal syndromes associated with P. falciparum infections. The major structural element of the knobs appears to be the

100


knob-associated histidine-rich protein (KAHRP). Also incorporated into this structure is one of a large family of highly polymorphic parasite exported proteins known P. falciparum erythrocyte membrane protein-1 (PfEMP1). PfEMP1 is anchored in the knob complex via interaction of its cytoplasmic domain with KAHRP and spectrin while its external domain acts as a ligand for binding to a number of endothelial cellexpressed receptors including CD36, ICAM-1 and chondroitin sulphate A [49, 82]. In addition to binding to the vascular endothelium, PfEMP1 also mediates adhesion of PRBC to other PRBC, a phenomenon known as autoagglutination, and to normal RBC, termed rosetting [85]. The importance of these cellular adhesive phenomena in vivo has been intensively studied. Although a good deal of controversy exists, on balance it appears that rosetting and agglutinating parasites are generally responsible for more severe disease. A number of counter receptors on the surface of RBC have been described to which PfEMP1 can bind including complement receptor 1 (CR1), heparin sulphate or heparin sulphate-like glycosaminoglycans, and the ABO blood group antigens, particularly blood group A. The physical forces binding RBC into a rosette have been measured using both dual micropipetting techniques and viscometry and are estimated to be at least five times higher than those involved in cytoadherence to endothelial cells [50]. In addition to PfEMP1, band 3 has also long been implicated in the altered adhesive properties of PRBC. Intramembranous particles, primarily due to tetramers of band 3, are specifically redistributed in the region of the knob. Monoclonal antibodies that recognize band 3 only in PRBC, termed Pfalhesin, have also been prepared and these are capable of blocking cytoadherence of PRBC to CD36, as are synthetic peptides based on the sequence of the altered band 3 [86]. It is not known precisely what processes lead to modification of band 3 but they are probably related to structural changes in the RBC secondary to malaria infection: antibodies that react with altered band 3 on the surface of PRBC also react with the surface of sickle RBC and reduce their adhesiveness to cultured endothelial cells [41]. Exported parasite proteins also have major effects on the alteration of the mechanical properties of RBC. When compared to normal RBC, the shear elastic modulus of RBC infected with P. falciparum parasites is profoundly increased [87]. Of the exported malaria proteins characterized so far, the effect of KAHRP appears to be the most profound, contributing to approximately 50% of the increased membrane rigidification seen in PRBC [52]. Second to this is PfEMP3, which appears to contribute an additional 15%. The proteins or mechanisms which contribute the remainder of the rigidification remain to be identified. However, since some proteins exported by parasites appear to be putative kinases (Table 2), alteration of the phosphorylation status of normal RBC integral membrane proteins may also play a role.

5. Established & Emerging Techniques for Determining the Mechanical and Adhesive Properties of RBC Numerous techniques have been developed to quantify the mechanical and adhesive properties of blood cells. Each have their own set of particular advantages and disadvantages, and frequently a combination of approaches is necessary, particularly to add biological relevance to experimental data. Measurements can be made on bulk populations of RBC as either dilute suspensions or in whole blood, or on individual cells. Recent and exciting advances in nanotechnology and molecular biomechanics


101

have led to the development of techniques such as optical traps or laser tweezers, atomic force microscopy and optical stretchers. These techniques, together with advances in microfluidics, now allow direct, real-time mechanical probing and manipulation of single RBC at the micro- and nanometer level and with nano- and pico-Newton force resolution. While an exhaustive review of the technical details of these methods is outside the scope of this chapter, some of the most commonly used and currently emerging techniques are summarized below with particular reference to their utility in studies of sickle or malaria-infected RBC. 5.1. Established Techniques 5.1.1. Bulk Cell Filtration, Viscometry and Rheoscopy Assessment of the bulk mechanical properties of RBC by measurement of their transit time through pores in polycarbonate filters having either single or multiple pores has been a popular method due to its relative simplicity and ease of performance. Likewise, the use of a variety of different rotational viscometers, such as cone/plate or bob/cup, to derive a viscosity profile of a suspension of RBC under defined shear can provide useful information about the mechanical properties of individual cells in the suspension. These techniques formed the basis of pioneering studies by Miller et al [88] in 1971, who were the first to begin to accurately quantify the altered mechanical properties of monkey RBC infected with malaria parasites. Direct or indirect visualization of individual RBC under defined fluid shear stress (e.g., rheoscopy or ektacytometry) yields a shear-induced elongation or deformation index; rheoscope techniques using dilute suspensions provide data for individual cells, and as such allows more precise quantitation at the single cell level. For example, direct visualization of individual malaria-infected RBC in sheared bulk suspension in a rheoscope allowed Cranston et al. [89] to extend the earlier viscometric studies of Miller [88] and to relate the extent of mechanical modification to different life cycle stages of the parasite. 5.1.2. Laminar Shear Flow Systems The use of parallel-plate flow chambers allows visualization and quantitation of the interactions of flowing blood cells with vascular endothelial cells or purified adhesion proteins under conditions of flow that mimic those in the circulation. Moreover, unlike in vivo models, laminar flow systems allow fine control over both the biophysical conditions (e.g., shear stress, shear rate and viscosity) as well as the biochemical composition of the environment. First designed to study the adhesive properties of leukocytes [90], they have been used, in one form or another, quite extensively for studies of other blood cells, particularly sickle and malaria-infected RBC [15, 91-93]. Direct microscopic observation of adherent cells under flow conditions allows both quantitative and qualitative aspects of adhesion to be monitored. Interestingly, studies of malaria-infected RBC in flow chambers have revealed that, like leucocytes, they also show different adhesive behavior (i.e., rolling or stationary) depending on the identity of the endothelial cell receptor with which they interact [92, 94]. In addition to studies of adhesion, flow chambers have also been used to determine the mechanical properties of RBC. Estimates of the membrane shear elastic modulus of single cells can be derived from shear-induced elongation indices for point attached RBC [95, 96]. Such measurements are also influenced, however, by cellular factors

102


such as hemoglobin concentration and the presence of inclusion bodies and may therefore differ quite substantially from the absolute value obtained for this parameter when determined by micropipette aspiration techniques. 5.1.3. Micropipette Aspiration The micropipette aspiration technique was first developed by Mitchison and Swann [97] in 1954 to probe the elasticity of sea urchin eggs. Since then, this method has been used by others to measure the membrane elasticity of many types of cells including leukocytes and RBC (see [98] for review). In this technique, a negative hydrostatic pressure is used to partially or completely aspirate a single cell into a glass micropipette with diameter ranging from less than 1 μm to 10 μm in diameter. The time taken for a whole cell to enter a micropipette with a diameter smaller than the diameter of the cell being aspirated is related to the overall deformability of the whole cell. Aspiration of a membrane tongue into § 1 μm pipettes and measurement of tongue length as a function of aspiration pressure allows derivation of the shear elastic modulus of the membrane skeleton itself and is not influenced by other cellular properties that contribute to whole cell deformability (Figure 2).

Figure 2. (a) Illustration of micropipette aspiration to determine membrane shear elastic modulus, and (b) an optical image showing aspiration of membrane from a human RBC. (Taken from [99] with permission).

The technique has been used to quantify mechanical alterations in the membrane skeleton of a variety of abnormal RBC, most notably sickle and malaria-infected cells. [52, 87, 100-103]. Collectively, these studies on malaria-infected RBC have revealed that there is progressive rigidification of the membrane skeleton of the RBC that is related in magnitude to the life-cycle stage of the parasite. The normal elastic modulus of uninfected RBC increases more than 3 fold (from 5 PN/m to greater than 15 PN/m) as parasites mature inside the RBC; this rigidification is related to the presence of knob structures that appear on the RBC membrane due to proteins that are exported from the parasite and interact specifically with the RBC membrane skeleton [52, 101]. Micropipettes have also been used to quantify the forces involved in the adhesive interactions between blood cells and the endothelium. For example, the force of interaction between normal RBC and a malaria-infected RBC in a rosette is at least 5 times higher than those between parasitized RBC and vascular endothelial cells indicating that these abnormal aggregates of RBC could easily exist in the microcirculation in humans infected with malaria parasites [91, 104]. Clearly, further


103

investigations of the precise function and contribution of these malaria proteins to the altered rheological properties of RBC is warranted since they are likely to play a pivotal role in the vaso-occlusive pathology that accompany this disease in humans. 5.2. Emerging Technologies In recent years, advances in high-resolution imaging, nanotechnology, molecular engineering and microfluidics, together with new techniques in cellular and molecular biology, have revolutionized our ability to measure the adhesive and mechanical properties of cells and relate these changes to the action of individual genes and proteins. Some examples are discussed below with particular reference to how they might, or have already, advanced our understanding of human malaria. 5.2.1. Atomic Force Microscopy Recently, the atomic force microscope (AFM) has been used as a powerful imaging tool with nanometer and sub-nanometer scale, and also as a force sensor with picoNewton resolution [99]. The AFM uses a very sharp tiny tip mounted at the end of a flexible cantilever to directly probe the surface of a sample. Precise relative lateral and vertical displacement between the sample and AFM tip can be obtained through a computer-controlled piezoelectric stage or cantilever holder. The interaction force between the tip and sample surface can also be obtained from the deflection of the cantilever. Furthermore, coating the tip with purified proteins also permits forces between specific ligands and their cell-expressed receptors to be accurately quantified, often down to the level of a single molecular interaction. The technique has been used to study adhesive interactions at the molecular level of both leucocytes and RBC [105, 106]. One distinct advantage of the AFM is that samples can be tested in both dry and wet environments as well as under or near physiologic conditions. Hence, it is an excellent tool for probing living cells and biomolecules such as proteins and nucleic acids. When infected by P. falciparum malaria parasites, the surface of an infected RBC becomes punctuated by as many as up to 10,000 electron-dense knob-like elevations (Figure 1). These knobs appear to be located over the junctional complexes of the RBC membrane skeleton and can vary in size from 70 nm to 150 nm in diameter as well as density (e.g., 10 /μm2 to 70 /μm2) [107]. AFM provides an excellent tool by which to image these structures in detail as well as to obtain a better understanding of the proteins responsible for their formation [70, 108]. Using this technique, it will be of interest to determine if and how the morphology of these structures differ from individuals with abnormal RBC such as hemoglobin C RBC or South East Asian ovalocytes, since these conditions are known to be protective against the severe vasoocclusive complications that accompany malaria in humans [109, 110]. AFM can also be applied to measure the membrane mechanical properties of RBC. For example, a recent study has revealed that RBC from individuals with diabetes mellitus have a shear elastic modulus that is up to 3 times higher than normal RBC [111]. 5.2.2. Optical Traps or Laser Tweezers Optical traps or laser tweezers use high intensity laser light to trap, control and manipulate minute particles in a suspending medium. When laser light passes through a dielectric particle which has a higher refractive index than the suspending medium, a

104


net light pressure or gradient force occurs which pushes the particle towards the focal point of the laser. As a result, the particle becomes “trapped” by the laser. The optical or laser trap has now become an important research tool in both physics and biology. RBC can be stretched over a large range using optical tweezers by non-specifically attaching dielectric silica microbeads of 1-4 μm diameter to the surface of the cell. These beads form ‘handles’ by which a RBC can be dragged through a viscous suspending medium in order to determine its overall elasticity [112]. Alternatively, two beads, attached at diametrically opposite points of the cell, can form a trap in which the cell can be uniaxially stretched (Figure 3) [113-115].

Figure 3. Stretching of a normal RBC using optical trap or laser tweezers. Two silica microbeads (4 μm diameter) are non-specifically attached to two opposite points of a RBC. (a) As a single trap is used, the left bead is immobilised to the glass slide surface while the right bead remains free. (b) The cell is stretched by first trapping and holding the right bead stationary while moving the glass slide towards the left. (Taken from [115] with permission).

The in-plane shear elastic modulus of the cell can be inferred from force/length measurements, and has been estimated to be between 3 to 8 PN/m for normal RBC, and thus within the range previously determined by others using either optical traps or micropipette aspiration [98, 116-118]. The method has recently been applied to study the altered mechanical properties of RBC infected with different life-cycle stages of P. falciparum malaria parasites [119] (Figure 4). 5.2.3. Microfluidics Microfluidics is the study of fluid flow at the microscale with the volume of fluid involved normally thousands of times smaller than that of a droplet. Such microscale systems can provide very precise control and manipulation of small volumes of fluids. Recently, there has been a tremendous increase in the use of such microfluidic devices in the areas of biology, biochemistry and medicine [120, 121]. There are several advantages to using microfluidic devices to study the rheological properties of RBC. They can be easily manufactured with a high degree of accuracy, sophistication and reproducibility using microfabrication and lithography techniques to mimic the architecture and flow dynamics of the in vivo microvasculature. Materials such as silicone elastomer can be used to fabricate channels with structural properties


105

that are close to those of human capillaries [122]. Furthermore, experiments can be performed using extremely small sample volumes and are highly suited for highthroughput screening and analysis. Elastomeric microfluidic channels with a height of 2 μm and width of 2-8 μm have been used recently to study the abnormal flow properties of malaria-infected RBC [123] (Figure 5). Providing further functionality to microfluidic channels by coating them with proteins expressed on the surface of vascular endothelial cells, together with more accurate quantitation, will further increase the utility of such devices for more advanced studies of both the mechanical and adhesive properties of RBC.

Figure 4: Analysis of the mechanical properties of P. falciparum (Pf) parasitized RBC (pRBC) by stretching using laser tweezers. Optical images of healthy RBC (H-RBC) and control uninfected RBC (Pf-U-RBC) are compared with RBC infected with ring stage (Pf-pRBC), trophozoite stage (Pf-T-pRBC) or schizont stage (Pf-S-pRBC) P. falciparum parasites either before stretching (left column), or under an applied force of 68 pN (middle column) or 151 pN (right column). (Taken from [119] with permission).

6. Theoretical and Computational Models of RBC Rheology Computational modeling and simulation is an important tool that can be used to complement physical or biological experiments, or provide vital information in situations where experiments are impossible or impractical to perform. It has provided insight into the mechanics of bulk blood flow, the shape, motion and deformation of individual blood cells and their dynamic cellular responses arising from mechanical deformation [124]. An interesting mathematical model developed recently is the first attempt to model the behavior of normal RBC in 3-dimensions in order to mimic their deformation when stressed in the circulation. The mathematical model, based on a novel concept to represent the actin protofilaments of the membrane skeleton as tensegrity structures within a geodesic dome, have given the first real insight into

106


precisely how junctional complexes interact with multiple spectrin molecules and maintain such precise 3-dimensional organization of the RBC membrane skeleton [125]. More conventional methods such as finite element modeling have also been applied to both normal RBC and RBC infected with malaria parasites to model RBC deformation induced by laser tweezers or micropipette aspiration [102, 113, 115]. Unfortunately, the great complexity of biological systems and biomaterials and the rapidity at which they respond to mechanical stress has meant that advances in theoretical and computational biology have lagged behind the recent development of new experimental methods. Clearly, new tools and more complex algorithms are necessary for the future development of this important complimentary discipline.

Figure 5: Analysis of the flow properties of malaria-infected RBC in microfluidic channels. Sequential video images showing RBC infected with different stages of P. falciparum malaria parasites passing through microchannels with widths between 2 and 8 μm. Complete blockage of channels smaller than 8 μm occurs with RBC infected with mature-stage parasites (Taken from [123] with permission).

7. Therapeutics for RBC Rheological Diseases Theoretically, rationally designed therapeutics that prevent or reverse specific mechanical or adhesive abnormalities in RBC should significantly ameliorate the severity of diseases associated with these cellular alterations. The feasibility and clinical benefit of such an approach has probably best been demonstrated for leukocyte or platelet adhesion disorders with the development of anti-adhesion therapies designed to antagonise integrin receptors [126]. For RBC disorders, sickle cell disease and malaria are likely to benefit most from the development of such strategies. Increased understanding of the molecular interactions of sickle or malaria-infected RBC with the vascular wall has identified a number of ligands which may be potential


107

targets for small molecule or antibody-based adhesion inhibitors. Reagents with therapeutic potential have been identified and tested in vitro including recombinant proteins based on the malaria adhesin PfEMP1 [127] or peptides that block ĮVȕ3 integrin mediated adhesion of sickle RBC [128]. Since the efficacy of such drugs in vivo would be dependent on high and sustained levels in the plasma, such small peptides would require chemical modification to form N-methylated, cyclic or mimetic peptides in order to increase their in vivo stability and maximize their plasma half life. Therapeutic modification of the mechanical properties of sickle RBC using ionchannel blockers such as ICA-17043 (clotrimazole), pentoxifylline, cetiedil, bepridil or, albeit to a lesser extent, L-arginine may also be of clinical benefit. These compounds reduce the level of dehydration in sickle cells resulting in decreased cellular hemoglobin concentration. This decreased concentration markedly reduces the tendency for HbS to polymerize and the extent of binding of hemoglobin to the RBC membrane skeleton. Together, these improve RBC morphology and increase whole cell deformability; such changes may reduce the trapping of mechanically-impaired RBC in the microcirculation. Whether the mechanical properties of malaria-infected RBC can be modified in vivo, and whether this would have a significant effect on progression or severity of the disease, remains unknown. However, recent identification of the specific domains within a number of exported malaria proteins that mediate their interaction with specific proteins of the RBC membrane skeleton, together with knowledge of their functional consequences at the cellular level, open up the possibility to develop small molecule inhibitors which could interfere specifically with these interactions. For example, preventing the interaction of KAHRP with spectrin using small peptides [71] could have major beneficial effects on reducing both the mechanical and the adhesive abnormalities of malaria-infected RBC.

8. Conclusions and Outlook Arguably, advances in the biomedical and biophysical sciences in recent years have provided the impetus for scientists to develop renewed interest in the rheological properties of RBC. New techniques in cellular and molecular biology, genomics and proteomics, imaging, micro- and nano-technology and molecular bioengineering have been put to use to obtain a much greater level of understanding at the molecular level of how structure and function in RBC are related, how their unique rheological properties are maintained and, more precisely, how these are perturbed during infection or disease. With this increased knowledge come new opportunities for research but also new challenges. Further improvements in bio-imaging techniques with greater resolution and smaller and more precisely fabricated microfluidic technologies will be fundamental to our continued progress toward new therapeutics designed to prevent or reverse cellular rheological phenomena associated with a variety of human diseases. Use of newly combined techniques, such as atomic force and/or near field microscopy with Raman spectroscopy, to scan the surface of P. falciparum parasitized RBC will enable the knob-like structures to be imaged and directly correlated with spectroscopic data, thereby providing further information on the precise composition of these complexes. The development of an in vitro culture system for other species of human malaria parasites, particularly P. vivax, is urgently required so that we can begin to study their effects on RBC and elucidate the molecular mechanisms by which they

108


cause such an important and widespread human disease. Finally, as new techniques develop and our knowledge in molecular rheology increases, the RBC will continue to provide a basis for understanding and predicting the mechanical and adhesive behavior of more complex cells and tissues in health and disease.

References [1]

[2] [3] [4]

[5] [6]

[7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]

[20] [21] [22] [23] [24]

J.A. Chasis and N. Mohandas, Erythrocyte membrane deformability and stability: two distinct membrane properties that are independently regulated by skeletal protein associations, J. Cell. Biol. 103 (1986), 343-350. E.A. Evans and R. M. Hochmuth, A solid-liquid composite model of the red cell membrane, J. Membr. Biol. 30 (1977), 351-362. N. Mohandas, Molecular basis for red cell membrane viscoelastic properties, Biochem. Soc. Trans. 20 (1992), 776-782. N. Mohandas and J.A. Chasis, Red blood cell deformability, membrane material properties and shape: regulation by transmembrane, skeletal and cytosolic proteins and lipids, Semin. Hematol. 30 (1993), 171-192. N. Mohandas and E. Evans, Mechanical properties of the red cell membrane in relation to molecular structure and genetic defects, Ann. Rev. Biophys. Biomol. Struct. 23 (1994), 787-818. L.D. Walensky, N. Mohandas and S.E.I. Lux, in Blood. Principles and practice of hematology, Second ed. R. I. Handin, S. E. I. Lux and T. P. Stossel, Eds., Lippincott Williams & Wilkins, Philadelphia, 2003, 1709-1858. T.M. Fischer, M. Stohr-Lissen and H. Schmid-Schonbein, The red cell as a fluid droplet: tank tread-like motion of the human erythrocyte membrane in shear flow, Science 202 (1978), 894-896. P. Gaehtgens, C. Duhrssen and K.H. Albrecht, Motion, deformation, and interaction of blood cells and plasma during flow through narrow capillary tubes, Blood Cells 6 (1980), 799-817. B. Bull, J. Stuart and I. Juhan-Vague, In: Investigative microtechniques in medicine and biology, J. Chayen and L. Bitensky, Eds., Marcel Dekker, New York, 1984, 257-295. J. Stuart, B. Bull and I. Juhan-Vague, In: Investigative microtechniques in medicine and biology, J. Chayen and L. Bitensky, Eds., Marcel Dekker, New York, 1984. J. Stuart, Erythrocyte rheology, J. Clin. Pathol. 38 (1985), 965-977. E.A. Evans, Structure and deformation properties of red blood cells: concepts and quantitative methods, Methods Enzymol. 173 (1989), 3-35. U. Bagge, P.I. Branemark, R. Karlsson and R. Skalak, Three-dimensional observations of red blood cell deformation in capillaries, Blood Cells 6 (1980), 231-239. T.W. Secomb and R. Skalak, A two-dimensional model for capillary flow of an asymmetric cell, Microvasc. Res 24 (1982), 194-203. R.P. Hebbel, Adhesive interactions of sickle erythrocytes with endothelium, J. Clin. Invest. 100 (1997), S83-86. G. Fairbanks, T.L. Steck and D.F. Wallach, Electrophoretic analysis of the major polypeptides of the human erythrocyte membrane, Biochemistry 10 (1971), 2606-2617. V. Bennett and D.M. Gilligan, The spectrin-based membrane skeleton and micron-scale organization of the plasma membrane, Ann. Rev. Cell Biol. 9 (1993), 27-66. E. Ling, Y.N. Danilov and C.M. Cohen, Modulation of red cell band 4.1 function by cAMP-dependent kinase and protein kinase C phosphorylation, J. Biol. Chem. 263 (1988), 2209-2216. S. Manno, Y. Takakuwa, K. Nagao and N. Mohandas, Modulation of erythrocyte membrane mechanical function by beta-spectrin phosphorylation and dephosphorylation, J. Biol. Chem. 270 (1995), 5659-5665. M.A. De Matteis and J.S. Morrow, Spectrin tethers and mesh in the biosynthetic pathway, J. Cell Sci. 113 ( Pt 13) (2000), 2331-2343. J.G. Conboy, Structure, function, and molecular genetics of erythroid membrane skeletal protein 4.1 in normal and abnormal red blood cells, Semin. Hematol. 30 (1993), 58-73. N.Mohandas and P. Gascard, What do mouse gene knockouts tell us about the structure and function of the red cell membrane?, Baillieres Best Pract. Res. Clin. Haematol. 12 (1999), 605-620. E.M. Pasini, M. Kirkegaard, P. Mortensen, H.U. Lutz, A.W. Thomas and M. Mann, In-depth analysis of the membrane and cytosolic proteome of red blood cells, Blood 108 (2006), 791-801. G.B. Nash and H.J. Meiselman, Red cell and ghost viscoelasticity. Effects of hemoglobin concentration and in vivo aging, Biophys. J. 43 (1983), 63-73.


109

[25] M.R. Clark, Mean corpuscular hemoglobin concentration and cell deformability, Ann. NY Acad. Sci. 565 (1989), 284-294. [26] W.H. Reinhart, L.P. Sung and S. Chien, Quantitative relationship between Heinz body formation and red blood cell deformability, Blood 68 (1986), 1376-1383. [27] N. Mohandas, L.E. Lie-Injo, M. Friedman and J.W. Mak, Rigid membranes of Malayan ovalocytes: a likely genetic barrier against malaria, Blood 63 (1984), 1385-1392. [28] J. Delaunay, Genetic disorders of the red cell membrane, Crit. Rev. Oncol. Hematol. 19 (1995), 79-110. [29] W.T. Tse and S.E. Lux, Red blood cell membrane disorders, Brit. J. Haematol. 104 (1999), 2-13. [30] S. Suresh, Mechanical response of human red blood cells in health and disease: Some structureproperty-function relationships, J. Mat. Res. 21 (2006), 1871-1877. [31] S. Eber and S.E. Lux, Hereditary spherocytosis--defects in proteins that connect the membrane skeleton to the lipid bilayer, Semin. Hematol. 41 (2004), 118-141. [32] A.E. Schofield, M.J. Tanner, J.C. Pinder, B. Clough, P.M. Bayley, G.B. Nash, A.R. Dluzewski, D.M. Reardon, T.M. Cox, R.J. Wilson and W.B. Gratzer, Basis of unique red cell membrane properties in hereditary ovalocytosis, J. Mol. Biol. 223 (1992), 949-958. [33] N. Mohandas, R. Winardi, D. Knowles, A. Leung, M. Parra, E. George, J. Conboy and J. Chasis, Molecular basis for membrane rigidity of hereditary ovalocytosis. A novel mechanism involving the cytoplasmic domain of band 3, J. Clin. Invest. 89 (1992), 686-692. [34] A.E. Schofield, D.M. Reardon and M.J. Tanner, Defective anion transport activity of the abnormal band 3 in hereditary ovalocytic red blood cells, Nature 355 (1992), 836-838. [35] M.H. Steinberg, B.G. Forget, D.R. Higgs and R.L. Nagel, Disorders of hemoglobin, Cambridge University Press, New York, 2001. [36] D.J. Weatherall and J.B. Clegg, The thalassaemia syndromes, Fourth ed., Blackwell Science, Oxford, 2001. [37] S.K. Ballas and N. Mohandas, Sickle red cell microrheology and sickle blood rheology, Microcirculation 11 (2004), 209-225. [38] E. Evans, N. Mohandas and A. Leung, Static and dynamic rigidities of normal and sickle erythrocytes. Major influence of cell hemoglobin concentration, J. Clin. Invest. 73 (1984), 477-488. [39] T.M. Wick and J.R. Eckman, Molecular basis of sickle cell-endothelial cell interactions, Curr. Opin. Hematol. 3 (1996), 118-124. [40] N.M. Matsui, L. Borsig, S.D. Rosen, M. Yaghmai, A. Varki and S.H. Embury, P-selectin mediates the adhesion of sickle erythrocytes to the endothelium, Blood 98 (2001), 1955-1962. [41] B.J. Thevenin, I. Crandall, S.K. Ballas, I.W. Sherman and S.B. Shohet, Band 3 peptides block the adherence of sickle cells to endothelial cells in vitro, Blood 90 (1997), 4172-4179. [42] P.G. Rowland, G.B. Nash, B.M. Cooke and J. Stuart, Comparative study of the adhesion of sickle cells and malarial-parasitized red cells to cultured endothelium, J. Lab. Clin. Med. 121 (1993), 706-713. [43] J.G. Breman, The ears of the hippopotamus: manifestations, determinants, and estimates of the malaria burden, Am. J. Trop. Med. Hyg. 64 (2001), 1-11. [44] K. Mendis, B.J. Sina, P. Marchesini and R. Carter, The neglected burden of Plasmodium vivax malaria, Am. J. Trop. Med. Hyg. 64 (2001), 97-106. [45] C. Newbold, A. Craig, S. Kyes, A. Rowe, D. Fernandez-Reyes and T. Fagan, Cytoadherence, pathogenesis and the infected red cell surface in Plasmodium falciparum, Int. J. Parasitol. 29 (1999), 927-937. [46] P.E. Duffy and M. Fried, Plasmodium falciparum adhesion in the placenta, Curr. Opin. Microbiol. 6 (2003), 371-376. [47] G.G. MacPherson, M.J. Warrell, N.J. White, S. Looareesuwan and D.A. Warrell, Human cerebral malaria. A quantitative ultrastructural analysis of parasitized erythrocyte sequestration, Am. J. Pathol. 119 (1985), 385-401. [48] C. Raventos-Suarez, D. K. Kaul, F. Macaluso, R. L. Nagel, Membrane knobs are required for the microcirculatory obstruction induced by Plasmodium falciparum-infected erythrocytes, Proc. Natl. Acad. Sci. U S A 82 (1985), 3829-3833. [49] B.M. Cooke, N. Mohandas and R.L. Coppel, Malaria and the red blood cell membrane, Semin. Hematol. 41 (2004), 173-188. [50] B.M. Cooke, N. Mohandas and R.L. Coppel, The malaria-infected red blood cell: structural and functional changes, Adv. Parasitol. 50 (2001), 1-86. [51] B.M. Cooke, N. Mohandas, A. F. Cowman, R. L. Coppel, Cellular adhesive phenomena in apicomplexan parasites of red blood cells, Vet. Parasitol. 132 (2005), 273-295. [52] F.K. Glenister, R.L. Coppel, A.F. Cowman, N. Mohandas and B.M. Cooke, Contribution of parasite proteins to altered mechanical properties of malaria-infected red blood cells, Blood 99 (2002), 10601063.

110


[53] M.J. Gardner, N. Hall, E. Fung, O. White, M. Berriman, R.W. Hyman, J.M. Carlton, A. Pain, K.E. Nelson, S. Bowman, I.T. Paulsen, K. James, J.A. Eisen, K. Rutherford, S.L. Salzberg, A. Craig, S. Kyes, M.S. Chan, V. Nene, S.J. Shallom, B. Suh, J. Peterson, S. Angiuoli, M. Pertea, J. Allen, J. Selengut, D. Haft, M.W. Mather, A.B. Vaidya, D.M. Martin, A.H. Fairlamb, M.J. Fraunholz, D.S. Roos, S.A. Ralph, G.I. McFadden, L.M. Cummings, G.M. Subramanian, C. Mungall, J.C. Venter, D J. Carucci, S.L. Hoffman, C. Newbold, R.W. Davis, C.M. Fraser and B. Barrell, Genome sequence of the human malaria parasite Plasmodium falciparum, Nature 419 (2002), 498-511. [54] T.J. Sargeant, M. Marti, E. Caler, J.M. Carlton, K. Simpson, T.P. Speed and A.F. Cowman, Lineagespecific expansion of proteins exported to erythrocytes in malaria parasites, Genome Biol. 7 (2006), R12. [55] M. Foley, L. Corcoran, L. Tilley and R. Anders, Plasmodium falciparum: mapping the membranebinding domain in the ring-infected erythrocyte surface antigen, Exp. Parasitol. 79 (1994), 340-350. [56] M. Foley, L.J. Murray and R.F. Anders, The ring-infected erythrocyte surface antigen protein of Plasmodium falciparum is phosphorylated upon association with the host cell membrane, Mol. Biochem. Parasitol. 38 (1990), 69-75. [57] M. Foley, L. Tilley, W.H. Sawyer and R.F. Anders, The ring-infected erythrocyte surface antigen of Plasmodium falciparum associates with spectrin in the erythrocyte membrane, Mol Biochem Parasitol 46 (1991), 137-147. [58] E. Da Silva, M. Foley, A.R. Dluzewski, L.J. Murray, R.F. Anders and L. Tilley, The Plasmodium falciparum protein RESA interacts with the erythrocyte cytoskeleton and modifies erythrocyte thermal stability, Mol. Biochem. Parasitol. 66 (1994), 59-69. [59] M.D. Silva, B.M. Cooke, M. Guillotte, D.W. Buckingham, J.P. Sauzet, C. Le Scanf, H. Contamin, P. David, O. Mercereau-Puijalon and S. Bonnefoy, A role for the Plasmodium falciparum RESA protein in resistance against heat shock demonstrated using gene disruption, Mol. Microbiol. 56 (2005), 9901003. [60] K.L. Waller, W. Nunomura, X. An, B.M. Cooke, N. Mohandas and R.L. Coppel, Mature parasiteinfected erythrocyte surface antigen (MESA) of Plasmodium falciparum binds to the 30-kDa domain of protein 4.1 in malaria-infected red blood cells, Blood 102 (2003), 1911-1914. [61] S. Lustigman, R.F. Anders, G.V. Brown and R.L. Coppel, The mature-parasite-infected erythrocyte surface antigen (MESA) of Plasmodium falciparum associates with the erythrocyte membrane skeletal protein, band 4.1, Mol. Biochem. Parasitol. 38 (1990), 261-270. [62] R.L. Coppel, S. Lustigman, L. Murray and R.F. Anders, MESA is a Plasmodium falciparum phosphoprotein associated with the erythrocyte membrane skeleton, Mol. Biochem. Parasitol. 31 (1988), 223-231. [63] R.L. Coppel, Repeat structures in a Plasmodium falciparum protein (MESA) that binds human erythrocyte protein 4.1, Mol. Biochem. Parasitol. 50 (1992), 335-347. [64] C. Magowan, R.L. Coppel, A.O. Lau, M.M. Moronne, G. Tchernia and N. Mohandas, Role of the Plasmodium falciparum mature-parasite-infected erythrocyte surface antigen (MESA/PfEMP-2) in malarial infection of erythrocytes, Blood 86 (1995), 3196-3204. [65] B.M. Cooke, F.K. Glenister, N. Mohandas and R.L. Coppel, Assignment of functional roles to parasite proteins in malaria-infected red blood cells by competitive flow-based adhesion assay, Brit. J. Haematol. 117 (2002), 203-211. [66] K.L. Waller, B.M. Cooke, W. Nunomura, N. Mohandas and R.L. Coppel, Mapping the binding domains involved in the interaction between the Plasmodium falciparum knob-associated histidine-rich protein (KAHRP) and the cytoadherence ligand P. falciparum erythrocyte membrane protein 1 (PfEMP1), J. Biol. Chem. 274 (1999), 23808-23813. [67] K.L. Waller, W. Nunomura, B.M. Cooke, N. Mohandas and R.L. Coppel, Mapping the domains of the cytoadherence ligand Plasmodium falciparum erythrocyte membrane protein 1 (PfEMP1) that bind to the knob-associated histidine-rich protein (KAHRP), Mol. Biochem. Parasitol. 119 (2002), 125-129. [68] A. Kilejian, M.A. Rashid, M. Aikawa, T. Aji and Y.F. Yang, Selective association of a fragment of the knob protein with spectrin, actin and the red cell membrane, Mol. Biochem. Parasitol. 44 (1991), 175181. [69] B.S. Crabb, B.M. Cooke, J.C. Reeder, R.F. Waller, S R. Caruana, K.M. Davern, M.E. Wickham, G.V. Brown, R.L. Coppel and A.F. Cowman, Targeted gene disruption shows that knobs enable malariainfected red cells to cytoadhere under physiological shear stress, Cell 89 (1997), 287-296. [70] M. Rug, S.W. Prescott, K.M. Fernandez, B.M Cooke and A.F. Cowman, The role of KAHRP domains in knob formation and cytoadherence of P falciparum-infected human erythrocytes, Blood 108 (2006), 370-378. [71] X. Pei, X. An, X. Guo, M. Tarnawski, R. Coppel and N. Mohandas, Structural and functional studies of interaction between Plasmodium falciparum knob-associated histidine-rich protein (KAHRP) and erythrocyte spectrin, J. Biol. Chem. 280 (2005), 31166-31171.


111

[72] P. Horrocks, R.A. Pinches, S.J. Chakravorty, J. Papakrivos, Z. Christodoulou, S.A. Kyes, B.C. Urban, D.J. Ferguson and C.I. Newbold, PfEMP1 expression is reduced on the surface of knobless Plasmodium falciparum infected erythrocytes, J. Cell Sci. 118 (2005), 2507-2518. [73] B.L. Pasloske, D.I. Baruch, M.R. van Schravendijk, S.M. Handunnetti, M. Aikawa, H. Fujioka, T.F. Taraschi, J.A. Gormley and R.J. Howard, Cloning and characterization of a Plasmodium falciparum gene encoding a novel high-molecular weight host membrane-associated protein, PfEMP3, Mol. Biochem. Parasitol. 59 (1993), 59-72. [74] J.G. Waterkeyn, M.E. Wickham, K.M. Davern, B.M. Cooke, R.L. Coppel, J.C. Reeder, J.G. Culvenor, R.F. Waller and A.F. Cowman, Targeted mutagenesis of Plasmodium falciparum erythrocyte membrane protein 3 (PfEMP3) disrupts cytoadherence of malaria-infected red blood cells, Embo J. 19 (2000), 2813-2823. [75] E. Knuepfer, M. Rug, N. Klonis, L. Tilley and A.F. Cowman, Trafficking determinants for PfEMP3 export and assembly under the Plasmodium falciparum-infected red blood cell membrane, Mol. Microbiol. 58 (2005), 1039-1053. [76] D. Mattei and A. Scherf, The Pf332 gene of Plasmodium falciparum codes for a giant protein that is translocated from the parasite to the membrane of infected erythrocytes, Gene 110 (1992), 71-79. [77] K. Hinterberg, A. Scherf, J. Gysin, T. Toyoshima, M. Aikawa, J.C. Mazie, L.P. da Silva and D. Mattei, Plasmodium falciparum: the Pf332 antigen is secreted from the parasite by a brefeldin A-dependent pathway and is translocated to the erythrocyte membrane via the Maurer's clefts, Exp. Parasitol. 79 (1994), 279-291. [78] J. Wiesner, D. Mattei, A. Scherf and M. Lanzer, Biology of giant proteins of Plasmodium: resolution on polyacrylamide-agarose composite gels, Parasitol. Today 14 (1998), 38-40. [79] J.F. Kun, A.R. Hibbs, A. Saul, D.J. McColl, R.L. Coppel and R.F. Anders, A putative Plasmodium falciparum exported serine/threonine protein kinase, Mo.l Biochem. Parasitol. 85 (1997), 41-51. [80] D.I. Baruch, B.L. Pasloske, H.B. Singh, X. Bi, X. C. Ma, M. Feldman, T.F. Taraschi and R.J. Howard, Cloning the P. falciparum gene encoding PfEMP1, a malarial variant antigen and adherence receptor on the surface of parasitized human erythrocytes, Cell 82 (1995), 77-87. [81] N. Kriek, L. Tilley, P. Horrocks, R. Pinches, B.C. Elford, D.J. Ferguson, K. Lingelbach and C.I. Newbold, Characterization of the pathway for transport of the cytoadherence-mediating protein, PfEMP1, to the host cell surface in malaria parasite-infected erythrocytes, Mol. Microbiol. 50 (2003), 1215-1227. [82] D.I. Baruch, S.J. Rogerson and B.M. Cooke, Asexual blood stages of malaria antigens: cytoadherence, Chem. Immunol. 80 (2002), 144-162. [83] J.D. Smith, B. Gamain, D.I. Baruch, S. Kyes and Decoding the language of var genes and Plasmodium falciparum sequestration, Trends. Parasitol. 17 (2001), 538-545. [84] S.S. Oh, S. Voigt, D. Fisher, S.J. Yi, P.J. LeRoy, L.H. Derick, S. Liu and A.H. Chishti, Plasmodium falciparum erythrocyte membrane protein 1 is anchored to the actin-spectrin junction and knobassociated histidine-rich protein in the erythrocyte skeleton, Mol. Biochem. Parasitol. 108 (2000), 237247. [85] V. Fernandez and M. Wahlgren, Rosetting and autoagglutination in Plasmodium falciparum, Chem. Immunol. 80 (2002), 163-187. [86] I.W. Sherman, S. Eda and E. Winograd, Cytoadherence and sequestration in Plasmodium falciparum: defining the ties that bind, Microbes. Infect. 5 (2003), 897-909. [87] G.B. Nash, E. O'Brien, E.C. Gordon-Smith and J.A. Dormandy, Abnormalities in the mechanical properties of red blood cells caused by Plasmodium falciparum, Blood 74 (1989), 855-861. [88] L.H. Miller, S. Usami and S. Chien, Alteration in the rheologic properties of Plasmodium knowlesiinfected red cells. A possible mechanism for capillary obstruction, J. Clin. Invest. 50 (1971), 1451-1455. [89] H.A. Cranston, C.W. Boylan, G.L. Carroll, S.P. Sutera, J.R. Williamson, I.Y. Gluzman and D.J. Krogstad, Plasmodium falciparum maturation abolishes physiologic red-cell deformability, Science 223 (1984), 400-403. [90] M.B. Lawrence, L.V. McIntire and S.G. Eskin, Effect of flow on polymorphonuclear leukocyte/ endothelial cell adhesion, Blood 70 (1987), 1284-1290. [91] G.B. Nash, B.M. Cooke, K. Marsh, A. Berendt, C. Newbold and J. Stuart, Rheological analysis of the adhesive interactions of red blood cells parasitized by Plasmodium falciparum, Blood 79 (1992), 798807. [92] B.M. Cooke, R.L. Coppel, Cytoadhesion and falciparum malaria: going with the flow, Parasitol. Today 11 (1995), 282-287. [93] B.M. Cooke, S. Usami, I. Perry and G.B. Nash, A simplified method for culture of endothelial cells and analysis of adhesion of blood cells under conditions of flow, Microvasc. Res. 45 (1993), 33-45.

112


[94] B.M. Cooke, A.R. Berendt, A.G. Craig, J. MacGregor, C.I. Newbold and G.B. Nash, Rolling and stationary cytoadhesion of red blood cells parasitized by Plasmodium falciparum: separate roles for ICAM-1, CD36 and thrombospondin, Brit. J.Haematol. 87 (1994), 162-170. [95] G.B. Nash and H.J. Meiselman, Alteration of red cell membrane viscoelasticity by heat treatment: effect on cell deformability and suspension viscosity, Biorheology 22 (1985), 73-84. [96] R. Suwanarusk, B.M. Cooke, A.M. Dondorp, K. Silamut, J. Sattabongkot, N.J. White and R. Udomsangpetch, The deformability of red blood cells parasitized by Plasmodium falciparum and Pvivax, J. Infec. Dis. 189 (2004), 190-194. [97] J.M. Mitchison and M.M. Swann, The mechanical properties of the cell surface I. The cell elastimeter, J. Exp. Biol. 31 (1954). [98] R.M. Hochmuth, Micropipette aspiration of living cells, J. Biomech. 33 (2000), 15-22. [99] C.T. Lim, E.H. Zhou, A. Li, S. R.K. Vedula and H.X. Fu, Experimental techniques for single cell and single molecule biomechanics, Mater. Sci. Eng. C 26 (2006), 1278-1288. [100] G.B. Nash, C.S. Johnson and H.J. Meiselman, Mechanical properties of oxygenated red blood cells in sickle cell (HbSS) disease, Blood 63 (1984), 73-82. [101] M. Paulitschke and G.B. Nash, Membrane rigidity of red blood cells parasitized by different strains of Plasmodium falciparum, J Lab Clin Med 122 (1993). [102] C. T. Lim, E. H. Zhou and S. T. Quek, Mechanical models for living cells - A review, J. Biomech. 39 (2006), 195-216. [103] T. Itoh, S. Chien and S. Usami, Effects of hemoglobin concentration on deformability of individual sickle cells after deoxygenation, Blood 85 (1995), 2245-2253. [104] G.B. Nash, B.M. Cooke, J. Carlson and M. Wahlgren, Rheological properties of rosettes formed by red blood cells parasitized by Plasmodium falciparum, Brit. J. Haematol. 82 (1992), 757-763. [105] W. Hanley, O. McCarty, S. Jadhav, Y. Tseng, D. Wirtz and K. Konstantopoulos, Single molecule characterization of P-selectin/ligand binding, J. Biol. Chem. 278 (2003), 10556-10561. [106] S. Sen, S. Subramanian and D.E. Discher, Indentation and adhesive probing of a cell membrane with AFM: theoretical model and experiments, Biophys. J. 89 (2005), 3203-3213. [107] J. Gruenberg, D.R. Allred and I.W. Sherman, Scanning electron microscope-analysis of the protrusions (knobs) present on the surface of Plasmodium falciparum-infected erythrocytes, J. Cell. Biol. 97 (1983), 795-802. [108] A. Li, A.H. Mansoor, K.S.W. Tan and C.T. Lim, Observations on the internal and surface morphology of malaria-infected blood cells using optical and atomic force microscopy, J. Microbiol. Methods 66 (2006), 434-439. [109] A. Cortes, M. Mellombo, C.S. Mgone, H.P. Beck, J.C. Reeder and B.M. Cooke, Adhesion of Plasmodium falciparum-infected red blood cells to CD36 under flow is enhanced by the cerebral malaria-protective trait South-East Asian ovalocytosis, Mol. Biochem. Parasitol. 142 (2005), 252-257. [110] R.M. Fairhurst, D.I. Baruch, N.J. Brittain, G.R. Ostera, J.S. Wallach, H.L. Hoang, K. Hayton, A. Guindo, M.O. Makobongo, O.M. Schwartz, A. Tounkara, O.K. Doumbo, D.A. Diallo, H. Fujioka, M. Ho and T.E. Wellems, Abnormal display of PfEMP-1 on erythrocytes carrying haemoglobin C may protect against malaria, Nature 435 (2005), 1117-1121. [111] M. Fornal, M. Lekka, G. Pyka-Fosciak, K. Lebed, T. Grodzicki, B. Wizner, J. Styczen, Erythrocyte stiffness in diabetes mellitus studied with atomic force microscope, Clin. Hemorheol. Microcirc. 35 (2006), 273-276. [112] M.M. Brandao, A. Fontes, M.L. Barjas-Castro, L.C. Barbosa, F.F. Costa, C.L. Cesar and S.T. Saad, Optical tweezers for measuring red blood cell elasticity: application to the study of drug response in sickle cell disease, Eur. J. Haematol. 70 (2003), 207-211. [113] M. Dao and C.T. Lim, S. Suresh, Mechanics of the human red blood cell deformed by optical tweezers, J. Mech. Phys. Solids (2003), 2259-2280. [114] C.T. Lim, M. Dao, S. Suresh, C.H. Sow and K.T. Chew, Large deformation of living cells using laser traps, Acta Materialia 52 (2004), 1837-1845. [115] J.P. Mills, L. Qie, M. Dao, C.T. Lim and S. Suresh, Nonlinear elastic and viscoelastic deformation of the human red blood cell with optical tweezers, MCB 1 (2004), 169-180. [116] D.E. Discher, D.H. Boal and S.K. Boey, Simulations of the erythrocyte cytoskeleton at large deformation. II. Micropipette aspiration, Biophys. J. 75 (1998), 1584-1597. [117] S. Henon, G. Lenormand, A. Richert and F. Gallet, A New Determination of the shear modulus of the human erythrocyte membrane using optical tweezers, Biophys. J. 76 (1999), 1145-1151. [118] J. Sleep, D. Wilson, R. Simmons and W. Gratzer, Elasticity of the red cell membrane and its relation to hemolytic disorders: an optical tweezers study, Biophys. J. 77 (1999), 3085-3095. [119] S. Suresh, J. Spatz, J. P. Mills, A. Micoulet, M. Dao, C.T. Lim, M. Beil and T. Seufferlein, Connections between single-cell biomechanics and human disease states: gastrointestinal cancer and malaria, Acta Biomaterialia 1 (2005), 15-30.


113

[120] G.M. Whitesides, E. Ostuni, S. Takayama, X.Y. Jiang and D.E. Ingber, Soft lithography in biology and biochemistry, Ann. Rev. Biomed. Eng. 3 (2001), 335-373. [121] J. Voldman, M.L. Gray and M.A. Schmidt, Microfabrication in biology and medicine, Ann. Rev. Biomed. Eng. 1 (1999), 401-425. [122] S.C. Gifford, M.G. Frank, J. Derganc, C. Gabel, R.H. Austin, T. Yoshida and M.W. Bitensky, Parallel microchannel-based measurements of individual erythrocyte areas and volumes, Biophys. J. 84 (2003), 623-633. [123] J.P. Shelby, J. White, K. Ganesan, P.K. Rathod and D.T. Chiu, A microfluidic model for single-cell capillary obstruction by Plasmodium falciparum infected erythrocytes, Proc. Natl. Acad. Sci. U S A 100 (2003), 14618-14622. [124] V. Cristini and G.S. Kassab, Computer modeling of red blood cell rheology in the microcirculation: a brief overview, Ann. Biomed. Eng. 33 (2005), 1724-1727. [125] C. Vera, R. Skelton, F. Bossens and L.A. Sung, 3-D nanomechanics of an erythrocyte junctional complex in equibiaxial and anisotropic deformations, Ann. Biomed. Eng. 33 (2005), 1387-1404. [126] B.S. Coller, Platelet GPIIb/IIIa antagonists: the first anti-integrin receptor therapeutics, J. Clin. Inves.t 99 (1997), 1467-1471. [127] B.M. Cooke, C.L. Nicoll, D.I. Baruch and R.L. Coppel, A recombinant peptide based on PfEMP-1 blocks and reverses adhesion of malaria-infected red blood cells to CD36 under flow, Mol. Microbiol. 30 (1998), 83-90. [128] A. Kumar, J.R. Eckman and T.M. Wick, Inhibition of plasma-mediated adherence of sickle erythrocytes to microvascular endothelium by conformationally constrained RGD-containing peptides, Am. J. Hemato.l 53 (1996), 92-98.

114


Red Blood Cell Aggregation a

Björn NEUa,1 and Herbert J. MEISELMANb School of Chemical and Biomedical Engineering, Nanyang Technological University, Singapore b Keck School of Medicine, University of Southern California, Los Angeles, USA

Introduction The reversible aggregation of human red blood cells (RBC) continues to be of interest in the field of hemorheology [1-12], in that RBC aggregation is a major determinant of the in vitro rheological properties of blood. In addition, the in vivo flow dynamics and flow resistance of blood are influenced by RBC aggregation [13]. Measures of RBC aggregation, such as the erythrocyte sedimentation rate (ESR), are commonly used as diagnostic tests and as one index to the efficacy of therapy (e.g., following drug therapy in rheumatoid arthritis); in diabetes mellitus RBC aggregation is normalized by improved glycemic control [14] There is now general agreement regarding the correlations between elevated levels of fibrinogen or other large plasma proteins and enhanced RBC aggregation, and the effects of molecular mass and concentration for neutral polymers such as dextran [15]. However, the specific mechanisms involved in RBC aggregation have not yet been elucidated, and thus it is not yet possible to fully understand the relations between pathology and altered RBC aggregation. RBC form multi-cell linear or branched aggregates in vitro when they are suspended in either plasma or solutions containing large polymers (e.g., dextran ≥ 40 kDa); the linear forms are often termed rouleaux since they resemble a stack of coins. In vivo RBC aggregation occurs at low shear forces or stasis and is a major determinant of low shear blood viscosity and thus in vivo flow dynamics [13]. It is important to note that RBC aggregation is a reversible process, with aggregates dispersed by mechanical or fluid flow forces, and then reforming when the forces are removed. Conversely, RBC agglutination and blood coagulation are irreversible processes due to either protein polymerization or strong antigen-antibody attractive forces. Abnormal increases of RBC aggregation have been observed in several diseases associated with vascular disorders (e.g., diabetes mellitus or hypertension). RBC aggregation is primarily determined by RBC aggregability (i.e., the intrinsic cell characteristics affecting RBC aggregation) and by the concentration of the inducing macromolecule or the plasma level of large proteins [16]. In blood, fibrinogen is one of the most important determinants of blood viscosity due to its strong tendency to increase both plasma viscosity and RBC aggregation [17]. In the past, most reports dealt primarily with the ability of plasma proteins to promote aggregation; for example higher fibrinogen levels have been linked to elevated blood viscosities in hypertensive patients [18]. 1 Corresponding Author: Division of Bioengineering, School of Chemical and Biomedical Engineering, Nanyang Technological University, 70 Nanyang Drive, Singapore 637457; E mail: [email protected]

B. Neu and H.J. Meiselman / Red Blood Cell Aggregation

115

Figure 1. Schematic drawing of polymer depletion at the red blood cell surface; if two cells approach solvent is displaced from the depletion zone into the bulk phase.

At present, there are two co-existing models for RBC aggregation: bridging and depletion. In the bridging model, red cell aggregation is proposed to occur when the bridging forces due to the adsorption of macromolecules onto adjacent cell surfaces exceed disaggregating forces due to electrostatic repulsion, membrane strain and mechanical shearing [15, 19-23]. This model seems to be similar to other cell interactions like agglutination, with the only difference being that the proposed adsorption energy of the macromolecules is much smaller in order to be consistent with the relative weakness of these forces. In contrast, the depletion model proposes quite the opposite. In this model RBC aggregation occurs as a result of a lower localized protein or polymer concentration near the cell surface compared to the suspending medium (i.e., relative depletion near the cell surface). This exclusion of macromolecules near the cell surface leads to an osmotic gradient and thus depletion interaction [24]. As with the bridging model, disaggregation forces are electrostatic repulsion, membrane strain and mechanical shearing. Several previous reports have dealt with the experimental and theoretical aspects of depletion aggregation, often termed depletion flocculation, as applied to the general field of colloid chemistry [25-28]. However, polymer depletion as a mechanism for red blood cell aggregation has received much less attention, with only a few literature reports relevant to this approach [24, 29-33].

1. Macromolecular Depletion at the Red Blood Cell Surface The above mentioned two models are in conflict: The bridging model predicts increased aggregation consequent to increased protein or polymer concentration at the RBC surface, whereas the depletion model predicts the opposite. As described in the previous section, the preferential exclusion of macromolecules leading to depletion

116


flocculation has received considerable attention by colloid scientists during the past decades. This, in turn, makes it surprising that this phenomenon has been almost ignored when considering the stability of cell suspensions. In part, this lack of attention is due to the previous and current problems associated with deciding if a depletion layer has occurred at cell surfaces. One reason is the small extension of such depletion layers. Depletion layers are in the same size range as the hydrated size of the depleted macromolecule, meaning for most plasma proteins known to induce aggregation, it is only a few nanometers; the thickness of the RBC glycocalyx has been determined to also be a few nanometers. Thus, it is not only impossible to detect such a layer with direct optical observations but, in addition, it is also quite challenging to distinguish between weak absorption or a depletion effect since the latter might also involve some intermixing (i.e., penetration) of the macromolecule and the glycocalyx. So even though past reports have described surface adsorption of dextran, and specific binding mechanism between fibrinogen and RBC have been outlined [1, 19, 20, 34], such results should be interpreted with great care. In fact, an extensive review of literature values by Janzen and Brooks [35] has detailed likely technical artifacts (e.g., trapped fluid between RBC) and thus the extremely wide range of reported data for fibrinogen and dextran binding. Several studies have investigated the structure and extension of depletion layers using various techniques [27, 36, 37]. One method, which is also applicable to macromolecular depletion at biological interfaces, is to investigate the depletion layer by means of electrophoresis. In solutions of neutral soluble polymers that give rise to depletion layers, particles have an unexpectedly high mobility [19, 23, 30]. This effect is due to the reduced viscosity near the particle surface due to the depletion effect [32]: electro-osmotic flow decreases rapidly outside the electric double layer, so if the double layer thickness is comparable to or larger than that of the depletion layer, the influence of suspending phase viscosity is reduced. Figure 2 presents results from a study which was directed toward validating the existence of the depletion layer by employing measurements of unit-gravity cell sedimentation and of cell mobility for RBC in various polymer solutions [30]. These studies thus tested the hypothesis that regardless of polymer molecular weight, the effects of suspending medium viscosity on sedimentation could be predicted via the Stokes Equation, whereas electrophoretic mobility (EPM) would become less sensitive to medium viscosity with increasing molecular weight (i.e., with increasing depletion layer thickness). As shown in Figure 2, the Stokes Equation is applicable regardless of polymer weight, whereas the electrophoretic mobility follows the expected inverse relation only for small polymers (i.e., 10 kDa dextran) with RBC electrophoretic mobility essentially independent of medium viscosity for large polymers (i.e., 500 kDa dextran). The latter point is of particular interest in that it demonstrates the existence of the depletion layer and its dependence on molecular size; dextran 10 kDa has a hydrated radius of about 3 nm compared to about 20 nm for the 500 kDa dextran [30]. Consequently, particle electrophoresis has been used extensively to study macromolecules at biological interfaces. Via changing the ionic strength of the suspending medium it is even possible to probe the extent of the depletion layer. Variation of ionic strength varies the thickness of the double layer, and thus measurement of mobility at different ionic strengths allows the estimation of depletion layer thickness [36]. As long as the double layer is significantly thinner


117

than the depletion layer, the mobility will be higher than predicted based upon media viscosity. With increasing double layer thickness the influence of bulk viscosity increases and when the double layer thickness becomes significantly greater than the depletion layer, the measured mobilities or ζ-potentials are unaffected by viscosity in the depletion region. Using this approach it has been confirmed that the thickness of such depletion layers is in the same range as expected for the hydrated size of these polymers, thereby agreeing with the concept of polymer depletion near the RBC surface and lending strong support to a “depletion model” mechanism for reversible RBC aggregation [24, 37-39].

EPM(dextran)/EPM(polymer free)

a)

1.0

Dextran 500 kDa

0.8 Dextran 10 kDa 0.6

0.4

0.2

theo retic al ra t

1.0

1.5 2.0 2.5 3.0 solvent viscosity (mPa·s)

1.0

1.5 2.0 2.5 3.0 solvent viscosity (mPa·s)

b) SED(dextran)/SED(polymer free)

io

1.0

0.8

0.6

0.4

0.2

Figure 2. a) The relative electrophoretic mobility (EPM) and b) the sedimentation rate (SED) of cells suspended in dextran solutions (i.e., values relative to those for cells suspended in dextranfree phosphate buffered saline) plotted against medium viscosity. The dextrans used had molecular weights of 10 and 500 kDa. The predicted relationships are shown as curved lines; separate lines are shown for the 10 and 500 kDa dextrans predicted sedimentation (SED) behavior due to different increases of medium density caused by these two polymers (modified from [30]).

118


It should also be noted that some studies have evaluated depletion of proteins and polyelectrolytes at RBC surfaces by means of particle electrophoresis. However, the details of such experiments as well as interpretation of the data are not as straightforward as with neutral polymers. For charged polyelectrolytes and hence also for proteins, it is also necessary to consider electrostatic forces which can also affect depletion or adsorption [40]. In addition, forces between the particle surface and the polyelectrolyte differ with ionic strength, and thus adsorption and depletion can vary with the extent of the double layer as well as the conformation of the charged macromolecule. For example, for polymers and surfaces of the same sign, only slight or no adsorption is often observed at low ionic strength, whereas with increasing ionic strength, the adsorbed amount increases. This behavior is due to the electrostatic screening, which increases with increasing salt concentration; in the case of pure electro-adsorption (i.e., when the surface charges and the polymer charges have opposite signs), the opposite effects of ionic strength are observed [40]. One consequence of the above-mentioned phenomena is that it is not possible to determine the thickness of the depletion layer by simply changing the ionic strength of the suspending media as is utilized for neutral polymers such as dextran. Further, if only measuring at a single ionic strength, it is sometimes not possible to decide if apparent changes of the particle mobility should be attributed to depletion or adsorption. Thus, even though protein and polyelectrolyte depletion can be detected with particle electrophoresis, it is also clear that for the investigation of plasma proteins at bio-interfaces it is still necessary to develop other techniques in order to obtain quantitative results.

2. Red Blood Cell Aggregability Over the past several decades, studies of RBC aggregation have focused primarily on the effects of protein or polymer concentration and molecular weight and, until fairly recently, much less attention has been focused on the effects of cellular factors on RBC aggregation. Work in the area of cellular factors affecting RBC aggregation seems to have been prompted by the salient observation of Nordt in 1983 [41]. He showed that density-separated (i.e., age-separated) RBC exhibited different degrees of aggregation when suspended in autologous plasma: older, denser cells exhibited greater aggregation than younger, less-dense RBC. This observation has been confirmed, and correlations between RBC aggregation in autologous plasma and in a standard polymer solution have been demonstrated [42, 43]. These results indicate that there is a large variation in the aggregating potential of cells from different normal, healthy subjects: high responders to plasma also tend to be high responders to the polymer dextran, whereas low responders in one medium are low in the other. Hence, there must be intrinsic cell–specific factors that affect the aggregation tendency of red cells, with these factors differing greatly between individuals. Note that these observations regarding intrinsic cell-specific factors were initially surprising, since it previously had been tacitly assumed that all red cells had the same aggregating tendency. That is, RBC aggregation differences between subjects for cells in plasma were assumed to be due only to differences in plasma concentrations of various aggregating agents; donor differences for cells in the same polymer solution were, when considered, usually ascribed to experimental error or measurement variability.


119

In order to distinguish between the effects of suspending media properties (e.g., polymer or protein type and concentration) and those intrinsic to the red cell, the term aggregability has been coined to indicate the intrinsic tendency of RBC to undergo aggregation. There are thus two important definitions: 1) aggregation refers to the measured rate, extent or strength of RBC aggregation in any medium; 2) aggregability refers to the measured rate, extent or strength of RBC aggregation in a defined medium and is used to compare RBC populations when each population is tested in the same defined medium. Thus, greater aggregation in a defined medium for a given cell population would indicate greater RBC aggregability than for another cell population exhibiting less aggregation. Note that the material in this section deals primarily with red cell aggregability for various RBC populations, including those from various healthy neonatal and adult donors, from individuals with chronic or acute diseases, from experimental studies that altered the physicochemical characteristics of red blood cells and from non-human mammals. 2.1. Donor–Specific Effects (Healthy Human Adults) The first major report in this area was that of Sowemimo-Coker and co-workers [43], who studied erythrocytes from different healthy human adult subjects suspended in autologous plasma or, after washing the cells free of plasma proteins, in phosphate buffered saline containing 3% dextran 70 (MW = 73 kDa). All suspensions were made to the same haematocrit, and the degree of aggregation measured using a light transmission technique (i.e., Myrenne Erythrocyte Aggregometer, see [44]). Paired data (i.e., extent of aggregation in plasma and in dextran) for each subject were plotted against one another, with typical results shown in Figure 3. Analysis of the results in Figure 3 leads to two important observations. First, the difference in aggregation between subjects is large, approximately two-fold in the dextran solution and five-fold in autologous plasma. Such ranges are far beyond the approximately 3-7% variation associated with experimental and instrumental uncertainty [45]. Thus, in the case of the dextran data, the range can only be explained by subject-to-subject variations in cell-specific factors that affect RBC aggregation in a defined solution. That is, subject-to-subject differences in RBC aggregability. Conversely, for the plasma data, it is not possible to eliminate suspending phase properties inasmuch as the level of pro-aggregant plasma proteins (e.g., fibrinogen) vary from donor to donor. Unfortunately, plasma protein levels were not reported in this study and therefore both cell and medium properties most likely contribute to the large range observed for RBC in autologous plasma [46]. Second, high responders in dextran tend to be high responders in autologous plasma and vice versa (i.e., a linear correlation between aggregation in the two media, p<0.001). This latter finding suggests that cell-specific factors may be of similar importance for any aggregating agent. Experimental evidence supporting the suggestion that aggregability differences are evident for cells suspended in various polymer solutions has been presented by Whittingstall and co-workers [47]. Using techniques similar to those employed by Nash and co-workers [42] and Sowemimo-Coker, et.al [43], they investigated RBC aggregation induced by isotonic solutions of polyvinylpyrrolidone 360 (PVP, MW = 360 kDa, 0.5%), poly-l-glutamic acid (P-L-Glu, MW = 61.2 kDa, 0.6%) and sodium heparin (MW = 17.0 kDa, 6%) as well as dextran (dextran 70, MW = 70 kDa, 3%) and autologous plasma. For each of the four polymers the extent of aggregation

120


RBC AGGREGATION INDEX (PLASMA)

produced by the aggregant was correlated with that produced by the plasma, with the correlation results shown in Table 1. These results clearly indicate that in spite of differences in size and charge of the molecules (dextran and PVP are neutral while PL-Glu and heparin are negative), aggregation induced by each aggregant exhibits a significant correlation with that induced by plasma. In addition, significant correlations were found between aggregation induced by any two polymers [47], again supporting the concept of similar aggregation behavior regardless of the aggregant. 35

r = 0.674 p < 0.001

30 25 20 15 10 5 0 15

20

25

30

35

40

45

50

RBC AGGREGATION INDEX (3% DEXTRAN 70) Figure 3. Normal RBC aggregation in autologous plasma versus aggregation for the same cells in a defined isotonic polymer solution (3% solution of 70 kDa dextran); each square represents a different donor. Cells in dextran were washed twice in protein-free buffer the re-suspended in the dextran solution. Note the wide range of aggregation in dextran (i.e., wide range of aggregability) and the correlation between aggregation in the two media.

Rampling and co-workers [48] have also demonstrated donor-specific factors using several methods to study RBC aggregation. RBC from 43 healthy adult subjects were washed and re-suspended in phosphate buffered saline containing either PVP 360 (2.6 g/l), dextran 70 (15 g/l), heparin (20 g/l) or fibrinogen (6 g/l); RBC in autologous plasma were also studied. Aggregation was determined using the erythrocyte sedimentation rate (ESR), the Myrenne Aggregometer, and low shear rate (0.28 s-1) apparent viscosity measured in a co-axial viscometer [49]. Their results indicated that for each polymer, the variability among the 43 subjects (mean CV = 25 ± 7%) was much larger than that due solely to experimental error: the CV for washed and re-suspended RBC equaled that for plasma, yet the cells were tested in defined and constant media. Associations between the degree of aggregation generated by the four polymer aggregants were also reported [16]: highly significant correlations (p<0.025 or better) were observed between each of the polymer aggregants. Their results thus confirm and extend the earlier findings of Sowemimo–Coker, et al. [43],

121


and allow the conclusion that cell-specific factors affecting RBC aggregability vary widely between human subjects and that the aggregation response to one aggregant is predictive of the aggregation response to another. Table 1. Correlations between plasma- and polymer induced RBC aggregation. Polymer

Correlation Coefficient

p-Value

Dextran 70

0.67

<0.001

PVP 360

0.88

<0.01

P-L-Glu

0.93

<0.001

Heparin

0.65

<0.05

Results are linear correlation coefficients with significance of slope tested versus zero.

2.2. Donor-Specific Results (Neonates) Several reports (e.g., [48, 50]) have indicated that, compared to adult blood, RBC from full-term neonates in autologous plasma exhibit significantly less aggregation. Further, such aggregation depends on gestational age, in that blood from infants at 24 to 28 weeks of gestation does not show significant RBC aggregation [50]. Red cell aggregation behavior of both preterm and term blood has been postulated to be due to age-related increases in the concentration of proteins known to induce aggregate formation as well as to variation of fibrinogen sialination with gestational age [51]: neonates generally have lower fibrinogen levels than adults and have hypersialinated fibrinogen. Studies dealing with the aggregability of neonatal RBC in polymer solutions give results that are more complicated than for adults: the neonatal data appear to depend critically upon the specific polymer. Using washed cells suspended in an isotonic solution of dextran 500 (MW = 500 kDa, 1%), Linderkamp and co-workers [50] indicate no differences versus adult cells in the rate or extent of aggregation regardless of gestational age. Further, neonatal RBC suspended in compatible adult plasma showed aggregation similar to that for adult cells in the plasma. Conversely, studies of term neonatal RBC in a lower molecular weight dextran indicated marked and highly significant differences from adult cells: using 3% dextran 70 (MW = 70 kDa), Whittingstall, et.al [52] report a 35% lower extent and a 60% lower strength of aggregation for the neonatal cells. These authors [24] confirmed that dextran 500 caused equal aggregation for neonatal and adult RBC, but only up to a concentration of 1%, the same concentration of dextran 500 employed by Linderkamp and coworkers [50]: neonatal cells had significantly lower aggregation at higher dextran 500 levels. Thus, both polymer type and polymer concentration affect aggregation differences between neonatal and adult RBC aggregation, thereby precluding generalized statements regarding cellular factors affecting polymer-induced neonatal RBC aggregation. 2.3. Effects of In Vivo Cell Age As indicated above in the introductory comments for this section of the chapter, the first report related to RBC aggregability appears to be that by Nordt [41], who employed high-speed centrifugation to achieve age separation of red cells: in normal

122


individuals, more dense cells are chronologically older whereas less dense cells are younger and have thus have spent less time in the circulation [53]. Nordt tested these age-separated cells in autologous plasma and observed that the 10% densest cells had more than two-fold greater aggregation compared to the 10% least dense cells, with cells of intermediate density exhibiting aggregation between these limits.

Figure 4. Aggregation behavior of age-separated (i.e., density-separated) RBC in autologous plasma and in isotonic 3% 70 kDa dextran solution. TOP RBC are younger, 10% least-dense cells and BOT RBC are older, 10% most-dense cells; MID RBC are remaining 80% of cell population. Note the greater aggregation for TOP, older RBC in both plasma and dextran.

Nash and co-workers [42] extended the seminal observations of Nordt to studies of age-separated RBC re-suspended in isotonic solutions of 70 kDa dextran; in their studies RBC from normal donors were fractionated by centrifugation into the top 10% (TOP), middle 80% (MID) and bottom (BOT) of the centrifuged red cell column. Confirming previous reports they also observed about two-fold greater aggregation in plasma for BOT versus TOP cells, with MID RBC aggregation behavior between these limits. Further, cells washed and re-suspended in 70 kDa dextran also demonstrated the effects of cell age: as shown in Figure 4, the order was BOT>MID>TOP, with again the BOT/TOP ratio being about two or greater over the entire range of dextran concentrations. The results shown in Figure 4 are therefore consistent with those for un-fractionated cells shown in Figure 3, in that enhanced aggregation in plasma predicts enhanced aggregation in dextran 70 solutions. Whittingstall, et al. [47] also tested age-separated RBC into TOP and BOT fractions and reported aggregation results for plasma, two neutral polymers (3% 70 kDa dextran and 0.5% 360 kDa PVP) and two negatively charged polymers (0.6% 61.2 kDa poly-l-glutamic acid and 6% sodium heparin). Salient results included: 1) As anticipated from Figure 4, BOT RBC exhibit significantly greater aggregation (p<0.01) in plasma and in 70 kDa dextran than TOP RBC, with the ratio of BOT/TOP for being 2.6 in plasma and 2.7 in dextran; 2) BOT RBC also exhibit significantly


123

greater aggregation (p<0.01) when suspended in the PVP, poly-l-glutamic acid or heparin media, the BOT/TOP ratios were 3.1 for PVP, 2.1 for heparin and 1.7 for poly-l-glutamic acid. Thus for both plasma and for neutral and negatively charged polymers, denser, older RBC exhibit significantly greater extent and strength of aggregation compared to less dense, younger cells. In brief, older cells have greater aggregability.

Figure 5. Aggregation behavior of age-separated (i.e., density-separated) adult and neonatal RBC in a 3% solution of 70 kDa dextran. TOP RBC are younger, 10% least-dense cells and BOT RBC are older, 10% most-dense cells. No differences were found for the TOP, younger cells whereas adult BOT, older cells had significantly greater aggregation.

An extension of the abovementioned studies employing age-separated adult RBC is an investigation of possible similar effects for neonatal RBC. Such studies are of particular value in addressing the effects of in vivo exposure time within the circulation: normal neonatal RBC have a lifetime that is about 60 days, and thus onehalf of the 120 day period for adult red cells [54]. Using a density-separation protocol identical to that employed for adult cells and 70 kDa dextran as the aggregant, it was found that both neonatal and adult RBC have a BOT/TOP aggregation ratio significantly greater than unity (Figure 5). The ratio of BOT/TOP for neonatal RBC, however, is markedly less than for adult RBC. Additionally, no significant difference (p>0.5) was observed between TOP neonatal and TOP adult cells, whereas this difference was highly significant for BOT neonatal versus adult cells. The results in Figure 5 thus suggest that the length of exposure to the in vivo circulatory environment may affect RBC aggregation tendency. However, other possibilities should be kept in mind since neonatal RBC are more prone to oxidant damage, have increased mechanical fragility, and have age-dependent RBC enzymes that decline more rapidly than in adult cells [50].

124


2.4. Clinical Conditions The body of literature dealing with clinical hemorheology and RBC aggregation in various clinical states and diseases is voluminous, with most studies employing either whole blood or RBC-plasma suspensions; essentially all results indicate enhanced aggregation for cells in autologous plasma [2, 3]. These enhanced levels of RBC aggregation raise the following question: Is RBC aggregability altered in clinical states and diseases, in particular those characterized by enhanced aggregation in plasma? Correlations between RBC aggregation in plasma and 70 kDa dextran have been demonstrated for normal subjects (Figure 3) and for subjects with Hansen’s disease [46], but do not directly address the question of cell-specific changes induced by the clinical conditions. However, there appear to be very few studies dealing with altered RBC aggregability in disease, and almost none related to possible rheological changes consequent to therapy. Table 2 presents a listing of some literature reports specifically dealing with comparisons of RBC aggregation in plasma and in a polymer solution and hence with RBC aggregability. It is clear that the extent of RBC aggregability alterations varies from condition to condition, and is subject to variations for the same disease: studies in Type 2 diabetes indicate either no alteration of aggregability [55] or significantly greater aggregation in a dextran solution [14]. Thus, as mentioned above for neonatal cells [50], polymer type and polymer concentration can affect aggregation differences between RBC populations, and therefore differences of aggregability in one polymer/concentration system may not always be applicable to another system. In general, altered aggregability seems to be associated with chronic clinical states (e.g., diabetes, unstable angina, beta-thalassemia, sickle cell disease) or due to relatively acute events that involve a major inflammatory state (e.g., sepsis). The study by Chong and co-workers [14] represents one of only a few in which RBC aggregation in plasma and in a defined polymer solution was followed over a period of therapy. In their study, blood was sampled from 55 subjects with Type 2 diabetes prior to and following a 14-week period of intensified management. During treatment, fasting glucose fell 27%, glycated hemoglobin fell 21% and serum lipids by 12-28% (p<0.0001 for each). Over the same 14-week period the extent and strength of RBC aggregation in plasma fell by 10-13% (p<0.002) and similar changes were seen for cells in a 3% solution of 70 kDa dextran (p<0.002). Note that the 14-week interval (i.e., 98 days) between samples represents a significant portion of the 120 day in vivo life span of RBC, and thus the decrease in aggregability occurred for a major proportion of RBC: whether this decrease was due to exposure of RBC to lower glucose levels during their time in the circulation or due to different RBC being released into the circulation is not known. 2.5. Altered Physicochemical Properties There are now several literature reports in which RBC physicochemical properties are altered, usually by chemical means, and the effects on aggregation in plasma or polymer evaluated. Very often differences in RBC aggregability were observed; some of these differences follow from current knowledge about RBC aggregation mechanisms, while a few remain unexpected and/or unexplained [67].


125

Table 2. Literature reports of RBC aggregability in various clinical states. Condition

Specifics

Polymer Aggregant

Aggregation Findings

Ref.

Diabetes

Poorly controlled adult onset (Type 2)

1% dextran 110 kDa

Same as controls in plasma and dextran

[55]

Diabetes

Type 2 with intense management

3% dextran 70 kDa

Decrease in plasma and dextran with improved control

[14]

Cardiac Ischemia

Acute Myocardial Infarction (AMI), Unstable Angina (UA)

0.5% dextran 500 kDa

All increase due to plasma for AMI, only 50% for UA

[56]

Hemoglobinopathies

Beta-Thalassemia


Increase primarily due to RBC

[57]

Hemoglobinopathies

Hemolytic anemia model in rat


No effects noted

[58]

Hemoglobinopathies

Sickle cell disease


Enhanced versus normal controls

[59]

Hypertension

Rat renal model


Similar increases in plasma and dextran

[60]

Lipid Abnormalities

LDL apheresis


All decrease due to plasma factors

[61]

Muscle Ischemia

Rat hind limb


Similar decrease in plasma and dextran

[62]

Nephrotic Syndrome

Renal patients

1% dextran 500 kDa

All increase due to plasma factors

[63]

Pregnancy

Normal pregnancy

1% dextran 500 kDa

All increase due to plasma factors

[64]

Pregnancy

Hyper- and normotensive subjects


75% (hyper) and 88% (normo) of increase due to plasma factors

[65]

Sepsis

Rat model

3% dextran 70 kDa

Same 3-fold increase in plasma and dextran

[66]

2.5.1. Enzyme Treatment A promising approach to investigating cellular factors affecting RBC aggregability is to study the effects of altering the cell’s membrane surface using specific enzymes. Jan and Chien [68] reported that treating red cells with neuraminidase to remove membrane bound sialic acid, and thus to reduce membrane charge density, greatly increased the aggregability of human erythrocytes. Similar results have been presented by Maeda and co-workers [69]. Nash, et al. [42] extended this approach to age-separated cells (i.e., TOP, MID, BOT), with their results indicating that neuraminidase treatment slightly reduced aggregation in plasma but markedly increased aggregation in 3% 70 kDa dextran, and that neuraminidase treatment did not alter the BOT/TOP aggregation ratio for cells in either media. The enhanced aggregation in dextran following enzyme treatment is consistent with the earlier report [70], while the failure to see a post-neuraminidase treatment increase of aggregation for cells in autologous plasma remains an enigma.

126


Rampling and Pearson [71] have compared the effects of digesting the RBC glycocalyx with a variety of proteolytic enzymes including neuraminidase, with results indicating: 1) increased aggregation for cells in 40 and 70 kDa dextran (rank order chymotrypsin to trypsin to bromelain); 2) that compared to the most effective proteolytic enzyme (i.e., bromelain), neuraminidase markedly increases aggregation for cells in both 40 and 70 kDa dextran solutions but has essentially no extra effect on fibrinogen-induced aggregation. These results again indicate that the effects of cellular factors on RBC aggregation may, under some circumstances, be specific to the aggregant. 2.5.2. Heat Treatment and Aldehyde Fixation Heating dilute RBC suspensions at 48 °C results in RBC with normal morphology but with increased membrane elastic shear modulus and viscosity and with reduced cellular deformability [72]. The effects of heating are time dependent and are minimal or non-existent for short time periods. Rakow and co-workers [73] report no changes of aggregation in 80 kDa dextran for cells heated for two minutes whereas heating for nine minutes decreases the extent and rate of aggregation in plasma yet increases the extent and rate in 3% 70 kDa dextran. Reinhart and Singh also report significant time effects for cells in plasma and 70 kDa dextran: slight increases of aggregation in both media at two and five minutes followed by progressive decreases over 10 to 80 minutes [74]. Treatment of RBC with aldehydes such as formaldehyde or glutaraldehyde also results in RBC with normal morphology but with decreased cellular deformability [75]. Overnight fixation with 2% formaldehyde increased aggregation in 70 kDa dextran [76] whereas a biphasic concentration effect was observed for glutaraldehydefixed cells in the same molecular weight dextran: aggregation increased up to about 0.03% glutaraldehyde then decreased to essentially zero at and above 0.05% [74]. 2.5.3. Miscellaneous The composition of the RBC phospholipid bilayer differs between the inner and outer leaflet and the absolute level of each component can be varied by several approaches. In particular, the ratio of phosphatidylcholine (PC) to phosphatidylserine (PS) in the outer leaflet affects aggregation in dextran solutions: incorporation of additional PC and hence raising the PC/PS ratio promotes aggregation whereas incorporation of PS has no effect on aggregability [77, 78]. DIDS, an inhibitor of anion transport across the RBC membrane, inhibits aggregation at 100-200 μM levels [79, 80]; the mechanism of action is not yet resolved, although a change of RBC morphology from biconcave disc to crenated sphere may be involved. The cationic local anesthetic procaine hydrochloride also affects aggregation in dextran: it has been reported to inhibit aggregation at low concentrations (80 to 800 μM) and to markedly increase aggregation at higher levels [81], or to always inhibit aggregation at levels of 4 μM and greater [82]. Storage of blood intended for transfusion affects aggregability in 0.5% 500 kDa dextran, with an 8-fold increase of aggregation noted over a 42-day period [7].


127

2.6. Non-Human Mammals Several literature reports have described blood rheology and the aggregation behavior of RBC for non-human mammals, with this area of hemorheology considered in Chapter II.7. The majority of studies presenting RBC aggregation data employed cells in autologous plasma, with salient results indicating large differences between species. For example, horse, leopard and rhinoceros exhibit very strong aggregation, no aggregation is observed for cattle, sheep and goat, with other mammals demonstrating intermediate behavior [83, 84]. Studies of the aggregation behavior of RBC washed and re-suspended in various polymer solutions have also demonstrated differences between species: rat red blood cells exhibit low aggregation in concentrations of 70 kDa dextran that markedly aggregate human RBC, and bovine cells show negligible aggregation in 70, 500 or 2000 kDa dextran or in 360 kDa polyvinylpyrrolidone [42]. Baskurt and co-workers [85] compared horse, human and rat RBC in 3% 70 kDa dextran and indicate that the horse RBC have a high response in both plasma and dextran, rat RBC are low in both, and human RBC are intermediate in behavior. Similar patterns have been reported for horse, bovine and human cells re-suspended in 2-3% solutions of 464 kDa dextran [86]. Marine mammals also demonstrate cell-specific effects, in that blood from Weddell seals (Leptonychotes weddellii) and from bowhead whales (Balaena mysticetus) exhibits intense RBC aggregation in autologous plasma and also in solutions of 57.5 kDa poly-l-glutamic acid, 70 kDa dextran and 360 kDa polyvinylpyrrolidone [87].

3. Depletion Mediated Red Blood Cell Interaction It is clear from the above discussion that there are cellular factors affecting RBC aggregation, and that these factors vary between individuals, between cell fractions from a given individual, and between various pathological states; RBC aggregability can also be modified by intentionally altering RBC characteristics. What is less clear, at present, is the detailed nature of these cellular factors and their specific contributions to the aggregation behavior of red blood cells. In general, however, it seems likely that differences in aggregability arise due to differences in physicochemical properties of the RBC glycocalyx. A major step for understanding how cellular factors might effect aggregation is to relate changes on the cell surface to changes in cell affinity. In order to calculate surface affinities between RBC when suspended in polymer solutions, it is first necessary to define the nature of the cell-cell interaction. The exterior RBC surface, termed the glycocalyx, consists of a complex layer of proteins and glycoproteins and bears a net negative charge that is primarily due to ionized sialic acid groups [88]. In the theoretical model employed herein, only depletion and electrostatic interactions are considered. As shown below, owing to the high electrostatic repulsion, cell-cell distances at which minimal interaction energy (i.e., maximal surface affinity) occur are always greater than twice the thickness of the cell’s glycocalyx. Thus, steric interactions between glycocalyx on adjacent RBC can be neglected. Further, calculated total interaction energies are in the order of 1-10 μJ/m2, whereas for cell separations greater than twice the glycocalyx thickness van

128


der Waals interactions are in the range of 10-2 μJ/m2 [89] and thus can also be neglected. 3.1. Depletion Interaction Energy If a surface is in contact with a polymer solution and the loss of configurational entropy of the polymer is not balanced by adsorption energy, a depletion layer develops near the surface. Within this layer the polymer concentration is lower than in the bulk phase. Thus, as two RBC approach, the difference of solvent chemical potential (i.e., the osmotic pressure difference) between the intercellular polymer-poor depletion zone and the bulk phase results in solvent displacement into the bulk phase and hence depletion interaction. Due to this interaction, an attractive force develops that tends to minimize the polymer-reduced space between the cells [40] . Examination of the energetics of depletion layers requires distinguishing between so-called “hard” and “soft or hairy” surfaces. Hard surfaces are considered to be smooth and do not allow polymer penetration into the surface, whereas soft surfaces, such as the RBC glycocalyx, are characterized by a layer of attached macromolecules that can be penetrated in part or entirely by the free polymer in solution [27, 90]. The depletion interaction energy wD can be calculated by assuming a step profile for the free polymer [27, 91]. Given a depletion layer thickness Δ, a polymer thickness δ, a separation distance of d between adjacent surfaces and a penetration of polymer into the glycocalyx p, wD is given by: d ⎛ ⎞ w D = −2Π⎜ Δ − + δ − p ⎟ 2 ⎠ ⎝

(1)

when (d/2-δ+p)<Δ and equals zero for (d/2-δ+p)>Δ. The osmotic pressure term Π is calculated using a viral equation neglecting all coefficients higher than the second (B2): Π=

( μ − μ10 ) RT b c 2 + B2 (c 2b ) 2 = − 1 M2 v1

(2)

where R, T, v1 and M2 are the gas constant, absolute temperature, molecular volume of the solvent and the molecular weight of the polymer. The chemical potential of the solvent in the polymer solution is μ1 and is μ10 in polymer free solution; cb2 represents the bulk polymer concentration. 3.2. Depletion Layer Thickness An approach introduced by Vincent [26] is used to calculate the depletion layer thickness (Δ). This approach is based upon calculation of the equilibrium between the compressional or elastic free energy and the osmotic force experienced by polymer chains at a non-absorbing surface and yields: 2

Δ=−

1 Π 1 ⎛Π⎞ 2 + ⎜ ⎟ + 4Δ 0 2 D 2 ⎝D⎠

(3)

129


where Π is the osmotic pressure of the bulk polymer solution. The parameter D is a function of the bulk polymer concentration (cb2): 2k T ⎛ c b N D = B2 ⎜ 2 a Δ 0 ⎜⎝ M 2

2

⎞3 ⎟ ⎟ ⎠

(4)

where kB and Na are the Boltzmann constant and Avogadro number. Δ0 is the depletion thickness for vanishing polymer concentration and is equal to 1.4·Rg, where Rg is the polymer’s radius of gyration [26]. 3.3. Penetration Depth Intuitively, the penetration depth p of the free polymer into the attached layer should depend on the polymer type, concentration and molecular size, and would be expected to be larger for small molecules and to increase with increasing polymer concentration due to increasing osmotic pressure. One possibility is to calculate p by assuming that penetration proceeds until the local osmotic pressure developed in the attached layer is balanced by the osmotic pressure of the bulk solution [27]. In this approach δ is assumed to be independent of bulk polymer concentration whereas a more realistic approach would also consider that the attached polymers, i.e. the glycocalyx, collapse under the osmotic pressure of the bulk polymer [90]. However, it is difficult to accurately apply such a model to RBC in polymer or a protein solution since too little is known about the physicochemical properties of the glycocalyx, and in particular, about the interaction between the glycocalyx and different polymers or proteins. Thus in the past an exponential approximation for the concentration dependence of the penetration depth has been applied [67, 92]: b p = δ ⎛⎜1 − e − c2 ⎝

c2p

⎞⎟ ⎠

(5)

where cp2 is the penetration constant of the polymer in solution (i.e., when cp2 equals cb2 , p is 63% of δ). In this approach δ is assumed to be independent of bulk polymer concentration. Therefore p is essentially a linear function of cb2 at low concentrations (relative to cp2) and asymptotically approaches δ at high concentrations. 3.4. Electrostatic Interaction Because of the strong electrostatic repulsion between adjacent cells other forces than depletion and electrostatic can be neglected [67, 92]. The electrostatic free energy of two cells can be calculated by considering an isothermal charging process as discussed previously. For d ≥ 2δ the electrostatic repulsion is then given by [92]: wE =

σ2 2

δ εε 0κ

3

(

sinh(κδ ) e κδ −κd − e −κd

)

(6)

where ε, εo, κ and σ are the relative and absolute permitivities, the inverse of the Debye-Hückel length and the surface charge density (i.e., charge per surface area).

130


Finally, the total interaction energy (wT) is given by the sum of the electrostatic interaction energy (wE) and the depletion interaction energy (wD) [92]: wT = w D + w E

(7)

Figure 6 presents calculated interaction energies between RBC as a function of dextran molecular weight and concentration. Qualitatively and quantitatively, calculations based upon the model presented herein show excellent agreement with experimental observations: 1) Depending on the method employed, aggregation in a given dextran usually shows a maximum between concentrations of 2 to 4%; 2) For human RBC from healthy donors, dextran 40 kDa does not promote aggregation whereas higher energies and greater aggregation are expected for larger polymers. It is also important to note that the calculated energies are in the expected range since experimental measurements of the energies leading to aggregation are usually on the order of a few μJ/m2 [93].

-6 100 molecular weight (kDa)

-4 -2 -1 50 wtotal (μJ/m2)

0

0

2 4 concentration (g/dL)

6

8

Figure 6. Lines of constant total interaction energy (wT) between adjacent RBC induced by dextran as a function of molecular weight and concentration (δ: 5nm, p: 5nm, σ: 36mC/m2).

4. Theoretical Interpretation of Aggregability As noted above, age-separated RBC exhibit significant differences in aggregability in either plasma or polymer solutions (i.e., older > younger), yet the differences are not related to altered cell volume, deformability or surface IgG levels. Based on the presented model one possibility for such changes would be an age dependent loss of surface charges and thus less repulsion between the cells and another possibility would be that the depletion layer increases at old cell surfaces due to surface alterations. With particle electrophoresis it is possible to distinguish between these two possibilities [24, 30].


131

EPM (10-8m 2V -1s -1)

1.2

1.1

+ 1%

TOP RBC BOT RBC

1 + 5%

0.9

0.8 Polymer free

3% Dextran 70

Figure 7. Electrophoretic mobility (EPM) of young (TOP) and old (BOT) RBC in polymer free phosphate buffered saline (PBS) and in PBS with 3% dextran 70 kDa (modified from [67]).

In Figure 7 electrophoretic mobilities of young and old cells in polymer free solution and in a solution containing 3 g/dl of dextran 70 kDa are compared. The mobilities in polymer-free solution show no significant difference and since the mobilities are determined by the surface charge there should not be an age dependent alteration of the electrostatic repulsion. In contrast, senescent RBC in 3 g/dl dextran 70 show a small but significant greater EPM (+5%) than young cells. This is most likely due to a reduced local viscosity at the cell surface, and thus an increased depletion layer at the surface of the old cells. The question remaining is if the changes in the depletion layer leading to a 5% difference in the mobilities are sufficient to explain the dramatic increase in aggregation. There are basically two possibilities (Figure 8) why the depletion layer increases: a decreasing polymer penetration or a decreasing glycocalyx thickness with cell age, both leading to lower viscosities at the cell surface and thus higher mobilities in polymer solution. Theoretical calculations showed that a decrease of 15% for the penetration (p) or the glycocalyx thickness (δ) would lead to 5% increase in mobilities and to an increase of 70-80% of the cell surface affinity, which would explain the differences in aggregation [67]. In overview, the model for depletion-mediated aggregation provides a rational framework for explaining the observed equal EPM values in buffer, the different EPM values in polymer solutions, and the marked increases of polymer-induced aggregation for age-separated RBC. That is, compared to less-dense TOP cells, denser BOT cells may have either a slightly thinner glycocalyx or a slightly less polymer penetration into their glycocalyx, and thus greater polymer depletion and larger osmotic forces favoring aggregation.

132

b)

pBOT pTOP

δ

distance from cell surface

polymer concentration

a)

polymer concentration


δBOT δTOP distance from cell surface Figure 8. Schematic representations of two approaches used to predict the effect of altered glycocalyx physicochemical properties on cell-cell affinity. a) Differing polymer penetration of the glycocalyx (i.e., pBOT versus pTOP) and b) differing thickness of the RBC glycocalyx (i.e., δBOT versus δTOP).

5. Outlook Even though the presented model appears robust, there are areas that require further attention. For example, the effects of abnormal RBC rheological behavior have not been considered [94], although they are acknowledged to potentially affect relations between calculated interaction energies and measured RBC aggregation. Red blood cells rigidified by heat treatment or chemical fixation are known to exhibit markedly decreased aggregation [42]. However, it seems evident that the above approach, which considers polymer depletion and electrostatic repulsion, is in qualitative and quantitative agreement with experimental measures of cell-cell affinity and RBC aggregation. Nevertheless, additional theoretical and experimental focus on more realistic treatments of the RBC glycocalyx and on interactions between the glycocalyx and charged or neutral polymers or proteins is needed. Charge distribution within the glycocalyx also needs to be considered since it has a significant effect on electrostatic interactional energy [89]. Although there are now several reports dealing with cellular factors affecting red blood cell aggregation (i.e., RBC aggregability), the field is still in its early stages and thus many questions remain to be answered. In particular, reasons for differences between subjects and between age-separated cells remain unclear. Further, there are


133

several experimental findings that are currently difficult to explain in terms of the depletion layer hypothesis, especially those where enzymatic manipulation of the cell surface is involved. Lastly, the authors wish to stress the potential clinical importance of understanding cellular factors affecting RBC aggregation. Disturbed in vivo blood flow consequent to elevated RBC aggregation has been observed in clinical states such as diabetes mellitus, myocardial infarction, and renal disease [2], and hence a clearer understanding of polymer-glycocalyx interactions should allow rational development of therapeutic agents. Detailed information about such interactions should aid in the development of methods for cellular modification to improve in vivo haemodynamics in the haemorheologically-compromised patient.

Acknowledgement This work was supported by grants from the National Institute of Health (USA, grants HL015722 and HL070595), the Deutsche Forschungsgemeinschaft (Germany, grant Ne 784/1) the Ministry of Education (Singapore, NTU AcRF grant RG36/05) and from A*Star (Singapore, BMRC grant 05/1/22/19/382).

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

[11] [12] [13] [14]

[15] [16]

S. Chien, S. Simchon, R. E. Abbot and K. M. Jan, Surface adsorption of dextrans on human red cell membrane, J. Colloid Interface. Sci. 62 (1977), 461-470. G. D. O. Lowe, Clinical Blood Rheology, CRC Press, Florida, 1988. J. F. Stoltz, M. Singh and P. Riha, Hemorheology in Practice, IOS Press, Amsterdam, 1999. H. J. Meiselman, O. K. Baskurt, S. O. Sowemimo-Coker and R. B. Wenby, Cell electrophoresis studies relevant to red blood cell aggregation, Biorheology 36 (1999), 427-432. B. Lim, P. A. J. Bascom and R. S. C. Cobbold, Simulation of red blood cell aggregation in shear flow, Biorheology 34 (1997), 423-441. N. B. Kounov and V. G. Petrov, Determination of erythrocyte aggregation, Math. Biosci. 157 (1999), 345-356. T. Hovav, S. Yedgar, N. Manny and G. Barshtein, Alteration of red cell aggregability and shape during blood storage, Transfusion 39 (1999), 277-281. L. Holley, N. Woodland, W. T. Hung, K. Cordatos and A. Reuben, Influence of fibrinogen and haematocrit on erythrocyte sedimentation kinetics, Biorheology 36 (1999), 287-297. G. Cloutier and Z. Qin, Ultrasound backscattering from non-aggregating and aggregating erythrocytes - A review, Biorheology 34 (1997), 443-470. A. Sennaoui, M. Boynard and C. Pautou, Characterization of red blood cell aggregate formation using an analytical model of the ultrasonic backscattering coefficient, IEEE Trans. Biomed. Eng. 44 (1997), 585-591. E. A. Evans and V. A. Parsegian, Energetics of membrane deformation and adhesion in cell and vesicle aggregation, Ann. N.Y. Acad. Sci. 416 (1983), 13-33. E. Evans and K. Buxbaum, Affinity of red blood cell membrane for particle surfaces measured by the extent of particle encapsulation, Biophys. J. 34 (1981), 1-12. M. Cabel, H. J. Meiselman, A. S. Popel and P. C. Johnson, Contribution of red blood cell aggregation to venous vascular resistance in skeletal muscle, Am. J. Physiol. 41 (1997), H1020-H1032. B. Chong-Martinez, T. A. Buchanan, R. Wenby and H. J. Meiselman, Decreased red blood cell aggregation subsequent to improved glycemic control in type 2 diabetes mellitus, Diabetic Med. 20 (2003), 301-306. S. Chien and L. A. Lang, Physicochemical basis and clinical implications of red cell aggregation, Clin. Hemorheol. 7 (1987), 71-91. M. W. Rampling, H. J. Meiselman, B. Neu and O. K. Baskurt, Influence of cell-specific factors on red blood cell aggregation., Biorheology 41 (2004), 91-112.

134


[17] A. J. Lee, The role of rheological and haemostatic factors in hypertension, J. Hum. Hypertens. 11 (1997), 767-776. [18] R. L. Letcher, S. Chien, T. G. Pickering, J. E. Sealey and J. H. Laragh, Direct relationship between blood pressure and blood viscosity in normal and hypertensive subjects. Role of fibrinogen concentration, Am. J. Med. 70 (1981), 1195-1202. [19] D. E. Brooks, The effect of neutral polymers on the electrokinetic potential of cells and other charged particles, J. Colloid Interface Sci. 43 (1973), 700-713. [20] D. E. Brooks, Mechanism of red cell aggregation, In: Blood Cells, Rheology and Aging, D. Platt, Ed., Springer Verlag, New York, 1988, pp .158-162. [21] S. Chien, Biophysical behavior of red cells in suspensions, In: The Red Blood Cell, D. M. Surgenor, Ed., Academic Press, Ne York, 1975, pp. 1031-1133. [22] S. Chien, R. J. Dellenback, S. Usami, D. A. Burton, P. F. Gustavson and V. Magazinovic, Blood volume, hemodynamic, and metabolic changes in hemorrhagic shock in normal and splenectomized dogs, Am. J. Physiol. 225 (1973), 866-879. [23] P. Snabre and P. Mills, Effect of Dextran Polymer on Glycocalyx Structure and Cell Electrophoretic Mobility, Colloid Polym. Sci. 263 (1985), 494-500. [24] H. Bäumler, E. Donath, A. Krabi, W. Knippel, A. Budde and H. Kiesewetter, Electrophoresis of human red blood cells and platelets. Evidence for depletion of dextran, Biorheology 33 (1996), 333351. [25] P. Jenkins and B. Vincent, Depletion flocculation of nonaqueous dispersions containing binary mixtures of nonadsorbing polymers. Evidence for Non-equilibrium effects, Langmuir 12 (1996), 3107-3113. [26] B. Vincent, The Calculation of Depletion Layer Thickness as a Function of Bulk Polymer Concentration, Colloids Surfaces 50 (1990), 241-249. [27] B. Vincent, J. Edwards, S. Emmett and A. Jones, Depletion flocculation in dispersions of stericallystabilised particles ("soft spheres"), Colloids and Surfaces 18 (1986), 261-281. [28] R. I. Feign and D. H. Napper, Depletion stabilization and depletion flocculation, J. Colloid Interface Sci. 75 (1980), 525-541. [29] J. K. Armstrong, H. J. Meiselman, R. B. Wenby and T. C. Fisher, Modulation of red blood cell aggregation and blood viscosity by the covalent attachment of Pluronic copolymers, Biorheology 38 (2001), 239-247. [30] B. Neu and H. J. Meiselman, Sedimentation and electrophoretic mobility behavior of human red blood cells in various dextran solutions, Langmuir 17 (2001), 7973-7975. [31] C. J. van Oss, K. Arnold and W. T. Coakley, Depletion flocculation and depletion stabilization of erythrocytes, Cell Biophys. 17 (1990), 1-10. [32] H. Bäumler and E. Donath, Does dextran indeed significantly increase the surface potential of human red blood cells?, Studia Biophysica 120 (1987), 113-122. [33] B. Neu, H. J. Meiselman and H. Bäumler, Electrophoretic mobility of human erythrocytes in the presence of poly(styrene sulfonate), Electrophoresis 23 (2002), 2363-2368. [34] D. Lominadze and W. L. Dean, Involvement of fibrinogen specific binding in erythrocyte aggregation, Febs Letters (1-3) 517 (2002), 41-44. [35] J. Janzen and D. E. Brooks, A critical reevaluation of the nonspecific adsorption of plasma proteins and dextrans to erythrocytes and the role of these in rouleaux formation, In: Interfacial Phenomena in Biological Systems, M. Bender, Ed., Marcel Dekker, New York, 1991, pp. 193-250. [36] E. Donath, P. Kuzmin, A. Krabi and A. Voigt, Electrokinetics of Structured Interfaces with Polymer Depletion - a Theoretical-Study, Colloid Polym. Sci. 271 (1993), 930-939. [37] A. Krabi and E. Donath, Polymer Depletion Layers as Measured by Electrophoresis, Colloid Surf. APhysicochem. Eng. Asp. 92 (1994), 175-182. [38] E. Donath, A. Krabi, M. Nirschl, V. M. Shilov, M. I. Zharkikh and B. Vincent, Stokes friction coefficient of spherical particles in the presence of polymer depletion layers - Analytical and numerical calculations, comparison with experimental data, J. Chem. Soc.-Faraday Trans. 93 (1997), 115-119. [39] E. Donath, L. Pratsch, H. Bäumler, A. Voigt and M. Taeger, Macromolecule Depletion at Membranes, Studia Biophysica 130 (1989), 117-122. [40] G. J. Fleer, M. A. Cohen Stuart, J. H. M. H. Scheutjens, T. Cosgrove and B. Vincent, Polymers at Interfaces, Chapman & Hall, New York, 1993. [41] F. J. Nordt, Hemorheology in cerebrovascular diseases: approaches to drug development, Ann. N. Y. Acad. Sci. 416 (1983), 651-661. [42] G. B. Nash, R. B. Wenby, S. O. Sowemimo-Coker and H. J. Meiselman, Influence of cellular properties on red cell aggregation, Clin. Hemorheol. Microcirc. 7 (1987), 93-108.


135

[43] S. O. Sowemimo-Coker, P. Whittingstall, L. Pietsch, R. M. Bauersachs, R. B. Wenby and H. J. Meiselman, Effects of cellular factors on the aggregation behavior of human, rat and bovine erythrocytes, Clin. Hemorheol. Microcirc. 9 (1989), 723-737. [44] H. Schmid-Shönbein, E. Volger and H. J. Klose, Microrheology and light transmission of blood, Pflügers Archive 333 (1972), 140-155. [45] R. M. Bauersachs, R. B. Wenby and H. J. Meiselman, Determination of Specific Red Blood-Cell Aggregation Indexes Via an Automated-System, Clin. Hemorheol. 9 (1989), 1-25. [46] H. J. Meiselman, Red-blood-cell role in RBC aggregation, Clin. Hemorheol. Microcirc. 13 (1993), 575-592. [47] P. Whittingstall, K. Toth, R. B. Wenby and H. J. Meiselman, Cellular factors in RBC aggregation: effects of autologous plasma and various polymers, In: Hemorheologie et Aggregation Erythrocytaire, J. F. Stoltz, Ed., Editions Medicales Internationales, Paris, 1994, pp. 21-30. [48] M. W. Rampling, G. Martin, S. Bignall, R. P. Rivers, T. J. Lissauer and P. C. Bailey, A comparison of the rheologic properties of neonatal and adult blood, Pediatr. Res. 25 (1989), 457-460. [49] M. W. Rampling and P. Whittingstall, A comparison of five methods for estimating red cell aggregation, Klinische Wochenschrift 64 (1986), 1084-1088. [50] O. Linderkamp, P. Ozanne, P. Y. Wu and H. J. Meiselman, Red blood cell aggregation in preterm and term neonates and adults, Pediatr. Res.) 18 (1984), 1356-1360. [51] M. A. Anwar, M. W. Rampling, S. Bignall and R. P. Rivers, The variation with gestational age of the rheological properties of the blood of the new-born, Brit. J. Haematol. 86 (1994), 163-168. [52] P. Whittingstall and H. J. Meiselman, Effect of galactose incubation on the aggregation behavior of density-separated human erythrocytes, In Hemorheologie et Aggregation Erythrocytaire, J. F. Stoltz, M. Donner and A. L. Copley, Eds., Editions Medicales Internationales, New York, 1991, pp. 111-122. [53] J. R. Murphy, Influence of temperature and method of centrifugation on the separation of erythrocytes, J. Lab. Clin. Med. 82 (1973), 334-341. [54] O. Linderkamp, P. Y. Wu and H. J. Meiselman, Deformability of density separated red blood cells in normal newborn infants and adults, Pediatric Research. 16 (1982), 964-968. [55] R. M. Bauersachs, S. J. Shaw, A. Zeidler and H. J. Meiselman, Red blood cell aggregation and blood viscoelasticity in poorly controlled type 2 diabetes mellitus., Clin. Hemorheol. 9 (1989), 935-952. [56] R. B. Ami, G. Barsthein, D. Zelster, Y. Goldberg, I. Shapira, A. Roth, G. Keren, H. Miller, V. Prochorov, A. Eldor, S. Berliner and S. Yedgar, Parameters of RBC aggregation as correlates of the inflammatory state, Am. J. Physiol. 280 (2001), H1982-1988. [57] S. Chen, A. Eldor, G. Barshtein, S. Zhang, A. Goldfarb, E. Rachmilewitz and S. Yedgar, Enhanced aggregability of red blood cells of beta-thalassemia major patients, Am. J. Physiol. 270 (1996), H1951-H1956. [58] A. Koshkaryev, G. Barsthein, A. Nyska, N. Ezov, T. Levin-Harrus, S. Shabat, M. Redlich, T. F. and S. Yedgar, 2-Butoxyethanol enhances the adherence of red blood cells, Arch. Toxicol. 77 (2003), 465469. [59] P. C. M. Obiefuna and D. P. Photiades, Sickle discocytes form more rouleaux in vitro than normal erythrocytes, J. Trop. Med. Hygiene 93 (1990), 210-214. [60] G. Hacioglu, O. Yalcin, M. Bor-Kucukatay, G. Ozkaya and O. K. Baskurt, Red blood cell rheological properties in various rat hypertension models, Clin.Hemorheol.Microcirc. 26 (2002), 27-32. [61] V. Schechner, R. Ben-Ami, T. Hershcovici, S. Yedgar, Y. Beigel, I. Shapira, B. S. and G. Barsthein, Plasma dependent reduction in red blood cell aggregation after dextran sulfate low-density lipoprotein apheresis, Ther. Apheresis Dialysis 9 (2005), 379-384. [62] E. Kayar, F. Mat, H. J. Meiselman and O. K. Baskurt, Red blood cell rheological alterations in a rat model of ischemia-reperfusion injury, Biorheology 38 (2001), 405-414. [63] P. Ozanne, R. B. Francis and H. J. Meiselman, Red blood cell aggregation in nephrotic syndrome, Kidney Int. 23 (1983), 519-525. [64] P. Ozanne, O. Linderkamp, F. C. Miller and H. J. Meiselman, Erythrocyte aggregation during normal pregnancy, Amer. J. Obst. Gyne. 147 (1983), 576-583. [65] R. Gamzu, G. Barsthein, F. Tsipis, J. B. Lessing, A. S. Berliner, M. J. Kupferminc, A. Eldor and S. Yedgar, Pregnancy-induced hypertension is associated with elevation of aggregability of red blood cells, Clin. Hemorheol. Microcirc. 27 (2002), 163-169. [66] O. K. Baskurt, A. Temiz and H. J. Meiselman, Red blood cell aggregation in experimental sepsis, Journal of Laboratory and Clinical Medicine. 130 (1997), 183-190. [67] B. Neu, S. Sowemimo-Coker and H. Meiselman, Cell-Cell Affinity of Senescent Human Erythrocytes, Biophys. J. 85 (2003), 75-84. [68] K. M. Jan and S. Chien, Role of surface electric charge in red blood cell interactions, J. Gen. Physiol. 61 (1973), 638-654.

136


[69] N. Maeda, K. Imaizumi, M. Sekiya and T. Shiga, Rheological characteristics of desialyated erythrocytes in relation to fibrinogen-induced aggregation, Biochim. Biophys. Acta 776 (1984), 151158. [70] K. M. Jan and S. Chien, Role of the electrostatic repulsive force in red cell interactions, Bibl. Anat. 11 (1973), 281-288. [71] M. W. Rampling and M. J. Pearson, Enzymatic degradation of the red cell surface and its effect on rouleaux formation, Clin. Hemorheol. 14 (1994), 531-538. [72] G. B. Nash and H. J. Meiselman, Alteration of red cell membrane viscoelasticity by heat treatment: effect on cell deformability and suspension viscosity, Biorheology 22 (1985), 73-84. [73] A. Rakow, S. Simchon, L. A. Sung and S. Chien, Aggregation of red cells with membrane altered by heat treatment, Biorheology 18 (1981), 3-8. [74] W. H. Reinhart and A. Singh, Erythrocyte aggregation: the roles of cell deformability and geometry, Eur. J. Clin. Invest. 20 (1990), 458-462. [75] W. D. Corry and H. J. Meiselman, Modification of erythrocyte physicochemical properties by millimolar concentrations of glutaraldehyde, Blood Cells 4 (1978), 465-480. [76] R. J. Knox, F. J. Nordt, G. V. F. Seaman and D. E. Brooks, Rheology of erythrocytes suspensions: dextran-mediated aggregation of deformable and non-deformable cells, Biorheology 14 (1977), 75-84. [77] E. Evans and B. Kukan, Free energy potential for aggregation of erythrocytes, Biophys. J. 44 (1983), 255-260. [78] A. Othmane, M. Bitbol, P. Snabre and P. Mills, Influence of altered phospholipid composition of the membrane outer layer on red blood cell aggregation, Eur. Biophys. J. 18 (1990), 93-99. [79] T. L. Fabry, Mechanism of erythrocyte aggregation and sedimentation, Blood 70 (1987), 1572-1576. [80] S. S. Norris, D. D. Allen, T. P. Neff and S. L. Wilkinson, Evaluation of DIDS in the inhibition of rouleaux formation, Transfusion 36 (1996), 109-112. [81] S. O. Sowemimo-Coker and H. J. Meiselman, Effect of procaine hydrochloride on the electrophoretic mobility of human red blood cells, Cell Biophys. 15 (1989), 235-248. [82] S. Ramakrishnan, R. Grebe, M. Singh and H. Schmid-Schobein, Influence of local anesthetics on aggregation and deformability of erythrocytes, Clin. Hemorheol. Microcirc. 20 (1999), 21-26. [83] H. Johnn, C. Phipps, S. C. Gascoyne, C. Hawkey and M. W. Rampling, A comparison of the viscometric properties of the blood from a wide range of mammals, Clin. Hemorheol. 12 (1992), 639647. [84] A. S. Popel, P. C. Johnson, M. V. Kameneva and M. A. Wild, Capacity for red blood cell aggregation is higher in athletic mammalian species than in sedentary species, J. Appl. Physiol. 77 (1994), 17901794. [85] O. K. Baskurt, R. A. Farley and H. J. Meiselman, Erythrocyte aggregation tendency and cellular properties in horse, human, and rat: a comparative study, Am. J. Physiol. 273 (1997), H2604-H2612. [86] H. Bäumler, B. Neu, R. Mitlohner, R. Georgieva, H. J. Meiselman and H. Kiesewetter, Electrophoretic and aggregation behavior of bovine, horse and human red blood cells in plasma and in polymer solutions, Biorheology 38 (2001), 39-51. [87] M. Castellini, R. Elsner, O. K. Baskurt, R. Wenby and H. J. Meiselman, Blood rheology of Weddell seals and bowhead whales, Biorheology 43 (2006), 57-69. [88] G. V. F. Seaman, Electrokinetic behavior of red cells, In: The Red Blood Cell, D. M. Surgenor, Ed., Academic Press, New York, 1975, pp. 1135-1229. [89] D. Lerche, Electrostatic fixed charge distribution in the RBC-glycocalyx and their influence upon the total free interaction energy, Biorheology 21 (1984), 477-492. [90] A. Jones and B. Vincent, Depletion Flocculation in Dispersions of Sterically-Stabilized Particles .2. Modifications to Theory and Further-Studies, Colloids Surfaces 42 (1989), 113-138. [91] G. J. Fleer, J. H. M. H. Scheutjens and B. Vincent, The stability of dispersions of hard spherical particles in the presence of nonadsorbing polymer, In: Polymer Adsorption and Dispersion Stability, E. D. Goddard and B. Vincent, Eds., ACS, New York, 1984, pp. 110-118. [92] B. Neu and H. J. Meiselman, Depletion-mediated red blood cell aggregation in polymer solutions, Biophys. J. 83 (2002), 2482-2490. [93] K. Buxbaum, E. Evans and D. E. Brooks, Quantitation of surface affinities of red blood cells in dextran solutions and plasma, Biochemistry 21 (1982), 3235-3239. [94] E. A. Evans, Structure and deformation properties of red blood cells: concepts and quantitative methods, Methods Enzymol. 173 (1989), 3-35.


137

Mechanical Properties of Leukocytes and Their Effects on the Circulation Roger TRAN-SON-TAYa and Gerard B. NASHb,1 Department of Mechanical & Aerospace Engineering, University of Florida, Gainesville, FL 32611, USA and bCentre for Cardiovascular Sciences, Medical School, University of Birmingham, Birmingham B15 2TT, UK a

Introduction Leukocytes circulate in small numbers compared to red cells and have little effect on bulk blood viscosity except in some leukemias where the ‘leukocrit’ approaches hematocrit. Their main functions are carried out in tissue, and they have evolved specialized adhesive and migratory capabilities to allow recruitment from the blood across vascular endothelium. However, although the cardiovascular system essentially acts as a dispersal system for leukocytes, this does not mean that their mechanical properties are unimportant, or that they cannot influence blood flow. Their slow motion through blood capillaries was recognized early and modern interest in their rheological behavior was spurred by the observations of deformation in human microvessels and glass capillaries made by Bagge, Branemark, Skalak and colleagues [1-3]. It has become evident that these ‘normal’ slow flowing leukocytes can hold up and modify the capillary transit of red cells and influence perfusion and resistance in the microcirculation [4-6]. If perfusion pressures are reduced (e.g., in shock) this slow transit may be reduced to leukostasis [7]. In addition, leukocytes, and especially the most numerous neutrophilic granulocytes, can dramatically change their mechanical properties when ‘activated’ by a variety of agents [8]. Activation at the vessel wall is a necessary part of their physiological migratory response, but if it occurs inappropriately, circulating cells have the potential to cause pathogenic microvascular occlusion [9]. Because of the importance of the mechanical properties of resting and activated leukocytes in the physiology and pathology of the microcirculation, they have been widely studied using rheological techniques. Here we review the theoretical and experimental analyses of leukocyte deformation, and the structural elements that influence the cellular rheology. Physico-chemical factors that influence leukocyte deformation are then described, and the impact of flow resistance on normal and pathological microcirculation is considered. 1 Corresponding author: Centre for Cardiovascular Sciences, Medical School, University of Birmingham, Birmingham B15 2TT, UK; E mail: [email protected]

138

R. Tran-Son-Tay and G.B. Nash / Mechanical Properties of Leukocytes

1. Experimental Methods for Investigating Leukocyte Rheology The ability of a cell to constantly change its shape, flow, and migrate in tissues is dictated by its rheological properties. These properties, arising from its structural characteristics, determine, for instance, the deformation and recovery of the leukocyte passing through the microcirculation. Deformability is a term used to describe the ability of a cell to change its shape in response to an applied force. A very important characteristic of a normal red or white blood cell that allows its deformation is that it has a surface area that is much greater, about 35% and 130% respectively, than that of a sphere of equal volume. Other major determinants of cell deformability include the mechanical, viscoelastic properties of the cell membrane, and intracellular fluid. In this section, we will review the most common methods used to characterize these determinants and analyze the rheological behavior of leukocytes. 1.1. Micropipette Techniques Leukocyte rheology has been studied extensively with the micropipette technique. Two typical types of experiments, aspiration and recovery (Figure 1), are usually performed to determine the mechanical properties of individual white blood cells [1013]. The aspiration method consists of aspirating a leukocyte inside a micropipette with a diameter smaller than that of the leukocyte, and tracking the extension of the cell inside the pipette as a function of time. The evolution of the cell extension is recorded through the acquisition of optical microscopy images as a function of time, aspiration pressure, and micropipette radius. The recovery experiment consists of aspirating the cell inside the micropipette and expelling the cell out of the pipette, and tracking the geometry of the cell as it recovers its initially spherical shape as a function of time. Aspiration

LP (a)

Recovery

Di

Do

D

(b) Wi

W

Figure 1. Micropipette Technique. (a) aspiration, (b) recovery

139


1.1.1. Aspiration Cells are aspirated at a constant pressure into a micropipette (Figure 1a). The length of the aspirated cell, Lp, is measured over time to generate an aspiration curve. Viscosity values, P, can be derived from the slope, dLp/dt, of the aspiration curve [11]: 'P R p

P

(1)

dLp / dt m§¨©1 R1 ·¸¹

where 'P is the aspiration pressure, Rp is the pipet radius, R = R/Rp, R is the radius of the cell outside the pipette, and m=6. 1.1.2. Recovery In this experiment, white blood cells are drawn by a small suction pressure into a micropipette, held there for about 15 seconds, and quickly expelled. The changing length of the cell, L, as it recovers its spherical shape (Figure 1b), is recorded as a function of time, t, and for a liquid drop, is described by a polynomial [12]: L D

o

L

i

D

At B(t )2 C (t )3

(2)

o

where A, B, C are known functions of (Li/Do). Li and Do are the initial deformed length and resting diameter of the cell, respectively.

The variable

t

2t ( P / To ) Do

represents a dimensionless time, where P is the cell viscosity and To is the surface tension of the membrane. From Equation (2), curve fitting of the experimental data provides the ratio

P To

,

and consequently the cell viscosity if the surface tension is known. 1.2. Dynamic Viscoelastic Testing Technique Another popular technique used in the estimation of the rheological properties of cells and their components is the dynamic (i.e., oscillatory) testing method [14]. This method consists of imposing an applied force on a marker (cell or body) and measuring its displacement, or vice versa; the force/indentation relationship is a characterization of the cellular component properties of the cell. Methods based on this principle provide a complex modulus, G* G ' iG" , where G’ and G” are the elastic and viscous components of the modulus G*, and represent, respectively, the energy that is stored and permanently lost in the material during an oscillatory deformation. These methods are usually done under small oscillatory deformations so that the simpler linear viscoelastic theory can be applied. For the dynamic testing techniques, we will focus only on recent methods that use a focal point force to locally indent/probe the surface of a cell, and can provide information on both the mechanical properties of adherent cells and their internal cytoskeleton. Dynamic testing techniques have been used to study a variety of cells.

140


1.3. Magnetic Twisting Cytometry (MTC) This technique probes the cell mechanical properties by binding magnetic microbeads, about 5 Pm in diameter, to the cell membrane with specific cytoskeletal receptors [15]. It consists of oscillating the beads horizontally with an external magnetic field and measuring the resulting bead displacement. 1.4. Atomic Force Microscopy (AFM) This method provides the complex modulus, G*(Z), of single cells by applying oscillatory indentations to the cell with a microtip attached to a cantilever of a known bending constant and then measuring the resulting force [16]. The use of microspheres as cantilever tips allows the application of the simple two-sphere Hertz contact equation, which has few assumptions and provides a direct estimation of G*(Z). The probe tip has a size comparable or smaller to that of a leukocyte and provides a measurement at the cellular or sub-cellular level. For oscillations around an operating indentation, G o , it can be shown that the complex modulus, G*(Z), is equal to [17]: * G (Z )

1X 4 R1/ 2G o1/ 2

ª F (Z ) º iZb (0) » « ¬ G (Z ) ¼

(3)

where Z 2S f is the angular frequency and iZb (0) is the correction for the viscous drag force exerted by the liquid medium on the cantilever. The applied force is F

4E 2

3(1X )

G

R

1/ 2

G

3/ 2

, Q is the Poisson ratio, G is the position of the indentation,

[( z z c ) d ] , d is the cantilever deflection, and

1 R

1 1 r r t n

with rt = radius of the

spherical cantilever tip and rn = neutrophil radius. The contact point zc and the apparent Young’s modulus E are determined by least squares fitting of the Hertzian force-indentation relationship of a spherical punch with infinite stiffness indenting an elastic sphere. Although it is difficult to extract the mechanical properties of the cell with dynamic testing methods, AFM methods have the capability to study passive as well as activated cells. Neutrophils can be maintained in a passive state by plating them on substrates rendered almost nonadhesive through coating with poly(HEMA) [18] and by applying only small deformations. 1.5. Filtration/Cell Transit Analysis The average resistance to flow of large numbers of leukocytes and their tendency to block ‘microvessels’ can be studied using filtration techniques, where suspensions are forced through micropore filters with discrete 5 Pm or 8 Pm diameter pores. Leukocytes present in blood greatly influence flow rate through such filters [19], and studies where filtration of blood was found to change in clinical conditions are likely to have been reporting changes in the number or properties of the leukocytes rather than red cells as was thought. Analysis of purified leukocyte suspensions can give less


141

equivocal information, with measurements based on detecting changes in flow rate for a given pressure, or changes in pressure for constant flow rate. Typically, pressure increases or flow rate decreases, initially at a rapid rate, as cells enter the filter until a steady state may be reached where entry and exit is balanced. However, there is usually a degree of pore clogging which imposes a slower rate of rise in pressure or reduction in flow rate. Early studies demonstrated the poor filterability of leukocytes with different types of leukemia [20], while filtration of purified normal human granulocytes and mononuclear cells suggested that monocytes were highly resistive to micropore flow [21]. In principle, for a known pore number and cell concentration, analysis of filtration rate curves can reveal the concentration of cells unable to transit the pores, and average cell resistance to flow or transit time for other sub-populations [22]. These methods showed, for instance, that transit times through 5 Pm pores were in the order monocytes > granulocytes > lymphocytes, and that purposeful activation of granulocytes greatly increased their flow resistance [23]. The filtration method was later adapted, using filters with only ~30 pores, to measure changes in electrical resistance as individual leukocytes flow through the pores (i.e., the Cell Transit Analyser). The width of the electrical pulses was taken to represent the pore transit time. In this way, distribution of cell transit times could be directly analyzed. While some difficulty arises because of the need to detect and correct for co-incidence when a cell enters a pore when another is still occupied, and because no direct detection of the number of “pore-blocking particles” is possible [22], the CTA does avoid errors due to plugged pores since only events associated with open pores are recorded. This method was able to show that resistance to flow of neutrophils increased within a minute of delivery of an activating stimulus [24], and with separate fast and slow phases upon rapid cooling of these cells [25].

2. Theory of Leukocyte Deformation To date, existing physical models of leukocytes are not capable of quantitatively explaining the wide range of deformation and recovery behaviors observed in experiments. However, relatively simple models have been able to describe the general rheological behavior of leukocytes. We will review the most popular ones here. 2.1. Liquid Drop Model The simplest model to describe the rheological behavior of a cell is the one of a droplet of Newtonian fluid with a constant surface-tension membrane (Figure 2a) equal to about 0.035 dyn/cm [26]. Despite the success of a simple Newtonian fluid model for the cell interior, the cell is seen by electron microscopy to contain a cytoskeleton of variable properties concentrated in a cortical region adjacent to the cell membrane. Cells may have a single nucleus, or multiple nucleoli depending on the type of leukocyte, with a denser structure than the surrounding cytoplasm. However, the simple Newtonian droplet model is capable of capturing many of the characteristic features observed in experiments [26, 27].

142

R. Tran-Son-Tay and G.B. Nash / Mechanical Properties of Leukocytes membrane

T

T P

P

k cytoplasme

(a)

(b)

(T1 ) ( P1 ) ( P2 )

P1 k2

k1

(T2 ) (d)

(c)

Figure 2. Leukocyte Models – Cross Sectional Cut (a) liquid drop with surface tension T, (b) Maxwell liquid drop with surface tension T, (c) standard solid, and (d) compound drop with cellular

and nuclear surface tensions,

T1 and T2 , respectively.

2.2. Non-Newtonian Models Non Newtonian models have also been proposed to describe the elastic and viscous characteristics of the cytoplasm. Maxwell fluid (Figure 2b) [28] and standard linear solid (Figure 2c) [29] models have been developed. Others methods have treated the cytoplasm as a pseudoplastic or power-law fluid [30] and have shown that, under certain circumstances, the cortex exhibits a variable surface tension [31] and that membrane bending can be important in pipettes smaller than 1.2 Pm in diameter [32]. It has been determined that when the dimensionless shear modulus G* t 20 the Maxwell model reduces to the Newtonian fluid model [33]. In this Newtonian limit, the shear modulus has little effect on the time it takes a cell to enter the pipette of minimum radius Rmin. This entry time is given by: t*

0.3(a* ) 0.55 ª¬( R* ) 6.5 1º¼

(4)

where the dimensionless entry time t*, shear modulus G*, cell radius R*, and entrance t ('P 'Pcrit ) Gcell ; G* ; radius of curvature a* are defined as follows: t * 'P 'Pcrit Pcell Rmin a ; and a* , with Pcell as the viscosity of the cytoplasm, and a is the Rcell Rcell entrance radius of curvature. The importance of the entrance geometry, a*, can also be R*


143

seen in the analytical study of Tran-Son-Tay, et al. [34] who considered a droplet entering into a tapered tube. The major problem with the above models is that they do not account for the inhomogeneities in the cell internal structure, and the fact that the cell actively responds to deformation by altering the matrix that contributes to its rigidity [35] and its strength of adhesion to the endothelium [36]. To date, only one continuum model, the compound drop [27, 37, 38], accounts for the presence of a cell nucleus of different rheological properties. 2.3. Compound Drop Based on experimental results, the rheological behavior of leukocytes modeled as a compound drop can explain several main observations and reconcile many of the apparent discrepancies reported in the literature for the values of the cell viscosity. However, because of a lack of information on the rheological properties of the cell substructure and cytoplasmic components, existing models can only describe qualitatively the rheological behavior of leukocytes. The problem considered by Kan et al. [27] consists of the deformation of a compound liquid drop subjected to a uniaxial extensional flow and its subsequent recovery when the imposed flow is stopped. The physical situation corresponds to three incompressible Newtonian fluid layers of density Ui and viscosity P i which occupy the regions :i (i=1,2,3). The regions :̧, :, and : represent, respectively, the suspending cytoplasm, cell membrane, and nucleus as shown in Figure 2d. The surface tension 7 and 7 at the two interfaces are assumed to be constant. Marella and Udaykumar [38] included non Newtonian properties into the Newtonian compound drop model of Kan et al. [27] and treated the membrane as an elastic shell and the cytoplasm as a power law shear thinning fluid. 2.4. Tensegrity Model In addition to the above continuum models, it is worth noting the tensegrity model [39]. This model has been developed in order to study the transmission and distribution of mechanical signals and the conversion of these signals into biological and chemical responses in the cell [40, 41]. The idea of the cellular tensegrity model is that the cell is a tensegrity structure in which microfilaments and intermediate filaments act as tensional elements, and microtubules and extracellular matrix act as compressive elements [41]. One of the key features of the cellular tensegrity model is that the cells are considered as pre-stressed structures. There have been experimental evidence which support the idea of cellular tensegrity model, such as the proportional relationship between the cell stiffness and the cytoskeletal stress [42]. There have also been attempts to construct mathematical models incorporating tensegrity theory. One of the most frequently used models is the cable and strut model which is composed of six struts and twenty-four cables [43]. This model was also used in studying the characteristics of adherent cells by attaching some of the nodes to substrate [44]. Although this six-strut model is an oversimplification of the real cytoskeletal structure, it captures many characteristics of living cells. The tensegrity model has been successfully used on adherent cells. However, its application on suspended cells is very limited, and nonexistent for leukocytes. A

144


potential problem with the tensegrity model is the lack of a temporal dimension; the mechanical properties of the strings and struts do not change with time, so the model can simulate viscoelastic behavior of some cells only at some times and under some conditions [45]. 2.5. Dynamic Viscoelastic Models As mentioned earlier, the dynamic testing method consists of imposing an applied force on a marker and measuring its displacement. Methods based on this principle provide a complex modulus, G* , G* G ' iG" . These methods are usually done under small oscillatory deformations so that the simpler linear viscoelastic theory can be applied. Although these techniques provide valuable information on the viscoelastic properties of the cell and its internal contents, values obtained are usually difficult to relate to the cell intrinsic mechanical properties. 2.6. Liquid Drop Model Fitting liquid drop models to experimental data is not straightforward because the cell is treated as a single body without making a distinction between the cortex and the cell interior. This distinction can be made by considering that the cell cortex exerts a force on the spherical tip given by [46]: Fmembrane

2T S a2 rt

(5)

where a is the contact radius between the tip and the cell, which is related to the indentation G as a 2 G R for small indentations [47]. Neglecting the cortex thickness, the total force on the cantilever is the sum of the forces exerted by the cortex and the cell interior: F

4 Ei 2T R1/ 2G 3 / 2 SG R 3(1 X 2 ) rt

(6)

where Ei is the Young’s modulus of the interior of the cell. It can be shown that, around an indentation point, G o , the total force yields: G* (Z )

Gi* (Z ) kT

(7)

where 1/ 2

kT

S (1 X ) § R · ¨ ¸ T 2rt © G o ¹

(8)

Gi* (Z ) is the complex shear modulus of the cell interior and kT is a frequency-

independent term given by cortical tension. kT depends on the specific measurement conditions and is thus a function of G o rt , and rn .


145

2.7. Standard Viscoelastic Solid Model If the leukocytes are modeled as a whole standard viscoelastic solid body with two spring constants, k1 and k2 and a dashpot P1 (Figure 2c), the complex shear modulus can be expressed as [48]:

G* (Z )

k1

P12 k2Z 2 P k 2Z i 2 1 22 2 2 2 2 k2 P1 Z k2 P1 Z

(9)

2.8. Power Law Structural Damping Model The power law structural damping model yields the following expression for G* (Z ) [15]: § Z · G* (Z ) D [Go ¨ ¸ © )o ¹

x 1

(1 iK )*(2 x) cos

S 2

( x 1) iZP ]

(10)

where D , Go and ) o are scaling factors for geometry, stiffness, and frequency, respectively, K

§ x 1 S · tan ¨ ¸ 2 © ¹

is the structural damping coefficient, (x-1) is the power

law exponent, and iZP is an additional Newtonian viscous term. In the model, x = 1 corresponds to a purely elastic solid, and x = 2 corresponds to a Newtonian liquid. Fits to the data are obtained by minimizing the sum of squares of the differences between the logarithm of G* (Z ) and that of the predictions of the model for each cell. 2.9. Discussion To date, in spite of intensive research efforts, a variety of behaviors of the leukocyte can still not be satisfactorily explained. For example: (i) when a neutrophil is rapidly aspirated into a micropipette with a diameter smaller than the cell diameter, it enters at an initial velocity much faster than the steady-state velocity; (ii) as the rate of deformation increases, the cell viscosity is found to decrease, seemingly due to a shear-thinning effect [11, 30]; (iii) following small but rapid deformations of the cell in less than 0.3 s, a rapid initial recovery during around 0.3 s is obtained, followed by a slower phase [49]; (iv) when the neutrophil is aspirated and held inside the pipette for a short time period of 5 s or less, and then expelled, the cell recovers with an initial rapid recoil [12], and is faster than when held for a longer time; (v) neutrophils that are only slightly deformed in a pipette before ejection yield much smaller values of apparent viscosity than cells that are significantly deformed. This is analogous to the finding that the cell appears to be more viscous when it flows into smaller diameter pipettes [26]. It can be observed that the rate of deformation, the extent of deformation, the excess aspiration pressure, and the holding time all affect the cell recovery characteristics, resulting in different apparent viscosities as shown in Table 1. These curious properties of the cell have been explained by the compound drop model. Kan et al. [27] provide a consistent framework for describing the variety of behaviors, Newtonian as well as non-Newtonian, exhibited by the cells, as summarized in Table 2.

146


The compound liquid drop model has been refined by incorporating an elastic cytoskeleton and a power law, shear thinning cytoplasm [38] but still cannot account for all the reported behaviors. However, the essence of the rheological behavior of leukocytes is shown to result from multiple time scales, with this characteristic well captured by the compound drop model. Table 1: Published Neutrophil Viscosity Values Under Different Flow Conditions Cell Apparent Source Experiments Operating Conditions Viscosity μ (Poise) Dong et al. [50] Aspiration Small deformation 100-300 Recovery Evans and Yeung [26] Aspiration Large deformation 1000-2000 Needham and Hochmuth Aspiration High deformation rate 1000 [11] Low deformation rate 2000 Dong et al. [49] Aspiration High deformation rate 200-2000 Recovery Low deformation 800-1500 Tran-Son-Tay et al. [12] Recovery Large deformation (short holding time) 1000-2000 Large deformation (long holding time) Dong and Skalak [28] Aspiration Large deformation 100-700 Hochmuth et al. [51] Recovery Small deformation 600 Tsai et al. [30] Aspiration Low aspiration pressure 5000 High aspiration pressure 500 Waugh and Tsai [52] Aspiration High deformation rate 1000-1500 Table 2: Explanation of Leukocyte Rheological Behavior Based On The Compound Drop Model Operation Conditions Aspiration experiment: High deformation rate (High aspiration pressure)

Observed Behavior

Compound Drop Model

Low apparent viscosity (Shear-thinning)

Cytoplasmic fluid responds immediately while nucleus deforms gradually.

Low apparent viscosity

Nucleus is only slightly deformed (Resistance from cytoplasm)

Small deformation (Large pipette) High apparent viscosity

Nucleus is highly deformed

Large deformation (Small pipette) Recovery experiment: Short holding time

Fast recoil (Non-Newtonian)

Nucleus does not have time to deform

Long holding time

Slow recovery (Newtonian)

Small deformation

Low apparent viscosity

Large Deformation

High apparent viscosity

Nucleus has time to deform Nucleus is only slightly deformed. Nucleus is highly deformed

3. Relation between Cell Structure and Resistance to Deformation

Studies of cell deformation initially neglected the role of the cell internal contents (e.g., nucleus, microfilaments, microtubules) because of their relatively small size. In these


147

earlier studies, the main resistance to flow was thought to be due to the cytoskeleton with some contribution from the cytoplasm. However, it is now clear that in processes requiring large deformations like in cell extravasation or flow in capillaries, these internal elements can affect the overall cell behavior [53]. The cytoskeleton is composed of three main types of proteins: actin, microtubules, and intermediate filaments. These proteins are involved in the way a cell responds to deformation [54]. When a cell flows into a capillary, all the excess membrane (i.e., microvilli) is pulled out before the cell membrane has to deform and stretch. The cell membrane offers the first resistance to deformation, with the role of the membrane primarily passive when the cell is responding to cytoskeletal deformations [55]. However, the actin cytoskeleton is not a static structure, and undergoes dramatic reorganization when leukocytes are induced to migrate by chemotactic stimuli. Chemotaxis requires continual, directed cycling between monmeric golbular (G-) actin and polymerized, filamentous (F-) actin. In neutrophils, following stimulation there is a rapid early increase in cortical F-actin before a gradual redistribution into pseudopodia over minutes. The intrinsic rigidity of F-actin polymer formed under these circumstances causes marked changes in resistance to deformation as noted below. The cytoplasm consists of the cytosol which is a water solution of enzymes and organic molecules and cellular organelles plus the cell nucleus. Although the cytoplasm is composed mainly of water, it possesses some non-Newtonian characteristics [47]. The complete rheological properties of the cytoplasm are still not known. Methods using a microinjection technique to shoot nanobeads directly into the cytoplasm of living cells and use the thermal motion of the beads to probe the rheological behavior of the cytoplasm have recently been developed toward that purpose [56, 57]. However, these studies have mainly been conducted on adherent cells with a very few on leukocytes [58]. The nucleus is a cellular organelle but is not considered part of the cytoplasm. It is the cell internal component that provides the most resistance to deformation in passive, non activated cells. The importance of the nucleus in cell deformation has been reported by several groups [12, 37, 49, 57, 59]. Because of the non negligible mechanical impact of the cell sub-structures, a wave of research in microrheology at the micro and nano scale levels is emerging and receiving increasing attention [15, 5660]. The application of these relatively new methods to leukocytes is still limited.

4. Physico-Chemical and Inflammatory Factors Influencing Deformation Resistance

A variety of environmental and soluble factors can modify resistance to deformation of leukocytes, with these factors most extensively studied for neutrophils. As mentioned above, resistance depends on cellular ‘activation’ because factors such as chemokines (e.g., interleukin-8) or formylated bacterial peptides (fMLP) cause receptor-mediated modifications in the cytoskeleton, including a shift in the balance from G- to F-actin, and formation of pseudopodia. Even in spontaneously active neutrophils, the pseudopodia themselves are more rigid than the cell body [61]. Purposeful addition of agents such as fMLP magnifies this effect as judged by aspiration of small portions of cells [62] or filtration studies [8, 24]. Exposure to water-soluble components of cigarette smoke causes a comparable increase in rigidity in neutrophils that is

148


associated with more disorganized changes in F-actin content and cell shape [63]. Interestingly, in vivo, smoking of cigarettes is associated with an essentially immediate increase in the transit time of neutrophils flowing through the lungs [64]. These ‘activation’ responses can be inhibited by the presence of cytochalasins, which bind the ‘barbed’ end of growing actin filaments and shift the balance back from F- to G-actin. Indeed, even in unstimulated neutrophils, cytochalasin-treament reduces structural rigidity [65]. Changes in osmolarity, pH and temperature all influence neutrophil rheology as judged from analysis of small scale deformation when short tongues were aspirated into a micropipette [66]. Hypertonic or alkali media both tend to increase structural viscoelastic moduli, while hypotonic or acid media have relatively little effect. Cooling causes an increase in the viscous component of resistance [66]; interestingly, at around 10 °C, there is an anomalous increase in resistance to flow of whole cells through pores (e.g., greater than that at 0°C) [25, 67]. This is inhibited by cytochalasin treatment, implying a structural change in the actin cytoskeleton. Cooling to 10°C followed by rewarming has been investigated in order to simulate what might happen as cells moved from cold skin to warm core [68]: on rewarming to 37°C, cells remained relatively poorly deformable for a brief period, but recovered within ~5 minutes.

5. Leukocyte Rheology and the Microcirculation - Physiology and Pathology

A variety of intravital studies have observed the passage of leukocytes through the microcirculation, and remarked on their erratic transit through capillaries, markedly slower rates of deformation than red cells, and tendency to hold up red cells [1, 4]. Direct microscopic and indirect observations based on analysis of flow resistance or lack of outflow have also shown that if perfusion pressure is reduced as in cases of shock or arterial occlusion, then leukocytes tend to become stuck in the microcirculation and may fail to be washed out upon return of pressure (i.e., the noreflow phenomenon) [7, 69, 70]. Even in normally perfused microvascular beds, leukocytes make a significant contribution to the overall flow resistance [5], and it is thought that if they become activated they can have local occlusive effects [71]. In addition to the effects arising from their mechanical rigidity, it is also worth pointing out that leukocytes adhering to the walls of post-capillary venules (e.g., during inflammation) will effectively increase flow resistance by restricting without fully occluding the lumen. This cannot necessarily be considered an abnormal phenomenon, although uncontrolled deposition of leukocytes which fail to clear from the vessel might be pathogenic. Roles for abnormal leukocyte rheology in microvascular pathology were first demonstrated in animal models in the context of myocardial ischemia induced by arterial ligation and hemorrhagic shock [7, 72]. Deposition of granulocytes was demonstrated, with failure to clear upon return of flow/pressure. Note that mechanical effects may have been allied with increased adhesion in the previously ischemic tissue, and it is thus difficult to definitively separate the two. In a number of studies, changes in granulocyte deformability after activation by endotoxin or chemoattractants were linked with observations on microvascular occlusions in the lungs induced by the activated cells [73-75]. These studies suggested that increased rigidity was a primary factor in trapping cells in capillaries, although changes in adhesive behavior also play a role in retention [76]. In humans, a series of studies showed that the flow properties of


149

circulating neutrophils were impaired in patients with critical ischemia of the leg, intermittent claudication following exercise to pain, and those recovering from myocardial infarction or stroke [9, 77-79]. These responses may have been secondary to the ischemic events, but nevertheless have the potential to contribute to the progress or outcome of disease. Changes in granulocyte rheology can be induced by autoantibodies, termed anti-neutrophil cytoplasm antibodies (ANCA), and linked with the pathology of small-vessel vasculitis [80]. ANCA cause an increase in cell adhesiveness as well as rigidity, and pathology in the lungs and kidneys might be influenced by both responses, which again are difficult to separate in vivo. In summary, the structure of leukocytes causes them to be much more resistive to deformation than red cells. Consequently, leukocytes flow relatively slowly through capillaries, and can hold up and influence the distribution of flow of red cells. Moreover, the requirement for leukocytes to migrate into tissue means that they are able to actively modify their structure and consequently their mechanical properties. Migration is normally a tightly regulated response only happening at the wall of inflamed venules or in lymphatic tissue, but if this process is uncontrolled, or there is spill-over of activating agents or activated cells into the systemic vasculature, then problems may ensue. Two main scenarios can be envisaged: reduction in perfusion pressure (e.g., in vascular occlusive ischemia or shock) may cause initially resting leukocytes to be trapped in capillaries. In this situation the cells may become activated in the poorly perfused region, change their mechanical and adhesive properties, and fail to wash out upon reperfusion or recovery of pressure. Alternatively, agents or activated cells released from inflamed or traumatized tissue, or presence of pathogenic agents in blood (e.g., bacterial toxins or autoantibodies) may cause systemic changes in rheology of leukocytes, which causes trapping in vulnerable tissue such as the lungs. Thus, understanding of the structural basis of leukocyte rheology and its regulation gives important insight into the physiological and pathological responses of these cells.

References U. Bagge, and P.I. Branemark, White blood cell rheology: an intravital study in man, Adv. Microcirc. 7 (1977), 1-17. [2] U. Bagge, R. Skalak and R. Attefors, Granulocyte rheology: experimental studies in an in vitro microflow system, Adv. Microcirc. 7 (1977), 29-48. [3] G.W. Schmid-Schonbein, K.L. Sung, H. Tozeren, R. Skalak and S. Chien, Passive mechanical properties of human leukocytes, Biophys. J. 36 (1981), 243-256. [4] G.W. Schmid-Schonbein, S. Usami, R. Skalak and S. Chien, The interaction of leukocytes and erythrocytes in capillary and postcapillary vessels, Microvasc. Res.19 (1980), 45-70. [5] D.W. Sutton and G.W. Schmid-Schonbein, Elevation of organ resistance due to leukocyte perfusion, Am. J. Physiol. 262 (1992), H1646-H1650. [6] U. Nobis, A.R. Pries, G.R. Cokelet and P. Gaehtgens, Radial distribution of white cells during blood flow in small tubes, Microvasc. Res. 29 (1985), 295-304. [7] U. Bagge, B. Amundson and C. Lauritzen, White blood cell deformability and plugging of skeletal muscle capillaries in hemorrhagic shock, Acta Physiol. Scand. 108 (1980), 159-163. [8] S.M. Buttrum, E.M. Drost, W. MacNee, E. Goffin, C.M. Lockwood, R. Hatton and G.B. Nash, Rheological response of neutrophils to different types of stimulation, J. Appl. Physiol. 77 (1994), 18011810. [9] Nash, G.B., P.R. Thomas & J.A. Dormandy. (1988). Abnormal flow properties of white blood cells in patients with severe ischaemia of the leg, Br. Med. J. 296, 1699-1701. [10] E. Evans and B. Kukan, Passive material behavior of granulocytes based on large deformation and recovery after deformation tests, Blood 64 (1984), 1028-1035. [1]

150


[11] D. Needham and R.M. Hochmuth, Rapid flow of passive neutrophils into a 4 microns pipet and measurement of cytoplasmic viscosity, J. Biomech. Eng. 112 (1990), 269-276. [12] R. Tran-Son-Tay, D. Needham, A. Yeung and R.M. Hochmuth, Time-dependent recovery of passive neutrophils after large deformation, Biophys. J.60 (1991), 856-866. [13] R. Tran-Son-Tay, H.P. Ting-Beall, D.V. Zhelev and R.M. Hochmuth, Viscous behavior of leukocytes. In: Cell Mechanics and Cellular Engineering, V.C. Mow, F. Guilak, R. Tran-Son-Tay and R. M. Hochmuth, Eds., Springer-Verlag, New York, 1994. [14] R. Tran-Son-Tay, G. B. Nash and H. J. Meiselman, Oscillatory viscometry of red blood cell suspensions: relations to cellular viscoelastic properties, J. Rheol. 30 (1986), 231-249. [15] B. Fabry, G.N. Maksym, J.P. Butler, M. Glogauer, D. Navajas and N.A. Taback, E.J. Millet and J.J. Fredberg, Time scale and other invariants of integrative mechanical behavior in living cells, Phys. Rev. E, Stat.Nonlin.Soft Mat. Phys. 68 (2003), 041914. [16] J. Alcaraz, L. Buscemi, M. Grabulosa, X. Trepat, B. Fabry and R. Farre, D. Navajas, Microrheology of human lung epithelial cells measured by atomic force microscopy, Biophys. J. 84 (2003), 2071-2079. [17] P. Roca-Cusachs, I. Almendros, R. Sunyer, N. Gavara, R. Farre and D. Navajas, Rheology of passive and adhesion-activated neutrophils probed by atomic force microscopy, Biophys. J. 91 (2006), 35083518. [18] J. Folkman and A. Moscona, Role of cell shape in growth control, Nature 273 (1978), 345-349. [19] S. Chien, E.A. Schmalzer, M.M. Lee, T. Impelluso and R. Skalak, Role of white blood cells in filtration of blood cell suspensions, Biorheology 20 (1983), 11-27. [20] M.A. Lightman, Rheology of leukocytes, leukocyte suspensions, and blood in leukemia. Possible relationship to clinical manifestations, J. Clin. Invest. 52 (1973), 350-358. [21] E.A. Schmalzer and S. Chien, Filterability of subpopulations of leukocytes: effect of pentoxifylline, Blood 64 (1984), 542-546. [22] G.B. Nash, J.G. Jones, J. Mikita and J.A. Dormandy, Methods and theory for analysis of flow of white cell subpopulations through micropore filters, Brit. J. Haematol. 70 (1988), 165-170 [23] G.B. Nash, J.G. Jones, J. Mikita, B. Christopher and J.A. Dormandy, Effects of preparative procedures and of cell activation on flow of white cells through micropore filters, Br. J. Haematol. 70 (1988), 171176. [24] Z. Pecsvarady, T.C. Fisher, A. Fabok, T.D. Coates and H.J. Meiselman, Kinetics of granulocyte deformability following exposure to chemotactic stimuli, Blood Cells 18 (1992), 333-352; discussion 353-358. [25] G.B. Nash, K.B. Abbitt, K. Tate, K.A. Jetha and S. Egginton, Changes in the mechanical and adhesive behaviour of human neutrophils on cooling in vitro, Eur. J. Physiol. 442 (2001), 762-770. [26] E. Evans and A. Yeung, Apparent viscosity and cortical tension of blood granulocytes determined by micropipet aspiration, Biophys. J. 56 (1989), 151-160. [27] H. Kan, H.S. Udaykumar, W. Shyy and R. Tran-Son-Tay, Hydrodynamics of a compound drop with application to leukocyte modeling, Phys. Fluid. 10 (1998), 760-774. [28] C. Dong, C and R. Skalak, Leukocyte deformability: finite element modeling of large viscoelastic deformation, J. Theor. Biol. 158 (1992), 173-193. [29] G.W. Schmid-Schonbein, K.L. Sung, H. Tozeren, R. Skalak and S. Chien, Passive mechanical properties of human leukocytes. Biophys. J. 36 (1981), 243-256. [30] M.A. Tsai, R.S. Frank and R.E. Waugh, Passive mechanical behavior of human neutrophils: power-law fluid, Biophys. J. 65 (1993), 2078-2088. [31] D. Needham and R.M. Hochmuth, A sensitive measure of surface stress in the resting neutrophil. Biophys. J. 61 (1992), 1664–70. [32] D.V. Zhelev, D. Needham and R.M. Hochmuth, Role of the membrane cortex in neutrophil deformation in small pipets, Biophys. J. 67 (1994), 696-705. [33] M. Bathe, A. Shirai, C.M. Doerschuk and R.D. Kamm, Neutrophil transit times through pulmonary capillaries: the effects of capillary geometry and fMLP-stimulation, Biophys. J. 83 (2002), 1917-1933. [34] R. Tran-Son-Tay, T.F. Kirk 3rd, D.V. Zhelev and R.M. Hochmuth, Numerical simulation of the flow of highly viscous drops down a tapered tube, J. Biomech. Eng. 116 (1994), 172-177. [35] G.S. Worthen, B. Schwab 3rd, E.L. Elson and G.P. Downey, Mechanics of stimulated neutrophils: cell stiffening induces retention in capillaries, Science 245 (1989), 183-186. [36] Y. Kitagawa, S.F. Van Eeden, D.M. Redenbach, M. Daya, B.A. Walker and M.E. Klut, B.R. Wiggs, J.C. Hogg, Effect of mechanical deformation on structure and function of polymorphonuclear leukocytes, J. Appl. Physiol.82 (1997), 1397-1405. [37] R. Tran-Son-Tay, H.C. Kan, H.S. Udaykumar, E. Damay and W. Shyy, Rheological modelling of leukocytes, Med. Biol. Eng. Comput. 36 (1998), 246-250. [38] S.V. Marella and H.S. Udaykumar, Computational Analysis of the deformability of leukocytes modeled with viscous and elastic structural components, Phys. Fluid. 16 (2004), 244-264.


151

[39] D.E. Ingber, Cellular tensegrity: defining new rules of biological design that govern the cytoskeleton, J. Cell Sci. 104 (1993), 613-627. [40] N. Wang, J.P. Butler and D.E. Ingber, Mechanotransduction across cell surface and through the cytoskeleton, Science 26 (1993), 1124-1127. [41] D.E. Ingber, Tensegrity I. Cell structure and hierarchical systems biology, J. Cell Sci. 116 (2003), 11571173. [42] N. Wang, K. Naruse, D. Stamenovic, J.J. Fredberg, S.M. Mijailovich, I.M. Tolic-Norrelykke, T. Polte, R. Mannix and D.E. Ingber, Mechanical behavior in living cells consistent with the tensegrity model, Proc. Natl. Acad. Sci. USA 98 (2001), 7765-7770. [43] D. Stamenovic, J.J. Fredberg, N. Wang, J.P. Butler and D.E. Ingber, A microstructural approach to cytoskeletal mechanics based on tensegrity, J. Theor. Biol. 181 (1996), 125-136. [44] M.F. Coughlin and D. Stamenovic, A tensegrity model of the cytoskeleton in spread and round cells, J. Biomech. Eng. 120 (1998), 770-777. [45] D.E. Ingber, S.R. Heidemann, P. Lamoroux and R.E. Buxbaum, Opposing views on tensegrity as a structural framework for understanding cell mechanics, J. Appl. Physiol. 89 (2000), 1663-1670. [46] E.B. Lomakina, C. M. Spillmann, M. R. King and R. E. Waugh, Rheological analysis and measurement of neutrophil indentation, Biophys. J. 87 (2004), 4246–4258. [47] K.L. Johnson, Contact mechanics, Cambridge University Press, Cambridge, 1985. [48] Y.C. Fung, Biomechanics, Springer-Verlag, New York, 1993. [49] C.R. Dong, R. Skalak and K.L. Sung, Cytoplasmic rheology of passive neutrophils, Biorheology 28 (1991), 557-567. [50] C. Dong, R. Skalak, K.-L. P. Sung, G. W. Schmid-Schnbein and S. Chien, Passive deformation analysis of human leukocytes, J. Biomech. Eng. 110 (1988), 27-36. [51] R.M. Hochmuth, H. P. Ting-Beall, B. B. Beaty, D. Needham and R. Tran-Son-Tay, Viscosity of Passive Neutrophils Undergoing Small Deformations, Biophys. J. 64 (1993), 1596-1601. [52] R.E. Waugh and M.A. Tsai, Shear rate dependence of leukocyte cytoplasmic viscosity. In: Cell Mechanics and Cellular Engineering, V.C. Mow, F. Guilak, R. Tran-Son-Tay and R. M. Hochmuth, Eds., Springer-Verlag, New York, 1994. [53] H.-C. Kan, H.S. Udaykumar, W. Shyy, P. Vigneron and R. Tran-Son-Tay, Effects of nucleus on leukocyte recovery, Ann. Biomed. Eng. 27 (1999), 648-655. [54] M. Radmacher, M. Fritz, C.M. Kacher, J.P. Cleveland and P.K. Hansma, Measuring the viscoelastic properties of human platelets with the atomic force microscope, Biophys. J. 70 (1996), 556-567. [55] M.P. Sheetz, Glycoprotein motility and dynamic domains in fluid plasma membranes, Ann. Rev. Biophys. Biomolec. Struc. 22 (1993), 417-431. [56] D. Weihs, T.G. Mason and M.A. Teitell, Bio-microrheology: a frontier in microrheology, Biophys. J. 91 (2006), 4296-4305. [57] Y. Tseng, J. S. H. Lee, T. P. Kole, I. Jiang and D. Wirtz, Micro-organization and visco-elasticity of the interphase nucleus revealed by particle nanotracking, J. Cell Sci. 17 (2004), 2159-2167. [58] M. Yanai, J. P. Butler, T. Suzuki, H. Sasaki and H. Higuchi, Regional rheological differences in locomoting neutrophils, Am. J. Physiol. 287 (2004), C603-C611. [59] K.N. Dahl, A.J. Engler, J.D. Pajerowski and D.E. Discher, Power-law rheology of isolated nuclei with deformation mapping of nuclear sub-structures, Biophys. J. 89 (2005), 2855-2864. [60] E. Canetta, A. Duperray, A. Leyrat and C. Verdier, Measuring cell viscoelastic properties using a forcespectrometer: Influence of protein-cytoplasm interactions, Biorheology 42 (2005), 321-333. [61] S. Chien, G.W. Schmid-Schonbein, K.L. Sung, E.A. Schmalzer and R. Skalak, Viscoelastic properties of leukocytes, Kroc Foundation Series 16 (1984), 19-51. [62] R.S. Frank, Time-dependent alterations in the deformability of human neutrophils in response to chemotactic activation, Blood 76 (1990), 2606-2612. [63] E.M. Drost, C. Selby, S. Lannan, G.D. Lowe and W. MacNee, Changes in neutrophil deformability following in vitro smoke exposure: mechanism and protection, Am. J. Respir. Cell Molec. Biol. 6 (1992), 287-295. [64] W. MacNee, B. Wiggs, A.S. Belzberg and J.C. Hogg, The effect of cigarette smoking on neutrophil kinetics in human lungs, New Eng. J. Med. 321 (1989), 924-928. [65] M.A. Tsai, R.S. Frank and R.E. Waugh, Passive mechanical behavior of human neutrophils: effect of cytochalasin B, Biophys. J. 66 (1994), 2166-2172. [66] K.L. Sung, G.W. Schmid-Schonbein, R. Skalak, G.B. Schuessler, S. Usami and S. Chien, Influence of physicochemical factors on rheology of human neutrophils, Biophys. J. 39 (1982), 101-106. [67] R.A. Adams, S.A. Evans, F. Kooshesh and J.G. Jones, The effects of temperature on the filtration of diluted blood through 3 μm and 5 μm filters, Biorheology 32 (1995), 643-653. [68] K.A. Jetha, S. Egginton and G.B. Nash, Increased resistance of neutrophils to deformation upon cooling and rate of recovery on rewarming, Biorheology 40 (2003), 567-576.

152


[69] M. Braide, B. Amundson, S. Chien and U. Bagge, Quantitative studies on the influence of leukocytes on the vascular resistance in a skeletal muscle preparation, Microvasc. Res. 27 (1984), 331-352. [70] M. Braide, A. Blixt and U. Bagge, Leukocyte effects on the vascular resistance and glomerular filtration of the isolated rat kidney at normal and low flow states, Circ. Shock 20 (1986), 71-80. [71] D.H. Adams and G.B. Nash, Disturbance of leucocyte circulation and adhesion to the endothelium as factors in circulatory pathology, Br. J. Anaesth. 77 (1996), 17-31. [72] R.L. Engler, G.W. Schmid-Schonbein and R.S. Pavelec, Leukocyte capillary plugging in myocardial ischemia and reperfusion in the dog, Am. J. Pathol. 111 (1983), 98-111. [73] C.M. Doerschuk, M.F. Allard and J.C. Hogg, Neutrophil kinetics in rabbits during infusion of zymosanactivated plasma, J. Appl. Physiol. 67 (1989), 88-95. [74] S.C. Erzurum, G.P. Downey, D.E. Doherty, B. Schwab 3rd, E.L. Elson and G.S. Worthen, Mechanisms of lipopolysaccharide-induced neutrophil retention. Relative contributions of adhesive and cellular mechanical properties, J. Immunol. 149 (1992), 154-162. [75] G.P. Downey, G.S. Worthen, P.M. Henson and D.M. Hyde, Neutrophil sequestration and migration in localized pulmonary inflammation. Capillary localization and migration across the interalveolar septum, Am. Rev. Resp. Dis. 147 (1993), 168-176. [76] C.M. Doerschuk, Mechanisms of leukocyte sequestration in inflamed lungs. Microcirculation 8 (2001), 71-88. [77] G.B. Nash, B. Christopher, A.J. Morris and J.A. Dormandy, Changes in the flow properties of white blood cells after acute myocardial infarction, Br. Heart J. 62 (1989), 329-334. [78] G. Ciuffetti, R. Balendra, S.E. Lennie, J. Anderson and G.D. Lowe, Impaired filterability of white cells in acute cerebral infarction, Br. Med. J. 298 (1989), 930-931. [79] F.J. Neumann, W. Waas, C. Diehm, T. Weiss, H.M. Haupt and R. Zimmermann, H. Tillmanns, W. Kubler, Activation and decreased deformability of neutrophils after intermittent claudication, Circulation 82 (1990), 922-929. [80] W.Y. Tse, G.B. Nash, P. Hewins, C.O. Savage and D. Adu, ANCA-induced neutrophil F-actin polymerization: implications for microvascular inflammation, Kidney Int. 67 (2005), 130-139.


153

Adhesion of Circulating Leukocytes and Platelets to the Vessel Wall Susan L. CRANMERa and Gerard B. NASHb,1 Australian Centre for Blood Diseases, Monash University, Melbourne VIC 3004, Australia, and bCentre for Cardiovascular Sciences, Medical School, University of Birmingham, Birmingham B15 2TT, UK a

Introduction: Leukocyte and Platelet Adhesion as Rheological Phenomena Circulating leukocytes and platelets must adhere to the wall of blood vessels in order to carry out the protective function of immunity for leukocytes and hemostasis for platelets: failure or lack of control in recruitment can be pathogenic. Given the importance of these adhesive processes, it is not surprising that they have been widely studied both in vivo using intravital microscopy and in vitro using flow-based models. It has become increasingly recognized that adhesion is constrained by the local hemodynamic environment and modulated by the rheological properties of the blood. The rate of motion of the cells before capture and the shear forces acting on them during adhesion critically control the efficiency of attachment. The rheology of the blood influences these hemodynamic parameters. It also affects the efficiency with which cells are brought into contact with the wall because margination in the flow depends on the concentration of the red cells and their flow-dependent tendency to aggregate [1]. Thus, physiological and pathological mechanisms of leukocyte and platelet adhesion represent important rheological phenomena requiring understanding at the biomechanical as well as molecular-biological levels. Flow-based studies have revealed that in each case, leukocytes and platelets use a multi-step process to achieve controlled recruitment [2, 3]. Broadly, specialized receptors support capture of fast-moving cells, separate receptors support stable attachment, and activating signals are required for transition between the states. The actual substrates and receptors differ; leukocytes usually adhere to intact endothelium while platelets typically adhere to sub-endothelial matrix exposed in damaged vessels. In the case of leukocytes, attachment is followed by migration through the endothelium, while platelets undergo spreading and an aggregation phase, and act as a surface for coagulation and fibrin deposition. Leukocyte adhesion is mainly restricted to post capillary venules where shear rates and stresses are relatively low. However, platelet adhesion is possible in all vessels in order to inhibit blood loss, and can occur in arteries at much higher shear rates and stresses. Clearly, there are interesting parallels but important distinctions between the behaviors of the two cell types. 1 Corresponding author: Centre for Cardiovascular Sciences, Medical School, University of Birmingham, Birmingham B15 2TT, UK; E mail: [email protected]

154 S.L. Cranmer and G.B. Nash / Adhesion of Circulating Leukocytes and Platelets to the Vessel Wall

Here we review basic concepts of dynamic cellular adhesion and experimental approaches to their investigation that are largely common between the platelets and leukocytes. The specific mechanisms underlying adhesion in the vascular system are then described for the different cells, and the major physico-chemical modulators of adhesion are outlined. Finally, ways in which abnormal adhesive responses can contribute to pathology are considered with particular attention to how these might influence blood circulation.

1. Concepts in Dynamic Cell Adhesion The adhesive interactions between leukocyte or platelets and endothelium have evolved to solve the problems of ensuring that the circulating fast-moving cells can be specifically captured and immobilized in a controlled manner at the desired site. The typical flow velocities of cells near the wall are in the mm/s range, and are thus “large” versus the scale of the cells and of the receptors and ligands that must interact. Effectively, this means that there is little time for bond formation between individual pairs. Moreover, once bonds are formed, the cell is essentially immobilized, and the shear forces exerted by the continuing flow of fluid will pull on the bonds and tend to disrupt them. Other cells in the body do not require such rapid bond kinetics and can form stable interactions involving many bonds over time. At the early stage of capture, small numbers of bonds will be generated, and a probabilistic description, considering the properties of the individual receptor-ligands, is needed to explain phenomena observed [4-9]. The likelihood of initial bond formation will depend on the density of receptors and ligands, the forward rate constant for the interaction and the velocity of the flowing cells (Figure 1). The velocity in turn depends on the wall shear rate and cell size. The survival of a formed bond will depend on its reverse rate constant, the sensitivity of this rate to applied force (i.e., reactive compliance), and the force applied to the cells as determined by the wall shear stress and their projected surface area (Figure 1). Specialized receptors and ligands have evolved for leukocytes and platelets (e.g., selectins and the glycoprotein (GP) Ib/V/IX complex, respectively - see below) that have sufficiently rapid kinetics to allow efficient capture. Interestingly, these bonds also have rapid reverse rate constants, so that the cells roll forward repeatedly making and breaking bonds [10, 11]. This motion can be quite jerky, especially when there are low densities of receptors on the surface, and cells may intermittently detach and re-attach. Separate receptors are used to stabilize the attachment, and transition from one state to the other occurs through ”activation” of members of the integrin family on the leukocytes or platelets. This requires signals from chemical mediators or from other prior-engaged surface receptors. The binding kinetics of these receptors are slower and they are not adequate for capture except at low shear rates [12]. They are, however, more long-lived, and typically the adherent cells undergo morphological changes which increase surface contact area, allow increased integrin engagement, and further stabilize the adhesion. Such cells are well-able to withstand the fluid shear forces in the circulation without detachment. Since the capture receptors and activators are exposed in a controlled manner only at the specific inflammatory or haemostatic site, and integrin activation occurs rapidly after signaling, the entire adhesive process can be confined to a restricted area. The rolling phase enables more efficient contact with activators and slows the cells

S.L. Cranmer and G.B. Nash / Adhesion of Circulating Leukocytes and Platelets to the Vessel Wall 155

sufficiently to allow integrin engagement with ligands. Nevertheless, the adhesion may only occur under a restricted range of shear rates or stresses, and leukocytes are usually only able to bind in post-capillary venules where these parameters have their lowest values in the microcirculation [13]. Here leukocytes have velocities on the order of mm/s and the wall shear stress is ~0.1 to 1.0 Pa. Platelets are able to adhere in all vessels, which may be due in part at least to their smaller size which is about ¼ of the diameter of leukocytes. For given flow conditions, their velocity at the wall will be slower and the shear drag exerted upon them markedly reduced compared to leukocytes.

Figure 1. Schematic of events as a leukocyte (upper panels) or platelet (lower panels) initially binds to the vessel wall. The local wall shear rate Jw and cell radius r determine velocity before attachment, while the wall shear stress tw determines the force applied to the cells and hence to the initial bonds formed immediately after capture [14]. Cells tend to roll erratically forward as unstable bonds are formed and broken. PSGL1 = P-selectin glycoprotein ligand 1; vWF = von Willebrand factor.

The other rheological factor to be considered is the flow behavior of the blood in which leukocytes are outnumbered by ~1000:1 by red cells and platelets at about 10:1. Contact of the less numerous cells with the wall is promoted by their margination in the blood stream. Margination is enhanced for leukocytes by slower flow and red cell aggregation, but in the case of platelets, the trends are in the opposite direction [15, 16]. In the case of leukocytes, it appears that red cells not only influence margination but their presence also stabilizes adhesion, even though it raises shear stress [17]. This may occur through their application of a force normal to the wall, but the mechanisms underlying both phenomena in blood remain uncertain.

2. Experimental Investigation of Dynamic Adhesion of Leukocytes and Platelets 2.1. Intravital Microscopy Quantitative analysis of the adhesion of flowing leukocytes in the modern age started with the studies of Atherton and Born [13], who characterized rolling adhesion in


post-capillary venules. Numerous intravital studies have followed, typically examining exteriorized tissues such as the mesentery or cremaster muscle in rabbits, rats and, increasingly, mice, or using inserted skin windows. In general, the trauma of surgery induces a level of adhesion, but specific agonists such as tumor necrosis factor or chemotactic agents can be used to increase recruitment. These analyses have demonstrated molecular mechanisms underlying adhesion, and the dependence of adhesion on rheological factors [18, 19]. It is difficult to vary hemodynamic conditions controllably in vivo, although, for example, the effects of modulating red cell aggregation have been analyzed [20]. In the case of platelet adhesion, intravital microscopic methods have also been used, but in combination with various techniques employed to generate injury to the vessel wall. One of the earliest experimental thrombosis models, used in combination with intravital videomicroscopy, was simply to puncture the wall of mesenteric vessels in the rabbit using a glass micropipette [21], thereby allowing visualization of thrombus formation and embolization. Since then, several more refined experimental models have been developed and have been reviewed elsewhere [22, 23]. Among the most commonly used methods are photochemical induced oxidative injury, chemical injury (e.g., most often ferric chloride) causing severe endothelial damage, and mechanical or electrical trauma causing direct endothelial injury. More recently, the development of a more spatially and temporally defined laser-induced injury model, combined with realtime confocal imaging, has begun to provide an increasingly detailed analysis of the mechanisms involved in the initiation and promotion of thrombus formation [24]. It should be noted, however, that it is unlikely that any of the currently used models exactly recapitulate the natural process of thrombus formation. Therefore, some caution should be exercised in the interpretation of results from the various in vivo models: the type and extent of injury are likely to vary significantly, and different areas of the circulation may also respond to injury in different ways and rely on different adhesive proteins and signaling mechanisms.

2.2. In Vitro Rheological Analysis Using Flow Systems Since leukocyte and platelet adhesion both occur under flow conditions in vivo, and since the activation of these cells may be influenced by fluid flow, flow-based perfusion chambers have been developed to investigate the mechanisms contributing to these processes in vitro and ex vivo (Figure 2). These experimental systems represent simplified models compared to the in vivo environment, but are essential in allowing the specific roles of various receptors, ligands and other vascular components to be investigated. The approaches used to investigate cell adhesion under dynamic flow conditions have become increasingly sophisticated in recent years, with the combination of well established flow models with high resolution, real-time confocal or multiphoton imaging techniques providing increased insight into the dynamics and mechanisms contributing to cell adhesion under flow. Probably the most widely used system for leukocytes and platelets has been the parallel-plate perfusion chamber, developed by Sakariassen and colleagues in the 1980’s to study hemostasis and thrombosis in flowing whole blood [25] and McIntire and colleagues for studies of red cell as well as white cell adhesion [26]. This rectangular flow chamber can be used to measure adhesion at different shear rates by changing either the chamber height or the flow rate, and microscopic visualization of


adhering cells is easily achievable owing to the transparent parallel plate geometry of the chamber (see Figure 2). One of the main advantages provided by this system over earlier ex vivo annular perfusion chambers is the ability to prepare a wide range of different adherent surfaces, including purified extracellular matrix proteins (e.g., collagen, vWf) and adherent cells (e.g., endothelium, platelets), which can then be exposed to flowing cells. Leukocytes and platelets in flowing whole blood can also be visualized using fluorescence microscopy with no interference from red blood cells since hemoglobin quenches the fluorescence from commonly used fluorescent markers which include calcein, rhodamine 6G, acridine orange and mepacrine.

Figure 2. Illustration of flow systems used for analysis of leukocyte or platelet adhesion in vitro. A). Prefabricated glass capillary ‘microslide’. B). Parallel-plate flow-chamber. C). Schematic of perfusion system. D). Fluorescently-labeled platelets adherent individually and forming aggregates during deposition on fibrillar collagen from flowing blood. E). Fluorescently-labeled leukocytes rolling on P-selectin during deposition from flowing blood. F. Isolated neutrophils adhering and migrating on cultured endothelial cells which were pre-treated with tumor necrosis factor-D.

Various modifications to this system have been developed to generate flow patterns that deviate from simple laminar flow; the goal of these modifications is to model the more complex flow conditions that occur in vivo such as at junctions or sharp bends in arteries and at sites of atheroma or vessel wall injury. An abrupt


expansion of a cylindrical capillary or a stepped flow chamber have been used to examine the promotion of platelet or leukocyte adhesion under disturbed flow conditions mimicking arterial zones of flow recirculation [27]. Another modification was introduced by Barstad and colleagues [28] where various eccentric stenoses coated with collagen were introduced into the flow chamber to examine thrombus formation at the apex of a stenosis where shear rates reached 32,000 s-1. There has also been significant progress in the development of clinical flow devices, particularly for the assessment of platelet function and the effects of potential anti-thrombotic drugs (see Zwaginga, et al. [29]). These include the PFA-100 platelet function analyzer by Dade Behring, where a citrated blood sample is perfused at a defined flow rate and therefore shear rate through an aperture in a collagen coated membrane. The time taken for a platelet plug to occlude the aperture is defined as the platelet haemostatic capacity of the sample. Currently, its clinical use is limited to the detection of various forms of von Willebrand disease owing to its limited sensitivity. A cone and plate device termed the IMPACT (Image Analysis Monitoring Platelet Adhesion Cone & Plate Technology) is also now available from Diamed; it allows imaging of platelet adhesion and aggregation at arterial shear rates, and so called shear induced platelet activation (SIPA).

3. Specific Receptors and Processes Underlying Leukocyte Adhesion from Flow The multi-step process whereby leukocytes bind to the wall of blood vessel has been well reviewed ([2]; see also Figure 3A). In post-capillary venules of inflamed tissue, upregulation of endothelial selectins allows capture of fast flowing cells through carbohydrate ligands borne on leukocyte proteins such as PSGL-1 (P-selectin glycoprotein ligand-1). Cells may roll for a prolonged period, but if an adequate activating signal is presented at the endothelial surface, they can immobilize. These signals (e.g., from chemokines specific for different leukocyte subsets) induce a conformational change in integrin receptors on the leukocytes, which then bind to their cognate receptors on the endothelium (e.g., neutrophils use E2-integrins to bind the immunoglobulin superfamily (IGSF) member, ICAM-1). This binding is relatively stable and is able to anchor the leukocytes while they spread on the surface, migrate over it and then through the endothelium. Migration requires periodic turning off and on of integrin-mediated adhesion, coordinated with directed changes in the actin cytoskeleton. Most cells migrate between endothelial cells, guided by localized receptors such as CD31, CD99 and junctional adhesion molecules such as JAM-A and JAM-C) [30], although migration through the endothelial cell body has also been described [31]. The underlying endothelial cell responses may be driven by inflammatory cytokines such as tumor necrosis factor-D (TNF) or interleukin-1E (IL1), or fast-acting agents such as histamine and thrombin. The broad process appears similar for all leukocyte subclasses, but the specific selectins, integrins and immunoglobulins, and activatory chemokines may differ. This allows selective recruitment of subsets at different times during the evolution of a response and in different tissues. For instance, in high endothelial venules of lymph nodes, there is continuous recruitment of lymphocytes but not neutrophils as part of immune surveillance.


A

B

Figure 3. Comparison of multistep adhesion processes used by: A) Leukocytes; B) platelets. CHO = carbohydrate; IgSF = immunoglobulin superfamily.

While leukocyte recruitment is typically restricted to post capillary venules where shear stresses are low, it can also occur in arterioles after disturbances such as ischemia and reperfusion [32, 33]. In addition, atherosclerotic plaques in large arteries contain monocytes and T-cells recruited from the blood. While the shear rates and stresses in arteries are generally too high to allow attachment, the lesions are typically found in regions of bends and bifurcations where flow is disturbed and vortices or reversing flow exists. These regions experience relatively low wall shear stresses, and experimentally disturbance of flow can be demonstrated to allow adhesion from blood in vessels where it would not otherwise occur [27].

4. Specific Receptors and Processes Underlying Platelet Adhesion from Flow Unlike leukocytes, platelets can adhere over the full range of physiological shear rates and, importantly, can adhere under the highly elevated shear rates (e.g., 20-40,000 s-1) or disturbed flow conditions associated with partial vessel occlusion [34, 35]. In all cases, there are two distinct stages of adhesion: the initial platelet-vessel wall interaction, followed by platelet-platelet cohesion (aggregation) required for stable thrombus formation (Figure 3B). The receptor-ligand interactions mediating these events are strongly influenced by the local hemodynamic conditions. The two most important components of the exposed extracellular matrix required for initial platelet adhesion are collagen which binds to platelet integrin DE and glycoprotein (GP)VI, and von Willebrand factor (vWf) which binds to platelet GPIb/V/IX and collagen. For platelet aggregation, the major adhesive ligands are vWf and fibrinogen which both bind to the platelet integrin DIIbE3. Recently, additional vascular adhesive proteins (eg. fibronectin, laminin) have also been shown to be able to contribute to the adhesive process.


4.1. Adhesion to Collagen Collagen is considered the most thrombogenic ECM protein mediating platelet adhesion/activation at sites of vessel injury, although acting alone it can recruit platelets only at shear rates of up to ~500 s-1 [36]). At shear rates above this threshold, vWf becomes the sole ligand able to initiate adhesion, either through exposure of subendothelial vWf (secreted from Weibel-Palade bodies of endothelial cells), or adsorption of circulating plasma vWf to exposed collagen. Nine different types of collagen are known to be expressed in the vessel wall, with fibrillar types I, III and VI being the most abundant and well studied. Two platelet membrane glycoproteins, integrin DE and GPVI (a member of the IgSF) are the major receptors for collagen, although a number of other putative receptors (of uncertain importance) have been identified, including GPIV (CD36) (reviewed by Watson & Gibbins) [37]. Until recently, a relatively simplistic ‘2-site, 2-step’ model for thrombus formation in response to collagen had been described where integrinDE was considered essential for initial platelet adhesion, with GPVI being required for platelet activation leading to integrin DIIbE3 activation and thrombus formation. With the recent development of gene-targeted mice individually lacking the two collagen receptors, the emerging picture is one in which both glycoproteins appear to play a co-operative role, each contributing to adhesion and activation. 4.2. Adhesion to vWf Platelet recruitment under elevated shear conditions (>1000 s-1) becomes completely dependent on the interaction between vWf and the platelet glycoprotein (GP) Ib/V/IX receptor. Plasma vWf rapidly binds to exposed collagen, and sub-endothelial vWf may also be present at the site of vessel damage to support platelet adhesion. Platelets bind to the A1 domain of vWf through an interaction with the GPIbD subunit of the GPIb/V/IX receptor complex. This initial step is analogous to the P-selectin-PSGL-1 interaction, and is characterized by rapid but reversible binding kinetics resulting in platelet translocation or rolling. The GPIbD-vWf interaction is unique in its ability to mediate platelet recruitment at all shear rates tested up to ~40,000 s-1 [38], and the specific characteristics of both ligand and receptors: vWf is a large multimeric protein which can present multiple binding sites to flowing platelets, and platelets express around 30,000 copies of GPIb/V/IX on their surface, allowing the rapid formation of multiple bonds in sufficient numbers to overcome the local flow conditions to mediate platelet tethering. Irreversible platelet adhesion is then dependent on activation of the major platelet integrin DIIbE3 which undergoes a conformational change enabling binding to the Arg-Gly-Asp motif within vWf. Where vWf is bound to collagen, initial platelet tethering through GPIbD also slows the velocity of the platelet enough to allow GPVI andDE-integrin to interact with collagen, thus promoting more stable platelet adhesion and downstream platelet activation. 4.3. Platelet Aggregation Subsequent to the initial platelet-vessel wall interaction and platelet activation, homotypic platelet cohesion occurs to promote thrombus formation. Part of the platelet activation process involves the release of platelet proteins including vWf and


fibrinogen which become localized on the platelet surface to enable the recruitment of additional platelets via GPIb/V/IX and DIIbE3-integrin. Plasma vWf and fibrinogen also contribute to this process by binding to the surface of adherent platelets. Under lower shear conditions, fibrinogen is considered the most important ligand, whereas vWf begins to play an increasingly important role as the shear rate increases. Thrombus growth, stabilization and consolidation is, of course, also dependent on numerous signaling pathways activated downstream of the major platelet adhesion receptors and has been reviewed extensively elsewhere [39-42]. There is increasing evidence that additional signals from soluble agonists for various G-protein coupled receptors (e.g., ADP, thrombin, TXA2), receptor tyrosine kinases (e.g., Gas-6, eph kinases) and other cell adhesion molecules (e.g., JAM-A, JAM-C), contribute to sustaining integrin DIIbE3 activation and preventing thrombus destabilization [43]. 4.4. Co-operation Between Receptors In vivo, platelets will probably never be exposed to a single adhesive protein in isolation. Thus, the adhesive process will be dependent on simultaneous and multiple receptor-ligand interactions acting in concert. As an example, the transition from rolling to stationary adhesion occurs much more efficiently when vWf is bound to collagen than on a pure vWf surface[44]. Additional adhesive proteins not traditionally thought to have a major role have recently been suggested to participate in platelet adhesion and thrombus formation. Studies using fibrinogen and vWf double knockout mice have demonstrated that occlusive arterial thrombi are still able to form [45], suggesting a role for other adhesive proteins. The same authors have demonstrated a role for plasma fibronectin in the promotion of thrombus growth and stability [46], potentially through interactions with integrin DIIbE3 or integrin DE[47]. Laminin which is expressed in the basement membrane is also a potential ligand [48]. Finally, thrombospondin-1 has been suggested to mediate platelet adhesion at high shear in vitro through an interaction with GPIb/V/IX, this being proposed as a potential backup mechanism to vWf [49]. However, the relevance of these interactions for in vivo platelet adhesion is currently not well defined.

5. Physical Factors Influencing Adhesion Hemodynamic and rheological factors can influence leukocyte and platelet adhesion in a direct manner through effects on the adhesion process itself, and indirectly through changes in receptor structure or mechano-transduction of forces into signals which affect cell behavior. 5.1. Direct Effects on Margination, Capture and Rolling For leukocytes, margination in the blood stream, and hence contact with the vessel wall, are encouraged as shear rate decreases and red cell aggregation increases [15, 16]. Perhaps surprisingly, the opposite applies for platelets [50, 51]. It is likely that relative ”particle” size is the critical factor, with inward-moving large red cell aggregates excluding leukocytes at low shear, but there is no exact theoretical framework for these processes (see chapter II.3.a).


The shear rate- or stress-dependence of leukocyte capture from flow by selectins has been well-described. Capture operates up to wall shear rates of ~300 s-1 both in isolated suspensions and whole blood [52-56]. In isolated suspensions, this equates to velocities near the wall on the order of mm/s and wall shear stresses of about 0.2-0.3 Pa. However, in blood, the shear rate and velocity at the wall are greater than expected from the predictions of Poiseuille flow. Because of plug flow of red cells and blunting of the velocity profile, the shear rate and velocity at the wall increase with increasing hematocrit [17]. The shear stress cannot accurately be computed because the local viscosity at the wall is unknown, but it is higher than in isolated suspensions. Given the ability to adhere from flow in blood at higher velocities and stresses than in dilute suspension [17], it seems that red cells may help stabilize attachment, perhaps through application of a normal force. As noted above, platelets can adhere to the vessel wall at much higher shear rates than leukocytes, in part because of their smaller size which causes them to travel slower at the vessel wall and experience less shear force (see Figure 1). Adhesion of flowing platelets is typically studied in vitro using whole blood, and indeed, in the absence of red cells, few will adhere. Thus, again, red cells influence the adhesion behaviour of platelets. The interesting and puzzling existence of a ‘shear-threshold’ for leukocyte attachment was first described for adhesion via leukocyte L-selectin and its carbohydrate ligand(s) [57]. Adhesion appears ineffective at shear stresses below ~0.05Pa, and similar but lesser effects have been described for binding to P- and Eselectin [55]. The existence of ‘catch-bonds’ between selectins and ligands, whose reverse rate constant goes down with loading, might explain this phenomenon [58]. Once captured from flow, leukocytes roll more or less steadily on surfaces coated with selectins, and in the case of lymphocytes on the IgSF member VCAM-1, at a velocity that decreases as receptor density increases. The velocity tends to increase if wall shear stress is increased, but this effect is less marked in whole blood, again perhaps due to a stabilizing effect of red cells [52, 56]. Direct observations of shortlived adhesive contacts of leukocytes with surfaces coated with a low density of selectin receptors show that adhesion lifetimes at low bond numbers are <1 s and are reduced by shear stress [10]. However, the ability of bonds to withstand stress is influenced not only by intrinsic bond kinetics but also by the ability of the cell to deform and allow membrane tethers to be drawn out. This “give” in the system tends to effectively reduce the off-rate, so that intrinsic bond kinetics may be faster than simple adhesion observations suggest [59]. Platelets also translocate when bound through the GPIbD-vWf interaction. Platelet rolling velocity on immobilized vWf increased non-linearly with increasing shear stress, but adhesion was considered to be positively regulated in the sense that increasing numbers of platelets adhered [60]. However, the latter effect did not take into account the increasing rate of delivery of platelets to the surface at the higher flow. The GPIbD-vWf interaction has recently been directly demonstrated to share some of the biomechanical properties associated with the selectin family. These properties include the requirement for a shear threshold and rapid bond kinetics, with similar values for the reverse rate constant and reactive compliance as selectin-ligand bonds [61]. One caveat to this model is that these experiments were performed with purified platelets and vWf bound to spherical beads, which may not accurately represent the adhesive events in vivo. Interestingly, the use of known ”gain-of-function” vWf mutations responsible for type 2B von Willebrand disease in humans which results in


spontaneous binding of platelets to plasma vWf, a thrombocytopaenia and ultimately a propensity to bleed, has demonstrated that altered kinetic properties of the GPIbD-vWf bond are likely to be responsible for the observed ”gain-of-function” phenotype [62]. Shear-dependent formation of membrane tethers has been demonstrated in platelets as well as neutrophils, and the evidence suggests that tethers play an important stabilizing role in the early, translocating phase of adhesion prior to cell activation and stable arrest by decreasing the force on adhesive bonds. Neutrophils form membrane tethers while rolling on P-selectin [63], with the numbers of tethers increasing with increasing shear stress and correlating with periods of slower, more uniform rolling [64]. Platelets translocating on vWf under flow also extend membrane tethers [65, 66], and, similar to neutrophils, the rate and extent of tether formation is sheardependent and regulates the stop-start phases of platelet translocation [65]. 5.2. Indirect Effects on Activation and Stable Adhesion Both neutrophils and platelets may ”roll” for a period before stopping, or stop almost immediately after capture, depending on the efficiency of delivery of the activating signal. In the case of chemotactic agents presented on endothelial cells, captured neutrophils can bind the agent through specific receptors, transform their integrins and convert to stable adhesion in under one second [67]. Subsequent shape change then takes less than a minute, and migration over and through the endothelial surface requires a further 2-3 minutes [68]. Shear forces have several different effects on the later stages of adhesion, apparently through ”mechano-transduction” of signals. Neutrophils sense shear stress, which actively guides migration through receptor-dependent mechanisms [69]. It has also been suggested that applied stress can accelerate the movement of T-cells and neutrophils through the endothelial cell barrier [70, 71]. Mechano-transduction of shear forces by various platelet receptors has been demonstrated to contribute to platelet adhesion. Both GPIb/V/IX [72] and integrin DIIbE3[73] have been suggested as platelet mechanoreceptors through a mechanism involving intracellular calcium signaling, an event essential for sustained platelet activation and adhesion. At physiological shear rates (up to ~2000 s-1) platelets circulate freely (in the presence of vWf) in their resting state. However, elevated shear rates (> 3000 s-1) are sufficient to activate platelets, resulting in shear induced platelet aggregation (SIPA) [74]. This process is dependent on both GPIb/V/IX and DIIbE3-integrin binding to vWf, leading to signaling events involving calcium influx and ADP release. However, much of the evidence suggesting that elevated shear per se activates platelets has been derived from in vitro studies using platelets sheared in suspension. There is currently no evidence that SIPA occurs in vivo, so the precise events initiating pathological platelet activation which may lead to occlusive thrombus formation remain elusive. What initiates SIPA at the molecular level is also still unknown. Hydrodynamic shear may influence the 3-dimensional morphology of vWf, resulting in an unfolding of vWf molecules[75], although the significance of this phenomena is unknown and has since been disputed, at least for collagen-bound vWf [76]). vWf in suspension has also been shown to self-associate in the presence of high shear [77], a phenomenon that could influence platelet adhesion under elevated flow conditions in the circulation. Interestingly, the crystal structures of vWf and GPIbD have revealed an ”unmasking” mechanism for GPIbD binding to vWf [78]. This finding suggests that shear-dependent


changes in the morphology of either vWf or GPIbD could well be a key initiating event for platelet adhesion under elevated, pathological shear conditions, although to demonstrate this experimentally will be a challenge. It is also notable that endothelial cells respond to changes in the shear stress that they experience, and can adopt different phenotypes depending on the local stress in the circulation [79]. Of relevance here, this leads to modulation of their responses to inflammatory cytokines and hence ability to capture flowing leukocytes [80]. Moreover, oscillatory shear stress with a reversing component appears to be proadhesive per se by causing upregulation of adhesion receptors [81].

6. Abnormal vs. Physiological Adhesion In general, the adhesive processes described above are protective responses against traumatic tissue damage and infection. The physiological multi-step responses have evolved to allow a high degree of control, but uncontrolled adhesion of either leukocytes or platelets is highly dangerous. Undesirable, continuous recruitment of leukocytes to tissue occurs in chronic inflammatory disorders such as rheumatoid arthritis, but is not a ”rheological” problem in the sense of causing impairment of perfusion or resulting from flow disturbance. With respect to microcirculatory impairment, neutrophilic granulocytes have been the most studied leukocytes because they have the potential to acutely occlude microvessels, and may also induce damage through release of lytic enzymes and reactive oxygen species normally used for bacterial killing and tissue remodeling. Unwanted adhesion, probably occurring along with mechanical changes due to cell activation (see Chapter II.4.c), may lie at the heart of the no-reflow phenomenon observed after ischemia and reperfusion or shock [82-84]. In either case, trapping of cells when perfusion pressure is reduced may be exacerbated by adhesion to the wall, so that upon return of pressure cells do not washout and a microcirculatory deficit persists. For instance, the extent of tissue necrosis after myocardial infarction is much reduced in animals depleted of neutrophils [85]. Another example of unwanted activation is in small-vessel vasculitis, where circulating autoantibodies against neutrophils (i.e., anti-neutrophil cytoplasm antibodies, ANCA) bind to surface antigen(s) and signal the cells through ligation of Fc-receptors. This results in changes in both mechanical and adhesive properties of the leukocytes which appear to potentiate their deposition in microvessels of vital organs such as the lungs and kidneys damaged in these disorders[86]. Platelet adhesion, while essential for hemostasis at sites of vascular injury to minimize hemorrhage, is also the initiating step in the formation of pathological thrombi that usually occur at sites of vessel stenosis or atherosclerotic plaque rupture. These occlusive arterial thrombi, often resulting in acute myocardial infarction or stroke, are the leading cause of death in industrialized countries. The adhesive events underlying pathological thrombosis appear to be similar to those required for hemostasis and to involve the same receptor-ligand interactions, but in some way functioning in an uncontrolled manner. The factors contributing to occlusive thrombosis are likely to be many, but one of the features thought to play an important role in thrombosis is the elevation and/or alteration in blood flow that occurs particularly at sites of vessel stenosis and atherosclerotic plaque rupture. Complex patterns of flow can develop, deviating significantly from constant laminar flow, such


that at the apex of a stenosis shear rates can be greatly elevated, whereas downstream areas of flow reversal and recirculation occur. Since the vWf-GPIbD interaction is positively regulated by increasing shear, and appears to be highly sensitive to changes in shear, it is likely that this interaction plays a critical role in the development of occlusive thrombi. There is now some evidence in the literature suggesting that rapid accelerations or decelerations in flow can elicit an enhanced platelet activation response that is absent at constant high shear [72, 73], and that short bursts of high shear stress are sufficient to induce platelet activation and aggregation [87]. The composition of an atherosclerotic plaque differs from a healthy vessel wall, such that changes in adhesion molecules (e.g., increased collagen), combined with the loss of negative platelet regulators that are normally expressed by intact endothelial cells (e.g., nitric oxide, prostaglandins, CD39 ecto-ADPase), may lead to exaggerated platelet adhesion and activation. The particular geometry and surface irregularity of the plaque may also modulate flow patterns, potentially contributing to the likelihood of both plaque rupture and occlusive thrombosis [88]. Finally, it may be noted that platelets and platelet-leukocyte interactions are increasingly recognized as contributing to inflammatory pathology. Platelets have the potential to interact with both endothelial cells and leukocytes and appear to link inflammation, atherogenesis and thrombosis. Platelet interaction with intact endothelium has been demonstrated in several experimental conditions, including ischemia/reperfusion and hypercholesterolemia, and two common features of these conditions are endothelial cell dysfunction and leukocyte recruitment [89]. Adhesion of leukocytes to platelets, themselves attached to highly polymerised vWF, has been demonstrated at high shear stress and may represent a mechanism supporting leukocyte tethering to arterial endothelial cells [90]. Thus, the rheology of leukocyte-platelet interactions is important and worthy of investigation, as well as the adhesion of the separate cells to the vessel wall. References [1]

H.L. Goldsmith and V.T. Turitto, Rheological aspects of thrombosis and haemostasis: basic principles and applications., Thrombos Haemost 55 (1986), 415-35. [2] T.A. Springer, Traffic signals on endothelium for lymphocyte recirculation and leukocyte emigration, Ann Rev Physiol 57 (1995), 827-72. [3] Z.M. Ruggeri, Platelet interactions with vessel wall components during thrombogenesis, Blood Cells Mol Dis 36 (2006), 145-7. [4] C. Cozens-Roberts, D.A. Lauffenburger and J.A. Quinn, Receptor-mediated cell attachment and detachment kinetics. I. Probabilistic model and analysis, Biophys. J. 58 (1990), 841-856. [5] D.A. Hammer and S.M. Apte, Simulation of cell rolling and adhesion on surfaces in shear flow: general results and analysis of selectin-mediated neutrophil adhesion, Biophys J 63 (1992), 35-57. [6] A. Tozeren and K. Ley, How do selectins mediate leukocyte rolling in venules?, Biophys J 63 (1992), 700-9. [7] M.D. Ward, M. Dembo and D.A. Hammer, Kinetics of cell detachment: peeling of discrete receptor clusters, Biophys J 67 (1994), 2522-34. [8] K.C. Chang and D.A. Hammer, The forward rate of binding of surface-tethered reactants: effect of relative motion between two surfaces, Biophys. J. 76 (1999), 1280-1292. [9] C.E. Orsello, D.A. Lauffenburger and D.A. Hammer, Molecular properties in cell adhesion: a physical and engineering perspective, Trends Biotechnol. 19 (2001), 310-316. [10] R. Alon, D.A. Hammer and T.A. Springer, Lifetime of the P-selectin-carbohydrate bond and its response to tensile force in hydrodynamic flow [published erratum appears in Nature 1995 Sep 7;376(6544):86], Nature 374 (1995), 539-42.

166 S.L. Cranmer and G.B. Nash / Adhesion of Circulating Leukocytes and Platelets to the Vessel Wall [11] T.A. Doggett, G. Girdhar, A. Lawshe, D.W. Schmidtke, I.J. Laurenzi, S.L. Diamond and T.G. Diacovo, Selectin-like kinetics and biomechanics promote rapid platelet adhesion in flow: the GPIb(alpha)-vWF tether bond, Biophys J 83 (2002), 194-205. [12] M.B. Lawrence, C.W. Smith, S.G. Eskin and L.V. McIntire, Effect of venous shear stress on CD18mediated neutrophil adhesion to cultured endothelium, Blood 75 (1990), 227-37. [13] A. Atherton and G.V. Born, Quantitative investigations of the adhesiveness of circulating polymorphonuclear leucocytes to blood vessel walls, J Physiol 222 (1972), 447-74. [14] A.J. Goldman, R.G. Cox, and H. Brenner, Slow viscous motion of a sphere parallel to a plane wall. II Couette flow., Chemical Engineering Science 22 (1967), 653-660. [15] H.L. Goldsmith and S. Spain, Margination of leukocytes in blood flow through small tubes, Microvasc Res 27 (1984), 204-22. [16] U. Nobis, A.R. Pries, G.R. Cokelet and P. Gaehtgens, Radial distribution of white cells during blood flow in small tubes, Microvasc Res 29 (1985), 295-304. [17] K.B. Abbitt and G.B. Nash, Rheological properties of the blood influencing selectin-mediated adhesion of flowing leukocytes, Am J Physiol 285 (2003), H229-H240. [18] S.D. House and H.H. Lipowsky, Leukocyte-endothelium adhesion: microhemodynamics in mesentery of the cat, Microvasc Res 34 (1987), 363-79. [19] H.H. Lipowsky, D.A. Scott and J.S. Cartmell, Leukocyte rolling velocity and its relation to leukocyteendothelium adhesion and cell deformability, Am J Physiol 270 (1996), 1371-80. [20] M.J. Pearson and H.H. Lipowsky, Influence of erythrocyte aggregation on leukocyte margination in postcapillary venules of rat mesentary, Am J Physiol 279 (2000), 1460-1471. [21] M.G. Oude Egbrink, G.J. Tangelder, D.W. Slaaf and R.S. Reneman, Thromboembolic reaction following wall puncture in arterioles and venules of the rabbit mesentery, Thromb. Haemost. 59 (1988), 23-28. [22] S.M. Day, J.L. Reeve, D.D. Myers and W.P. Fay, Murine thrombosis models, Thromb Haemost 92 (2004), 486-94. [23] B. Furie and B.C. Furie, Thrombus formation in vivo, J Clin Invest 115 (2005), 3355-62. [24] S. Falati, P. Gross, G. Merrill-Skoloff, B.C. Furie and B. Furie, Real-time in vivo imaging of platelets, tissue factor and fibrin during arterial thrombus formation in the mouse, Nat. Med. 8 (2002), 11751181. [25] K.S. Sakariassen, V.T. Turitto and H.R. Baumgartner, Recollections of the development of flow devices for studying mechanisms of hemostasis and thrombosis in flowing whole blood, J Thromb Haemost 2 (2004), 1681-90. [26] M.B. Lawrence, L.V. McIntire and S.G. Eskin, Effect of flow on polymorphonuclear leukocyte/endothelial cell adhesion, Blood 70 (1987), 1284-90. [27] C. Skilbeck, S.M. Westwood, P.G. Walker, T. David, and G.B. Nash, Population of the vessel wall by leukocytes binding to P-selectin in a model of disturbed arterial flow, Arteriosclerosis, thrombosis and vascular biology 21 (2001), 1294-1300. [28] R.M. Barstad, H.E. Roald, Y. Cui, V.T. Turitto and K.S. Sakariassen, A perfusion chamber developed to investigate thrombus formation and shear profiles in flowing native human blood at the apex of welldefined stenoses, Arterioscler Thromb 14 (1994), 1984-91. [29] J.J. Zwaginga, K.S. Sakariassen, G. Nash, M.R. King, J.W. Heemskerk, M. Frojmovic and M.F. Hoylaerts, Flow-based assays for global assessment of hemostasis. Part 2: current methods and considerations for the future, J Thromb Haemost 4 (2006), 2716-7. [30] C. Johnson-Leger, M. Aurrand-Lions and B.A. Imhof, The parting of the endothelium: miracle or simply a junctional affair., J Cell Sci 113 (2000), 921-933. [31] D. Feng, J.A. Nagy, K. Pyne, H.F. Dvorak and A.M. Dvorak, Neutrophils emigrate from venules by a transendothelial cell pathway in response to FMLP., Journal of Experimental Medicine. 187 (1998), 903-915. [32] E.J. Kunkel, U. Jung and K. Ley, TNF-alpha induces selectin-mediated leukocyte rolling in mouse cremaster muscle arterioles, Am. J. Physiol 272 (1997), H1391-H1400. [33] Liu X, Peter FW, Barker JH, Pierangeli SS, Harris EN, and Anderson GL, Leukocyte-endothelium interaction in arterioles after ischemia and reperfusion., Journal of Surgical Research 87 (1999), 77-84. [34] A. Mailhac, J.J. Badimon, J.T. Fallon, A. Fernandez-Ortiz, B. Meyer, J.H. Chesebro, V. Fuster and L. Badimon, Effect of an eccentric severe stenosis on fibrin(ogen) deposition on severely damaged vessel wall in arterial thrombosis. Relative contribution of fibrin(ogen) and platelets, Circulation 90 (1994), 988-96. [35] J.M. Siegel, C.P. Markou, D.N. Ku and S.R. Hanson, A scaling law for wall shear rate through an arterial stenosis, J Biomech Eng 116 (1994), 446-51. [36] Z.M. Ruggeri, Platelets in atherothrombosis, Nat Med 8 (2002), 1227-34.

S.L. Cranmer and G.B. Nash / Adhesion of Circulating Leukocytes and Platelets to the Vessel Wall 167 [37] S.P. Watson and J. Gibbins, Collagen receptor signalling in platelets: extending the role of the ITAM, Immunol. Today 19 (1998), 260-264. [38] B. Savage, F. Almus-Jacobs and Z.M. Ruggeri, Specific synergy of multiple substrate-receptor interactions in platelet thrombus formation under flow, Cell 94 (1998), 657-66. [39] S.P. Jackson, W.S. Nesbitt and S. Kulkarni, Signaling events underlying thrombus formation, J Thromb Haemost 1 (2003), 1602-12. [40] S.P. Watson, J.M. Auger, O.J. McCarty and A.C. Pearce, GPVI and integrin alphaIIb beta3 signaling in platelets, J Thromb Haemost 3 (2005), 1752-62. [41] Y. Ozaki, N. Asazuma, K. Suzuki-Inoue and M.C. Berndt, Platelet GPIb-IX-V-dependent signaling, J Thromb Haemost 3 (2005), 1745-51. [42] J.M. Gibbins, Platelet adhesion signalling and the regulation of thrombus formation, J Cell Sci 117 (2004), 3415-25. [43] L.F. Brass, H. Jiang, J. Wu, T.J. Stalker and L. Zhu, Contact-dependent signaling events that promote thrombus formation, Blood Cells Mol Dis 36 (2006), 157-61. [44] B. Savage, M. Cattaneo and Z.M. Ruggeri, Mechanisms of platelet aggregation, Curr Opin Hematol 8 (2001), 270-6. [45] H. Ni, C.V. Denis, S. Subbarao, J.L. Degen, T.N. Sato, R.O. Hynes and D.D. Wagner, Persistence of platelet thrombus formation in arterioles of mice lacking both von Willebrand factor and fibrinogen, J Clin Invest. 106 (2000), 385-92. [46] H. Ni, P.S. Yuen, J.M. Papalia, J.E. Trevithick, T. Sakai, R. Fassler, R.O. Hynes and D.D. Wagner, Plasma fibronectin promotes thrombus growth and stability in injured arterioles, Proc Natl Acad Sci U S A 100 (2003), 2415-9. [47] S. Gruner, M. Prostredna, V. Schulte, T. Krieg, B. Eckes, C. Brakebusch and B. Nieswandt, Multiple integrin-ligand interactions synergize in shear-resistant platelet adhesion at sites of arterial injury in vivo, Blood 102 (2003), 4021-7. [48] O. Inoue, K. Suzuki-Inoue, O.J. McCarty, M. Moroi, Z.M. Ruggeri, T.J. Kunicki, Y. Ozaki and S.P. Watson, Laminin stimulates spreading of platelets through integrin alpha6beta1-dependent activation of GPVI, Blood 107 (2006), 1405-12. [49] K. Jurk, K.J. Clemetson, P.G. de Groot, M.F. Brodde, M. Steiner, N. Savion, D. Varon, J.J. Sixma, H. Van Aken and B.E. Kehrel, Thrombospondin-1 mediates platelet adhesion at high shear via glycoprotein Ib (GPIb): an alternative/backup mechanism to von Willebrand factor, FASEB J 17 (2003), 1490-2. [50] B. Woldhuis, G.J. Tangelder, D.W. Slaaf and R.S. Reneman, Concentration profile of blood platelets differs in arterioles and venules, Am. J. Physiol 262 (1992), H1217-H1223. [51] B. Woldhuis, G.J. Tangelder, D.W. Slaaf and R.S. Reneman, Influence of dextrans on platelet distribution in arterioles and venules, Pflugers Arch. 425 (1993), 191-198. [52] M.B. Lawrence and T.A. Springer, Leukocytes roll on a selectin at physiologic flow rates: distinction from and prerequisite for adhesion through integrins, Cell 65 (1991), 859-73. [53] M.B. Lawrence and T.A. Springer, Neutrophils roll on E-selectin, Journal of Immunology. 151 (1993), 6338-6346. [54] E.B. Finger, K.D. Puri, R. Alon, M.B. Lawrence, U.H. von Andrian and T.A. Springer, Adhesion through L-selectin requires a threshold hydrodynamic shear, Nature 379 (1996), 266-269. [55] Lawrence MB, Kansas GS, Kunkel EJ, and Ley K, Threshold levels of fluid shear promote leukocyte adhesion through selectins (CD62L,P,E)., Journal of Cell Biology 136 (1997), 717-727. [56] K.B. Abbitt and G.B. Nash, Characteristics of leucocyte adhesion directly observed in flowing whole blood in vitro, Br J Haematol 112 (2001), 55-63. [57] E.B. Finger, K.D. Puri, R. Alon, M.B. Lawrence, U.H. von Andrian and T.A. Springer, Adhesion through L-selectin requires a threshold hydrodynamic shear, Nature 379 (1996), 266-269. [58] B.T. Marshall, M. Long, J.W. Piper, T. Yago, R.P. McEver and C. Zhu, Direct observation of catch bonds involving cell-adhesion molecules, Nature 423 (2003), 190-193. [59] M.R. King, V. Heinrich, E. Evans and D.A. Hammer, Nano-to-Micro Scale Dynamics of P-Selectin Detachment from Leukocyte Interfaces. III. Numerical Simulation of Tethering under Flow, Biophys. J. 88 (2005), 1676-1683. [60] B. Savage, E. Saldivar and Z.M. Ruggeri, Initiation of platelet adhesion by arrest onto fibrinogen or translocation on von Willebrand factor, Cell 84 (1996), 289-97. [61] T.A. Doggett, G. Girdhar, A. Lawshe, D.W. Schmidtke, I.J. Laurenzi, S.L. Diamond and T.G. Diacovo, Selectin-like kinetics and biomechanics promote rapid platelet adhesion in flow: the GPIb(alpha)-vWF tether bond, Biophys. J. 83 (2002), 194-205. [62] T.A. Doggett, G. Girdhar, A. Lawshe, D.W. Schmidtke, I.J. Laurenzi, S.L. Diamond and T.G. Diacovo, Selectin-like kinetics and biomechanics promote rapid platelet adhesion in flow: the GPIb(alpha)-vWF tether bond, Biophys. J. 83 (2002), 194-205.

168 S.L. Cranmer and G.B. Nash / Adhesion of Circulating Leukocytes and Platelets to the Vessel Wall [63] D.W. Schmidtke and S.L. Diamond, Direct observation of membrane tethers formed during neutrophil attachment to platelets or P-selectin under physiological flow, J Cell Biol 149 (2000), 719-30. [64] V. Ramachandran, M. Williams, T. Yago, D.W. Schmidtke and R.P. McEver, Dynamic alterations of membrane tethers stabilize leukocyte rolling on P-selectin, Proc Natl Acad Sci U S A 101 (2004), 13519-24. [65] S.M. Dopheide, M.J. Maxwell and S.P. Jackson, Shear-dependent tether formation during platelet translocation on von Willebrand factor, Blood 99 (2002), 159-67. [66] A.J. Reininger, H.F. Heijnen, H. Schumann, H.M. Specht, W. Schramm and Z.M. Ruggeri, Mechanism of platelet adhesion to von Willebrand factor and microparticle formation under high shear stress, Blood 107 (2006), 3537-45. [67] G.E. Rainger, A.C. Fisher and G.B. Nash, Neutrophil rolling is rapidly transformed to stationary adhesion by IL-8 or PAF presented on endothelial surfaces., Am J Physiol 272 (1997), H114-H122. [68] N.T. Luu, G.E. Rainger and G.B. Nash, Kinetics of the different steps during neutrophil migration through cultured endothelial monolayers treated with tumour necrosis factor-alpha, J Vasc Res 36 (1999), 477-85. [69] G.E. Rainger, C.D. Buckley, D.L. Simmons and G.B. Nash, Neutrophils sense flow-generated stress and direct their migration through alphaVbeta3-integrin, Am J Physiol 276 (1999), 858-64. [70] G. Cinamon, V. Shinder and R. Alon, Shear forces promote lymphocyte migration across vascular endothelium bearing apical chemokines., Nature Immunol 2 (2001), 515-522. [71] G. Cinamon, V. Shinder, R. Shamri and R. Alon, Chemoattractant signals and beta 2 integrin occupancy at apical endothelial contacts combine with shear stress signals to promote transendothelial neutrophil migration, J. Immunol. 173 (2004), 7282-7291. [72] M. Mazzucato, P. Pradella, M.R. Cozzi, L. De Marco and Z.M. Ruggeri, Sequential cytoplasmic calcium signals in a 2-stage platelet activation process induced by the glycoprotein Ibalpha mechanoreceptor, Blood 100 (2002), 2793-800. [73] I. Goncalves, W.S. Nesbitt, Y. Yuan and S.P. Jackson, Importance of temporal flow gradients and integrin alphaIIbbeta3 mechanotransduction for shear activation of platelets, J Biol Chem 280 (2005), 15430-7. [74] M.H. Kroll, J.D. Hellums, L.V. McIntire, A.I. Schafer and J.L. Moake, Platelets and shear stress, Blood 88 (1996), 1525-41. [75] C.A. Siedlecki, B.J. Lestini, K.K. Kottke-Marchant, S.J. Eppell, D.L. Wilson and R.E. Marchant, Shear-dependent changes in the three-dimensional structure of human von Willebrand factor, Blood 88 (1996), 2939-50. [76] L. Novak, H. Deckmyn, S. Damjanovich and J. Harsfalvi, Shear-dependent morphology of von Willebrand factor bound to immobilized collagen, Blood 99 (2002), 2070-6. [77] H. Shankaran, P. Alexandridis and S. Neelamegham, Aspects of hydrodynamic shear regulating shearinduced platelet activation and self-association of von Willebrand factor in suspension, Blood 101 (2003), 2637-45. [78] S. Uff, J.M. Clemetson, T. Harrison, K.J. Clemetson and J. Emsley, Crystal structure of the platelet glycoprotein Ib(alpha) N-terminal domain reveals an unmasking mechanism for receptor activation, J Biol Chem 277 (2002), 35657-63. [79] S. Chien, S. Li, and Shyy.Y.-J., Effects of mechanical forces on signal transduction and gene expression in endothelial cells., Hypertension 31 (1998), 162-169. [80] S. Sheikh, G.E. Rainger, Z. Gale, M. Rahman and G.B. Nash, Exposure to fluid shear stress modulates the ability of endothelial cells to recruit neutrophils in response to tumor necrosis factor-alpha: a basis for local variations in vascular sensitivity to inflammation, Blood 102 (2003), 2828-2834. [81] A. Barakat and D. Lieu, Differential responsiveness of vascular endothelial cells to different types of fluid mechanical shear stress, Cell Biochem. Biophys. 38 (2003), 323-343. [82] R.L. Engler, G.W. Schmid-Schonbein and R.S. Pavelec, Leukocyte capillary plugging in myocardial ischemia and reperfusion in the dog, Am. J. Pathol. 111 (1983), 98-111. [83] U. Bagge, B. Amundson and C. Lauritzen, White blood cell deformability and plugging of skeletal muscle capillaries in hemorrhagic shock, Acta Physiol Scand. 108 (1980), 159-163. [84] D.L. Carden and D.N. Granger, Pathophysiology of ischaemia-reperfusion injury, J. Pathol. 190 (2000), 255-266. [85] J.L. Romson, B.G. Hook, S.L. Kunkel, G.D. Abrams, M.A. Schork and B.R. Lucchesi, Reduction of the extent of ischemic myocardial injury by neutrophil depletion in the dog, Circulation 67 (1983), 10161023. [86] C.O. Savage, The evolving pathogenesis of systemic vasculitis, Clin. Med. 2 (2002), 458-464. [87] J.N. Zhang, A.L. Bergeron, Q. Yu, C. Sun, L.V. McIntire, J.A. Lopez and J.F. Dong, Platelet aggregation and activation under complex patterns of shear stress, Thromb Haemost 88 (2002), 817-21.

S.L. Cranmer and G.B. Nash / Adhesion of Circulating Leukocytes and Platelets to the Vessel Wall 169 [88] J.S. Stroud, S.A. Berger and D. Saloner, Influence of stenosis morphology on flow through severely stenotic vessels: implications for plaque rupture, J Biomech 33 (2000), 443-55. [89] A. Tailor, D. Cooper and D.N. Granger, Platelet-vessel wall interactions in the microcirculation, Microcirculation 12 (2005), 275-85. [90] A. Bernardo, C. Ball, L. Nolasco, H. Choi, J.L. Moake and J.F. Dong, Platelets adhered to endothelial cell-bound ultra-large von Willebrand factor strings support leukocyte tethering and rolling under high shear stress, J Thromb Haemost 3 (2005), 562-70.

170


Mechanisms of Blood Rheology Alterations Oguz K. BASKURT1 Department of Physiology, Akdeniz University Faculty of Medicine, Antalya, Turkey

Introduction The factors that determine the rheological behavior of blood, a two-phase fluid, include the relative volume of each phase as reflected by the hematocrit value, plasma composition and the properties of cellular elements. The contribution of these factors to blood rheology has been detailed in previous chapters (Chapters II.2 and II.3.a). Thus it is not surprising that the mechanisms of blood rheology alterations are also related to these factors. In a classification published by P.F. Leblond in an early, 1987 textbook of hemorheology, hyperviscosity syndromes were discussed from a pathophysiological point of view [1]: a) Polycythemic disorders to include erythrocytosis and leukocytosis. b) Sclerocythemic disorders to include conditions with impaired red blood cell (RBC) deformability. c) Plasmatic disorders to include alterations in plasma viscosity and red cell aggregation. Similar approaches have been used by others [2]. J-F. Stoltz classified hyperviscosity syndromes into five groups [3]: a) Increase in number of blood cells; b) Altered plasma protein levels; c) Increased RBC internal rigidity; d) Changes in RBC viscoelastic properties; e) Enhanced RBC aggregation. These classifications, based on the mechanisms of hemorheological alterations, need to be revised in light of developments in the past twenty years. There has been a significant change in our understanding of the factors that determine the degree of RBC aggregation: in addition to plasma factors and local shear forces, RBC properties are now well established as an additional factor affecting the intensity of aggregation. Thus, enhanced RBC aggregation may not always be related to plasma composition changes, but can co-exist with plasma viscosity alterations. RBC properties therefore influence both cell deformability and aggregation. Table 1 presents a classification of hemorheological alterations based upon the considerations described above.

1. Alterations of Blood Composition As detailed in the previous chapters (Chapter II.2), blood is a suspension of cellular elements in the liquid phase plasma. Normally, the vast majority of cellular elements 1

Corresponding author: Department of Physiology, Akdeniz University Faculty of Medicine, Antalya, Turkey; E Mail: [email protected].

O.K. Baskurt / Mechanisms of Blood Rheology Alterations

171

are RBC, with a 103 times higher relative contribution to total cell population compared to leukocytes. Therefore, the hematocrit value as an index of the relative contribution of RBC plays the most significant role under physiological conditions. Leukocytes may also play a role in blood fluidity, but only if their numbers increase very significantly due to pathological processes such as leukemia. Plasma composition contributes to blood rheology by affecting plasma viscosity and/or RBC aggregation. Table 1. Classification of hemorheological alterations according to their mechanism. I. Alterations of blood composition 1. Hematocrit alterations 2. Increased number of leukocytes 3. Plasma composition changes II. Alterations in RBC properties 1. Alterations in RBC deformability a. Genetic defects b. Alterations in red blood cell metabolism c. Oxidative damage d. Micro-environmental influences on red blood cells e. Other factors 2. Alterations in RBC aggregation a. Change in plasma composition b. Alterations in RBC deformability c. Alterations in the surface properties of red blood cells III. Alterations of leukocytes 1. Leukocyte activation and microcirculatory function 2. Leukocyte-induced alterations in red blood cells rheology a. Effect on RBC deformability b. Effect on RBC aggregability

1.1. Hematocrit Alterations Hematocrit is the most important determinant of blood viscosity under bulk flow conditions (e.g., large diameter vessels or large geometry viscometers). However, it has been demonstrated that it is not the same throughout the circulatory system, with microvascular hematocrit influenced by various mechanisms (see Chapters II.3.a, III.1 and III.3). Additionally, it has been shown that the hematocrit of blood leaving the microvasculature (i.e., the venous hematocrit) may be affected by flow conditions in the preceding microvessels [4]. It is thus well-established that the hematocrit values measured in blood obtained from large blood vessels do not represent this parameter at the microcirculatory level [5]. The physiological meaning of hematocrit value should be carefully considered. On one hand, the hematocrit value reflects the oxygen carrying capacity of blood since higher hematocrit usually correlates with higher hemoglobin concentration and higher oxygen binding capacity. On the other hand, the hematocrit value is logarithmically related to blood viscosity, hence a determinant of flow resistance. Oxygen transfer to a given tissue is a function of both blood flow to that tissue and oxygen content of blood flowing to that tissue. This double role of hematocrit yields the concept of an optimal hematocrit at which the oxygen transfer to tissues should be maximum [6]. Above this critical value of hematocrit, increased blood viscosity and flow resistance would dominate and the main physiological function of blood flow (i.e., supplying oxygen to

172


tissues) would be impaired. This optimal hematocrit has been reported to correspond to normal physiological range of hematocrit value for mammals [6]. Being related to the physiological importance of an optimal hematocrit value, RBC production is a well-controlled process [7]. The main control factor of RBC production is the hormone erythropoietin, and its secretion is, in turn, controlled by the degree of tissue oxygenation [7]. Therefore, the main physiological stimulus for RBC production and hence increased hematocrit is tissue hypoxia: RBC production is kept at a constant rate to maintain normal tissue oxygenation through a negative-feedback control mechanism. Despite this tight physiological control of RBC production, hematocrit should be considered to be a dynamic property. The relative volume of RBC in blood may vary acutely, changing in minutes without an absolute change in total RBC mass. This might be the result of an alteration of extacellular fluid volume (e.g., dehydration), or even an acute alteration of the volume of the vascular compartment, (e.g., capillary leak syndrome). Rapid fluid transfer across the endothelial barrier can yield a contraction or expansion of plasma volume and, given a constant RBC mass, hematocrit alterations are inversely related to plasma volume [8]. An increase in hematocrit value due to such mechanisms has been termed stress or relative polycythemia [9]. In contrast with relative polycythemia, altered hematocrit might be the result of a true polycythemia due to increased RBC mass. Increased RBC production might be secondary to a variety of physiologic or pathologic conditions, such as exposure to high altitudes, various disease processes (e.g., pulmonary, cardiovascular diseases, obesity, hypertension), smoking and drug usage [8, 10]. Alternatively, polycythemia might be due to a primary disease process. Polycythemia vera is a clonal myeloproliferative disease involving multipotent hematopoietic progenitor cells [11, 12]. The resulting clinical picture is usually characterized by increased RBC mass and often increased leukocyte and platelet counts. Serum erythropoietin levels have been found to be significantly lower than normal in most patients with polycythemia vera [13]. Although several specific molecular markers have recently been described, the etiology of polycythemia vera is still unknown [11, 12]. 1.2. Increased Number of Leukocytes Normally, the contribution of leukocytes to blood viscosity under bulk flow conditions is negligible due to their relatively small contribution to the cell population of blood. However, if the number of leukocytes becomes comparable with that of the RBC, hyperviscosity might become manifest. Increased blood viscosity has been reported in various types of leukemia [14, 15]. Alternatively, the bulk viscosity of blood might be unaltered due to a decrease of RBC combined with an increased leukocrit [16]. However, even with an unaltered bulk viscosity of blood, increased number of leukocytes may seriously affect flow in the microcirculation as these cells are larger and more rigid compared to RBC (see Chapter II.4.c.) Slugging and aggregation of leukocytes may take place in the microcirculation and the phenomenon of leukostasis may occur [16, 17]. 1.3. Altered Plasma Composition An altered, usually increased, content of certain proteins of plasma is a well-known cause of hyperviscosity. Plasma viscosity is mainly determined by the protein content,


173

but the contribution of various protein fractions differ. Albumin contributes about 36% of the difference between water and plasma viscosity, while its contribution to total protein pool of plasma is about 60% [18]. Fibrinogen corresponds to only about 4% of total protein weight in plasma, while its contribution to plasma viscosity is about 22% under physiological conditions [18]. The difference in relative contribution between the protein fractions is due to their molecular size and shape, fibrinogen being significantly more asymmetric compared to other proteins [18]. Globulin fractions also contribute to plasma viscosity to a greater extent than albumin due to their higher molecular weight. Plasma composition and viscosity can also be altered due to physiological variations, mainly related to body fluid balance, which affect the concentration of plasma components without a significant alteration of their absolute amount (e.g., dehydration). Alternatively, the concentration of plasma components may change due to pathophysiological processes. The most common situation in which plasma protein composition and related properties are altered is the acute phase reaction. In this case, the concentrations of a variety of proteins are increased within hours during the course of inflammatory reactions. The concentration of some proteins can vary greatly: Creactive protein may increase by up to 1000-fold, and fibrinogen, another well-known acute phase reactant, may increase by 5-fold, thereby increasing plasma viscosity significantly [19]. Other acute phase reactants may also contribute to plasma viscosity. Immunoglobulin concentrations in plasma can be significantly increased in patients with monoclonal or polyclonal gammopathies [20]. Monoclonal hypergammaglobulinemia is manifest in multiple myeoloma and Waldenström’s macroglobulinemia. Polyclonal hypergammaglobulinemia is usually associated with autoimmune disorders such as rheumatoid arthritis and Sjögren’s syndrome. Plasma viscosity may also be increased due to elevated concentrations of IgG, IgA or IgM. Clinical findings that are related to hyperviscosity are reported to have plasma viscosities exceeding 5 mPa.s, corresponding to an IgM concentration of 3 g/dl or IgA concentration of 6 g/dl [20]. Cryoglobulins, which are immunoglobulins that precipitate at temperatures below 37° C, may also contribute to plasma viscosity alterations [21]. Cryoglobulins may be present in multiple myeloma, Waldenström’s macro-globulinemia and lymphoma, and can cause vasculitis in the clinical entity called mixed cryoglobulinemia [21]. The non-protein content of plasma is relatively less important as a determinant of the rheological behavior of plasma or blood. Although there are reports of significant correlations between blood viscosity at high shear rate and plasma triglycerides and cholesterol concentration [22], there is no strong evidence indicating that these lipids have a direct influence on rheology. However, it should be kept in mind that plasma is the natural environment for blood cells and, as such, its composition may affect the physiological status of RBC, thereby affecting their rheological behavior. More specifically, the concentrations of smaller organic and inorganic molecules (e.g., ionic content and osmotic/oncotic pressure, pH, nutrients) are among the factors affecting RBC behavior. It should be noted that the protein content of plasma is not only an important factor for determining the contribution of plasma viscosity to the overall hemodynamic resistance in circulation, but also has a strong impact on the rheological behavior of blood cells, especially RBC; this aspect is detailed later in this chapter.

174


2. Alterations in Red Blood Cell Properties As detailed in a previous chapter (Chapter II.3.a), the contribution of RBC to flow behavior of blood at various levels of circulatory system is closely related to cell deformability and aggregation. Cellular deformability is affected by intrinsic (e.g., cytoplasmic viscosity, membrane rigidity) and extrinsic properties (e.g., membrane surface area to volume ratio, cell shape), with aggregation influenced by both cellular factors and plasma properties. 2.1. Alterations in Red Blood Cell Deformability Deformability is the ability of RBC to undergo reversible shape changes in response to externally applied shear forces. This property is determined by geometric factors related to the cell’s special, biconcave-discoid shape, cytoplasmic viscosity reflecting cytoplasmic hemoglobin concentration, and the membrane skeletal protein network that determines the visco-elastic properties of the membrane [23]. Detailed analysis of these factors is presented in Chapter II.4.a. RBC deformability can be altered due to structural and/or functional changes resulting from genetic factors and/or “environmental” influences. Environmental influences in this case refer to the effects due to plasma alterations reflecting conditions at the tissue level, especially those in the microcirculation. 2.1.1. Hemoglobinopathies The cytoplasm of RBC can be regarded simply as a hemoglobin solution, with any alterations in the concentration and physical properties of this solution affecting RBC deformability. Cellular hemoglobin concentration depends strongly on the hydration status of the cell, and is closely regulated by membrane ion pumps that maintain osmotic/volumetric balance across the RBC membrane [24]. However, the molecular structure of hemoglobin is also important in determining the physical properties of the cytoplasm. The well-known example of this type of cytoplasmic alteration is sickle cell disease, in which hemoglobin S undergoes a physical change when deoxygenated, resulting in the formation of hemoglobin S fibers [25]. This drastic change in the solubility of hemoglobin results in greatly reduced deformability and leads to blockage of the microcirculation [26]. Another important group of hemoglobinopathies is thalassemias, characterized by genetic disorders that result in unmatched synthesis of globins. In addition to the anemia due to the defective globin synthesis, accumulation of excess, unmatched globin chains leads to increased oxidant damage in thalassemic RBC [27], resulting in reduced RBC deformability [28, 29]. It has been recently reported that L-carnitine deficiency is associated with E-thalasssemia [30, 31] and that the impairment of RBC deformability could be partly corrected with L-carnitine treatment [30]. There is limited published evidence for impaired RBC deformability with other abnormal hemoglobin variants, including hemoglobin C, D, E and a variety of unstable hemoglobins. However, it is clear that these hemoglobin variants are more prone to oxidant damage which can usually be manifested by the formation of hemichromes [32]. Therefore, deformability is likely to be impaired for RBC carrying abnormal hemoglobins.


175

2.1.2. Alterations of Red Blood Cell Membrane The RBC membrane is the only structure that regulates both cell volume and shape. Cation pumps located in RBC membrane maintain the physiological volume of RBC by regulating the osmotic balance and water movement across the membrane. The biconcave discoid shape of RBC is maintained by a special membrane skeletal network. This protein network also significantly contributes to the mechanical properties of the RBC membrane and hence affects deformability. These aspects of RBC physiology are described in more detail in Chapter II.4.a. 2.1.2.1. Genetic Modifications of Red Blood Cell Membrane Proteins Since the major role of the membrane skeletal protein network is maintaining the biconcave discoid shape, genetic modifications of RBC membrane proteins are characterized by altered cell shape. These genetic modifications are usually associated with hemolytic anemia and related clinical problems (e.g., icterus, splenomegaly) due to impaired membrane stability and RBC deformability [33]. The most common form of RBC genetic disorder is hereditary spherocytosis, which is a morphological manifestation of a wide variety of mutations affecting one or more membrane skeletal proteins. The affected proteins are reported to be band 3, D and E spectrins, ankyrin and band 4.2 [34, 35]. The shape change is due to the loss of membrane surface area, also resulting in impaired deformability (see Chapter II.4.a). Increased membrane fragility and membrane vesiculation are the underlying mechanism for membrane area shrinkage [34]. Hereditary elliptocytosis is characterized by cells with elliptical shapes and is associated with impaired mechanical stability. This impairment is thought to be related to a disruption in spectrin network integrity, possibly due to defects in spectrin, band 4.1 or glycophorin C [34]. Other genetic disorders of RBC membrane proteins include pyropoikilocytosis, characterized by extensive fragmentation of RBC, and ovalocytosis, associated with genetic modifications in band 3 [34]. Genetic modifications that lead to altered passive cation transport across the RBC membrane are also manifested by shape changes (e.g., stomatocytosis) [34], and RBC deformability is impaired in individuals with these genetic disorders. An osmotic deformability scan, which determines RBC deformability as a function of osmolality, is frequently used to monitor mechanical impairment in these cases [36]. 2.1.2.2. Modifications of RBC Membrane Proteins During Cell Life-span Membrane proteins can also be altered during the life-span of RBC by internal or external factors. The most widely encountered modification of RBC membrane proteins is due to free radical attack which, in turn, results from the disturbed balance between oxidant and anti-oxidant mechanisms (see section 2.1.4 of this chapter). Modifications of membrane proteins include cross-linking between spectrin subunits, spectrin and other membrane skeletal proteins (e.g., band 3, glycophorins) and also membrane skeletal proteins and hemoglobin [37, 38]. In particular, the cross linking within the membrane skeleton significantly affects RBC deformability, with reduced RBC deformability a common consequence of these cross linkages [37-39]. Oxidatively-damaged membrane proteins are susceptible to digestion by specialized enzymes in the RBC cytoplasm [40, 41], with altered cytosolic calcium concentration playing a role in activating proteolytic enzymes [42]. This activation can be regarded as a protective mechanism against further mechanical impairment [40], but

176


the damaged proteins cannot be replaced by new copies since the RBC lacks protein synthesizing mechanisms. 2.1.2.3. Alterations in RBC Membrane Lipid Composition RBC cannot synthesize lipids de novo, but lipids can be exchanged between the cell’s lipid bilayer and the plasma lipid pool [43]. It has been demonstrated that RBC membrane cholesterol correlates well with plasma cholesterol concentration [44]. However, compared to the major role played by the membrane skeletal network, membrane lipid composition has a very limited influence on RBC membrane mechanical properties, [45, 46]. For example, a fifty percent reduction of RBC membrane cholesterol content has been shown to have no influence on deformability of RBC from guinea pigs treated with lipid lowering agents [47]; conversely, hypercholesterolemia has been reported to impair RBC deformability [48]. 2.1.3. Alterations in Red Blood Cell Metabolism Despite its very simple structure, RBC have an active glycolytic metabolism. Glycolytic pathways serve to generate the ATP used mainly by the cation pumps located at the membrane. Glycolytic pathways are also important to maintain the cofactor pool (NADH and NADPH) that is used by the antioxidant mechanisms and enzymes that keep hemoglobin in a functional form. The factor that regulates the oxygen affinity of hemoglobin, 2,3 diphosphoglycerate (2,3 DPG), is also synthesized by an alternative pathway of glycolysis. About 90% of glucose entering RBC is utilized by the anaerobic, EmbdenMeyerhof pathway, including the 2,3 DPG synthesizing Rapoport-Luebering shunt. ATP, NADH and 2,3 DPG are the products of this pathway [49]. The remaining 10% goes to the aerobic pentose phosphate pathway, synthesizing NADPH, which is the cofactor of the glutathione reductase enzyme, a key element of the RBC antioxidant defense mechanism [43]. Glucose enters RBC by insulin-independent facilitated diffusion, and reduced glucose concentration in the environment of RBC results in a shortage of metabolic energy. For example, RBC metabolism can be significantly affected after prolonged storage of blood samples, and RBC volume and shape may be altered due to insufficient energy supplies for the cation pumps that maintain the physiological volume of RBC [50, 51]. The major mechanism that controls RBC volume is sodiumpotassium ATPase located on RBC membrane, and its impaired function due to ATP shortage results in increased sodium content and hence increased intracellular water. Another cation pump that can be affected is calcium ATPase, which maintains the low cytosolic calcium concentration of the RBC. Slowing of this pump results in increased cytosolic calcium concentrations, thereby tending to affect the organization of the membrane skeletal network, membrane mechanical properties and RBC deformability [52]. Oxidative stress may also be manifest under such conditions, as the synthesis of the co-factors NADH and NADPH would be impaired. The consequences of metabolic impairment in RBC can be observed in patients with genetic deficiencies of enzymes that take part in the glycolytic pathways. Genetic disorders related to Embden-Mayerhoff pathway may be related to various enzymes including hexokinase, pyruvate kinase, aldolase, glucose 6 phosphate isomerase, phosphofructokinase, and others [49]. A large number of mutations have been described for each of these enzymes [53]. These enzyme deficiencies are usually


177

characterized by hemolytic anemia of various degrees of severity; the life-span of RBC might be shortened [49] and RBC deformability has been reported to be altered [54]. The most widely known disorder related to the pentose phosphate pathway is glucose-6-phosphate dehydrogenase (G6PD) deficiency. G6PD is the key enzyme of this pathway which synthesizes NADPH from NADP, and insufficient generation of G6PD results in increased oxidative damage to RBC (see section 2.1.4, this chapter); reduced RBC deformability has been reported in G6PD enzyme deficiency [55]. 2.1.4. Oxidative Damage Although they are equipped with a powerful set of antioxidant defense mechanisms, RBC are susceptible to oxidative damage. Oxidative damage to both lipid and protein components of the RBC membrane may be responsible for mechanical deteriorations in many pathophysiological processes, including ischemia-reperfusion, inflammation, and various systemic diseases such as diabetes [56-58]. The RBC membrane is rich in polyunsaturated fatty acids, thus making it susceptible to oxidative damage. Exposure to high concentrations of oxygen and the existence of iron, a transitional metal, also increase the risk for oxidative damage. Reactions of hemoglobin with oxygen are continuous sources of free radicals [59]. When an oxygen molecule binds to the ferrous iron in hemoglobin, it shares an electron with iron, and normally this shared electron remains with iron during deoxygenation of hemoglobin. However, under certain conditions, an oxygen molecule can leave hemoglobin taking with it this shared electron: an oxygen free radical (superoxide) is thus formed and a ferric iron molecule is associated with hemoglobin (i.e., methemoglobin). The oxygen free radical is very reactive, and can readily react with biomolecules, lipids and proteins in RBC [59]. Furthermore, superoxide radicals may react with oxy or deoxy hemoglobin to generate new free radicals, including the extremely reactive hydroxyl radical. Free radicals start chain reactions when interacting with biomolecules: the products are new reactive free radicals and damaged biomolecules that are functionally impaired. Hemoglobin is the first RBC component to be affected by oxidant damage, with methemoglobin formation being the first step of oxidative damage to hemoglobin. More severe oxidant stress to hemoglobin may yield hemichromes that precipitate in the RBC cytoplasm, seriously affecting RBC mechanical properties [60]. Lipid peroxidation and protein oxidation are the consequences of oxidant damage, with both RBC membrane skeletal network proteins and membrane-bound or cytosolic enzymes susceptible to oxidation. RBC are well equipped with a set of antioxidant enzymes to defend themselves against oxidant-induced deterioration. Methemoglobin can be reduced to hemoglobin by the specific enzymes glutathione peroxidase and methemoglobin reductase [60]. Glutathione peroxidase uses reduced glutathione, which is generated by glutathione reductase using NADPH as a cofactor [60], and methemoglobin reductase needs NADH as a cofactor. Both of these cofactors are the products of glycolysis in the RBC, and therefore proper oxidative defense requires active metabolism within the RBC. This antioxidant defense system also includes superoxide dismutase, catalase and glutathione reductase [59]. Superoxide dismutase inactivates superoxide free radicals by generating hydrogen peroxide, which is, in turn, reduced to water by catalase or glutathione reductase. The degree of oxidative damage in RBC is thus determined by

178


the balance between the formation of oxygen free radicals and antioxidant defense mechanisms. Oxidant stress is one of the well documented mechanisms of RBC mechanical impairment [61], and there is now experimental evidence indicating that both internally and externally generated free radicals affect RBC properties [62]. It has been demonstrated that superoxide radicals that are generated in the cytoplasm of RBC are more effective in altering deformability [62], although externally generated superoxide may also affect RBC mechanical properties. The internal source of oxidants in RBC is usually the reactions of hemoglobin mentioned above, while external sources of oxidants include the hypoxanthine/xanthine oxidase system associated with ischemia/reperfusion injury [62], mitochondrial leakage [63] and activated leukocytes [39]. 2.1.5. Microenvironmental Influences RBC are exposed to a wide-variety of environmental conditions, frequently with extremes in terms of chemical and/or physical parameters, throughout their life-span in the circulatory system. These extreme conditions may include non-physiological osmotic pressure, high or low pH, and mechanical influences (e.g., shear forces). 2.1.5.1. Local Metabolic Changes As discussed in section 2.1.3, the RBC requires an active mechanism to maintain its biconcave discoid shape, to keep hemoglobin in the functional form, and to maintain adequate antioxidant defense. An insufficient supply of substrates such as glucose results in an impairment of mechanisms related to RBC metabolism. This impairment may not occur under normal in vivo circulatory conditions, yet prolonged storage of RBC suspensions may result in a metabolic insufficiency that leads to mechanical impairment [51, 64]: storage at lower temperatures (e.g., 4 °C) may slow this deterioration due to metabolic depletion. 2.1.5.2. Alterations in Osmotic Pressure RBC volume is primarily determined by intracellular cation and water content, with these two factors being osmotically linked to each other. Therefore, RBC volume is sensitive to alterations of osmotic pressure in their environment. Increased RBC volume due to lower osmotic pressure and hence osmotic water inflow results in a reduced surface area/volume ratio and reduced deformability. Conversely, increased osmotic pressure yields a decreased volume due to water outflow, resulting in an increased cytosolic hemoglobin concentration and viscosity and decreased deformability. However, it should be kept in mind that RBC volume changes due to osmotic changes are readily reversible. Therefore, alterations of RBC perfusing tissues with extreme osmotic pressures (e.g., renal medulla) are promptly reversed when the cells return to isotonic regions, and an alteration of RBC deformability can thus not be detected in the systemic circulation. Alternatively, if osmotic pressure is altered in a generalized fashion, RBC mechanical impairment can be detected in circulating blood. For example, severe diabetes may be manifest by a hyperosmolar environment, and RBC deformability impairment [65].


179

2.1.5.3. Alterations of pH The dependence of RBC membrane rigidity on the pH of the suspending medium has been demonstrated by micropipette aspiration studies [66]. The increased membrane rigidity at lower pH values is most likely related to altered relationships between membrane skeletal proteins due to protonation of negatively charged proteins [67]. However, pH alterations in the range that may have a pathophysiological meaning may have other influences on cell shape and volume [67], thereby affecting cytoplasmic viscosity. The overall response to an alteration of pH thus reflects the balance between the effects on membrane material properties and geometric alterations. In tissues/organs that are affected by pathophysiological processes such as ischemia, altered mechanical properties of RBC perfusing the affected tissue may be due, in part, to the altered pH of the microenvironment [68]. Alterations of pH in body fluids may also be an underlying cause of acute impairments of RBC deformability following strenuous exercise [69]. 2.1.5.4. Effect of Mechanical Factors RBC mechanical properties may be affected by shear forces acting on them during flow. Under normal physiological conditions, shear forces usually have a magnitude that is insufficient to damage the cell membrane. However, higher shear forces can be generated under non-physiological circulatory conditions. These conditions include flow in artificial circulatory environments such as artificial hearts and heart valves, circulatory assist devices, and extracorporeal circulation systems. Local shear forces may be high enough to disrupt the RBC membrane, causing hemolysis. The magnitude of shear stress that may cause RBC destruction has been reported to be in excess of 300 Pa [70], with shear stress levels below this magnitude inducing more subtle damage to RBC. Such sub-hemolytic trauma to RBC results in alterations of cellular metabolism [71, 72], membrane organization and ion transport across the membrane [73], and rheological properties [74]. Mechanical damage to RBC is covered in more detail in Chapter II.5.c. Another source of mechanical trauma to RBC is related to external mechanical impact on blood vessels. Such an effect might be manifest by various degrees of hemolysis related to repeated trauma, for example in blood vessels of feet during prolonged marching or running. This effect might contribute to the impairment of RBC deformability due to exercise in addition to the so called “sports anemia” [75]. 2.1.6. Others 2.1.6.1. Nitric Oxide Nitric oxide (NO) plays a significant role in vasomotor control, being synthesized in endothelial cells and inhibiting vascular smooth muscle tone [76]. Additionally, NO has been demonstrated to be synthesized in a variety of other cells, including RBC [77]. There is firm evidence indicating that either internally synthesized or externally generated NO plays a regulatory role in maintaining normal RBC mechanical properties [78]. However, the effect of NO is dose dependent and impairment of RBC deformability and alterations in cell shape accompanied with oxidative damage was reported in RBC exposed to concentrations higher than 10-5 M [77, 79]. The regulatory effect of NO in RBC has been demonstrated to be only partly dependent on guanylate cyclase [78], and there is evidence for an effect mediated by

180


potassium conductance across the RBC membrane [74, 80]. Endogenous inhibitors of NO synthases may be increased in a variety of clinical disorders, including hypercholesterolemia and ischemic vascular diseases [81]. Asymmetric dimethylarginine (ADMA), a potent, endogenous inhibitor of NO synthesis, has been proposed to impair RBC deformability [82]. 2.1.6.2. Other Endocrine-Paracrine Effects on RBC Deformability The effects of insulin on hemorheological parameters have been widely investigated due to the frequently observed hemorheological abnormalities in diabetic patients. While diabetes is usually associated with various degrees of vascular disorders that may explain the hemorheological impairments, it has also been reported that impairment in RBC deformability precedes vascular alterations in experimental diabetes, suggesting a more direct relationship between hormonal changes in diabetes and RBC mechanics [83]. Although no direct effects of insulin, glucagon and Cpeptide on blood viscosity at various shear rates have been observed [84], Kunt et al. demonstrated that human proinsulin C peptide restores RBC deformability in blood samples obtained from type I diabetics [85]. Juhan also reported increased RBC deformability with insulin for diabetic blood samples [86]. Alternatively, a study using an animal model of obesity and diabetes did not support this finding [87]. In brief, the literature data regarding the role of insulin in maintaining RBC mechanical properties are not conclusive. It has been reported that adrenaline affects RBC deformability in vitro [88]. RBC membrane fluidity and membrane acetylcholinesterase activity were also reported to be altered. Interestingly, the effects of adrenaline were found to be determined by the gender of the blood donor [88]. 2.1.6.3. Parasitic infections RBC infected with parasites, such as Falciparum Malaria, exhibit altered mechanical properties. Chapter II.4.a presents a detailed description of such alterations. 2.2. Alterations in Red Blood Cell Aggregation RBC aggregation is affected by both suspending phase and cellular properties, as detailed in Chapter II.4.b. Both suspending phase and cellular properties that are relevant to RBC aggregation have been demonstrated to be altered during physiopathological processes (Figure 1). Alterations in cellular properties affecting aggregation can be demonstrated by measuring aggregation of RBC re-suspended in a standard aggregating medium (e.g., aqueous solutions of high molecular weight dextran). The aggregation behavior of cells in such a defined polymer solution, reflecting the intrinsic properties of RBC that determine its tendency for aggregation, has been termed red cell “aggregability” [89]. 2.2.1. Alterations in Plasma Composition RBC aggregation occurs only if RBC are suspended in solutions containing macromolecules of sufficient molecular weight and hydrated size at appropriate concentration. Thus, the concentration of such macromolecules is a major determinant of the degree of RBC aggregation. Fibrinogen is the most important macromolecule of this kind in plasma, and strong correlations between plasma fibrinogen concentration


181

and the extent of RBC aggregation have been demonstrated [90]. Plasma fibrinogen is the major acute phase reactant and both its plasma level and RBC aggregation increase significantly during inflammatory processes [19]; fibrinogen concentrations can increase by 5-fold during acute phase reactions. Fibrinogen concentration affects various parameters related to RBC aggregation, including aggregate size (i.e., number of RBC per aggregate), yield stress, low-shear viscosity of RBC suspensions and erythrocyte sedimentation rate [90]. 12

Aggregation index

10

*

Septic

Control

*

8 6 4 2 0 Plasma

0.5% Dextran 500

Figure 1. RBC aggregation indexes measured in autologous plasma and 0.5% dextran 500 (MW: 500 kD) in control and septic rats. Sepsis was experimentally induced by cecal ligation-puncture and the animals were sampled 18 hours after the operation. Aggregation indexes measured in both plasma and standard dextran medium were significantly higher for septic blood samples, with the aggregation in dextran indicating the higher aggregability of RBC from septic animals (From [91]).

There is now increasing evidence for the heterogeneity of the fibrinogen molecule, and more than ten sub-fractions of fibrinogen have been demonstrated [92]. It has also been demonstrated that the relative proportion of these sub-fractions in the plasma fibrinogen pool may be changed during acute phase reactions, thereby affecting the average size and charge distribution of fibrinogen molecules. Therefore, not only the total concentration of fibrinogen might be altered during inflammatory processes, but also its effectiveness in RBC aggregation might be changed [92, 93]. Further studies are needed to clarify this aspect of the acute phase response. Other macromolecules such as D2 macroglobulin, IgM and IgG have also been demonstrated to affect RBC aggregation in plasma, but their effects on RBC aggregation were only evident at concentrations much higher than physiological levels [90]. 2.2.2. Alterations in Red Blood Cell Properties Aggregation of RBC in a given suspending medium is mainly determined by the cell’s surface properties. However, cellular shape and deformability also affect the extent of aggregation, such that any change from the normal biconcave discoid shape and normal deformability may interfere with physiological rouleaux aggregate formation.

182


2.2.2.1. Alterations in Red Blood Cell Surface Properties The role of RBC surface properties in aggregation has been studied extensively. The aim of this chapter is not to discuss this aspect of RBC aggregation in detail since a complete discussion of this area is presented in Chapter II.4.b. It has been demonstrated that RBC treated with proteolytic enzymes exhibited altered aggregation properties [94]. There is also experimental evidence indicating that oxygen free radicals can alter cellular properties of RBC, resulting in modified aggregability [62]. It is important to note that, compared to internally generated radicals, oxygen free radicals that are generated outside of the RBC are more effective in inducing alterations of aggregability [62]. These effects could be explained based upon a modification of RBC surface charge density [94]. Other possible surface targets for such treatments were proposed to be binding sites of aggregating macromolecules [62], with this suggestion based on the bridging hypothesis for RBC aggregation. However, recent reports have focused more on alterations of the RBC glycocalyx and modification of the macromolecule depletion layer as the mechanism responsible for the modified aggregability [91, 95-97]. Altered RBC aggregability has been reported in various clinical disorders as well as in animal models of disease processes. Inflammatory conditions that induce the acute phase reaction also affect RBC aggregability [98]: enhanced aggregation of RBC in standard suspending media has been reported in patients with thalassemia major [99], type II diabetes [100], obesity [101] and schizophrenia [102]. Investigations of usual biochemical and hematological parameters usually fail to reveal any specific alteration that predicts higher RBC aggregability in such clinical disorders. It has been suggested that altered RBC membrane lipid composition might be the underlying mechanism of the alterations in aggregability, as well as modified sialic acid content of the membrane surface [102]. Increased aggregability has been reported in a rat model of fecal sepsis induced by cecal ligation-puncture [91, 103]. Surface properties of RBC were investigated by either studying behavior in a two-phase, charge sensitive partitioning system [103]or by cell electrophoresis [91]. Although partition coefficients were found to be significantly decreased in the earlier study [103], suggesting decreased surface charge density, electrophoretic mobility measurements of septic and control RBC in isotonic phosphate buffer indicated no difference in surface charge density [91]. However, the electrophoretic behavior of septic and control RBC in 500 kDa dextran solutions indicated significant differences: the sensitivity of electrophoretic mobility to dextran concentration was significantly blunted for septic RBC, suggesting a more effective depletion of macromolecules from the RBC surface [91]. While it now seems certain that the depletion phenomena is the primary mechanism for RBC aggregation [104], and while there is now convincing experimental evidence indicating that differences of the depletion layer at the surface of RBC play an important role in RBC aggregation, the mechanisms of its alteration in pathophysiological processes remain uncertain. 2.2.2.2. Alterations in Red Blood Cell Shape and Deformability It is obvious that an extensive contact between the membranes of adjacent RBC is required during aggregation and that this contact is significantly facilitated by cellular deformability. Marked increases of RBC rigidity induced by treatment with glutaraldehyde at 0.01-0.02% concentrations are associated with inhibition of aggregation [105]. However, the degree of alteration of RBC deformability resulting


183

from such interventions is usually much greater than the impairments observed during clinical disorders. Most clinical studies fail to demonstrate a correlation between RBC deformability changes and altered aggregation. This lack of correlation may be due to the multiple pathophysiological processes that affect both RBC deformability and aggregability, hence masking the role of impaired deformability by enhanced aggregability. It has also been demonstrated that RBC shape is an important determinant of RBC aggregation. Experimental data indicate that any shift from normal-biconcave discoid shape towards either the echinocyte or stomatocyte form is associated with decreased aggregation [106]. Such shape changes tend to limit cell-to-cell membrane contact, thus reducing RBC aggregation and RBC aggregability. 2.2.3. Pathological Aggregation Physiological RBC aggregation is characterized by the formation of multi-cell structures, initially linear face-to-face, “stack of coin”t forms called rouleaux, followed by clustering and branching of rouleaux. However, in blood samples with enhanced aggregation due to certain pathophysiological processes, this morphology of RBC aggregates can be changed to more irregular shapes or clumps where the regular rouleaux formation is less evident. “Pathologic aggregates” are known to occur in diabetes, myocardial infarction, nephrotic syndrome and after prolonged storage of blood samples. Although there is no obvious evidence for separate mechanisms of aggregation under physiological and pathological conditions, it has been demonstrated that even RBC with very marked echinocytic transformation can participate in RBC aggregation when in strongly aggregating plasma [107, 108]. Pathological RBC aggregation is often characterized by increased aggregate strength, while parameters reflecting aggregate size (e.g., number of RBC per aggregate) can be unaltered or even decreased [62, 98]. Human RBC exposed to superoxide radical generated by xanthine oxidase in the presence of hypoxanthine exhibit decreased aggregation in autologous plasma and in 3% dextran 70, while disaggregation shear rate is significantly increased in both suspending media [62]. Further research is needed to clarify the basic differences in the characteristics of physiological and pathological RBC aggregation.

3. Alterations of Leukocytes Leukocytes are often ignored in discussions of hemorheological concepts, mostly due to their negligible contribution to the whole cell population of peripheral blood. The ratio of leukocytes to RBC is normally in the order of 1/1000, and thus the presence of leukocytes can usually be ignored when considering macroscopic, bulk flow conditions. However, leukocytes represent a highly variable cellular population, both in terms of quantity and quality of these cells. In addition to the relatively rapid alterations in the quantity of leukocytes during certain pathophysiological processes (e.g., inflammation), these cells undergo an activation process during which important structural and functional alterations take place [109]. The aim of this chapter is not to describe this activation process in detail, but rather to focus on the influence of activated leukocytes on neighboring RBC.

184


3.1. Leukocyte Activation and Microcirculatory Function The activation process of leukocytes leads to morphological and mechanical alterations [110]. This aspect of leukocyte activation has been covered extensively in Chapter II.4.c, and therefore will not be treated in this chapter. 3.2. Leukocyte-Induced Alteration of Red Blood Cell Rheology An important aspect of leukocyte activation is the enhanced secretion of various mediators and chemicals from the cell [109]. The secreted chemicals include free radicals and proteolytic enzymes that are normally useful for destroying and/or inactivating the agents causing the inflammatory process (e.g., microorganisms). However, these secretions of activated leukocytes also attack neighboring cells/tissues, with RBC being among the cells most prone to such an attack. RBC deformability is impaired and aggregability is enhanced for RBC suspensions incubated with polymorphonuclear leukocytes that were activated in vitro by TNF-D and fMLP [111]. Both alterations in deformability and aggregation were prevented by including a protease inhibitor or antioxidant enzymes (superoxide dismutase and catalase) in the suspension of activated leukocytes and RBC [111]. A similar effect on RBC deformability has been shown for leukocytes isolated from septic rats [39]. Leukocytes isolated from rats with sepsis induced by cecal ligation-puncture were activated as judged by increased elastase and phagocytic activity [39], with lipid peroxidation reported to be increased in RBC incubated with these leukocytes. Furthermore, SDSPAGE electrophoresis of RBC membrane proteins revealed widening of band 1 [39], suggesting crosslinking of hemoglobin with spectrin, and a finding similar to that demonstrated for RBC under oxidative stress [37]. Leukocyte activation is a common consequence of a variety of pathological and even physiological (e.g., exercise) processes. The effects induced in RBC by activated leukocytes may therefore explain the mechanisms of hemorheological alterations in many pathophysiological conditions [69, 112].

4. Hemorheological Vicious Cycle The relationship between hemorheological alterations and blood flow has been discussed in various chapters of this book. It is obvious from the discussions dealing with the factors affecting blood fluidity that these factors are closely related to the maintenance of homeostasis in the internal environment and hence to the environment of blood. Based on basic physiological concepts, it is also quite obvious that the maintenance of this internal homeostasis depends on an appropriate blood supply. These three closely related points can result in a hemorheological “vicious cycle” [113] as shown in Figure 2. Frequently, circulatory insufficiency may initially be confined to a small local area. This local circulatory problem is followed by local disturbances of homeostasis, such as hypoxia, acidosis, lack of required metabolites and accumulation of metabolic wastes. As discussed above, RBC properties are sensitive to such changes. Additionally, tissue injury following these local disturbances of homeostasis may affect RBC rheological behavior via an acute phase reaction and/or leukocyte activation. The acute phase reaction also affects plasma viscosity, resulting in increased

185


blood viscosity together with decreased RBC deformability and enhanced RBC aggregation. Since blood viscosity is an important determinant of flow resistance, increased blood viscosity would tend to worsen circulatory insufficiency. It is important to note that, under normal conditions, vasomotor control mechanisms tend to compensate for any hemorheological alterations, thereby maintaining adequate blood supply [114]. However, if a vascular problem underlies the starting point of this vicious cycle, vascular autoregulatory mechanisms may not be able to compensate for blood rheology abnormalities [23]. Vasomotor Control Local hypoxia

?

Local acidosis

Blood flow

Circulatory insufficiency

Metabolite debt

Waste accumulation

Blood viscosity Tissue injury Hematocrit

Dehydration

Plasma Viscosity

Inflammation

Acute phase reaction

Leukocyte activation

Fibrinogen

RBC aggregation

RBC deformability

Figure 2. Hemorheological vicious cycle (See the text for explanation).

References [1] [2] [3] [4] [5] [6] [7]

P.F. Leblond. Hemorheology and blood diseases. In: Clinical Hemorheology, S.Chien, J.Dormandy, E.Ernst and A.Matrai, Eds., Martinus Nijhoff Publishers, Dordrecht/Boston/Lancaster, 1987, pp. 227254. T. Somer and H.J. Meiselman, Disorders of blood viscosity, Ann. Med. 25 (1993), 31-39. J.F. Stoltz, Hemorheology: pathophysiological significance, Acta Med. Port. 6 (1985), S4-S13. A.R. Pries and T.W. Secomb, Microvascular blood viscosity in vivo and the endothelial surface layer, Am. J. Physiol. 289 (2005), H2657-H2664. O.K. Baskurt, O. Yalcin, F. Gungor and H.J. Meiselman, Hemorheological parameters as determinants of myocardial tissue hematocrit values, Clin. Hemorheol.Microcirc. 35 (2006), 45-50. L. Bogar, I. Juricskay, G. Kesmarky, P. Kenyeres and K. Toth, Erythrocyte transport efficacy of human blood: a rheological point of view, Eur. J. Clin. Invest. 35 (2005), 687-690. M.M. Wintrobe, G.R. Lee, D.R. Boggs, T.C. Bithell, J. Foerster, J.W. Athens and J.N. Lukens. Erythropoiesis. In: Clinical Hemotology, Lea-Febiger, Philadelphia, 1981, pp. 108-135.

186 [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25]

[26] [27]

[28] [29] [30] [31] [32] [33] [34] [35]

O.K. Baskurt / Mechanisms of Blood Rheology Alterations H. Kwaan and J. Wang, Hyperviscosity in polycythemia vera and other red cell abnormalities, Semin. Thromb. Hemost. 29 (2003), 451-458. J.P. Isbister, The stress polycythaemia syndromes and their haemorheological significance, Clin. Hemorheol. 7 (1987), 159-179. M. Messinezy and T.C. Pearson, The classification and diagnostic criteria of the erythrocytoses (polycythaemias), Clin. Lab. Haematol. 21 (1999), 309-316. M. Cao, R.J. Olsen and Y. Zu, Polycythemia vera: new clinicopathologic perspectives, Arch. Pathol. Lab. Med. 130 (2006), 1126-1132. A. Tefferi and J.L. Spivak, Polycythemia vera: scientific advances and current practice, Semin. Hematol. 42 (2005), 206-220. P. Mossuz, F. Girodon, M. Donnard, V. Latger-Cannard, I. Dobo, N. Boiret, J.C. Lecron, C. Binquet, C. Barro, S. Hermouet and V. Praloran, Diagnostic value of serum erythropoietin level in patients with absolute erythrocytosis, Haematologica 89 (2004), 1194-1198. D.H. Hild and T.J. Myers, Hyperviscosity in chronic granulocytic-leukemia, Cancer 46 (1980), 14181421. T.-H. Young, J. Cuttner, R. Meyer and S. Chien, Leukaphresis reduced blood viscosity in chronic myelogenous leukemia patients with hyperleukocytosis, Clin. Hemorheol. 10 (1990), 191-203. M.A. Lichtman and J.M. Rowe, Hyperleukocytic leukemias: rheological, clinical, and therapeutic considerations, Blood 60 (1982), 279-283. M. Rampling, Hyperviscosity as a complication in a variety of disorders, Semin. Thromb. Hemost. 29 (2003), 459-466. G.D.O. Lowe and J.C. Barbenel. Plasma and blood viscosity. In: Clinical blood rheology, G.D.O.Lowe, Ed., CRC Press,Inc., Florida, 1988, pp. 1-10. J.T. Whicher, Cytokines and acute phase proteins, Rev. Port. Hemorreol. 5 (1990), 49-56. J. Mehta and S. Singhal, Hyperviscosity syndrome in plasma cell dyscrasias, Semin. Thromb. Hemost. 29 (2003), 467-471. A. Della Rossa, A. Tavoni and S. Bombardieri, Hyperviscosity syndrome in cryoglobulinemia: clinical aspects and therapeutic considerations, Semin. Thromb. Hemost. 29 (2003), 473-477. R.S. Rosenson, S. Shott and C.C. Tangney, Hypertriglyceridemia is associated with an elevated blood viscosity Rosenson: triglycerides and blood viscosity, Atherosclerosis 161 (2002), 433-439. S. Chien, Red cell deformability and its relevance to blood flow, Ann. Rev. Physiol. 49 (1987), 177-192. N. Mohandas and S.B. Shohet, The role of membrane-associated enzymes in regulation of erythrocyte shape and deformability, Clin. Hematol. 10 (1981), 223-237. E. Beutler. The sickle cell diseases and related abnormalities. In: Williams Hematology, E. Beutler, M.A. Lichtman, B.S. Coller, T.J. Kipps and U. Seligsohn, Eds., McGraw-Hill, New York, 2001, pp. 581-605. D.K. Kaul, Microvascular flow behavior of red cells in sickle cell anemia, Clin. Hemorheol. 12 (1992), 191-202. M.D. Scott, J.J. van den Berg, T. Repka, P. Rouyer-Fessard, R.P. Hebbel, Y. Beuzard and B.H. Lubin, Effect of excess alpha-hemoglobin chains on cellular and membrane oxidation in model betathalassemic erythrocytes, J. Clin. Invest. 91 (1993), 1706-1712. A.M. Dundorp, K.T. Chotivanich, S. Fucharoen, K. Silamut, J. Vreeken, P.A. Kager and N.J. White, Red cell deformability, splenic function and anaemia in thalassemia, Br. J. Haematol. 105 (1999), 505508. A. Vaya, J. Iborra, C. Falco, I. Moreno, P. Bolufer, F. Ferrando, M.L. Perez and A. Justo, Rheological behaviour of red blood cells in beta and deltabeta thalassemia trait., Clin. Hemorheol. Microcirc. 28 (2003), 71-78. B. Toptas, A. Baykal, A. Yesilipek, M. Isbir, A. Kupesiz, O. Yalcin and K. Baskurt, L-carnitine deficiency and red blood cell mechanical impairment in beta-thalassemia major, Clin. Hemorheol. Microcirc. 35 (2006), 349-357. V. Tsagris and G. Liapi-Adamidou, Serum carnitine levels in patients with homozygous beta thalassemia: a possible new role for carnitine? Eur. J. Pediat. 164 (2005), 131-134. E. Beutler. Hemoglobinopathies associated with unstable hemoglobin. In: Williams Hematology, E.Beutler, M.A.Lichtman, B.S.Coller, T.J.Kipps and U.Seligsohn, Eds., McGraw Hill, New York, 2001, pp. 607-610. J. Delaunay, The molecular basis of hereditary red cell membrane disorders, Blood Reviews 21 (2007), 1-20. P.G. Gallagher, Red Cell Membrane Disorders, Hematology 2005 (2005), 13-18. A. Iolascon, E.M. del Giudice, S. Perrotta, N. Alloisio, L. Morle and J. Delaunay, Hereditary spherocytosis: from clinical to molecular defects, Haematologica 83 (1998), 240-257.


187

[36] R.M. Johnson and Y. Ravindranath, Osmotic scan ektacytometry in clinical diagnosis, J. Ped. Hematol. Oncol. 18 (1996), 122-129. [37] L.M. Snyder, N. Fortier, J. Trainor, J. Jacobs, L. Leb, S. Lubin, D. Chiu, S. Shohet and N. Mohandas, Effect of hydrogen peroxide exposure on normal human erythrocyte deformability, morphology, surface characteristics, and spectrin-hemoglobin cross-linking, J. Clin. Invest. 76 (1985), 1971-1977. [38] N. Uyesaka, S. Hasegawa, N. Ishioka, R. Ishioka, H. Shio and A.N. Schechter, Effect of superoxide anions on red cell deformability and membrane proteins, Biorheology 29 (1992), 217-229. [39] O.K. Baskurt, Activated granulocyte induced alterations in red blood cells and protection by antioxidant enzymes, Clin. Hemorheol. 16 (1996), 49-56. [40] M. Beppu, M. Inoue, T. Ishikawa and K. Kukugawa, Presence of membrane-bound proteinases that preferentially degrade oxidatively damaged erythrocyte membrane proteins as secondary antioxidant defense, Biochim. Biophys. Acta 1196 (1994), 81-87. [41] K.J.A. Davies and A.L. Goldberg, Oxygen radicals stimulate intracellular proteolysis and lipid peroxidation by independent mechanisms in erythrocytes, J. Biol. Chem. 262 (1987), 8220-8226. [42] M. Shields, P.L. La Celle, R.E. Waugh, M. Scholz, R. Peters and H. Passow, Effects of intracellular Ca2+ and proteolytic digestion of the membrane skeleton on the mechanical properties of the red blood cell membrane, Biochim. Biophys. Acta 905 (1987), 181-194. [43] M.M. Wintrobe, G.R. Lee, D.R. Boggs, T.C. Bithell, J. Foerster, J.W. Athens and J.N. Lukens. The mature erythrocyte, In: Clinical Hematology, Lea-Febiger, Philadelphia, 1981, pp. 75-107. [44] P. Lijnen, H. Celis, R. Fagard, J. Staessen and A. Amery, Influence of cholesterol lowering on plasma membrane lipids and cationic transport systems, J. Hypertension 12 (1994), 59-64. [45] R.M. Hochmuth and R.E. Waugh, Erythrocyte membrane elasticity and viscosity, Ann. Rev. Physiol. 49 (1987), 209-219. [46] N. Mohandas, J.A. Chasis and S.B. Shohet, The influence of membrane skeleton on red cell deformability, membrane material properties, and shape, Semin. Hematol. 20 (1983), 225-242. [47] M. Uyuklu, H.J. Meiselman and O.K. Baskurt, Effect of decreased plasma cholesterol by atorvastation treatment on erythrocyte mechanical properties., Clin. Hemorheol. Microcirc. 36 (2006), 25-33. [48] A. Vaya, M. Martinez, P. Solves, J.L. Barbera and J. Aznar, Red blood cell deformability determined by the Rheodyn SSD in familial hypercholesterolemia, Clin. Hemorheol. 16 (1996), 515-522. [49] R. van Wijk and W.W. van Solinge, The energy-less red blood cell is lost: erythrocyte enzyme abnormalities of glycolysis, Blood 106 (2005), 4034-4042. [50] T. Sollberger, R. Walter, B. Brand, J. Contesse, D.O. Meredith and W.H. Reinhart, Influence of prestorage leucocyte depletion and storage time on rheologic properties of erythrocyte concentrates, Vox Sanguinis 82 (2002), 191-197. [51] V. Nagaprasad and M. Singh, Sequential analysis of the influence of blood storage on aggregation, deformability and shape parameters of erythrocytes, Clin. Hemorheol. Microcirc. 18 (1998), 273-284. [52] E. Friederichs, R.A. Farley and H.J. Meiselman, Influence of calcium permeabilization and membraneattached hemoglobin on erythrocyte deformability, Am. J. Hematol. 41 (1992), 170-177. [53] A. Zanella, E. Fermo, P. Bianchi and G. Valentini, Red cell pyruvate kinase deficiency: molecular and clinical aspects, Br. J. Haematol. 130 (2005), 11-25. [54] M. Lakomek and H. Winkler, Erythrocyte pyruvate kinase- and glucose phosphate isomerase deficiency: Perturbation of glycolysis by structural defects and functional alterations of defective enzymes and its relation to the clinical severity of chronic hemolytic anemia, Biophys. Chem. 66 (1997), 269-284. [55] N. Gurbuz, O. Yalcin, T.A. Aksu and O.K. Baskurt, The relationship between the enzyme activity, lipid peroxidation and red blood cells deformability in hemizygous and heterozygous glucose-6-phosphate dehydrogenase deficient individuals, Clin. Hemorheol. Microcirc. 31 (2004), 235-242. [56] N. Nemeth, T. Lesznyak, M. Szokoly, I. Furka and I. Miko, Allopurinol prevents erythrocyte deformability impairing but not the hematological alterations after limb ischemia-reperfusion in rats, J. Invest. Surg. 19 (2006), 47-56. [57] S.M.S. Reshamwala and N.D. Patil, Biochemical changes in erythrocyte membrane in uncontrolled type 2 diabetes mellitus, Ind. J. Biochem. Biophys. 42 (2005), 250-253. [58] L. Czopf, R. Halmosi, G. Kesmarky, T. Habon, K. Toth, I. Juricskay, E. Roth and G. Mozsik, Lovastatin and nitrate therapy induced changes in hemorheological parameters and in free radical mediated processes in patients with ischemic heart disease, Perfusion 12 (1999), 50-57. [59] O.K. Baskurt and S. Yavuzer. Some hematological effects of oxidants. In: Environmental oxidants, J.O.Nriagu and M.S.Simmons, Eds., John Wiley and Sons. Inc., New York/Chichester/Brisbane/ Toronto/Singapore , 1994, pp. 405-423. [60] M.M. Wintrobe, G.R. Lee, D.R. Boggs, T.C. Bithell, J. Foerster, J.W. Athens and J.N. Lukens. Methemoglobinemia and other disorders usually accompanied by cyanosis. In: Clinical Hematology, Lea-Febiger, Philadelphia , 1981, pp. 1011-1020.

188


[61] L.M. Snyder, N. Sauberman, H. Condara, J. Dolan, J. Jacobs, I. Szymanski and N.L. Fortier, Red cell membrane response to hydrogen peroxide-sensitivity in hereditary xerocytosis and in other abnormal red cells, Br. J. Hematol. 48 (1981), 435-444. [62] O.K. Baskurt, A. Temiz and H.J. Meiselman, Effect of superoxide anions on red blood cell rheologic properties, Free Radic. Biol. Med. 24 (1998), 102-110. [63] L.L. Ji and S. Leichtweis, Exercise and oxidative stress: Sources of free radicals and their impact on antioxidant systems, Age 20 (1997), 91-106. [64] H.J. Meiselman and R.F. Baker, Flow behavior of ATP-depleted human erythrocytes, Biorheology 14 (1977), 111-126. [65] B.W. Nugent, Hyperosmolar hyperglycemic state, Emerg. Med. Clin. North Am. 23 (2005), 629-651. [66] E.D. Crandall, A.M. Critz, A.S. Osher, D.J. Keljo and R.E. Forster, Influence of Ph on Elastic Deformability of Human Erythrocyte-Membrane, Am. J. Physiol. 235 (1978), C269-C278. [67] D. Kuzman, T. Znidarcic, M. Gros, S. Vrhovec, S. Svetina and B. Zeks, Effect of pH on red blood cell deformability, Pflugers Archiv- Eur. J. Physiol. 440 (2000), R193-R194. [68] E. Kayar, F. Mat, H.J. Meiselman and O.K. Baskurt, Red blood cell rheological alterations in a rat model of ischemia-reperfusion injury, Biorheology 38 (2001), 405-414. [69] O. Yalcin, A. Erman, S. Muratli, M. Bor-Kucukatay and O.K. Baskurt, Time course of hemorheological alterations after heavy anaerobic exercise in untrained human subjects, J. Appl. Physiol. 94 (2003), 9971002. [70] C.G. Nevaril, E.C. Lynch, C.P. Alfrey Jr and J.D. Hellums, Erythrocyte damage and destruction induced by shear stress, J. Lab. Clin. Med. 71 (1968), 784-790. [71] F. Cavalieri, F. Meo, A. Scapigliati, M. Sciarra and R. Schiavello, Red blood cell energy metabolism during cardiopulmonary bypass, J. Cardiovasc. Surg. (Torino) 40 (1999), 653-657. [72] M. Kodicek, J. Suttnar, L. Mircevova and T. Marik, Red blood cells under mechanical stress, Gen. Physiol. Biophys. 9 (1990), 291-299. [73] K.M. Dao, E.A. O'Rear, A.E. Johnson and S.E. Peitersen, Sensitivity of the erythrocyte membrane bilayer to subhemolytic mechanical trauma as detected by fluorescence anisotropy, Biorheology 31 (1994), 69-76. [74] O.K. Baskurt, M. Uyuklu and H.J. Meiselman, Protection of erythrocytes from sub-hemolytic mechanical damage by nitric oxide mediated inhibition of potassium leakage, Biorheology 41 (2004), 79-89. [75] R.D. Telford, G.J. Sly, A.G. Hahn, R.B. Cunningham, C. Bryant and J.A. Smith, Footstrike is the major cause of hemolysis during running, J. Appl. Physiol. 94 (2003), 38-42. [76] L.J. Ignarro, G. Cirino, A. Casini and C. Napoli, Nitric oxide as a signaling molecule in the vascular system: an overview., J. Cardiovasc. Pharmacol. 34 (1999), 879-886. [77] P. Kleinbongard, R. Schutz, T. Rassaf, T. Lauer, A. Dejam, T. Jax, I. Kumara, P. Gharini, S. Kabanova, B. Ozuyaman, H.-G. Schnurch, A. Godecke, A.-A. Weber, M. Robenek, H. Robenek, W. Bloch, P. Rosen and M. Kelm, Red blood cells express a functional endothelial nitric oxide synthase, Blood 107 (2006), 2943-2951. [78] M. Bor-Kucukatay, R.B. Wenby, H.J. Meiselman and O.K. Baskurt, Effects of nitric oxide on red blood cell deformability, Am. J. Physiol. 284 (2003), H1577-H1584. [79] F.A. Carvalho, A.V. Maria, J.M. Braz Nogueira, J. Guerra, J. Martins-Silva and C. Saldanha, The relation between the erythrocyte nitric oxide and hemorheological parameters, Clin. Hemorheol. Microcirc. 35 (2006), 341-347. [80] A. Caramelo, J. Riesco, I. Outeirino, G. Milas, B. Blum, V. Monzu, L. Casado, J.R. Sanchez, S. Mosquera, S. Casado and A.L. Fare, Effects of nitric oxide on red blood cell: changes in erythrocyte resistance to hypotonic hemolysis and potassium efflux by experimental maneuvers that decrease nitric oxide, Biochem. Biophys. Res. Com. 15 (1994), 47-54. [81] V. Turchetti, L. Boschi, G. Donati, L. Trabalzini and S. Forconi, Impact of hemorheological and endothelial factors on microcirculation, Adv. Exp. Med. Biol. 578 (2006), 107-112. [82] K. Tsuda and I. Nishio, An association between plasma asymmetric dimethylarginine and membrane fluidity of erythrocytes in hypertensive and normotensive men - An electron paramagnetic resonance investigation, Am. J. Hypertens. 18 (2005), 1243-1248. [83] E.J. Diamantopoulos, C. Kittas, D. Charitos, M. Grigoriadou, G. Ifanti and S.A. Raptis, Impaired erythrocyte deformability precedes vascular changes in experimental diabetes mellitus, Hormone Metab. Res. 36 (2004), 142-147. [84] L. Schnyder, R. Walter, A. Rohrer and W.H. Reinhart, No influence of C-peptide, insulin, and glucagon on blood viscosity in vitro in healthy humans and patients with diabetes mellitus, Clin. Hemorheol. Microcirc. 24 (2001), 65-74.


189

[85] T. Kunt, S. Schneider, A. Pfutzner, K. Goitum, M. Engelbach, B. Schauf, J. Beyer and T. Forst, The effect of human proinsulin C-peptide on erythrocyte deformability in patients with Type I diabetes mellitus, Diabetologia 42 (1999), 465-471. [86] I. Juhan, P. Vague, M. Buonocore, J.P. Moulin, M.F. Calas, B. Vialettes and J.J. Verdott, Effects of insulin on erythrocyte deformability in diabetic-relationship between erythrocyte deformability and platelet aggregation, Scand. J. Clin. Lab. Invest. 156 (1981), 159-164. [87] L. Cinara, A. Bollini, M.C. Gayol and G.N. Hernandez, In vitro effect of insulin on rats erythrocytes rheological behaviour, Clin. Hemorheol. Microcirc. 35 (2006), 367-373. [88] S. Hilario, C. Saldanha and J.M.E. Silva, An in vitro study of adrenaline effect on human erythrocyte properties in both gender, Clin. Hemorheol. Microcirc. 28 (2003), 89-98. [89] O.K. Baskurt, M. Bor-Kucukatay, O. Yalcin, H.J. Meiselman and J.K. Armstrong, Standard aggregating media to test the "aggregability" of rat red blood cells, Clin. Hemorheol. Microcirc. 22 (2000), 161-166. [90] M.W. Rampling. Red cell aggregation and yield stress. In: Clinical blood rheology, G.D.O.Lowe, Ed., CRC Press, Inc., Florida , 1988, pp. 45-64. [91] O.K. Baskurt, E. Tugral, B. Neu and H.J. Meiselman, Particle electrophoresis as a tool to understand the aggregation behavior of red blood cells, Electrophoresis 23 (2002), 2103-2109. [92] M.W. Rampling, Haemorheology and the inflammatory process, Clin. Hemorheol. Microcirc. 19 (1998), 129-132. [93] M. Musielak and J. Koscielak, Different RBC aggregating properties of the Aa2Bb2gA2 and Aa2Bb2gAg' subpopulations of human fibrinogen, Clin. Hemorheol. Microcirc. 35 (2006), 303. [94] M.W. Rampling, H.J. Meiselman, B. Neu and O.K. Baskurt, Influence of cell-specific factors on red blood cell aggregation, Biorheology 41 (2004), 91-112. [95] J.K. Armstrong, H.J. Meiselman and T.C. Fisher, Evidence against macromolecular bridging as the mechanism of red blood cell aggregation induced by nonionic polymers, Biorheology 36 (1999), 433437. [96] H. Baumler, B. Neu, E. Donath and H. Kiesewetter, Basic phenomena of red blood cell rouleaux formation, Biorheology 36 (1999), 439-442. [97] B. Neu, S.O. Sowemimo-Coker and H.J. Meiselman, Cell-cell affinity of senescent human erythrocytes, Biophys. J. 85 (2003), 75-84. [98] R.B. Ami, G. Barshtein, D. Zeltser, Y. Goldberg, I. Shapira, A. Roth, G. Keren, H. Miller, V. Prochorov, A. Eldor, S. Berliner and S. Yedgar, Parameters of red blood cell aggregation as correlates of the inflammatory state, Am. J. Physiol. 280 (2001), H1982-H1988. [99] S. Chen, A. Eldor, G. Barshtein, S. Zhang, A. Goldfarb, E. Rachmilewitz and S. Yedgar, Enhanced aggregability of red blood cells of beta-thalassemia major patients., Am. J. Physiol. 270 (1996), H1951H1956. [100] B. Chong-Martinez, T.A. Buchanan, R. Wenby and H.J. Meiselman, Decreased red blood cell aggregation subsequent to improved glycemic control in type 2 diabetes mellitus, Diabetic Med. 20 (2003), 301-306. [101] D. Samocha-Bonet, R. Ben-Ami, I. Shapira, G. Shenkerman, S. bu-Abeid, N. Stern, T. Mardi, T. Tulchinski, V. Deutsch, S. Yedgar, G. Barshtein and S. Berliner, Flow-resistant red blood cell aggregation in morbid obesity, Int. J. Obesity 28 (2004), 1528-1534. [102] G. Barshtein, M. Ponizovsky, Y. Nechamkin, M. Ritsner, S. Yedgar and L.D. Bergelson, Aggregability of red blood cells of schizophrenia patients with negative syndrome is selectively enhanced, Schizoph. Bull. 30 (2004), 913-922. [103] O.K. Baskurt, A. Temiz and H.J. Meiselman, Red blood cell aggregation in experimental sepsis, J. Lab Clin. Med. 130 (1997), 183-190. [104] B. Neu and H.J. Meiselman, Depletion-mediated red blood cell aggregation in polymer solutions, Biophys. J. 83 (2002), 2482-2490. [105] O.K. Baskurt and H.J. Meiselman, Cellular determinants of low-shear blood viscosity, Biorheology 34 (1997), 235-247. [106] H.J. Meiselman, Rheologic behavior of shape-transformed human red cells, Biorheology 15 (1978), 225-237. [107] C. Berling, C. Lacombe, J.C. Lelievre, M. Allary and J. Saint-Blancard, The RBC morphological dependance of the RBC disaggregability, Biorheology 25 (1988), 791-798. [108] T. Hovav, S. Yedgar, N. Manny and G. Barshtein, Alteration of red cell aggregability and shape during blood storage, Transfusion 39 (1999), 277-281. [109] G.L.J. Vermeiren, M.J. Claeys, D. Van Bockstaele, B. Grobben, H. Slegers, L. Bossaert and P.G. Jorens, Reperfusion injury after focal myocardial ischaemia: polymorphonuclear leukocyte activation and its clinical implications, Resuscitation 45 (2000), 35-61.

190


[110] S.M. Buttrum, E.M. Drost, M.E. Goffin, C.M. Lockwood, R. Hatton and G.B. Nash, Rheological response of neutrophils to different types of stimulation, J. Appl. Physiol. 77 (1994), 1801-1810. [111] O.K. Baskurt and H.J. Meiselman, Activated polymorphonuclear leukocytes affect red blood cell aggregability, J. Leukoc. Biol. 63 (1998), 89-93. [112] S.M. Buttrum, G.B. Nash and R. Hatton, Changes in neutrophil rheology after acute ischemia and reperfusion in the rat hindlimb, J. Lab. Clin. Med. 128 (1996), 506-514. [113] L.Dintenfass. Hyperviscosity and Hyperviscosemia. MTP Press, Lancaster, 1985. [114] O.K. Baskurt, O. Yalcin and H.J. Meiselman, Hemorheology and vascular control mechanisms, Clin. Hemorheol. Microcirc. 30 (2004), 169-178.


191

Hemorheology of the Fetus and Neonate Otwin LINDERKAMP1 Department of Pediatrics, University of Heidelberg; Im Neuenheimer Feld 150, D-69120 Heidelberg, Germany.

Introduction The development of the fetus, the transition and adaptation at birth, and the subsequent maturation during infancy and childhood require considerable adaptation processes of the macro- and microcirculation. The fetus lives in a severely hypoxic atmosphere with an oxygen saturation of 65 to 70% in blood flowing to vital organs such as the brain. This borderline oxygen supply makes the fetus extremely vulnerable to hypoxicischemic events. At birth, the cardiovascular system undergoes eminent changes such as the sudden interruption of placental blood flow and the redistribution of cardiac output to the pulmonary arteries. Moreover, the neonate is at high risk to acquire disorders with strong impact on blood circulation (e.g., septicemia). The circulation continues to show marked changes with growth and further organ differentiation during infancy and childhood. Peculiar rheologic properties of blood appear to play an important role in the maintenance of high blood flow conditions in spite of very low blood pressure in the fetus and neonate [1]. On the other hand, sudden changes of rheologic properties may develop in the perinatal period with marked effects on circulation such as polycythemia resulting from placental transfusion [2]. This overview is designed to describe the physiology and pathophysiology of hemorheological parameters during the fetal and neonatal period. In addition, some information on developmental hemorheology during infancy and childhood will also be presented.

1. Normal Hemorheology in the Fetus, Preterm and Full-Term Newborn Infant In this chapter the developmental changes of the major determinants of blood viscosity (i.e., hematocrit, plasma viscosity, RBC aggregation and RBC deformability) in the fetus, neonate infant and child are described (Table 1). Hematocrit increases in the fetus from 33% in the 12th week of gestation to 45% in the 30th week and 50% in the 40th week at full-term [3]. At birth the hematocrit may markedly rise as a result of “blood transfusion” from the placenta to the neonate (see below) [2]. A postnatal hematocrit of 45-65% is considered normal in the healthy fullterm neonate.

1

Corresponding author: Director of Neonatology, Department of Pediatrics, University of Heidelberg, Medical School, Im Neuenheimer Feld 150, D-69120 Heidelberg, Germany; E-mail: [email protected]

192

O. Linderkamp / Hemorheology of the Fetus and Neonate

After birth less oxygen transport capacity is required than in the fetus since the arterial oxygen saturation increases from 70% to 95%. This results in cessation of erythropoiesis and a steady decrease of the hemoglobin concentration until it reaches about 11 g/dl at six weeks of postnatal age. The hemoglobin concentration later increases slowly and reaches adult values after puberty. Plasma viscosity depends on the total plasma protein concentration, but is more influenced by high molecular weight proteins (e.g., fibrinogen) than by smaller proteins (e.g., albumin). The concentrations of total plasma protein, plasma fibrinogen and other plasma proteins are very low in immature fetuses and preterm infants, but increase with gestational age. At 20 weeks of gestational age, total plasma protein is 3 g/dl and plasma viscosity is only slightly higher than the viscosity of water. In the full-term neonate, plasma viscosity is still markedly lower than in adults [3-9]. During infancy and childhood total plasma protein concentration and plasma viscosity increase steadily and begin to approximate adult values at 12 months (Table 1). RBC aggregation results from interactions between RBC and macromolecules such as fibrinogen. The aggregation process requires several seconds of contact between adjacent RBC and, therefore, occurs only during blood stasis or at low shear stresses. At high shear stresses, RBC aggregates are rapidly dispersed. Because the plasma concentrations of fibrinogen and other large proteins are low in extremely immature preterm infants, their blood shows very little or no aggregation during the first minute of stasis. Both the rate and extent of RBC aggregation increase with increasing gestational age and are closely related to the fibrinogen concentration. In full-term neonates, RBC aggregation as assessed by low shear viscometry, light transmission or laser backscattering methods is still markedly reduced when compared with adults [812]. In addition to the low plasma protein concentrations, neonatal RBC may show decreased aggregability [13, 14]. This phenomenon may be a result of increased echinocyte formation of the neonatal RBC when compared with adult cells [15]. After birth RBC aggregation increases as a result of rising plasma fibrinogen. RBC deformability is not a mathematically defined mechanical parameter, but the result of various geometric (e.g., shape, volume and excess surface area) and mechanical (e.g., membrane elastic moduli, membrane and internal viscosity) properties. The excess surface area beyond that required to enclose the cellular volume determines the maximum extent of RBC deformation. The cell shape determines how much of the excess surface is available for deformation. Neonatal RBC are markedly larger than adult RBC, but have the same excess surface area and swelling capacity [16, 17]. However, blood in preterm and full-term infants contains more RBC with irregular shape (e.g., keratocytes, spherocytes, acanthocytes, elliptocytes, echinocytes) than adult blood. Geometric and mechanical properties of some of these cells deviate markedly from normal discocytes [15]. Elastic moduli define the forces that are required to achieve a given membrane deformation. Three different elastic moduli (i.e., bending, shear, and area compressibility modulus) have been determined by means of a micropipette system for neonatal and adult RBC [17, 18]. Compared to adults, the elastic moduli of neonatal RBC are decreased by 16 to 32%. This suggests that the resistance of the neonatal RBC membrane to various types of elastic deformation is lower than in adults. In preterm infants, the membrane shear elastic modulus is even smaller than in full-term infants, indicating that their RBC membrane is very flexible. Membrane and hemoglobin viscosity of neonatal and adult RBC are similar [18].


193

In spite of these favorable membrane mechanical properties, studies on the deformability of neonatal RBC give conflicting results. The widely used filtration methods showed markedly reduced filterability of neonatal RBC when compared with adult cells due to the larger RBC volume in neonates [19, 20]. The pressure required to aspirate single RBC completely into 3.3 μm diameter pipettes was also higher for neonatal than for adult RBC, but the aspiration pressure tended to be lower for neonatal RBC when cells with the same volume were compared [18]. Counter-rotating devices based on the method of cone-plate viscometry apply well-defined shear forces and are little influenced by the cell volume. Studies using these devices (e.g., rheoscope, Rheodyne, ectacytometer) found similar RBC deformability in preterm and full-term neonates and adults [21, 22]. This may be explained by a balance of favorable properties (e.g., decreased membrane elastic modulus) and unfavorable properties (e.g., more cells with abnormal shape) of neonatal RBC. When RBC deformability in a counter-rotating device was studied within one hour following blood sampling, RBC deformability was better in preterm and small-for-gestational age infants than in healthy full-term neonates who, in turn, showed better RBC deformability than adults [15, 23]. Infants and children show RBC deformability levels that are similar to adults. Whole blood viscosity measured in rotational viscometers is determined by the hematocrit, plasma viscosity and, at low shear rate, RBC aggregation. At a given hematocrit, whole blood viscosity decreases with decreasing maturity of the neonate due to the decrease in plasma viscosity [4-8, 11]. Moreover, the rise in blood viscosity at low shear rates is less pronounced for neonatal blood as a result of reduced RBC aggregation [4, 8, 11]. Mixing neonatal RBC with adult plasma and adult RBC with neonatal plasma changes blood viscosity at given hematocrit to values for adult and neonatal suspensions (i.e., neonatal RBC in adult plasma behave like adult blood) [4]. Moreover, fetal transfusion of adult blood increased fetal blood viscosity in a manner that was dependent on both the rise in hematocrit and in plasma protein concentration [24]. Thus, the effects of plasma proteins on hemorheological properties have to be considered when plasma is transfused together with RBC. Blood viscosity in tube viscometers depends on the tube diameter in addition to the other determinants of blood viscosity. Both the tube hematocrit and the blood viscosity decrease with decreasing tube diameter, with these diameter-dependent changes termed the Fåhraeus Effect and Fåhraeus-Lindqvist Effect, respectively. Both effects are more pronounced for neonatal RBC [7, 25, 26] due to their larger mean cell volume (MCV) and increased membrane elastic deformability [17, 18]. The extent of the FåhraeusLindqvist Effect increases with rising hematocrit: when going from a 500 μm tube to a 50 μm tube, viscosity reductions at a hematocrit of 70% were 56% in preterm infants, 50% in full-term neonates, and 39% in adults, whereas the viscosity reductions at a hematocrit of 30% were only 35, 29, and 19%, respectively [7]. Because of the enhanced Fåhraeus-Lindqvist Effect and decreased plasma viscosity, blood viscosity in 50-μm tubes is similar in neonates with a hematocrit of 70% to adults with a hematocrit of 50% (Figure 1) [7]. These findings may explain why circulation in neonates is less affected by a high hematocrit than circulation in adults (see below).

194


Figure 1. Viscosity of neonatal and adult blood at hematocrits of 50% (i.e., 0.50) and 70% (i.e., 0.70) measured in tubes with diameters of 50, 100 and 500 μm.

2. Hemorheology and Circulation in the Fetus and Neonate The effect of blood viscosity on blood flow in various vessels can principally be derived from Poiseuille’s law. In a circular vessel or tube, the resistance (R) increases with increasing viscosity (K) of the moving fluid and with the hindrance resulting from vessel geometry (Z): R = (Z)(K). The hindrance increases with increasing vessel length and decreasing vessel radius (r) to the fourth power (r4). The length of large vessels increases with body length. However, it is not known if the small vessels (i.e., arterioles, capillaries, venules) show differences in geometry between neonates and adults. Circulation in the fetus and neonate is characterized by high flow conditions in spite of low vascular pressures. Systolic blood pressure (P) increases by 100% from birth to adulthood, whereas systemic cardiac output (CO) is 150 ml/kg/min in the neonate and 80 ml/kg/min in adults. However, the cardiac index (CI) is about 3 l/kg/m2 from birth to adulthood [1,27]. Thus, systemic flow resistance R, where R=P/CO, increases by the same extent as blood pressure from birth to adults. Since the blood viscosity in narrow tubes with a diameter of 50 μm (shown in Table 1), corresponding to resistant vessels in vivo, increases markedly during maturation, the ratio of systemic flow resistance (i.e. vascular hindrance) to blood viscosity increases less than the systemic flow resistance (Table 1). This suggests that blood viscosity and vascular hindrance each contribute about 50% to the age-dependent rise in systemic flow resistance. In other words, low blood viscosity is an important prerequisite for the maintenance of the fetal and neonatal circulation in spite of low vascular pressures.

195


Note that the blood viscosity values measured in 50 μm tubes appear to be relatively low due to the Fåhraeus-Lindqvist Effect (Figure 1). Steel, et al. have shown that both plasma viscosity and the resistance index increase with gestational age of the fetus and that 55% of the variance of the resistance index can be explained by changes of plasma viscosity [6]. Thus, plasma viscosity appears to play a major role for the flow resistance in the fetus and during postnatal development. Microcirculation studies in human neonates have been limited to superficial structures such as the skin. Studies of flow velocity of RBC in nailfold capillaries by means of a television-video microscope demonstrated similar RBC velocities in neonates and adults [28]. Laser Doppler blood flow velocimetry in cutaneous microcirculation revealed less vasomotion during the first postnatal days [29] and marked vasoconstriction of skin vessels in neonates in response to septicemia [30]. Table 1. Hemorheology and circulation in preterm and full-term neonates and adults

Age

Hematocrit %

Plasma Viscosity

RBC Aggregation

mPa.s

Blood Viscosity (K)

Cardiac Index (CI)

Blood Pressure (P)

mPa.s

mL/min/m2

mmHg

Vascular Hindrance P/(CI)(K)

25 wks

45

0.83

1

1.6

2.8

45

10

32 wks

48

0.93

2.4

1.7

3.1

55

10

At birth

51

1.04

3.8

2.1

3.1

65

10

1 month

34

1.07

4.4

1.6

3.2

75

14

6 months

35

1.10

5.0

1.7

3.3

80

14

1 year

36

1.14

5.8

1.9

3.5

90

14

10 years

39

1.22

8.9

2.2

3.3

100

14

25 years

42.5

1.26

10.3

2.6

3.0

120

15

50 years

42.5

1.35

12.3

2.7

3.0

130

16

75 years

41

1.41

13.6

2.8

3.0

140

17

* The following measurement techniques were used: plasma viscosity, 100 μm tubes; RBC aggregation, Myrenne aggregometer; blood viscosity, 50-μm tubes (see Figure 1); cardiac index, Doppler ultrasound; systolic blood pressure, oscillometry; vascular hindrance = blood pressure/cardiac index/blood viscosity

We have used a mathematical model to describe blood flow in capillaries with diameters of 3 to 6 μm. The model requires measurements of RBC surface area and volume, plasma viscosity and capillary diameter. Plasma viscosity determines the friction in the gap between RBC and capillary wall (i.e., lubrication). If the RBC are suspended in the same medium, the large neonatal RBC from a full-term infant require 27% higher driving pressures than adult RBC to achieve the necessary elongation for passing through a 4.5-μm capillary [31]. However, all RBC require similar driving pressures if the cells are suspended in the corresponding autologous plasmas. Below a critical vessel diameter, 3.3 μm for adults and 3.6 μm for full-term neonates, the driving pressure and blood viscosity increase steeply as a result of rising friction between RBC and the vessel wall. This suggests that in narrow capillaries, the disadvantage of the large size of neonatal RBC is compensated for by the low plasma viscosity.

196


3. Accelerated Aging of Neonatal RBC The life-span of fetal and neonatal RBC (45-70 days) is considerably shorter than that of adult RBC (120 days) [32]. It is generally assumed that RBC aging is a result of membrane damage associated with decreases in surface area, water, volume, potassium and several enzyme activities, and increases in hemoglobin concentration (i.e., mean cellular hemoglobin concentration, MCHC), membrane and hemoglobin viscosity [33, 34]. Because RBC density rises with increasing MCHC, density separation is frequently used to study RBC aging. The age-dependent membrane loss has been explained by various mechanisms such as endocytosis, damage of RBC proteins (e.g., enzymes, hemoglobin, membrane proteins) by oxygen radicals, removal of RBC membrane fragments by immunologic mechanisms, and the continuous mechanical stress acting on the membrane of circulating RBC. As final steps in RBC removal the dense and rigid aged RBC bind IgG, are captured in the spleen and phagocytosed. In vivo aging of RBC is clearly associated with a steady rise in density and decrease in deformability of RBC [35-37]. Neonatal RBC show decreased membrane resistance to both deformation and fragmentation when compared with adult RBC [17, 18, 38, 39]. In particular, tether formation by the neonatal RBC membrane requires markedly lower pressure than adult RBC. Approximately 20% of un-separated neonatal RBC, compared to only 3% of adult RBC, have irregular shapes including spherocytes and acanthocytes; these abnormal shapes are mainly found in the densest RBC fraction and have a small volume and surface area and a high membrane viscosity when compared with normal discocytes [15]. This agrees with the finding of a larger portion of extremely dense RBC in un-fractionated neonatal RBC when compared with adult cells [21]. Studies on density-separated RBC indicate that the densest 3% of neonatal RBC show a greater increase in density, MCHC, membrane viscosity and a more pronounced decrease in enzyme activities, potassium, volume, surface area, and deformability than the densest 5% of adult RBC [17].

4. Leukocytes At birth the white blood cell (WBC) count averages 15,000/μl in the full-term neonate [19]. The percentage of neutrophils is similar in neonates and adults, but neonates show a marked left shift toward younger cells, with 7% for full-term to 20% for preterm immature neutrophils [19, 40]. Comparison of filtration rates of whole blood and RBCbuffer suspensions is a simple method for assessment of WBC deformability. In adults, the presence of WBC decreased the filtration rate by 41%, whereas in neonates the reduction was 53% for full-term to 77% for preterm [19]. This can be explained by the 2-fold higher WBC count and the presence of more immature neutrophils in neonatal blood when compared with adults. Immature neutrophils are less deformable than mature neutrophils [41]. Miller was the first to study the deformability of leukocytes in neonates by means of a micropipette system. He aspirated neutrophils into micropipettes with internal diameters of 3-5 μm and measured the aspiration pressure for total aspiration [42]. Miller reported that neonatal neutrophils require higher aspiration pressures than adult cells, thereby concluding that neonatal neutrophils are less deformable and that decreased neutrophil deformability might contribute to functional defects of neonatal

197


neutrophils. However, Miller did not distinguish between mature and immature neutrophils. We have studied volume, membrane and cellular deformability of various neonatal and adult WBC [43]. Membrane deformability was studied by aspiration of membrane-cytoplasm tongues into 2.5-μm diameter micropipettes over a period of 1 min. In addition, neonatal and adult neutrophils were totally aspirated into 5-μm pipettes. In contrast to the results of Miller, neonatal and adult WBC were found to have similar volume, membrane and cellular deformability when the same cell types were compared (Table 2). Nevertheless, WBC may have a greater impact on neonatal circulation because the higher total WBC count and the higher percentage of immature WBC when compared to adults. Neonatal neutrophils show several functional defects when compared with adult neutrophils such as impaired adherence, lower bactericidal activity, and diminished actin polymerization and chemotaxis during activation [44]. However, in vitro stimulation of neonatal and adult PMN with activators (e.g., fMLP, tumor necrosis factor-Į, and interleukin-8) resulted in similar pseudopod formation and loss of deformability [44]. As in adults, pseudopods of neonatal neutrophils are extremely rigid [40]. Table 2. Comparison of neonatal and adult white blood cells Volume

Tongue Length*

Aspiration Time**

(fL)

(μm)

(s)

Adult

370±41

6.63±1.18

4.63±2.17

Neonatal

360±38

6.12±1.15

4.42±1.96

Neonatal

421±68

4.20±1.16

Adult

233±29

3.58±0.68

Neonatal

235±35

3.22±0.53

Adult

532±60

3.75±0.62

Neonatal

519±66

3.38±0.70

Cell type

Neutrophilic PMN Immature neutrophils Lymphocytes

Monocytes Values are mean ± 1 SD *Membrane tongue aspiration into a 2.5-μm diameter pipette using a pressure of -2 cm H2O for 20 s; **Aspiration time of entire RBC into a 5-μm diameter pipette using a pressure of -2 cm H2O

5. Pathologic Blood Flow Properties in the Fetus and Neonate 5.1. Polycythemia in the Neonate Neonatal polycythemia, defined as a venous hematocrit greater than or equal to 65%, occurs in about 3% of all newborn infants [46, 47]. In infants with late cord-clamping, the incidence of polycythemia is about 25% [48]. Increased intrauterine erythropoiesis due to chronic hypoxia is another major cause of neonatal polycythemia. In some disorders such as maternal diabetes and asphyxia, increased plasma viscosity, decreased RBC deformability and increased RBC aggregation may contribute to increased blood viscosity in addition to polycythemia.

198


5.1.1. Placental Transfusion Blood transfusion from the placenta to the neonate occurs when the umbilical cord is clamped at some time after 5 seconds of birth. Before birth, the fetal blood volume is approximately 70 ml/kg of blood. Another 45 ml of blood per kg of fetal weight is contained in the placenta [2]. If the newly born infant is kept at or below the level of the placenta and the umbilical cord is clamped three minutes after birth or later, 35 ml/kg may flow into the neonate. The rapid increase in blood volume by 50% is counteracted by extravasation of plasma so that the hematocrit rises from approximately 50% at birth to 65% at 2-4 hours of birth. This increase in hematocrit is associated with a rise in blood viscosity by 50% [48] (Table 3). The resulting hyperviscosity may impair blood flow to various organs, thereby compromising their oxygen supply. For this reason, late cord clamping of infants, held at or below the level of the placenta, has become uncommon. However, the Leboyer method is widely used. The Leboyer birth method requires that the newly born infant is placed on the mother’s abdomen and the cord is clamped when it stops pulsating. Thus, the cords of these infants are clamped late, but the pressure gradient between placenta and infant is decreased by lifting the infant above the placenta. Consequently, the volume of placental transfusion and the rise in hematocrit in Leboyer deliveries is between that in infants with early and late cord-clamping [2, 49]. Enhanced uterine contractions (e.g., oxytocin treatment) may cause marked placental transfusion to the fetus and polycythemia before birth [2]. Late cord-clamping can not be recommended for neonates with a high risk for polycythemia [50]. 5.1.2. Perinatal Hypoxia and Acidosis. Prolonged intrauterine hypoxemia stimulates erythropoiesis and causes a shift of blood from the placenta to the fetus, thereby increasing the hematocrit [2, 51]. Impaired oxygen supply to the placenta (e.g., maternal gestosis, smoking, placental insufficiency) is therefore often associated with polycythemia of the fetus and the neonate [52, 53]. Moreover, prolonged intrauterine stress may cause a rise in plasma fibrinogen, thereby increasing plasma viscosity and RBC aggregation, and a rise in total leukocyte count and the percentage of immature neutrophils. Buchan proposes that increased blood viscosity in infants with intrauterine growth retardation may be an important risk factor for hypertension in adult life [54]. RBC deformability is not altered in infants with intrauterine growth retardation [23, 55]. However, hypoxia and acidosis cause a significant increase in hematocrit due to blood transfer from the placenta to the fetus [56] and a decrease in RBC filterability in the fetus and neonate [57]. Even the moderate hypoxic stress of normal vaginal delivery decreases RBC filterability below that of neonates born by primary Cesarean section. Plasma viscosity does not rise as a result of acute fetal hypoxemia [58]. Polycythemia in the neonate decreases cardiac output and blood flow to the brain, the gastrointestinal tract, the kidneys, the lungs, the limbs and the skin [59, 60]. However, systemic RBC transport, calculated as product of cardiac output times hematocrit, RBC transport to the brain (i.e., blood flow velocity times hematocrit), and cerebral oxygenation remain stable in the neonate over a hematocrit range of 40 to 70% [61, 62]. These in vivo findings agree with in vitro studies of RBC flow in narrow tubes (Figure 1) [7]. Nevertheless, clinical consequences of polycythemia in neonates have been reported within this range of hematocrit values. Wiswell and associates observed

199


clinical signs and symptoms in 50% of neonates with hematocrits of 65% or greater [63], while van der Elst, et al. found most infants healthy and unaffected by this hematocrit [64]. Acunas, et al. report a significant relationship between thrombocytopenia and the severity of clinical findings in polycythemic neonates [65]. Table 3. Effect of cord-clamping on hemorheology in full-term infants at two hours after birth*

Hematocrit (%) Plasma viscosity (mPa.s) RBC Aggregation RBC Deformation (at 3 Pa) Blood Viscosity (mPa.s)

Early [a]

Leboyer [b]

Late [c]

P<0.05

48±6

58±6

63±5

a
1.04±0.09

1.06±0.08

1.09±0.09

a=b=c

4.2±1.0

4.2±1.8

4.0±1.4

a=b=c

0.42±0.04

0.41±0.05

0.42±0.05

a=b=c

2.8±0.5

3.7±0.5

4.2±0.4

a
* From [48, 49]

Long-term studies on the incidence of developmental and neurological abnormalities at one to seven years also come to conflicting results (Table 4). Goldberg, et al. [66] and Black, et al. [67] report a markedly increased incidence of neurological problems in polycythemic infants at 9 months and at 1 and 2 years of age, respectively. At seven years only small, insignificant differences were observed between children with and without neonatal poycythemia [68]. However, Goldberg, et al. and Black, et al. included infants with additional risks for developmental problems. Moreover, hemodilution did not significantly influence the results. Investigators who included only polycythemic infants without additional risks found no effect of polycythemia and hemodilution on long-term outcome [64, 69-71]. Drew, et al. report that increased blood viscosity is a better predictor of poor outcome than a high hematocrit [72]. Two recent reviews came to the conclusion that there is no evidence of long term benefit from partial exchange in polycythemic infants [73, 74]. The incidence of necrotizing enterocolitits even increased by partial exchange transfusion, probably because the procedure requires a central catheter usually via the umbilical vein [74]. Although most studies failed to show a benefit of partial exchange transfusion, it is likely that a hematocrit of 70% or higher compromises the oxygen supply of vital organs. In neonates, isovolemic hemodilution has been done using plasma, 5% human serum albumin, serum free of activated clotting factors, Haemaccel, or crystalloids such as normal saline or Ringer solution [75-77]. Adult plasma increases the plasma viscosity and the RBC aggregation and should, therefore, not be used for hemodilution [4]. Moreover, little is known about the distribution and metabolism of plasma expanders such as Haemaccel or hydroxyethyl starch. A systematic review of the optimal fluid for partial exchange transfusion in polycythemic neonates concludes that there is no difference in effectiveness between various solutions [78]. Cristalloids are, therefore, presently recommended for hemodilution in the neonate.

200


Table 4. Incidence of developmental and neurological abnormalities: Influence of neonatal polycythemia and hemodilution

First Author, Year of Publication and Number

Incidence of Abnormalities

Child Age at Study

Neonatal Hematocrit (%)

(years)

With or Without Hemodilution

Control

Hemodilution

NoHemodilution

Neonates

0.67

>65

0/49 (0%)

0/24 (0%)

0/30 (0%)

v. d. Elst 1980 [60] 6

65

1/25 (4%)

-

3/67 (4%)

Høst 1982 [64]

2.5

65

1/25 (4%)

-

0/73 (0%)

Goldberg 1982 [61]

0.67

64

4/6 (67%)

5/10 (50%)*

1/6 (17%)

1

65

9/32 (28%)*

6/21 (29%)*

4/36 (11%)

2

65

16/29 (55%)*

10/27 (37%)*

9/46 (20%)

7

65

100±14**

98±15**

104±13**

2

63 ( HV)

90±13**

-

78±16**

2

63 (no HV)

85± 9**

88±13**

78±16**

1.5-2

65 (no HV)

11/25

4/15

7

BV 2SD**

-

16/23 (70%)*

Black 1985 [62] Delaney-Black 1989 [63] Bada 1992 [65] Ratrisawada 1994 [66] Drew 1997 [67]

5/22 (23%)

BV, blood viscosity; HV, symptomatic hyperviscosity; *Significantly different indices of development and neurological abnormalities compared to control group without neonatal polycythemia; no significant effect of partial exchange transfusion was observed in any study; **Blood viscosity more than 2 SD above normal mean

5.1.3. Infants of Diabetic Mothers Adverse effects of type I and type II diabetes on hemorheological properties have frequently been reported in the literature. In children with untreated, newly diagnosed insulin-dependent diabetes, both RBC and PMN deformability was markedly reduced when compared with healthy children [79]. Poorly controlled insulin-dependent diabetes during pregnancy may cause a variety of metabolic problems in the infant due to hyperglycemia and hyperinsulinism. The hematocrit may rise as a result of hyperinsulinism and placental insufficiency [80]. Moreover, plasma viscosity of these infants may be increased and RBC deformability may be decreased [81, 82]. The decrease in RBC deformability is a result of decreased membrane elasticity (i.e. increased membrane elastic shear modulus) [81]. These impaired hemorheological properties may contribute to the increased risk of thromboembolic complications for infants of diabetic mothers. 5.1.4. Neonatal Septicemia Neonates are highly susceptible to septicemia and septic shock due to immaturity of their immunological defense mechanisms. Disordered microcirculatory perfusion is the major cause of clinical manifestations and death in severe sepsis [83]. Signs of impaired microcirculation develop early in septic neonates, often before a rise of Creactive protein and a marked left-shift of the leukocytes become apparent [30]. Most septic neonates do not demonstrate a hyperdynamic flow state at the beginning of a


201

septic shock, but may be hypodynamic and hypotensive when the first clinical signs appear [45]. The sudden deterioration of the microcirculation may explain why many septic neonates show a sudden fall of the leukocyte count due to sequestration in small vessels [40]. Leukocyte stiffening and adherence to the endothelium of small vessels are major steps in the development of septic shock [84, 85]. Both pulmonary failure and brain damage in septic neonates have been explained by capillary obstruction with leukocytes [86, 87]. Several hemorheological abnormalities may contribute to impaired micro- and macro-circulatory blood flow in septic neonates [45]. RBC aggregation and plasma viscosity can be enhanced due to increased plasma fibrinogen, fibrin monomers and other plasma components. RBC deformability may be impaired as a result of direct interactions of bacterial components with the RBC membrane or cytoplasm, or by extrinsic factors such as activated WBC and platelets, oxygen radicals and cytokines. Impaired RBC deformability in septic rats [88] and after lipid A incubation [89] was associated with a decrease in membrane elasticity (i.e. increase in shear elastic modulus) studied by a micropipette technique. The increase in shear elastic modulus may be explained by an increase in cytoskeletal protein-protein interaction, an increase in free cytosolic calcium concentration and/or oxidation of hemoglobin [45]. We have recently demonstrated that endotoxin binds to RBC in human adults with gram negative septicemia using ȕ-hydroxymyristic acid (HMA) as natural label of LPS [90]. The HMA content of RBC was related to the impairment of cellular deformability. Neonatal RBC bind less lipid A than adult RBC [91]. Both neonatal and adult RBC show a significant reduction in RBC deformability upon incubation with lipid A. Neonatal RBC showed a fast recovery with almost normal deformability after 30 minutes, whereas adult RBC had not fully recovered after 60 minutes [92]. This may be explained by a higher amount of saturated fatty acids in neonatal RBC membranes. Todd, et al. [93] found a marked increase of membrane viscosity in neonatal and adult RBC membranes after in vitro incubation with E. coli endotoxin. Neonatal group B Streptococcus (GBS) septicemia caused a marked reduction of RBC deformability [94]. In vitro incubation of neonatal RBC with GBS caused a more pronounced reduction of deformability than was found for adult RBC [94]. This corresponds to the observation that GBS toxins bind more to neonatal cells than to adult cells [95]. In contrast to GBS, group-A streptolysin O causes less hemolysis and less impairment of RBC deformability in neonates than in adults [96]. This corresponds to the lower risk of neonates to acquire severe infection by group-A ȕ–hemolytic streptococcus. During bacterial infection, large numbers of mature and relatively rigid immature neutrophils are released from the bone marrow into the circulating blood. Moreover, the deformability of mature neutrophils is decreased significantly in septic adults [97], children [98] and neonates [40] due to pseudopod formation, alterations of the viscoelastic properties of the membrane and cytoplasm, and reduction of excess surface area during activation. In neonates with GBS or gram negative septicemia, one third of the circulating neutrophils are immature [40]. Moreover, up to 14% of the circulating neutrophils in septic neonates became activated during aspiration into mircropipettes, while only 4% are activated for cells from healthy adults and neonates. Neonatal and adult PMN showed equal activation and loss of deformability in response to activators [44].

202


References [1] [2] [3] [4] [5] [6]

[7] [8] [9] [10] [11] [12]

[13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30]

O. Linderkamp, Pathological flow properties of blood in the fetus and neonate, Clin. Hemorheol. Microcirc. 16 (1996), 105-116. O. Linderkamp, Placental transfusion: determinants and effects, Clin. Perinatol. 9 (1982), 559-592. C.R. Welch, M.W. Rampling, M.A. Anwar, D.G. Talbert and C.H. Rodeck, Gestational reference ranges for fetal haemorheological parameters, Clin. Hemorheol. 14 (1994), 93-103. O. Linderkamp, H.T. Versmold, K.P. Riegel and K. Betke, Contributions of red cells and plasma to blood viscosity in preterm and full-term infants and adults, Pediatrics 74 (1984), 45-51. W.H. Reinhardt, S.J. Danoff, R.G. King and S. Chien, Rheology of fetal and maternal blood, Pediatr. Res. 19 (1985), 147-153. S.A. Steel, J.M. Pearce, G. Nash, B. Christopher, J. Dormandy and J.M. Bland, Correlation between Doppler flow velocity waveforms and cord blood viscosity, Br. J. Obstet. Gynecol. 96 (1989), 11681172. O. Linderkamp, A.A. Stadler and E.P. Zilow, Blood viscosity and optimal hematocrit in preterm and full-term neonates in 50- to 500-μm tubes, Pediatr. Res. 32 (1992), 97-102. M.A. Anwar, M.W. Rampling, S. Bignall and R.P. Rivers, The variation with gestational age of the rheological properties of the blood of the new-born, Br. J. Haematol. 86 (1994), 163-168. L. Heilmann, W. Rath and K. Pollow, Fetal hemorheology in normal pregnancy and severe preeclampsia. Clin. Hemorheol. Microcirc. 32 (2005), 183-190. O. Linderkamp, P. Ozanne, P.Y.K. Wu and H.J. Meiselman, Red blood cell aggregation in preterm and term neonates and adults, Pediatr. Res. 18 (1984), 1356-1360. M.W. Rampling, P. Whittingstall, G. Martin, S. Bignall, R.P.A. River, T.J. Lissauer and P.C. Bailey, A comparison of the rheologic properties of neonatal and adult blood, Pediatr. Res. 25 (1989), 457-460. A. El Bouhmadi, P. Boulot, F. Laffargue and J.F. Brun, Rheological properties of fetal red cells with special reference aggregability and disaggregability analyzed by light transmission and laser backscattering techniques, Clin. Hemorheol. Microcirc. 22 (2000), 79-90. H.J. Meiselman, Red blood cell role in RBC aggregation: 1963-1993 and beyond, Clin. Hemorheol. 13 (1993), 575-592. M.W. Rampling, H.J. Meiselman, B. Neu and O.K. Baskurt, Influence of cell-specific factors on red blood cell aggregation, Biorheology 41 (2004), 91-112. P. Ruef and O. Linderkamp, Deformability and geometry of neonatal erythrocytes with irregular shapes, Pediatr. Res. 45 (1999), 114-119. O. Linderkamp, P.Y.K. Wu and H.J. Meiselman, Geometry of neonatal red blood cells, Pediatr. Res. 17 (1983), 250-253. O. Linderkamp, E. Friederichs and H.J. Meiselman, Mechanical and geometrical properties of densityseparated neonatal and adult erythrocytes, Pediatr. Res. 34 (1993), 688-693. O. Linderkamp, G.B. Nash, P.Y.K. Wu and H.J. Meiselman, Deformability and intrinsic material properties of neonatal red blood cells, Blood 67 (1986), 1244-1250. O. Linderkamp, B.J. Hammer and R. Miller, Filterability of erythrocytes and whole blood in preterm and full-term neonates and adults, Pediatr. Res. 20 (1986), 1269-1273. G. Buonocore, S. Bernie, D. Gioia, G. Garosi and R. Bracci, Whole blood filterability in the neonate, Clin. Hemorheol. 11 (1991), 41-48. O. Linderkamp, M. Güntner, W. Hiltl and V.M. Vargas, Erythrocyte in the fetus, preterm and term neonate, Pediatr. Res. 20 (1986), 93-96. L.M. Matovcik, D. Chiu, B. Lubin, W.C. Mentzer, P.A. Lane, N. Mohandas and S.L. Schrier, The aging process of human neonatal erythrocytes, Pediatr. Res. 20 (1986), 1091-1096. O. Linderkamp, U. Kiau and P. Ruef, Cellular and membrane deformability of red blood cells in preterm infants with and without growth retardation, Clin. Hemorheol. Microcirc. 17 (1997), 279-283. R. Welch, M.W. Rampling, A. Anwar, D.G. Talbert and C.H. Rodeck, Changes in hemorheology with fetal intravascular transfusion. Am. J. Obstet. Gynecol. 170 (1994), 726-732. E.P. Zilow and O. Linderkamp, Viscosity reduction of red blood cells from preterm and full-term neonates and adults in narrow tubes (Fahraeus-Lindqvist effect), Pediatr. Res. 25 (1989), 595-597. C.B. McKay, O. Linderkamp and H.J. Meiselman, Fahraeus and Fahraeus-Linqvist effects for neonatal and adult red blood cell suspensions, Pediatr. Res. 34 (1993), 538-543. O. Linderkamp, Blood viscosity of the neonate, Neoreview 5 (2004), e406-e416. M. Norman, P. Herin, B. Fagrell and R. Zetterström, Capillary blood cell velocity in full-term infants as determined in skin by videophotometric microscopy, Paediatr. Res. 23 (1988), 585-588. J. Pöschl, T. Weiss, C. Diehm and O. Linderkamp, Periodic variations in skin perfusion in full-term and preterm neonates using laser Doppler technique, Acta Paediatr. Scand. 80 (1991), 999-1007. J.M.B. Pöschl, T. Weiss, F. Fallahi and O. Linderkamp, Reactive hyperemia of skin microcirculation in


203

septic neonates, Acta Paediatr. 83 (1994), 808-811. [31] A. Stadler and O. Linderkamp, Flow behavior of neonatal and adult erythrocytes in narrow tubes, Microvasc. Res. 37 (1989), 267-279. [32] O. Linderkamp, Blood rheology in the newborn infant, Baill. Clin. Haematol. 1 (1987), 801-825. [33] O. Linderkamp, Meiselman HJ. Geometric, osmotic, and membrane mechanical properties of densityseparated human red cells, Blood 59 (1982), 1121-1127. [34] G.B. Nash and H.J. Meiselman, Red cell and ghost viscoelasticity. Effects of hemoglobin concentration and in vivo aging, Biophys. J. 43 (1983), 63-73. [35] O. Linderkamp, E. Friederichs, T. Böhler and A. Ludwig, Age dependency of red blood cell deformability and density: studies in transient erythroblastopenia of childhood, Br. J. Haematol. 83 (1993), 125-129. [36] R.E.Waugh, N. Mohandas, C.W. Jackson, T.J. Mueller, T. Suzuki and G.L. Dale, Rheologic properties of senescent erythrocytes: loss of surface area and volume with red blood cell age, Blood 79 (1992), 1351-1358. [37] M. Klipp, A.U. Holzwarth, J. Poeschl, M. Nelle and O. Linderkamp, Effects of erythropoietin on erythrocyte deformability in non-transfused preterm infants, Acta Paediatr. 96 (2007), 253-256. [38] T. Böhler, A. Leo, A. Stadler and O. Linderkamp, Mechanical fragility of erythrocyte membrane in neonates and adults, Pediatr. Res. 32 (1992), 92-96. [39] P. Ruef and O. Linderkamp, Plastic deformation of erythrocyte membrane in full-term neonates and adults, Clin. Hemorheol. 14 (1994), 659-662. [40] O. Linderkamp, P. Ruef, B. Brenner, E. Gulbins and F. Lang, Passive deformability of mature, immature, and active neutrophils in healthy and septicemic neonates, Pediatr. Res. 44 (1998), 946-950. [41] M.A. Lichtman and R.I. Weed, Alteration of the cell periphery during granulocyte maturation: relationship to cell function, Blood 39 (1972), 301-316. [42] M.E. Miller, Pathology of chemotaxis and random mobility, Semin. Hematol. 12 (1975), 59-82. [43] P. Ruef, T. Böhler and O. Linderkamp, Deformability and volume of neonatal and adult leukocytes, Pediatr. Res. 29 (1991), 128-132. [44] P. Ruef, J.M. Poeschl, C. Simon, F. Altfelder, E. Craciun and O. Linderkamp, Effect of activators and the phosphodiesterase inhibitors pentoxifylline and enoximone on the deformability of neutrophils in neonates and adults, Acta Paediatr. 93 (2004), 1288-1293. [45] O. Linderkamp, J. Poeschl and P. Ruef, Microcirculatory effects of red blood cells and white blood cells in neonatal sepsis, Neoreview 7 (2006), 517-523. [46] E.J. Werner, Neonatal polycythemia and hyperviscosity, Clin. Perinatol. 22 (1995), 693-710. [47] A. Pappas V. Delaney-Black, Differential diagnosis and management of polycythemia, Pediatr. Clin. North Am. 51 (2004), 1063-1086. [48] O. Linderkamp, M. Nelle, M. Kraus and E.P. Zilow, The effect of early and late cord-clamping on blood viscosity and other hemorheological parameters in full-term neonates, Acta Paediatr. 81 (1992), 745-750. [49] M. Nelle, E.P. Zilow, M. Kraus, G. Bastert and O. Linderkamp, The effect of Leboyer delivery on blood viscosity and other hemorheological parameters in term neonates, Am. J. Obstet. Gynecol. 169 (1993), 189-193. [50] L. Capasso, F. Raimondi, A. Capasso, V. Crivaro, R. Capasso R. Paludetto, Early cord clamping protects at-risk neonates from polycythemia, Biol. Neonate. 83 (2003), 197-200. [51] D. Mandel, Y. Littner, F.B. Mimouni, S. Dollberg, Nucleated red blood cells in polycythemic infants, Am. J. Obstet. Gynecol. 188 (2003), 193-195. [52] F.O. Awonusonu, T.H. Pauly and A.A. Hutchison, Maternal smoking and partial exchange transfusion for neonatal polycythemia, Am. J. Perinatal. 19 (2002), 349-354. [53] I. Kurlat and A. Sola, Neonatal polycythemia in appropriately grown infants of hypertensive mothers, Acta Paediatr. 81 (1992), 662-664 . [54] P.C. Buchan, Fetal deja vu of adult disease - haemorheological association between fetal intrauterine growth retardation and adult hypertension, Clin. Hemorheol. 14 (1994), 651-657. [55] B.D. Christopher, G.B. Nash, S.A. Steel, J.M.F. Pearce and J.A. Dormandy, Filterability of neonatal red cells after normal and complicated delivery, Clin. Hemorheol. 10 (1990), 113-118. [56] D. Carmi, B. Wolach, T. Dolfin and P. Merlob, Polycythemia of the preterm and full-term newborn infant: relationship between hematocrit and gestational age, total blood solutes, reticulocyte count, and blood pH, Biol. Neonate 61 (1992), 173-178. [57] S. Katoh, K. Yamamoto and M. Kitao, Fetal erythrocyte deformability in hypoxia during delivery, Clin. Hemorheol. 12 (1992), 297-308. [58] A. Munoz, J. Uberos, A. Bonillo, A. Valenzuela, A. Puertas, E. Narbona, R. Sanchez, J.A. Molina, Plasma and internal erythrocyte viscosity in umbilical artery and vein of premature infants with and without acute asphyxia, Clin. Hemorheol. 14 (1994),75-82.

204


[59] I.R. Holzman, B. Tabata and D.I. Edelstone, Blood flow and oxygen delivery to the organs of the neonatal lamb as a function of hematocrit, Pediatr. Res. 20 (1986), 1274-1279. [60] T.S. Rosenkrantz, Polycythemia and hyperviscosity in the newborn, Semin. Thromb. Hemost. 29 (2003), 515-527. [61] V.H.A. Mandelbaum, C.D. Guajardo, M. Nelle and O. Linderkamp, Effects of polycythaemia and haemodilution on circulation in neonates, Arch. Dis. Childh. 71 (1994), F53-F54. [62] K.D. Liem, J.C. Hopman, B. Oeseburg, A.F. de Haan And L.A. Kollee, The effect of blood transfusion and haemodilution on cerebral oxygenation and haemodynamics in newborn infants investigated by near infrared spectrophotometry, Eur. J. Pediatr. 156 (1997), 305-310. [63] T.E. Wiswell, J.D. Cornish and R.S. Northam, Neonatal polycythemia: Frequency of clinical manifestations and other associated findings, Pediatrics 78 (1986), 26-30. [64] C.W. Van der Elst, C.D. Molteno, A.F. Malan, H. Heese, The management of polycythaemia in the newborn infant, Early Hum. Dev. 4 (1980), 393-403. [65] B. Acunas, C. Celtik, U. Vatansever and S. Karasalihoglu, Thrombocytopenia: an important indicator for the application of partial exchange transfusion in polycythemic newborn infants, Pediatr. Int. 42 (2000), 343-347. [66] K. Goldberg, F.H. Wirth, W.E. Hathaway, M.A. Guggenehim, J.R. Murphy, W.R. Braithwaite and L.O. Lubchenco, Neonatal hyperviscosity II. Effect of partial plasma exchange transfusion, Pediatrics 69 (1982), 419-425. [67] V.D. Black, L.O. Lubchenco, B.L. Koops, R.L. Poland and D.P. Powell, Neonatal hyperviscosity: Randomized study of effect of partial plasma exchange transfusion on long-term outcome, Pediatrics 75 (1985), 1048-1053. [68] V. Delaney-Black, B.W. Camp, L.O. Lubchenco, C. Swanson, L. Roberts, P. Gaherty and B. Swanson, Neonatal hyperviscosity association with lower achievement and IQ scores at school age, Pediatrics 83 (1989), 662-667. [69] A. Host and M. Ulrich, Late prognosis in untreated neonatal polycythemia with minor or no symptoms, Acta Paediatr. Scand. 71 (1982), 629-633. [70] H.S. Bada, S.B. Korones, M. Pourcyrous, S.P. Wong, W.M. 3rd Wilson, H.W. Kolni and D.L. Ford, Asymptomatic syndrome of polycythemic hyperviscosity: effect of partial plasma exchange transfusion, J. Pediatr. 120 (1992), 579-585. [71] V. Ratrisawadi, R. Plubrukarn, K. Trachulchang and Y. Puapondh, Developmental outcome of infants with neonatal polycythemia, J. Med. Assoc. Thai 77 (1994), 76-80. [72] J.H. Drew, R.L. Guaran, M. Cichello and J.B. Hobbs, Neonatal whole blood hyperviscosity: the important factor influencing later neurologic function is the viscosity and not the polycythemia, Clin. Hemorheol. Microcirc. 17 (1997), 67-72. [73] M.S. Schimmel, R. Bromiker and R.F. Soll, Neonatal polycythemia: is partial exchange transfusion justified? Clin. Perinatal. 31 (2004), 545-553. [74] E.M. Dempsey and K. Barrington, Short and long term outcomes following partial exchange transfusion in the polycythaemic newborn: a systematic review, Arch. Dis. Child. Fetal Neonat. Ed. 91 (2006), F2-F6. [75] A. Roithmaier, R. Arlettaz, K. Bauer, H.U. Bucher, M. Krieger, G. Duc and H.T. Versmold, Randomized controlled trial of Ringer solution versus serum partial exchange transfusion in neonatal polycythaemia, Eur. J. Pediatr. 154 (1996), 53-56. [76] W. Wong, T.F. Fok, C.H. Lee, P.C. Ng, K.W. So, Y. Ou and K.L. Cheung, Randomised controlled trial: comparison of colloid or crystalloid for partial exchange transfusion for treatment of neonatal polycythaemia, Arch. Dis. Child. Fetal Neonat. Ed. 77 (1997), F115-F118. [77] S. Supapannachart, P. Siripoonya, W. Boonwattanasoontorn and S. Kanjanavanit, Neonatal polycythemia: effects of partial exchange transfusion using fresh frozen plasma, Haemaccel and normal saline, J. Med. Asoc. Thai 82 (1999), S82-S86. [78] K.A. De Waal, W. Baerts and M. Offringa, Systematic review of the optimal fluid for dilutional exchange transfusionin in neonatal poycythaemia, Arch. Dis. Child. Fetal Neonat. Ed. 91 (2006), F7F10 [79] O. Linderkamp, P. Ruef, E.P. Zilow and G.F. Hoffmann, Impaired deformability of erythrocytes and neutrophils in children with newly diagnosed insulin-dependent diabetes mellitus, Diabetologica 42 (1999), 865-869. [80] D.R. Salvesen, M.J. Brudenell and K.H. Nicolaides, Fetal polycythemia and thrombocytopenia in pregnancies complicated by maternal diabetes mellitus, Am. J. Obstet. Gynecol. 166 (1992), 1287-1292. [81] O. Linderkamp and H.J. Meiselman, Deformability of red blood cells from infants of diabetic mothers (abstr). Int. J. Microcirc. Clin. Exp. 1 (1982), 292. [82] K. Violaris, E. Bitzos and N. Rudolph, Increased blood viscosity and tachypnoea in infants of diabetic mothers, Arch. Dis. Childh. 61 (1986), 910-912.


205

[83] S. Trzeciak and E.P. Rivers, Clinical manifestations of disordered microcirculatory perfusion in severe sepsis, Crit. Care 9 (2005), S20-S26. [84] G.S. Worthen, B. Schwab, E.L. Elson and G.P. Downe, Mechanics of stimulated neutrophils: cell stiffening induces retention in capillaries, Science 245 (1989), 183-186. [85] H.H. Lipowsky, D. Riedel and G.S. Shi, In vivo mechanical properties of leukocytes during adhesion to venular endothelium, Biorheology 28 (1991), 53-64. [86] B. Engelhardt, K. Sandberg, D. Bratton, A. van den Abbele, J. Grogaard, C. Hellerqvist and H. Sundell, The role of granulocytes in the pulmonary response to group B streptococcal toxin in young lambs, Pediatr. Res. 21 (1987), 159-165. [87] R.G. Faix and S.M. Donn, Association of septic shock caused by early-onset group B streptococcal sepsis and periventricular leukomalacia in the preterm infant, Pediatrics 76 (1985), 415-419. [88] O.K. Baskurt, D. Gelmont and H.J. Meiselman, Red blood cell deformability in sepsis, Am. J. Respir. Crit. Care Med. 157 (1998), 421-427. [89] J.M.B. Pöschl and O. Linderkamp, Effect of lipid A on the deformability, membrane rigidity and geometry of human adult red blood cells, Eur. J. Clin. Invest. 22 (1992), 625-629. [90] J.M. Pöschl, C. Leray, P. Ruef, J.P. Cazenave and O. Linderkamp, Endotoxin binding to erythrocyte membrane and erythrocyte deformability in human sepsis and in vitro, Crit Care Med. 31 (2003), 924928. [91] J.M.B. Pöschl, M. Schnauffer, C. Galanos and O. Linderkamp, Lipid A binding to neonatal and adult red blood cells, Biol. Neonate 67 (1995), 109-112. [92] J.M.B. Pöschl, P. Ruef, M. Schnauffer and O. Linderkamp, The effect of lipid A on the deformability of neonatal and adult red blood cells, Clin. Hemorheol. 16 (1996), 129-133. [93] J. Todd, N.D. Poulos and Mollit, The effect of endotoxin on the neonatal erythrocyte, J. Pediatr. Surg. 28 (1993), 334-337. [94] J.M.B. Pöschl, M. Schnauffer, S. Pohl, H.G. Sonntag and O. Linderkamp, Group B Streptococcus impairs erythrocyte deformability in neonates more than in adults, Arch. Dis. Child. Fetal. Neonat. Ed. 74 (1996), F187-F190. [95] C.G. Hellerqvist, G. Thurman, D. Page, Y. Wang, B. Russel and C.A. Montgomery, Antitumor effect of group B-hemolytic streptococcus, J. Cancer Res. Clin. Oncol. 120 (1993), 63-70. [96] J.M.B. Pöschl, P. Ruef, M. Schnauffer and O. Linderkamp, Effects of group-A streptolysin O on neonatal and adult red blood cell deformability and hemolysis, Eur. J. Clin. Invest. 26 (1996), 461-464. [97] E.M. Drost, G. Kassabian, H.J. Meiselman, D. Gelmont and T.C. Fisher, Increased rigidity and priming of polymorphonuclear leukocytes in sepsis, Am. J. Respir. Crit. Care Med. 159 (1999), 1696-1702. [98] J.M. Poeschl, P. Ruef and O. Linderkamp, Deformability of passive and activated neutrophils in children with Gram-negative septicemia, Scand. J. Clin. Lab. Invest. 65 (2005), 333-339.

206


Mechanical Trauma to Blood Marina V. KAMENEVAa,1 and James F. ANTAKIb Department of Surgery and Bioengineering, McGowan Institute for Regenerative Medicine,University of Pittsburgh, Pittsburgh, PA, USA and bCarnegie Mellon University, Pittsburgh, PA,USA

a

Introduction Prosthetic heart valves, heart-assist devices, oxygenators, dialyzers and other biomedical devices that repair, replace or support various organ systems of the human body are in wide clinical use. These devices are responsible for saving, extending, and enhancing the lives of patients with otherwise hopeless medical conditions. The safety and efficacy of these blood-contacting biomedical devices strongly depends on the extent to which they damage blood. Unfortunately, in many cases these devices cause dangerous complications triggered by non-physiological factors within the blood flow.

RBC overstretch or fragmentation hemolysis elevated viscous forces turbulence, cavitation prolonged contact collisions

ANEMIA

Ĺ RBC aggregation Ļ RBC deformability

premature RBC removal

Sublethal RBC trauma

POOR CAPILLARY PERFUSION ENDOTHELIAL DYSFUNCTION

free hemoglobin release into plasma

HEMOGLOBINUREA

VASOCONSTRICTION activation or dysfunction of platelets and leukocytes

THROMBOSIS

altered hemodynamics: strain and concentration fields

BLEEDING

INFECTION

Ĺ concentration of inflammatory mediators complement activation

Figure 1. Mechanisms of blood damage: non-physiological hemodynamics results in cascade of events resulting in deleterious clinical endpoints. Synergistic or feedback relationships (dotted lines) exacerbate the effects.

The precise mechanisms of blood damage in blood-contacting devices are heterogeneous and are not well understood in spite of numerous investigations of blood trauma conducted over several decades by investigators worldwide. This body of research has revealed that blood trauma is related to non-physiological flow conditions 1 Corresponding author: Departments of Surgery and Bioengineering, McGowan Institute for Regenerative Medicine, University of Pittsburgh, Pittsburgh, PA, USA; E mail: [email protected]

M.V. Kameneva and J.F. Antaki / Mechanical Trauma to Blood

207

such as elevated shear forces, turbulence, cavitation, prolonged contact and collision between blood cells and foreign surfaces. These factors may induce a variety of damage mechanisms: overstretching or fragmentation of a subpopulation of erythrocytes causing free hemoglobin to be released into plasma (i.e., hemolysis); activation or dysfunction of platelets and leukocytes; increased concentrations of inflammatory mediators; complement activation; and sub-lethal blood trauma such as alterations in mechanical properties of erythrocytes as manifested by an increase in RBC aggregation and decrease in their deformability (see Figure 1). This latter sublethal RBC mechanical damage causes a shortening of RBC life span, a decrease in density of functioning capillaries and area of contact surface of RBC with capillary walls, and may lead to anemia, tissue hypoxia and other complications. Even moderate hemolysis, which is not an immediate threat to renal function, is an important warning sign of other potential blood cell damage such as platelet activation, white blood cell (WBC) dysfunction, and other serious complications such as scavenging of nitric oxide [1], damage to glycocalyx and endothelial cells, and impairment of the vascular smooth muscle tone [2]. Although damage to platelets and WBC is an extremely important topic, this chapter concentrates on the mechanical trauma to RBC and related changes in rheological properties of whole blood. In summary, in vitro experimental studies and clinical experience with artificial organs presented in this chapter substantiate the assertion that mechanical stress substantially impairs the mechanical properties of RBC and adversely affects whole blood rheology, and that this impairment may contribute to various medical complications in dialysis, prosthetic heart valve and circulation-assist device recipients.

1. Artificial organs and devices Blood-contacting biomedical devices are widely used to rehabilitate or replace various vital organs such as heart, lungs, kidneys and liver, as well as other tissues such as heart valves and vessels. In many instances these devices provide support to patients waiting for a donor organ. Even with the increase in organ donation, there is still a great need for organ transplant in the United States [3]. Over 92,000 patients are currently on the official waiting list, of which over 67,000 are waiting for a kidney transplant and over 25,000 are waiting for heart, liver, pancreas, or lung. While the demand continues to grow, the number of donors has not changed significantly in the last ten years [4], thereby motivating the continued development of prosthetic devices. While prosthetic heart valves, circulatory-assist devices, oxygenators, vascular grafts and blood dialysis systems help to save or significantly extend lives, these same devices can produce damage to blood cells that pass through them. Non-physiological hemodynamics, including regions of extremely high shear forces, flow stasis, microcavitation, and elevated Reynolds shear stresses may produce a variety of damage mechanisms. These include complete mechanical destruction of erythrocytes (i.e., hemolysis), activation of platelets and leukocytes, increased concentrations of inflammatory mediators, complement activation, and changes in mechanical properties of erythrocytes [5-8]. In recent years, the attention to red cell trauma in blood-wetted devices has diminished. This is due, in part, to the improvements made in blood pumps and prosthetic valves in which flow-induced hemolysis was most prevalent. One may also

208


suggest that the prevalence of more serious conditions, such as infection and thrombosis, have overshadowed the concern for low levels of hemolysis. Nevertheless, hemolysis still persists in these devices; although the levels may not be so great as to produce anemia, jaundice, or hematuria, its deleterious effects have been documented. Low levels of hemolysis, for example, drastically increase red blood cell aggregation at low shear conditions [9]. Cardio pulmonary bypass (CPB) has been found to decrease red cell deformability which may lead to impaired microcirculation and diminished tissue oxygen supply. Additionally, even low concentrations of plasma free hemoglobin may have a toxic effect on the cardiovascular system due to its ability to bind nitric oxide, an endothelium-derived relaxing factor [1]. In the case of CPB, the anomalous mechanical and environmental factors causing sub-lethal trauma are further exacerbated by decreased oncotic pressure caused by dilution of plasma, and moderate and deep hypothermia widely applied during CPB in infants [10]. Our recent study of blood trauma in clinical hemodialysis vascular access systems (Tesio-Cath, MedComp and LifeSite™ by Vasca, Inc.) have confirmed the sub-lethal damage to red blood cells as indicated by an increase in RBC mechanical fragility and, in one system, a decrease in RBC deformability [11]. 1.1. Artificial Hearts: Ventricular Assist Devices (VAD) Each year, approximately 50,000 Americans with end-stage congestive heart failure could benefit from a heart transplant while only 2000-2500 donor hearts are available for transplantation. Consequently, many people die while waiting for a donor heart. Currently, heart-assist devices are mostly employed as a bridge to transplantation and for extracorporeal circulatory support during heart surgery, yet their ultimate utility will be as a substitute for transplantation. Blood trauma remains arguably the number one cause of morbidity and mortality associated with the preponderance of these cardiovascular devices and indeed this has been a primary barrier to widespread acceptance of this technology [12]. Mechanical circulatory support, such as CPB or chronic ventricular assist devices (VADs), involves prolonged contact and collision between blood cells and foreign surfaces and extremely high mechanical forces which all can be sources of blood trauma. In addition, blood damage caused by small bore percutaneous cannulae in the context of emergency cardiopulmonary support, especially in pediatric patients, are sites in which flow-induced hemolysis may be found. Complications include bleeding, hemolysis, thrombosis, increased susceptibility to inflammation and infection, and transient immune compromise [13]. Chronic anemia has also been observed in patients supported with mechanical circulatory assist devices without remarkable hemolysis. This phenomenon may be due to decreased RBC circulating life span related to sublethal trauma to RBC. Additional clinical data on patients with circulatory support devices has shown alterations of blood rheology such as increased blood viscosity, RBC aggregation, and decreased RBC deformability. These observations further illustrate the prevalence of sub-lethal cell trauma and hence early RBC removal from circulation. [14-21].


209

1.2. Artificial Lungs: Extracorporeal Membrane Oxygenation (ECMO), Cardiopulmonary Support (CPS) Extracorporeal membrane oxygenation (ECMO) is a therapy most commonly used in children with inadequate cardiac or pulmonary function. It entails a combination of blood pump and oxygenator, and is typically applied acutely, limited primarily by the complications arising from blood trauma. Various blood oxygenators used during CPB in open-heart surgery can produce serious complications and in severe cases contribute to high patient mortality. This was especially true in the past era of bubble oxygenators before they were replaced with flat sheet membrane and hollow fiber oxygenators [22]. These latter types of oxygenators have been shown to possess better blood compatibility than bubble oxygenators. However, leukocyte activation still occurs during CPB, which induces a systemic inflammatory response and contributes to postoperative multiorgan dysfunction. The major described complications include bleeding, thromboembolic events, hemolysis, infections, renal and neurological complications, leukopenia [23-25]. Harmful effects of oxygenators on blood due to mechanical stresses and large blood-gas interface include significant decrease in RBC deformability and were considered to be a major cause of these complications [26]. Membrane oxygenators are mostly used in CPB operations. Microporous hollow-fiber membranes are primarily used for short-term CPB application, whereas nonmicroporous hollow-fiber membranes are primarily used for long-term extracorporeal membrane oxygenation application as the intensive therapy of respiratory distress syndrome which has been demonstrated to be beneficial for neonates but less so for older patients [27, 28]. During extracorporeal circulation in CPB open-heart surgery, blood is exposed to anomalous mechanical and environmental factors, such as high shear stress, turbulence, decreased oncotic pressure caused by dilution of plasma, and moderate and deep hypothermia widely applied during CPB in infants. These factors cause damage to RBC, which is manifest by immediate and delayed hemolysis and by changes in the mechanical properties of blood cells. CPB has been found to decrease RBC deformability, which may lead to impaired microcirculation, diminished tissue oxygen supply and contribute to other serious complications of CPB [10]. Significant impairment of RBC deformability was reported in many studies of blood of patients undergoing CPB open-heart surgery [29-32]. 1.3. Artificial Kidneys: Dialysis Machines Artificial kidney or hemodialyzers are used to remove waste and fluid from blood. Currently, over 300,000 Americans receive ongoing dialysis. Most conventional hemodialysis devices incorporate peristaltic roller pumps as the means to transport blood to and from the patient. In general, these devices are traditionally considered to produce blood damage at acceptable levels. However, Spry [33] and DeWachter, et al. [34] independently concluded that the hemolytic effects of cannulae in venipuncture dialysis at high blood flow rates are non-negligible. The major source of RBC damage in hemodialysis systems was associated with peripheral dialysis needles [11, 35, 36]. Consequently, anemia is very common in patients on hemodialysis. The major cause of anemia is a deficit of erythropoietin (EPO). However, it is known that anemia is persistent in hemodialysis patients in spite of EPO treatment. An additional common cause of anemia is a shortening of the RBC lifespan due to sub-lethal damage in dialysis systems. Moreover, thrombosis within the dialyzer itself is not an

210


uncommon occurrence. Several studies have confirmed the negative effects of hemodialysis blood-contacting equipment (e.g., pumps, vascular access systems) on RBC deformability [11, 37-40]. A decrease in RBC deformability in patients suffering from end stage renal failure was found to be significantly higher in a sub-group treated on hemodialysis compared to patients treated with continuous ambulatory peritoneal dialysis (CAPD) which does not contact blood directly [41]. In fact, RBC deformability in the latter group was not statistically significantly different from that measured in a large group of normal individuals. Significant improvement of RBC deformability in hemodialysis patients was found after treatment of these patients with an L-carnitine supplementation and gamma-linolenic acid (GLA) [42, 43]. Treatment of hemodialysis patients with recombinant human erythropoietin (rHuEPO) was associated mostly with a significant increase in whole blood viscosity due to increased hematocrit and not with any significant changes in RBC deformability or aggregation [44-46]. 1.4. Heart Valves Blood trauma has been, and remains an abiding problem for patients with mechanical heart valves. Complications arising from non-physiological fluid dynamics include hemolysis, thrombosis, bleeding due to chronic anticoagulation, and fatal valve failure related to extremely high shear stresses, cavitation and turbulence [47-53]. Hemolytic effects such as an increased plasma free hemoglobin concentration or serum levels of lactate dehydrogenase (LDH) or presence of hemolytic anemia are extensively reported changes in heart valve recipients [54-58]. In addition, subhemolytic damage to RBC which leads to a reduction of RBC deformability and shortening of their lifespan has also been reported [59, 60]. 1.5. Artificial Liver, Biohybrids Combined with Extracorporeal Blood Circulation While mechanical support to failing heart and kidney has been available for several decades, artificial liver support systems became available only recently. Similar to circulatory-assist devices and dialysis machines, these systems must provide patients having fulminant hepatic failure with physiologic support and "bridge" these patients to donor liver transplantation or allow the native liver to recover [61-64]. The liver support systems include traditional hemo- and peritoneal dialysis, charcoal hemoperfusion, hemodiabsorption, high volume plasma exchange, and combinations of all of these techniques, molecular readsorption recirculating system using albumin as the dialysate, bioartificial livers and extracorporeal liver assist devices [65-67]. Many issues are arising on biocompatibility of these devices which are related to blood contact with foreign/non-biological surfaces and potential mechanical damage to blood due to non-physiological flow conditions in these devices (e.g., damage to blood cells, complement and coagulation system activation and hemolysis). 1.6. Artificial Blood: Hemoglobin- and Perfluorochemical-Based Blood Substitutes The problems related to the safety of blood transfusion, such as risk of transmission of HIV and other dangerous viruses (e.g., hepatitis, herpes), possible errors in compatibility testing and bacterial infection, and the shortage of homologous blood have prompted the development of blood substitutes that could replace the gas transport properties of blood. Perfluorochemical (PFC) emulsions and modified


211

human and animal hemoglobin (Hb) solutions are being developed for this purpose and are now in clinical trials. PFCs are synthetic fluorinated hydrocarbons, capable of dissolving large quantities of oxygen at high (90-100%) inspired concentrations of O2, and of delivering this oxygen to the tissues. Since perfluorochemicals are not water/blood soluble they have to be emulsified. Oxygent, a 40% w/v PFC emulsion, developed by Alliance Pharmaceutical Corp, San Diego, CA, USA, has been shown to allow significant reduction or elimination of RBC/blood transfusions during surgery in a Phase III clinical trial. However, another Phase III clinical trial in cardiopulmonary bypass surgery was discontinued due to unacceptable rates of complications. Another blood substitute, modified (polymerized) bovine or human hemoglobin, consists of chemically stabilized polyhemoglobin formulated in an osmotically balanced solution. On a gram to gram basis, these hemoglobin-based oxygen carriers (HBOC) can transport the same amount of oxygen as the hemoglobin in RBC. These cross-linked hemoglobin (Hb) molecules are small and their solutions have lower viscosity than that of whole blood, so when they circulate in vivo they more quickly release oxygen to tissues than RBC. Consequently, they can provide gas transport (oxygen delivery and waste removal) at low blood pressure levels and can move through blood capillaries which RBC are not able to enter due to very low pressure caused by hemorrhagic shock or other pathological conditions. One of these products, Hemopure (Biopure, Cambridge, MA, USA), a glutaraldehyde-crosslinked bovine hemoglobin, has been approved and successfully used in clinics of South Africa for several years and is currently in the Phase III stage of clinical trials in the US. Very limited information is available on the effects of blood substitutes on rheological properties of blood. Oxygent demonstrated no effect of the emulsion on hemostasis, hemolysis and blood rheology during cardiopulmonary bypass surgery [68]. In vitro studies performed with a product of the first generation of PFCs, FluosolDA, showed that 20% plasma replacement in human blood significantly reduced low shear viscosity and erythrocyte sedimentation rate (ESR) and hence RBC aggregation. It also led to a significant reduction of hemolysis caused by blood exposure to mechanical stress. No effect on RBC deformability was found [69, 70]. Some experimental HBOCs such as polyoxyethylene (POE)-conjugated pyridoxalated Hb (POE-PLP-Hb) and hydroxyethylstarch-conjugated Hb (HES-XLHb) were found to induce strong aggregation of human RBC [71]. On the other hand, intra-molecular cross-linked hemoglobins (XLHb), and a cellular type of HBOC (Hb-vesicles, HbV) where the surface is modified with POE (POE-HbV), did not cause the formation of RBC aggregates. Stetter, et al. showed that the admixture of genetically engineered recombinant hemoglobin, consisting of two alpha chains and one beta chain (rHb 1.1), to plasma or to RBC suspensions at constant Hb concentration, led to a dose-dependent decrease in their viscosities due to the very low viscosity of rHb 1.1 solution (0.80 mPa.s) at 37°C at concentration of 50 g/L in phosphate-buffered saline. The rHb 1.1 did not affect red cell aggregation or the deformability of red cells or white cells, as measured by the cells' transit time through small pores [72]. Finally, Menu, et al. compared the rheological effects of dextran-benzene-tetra-carboxylate hemoglobin (Dex-BTC-Hb), stroma free hemoglobin and several plasma expanders (dextran 40, Plasmacair), modified fluid gelatin (MFG-Plasmion) or hydroxyethyl starch (HEAElohes)) and found that Dex-BTC-Hb increased RBC aggregation similarly to plasma expanders and did not affect RBC deformability [73, 74].

212


1.7. Blood Salvage Devices There has been a growing interest in autologous blood conservation due to a chronic shortage of allogenic blood [75, 76], increasing costs of blood products, immunosuppressive effects of allogeneic transfusion, increasing recognition of transfusion related acute lung injury [77], and the risk of various transfusiontransmitted diseases [78]. Multiple strategies can be applied to avoid allogeneic transfusion during surgery. The primary techniques involve preoperative erythropoietin administration and iron supplementation, preoperative autologous donation, acute normovolemic hemodilution, and the application of cell salvage (CS) systems which wash collected blood. Of these techniques, CS offers the greatest ability to avoid allogeneic transfusion if applied optimally [75]. Despite significant shear stresses, large blood-air interfaces, and negative pressure which can be as low as -300 mmHg,, RBC survival is quite high since one pass conditions are used. Hemolysis ranging from 0.2 to 2.3% was found in our recent study [75] and from 0.3 to 2.85% in Gregoretti’s study [79]. In both cases, the likely cause of hemolysis was mixing of air with the blood at a vacuum pressure of 300 mmHg. Moreover, washing and filtering RBC before return to the patients obviously contributed to the increased quality of these cells. The literature is virtually devoid of data on the effects of blood salvage procedures on the rheological properties of RBC. Durand et al, found a slight decrease in RBC deformability and a 20-fold increase in plasma free hemoglobin persisting despite successive washes [80]; these changes could be related to sub-lethal damage to RBC. On the other hand, Dalen and Engstrom found an increase in filterability of drained blood measured as 5 micronpore filterability which they explained by reduction of leukocyte count due to the leukocyte reduction during filtration of drained blood [81].

2. Mechanically Induced Hemolysis Mechanical trauma of flowing blood has been studied extensively in vivo and in vitro for over 30 years. The works of Nevaril, et al. in 1968 [82], Leverett, et al. in 1972 [83] and Williams in 1972 [84] were among the first to introduce the concept of critical lytic thresholds of shear stress. Historically, the most often studied aspect of blood trauma was shear-induced hemolysis [85-90]. The early experiments established the inverse relationship between critical shear stress and time exposure: blood is able to sustain greater shear for shorter periods of time. The threshold level of hemolysis was originally approximated to be 4000 N/m2 for exposure time as small as 10-5 s [85] and about 150 N/m2 for exposure times on the order of 102 s [83]. Most of the published works on blood trauma address lethal blood damage with destruction of over 50% of RBC, which is orders of magnitude greater than clinically acceptable blood contacting devices. These studies are inevitably fraught with challenges and limitations. The most vexing is the impossibility of uncoupling exposure time and shear stress in an independent experiment: shear, by definition, requires a gradient of velocity, and exposure time is inversely proportional to velocity. Consequently, investigators have devised numerous experiments, using a variety of flow geometries, such as cone-plate, capillary tube, annular Couette with crossflow [91], jet [51, 83, 92], nozzle [93], and piston [94]. Additional experimental perturbations have also been introduced to achieve desired conditions: for instance, to provide a very high shear stress, RBC may be suspended in an artificial medium with very high viscosity [95], such as 34% albumin


213

solution [83] or concentrated dextran solutions [85]. Unfortunately, some of these techniques may introduce artifacts by altering oncotic pressure which may independently cause RBC rupture. When RBC are overstretched or fragmented by non-physiological blood flow conditions, hemoglobin released into plasma is bound by the plasma protein haptoglobin and this hemoglobin-haptoglobin complex is removed from the vascular system by the reticuloendothelial system. This process of free Hb removal, in which the molecule is reduced to protein and iron with the iron recycled, is physiological since the aged RBC which become smaller and less deformable are continually removed from circulation. Approximately 0.1 mg/kg/min of free Hb, which is equivalent to 5 mg/dl/hr, can be removed by the reticuloendothelial system in adult subjects [96]. Circulating free hemoglobin up to 130 mg/dl may be bound to form hemoglobinhaptoglobin complexes, and a renal clearance of 14 g per day of free hemoglobin (~12 mg/dl/hr) is possible. Hemoglobinuria occurs at plasma Hb levels greater than 100 mg% [97], with renal damage produced at plasma Hb levels greater than 300 mg% [97]. The maximal physiological shear stress in the vascular system is ~10 N/m2 [98]. Although it is not directly related to the focus of this chapter, it is worth noting that one hour of exposure to wall shear stress exceeding ~40 N/m2 was found to damage endothelial cells leading to their destruction [99]. In addition to the direct effects of plasma free hemoglobin which can cause endothelial dysfunction and vasomotor instability by reducing nitric oxide bioavailability [1], hemolysis can cause an increase in blood pressure, renal damage, and activation of platelets. By the end of the last century, it was already established that hemolysis promotes hypercoagulation and intravascular thrombosis [100]. Experiments on platelet aggregation induced by the addition of stroma-free RBC lysate revealed that a concentration of lysate resulting in a hemoglobin concentration of 30 mg/dl and higher induced a spontaneous aggregation of platelets [101]; the level of platelet aggregation was found to linearly correlate with the concentration of lysate. Sutera’s review article [86] suggested that a leakage of ions and the larger molecules such as ATP and 2-3 DPG through 10-20 Å micropores of the stretched RBC membrane can occur much earlier than leakage of Hb molecules. In vitro studies by Alkhamis, et al. [102] suggest that RBC release a significant fraction of their ADP (2% at a shear rate of 5680 s-1), which is enough to induce platelet aggregation. Low levels of hemolysis have also been shown to drastically increase RBC aggregation at low shear conditions [9]. Finally, plasma free hemoglobin may have a toxic effect on the cardiovascular system due to its ability to bind nitric oxide, an endothelium-derived relaxing factor [1]. 2.1. Methods of Assessment of Mechanical Hemolysis The common parameter used to assess the hemolytic potential of circulatory-assist devices, oxygenators and other blood contacting devices in vitro and in vivo is an increase in the concentration of hemoglobin released into the plasma and not bound by haptoglobin. It is often represented by a Normalized Index of Hemolysis (NIH) which is proportional to the rate of plasma free hemoglobin (plfHb), volume of circulating blood, hematocrit, and inversely proportional to flow rate [103]: NIH ( g / 100 L) =

'Hb x V x ( 100 - Ht) / 100 [Q * T ] / 100

214


where 'Hb is an increase in plasma free Hb concentration (g/L) during the test time (min), V is a total circuit volume (L), Q is a flow rate (L/min), Ht is the blood hematocrit value (%) and T is the duration of the test (min). Recently, a revised index of hemolysis was proposed [104] since the traditional index (NIH) was found not to be scalable to the small devices required for pediatric applications in which flow rates as low as 0.2 - 0.5 L/min are used. This new Hemolysis Rate Index (HRI, g/hr)) does not depend on the pump flow rate and represents hemolysis obtained during testing of circulatory-assist devices (T is the duration of the test (hrs)). HRI ( g / hr ) =

'Hb x V x ( 100 - Ht) / 100 T

Figure 2 shows an example of plasma free Hb dynamics during hemolysis tests using a pediatric centrifugal pump and calculated NIH and HRI as functions of the flow rate [104].

0.020

0.20 NIH (g/100 L) HRI (g/hr) Delta plfHb (g/L) Linear (Delta plfHb (g/L))

0.15

Power (NIH (g/100 L))

NIH; HRI

Linear (HRI (g/hr))

0.010

0.10

0.005

0.05

0.000

Delta plfHb (g/L)

0.015

0.00 0

0.5

1

1.5

2

Flow rate (L/min)

Figure 2. Dynamics of plasma free Hb and calculated NIH and HRI as functions of flow rate.

3. Sub-Lethal RBC Damage It has been noticed for a long time that RBC could be altered during prolonged exposure to shear stresses, which are well below any hemolytic threshold [86]. The concept of sub-lethal RBC damage was first introduced by Galletti et al. in 1962 [105], who discovered the development of anemia and shortened RBC life spans in animals which had undergone extracorporeal perfusion for 10 to 48 hours. This sub-lethal trauma is much more difficult to detect and characterize than total lysis. The experiments of Sandza, et al. [106], in which an isolated rabbit spleen was perfused by a mixture of sheared and unsheared autologous RBC, showed that the spleen could


215

"recognize" and selectively remove cells that had been exposed to less than 10 N/m2 shear stress for two hours, suggesting some changes had occurred to the mechanical properties of the RBC. In the circulation, each RBC enters the spleen every 20-30 minutes and is “checked” for its ability to pass through tiny slits in the red pulp. Thus, the body applies a natural “rheometer” to measure RBC deformability and removes RBC which do not pass the “flexibility test”. Therefore, deformability of RBC appears to be the major determinant of their survival in the vascular system which is approximately 100120 days for normal human RBC. It has been found that naturally aged RBC are removed from the circulation by the spleen as they become less deformable [107]. Thus the RBC exposed to extensive mechanical stress would also be removed from the circulation because of their lower deformability, and a sub-lethal trauma could be considered as analogous the RBC aging process in this regard. In some reports, clinically-observed chronic anemia in patients supported with mechanical circulatory assist devices was thought to be related to undetermined mechanisms [108]; however, it might directly result from sub-lethal RBC damage and shortened lifespan. Table 1 presents reported values of sub-lethal shear stresses/exposure times collected from various sources. Table 1: Literature values for sub-lethal RBC damage 2

Exposure time, s

Source

4000

0.00001

Blackshear, 1972 [85]

300

0.4

Paul et al., 2003 [90]

200

0.6

Paul et al., 2003 [90]

120

15

Baskurt et al., 2004 [113]

56

180

Lee et al., 2004 [114]

Shear stress, N/m

10 144,000 Physiological stress* *It is assumed that maximal physiological stress of 10 N/m2 is applied during 1 second during the one minute blood circulation time and that the average RBC life span is ~100 days (1 s/min)*(60 min/hour)*(24 hour/day)*(100 days) = 144,000 s.

3.1. Effect of mechanical stress on RBC deformability Deformability is extremely important for the RBC to be able to flow through the smallest capillaries in the microcirculation and provide adequate transport of gases. Cardiopulmonary bypass (CPB) has been found to decrease red cell deformability, which may lead to impaired microcirculation and diminished tissue oxygen supply [10]. Long-term experience with mechanical circulatory-assist devices and other bloodcontacting artificial devices has also demonstrated the effects of sub-lethal damage [1621]. In addition to hemolysis, the extremely high fluid dynamic forces within these devices may cause denaturing of proteins, activation of platelets and leukocytes, increased concentrations of inflammatory mediators, complement activation, and unfavorable changes in the mechanical properties of blood cells including reduction in RBC deformability and increase in RBC aggregation [86, 109, 110]. Results of a number of basic hemorheological studies performed in vitro on the effects of mechanical stress on RBC deformability support observations in pre-clinical and clinical tests with blood-contacting artificial organs. The decrease of RBC

216


deformability after exposure of blood to mechanical stress has been reported in numerous studies [10, 111-116]. Baskurt, et al. evaluated the effects of mechanical stress on human RBC deformability by subjecting cells to a uniform fluid shear stress of 120 Pa for 15-120 seconds at 37°C. This level of stress induced significant impairment of RBC deformability as assessed by ektacytometry with no visible hemolysis observed in the sheared RBC suspensions [113]. Mizuno, et al. [115] examined the relative amount of membrane band-3 protein in relation to RBC deformability using flow cytometry. The continuous exposure to a low shear stress caused cell shrinkage, an increase in band 3 density, and a decrease in the deformability of the RBC membrane. It was concluded that under low shear stress, the RBC cytoskeleton shows gradual destruction which is exhibited as a disorder of band 3 distribution. The related membrane dysfunction includes decreases in RBC deformability and stability [115]. Kameneva, et al. showed that the decrease in RBC deformability induced by mechanical stress was aggravated by hypothermia and a reduction of plasma protein concentration [10]. Dao, et al., have found a reduction of human RBC filterability accompanied by significant changes in the RBC lipid bilayer after exposure of RBC suspensions to sub-lethal shear stress of 100 Pa for 120 s [117]. Some methods for reducing RBC sensitivity to mechanical stress and for improving their deformability are reviewed below. 3.2. Effect of Mechanical Stress on RBC Aggregation Red blood cells of humans and some mammalian species, such as horse, feline, and antelope, form aggregates which can be observed in blood both in vitro and in vivo under no or low flow conditions. There are at least two proposed mechanisms for RBC aggregation (i.e., bridging, depletion), with the depletion model currently the most popular; RBC aggregation is caused by fibrinogen and other high molecular weight protein molecules in blood plasma. However, in many other mammalian species such as bovine, ovine and goats, RBC do not aggregate at all in spite of often higher fibrinogen concentrations in plasma. The physiological significance of RBC aggregation is not yet fully defined, but it is known that RBC aggregation promotes the development of a cell-depleted plasma layer at the wall of blood vessels [118], thereby reducing wall shear stress loading and reducing release of endothelium-dependent vasodilators. On the other hand, RBC aggregation may help to maintain a balance between arterial and venous resistance in athletic species during the transition from exercise to rest and vice versa; the mechanism involves a significant increase in blood viscosity at low flow states and a decrease in blood viscosity with increasing flow rate [119]. Furthermore, it is well known that an increase in RBC aggregation relative to the normal physiological level is usually associated with development of an inflammatory reaction, infection or some other pathological state [120, 121], and is usually related to an increase in the concentrations of plasma fibrinogen, immunoglobulin, and other plasma macromolecules. Although RBC aggregation is of great basic scientific and clinical interest and a large number of clinical and experimental studies have aimed at finding and testing drugs to reduce RBC aggregation, the advantages and disadvantages of having low or no RBC aggregation are not well described. Figure 3 demonstrates RBC aggregation in normal human blood (left) and in the blood of cardiac patient (right).


217

Figure 3. Rouleaux formation in normal human blood (left) and severe RBC aggregation in the blood of cardiac patient (right).

Clinical observations show that RBC aggregation is usually decreased in patients undergoing CPB due to significant hemodilution [122-125]. Morariu, et al. hypothesized that a drop in RBC aggregation was a disadvantage contributing to a loss of shear stress signaling leading to endothelial activation [122]. On the other hand, recipients of implantable circulatory-assist devices demonstrate a very significant increase of blood viscosity, especially at low shear rates where RBC aggregation plays an important role [16, 19-21]. Our studies of blood from patients with heart-assist devices showed a significant increase in RBC aggregation compared to healthy controls as assessed by low shear viscosity and erythrocyte sedimentation rate [116]. This increase in low shear viscosity was highly correlated with increases of blood fibrinogen concentration. Increased RBC aggregation was also observed in hemodialysis patients in spite of significantly reduced hematocrit [126, 127]. Our experience during in vivo animal evaluation of an axial-flow ventricular assist device (VAD), the Nimbus AxiPump, revealed an interesting rheological phenomenon. Sheep RBC started to form RBC aggregates, typically on the third post-operative day, simultaneously with increasing fibrinogen levels, whereas normal sheep blood does not demonstrate RBC aggregation. To investigate the relative effects of mechanical stress and elevated fibrinogen levels on ovine RBC aggregation, we conducted in vitro studies using blood from control sheep. These studies indicated that neither mechanical trauma nor elevated fibrinogen alone caused RBC aggregation as seen in vivo. However, the combination of mechanical stress and elevated fibrinogen did cause this unusual effect for sheep blood [109]. Another example of such unusual RBC aggregation, observed in the blood of a calf implanted with a VAD, is presented in Figure 4. It is well known that in spite of a quite high concentration of fibrinogen in normal bovine blood (200-600 mg/dl), red blood cells in this blood do not aggregate. After device implantation, the animals usually develop an increased concentration of fibrinogen due to the normal post-surgical reaction (i.e., acute phase response). Figure 4A shows a microscopic picture of blood of one of the VAD implanted animals (“A”) where strong RBC aggregation is seen as the fibrinogen concentration reached 1000 mg/dl. Figure 4B shows blood of a normal calf (“B”) with no tendency to RBC aggregation. When RBC of this normal calf (“B”) were resuspended in the plasma of the implanted calf (“A”) they still did not show a tendency to aggregation (Figure 4C). However, when RBC of calf “A” were resuspended in the plasma of normal calf “B”, these RBC aggregated although the fibrinogen concentration (200 mg/dl) was much lower than one in their autologous plasma (Figure 4D). These observations illustrate that mechanical stress to RBC

218


increases their ability to aggregate, potentially due to a decrease in membrane sialic acid content and hence cell negative charge which reduces repulsive forces between RBC [116, 128].

A

B

C

D

Figures 4A-4D. RBC aggregation in blood from a calf with implanted VAD (A), from a control animal (B), for control cells in plasma from a VAD animal (C) and for VAD cells in control plasma. (see text).

4. Mechanical Fragility of Red Blood Cells RBC mechanical fragility is a critical parameter in the evaluation of sub-lethal mechanical blood trauma. It is a more sensitive measurement than lethal blood trauma (i.e., hemolysis), and can be used following implantation of circulatory-assist devices even without measurable hemolysis [129]. Probably the first method for RBC mechanical fragility measurements involved one ml of blood mixed with 60 mg of silica particles for 5 min, with the resulting hemolysis compared with samples which were mixed without silica particles [130]. A detailed review of various methods for assessment of RBC mechanical fragility published by Gu et al. [131] concluded that the simple method described in [116, 129, 132, 133] is perhaps the best for routine evaluation of mechanical fragility. This simple method is as follows: anticoagulated whole blood or RBC suspension samples (3 ml) are placed in standard 7 ml “red top” vacuum tubes for blood collection along with five 1/8” diameter stainless steel spheres. The tubes are rocked on a rocker platform for a constant exposure time, thereby subjecting the RBC to mechanical stress induced by the rolling of stainless steel spheres. Control samples of the same blood or RBC suspension are placed in similar tubes with stainless steel spheres added but are not rocked. For comparative studies all samples are prepared at a standard hematocrit of 30%, and thus for tests with whole blood, adjustment to a value of 30% is achieved by


219

adding or removing autologous plasma; whole blood should be anticoagulated with heparin or EDTA to minimize dilution effects. The amount of free hemoglobin released in each sample is measured after centrifugation of the samples and measuring the supernatant using a spectrophotometer at 540 nm. RBC mechanical fragility is then calculated using the formula:

MFI =

Hb final - Hbcont x100 Hbw.bl. - Hbcont

where Hbfinal is the plasma free hemoglobin (mg%) in the sample exposed to mechanical stress on the rocker, Hbcont is the plasma free hemoglobin (mg%) in control samples which were not rocked, and Hbw.bl. is the hemoglobin concentration in whole blood (mg%). Methods for RBC mechanical fragility measurements are susceptible to the same sources of errors as for those described for lethal blood trauma. The storage time and handling of blood during collection may induce changes in RBC mechanical properties including an increase in their fragility [134, 135]. In addition, RBC mechanical fragility has been found to depend on the medium in which RBC are suspended, and tends to increase with a reduction of plasma protein concentration [132, 136, 137]. Comparing the mechanical fragility of human and animal RBC, the results of animal tests might be used to predict mechanical blood trauma in patients implanted with devices. Most of in vitro studies of hemolysis and sub-lethal mechanical blood trauma in blood-contacting artificial devices are performed using animal blood, commonly adult bovine blood obtained from local slaughter houses. However, most of the pre-clinical in vivo tests of artificial devices are performed in calves. In vivo studies of pediatric devices require yet smaller experimental animals, such as young sheep, goats and pigs. In addition to age and maturity, the RBC of various species may have different sensitivity to mechanical stress [138, 139]. Therefore, the mechanical fragility of human RBC must be calibrated to these animals models to accurately predict mechanical blood trauma that may occur clinically. For example, the fragility of adult human RBC is about 2-fold less than that of ovine RBC [138], slightly higher than that of bovine RBC [132, 138] and about the same as that of porcine RBC [139]. Therefore, animal tests with a particular device will result in varying degrees of hemolysis depending on the animal model. It is worth noting that RBC mechanical fragility in human neonates is higher than that in adults [140]. Moreover, in human blood RBC mechanical fragility was shown to be higher in male compared to pre-menopausal female blood [133]. In the blood of all tested species, young RBC were found to be less fragile than senescent RBC [116, 133, 141].

5. Methods of Protection of Red Blood Cells from Mechanical Trauma Several studies have shown that there are potential ways to reduce blood trauma caused by extensive mechanical stresses in blood-contacting artificial devices. Kamada, et al. based their experience on using albumin additives to the priming solution (i.e., 50 g of human albumin added to the priming volume) during extracorporeal circulation in patients undergoing coronary bypass surgery. These authors suggested that maintaining adequate levels of plasma albumin is important in preventing erythrocyte crenation during EC and improving microcirculatory flow in patients undergoing open heart surgery [142, 143]. In a very elegant study of the effect of the nonionic surfactant

220


poloxamer 188 on RBC-induced platelet aggregation, Armstrong, et al. demonstrated that the polymer protected RBC from damage caused by mechanical agitation, thereby reducing leakage of ADP from red cells and inhibiting ADP-induced platelet aggregation [144]. Kameneva, et al. studied the effects of whole plasma and plasma proteins on RBC mechanical damage using bovine RBC suspended in different media inasmuch as earlier reports indicated the protective effects of certain plasma components [136]. In these studies, cells were exposed to the same mechanical stress using the standard RBC mechanical fragility test [132], with experiments performed at room temperature with control of osmolality and viscosity of the suspending media. The lowest hemolysis was obtained for RBC suspended in serum, plasma, and albumin solution, while hemolysis in phosphate buffered saline (PBS) and dextran suspensions was over 3 times higher than in plasma (p<0.001). The presence of relatively small amounts of plasma (i.e., 25%) in the suspension media significantly (p<0.001) decreased hemolysis. These results suggested that blood plasma has a protective effect on RBC exposed to mechanical stress, and that a decrease in the concentration of plasma proteins due to hemodilution may elevate blood damage during extracorporeal circulation. Protective effects of a 4% modified fluid gelatin solution and a 4% albumin solution on blood bank stored RBC exposed to mechanical stress has been reported [137]. In addition, protection of RBC from mechanical trauma and a decrease in low shear viscosity and erythrocyte sedimentation rate were found after replacement of 20% of blood plasma with a perfluorochemical-based blood substitute in vitro [70]. Roberts, et al. showed that RBC deformability was significantly improved and plasma free hemoglobin was significantly reduced in CPB patients who were administered urea [145, 146]. RBC deformability in 15 hemodialysis patients tested before and after three months on L-carnitine supplementation (i.e., 30 mg/Kg body wt/dialysis session) showed significant improvement whereas it was impaired in control dialysis patients [147]. Beneficial effects of L-carnitine and acetyl-L-carnitine on RBC membrane stability and deformability was previously demonstrated by Arduini, et al. in vitro [148], and most recently an improvement of RBC deformability using L-carnitine was confirmed in a clinical study by Toptas, et al. [149]. Finally, in vitro studies of the effects of sub-lethal mechanical stress on RBC deformability demonstrated that the nitric oxide donor sodium nitroprusside prevented reduction in RBC deformability caused by a uniform shear stress [113].

Summary Mechanical blood trauma is a widely recognized challenge in development and use of blood-contacting devices. While mechanical forces responsible for this blood damage are harmful for all elements including blood cells and plasma components, this chapter mostly dealt with lethal and especially sub-lethal trauma to red blood cells and the associated effects on rheological properties of RBC and whole blood. Extensive mechanical stress applied to blood in blood-contacting artificial devices causes hemolysis and a sub-lethal blood trauma which is measured by a deterioration of RBC deformability and an increase in RBC ability to aggregate. A decrease in RBC deformability causes, in turn, an impediment of RBC passage through the smallest capillaries, a reduction in density of functioning capillaries and a shortening of RBC life span.


221

References 1.

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

20. 21.

P.C. Minneci, K.J. Deans, H. Zhi, P.S. Yuen, R.A. Star, S.M. Banks, A.N. Schechter, C. Natanson, M.T. Gladwin and S.B. Solomon, Hemolysis-associated endothelial dysfunction mediated by accelerated NO inactivation by decompartmentalized oxyhemoglobin, J. Clin. Invest. 115 (2005), 34093417. R.P. Rother, L. Bell, P. Hillmen and M.T. Gladwin, The clinical sequelae of intravascular hemolysis and extracellular plasma hemoglobin: a novel mechanism of human disease. JAMA 293 (2005), 16531662. United Network for Organ Sharing, Released on 07/19/2006; http://www.unos.org/news/; accessed on 12/23/2006. National Kidney Foundation; http://www.kidney.org/transplantation/livingDonors/aboutHistory.cfm; accessed on 12/23/2006. L.J. Wursinger and R. Opitz, Hematological principles of hemolysis and thrombosis with special reference to rotary blood pumps. In: Proceeding of the International Workshop on Rotary Blood Pumps, H. Schima, H. Thoma, G. Weiselthaler and E. Wolner, Eds., Vienna, 1991, pp. 19-25. M. H. Kroll, J. D. Hellums, L. V. McIntire, A. I. Schafer and J. L. Moake, Platelets and shear stress, Blood 88 (1996), 1525-1541. P. L. Blackshear, Jr., F. D. Dorman and J. H. Steinbach, Some Mechanical Effects That Influence Hemolysis, ASAIO Trans. 11 (1965), 112-117. P. L. Blackshear, Jr., F. D. Dorman, J. H. Steinbach, E. J. Maybach, A. Singh and R. E. Collingham, Shear, wall interaction and hemolysis, ASAIO Trans. 12 (1966), 113-120. A. Seiyama, Y. Suzuki, N. Tateshi and N. Maeda. Viscous properties of partially hemolyzed erythrocyte suspension. Biorheology 28 (1991), 452. M.V. Kameneva, A. Undar, J.F. Antaki, M.J. Watach, J.H. Calhoon and H.S. Borovetz, Decrease in red blood cell deformability caused by hypothermia, hemodilution, and mechanical stress: factors related to cardiopulmonary bypass, ASAIO J. 45 (1999), 307-310. M.V. Kameneva, P..F Marad, J.M. Brugger, B.M. Repko, J.H. Wang, J. Moran and H.S. Borovetz, In vitro evaluation of hemolysis and sublethal blood trauma in a novel subcutaneous vascular access system for hemodialysis, ASAIO J. 48 (2002), 34-38. M.V. Kameneva, G.B. Burgreen, K. Kono, B.M. Repko, J.F. Antaki and M. Umezu, Effects of turbulent stresses on mechanical hemolysis. Experimental and computational analysis, ASAIO J. 50 (2004), 418-423. L.O. Thompson, M. Loebe and G.P. Noon, What price support? Ventricular assist device induced systemic response. ASAIO J. 49 (2003), 518-526. C.N. Pierce and D.F. Larson, Inflammatory cytokine inhibition of erythropoiesis in patients implanted with a mechanical circulatory assist device, Perfusion 20 (2005), 83-90. C.N. Pierce, D..F Larson, F.A. Arabia and J.G. Copeland, Inflammatory mediated chronic anemia in patients supported with a mechanical circulatory assist device, J. Extra Corpor. Technol. 36 (2004), 10– 15. R.L. Kormos, H.S. Borovetz, B.P. Griffith and T.C. Hung, Rheologic abnormalities in patients with the Jarvik-7 total artificial heart. ASAIO Trans. 33 (1987), 413-417. P.L. Frattini, C. Watcher, T.C. Hung, R.L. Kormos, B.P. Griffith and H.S. Borovetz, Erythrocyte deformability in patients on left ventricular assist systems, ASAIO Trans. 35 (1989), 733-735. H.S. Borovetz, R.L. Kormos, B.P. Griffith and T.C. Hung, Clinical utilization of artificial heart, Crit Rev Biomed Eng. 17 (1989), 179-201. T.C. Hung, D.B. Butter, R.L. Kormos, Z. Sun, H.S. Borovetz, B.P. Griffith and C.L. Yie, Characteristics of blood rheology in patients during Novacor left ventricular assist system support, ASAIO Trans. 35 (1989), 611–613. T.C. Hung, D.B. Butter, C.L. Yie, R.L. Kormos, H.S. Borovetz, B..P Griffith and R.L. Hardesty, Effects of long-term Novacor artificial heart support on blood rheology, ASAIO Trans. 37 (1991), M312–M313. T.C. Hung, D.B. Butter, C.L. Yie, Z. Sun, H.S. Borovetz, R.L. Kormos, B.P. Griffith and R.L. Hardesty, Interim use of Jarvik-7 and Novacor artificial heart: blood rheology and transient ischemic attacks (TIA's), Biorheology 28 (1991), 9-25.

222


22. L.K. von Segesser, Cardiopulmonary support and extracorporeal membrane oxygenation for cardiac assist, Ann. Thorac. Surg. 68 (1999), 672-677. 23. T. Kammerer, A. Fuchs, M. Mendonca, S.H. Dabritz and R. Kozlik-Feldmann, Extracorporeal membrane oxygenation (ECMO) as cardiac assist device in pediatrics, Anasthesiol. Intensivmed. Notfallmed. Schmerztherap. 38 (2003), 514-521. 24. Y.J. Gu, P.W. Boonstra, R. Graaff, A.A. Rijnsburger, H. Mungroop and W. van Oeveren, Pressure drop, shear stress, and activation of leukocytes during cardiopulmonary bypass: a comparison between hollow fiber and flat sheet membrane oxygenators, Artif. Organs 24 (2000), 43-48. 25. T.L. Zach, R.H. Steinhorn, M.K. Georgieff, M.M. Mills and T.P. Green. Leukopenia associated with extracorporeal membrane oxygenation in newborn infants, J. Pediatr. 116 (1990), 440-444. 26. A. Hakoshima, H. Goto, K. Abe, K.T. Benson, J.F. Moran and K. Arakawa, Alteration of red cell deformability during extracorporeal bypass: membrane versus bubble oxygenator, Cardiothorac. Anesth. 3 (1989), 189-192. 27. M.W. Lim, The history of extracorporeal oxygenators, Anaesthesia 61 (2006), 984-995. 28. H. Iwahashi, K. Yuri and Y. Nose, Development of the oxygenator: past, present, and future, Artif. Organs 7 (2004), 111-120. 29. A. Belboul, N. Al-Khaja, M. Gudmundsson, H. Karlsson, T. Uchino, B. Liu, A. El-Gatit, A.Bjell, D. Roberts and G. William-Olsson, The influence of heparin-coated and uncoated extracorporeal circuits on blood rheology during cardiac surgery, J. Extra Corpor. Technol. 25 (1993), 40-46. 30. T. Hirayama, D.G. Roberts, M. Allers, A. Belboul, N. al-Khaja and G. William-Olsson, Association between pulmonary dysfunction and reduced red cell deformability following cardiopulmonary bypass, Scand. J. Thorac. Cardiovasc. Surg. 22 (1988), 175-177. 31. N. al-Khaja, A. Belboul, P. Bergman, D. Roberts and G. William-Olsson, Cutaneous microcirculation and blood rheology following cardiopulmonary bypass. Laser Doppler flowmetric and blood cell rheologic studies, Scand. J. Thorac. Cardiovasc. Surg. 22 (1988), 149-153. 32. T. Hirayama, H. Yamaguchi, M. Allers, D. Roberts and G. William-Olsson, Changes in red cell deformability associated with anaesthesia and cardiopulmonary bypass in open-heart surgery, Scand. J. Thorac. Cardiovasc. Surg. 19 (1985), 257-262. 33. L. Spry, The diagnosis of hemolysis during hemodialysis, Seminars in Dialysis, 12(1999), 205. 34. D. S. De Wachter, P. R. Verdonck, J. Y. De Vos and R. O. Hombrouckx, Blood trauma in plastic haemodialysis cannulae, Int. J. Artif. Org. 20 (1997), 366-370. 35. D.S. De Wachter, P.R. Verdonck, R.F. Verhoeven and R.O. Hombrouckx, Red cell injury assessed in a numeric model of a peripheral dialysis needle. ASAIO J. 42 (1996), M524-M529. 36. M. Dhaene, B. Gulbis, N. Lietaer, N.Gammar, C. Thayse, H.A. Ooms and J.L. Vanherweghem, Red blood cell destruction in single-needle dialysis. Clin. Nephrol. 31 (1989), 327-331. 37. F.F. Ibrahim, M.M. Ghannam and F.M. Ali, Effect of dialysis on erythrocyte membrane of chronically hemodialyzed patients, Ren. Fail. 24 (2002), 779-790. 38. P. Zachee, A. Ferrant, R. Daelemans, L. Coolen, W. Goossens, R.L. Lins, M. Couttenye, M.E.De Broe and M.A. Boogaerts, Oxidative injury to erythrocytes, cell rigidity and splenic hemolysis in hemodialyzed patients before and during erythropoietin treatment, Nephron 65 (1993), 288-293. 39. M.D. Scott, F.A. Kuypers, P. Butikofer, R.M. Bookchin, O.E. Ortiz and B.H. Lubin, Effect of osmotic lysis and resealing on red cell structure and function. J. Lab. Clin. Med. 115 (1990), 470-480. 40. D. Docci, C. del Vecchio, P. Salvi, F. Turci, G. Salvi, L. Cenciotti, E. Pretolani, Osmotic fragility of erythrocytes, cell deformability and secondary hyperparathyroidism in uremic patients on maintenance hemodialysis, Clin. Nephrol. 23 (1985), 68-73. 41. N. Sotirakopoulos, T. Tsitsios, M. Stambolidou, G. Athanasiou, M. Peiou, V. Kokkinou, K. Mavromatidis, The red blood cell deformability in patients suffering from end stage renal failure on hemodialysis or continuous ambulatory peritoneal dialysis, Ren. Fail. 26 (2004), 79-83. 42. S. Nikolaos, A. George, T. Telemachos, S. Maria, M. Yannis and M. Konstantinos, Effect of Lcarnitine supplementation on red blood cells deformability in hemodialysis patients, Ren. Fail. 22 (2000), 73-80. 43. S. Iijima, F. Otsuka, H. Kikuchi, K. Yamada, T. Nakajima, K. Yahiro and A. Kondo, Oral supplementation with gamma-linolenic acid extracted from Mucor circinelloides improves the deformability of red blood cells in hemodialysis patients, Nephron 86 (2000), 122-128. 44. K. Ludat, M. Paulitschke, E. Riedel and H. Hampl, Complete correction of renal anemia by recombinant human erythropoietin, Clin. Nephrol. 53 (2000), 42-49. 45. B.I. Shand, A.L. Buttimore, K.L. Lynn, R.R. Bailey and R.A. Robson, Effect of hemodialysis and recombinant human erythropoietin on determinants of blood viscosity, Ren. Fail. 16 (1994), 407-413. 46. P. Zachee, A. Ferrant, R. Daelemans, L. Coolen, W. Goossens, R.L. Lins, M. Couttenye, M.E. De Broe and M.A. Boogaerts, Oxidative injury to erythrocytes, cell rigidity and splenic hemolysis in hemodialyzed patients before and during erythropoietin treatment, Nephron 65(1993), 288-293.


223

47. T.S. Andersen, P. Johansen, B.O. Christensen, P.K. Paulsen, H. Nygaard and J.M. Hasenkam, Intraoperative and postoperative evaluation of cavitation in mechanical heart valve patients, Ann. Thorac. Surg. 81 (2006), 34-41. 48. T. Saito, Y. Misawa, K. Fuse and H. Konishi, The CarboMedics prosthetic heart valve: experience with 180 implants, Artif. Organs 8 (2005), 51-55. 49. J.M. Vesey and C.M. Otto. Complications of prosthetic heart valves. Curr. Cardiol. Rep. 6 (2004), 10611. 50. W.L. Lim, Y.T.Chew, T.C. Chew and H.T. Low Pulsatile flow studies of a porcine bioprosthetic aortic valve in vitro: PIV measurements and shear-induced blood damage, J. Biomech. 34 (2001), 1417-1427. 51. R. Zimmer, A. Steegers, R. Paul, K. Affeld and H. Reul, Velocities, shear stresses and blood damage potential of the leakage jets of the Medtronic Parallel bileaflet valve, Int. J. Artif. Org. 23 (2000), 4148. 52. J.T. Ellis, T.M. Wick and A.P. Yoganathan, Prosthesis-induced hemolysis: mechanisms and quantification of shear stress, J. Heart Valve Dis. 7 (1998), 376-386. 53. T.C. Lamson, G. Rosenberg, D.B. Geselowitz, S. Deutsch, D.R. Stinebring, J.A. Frangos and J.M. Tarbell, Relative blood damage in the three phases of a prosthetic heart valve flow cycle, ASAIO J. 39 (1993):M626-M633. 54. N. Ishibashi, H. Kasegawa, T. Koyanagi and T. Ida, Mechanism of hemolysis after mitral valve repair and new surgical management: prosthetic annuloplasty ring covered with autologous pericardium, J. Heart Valve. Dis. 14 (2005), 588-591. 55. M. Suedkamp, A.J. Lercher, F. Mueller-Riemenschneider, K. LaRosee, P. Tossios and U. Mehlhorn, Hemolysis parameters of St. Jude Medical: Hemodynamic Plus and Regent valves in aortic position, Int. J. Cardiol. 95 (2004), 89-93. 56. B.K. Lam, D.M. Cosgrove, S.K. Bhudia and A.M. Gillinov, Hemolysis after mitral valve repair: mechanisms and treatment, Ann. Thorac. Surg. 77 (2004), 191-195. 57. Y. Shapira, O. Bairey, M. Vatury, H. Magen-Nativ, M. Prokocimer and A. Sagie, Erythropoietin can obviate the need for repeated heart valve replacement in high-risk patients with severe mechanical hemolytic anemia: case reports and literature review, J. Heart Valve Dis. 10 (2001), 431-435. 58. G. Ismeno, A. Renzulli, A. Carozza, M. De Feo, M. Iannuzzi, P. Sante, M. Cotrufo, Intravascular hemolysis after mitral and aortic valve replacement with different types of mechanical prostheses. Int. J. Cardiol. 69 (1999), 179-183. 59. F.A. Kuypers, Red cell membrane damage, J. Heart Valve Dis. 7 (1998), 387-395. 60. T. Hirayama, D. Roberts and G. William-Olsson, Mechanical trauma to red blood cells caused by Björk-Shiley and Carpentier-Edwards heart valves, Scand. J. Thorac. Cardiovasc. Surg. 19 (1985), 253-256. 61. N. Barshes, A. Gay, B. Williams, A. Patel and S. Awad, Support for the acutely failing liver: A comprehensive review of historic and contemporary strategies, J. Am. Coll. Surg. 201 (2005), 458-476. 62. A. Sechser, J. Osorio, C. Freise and R.W. Osorio, Artificial liver support devices for fulminant liver failure, Clin. Liver Dis. 5(2001), 415-430. 63. P. Rosenthal, Is There a Future for Liver-assist Devices? Curr. Gastroenterol. Rep. 2 (2000), 55-60. 64. E.S. Tzanakakis, D.J. Hess, T.D. Sielaff and W.S. Hu, Extracorporeal tissue engineered liver-assist devices, Ann. Rev. Biomed. Engin. 2(2000), 607-632. 65. J.C. Gerlach, Bioreactors for extracorporeal liver support, Cell Transplant. 15 (2006), Suppl 1, S91S103. 66. T.M. Chang, Artificial cells for replacement of metabolic organ functions, Artif. Cells Blood Substit. Immobil. Biotechnol. 31 (2003), 151-161. 67. G.V. Mazariegosa, J.F. Patzer II, R.C. Lopez, M. Giraldo, M.E. deVera, T.A. Grogan, Y. Zhu, M.L. Fulmer, B.P. Amiot and D.J. Kramer, First Clinical Use of a Novel Bioartificial Liver Support System (BLSS). Am. J. Transplant. 2 (2002), 260-266. 68. J.G. Riess, Perfluorocarbon-based oxygen delivery, Artif. Cells Blood Substit. Immobil. Biotechnol. 34 (2006), 567-580. 69. M.V. Kameneva, J.F. Antaki, H. Konishi, J.J. Whalen, J.P. Kerrigan, M.J. Watach, R.L. Kormos, B.P. Griffith, H.S. Borovetz, Effect of perfluorochemical emulsion (FLUOSOL®) on blood trauma and hemorheology, ASAIO Trans. 40 (1994), M576-M579. 70. M.V. Kameneva, H.S. Borovetz, J.F. Antaki, P. Litwak, W. Federspiel, R.L. Kormos and B.P. Griffith, Effect of perfluorochemical emulsion on hemorheology and hemodynamics. Possible mechanisms and future applications, Adv. Exp. Med. Biol. 411 (1997), 383-390. 71. H. Sakai, M. Yuasa, H. Onuma, S. Takeoka and E. Tsuchida, Synthesis and physicochemical characterization of a series of hemoglobin-based oxygen carriers: objective comparison between cellular and acellular types, Bioconjug. Chem. 11 (2000), 56-64.

224


72. M.N. Stetter, G.M. Baerlocher, H.J. Meiselman and W.H. Reinhart, Influence of a recombinant hemoglobin solution on blood rheology, Transfusion 37 (1997), 1149-1155. 73. P. Menu, D. Longrois, B. Faivre, M. Donner, P. Labrude, JF. Stoltz and C. Vigneron, Rheological behaviour of red blood cells suspended in hemoglobin solutions. In vitro study comparing dextranbenzene-tetra-carboxylate hemoglobin, stroma free hemoglobin and plasma expanders, Transf. Sci. 20 (1999), 5-16. 74. P. Menu, W. Bleeker, D. Longrois, A. Caron, B. Faivre-Fiorina, S. Muller, P. Labrude and J.F. Stoltz, In vivo effects of Hb solutions on blood viscosity and rheologic behavior of RBCs: comparison with clinically used volume expanders, Transfusion 40 (2000), 1095-1103. 75. J.H. Waters, B. Williams, M.H. Yazer, M.V. Kameneva, Modification of suction induced hemolysis during cell salvage, Anesth. Analg. 104 (2007), 684-687. 76. E.L. Wallace, W.H. Churchill, D.M. Surgenor, G.S. Cho and S. McGurk, Collection and transfusion of blood and blood components in the United States, 1994, Transfusion 38 (1998), 625-636. 77. S.B. Moore, Transfusion-related acute lung injury (TRALI): clinical presentation, treatment, and prognosis, Crit. Care Med. 34 (2006), S114-S117. 78. S Kleinman, P Chan and P Robillard, Risks associated with transfusion of cellular blood components in Canada, Transfus. Med. Rev. 17 (2003), 120-162. 79. S.Gregoretti, Suction-induced hemolysis at various vacuum pressures: implications for intraoperative blood salvage, Transfusion 36 (1996), 57-60. 80. F. Durand, M. Duchesne-Gueguen, J.Y. Le Bervet, P. Marcorelles, R. Tardivel, J.M. Vovan, M.C. Le Goff and B. Genetet, Rheologic and cytologic study of autologous blood collected with Cell Saver 4 during cesarean, Rev. Fr. Transfus. Hemobiol. 32 (1989), 179-191. 81. T. Dalen and K.G.Engstrom, Filterability of autotransfusion blood cells and plasma after total knee arthroplasty, Clin. Hemorheol. Microcirc. 19 (1998), 181-195. 82. C.G. Nevaril, E.C. Lynch, C.P. Alfrey, Jr. and J.D. Hellums, Erythrocyte damage and destruction induced by shearing stress, J. Lab. Clin. Med. 71 (1968), 784-790. 83. L.B. Leverett, J.D. Hellums, C.P. Alfrey and E.C. Lynch, Red blood cell damage by shear stress, Biophys. J. 12 (1972), 257-273. 84. A.R.Williams, Increased shear resistance of human erythrocytes in the presence of glucose, Nature 239 (1972), 162-163. 85. P.L. Blackshear, Mechanical hemolysis in flowing blood. In: Biomechanics: Its Foundation and Objectives. Y.C. Fung, N. Perrone, M. Anlikar, Eds., Prentice-Hall, Englewood Cliffs, N.J. 1972, pp. 501-528. 86. SP.Sutera, Flow induced trauma to blood cells, Circ. Res. 41(1977), 2-8. 87. A.M. Sallam and N.H.C. Hwang, Human red blood cell hemolysis in a turbulent shear flow: contribution of Reynolds shear stress, Biorheology 21 (1984), 783-797. 88. P.L. Blackshear and G.L. Blackshear, Mechanical Hemolysis, In: Handbook of Bioengineering, R. Skalak and S. Chien, Eds. McGraw-Hill Book Company, New York, 1986, pp. 15.1-15.19. 89. M. Geirsiepen, L.J. Wursinger, R. Opitz and H. Reul, Estimation of shear stress related blood damage in heart valve prosthesis - in vitro comparison of 25 aortic valves, Int. J. Artif. Org. 13(1990), 300-306. 90. R. Paul, J. Apel, S. Klaus, F. Schugner, P. Schwindke and H. Reul, Shear stress related blood damage in laminar couette flow, Artif. Org. 27(2003), 517-529. 91. G. Heuser and R. Opitz, A Couette viscometer for short time shearing in blood, Biorheology 17 (1980), 17-27. 92. R.J. Forstrom, A new measure of erythrocyte membrane strength: The jet fragility test, PhD thesis, University of Minnesota, Minneapolis, St. Paul, 1969. 93. M. Umezu, Y. Murayama and A. Nogawa, The effects of inner shapes of plastic connectors on blood in an extracorporeal circulation, In: 7th Int Conf on Biomed Eng, Singapore, 1992, pp. 197-199. 94. K.K. Yeleswarapu, J.F. Antaki, M.V. Kameneva and K.R. Rajagopal, A mathematical model for shearinduced hemolysis, Artif. Org. 19 (1995), 576–582. 95. J.V. Champion, P.F. North, W.T. Coakley and A.R. Williams, Shear fragility of human erythrocytes, Biorheology 8 (1971), 23-29. 96. E.F. Bernstein, P.L. Blackshear and K.H. Keller, Factors influencing erythrocyte destruction in artificial organs, Am. J. Surg. 114 (1967), 126-138. 97. G.H.R. Clowes, Extracorporeal maintenance of circulation and respiration, Physiol. Rev. 40 (1960), 826-919. 98. C.G. Caro, T.J. Pedley, R.C. Schroter and W.A. Seed, The mechanics of the circulation, Oxford University Press, CITY, 1978. 99. D.L. Fry, Acute vascular endothelial changes associated with increased blood velocity gradients, Circ. Res. 22 (1968), 165–197. 100. A.A. Schmidt, Zur Blutlehre, F. C. W. Vogel, Leipzig,1892, p. 234.


225

101. L.J. Wurzinger, P. Blasberg and H. Schmid-Schonbein, Towards a concept of thrombosis in accelerated flow: rheology, fluid dynamics, and biochemistry, Biorheology 22 (1985), 437-450. 102. T. M. Alkhamis, R. L. Beissinger and J. R. Chediak, Red blood cell effect on platelet adhesion and aggregation in low-stress shear flow. Myth or fact? ASAIO Trans. 34 (1988), 868-873. 103. K. Naito, K. Mizuguchi and Y. Nose, The need for standardizing the index of hemolysis, Artif. Org. 18 (1994), 7-10. 104. M.V. Kameneva, T.A. Snyder, P.J. Marascalco and J.F. Antaki, Validity of normalized index of hemolysis in pediatric mechanical circulatory assist devices, ASAIO J. 52(2006), 41A. 105. D.E. Brinsfield, M.A. Hoff, R.B. Geering and P.M. Galletti, Hematological changes in long-term perfusion, J. Appl. Physiol. 17 (1962), 531-538. 106. J.G. Sandza, R.E. Clark, C.S. Weldon and S.P. Sutera, Subhemolytic trauma of erythrocytes; recognition and sequestration by spleen as a function of shear, ASAIO Trans. 2 (1974), 457-462. 107. W.S. Ganong, Review of Medical Physiology, 15th ed. Appleton and Lange, Norwalk, CT, 1991, pp. 490-491. 108. C.N. Pierce, D.F. Larson, F.A. Arabia and J.G. Copeland, Inflammatory mediated chronic anemia in patients supported with a mechanical circulatory assist device, J. Extra Corpor. Technol. 36 (2004), 1015. 109. M.V. Kameneva, J.F. Antaki, K.C. Butler, M.J. Watach, R.L. Kormos, B.P. Griffith and H.S. Borovetz, A sheep model for the study of hemorheology with assisted circulation. Effect of an axial flow blood pump, ASAIO J. 40 (1994), 959-963. 110. S. Berger and E.W. Salzman, Thromboembolic complication of prosthetic devices, Progr. Hemostas. Thromb. 2 (1974), 273-309. 111. S.S. Lee, J.F. Antaki, M.V. Kameneva, K.H. Ahn and S.J. Lee, Strain hardening of red blood cells by accumulated cyclical supraphysiological stress, Artif. Org. 31 (2007), 80-86. 112. P.J. Marascalco, S.P. Ritchie, T.A. Snyder and M.V. Kameneva, Development of standard tests to examine viscoelastic properties of blood of experimental animals for pediatric mechanical support device evaluation, ASAIO J. 52 (2006), 567–574. 113. O.K. Baskurt, M. Uyuklu and H.J. Meiselman, Protection of erythrocytes from sub-hemolytic mechanical damage by nitric oxide mediated inhibition of potassium leakage, Biorheology 41 (2004), 79-89. 114. S.S. Lee, K.H. Ahn, S.J. Lee, K. Sun, P.T. Goedhart and M.R.Hardeman, Shear induced damage of red blood cells monitored by the decrease of their deformability, Korea-Aust Rheol. J. 16 (2004), 141–146. 115. T. Mizuno, T. Tsukiya, Y. Taenaka, E. Tatsumi, T. Nishinaka, H. Ohnishi, M. Oshikawa, K. Sato, K. Shioya, Y. Takewa and H. Takano, Ultrastructural alterations in red blood cell membranes exposed to shear stress, ASAIO J. 48 (2002), 668-670. 116. M.V. Kameneva, J.F. Antaki, H.S. Borovetz, B.P. Griffith, K.C. Butler, K.K. Yeleswarapu, M.J. Watach and R.L. Kormos, Mechanisms of red blood cell trauma in assisted circulation. Rheologic similarities of red blood cell transformations due to natural aging and mechanical stress, ASAIO J. 41 (1995), M457-M460. 117. K.M. Dao, E.A. O'Rear, A.E. Johnson and S.E. Peitersen, Sensitivity of the erythrocyte membrane bilayer to subhemolytic mechanical trauma as detected by fluorescence anisotropy, Biorheology 31 (1994), 69-76. 118. R. Fåhraeus and T. Lindquist, The viscosity of the blood in narrow capillary tubes, Am. J. Physiol. 96 (1931), 562-568. 119. A.S. Popel, P.C. Johnson, M.V. Kameneva and M.A. Wild, Capacity for red cell aggregation higher in athletic mammalian species than in sedentary species, J. Appl. Physiol. 77(1994), 1790-1794. 120. H. J. Meiselman, Red blood cell role in RBC aggregation: 1962–1993 and beyond, Clin. Hemorheol. 13 (1993), 575–592. 121. R.B. Ami, G. Barshtein, D. Zeltser, Y. Goldberg, I. Shapira, A. Roth, G. Keren, H. Miller V. Prochorov, A. Eldor, S. Berliner and S. Yedgar, Parameters of red blood cell aggregation as correlates of the inflammatory state, Am. J. Physiol. 280 (2001), H1982-H1988. 122. A.M. Morariu, M.H. Maathuis, S.A. Asgeirsdottir, H.G. Leuvenink, P.W. Boonstra, W. van Oeveren, R.J. Ploeg, I. Molema and G. Rakhorst, Acute isovolemic hemodilution triggers proinflammatory and procoagulatory endothelial activation in vital organs: role of erythrocyte aggregation, Microcirculation 13 (2006), 397-409. 123. A.M. Morariu, Y.J. Gu, R.C. Huet, W.A. Siemons, G. Rakhorst and W.V. Oeveren, Red blood cell aggregation during cardiopulmonary bypass: a pathogenic cofactor in endothelial cell activation? Eur. J. Cardiothorac. Surg. 26 (2004), 939-946. 124. Y.J. Gu, R. Graaff, E. de Hoog, N.J.G.M. Veeger, G. Panday, P.W. Boonstra and W. van Oeveren, Influence of hemodilution of plasma proteins on erythrocyte aggregability: an in vivo study in patients undergoing cardiopulmonary bypass. Clin. Hemorheol. Microcirc. 33 (2005), 95-107.

226


125. A. Undar, Effect of hypothermic cardiopulmonary bypass on blood viscoelasticity in pediatric cardiac patients, ASAIO J. 51 (2005), 522-524. 126. R. Koppensteiner, K. Derfler and H. Ehringer, Blood rheology after renal transplantation, Nephron 74 (1996), 328-332. 127. L. Houbouyan, J.F. Stoltz, A. Beauchet, N. Consoli, J. Prinseau, A. Baglin and A. Goguel, Modifications of hemorheologic parameters in the course of hemodialysis in chronic renal insufficiency, J. Mal. Vasc. 19 (1994), 132-136. 128. A.L. Hadengue, M. Del-Pino, A. Simon and J. Levenson, Erythrocyte disaggregation shear stress, sialic acid, and cell aging in humans, Hypertension 32 (1998), 324-330. 129. M.V. Kameneva, M.J. Watach, P. Litwak, J.F. Antaki, K.C. Butler, D.C. Thomas, L.P. Taylor, H.S. Borovetz, R.L. Kormos and B.P. Griffith, Chronic animal health assessment during axial ventricular assistance: importance of hemorheologic parameters, ASAIO J. 45 (1999), 183-188. 130. R.G. Cooper, R.A. Kahn, C.N. Cornell and M.E. Muhrer, Erythrocyte mechanical fragility test, J. Clin. Pathol. 21 (1968), 781-783. 131. L. Gu, W.A. Smith and G.P. Chatzimavroudis, Mechanical fragility calibration of red blood cells, ASAIO J. 51 (2005), 194-201. 132. M.V. Kameneva, J.F. Antaki, K.K. Yeleswarapu, M.J. Watach, B.P. Griffith and HS. Borovetz, Plasma protective effect on red blood cells exposed to mechanical stress, ASAIO J. 43 (1997), M571-M575. 133. M.V. Kameneva, K.O. Garrett, M.J. Watach, H.S and Borovetz, Red blood cell aging and risk of cardiovascular diseases, Clin. Hemorheol. Microcirc. 8(1998), 67-74. 134. E. Hegedus, V. Harsanyi and S.R. Hollan, The effect of increasing glucose in CPD on the quality of stored blood, Folia Haematol. 107 (1980), 928-933. 135. K. J. Halbhuber, H. Feuerstein, D. Stibenz and W.Linss, Membrane alteration during banking of red blood cells, Biomed. Biochim. Acta 42 (1983), S337-S341. 136. T. Butler, C.A. Bradley and J.E. Owensby, Plasma components protect erythrocytes against experimental haemolysis caused by mechanical trauma and by hypotonicity, Int. J. Exp. Pathol. 73 (1992), 27-33. 137. R. Sumpelmann, T. Schurholz, G. Marx and R. Zander, Protective effects of plasma replacement fluids on erythrocytes exposed to mechanical stress, Anaesthesia 55 (2000), 976-979. 138. T. Jikuya, T. Tsutsui, O. Shigeta, Y. Sankai and T. Mitsui, Species differences in erythrocyte mechanical fragility: comparison of human, bovine, and ovine cells, ASAIO J. 44 (1998), M452-M455. 139. S. Klaus, S. Korfer, K. Mottaghy, H. Reul and B. Glasmacher, In vitro blood damage by high shear flow: human versus porcine blood, Int. J. Artif. Org. 25 (2002), 306-312. 140. T. Bohler, A. Leo, A. Stadler and O. Linderkamp, Mechanical fragility of erythrocyte membrane in neonates and adults, Pediatric Res. 32 (1992), 92-96. 141. Z.Y. Wen, L.C. Song, Z.Y. Yan, Z.H. Lu, D.G. Sun, Y. Shi and S. Chien, An animal model to study erythrocyte senescence with a narrow time window of erythrocyte production: alterations in osmotic fragility and deformability of erythrocytes during their life span, Clin. Hemorheol. Microcirc. 19 (1998), 299-306. 142. T. Kamada, D.E. McMillan, J.J. Sternlieb, V.O. Bjork and S. Otsuji, Erythrocyte crenation induced by free fatty acids in patients undergoing extracorporeal circulation, Lancet 2 (1987), 818-821. 143. T. Kamada, D.E. McMillan, J.J. Sternlieb, V.O. Bjork and S. Otsuji, Albumin prevents erythrocyte crenation in patients undergoing extracorporeal circulation, Scand. J. Thorac. Cardiovasc. Surg. 22 (1988), 155-158. 144. J.K. Armstrong, H.J. Meiselman and T.C. Fisher, Inhibition of red blood cell-induced platelet aggregation in whole blood by a nonionic surfactant, poloxamer 188 (RheothRx injection), Thromb. Res. 79 (1995), 437-450. 145. D. Roberts, B. Bake and G. William-Olsson, Improved red blood cell survival after cardiac operations with administration of urea during cardiopulmonary bypass, J. Thorac. Cardiov. Surg. 89 (1985), 107114. 146. D. Roberts, L. Dernevik, T. Hirayama, H. Yamaguchi, M. Allers and G. William-Olsson, Reduced perand postoperative mortality following the use of urea during elective cardiopulmonary bypass. A proposed treatment for the prevention of reduced red cell deformability during open heart surgery, J. Cardiovasc. Surg. 28 (1987), 75-80. 147. S. Nikolaos, A. George, T. Telemachos, S. Maria, M. Yannis and M. Konstantinos, Effect of Lcarnitine supplementation on red blood cells deformability in hemodialysis patients, Ren. Fail. 22 (2000), 73-80. 148. A. Arduini, M. Rossi, G. Mancinelli, M. Belfiglio, R. Scurti, G. Radatti and S.B. Shohet, Effect of Lcarnitine and acetyl-L-carnitine on the human erythrocyte membrane stability and deformability, Life Sci. 47 (1990), 2395-2400.


227

149. B. Toptas, A. Baykal, A. Yesilipek, M. Isbir, A. Kupesiz, O. Yalcin and O.K. Baskurt, L-carnitine deficiency and red blood cell mechanical impairment in beta-thalassemia major, Clin. Hemorheol. Microcirc. 35 (2006), 349-357.

228


Hemorheological Considerations in Stored Blood Transfusion James P. ISBISTER1 Department of Haematology and Transfusion Medicine, Royal North Shore Hospital of Sydney, Faculty of Medicine, University of Sydney, St Leonards, New South Wales. Australia 2065

Introduction In most of the world, the term “blood transfusion” usually means the infusion of stored red blood cell (RBC) concentrates, often termed “packed cells”, into the circulatory system. Such transfusions are given in response to severe anemia, significant blood loss, or as therapy (e.g., sickle cell disease). While intended to be beneficial, infusion of stored RBC can impact blood rheology in both large vessels and the microcirculation. In part, the rheological effects are directly linked to altered hematocrit levels, but additional factors related to storage are also involved. Thus, understanding the practice and problems of transfusion medicine is a necessary prerequisite for exploring hemorheological changes that occur subsequent to the transfusion of stored cells.

1. Background Stored red cells have changes in their shape, deformability and hemoglobin function that may take several hours to return to normal following transfusion. Blood storage lesions impacting on the efficacy and safety of transfusion are most relevant in relationship to the labile blood components, especially the cellular components stored in the liquid state [1-5]. Red cell concentrates are the most widely used blood component in this respect and will be the focus of this chapter. Platelet concentrates, with a much shorter storage time and the requirement that they be stored at 22-24oC, are a constant challenge for blood supply agencies and present additional problems, especially in relationship to bacterial contamination; platelet concentrates will not be addressed in this chapter. The clinically significant hazards of blood component therapy have generally been classified as immunological or infectious in nature [6]. However, a much broader view of the safety and efficacy of blood component therapy is now being taken. The

1 Corresponding author: Department of Haematology and Transfusion Medicine, Royal North Shore Hospital of Sydney, Faculty of Medicine, University of Sydney, St Leonards, New South Wales. Australia; E mail: [email protected]

J.P. Isbister / Hemorheological Considerations in Stored Blood Transfusion

229

pathophysiology of blood transfusion reactions can broadly be divided into three categories [7]: 1. Reactions due to immunological differences between the donor and recipient resulting in varying degrees of blood component incompatibility. In general, in order for a reaction to occur the recipient needs to have been previously immunized to a cellular or plasma antigen. 2. A wide range of infectious agents that may be transmitted by homologous blood component therapy. 3. Alterations in blood products due to preservation and storage resulting in quantitative and/or qualitative deficiencies in the blood components which will reduce transfusion efficacy and expose the patient to potentially adverse consequences from storage-associated changes in the component . In terms of causation of an adverse clinical event the possible role of blood transfusion can be classified into three categories on the basis of probability: Definite – unifactorial – transfusion caused The well understood and reported hazards of transfusion (i.e., immunological, technical, infectious) are generally unifactorial with a 1:1 causal relationship between the blood component transfused, usually a specific individual unit, and the adverse consequence for the patient. ABO blood group incompatibility, transfusion related infection transmission, transfusion associated graft versus host disease (TAGVHD) and others are in this category. Probable – oligofactorial – transfusion initiated Some adverse consequences of transfusion result from interaction with other insults, pathophysiology or host factors, but the contribution of the transfusion can usually be specifically identified. Fever, allergic reactions, hypotensive reactions, pulmonary oedema, some cases of transfusion related acute lung injury (TRALI), hyper-bilirubinemia and CMV transmission are examples of this category. Possible – Multifactorial- transfusion associated Transfusion may be a contributor to a complication or poor clinical outcome. In these circumstances it is usually difficult to implicate transfusion directly in an individual case, or is it necessarily the major factor. Transfusion-induced immunomodulation (TRIM) and the clinical consequences of storage lesions fall into this category. The role of transfusion being associated with such adverse consequences has been well established in observational studies, but causation can only be confidently identified by large, well-conducted randomized controlled trials. Accumulating evidence, in the absence of such trials, supports causation thus demanding a more precautionary approach to allogeneic blood transfusion, examination of alternatives to transfusion and implementation of methods to improve the quality of transfusion blood components, especially red cell and platelet concentrates. Universal prestorage leucodepletion is currently the most important and effective strategy for minimizing the clinical impacts or TRIM and the blood storage lesions In categories 1 and 2 above, disturbances in microcirculatory function may occur due to changes in the patient’s blood, transfused blood or the microvasculature.

230


Depending on the fractionation methods used there may be various amounts of “debris” in stored blood (i.e., microaggregates) that have to be cleared by the pulmonary microcirculation and the reticuloendothelial system. Fortunately, in recent years the use of fractionation, buffy coat removal and leucoreduction filters have tended to minimize the problem. However, if blood is stored as un-fractionated whole blood, as it commonly is with preoperative autologous deposits, microaggregates up to 200 μm progressively form with increasing storage time. Ironically, the standard 170 μm filters introduced in the early days of blood transfusion were based on the finding that if large microaggregates were not removed there were difficulties with blockages during infusion. Microaggregate filters have been used to remove microaggregates down to 20 μm, especially when large volumes of blood were being rapidly transfused, but with the current red cell preparation methods used in most developed countries these filters are generally not used. Most research focusing on the preparation, preservation and storage of blood components has been in vitro, with analysis carried out on biochemical parameters (e.g., ATP, pH, 2,3 DPG) and occasionally membrane, morphological and rheological characteristics [8-13]. In vivo studies have generally focused on intravascular survival of transfused blood components and, to a lesser extent, post-transfusion hemorheology, function and clinical efficacy. It is only in recent years that more detailed attention is being given to assessing potential adverse clinical consequences of storage changes in the clinical use of blood components. Although concern has been expressed by clinicians for decades about the potential for the storage lesion to have adverse effects on patients, it has been difficult to prove in prospective clinical studies due to the multifactorial nature and complexity of many transfusion settings, especially hemorrhagic shock and trauma. In vitro data and animal studies have long raised concerns that the changes resulting from storage of blood components may have adverse clinical consequences. However, blood supply agencies and hospital transfusion services have been preoccupied with the provision of compatible and infection-free blood components. Although constant vigilance towards these safety issues is important, the immunological and infection transmission hazards of transfusion have largely been understood and effective risk management strategies implemented. More attention is now being directed towards questions of appropriateness of transfusion, the quality of blood products and the clinical consequences of transfusion-related immunomodulation (TRIM), blood storage lesions, bacterial contamination and transfusion related acute lung injury (TRALI) [6, 14-22]. In recent years there has been a reassessment of the management of the acutely hemorrhaging patient. Advances in patient retrieval, resuscitation protocols, techniques for rapid and real time diagnosis, trauma teams and early “damage control” surgery have all improved the management of acutely hemorrhaging patients [23]. Patients are now surviving with larger volumes of blood transfusion, but sepsis, acute lung injury and multi-organ failure remain major challenges and blood transfusion is increasingly being recognized as a two-edged sword and probably a contributory factor to these complications in which microcirculatory dysfunction is recognized as central in the pathophysiology [24]. Observational studies have identified blood transfusion as an independent risk factor for morbidity and mortality [25-29]. Correlation does not mean causation, but on critical assessment of the evidence, it is likely to be the case. This has lead to a re-analysis of guidelines for the management of acutely bleeding patients. Clinical practice guidelines should no longer primarily address the management of


231

massive blood transfusion, but rather the management of critical bleeding and quality/efficacy of blood components. The modus operandi is now pre-emptive, instituting measures to avoid getting into the massive transfusion and coagulopathy quagmire in which the patient spirals down into the “triad of death”: coagulopathy, acidosis and hypothermia. Stored red cells are pro-inflammatory and procoagulant [3034]. Recent studies have identified that there are questions about the role of red cell transfusions in many clinical settings and dogmas about the appropriateness of red cell transfusion are being questioned. This evidence includes: x Transfusion of red cell concentrates in the critical care setting to achieve a hemoglobin level >80 g/l does not improve outcome and may indeed be detrimental [35, 36]. x It has been difficult to confirm the benefit of red cell transfusions in the peri-operative period and there is no agreement as to the hemoglobin level at which outcome is improved by transfusion [36]. x In years past the mortality of trauma patients requiring massive transfusions of >20 units of blood declined dramatically [24]. Almost 50% of trauma patients requiring >50 units of blood are surviving, but there is disturbing evidence, on the basis of multivariate analysis, that the transfusions are an independent risk factor for poor survival, increased infection rates, prolongation of length of hospital stay and development of the multi-organ failure syndrome and adult respiratory distress syndrome. x The transfusion of stored red cells in the bleeding setting are effective in restoring the macrocirculation, but not the microcirculation and may be detrimental to microcirculatory function [37].

2. Blood Storage Lesions Blood is altered from the moment of collection. Anticoagulation, separation, cooling, preservation and storage compound and progressively increase storage changes until the date of expiry [11, 38]. The extent of these changes is determined by the collection technique, the specific blood component, the preservative medium, the container, storage time and storage conditions. The threshold storage time for blood components has generally been arbitrarily determined by in vitro studies and assessment of in vivo survival. In the case of red cell concentrates, greater than 70% of transfused cells should survive at 24 hours post-transfusion. Storage may result in quantitative and/or qualitative deficiencies in blood components which can reduce the efficacy of a transfusion. In parallel with these storage changes is an accumulation of degenerate material (e.g., microaggregates and procoagulant material), release of vasoactive agents, cytokine generation, and hemolysis (Figure 1) [32, 39, 40]. Many of the changes occurring during storage are related to the presence of leucocytes and can be minimized by pre-storage leucoreduction [32]. With storage red cells undergo a change from their biconcave disc shape to cells with spiky projections termed echinocytes, eventually becoming spiky spheres known as sphereo-echinocytes; these changes from the biconcave shape reduce cellular deformability. There are also changes in the red cell membrane resulting an in

232


increased tendency to adhere to endothelial cell surfaces in the microcirculation, especially if there is any activation of endothelial cells such as seen in the presence of the systemic inflammatory response (e.g., with shock or sepsis) [41]. There is thus good evidence that the immediate post-transfusion function of stored red cells and hemoglobin in delivering and unloading oxygen to the microcirculation is questionable and several hours are required for red cell oxygen carriage and delivery return to normal [42, 43].

Figure 1: Blood storage lesion, illustrating the range of alterations and the possible clinical consequences.

If you wish to start a heated debate, mention the use of fresh blood to a blood banker: it is an anathema to the philosophy of blood banking. The focused clinician, confronted with a critically bleeding patient, may state: "My patient bleeds fresh whole blood, therefore, he needs fresh whole blood." At the other end of the spectrum are the blood bankers, distanced from the clinical action, promoting the concept that the mix of stored red cells, platelet concentrates, factor concentrates and fresh frozen plasma is equivalent to fresh whole blood. If stored whole blood was submitted to the FDA in the United States for registration as a therapeutic agent, one suspects that the FDA would insist that all clinical studies should use fresh blood to demonstrate its efficacy and safety; following such a demonstration, consideration for extending the shelf-life could be submitted. The fact that the debate regarding the advantages of fresh whole blood continues highlights the fact that we do not have a definitive answer. When one is unable to prove, in a truly scientific manner based on clinical trials, that a form of therapy is efficacious or not, one has to fall back on a sound logical use of the currently available scientific understanding.


233

Figure 2: Consequences of storage on red cell membrane and haemoglobin functions

2.1. Clinical Significance of Storage Lesions of Red Cell Concentrates The clinical significance of blood storage lesions continues to be debated; in some cases the effects are widely accepted, while in others further studies are needed. Transfusion may be an additional risk factor for such conditions as adult respiratory distress syndrome (ARDS), multiorgan failure (MOF), vasoactive reactions and alterations in laboratory parameters (Table 1 and Figure 1) [42, 44-48]. In reviewing the potential adverse clinical effects of storage changes, the assumption is made that the blood component has been appropriately, collected, processed, stored, transported and transfused. Although there is now much greater attention to standard operating procedures and regulations applied to these aspects of the blood component supply chain, quality of the final product cannot be guaranteed [49]. The “assumed” quality of labile cellular blood products is based on research data and monitoring of standard operative procedures. There is rarely detailed individual product assessment prior to transfusion [43]. It is, however, generally accepted that the adverse effects of storage increase with time and an arbitrary “cut off” is mandated on the basis of research studies. In relation to the possible clinical significance of the storage lesion, the following should be considered: Quantitative and qualitative deficiency of blood component x Failure to achieve anticipated end points due to reduced quantity and/or quality of the blood product x Exposure to excessive numbers of donors in achieving efficacy Physical characteristics x Hypothermia Chemical characteristics x Citrate toxicity x Acid-base imbalance x Glucose

234


Contamination x Bacterial resulting in endotoxaemia or septicaemia x Plasticisers Accumulation of “toxic” or degenerate products x Role of the storage lesion in transfusion related immunomodulation o Role of cytokines o Role of reticuloendothelial system blockade x Accentuation of free radical pathophysiology due to free iron x Effects of transfusion on laboratory parameters (e.g., elevations in bilirubin, neutrophils, serum iron and lactic dehydrogenase) which may lead to incorrect interpretation x Large volume transfusions of longer-term stored components as a risk factor for multi-organ failure (MOF) and adult respiratory distress syndrome (ARDS) x Early hyperkalaemia, late hypokalaemia and hypocalcaemia x Activation and consumption of the haemostatic factors with possible contribution to disseminated intravascular coagulation and venous thromboembolism x Non haemolytic, non-febrile transfusion reactions x Hypotension and circulatory instability due to vasoactive substance (kinins, histamine) Table 1: Blood storage lesions, the clinical consequences and the role of pre-storage leucoreduction in their minimisation (TRALI: Transfusion Related Acute Lung Injury, MOF: Multiorgan failure, VTE: Venous Thromboembolism, TRIM: Transfusion Related Immunomodulation, DIC: Disseminated Intravascular Coagulation)

2.2. The Current Focus on Allogenic Transfusion-Related Poor Clinical Outcome Where There is Correlation Versus Causation: The Example of Acute Lung Injury As alluded to above, over the last decade experimental and clinical studies have identified blood transfusion per se as a possible independent risk factor for morbidity and mortality as well as for increased admission rates to intensive care units, increased length of hospital stay and additional costs [6, 29, 50].


235

In particular, TRALI is receiving attention as a potentially serious complication of blood transfusion [15, 16, 51]. In the classical plasma-neutrophil-antibody mediated form of the disease, symptoms usually arise within hours of a blood transfusion [52, 53]. The underlying pathophysiology of “classical” TRALI is due to the presence of leucoagglutinins in donor plasma. When complement is activated, C5a promotes neutrophil aggregation and sequestration in the microcirculation of the lung causing endothelial damage leading to interstitial edema, and acute respiratory failure. This classical form of transfusion-related acute lung injury has been recognized for over five decades, but was thought to be a rare complication of allogeneic plasma transfusion. It is now accepted that there has been under-recognition and under-diagnosis of TRALI, partly due to a lack of clinical awareness, but also a lack of a broad understanding of mechanisms by which blood transfusion may cause or be a contributory factor to lung injury [54-56]. The term TRALI is now being expanded to include cases in which transfusion is identified as an independent risk factor predisposing patients to lung injury. The reader is referred to recent reviews addressing this expanding area of concern in which transfusion per se is being identified as an independent risk factor for poor clinical poor outcomes, of which TRALI is only one [51]. Blood transfusion being implicated as part of the problem rather than optimal therapy has been a surprise to many clinicians, as it has always been assumed that blood transfusion can only be of benefit to the bleeding or anemic patient. These concerns have led to a major reanalysis of guidelines for the management of acutely bleeding and anemic patients [57-59]. This reappraisal is resulting in challenging of long-standing dogmas [60]. There is greater tolerance of hypotension until hemorrhage is controlled and lower hemoglobin levels are tolerated with closer attention to the clinical context of the anemia and its impact on systemic and local oxygen delivery, especially if there are compromises in cardiorespiratory function. The evidence that the immediate post-transfusion ability of stored red cells and hemoglobin to deliver and unload oxygen to microcirculation may be impaired is resulting in stored red cell concentrates not being automatically the first therapy for acutely enhancing oxygen delivery. More attention and research is being directed toward the role of clear fluids and the importance of plasma viscosity, colloid oncotic pressure and functional capillary density [61-67]. More recently the focus has moved towards targeted oxygen therapy with high inspired oxygen levels, hyperbaric oxygen and a re-evaluation of hemoglobin based oxygen carriers [68-73]. There are thus many questions being asked about the safety and efficacy of allogenic blood transfusion and the possibility/probability that transfusion may not only be part of the solution but be partly responsible for poorer clinical outcomes. 2.3. Prevention and/or Minimization of the Storage Lesion and its Clinical Consequences With the accumulating evidence that blood storage lesions are clinically significant it is important to differentiate between storage lesions being responsible for failure to achieve clinical/laboratory endpoints as a result of reduced survival and/or qualitative defects in cellular function, and the “toxic” effects of storage. Clearly, avoiding or minimizing the use of allogenic blood transfusions by appropriate clinical decision making and use of alternatives is an obvious approach to reducing the clinical impact of the storage lesion [74-76].

236


There are several approaches that can be taken to prevent or minimize the red cell storage lesion and its clinical consequences. Blood supply agencies in developed countries are generally externally regulated and have standard operating procedures using long established and tested methods for the collection and preservation of labile blood components. There are continuing efforts to improve blood preservation with the aim of supplying better quality products. At present however, assuming integrity of the current processes, pre-storage leucoreduction to <1 x 106/l is probably the most effective intervention to improve the survival and quality of red cell and platelet concentrates [77-84]. Cellular blood products contain variable numbers of donor leucocytes, and except in highly specific instances the transfusion of donor leucocytes is an unintended, undesirable and potentially hazardous consequence of blood component therapy. Leucoreduction is advocated to reduce the incidence and/or delay the onset of alloimmunization to leucocytes with the aim of preventing non-hemolytic febrile transfusion reactions, minimizing the development of refractoriness to platelet transfusions, and improving access to potential donors for tissue transplantation. Leucocytes also act as the principal reservoirs and/or transport vehicles for a range of cell associated viral, bacterial and protozoal pathogens. Evidence also supports a protective role for leucocyte-depletion against a range of transfusion transmitted infections especially CMV. 2.4. Prioritizing Patients for Fresher Blood Products Inventory management is always a challenge for blood supply agencies and hospital transfusion services. In determining priority for fresher red cell concentrates the volume, rate, frequency of infusion and clinical condition of the patient need to be considered. The following is a personal preference as to how patients can be prioritized into categories when considering the clinical significance of red cell storage lesion. 1. High priority a. Neonates, especially exchange transfusion b. Massive and rapid blood transfusion >5 units c. Cardiac bypass patients d. Critically ill and/or unstable patients with cardiorespiratory failure, MOF, ARDS 2. Medium priority a. Chronic anemia b. Neurosurgical patients c. Patients with vascular disease (coronary, cerebral, peripheral) d. Stable ICU patients 3. Low priority a. Low volume “one off” transfusions (e.g., perioperative, GIT hemorrhage)

3. Conclusions There is now evidence that the blood storage lesion is clinically significant. It is important to differentiate between the storage lesion that is responsible for failure to achieve clinical/laboratory endpoints: reduced survival, qualitative defects in cellular function, and the “toxic” effects of storage. Taking a precautionary approach, it is


237

reasonable to conclude that in the case of red cell concentrates, significant changes occur in non leucoreduced products after 7-10 days. As leucocytes are largely responsible for the clinically significant storage changes, the wider use of pre-storage leucoreduction should minimize the problem. Buffy coat removal is an inadequate method of leucoreduction and pre-storage filtration is necessary. Blood filters have always been a subject of debate in transfusion medicine. The use of blood filters, whatever type, is an acknowledgement of the existence of the blood storage lesion and its probable clinical significance. The 170 μm filters in standard blood sets were originally introduced into transfusion medicine to prevent blood sets obstructing, not because of a concern that the fibrin clots may harm the patient. Fortunately the human lung is one of Nature's remarkable filters and receives all intravenous infusions on their “first pass”. One presumes it was concluded, by trial and error, that 170 μmҏ filters seemed to be the best compromise between using a larger needle and keeping the blood transfusion flowing. In the 1960’s and 1970’s interest arose in adult respiratory distress syndrome, mainly stimulated by experiences during the Vietnam War. Logic and animal data suggested that the unfiltered microaggregates accumulating during storage may be a contributor to the development of ARDS, and that microfilters to remove microparticles 20-40 μm in size may be protective [85-87]. Unfortunately, it was difficult to prove this in clinical trials for several reasons, especially the multifactorial nature of the clinical problem of massive blood transfusion and the less-sophisticated statistical methods than are currently available. Nevertheless, microaggregate filters are inadequate to significantly address the problem of the storage lesion and its clinical significance. Preventing the development of the storage lesion from its inception, using prestorage leucodepletion filters, is a more logical and scientific approach to the problem. The storage lesions of the commonly used labile blood components (i.e., red cell and platelet concentrates) and the clinical consequences, especially in the critical bleeding setting, are now receiving appropriate attention. Universal pre-storage leucoreduction is now standard practice in many countries. Ironically, its introduction was for other reasons than minimizing the storage lesions, preventing leucocyte immunization and immunomodulation and not on sound scientific evidence. Rather, they were introduced as a precautionary measure against the possible transfusion transmission of vCJD that subsequently turned out to be justified [88, 89]. The efficacy of red cell transfusions has been assumed and risks regarded as minimal. Allogenic blood for transfusion in western countries is now extremely safe from the viral transmission perspective, but there is now increasing evidence and realization that red cell transfusions may be a risk factor for poor clinical outcomes. There is an ongoing reassessment of the evidence base for many transfusion practices, especially in relationship to the use of red cell concentrates and fresh frozen plasma. With this move from a transfusion-focused to a patient-focused blood management approach, there is improved understanding of the benefits and risks of transfusion and an awareness of techniques for minimizing or avoiding allogenic transfusions. The global costs associated with blood component therapy and adequate supply are promoting greater interest in transfusion alternatives, recombinant products and blood substitutes. In particular, the concept of targeted oxygen therapy is gaining ground; this is an area of research in which hemoglobin-based oxygen carriers and hemoglobin substitutes, with their long awaited promise, may have a role. Advances in our understanding of the microcirculation and hemorheology are giving this area of research a much sounder scientific base.

238


References [1]

[2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]

[25] [26] [27] [28]

[29]

G.H. Longster, T. Buckley, J. Sikorski and L.A. Derrick-Tovey, Scanning electron microscope studies of red cell morphology. Changes occurring in red cell shape during storage and post transfusion, Vox. Sang, 22 (1972), 161-170. B.G. Solheim, O. Flesland, J. Seghatchiam and F. Brosstad, Clinical implications of red blood cell and platelet storage lesions: an overview. Transfus. Apheresis Sci., 31 (2004), 185-189. C.F. Hogman and H.T. Meryman, Storage parameters affecting red blood cell survival and function after transfusion, Transfus. Med. Rev. 13 (1999), 275-296. C.F. Hogman, Liquid-stored red blood cells for transfusion. A status report. Vox. Sang, 76 (1999), 6777. C.F. Hogman, Storage of blood components, Curr. Opin. Hematol. 6 (1999), 427-431. A. Shander, Emerging risks and outcomes of blood transfusion in surgery, Semin. Hematol. 41 (2004), 117-124. J.P. Isbister, Hazards of homologous blood transfusion. Transfusion Alternative in Transfusion Medicine 4 (2002), 129-137. S.L. Schrier, B. Hardy, K. Bensch, I. Junga and J. Krueger, Red blood cell membrane storage lesion, Transfusion 19 (1979), 158-165. J.P. McCue and J.M. Vincent, Changes in red blood cell membrane phosphate concentration during blood bank storage, Transfusion 21 (1981), 107-112. S.L. Schrier, P.R. Sohmer, G.L. Moore, L. Ma and I. Junga, Red blood cell membrane abnormalities during storage: correlation with in vivo survival, Transfusion 22 (1982), 261-265. K.L. Scott, J. Lecak and J.P. Acker, Biopreservation of red blood cells: past, present, and future, Transfus. Med. Rev. 19 (2005), 127-142. R.B. Dawson, Preservation of red blood cells for transfusion, Hum. Pathol. 14 (1983), 213-7. J.R. Hess, An update on solutions for red cell storage, Vox. Sang. 91 (2006), 13-19. J.P. Wallis, Transfusion-related acute lung injury (TRALI)- under-diagnosed and under-reported, Br. J. Anaesth. 90 (2003), 573-576. K. Swanson, D.M. Dwyre, J. Krochmal and T.J. Raife, Transfusion-related acute lung injury (TRALI): current clinical and pathophysiologic considerations, Lung 184 (2006), 177-185. M.A. Popovsky, Pulmonary consequences of transfusion: TRALI and TACO, Transfus. Apher. Sci. 34 (2006), 243-244. S.B. Moore, Transfusion-related acute lung injury (TRALI): clinical presentation, treatment, and prognosis, Crit. Care Med. 34 (2006), S114-S117. M.A. Blajchman, Immunomodulation and blood transfusion, Am. J. Ther. 9 (2002), 389-395. M. Raghavan and P.E. Marik, Anemia, allogenic blood transfusion, and immunomodulation in the critically ill, Chest 127 (2005), 295-307. M.S. Mincheff and H.T. Meryman, Blood transfusion, blood storage and immunomodulation, Immunol. Invest. 24 (1995), 303-309. J.W. Pattison, Ensuring appropriate use of blood transfusion in anaemia, Nurs. Times 101 (2005), 3033. A. Shander and L.T. Goodnough, Objectives and limitations of bloodless medical care, Curr. Opin. Hematol. 13 (2006), 462-470. D. Bose and N.C. Tejwani, Evolving trends in the care of polytrauma patients, Injury 37 (2006), 20-28. P. Hakala, S. Hiippala, M. Syrjala and T. Randell, Massive blood transfusion exceeding 50 units of plasma poor red cells or whole blood: the survival rate and the occurrence of leukopenia and acidosis, Injury 30 (1999), 619-622. A.B. Nathens, Massive transfusion as a risk factor for acute lung injury: association or causation? Crit. Care Med. 34 (2006), S144-S150. F.A. Moore, E.E. Moore and A. Sauaia, Blood transfusion. An independent risk factor for postinjury multiple organ failure, Arch. Surg. 132 (1997), 624-625. A. Charles, A.A. Shaikh, M. Walters, S. Huehl and R. Pomerantz, Blood transfusion is an independent predictor of mortality after blunt trauma, Am. Surg. 73 (2007), 1-5. W.P. 3rd. Robinson, J. Ahn, A. Staffler, E.J. Rutherford, H. Hurd, B.L. Zarzaur, C.C. Baker, A.A. Meyer and P.B. Rich, Blood transfusion is an independent predictor of increased mortality in nonoperatively managed blunt hepatic and splenic injuries. J. Trauma -Inj. Inf. Crit. Care 58 (2005), 437-445. D.L. Malone, J. Dunne, J.K. Tracy, A.T. Putnam, T.M. Scalea and L.M. Napolitano, Blood transfusion, independent of shock severity, is associated with worse outcome in trauma, J. Trauma 54 (2003), 898905.


239

[30] K.M. Twomley, S.V. Rao and R.C. Becker, Proinflammatory, immunomodulating, and prothrombotic properties of anemia and red blood cell transfusions, J. Thromb. Thrombolysis 21 (2006), 167-174. [31] G.J. Despotis, L. Zhang and D.M. Lublin, Transfusion Risks and Transfusion-related Pro-inflammatory Responses, Hematol. Oncol. Clin. North Am. 21 (2007), 147-161. [32] R.L. Sparrow and K.A. Patton, Supernatant from stored red blood cell primes inflammatory cells: influence of prestorage white cell reduction, Transfusion 44 (2004), 722-730. [33] J.R. Dunne, D.L. Malone, J.K. Tracy and L.M. Napolitano, Allogenic blood transfusion in the first 24 hours after trauma is associated with increased systemic inflammatory response syndrome (sirs) and death, Surg. Infect. (Larchmt) 5 (2004), 395-404. [34] M.H. Friedlander, R. Simon and G.W. Machiedo, The relationship of packed cell transfusion to red blood cell deformability in systemic inflammatory response syndrome patients, Shock 9 (1998), 84-88. [35] P.C. Hebert, G. Wells, M.A. Blajchman, J. Marshall, C. Martin, G. Pagliarello, M. Tweeddale, I. Schweitzer and E. Yetisir, A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion requirements in critical care investigators, Canadian Critical Care Trials Group, N. Engl. J. Med. 340 (1999), 409-417. [36] J.L. Carson, S. Hill, P. Carless, P. Hebert and D. Henry, Transfusion triggers: a systematic review of the literature, Transfus. Med. Rev. 16 (2002), 187-199. [37] A.G. Tsai, P. Cabrales and M. Intaglietta, Microvascular perfusion upon exchange transfusion with stored red blood cells in normovolemic anemic conditions, Transfusion 44 (2004), 1626-1634. [38] C.R. Valeri and G. Ragno, Cryopreservation of human blood products, Transfus. Apher. Sci. 34 (2006), 271-287. [39] R.L. Sparrow, G. Healey, K.A. Patton and M.F. Veale, Red blood cell age determines the impact of storage and leukocyte burden on cell adhesion molecules, glycophorin A and the release of annexin V, Transfus. Apher. Sci. 34 (2006), 15-23. [40] A.M. Anniss, K.M. Glenister, J.J. Killian and R.L. Sparrow, Proteomic analysis of supernatants of stored red blood cell products, Transfusion 45 (2005), 1426-1433. [41] A.M. Anniss and R.L. Sparrow, Storage duration and white blood cell content of red blood cell (RBC) products increases adhesion of stored RBCs to endothelium under flow conditions, Transfusion 46 (2006), 1561-1567. [42] A. Tinmouth, D. Fergusson, I.C. Yee and P.C. Hebert, Clinical consequences of red cell storage in the critically ill, Transfusion 46 (2006), 2014-2027. [43] M.D. Elfath, Is it time to focus on preserving the functionality of red blood cells during storage? Transfusion 46 (2006), 1469-1470. [44] J.P. Isbister, Is the clinical significance of blood storage lesions underestimated? Transfusion Alternatives in Transfusion Medicine 5 (2003), 356-362. [45] S. Basran, R.J. Frumento, A. Cohen, S. Lee, Y.L. Du, E. Nishanian, H.S. Kaplan, M. Stafford-Smith and E. Bennett-Guerrero, The association between duration of storage of transfused red blood cells and morbidity and mortality after reoperative cardiac surgery, Anesth. Analg. 103 (2006), 15-20. [46] J. Ho, W.J. Sibbald and I.H. Chin-Yee, Effects of storage on efficacy of red cell transfusion: when is it not safe? Crit. Care Med. 31 (2003), S687-S697. [47] T. Eastlund, Vasoactive mediators and hypotensive transfusion reactions, Transfusion 47 (2007), 369372. [48] J.P. Isbister and J.C. Biggs, Reactions to rapid infusion of stable plasma protein solution during large volume plasma exchange, Anaesth. Intensive Care. 4 (1976), 105-107. [49] C.F. Hogman and H.T. Meryman, Red blood cells intended for transfusion: quality criteria revisited, Transfusion 46 (2006), 137-142. [50] B.D. Spiess, Risks of transfusion: outcome focus, Transfusion 44 (2004), 4S-14S. [51] M. Goldman, K.E. Webert, D.M. Arnold, J. Freedman, J. Hannon and M.A. Blajchman, Proceedings of a consensus conference: towards an understanding of TRALI, Transfus. Med. Rev. 19 (2005), 2-31. [52] B.R. Curtis and J.G. McFarland, Mechanisms of transfusion-related acute lung injury (TRALI): antileukocyte antibodies, Crit. Care Med. 34 (2006), S118-S123. [53] J. Bux, Transfusion-related acute lung injury (TRALI): a serious adverse event of blood transfusion, Vox. Sang. 89 (2005), 1-10. [54] C.C. Silliman and N.J. McLaughlin, Transfusion-related acute lung injury, Blood Rev. 20 (2006), 139159. [55] C.C. Silliman, The two-event model of transfusion-related acute lung injury, Crit. Care Med. 34 (2006), S124-S131. [56] C.C. Silliman and M. Kelher, The role of endothelial activation in the pathogenesis of transfusionrelated acute lung injury, Transfusion 45 (2005), 109S-116S. [57] B.D. Spiess, Red cell transfusions and guidelines: a work in progress, Hematol. Oncol. Clin. North Am. 21 (2007), 185-200.

240


[58] Practice guidelines for perioperative blood transfusion and adjuvant therapies: an updated report by the American Society of Anesthesiologists Task Force on Perioperative Blood Transfusion and Adjuvant Therapies. Anesthesiology 105 (2006), 198-208. [59] S.C. Rao, J.G. Jollis, R.A. Harrington, C.B. Granger, L.K. Newby, P.W. Armstrong, D.J. Moliterno, L. Lindblad, K. Pieper, E.J. Topol, J.S. Stamler and R.M. Califf, Relationship of blood transfusion and clinical outcomes in patients with acute coronary syndromes, JAMA 292 (2004), 1555-1562. [60] S. Huber-Wagner, M. Ovick, T. Mussack, E. Euler, M.V. Kay, W. Mutschler and K.G. Kanz, Massive blood transfusion and outcome in 1062 polytrauma patients: a prospective study based on the Trauma Registry of the German Trauma Society, Vox. Sang. 92 (2007), 69-78. [61] R. Wettstein, D. Erni, M. Intaglietta and A.G. Tsai, Rapid restoration of microcirculatory blood flow with hyperviscous and hyperoncotic solutions lowers the transfusion trigger in resuscitation from hemorrhagic shock, Shock 25 (2006), 641-646. [62] P. Cabrales, J. Martini, M. Intaglietta and A.G. Tsai, Blood viscosity maintains microvascular conditions during normovolemic anemia independent of blood oxygen carrying capacity, Am. J. Physiol. 291 (2006), H581-H590. [63] A.G. Tsai, P. Cabrales and M. Intaglietta, Blood viscosity: a factor in tissue survival? Crit. Care Med. 33 (2005), 1662-1663. [64] P. Cabrales, A.G. Tsai, R.M. Winslow and M. Intaglietta, Extreme hemodilution with PEG-hemoglobin vs. PEG-albumin, Am. J. Physiol. 289 (2005), H2392-400. [65] P. Cabrales, M. Intaglietta and A.G. Tsai, Increased plasma viscosity sustains microcirculation after resuscitation from hemorrhagic shock and continuous bleeding, Shock 23 (2005), 549-555. [66] P. Cabrales, J. Martini, M. Intaglietta and A.G. Tsai, Blood viscosity maintains microvascular conditions during normovolemic anemia independent of blood oxygen-carrying capacity, Am. J. Physiol. 291 (2006), H581-H590. [67] R. Wettstein, A.G. Tsai, D. Erni, A.N. Lukyanov, V.P. Torchilin and M. Intaglietta, Improving microcirculation is more effective than substitution of red blood cells to correct metabolic disorder in experimental hemorrhagic shock, Shock 21 (2004), 235-240. [68] C. Thyes and D.R. Spahn, Current status of artificial O2 carriers, Anesthesiol. Clin. North Am. 23 (2005), 373-389. [69] R.M. Winslow, Targeted O2 delivery by low-p50 hemoglobin: a new basis for hemoglobin-based oxygen carriers, Artif. Cells Blood Substit. Immobil. Biotechnol. 33 (2005), 1-12. [70] R. Wettstein, P. Cabrales, D. Erni, A.G. Tsai, R.M. Winslow, M. Intaglietta, Resuscitation from hemorrhagic shock with MalPEG-albumin: comparison with MalPEG-hemoglobin, Shock 22 (2004), 351-357. [71] A.G. Tsai, P. Cabrales and M. Intaglietta, Oxygen-carrying blood substitutes: a microvascular perspective, Expert Opin. Biol. Ther. 4 (2004), 1147-1157. [72] K.W. van Meter, A systematic review of the application of hyperbaric oxygen in the treatment of severe anemia: an evidence-based approach, Undersea Hyperb. Med. 32 (2005), 61-83. [73] J. Meier, G.I. Kemming, H. Kisch-Wedel, J. Blum, A. Pape and O.P. Habler, Hyperoxic ventilation reduces six-hour mortality after partial fluid resuscitation from hemorrhagic shock, Shock 22 (2004), 240-247. [74] J.P. Isbister, Decision making in perioperative transfusion, Transfus. Apheresis. Sci. 27 (2002), 19-28. [75] C.V. Madjdpour, V. Heindl and D.R. Spahn, Risks, benefits, alternatives and indications of allogenic blood transfusions, Minerva Anestesiol. 72 (2006), 283-298. [76] D.R. Spahn, Strategies for transfusion therapy, Best Pract. Res. Clin. Anaesthesiol. 18 (2004), 661-673. [77] S.M. Picker, S.S. Sturner, L. Oustianskaja, B.S. Gathof, Leucodepletion leads to component-like storage stability of whole blood--suggesting its homologous use? Vox. Sang. 87 (2004), 173-181. [78] D. Bratosin, S. Leszczynski, C. Sartiaux, O. Fontaine, J. Descamps, J.J. Huart, J. Poplineau, F. Goudaliez, D. Aminoff and J Montreuil, Improved storage of erythrocytes by prior leukodepletion: flow cytometric evaluation of stored erythrocytes, Cytometry 46 (2001), 351-356. [79] S.J. Wagner and A.C. Myrup, Prestorage leucoreduction improves several in vitro red cell storage parameters following gamma irradiation, Transfus. Med. 16 (2006), 261-265. [80] P.C. Frake, H.E. Smith, L.F. Chen and W.L. Biffl, Prestorage leukoreduction prevents accumulation of matrix metalloproteinase 9 in stored blood, Arch. Surg. 141 (2006), 396-400. [81] M.I. Gyongyossy-Issa, S.L. Weiss, S.O. Sowemimo-Coker, R.B. Garcez and D.V. Devine, Prestorage leukoreduction and low-temperature filtration reduce hemolysis of stored red cell concentrates, Transfusion 45 (2005), 90-96. [82] H. Bessos and J. Seghatchian, Red cell storage lesion: the potential impact of storage-induced CD47 decline on immunomodulation and the survival of leucofiltered red cells, Transfus. Apheresis. Sci. 32 (2005), 227-232.


241

[83] S. Obara and H. Iwama, Prestorage leukocyte reduction prevents the formation of microaggregates that occlude artificial capillary vessels, J. Crit. Care 19 (2004), 179-186. [84] C.S. Luk, L.A. Gray-Statchuk, G. Cepinkas and I.H. Chin-Lee, WBC reduction reduces storageassociated RBC adhesion to human vascular endothelial cells under conditions of continuous flow in vitro, Transfusion 43 (2003), 151-156. [85] M.D. Rosario, E.W. Rumsey, G. Arakaki, R.E. Tanoue, J. McDanal and J.J. McNamara, Blood microaggregates and ultrafilters, J. Trauma 18 (1978), 498-506. [86] O.F. James, The occurence and significance of microaggregates in stored blood, Eur. J. Intensive Care. Med. 2 (1976), 163-166. [87] J. Girdano, M. Zinner, R.W. Hobson and A. Gervin, The effect of microaggregates in stored blood on canine pulmonary vascular resistance, Surgery 80 (1976), 617-623. [88] K. Wilson, M. Wilson, P.C. Hebert and I. Graham, The application of the precautionary principle to the blood system: the Canadian blood system's vCJD donor deferral policy, Transfus. Med. Rev. 17 (2003), 89-94. [89] J. Seghatchian, nvCJD and leucodepletion: an overview, Transfus. Sci. 22 (2000), 47-48.

242


Methods in Hemorheology a

Max R. HARDEMANa,1, Peter.T. GOEDHARTa, Sehyun SHINb Department of Physiology, Academic Medical Center, Amsterdam, The Netherlands b Department of Mechanical Engineering, Korea University, Seoul, Korea.

Introduction In 1678, ages before there was any concept of blood viscosity, it was appreciated by Anthony van Leeuwenhoek (Delft, the Netherlands) that red blood cells (RBC) have to deform in order to negotiate capillary passages [1]. Using his home-made microscopes he also noticed the phenomenon of reversible red cell aggregation in relation to slow and stagnant in vivo blood flow [2]. Van Leeuwenhoek was far ahead of his contemporary scientists, and although the existence and importance of blood circulation was recognized 175 years later (i.e., around 1850 by Harvey and Virchof), studies of the flow behavior of blood, hemorheology, were neglected for at least another 80 years. In 1931, Fåhraeus and Lindqvist published their classic study entitled “The Viscosity of Blood in Narrow Capillary Tubes” [3] that re-awakened scientific interest in hemorheology. However, until fairly recently, progress in this branch of science has been relatively slow, primarily due to the absence of reliable hemorheological laboratory instruments. Because of the non-Newtonian flow behavior of whole blood, viscosity measurements were initially limited to studies of plasma and serum. In fact, the clinical condition known as “Hyperviscosity Syndrome” was originally defined solely by high plasma viscosity. However, special instruments developed in the last few decades and dedicated to the measurement of various hemorheological parameters (e.g., whole blood viscosity, RBC aggregation, RBC deformability) greatly improve our measurement abilities and thus are a welcome addition to the early plasma viscometers. In this chapter no comprehensive treatment of all described methods and techniques has been attempted. Rather, instead of listing all home-made techniques and instruments that have appeared in the literature and sometimes used only once by a single group, a survey of generally accepted methods and/or techniques as well as some recent developments are presented. Relevant website addresses are provided in the hope that these will be updated regularly by the manufacturers and vendors. However, in order to increase the direct practical value of this chapter, more detailed descriptions of relevant, commercially available instruments are given. Comparative studies of these methods and/or instruments are briefly discussed. Note that the latest version of the “Guidelines for Measurement of Blood Viscosity and Erythrocyte Deformability” [4] dates back to 1986 and a more recent version, comparing the newer techniques and instruments and giving generally agreed upon recommendations, is urgently needed. Finally, some general practical hemorheological laboratory techniques are described. 1

Corresponding author: Department of Physiology, Academic Medical Center, Amsterdam, The Netherlands; E mail: [email protected]

M.R. Hardeman et al. / Methods in Hemorheology

243

1. Whole Blood Viscosity Unlike “static” clinical chemistry values such as plasma sodium content or RBC hemoglobin concentration, the viscosity of blood is a “dynamic” property. This means that in viscometry, the suspension must flow for this property to be measured. The rate of flow that occurs as a result of a pressure difference depends on the sample’s viscosity. Elaboration of nomenclature as well as a detailed treatment of the matter can be found elsewhere in this volume (see Chapter II.3.a). It is important to point out that, in contrast to plasma, whole blood is a non-Newtonian suspension and thus its viscosity is dependent on the shear rate (i.e., flow rate) at which it is measured. Therefore, we can not speak about a single viscosity of blood, but rather should always associate this parameter with the corresponding shear rate that has been used in its measurement. Whole blood viscometers should therefore be able to operate at accurately defined shear rates. Both tube flow and rotational viscometers (e.g. cone-plate geometry or two concentric cylinders known as a Couette system) are usable, although determining shear rate is more direct when using rotational systems. Since the viscosity of blood is dependent on a range of internal and external variables which are different in various parts of the macro- and micro-circulation, there can be some concern about the clinical relevancy of in vitro assays (see below). Furthermore, two different blood samples may have almost equal high shear viscosities but appear to diverge significantly at low shear rates due to a difference in RBC aggregation. Possible confusion associated with such findings can be avoided if we define the behavior of blood viscosity as the laboratory parameter for a particular blood sample (i.e., always state viscosity values for at least two shear rates, one at a high and one at a low shear rate). The exact level of shear rates has still to be defined, but it is necessary that the lowest shear rate should be low enough (e.g. < 5 s-1) to allow a significant degree of RBC aggregation. Unfortunately, many of the current industrial viscometers do not meet this criterion (see [4]). Because the viscosity of a fluid is a temperature-dependent property, temperature control is an essential feature of any viscometer. Therefore, the measuring temperature should always be specified in communications concerning blood viscosity measurements. Usually, such measurements are conducted at 37 °C although some tests (e.g., erythrocyte sedimentation rate) are performed at room temperature.

2. Plasma or Serum Viscosity Unlike suspensions of red cells in plasma or whole blood, plasma and serum are Newtonian fluids with their viscosity independent of shear rate. This means that their viscosity is an intrinsic property of the liquid itself and there is no need to measure plasma or serum viscosity at defined shear rates. Consequently, the viscometer can be rather simple (e.g., capillary type, see below), but should be designed to avoid surface film artifacts and the appearance of “non-Newtonian” serum or plasma (see chapters II 1 and II 3.a). Since the viscosity of normal plasma increases by 2-3% for every degree decrease in temperature from 37 °C to 15 °C, temperature control is critical [5].

244


3. Survey of Currently Used Methods and Instruments 3.1.1. Viscometry Various types of viscometer have been developed and applied in hemorheological studies (e.g., capillary, rotational, falling ball, oscillating flow). Most of the viscometers until the early 1960`s were of the capillary type and based on the principle of measuring the rate of flow through a glass tube of specified dimensions. The shear stress in a capillary tube, however, is not uniform but greatest at the wall and zero at the center. Moreover, in simple U-tube viscometers, the pressure gradient (i.e., pressure difference across the tube generated by two fluid-filled vertical tubes of unequal height), and hence flow decrease with time. The sample is exposed to a range of shear rates making it difficult to define a single shear rate at which the particular viscosity is measured. For these reasons, simple U-tube capillary viscometers are generally unsuitable for absolute measurements of non-Newtonian whole blood viscosity. However, this type of relatively inexpensive viscometer has been frequently used for the determination of plasma and serum viscosities. The well known Harkness tube viscometer which has been recommended by the I.C.S.H. as the standard method for measuring serum and plasma viscosity [6] and its successor the Coulter Viscometer II tube viscometer are, unfortunately, no longer in production. A new computerized tube viscometer, calculating viscosity-shear rate data from time-dependent pressure gradient and flow data (RheologTM, see below) is in development. In rotational viscometers the sample is sheared in a narrow gap between two surfaces, one of them being static and the other rotating at various speeds thereby producing specific shear rates. Usually the stress exerted on the static surface is measured and used for calculation of viscosity as shear stress/shear rate, although some devices (e.g., cone-plate) measure stress on the rotating element. Both plate-plate and cone-plate geometry can be used (Figure 1). The cone-plate configuration ensures that the entire sample is exposed to the same shear rate. In this geometry, the velocity difference between the two surfaces increases towards the periphery but so does the distance between them, thereby ensuring that the velocity gradient or shear rate remains constant throughout the sample Another frequently used geometry in rotational viscometry consists of two coaxial cylinders (i.e., Couette system) with one fixed and one rotating cylinder. While the uniformity of the shear field is less than in the cone-plate geometry (i.e., variation across the gap between cylinders proportional to square of radius ratio), it can be shown that for a large internal radius - gap width ratio, shear rate variations are minimized and blood viscosity is not significantly affected [7]. Note that measurements using both the cone-plate and Couette geometries can be affected by surface film artifacts at the interface between air and blood or plasma. In addition, both can be affected by RBC aggregation that causes settling away from the upper surface in a cone-plate or movement away from both cylindrical surfaces in a Couette device. In addition, torque-time artifacts are possible for each geometry (see chapters II.1 and II.3.a). The sensitivity and precision of rotational viscometers depends strongly on the torque measuring system (e.g., its sensitivity and dynamic response) and on accurate rotational speeds [8].


245

Figure 1. Cone-plate (left) and plate-plate (right) geometry often used in rotational viscometry.

Figure 2. Principles of oscillatory viscometry. The variations of shear stress (Ĳ), shear deformation (Ȗ) and shear rate (Ǥ) are all sinusoidal in time, but have different phases. The phase of Ǥ is always 90o ahead of Ȗ but the phase of Ĳ depends on the fluid properties [9].

Oscillatory-flow viscometers are employed to study the viscous and elastic properties of a blood sample, with this technique used in both rotational and tube viscometers. The viscous component (Ș’) is calculated from the measured amplitude of the shear stress and the elastic component (Ș’’) from the phase lag of shear stress relative to shear rate (see Figure 2 and chapter II.3.b). The basic concept employed in a falling ball viscometer is to measure the time a steel or glass ball needs to travel over a certain distance through a vertical tube filled with the sample. A variation is the rolling ball viscometer with a tilted tube. The speed of the ball depends on the density and viscosity of the fluid, and the shear rate is difficult to calculate and is not constant: the faster the ball moves, the higher the shear rate. Accurate calibration, as well as known sample density, is essential for an accurate measurement of viscosity. Note that this method is unsuitable for a non-Newtonian

246


fluid such as whole blood but may be of use for plasma, serum, or any other simple fluid (e.g., sugar solutions, some polymer solutions). Viscometers most used and suitable for blood studies include: x Wells-Brookfield (www.brookfieldengineering.com). One of the first rotational viscometers was described in 1961 by Wells et al. [10], and several cone-plate models are now marketed under the name Wells-Brookfield DV I, II, and III. Inasmuch as the instrument has a single torque measuring range, useful shear rates depend on the sample’s viscosity (i.e., either too low a reading or an off scale reading if the shear rate is too low or too high). The nominal shear rate range is a function of cone angle and motor speed; blood studies often use 75-1500 s-1. Approximate price (2006): Euro 3.900 – 5.900, US$ 5,000 - 7,000. x Contraves ProRheo (www.proRheo.de). The rotational low shear viscometer (LS-30) first produced by Contraves is still the most frequently used blood viscometer. It is usually operated with a Mooney sample chamber (i.e., Couette plus cone-plate at bottom of inner cylinder), requires a small sample volume (e.g., one ml), and has a shear rate range of 0.01 to 95 s-1. Unfortunately, after production of an updated model (LS-40), Contraves stopped production, and then subsequently transferred ownership to proRheo GmbH (Althengstett, Germany). This company now provides spare parts, calibration and repair, and is presently developing a LS-300 viscometer as a successor to the LS30. Anticipated price for LS-300: Euro 50.000, US$ 65,000. x RheologTM, Rheologics Inc. (www.rheologics.com). This new automated scanning capillary viscometer has been tested [11] and compared to other viscometers (see below) but is currently not commercially available. In this viscometer the pressure gradient is the height difference of blood in a U-shaped assembly with two vertical blood-filled columns connected at their lower ends by a calibrated capillary tube. The heights of the two blood columns are continuously measured by LED scanners and digitally recorded; the vertical columns and the calibrated tube are temperature controlled at 37 qC. Flow is rapid at the beginning of the measurement since the height difference is maximal, then decreases as the height difference becomes smaller. This gives a flow versus pressure curve that is calculated from the time and heights measured by the LED scanners. The results are then utilized to calculate blood viscosity versus shear rate data using proprietary software based on the Casson relationship. This viscometer is portable, uses disposable measuring tubes and can be used to measure viscosity at a patient’s bedside. Price has not yet been established by the manufacturer. x Vilastic Bioprofiler (www.vilastic.com). The Bioprofiler is the successor of the Vilastic-3 Analyzer and is based on the oscillatory tube flow principle. This viscometer is designed for measurements of blood viscoelasticity and of plasma viscosity (see chapter II.3.b). It requires 0.5 to 1 ml of sample, operates at a fixed frequency of 2 Hz, and can generate a wide range of shear rates by changing the amplitude of oscillation. Viscosity values are measured at


247

room temperature and automatically adjusted to another temperature (e.g., 37 oC). Approximate price (2006): US$ 19,500. x ViscoLab 450 (www.cambridgeviscosity.com). This instrument is a variation on the falling-ball concept but uses a piston moved vertically up and down in the sample by magnets. The applied force is held constant and the travel time of the piston measured electronically. It is not yet clear if this instrument is useful for blood measurements, although it may provide an apparent viscosity at a given shear rate if the piston travel time is utilized to calculate a steady shear. Approximate price (2006): US$ 4,600. For other viscometers which may be suitable for blood rheology studies but have not yet been widely evaluated, see below (5. New Developments). 3.1.2. Comparison of Viscometers Because some widely used instruments are no longer available, and because experience with some new, commercially available machines is limited, only recent publications are provided and briefly discussed below. The most recent comparative survey, mainly restricted to capillary and fallingbody viscometers not intended primarily for clinical laboratory use, is the publication by Rosencranz and Bogen [12]. Discussed are: Ostwald U-tube viscometer, Harkness viscometer and its successor Coulter Viscometer II, Schott AVS automated capillary viscometers, Stony Brook falling-needle viscometer, Anton Paar AMVn automated viscometer, ThermoHaake Micro Visco 2 falling ball viscometer, Cambridge ViscoLab 410, Wells-Brookfield cone-plate viscometer. They concluded that of the viscometers tested, the Viscolab 410 was the best choice for their clinical laboratory. Rosenson and Wolff compared the RheologTM scanning capillary viscometer and the Contraves LS-40 rotational viscometer [13]. A validation study of the RheologTM was recently carried out by Alexy, et al. and the results were compared to those obtained with the Wells-Brookfield cone-plate viscometer and Contraves LS-30 rotational viscometer [14]. Both studies concluded that, within the hematocrit and shear rate limits employed (i.e. 28-56% and 1-1500 s-1), this capillary viscometer has potential for both clinical and research applications. 3.2. Erythrocyte Deformability The resting erythrocyte has a biconcave discoid shape with an average diameter of approximately 7-8 μm. Since the diameter of the capillaries average 3-5 μm, the cell must therefore be able to undergo rapid, large scale deformations in order to traverse the microcirculation and nutritive capillaries. During deformation both the cell volume and the surface area remain constant [15]; this is only possible because the erythrocyte has an excess surface area of about 40% compared to a sphere of equivalent volume. Beside the surface/volume ratio, deformability is dependent on intracellular viscosity, mainly determined by the Mean Cell Hemoglobin Concentration (MCHC). In the normal erythrocyte internal viscosity is about 6 mPa.s at 37 °C [16]. The internal viscosity can be greatly increased in the presence of abnormal hemoglobins (e.g., HbSS in sickle cell anemia [17]) or in the presence of the tropical malaria parasite, Plasmodium Falciparum [18]. Finally, erythrocyte deformability is influenced by the visco-elastic mechanical properties of the cell membrane [19].

248


Methods based on the measurement of individual RBC deformability have been of great value for obtaining a deeper insight into the fundamental mechanisms involved in this cellular property. For routine clinical applications, however, tests performed on suspensions are of practical value since they provide the average behavior of a large population of cells. Separation of RBC sub-populations prior to deformability measurements (e.g., on the basis of their density) increases the information gained from such studies; methods that provide a RBC deformability distribution histogram help identify sub-populations of RBC with different deformability. 3.2.1. Filterability Initial measurements of erythrocyte deformability were performed by determining the ability of cells to traverse polycarbonate membranes with pore sizes of 3-5 μm. Many home-made variations of the original method of Reid, et al. [20] have been published, and include filtration under gravity force or by application of positive or negative pressure; studies recording pressure changes at constant flow have also been carried out. Quantification of the process is achieved either by the time required for passage of a certain volume of a RBC suspension or by the pressure-flow relationship. Practical problems encountered with this technique are blockage of the pores by more rigid leukocytes or by the presence of platelet micro aggregates; variations in the filters themselves can also cause poor reproducibility. An automated, computer-assisted micropore system (Cell Transit Analyzer, “CTA”, ABX, France), using special 30-pore membranes and generating a pore transit-time histogram, is generally considered to be one of the best of its kind. Unfortunately, production of this instrument was terminated a few years ago. 3.2.2. Micropipette Techniques With this technique segments of the cell membrane or the entire erythrocyte are partially or completely aspirated into glass capillaries of 1-5 μm. in diameter [21]. Quantification is based on the amount of negative pressure needed to aspirate part or the entire cell: partial aspiration of a portion of the membrane yields the shear elastic modulus of the membrane, while complete aspiration measures cellular deformability. These single cell measurements can provide relevant information on the visco-elastic properties of the membrane and of the entire cell. However, a high degree of technical skill is required and micropipette fabrication reproducibility can be problematic. Because cells are studied individually and the method is time-consuming, only small numbers of cells can be studied, and thus the overall effectiveness of the micropipette technique for extensive hemorheological studies is limited (see chapter II.4.a). 3.2.3. Micro Channels A special form of filtrometry is made possible via the use of an array of parallel micro channels produced by a technique employing micromachining. With this technique the deformation of whole cells can be observed visually. However, red cell deformability as well as the presence of leukocytes and/or platelet aggregates affects flow through the channels (see chapter II.4.a). x MC FAN HR300 Blood Fluidity Analyzer (www.mclab.co.jp/en). Distributed by Arkray Inc. Japan. Price indication (2006): US$ 61,500.


249

3.2.4. LaserDiffraction Ektacytometry Another technique for the assessment of RBC deformability is based on laser diffraction ellipsometry or ektacytometry. In this technique, a suspension of erythrocytes in a high viscosity medium is subjected to flow stress, and analysis of the laser diffraction pattern resulting from a laser beam transversing the suspension provides a measure of cell deformability. Useful geometries for shearing the suspension include cone-plate or plate-plate, a parallel plate flow system and a Couette system (i.e., the gap between two coaxial cylinders). In the Couette system, the inner cylinder (bob) can be static while the outer cylinder (cup) rotates driven by a motor; the reverse static/rotate operation is also possible. Since the rotational speed of the moving element, the gap width between the two cylinders and the viscosity of the suspension medium are all known, the applied shear stress can be calculated (Figure 3). However, because the shear rate in a Couette system varies across the gap as the square of the radii ratio, and since there can be optical averaging through the gap, proper design favors large radii and a narrow gap.

Figure 3. Schematic presentation of rotational ektacytometric measurement of RBC deformability. The diffraction image can be analyzed using diode pairs on the screen or via video image analysis. EI is a geometric parameter and increases with cellular deformation. Photo diodes within the inner cylinder are not used for deformability studies.

As shown in Figure 4, the laser diffraction pattern is circular for cells at rest and becomes elongated as the cells are deformed and aligned parallel to flow. In the system schematically shown in Figure 3, the diffraction pattern is projected on a screen and recorded with a video camera. Subsequent computer analysis involving ellipse fitting of the pattern allows the calculation of the Elongation Index (EI). An alternative method for analyzing the diffraction pattern utilizes four light sensitive diodes equidistantly positioned along the two orthogonal axes of the diffraction pattern. Normal RBC exhibit a S-shaped deformation curve as shown in Figure 5. Various authors have investigated parameterization of such ektacytometric deformation curves in order to define one or more global parameters of deformability [22, 23]. Two approaches to simplify data presentation have been compared in terms of their power to detect a difference between groups, with results indicating equal ability to identify differences [24].

250


Figure 4. Typical laser diffraction patterns obtained with a Couette shearing system. Images shown are for cells at rest (left panel) and for cells subjected to a shear stress of 30 Pa (center). The right panel demonstrates the effects of the presence of a 25% subpopulation of rigid cells at a shear stress of 30 Pa.

Figure 5. Typical RBC deformation-shear stress curve for normal RBC suspended in an isotonic viscous medium. EI is the calculated Elongation Index.

3.2.4.1. Ektacytometers x Laser-Assisted Optical Rotational Cell Analyzer (LORCA); produced under the name LoRRca by R&R Mechatronics, Hoorn, the Netherlands, Price indication (2006): Euro 35.600, (www.mechatronics.nl ). US$ 46,000 (Figure 6). This instrument has multiple hemorheological applications: RBC deformability, RBC aggregation (see 4.6) and the rate of RBC relaxation or shape recovery (see 4.4) [25, 26]. The shearing system is of the Couette type, and 25 μl of blood is needed. The cells are exposed to a constant, user-adjustable temperature and defined shear stresses approximating those observed in the circulation [27]. Diffraction patterns are recorded by a CCD video camera and analyzed by computercontrolled ellipse fitting.


251

Figure 6. Laser-assisted Optical Rotational Cell Analyzer (LORCA) for measurement of RBC deformability, RBC aggregation and RBC shape recovery

x

Rheodyn-SSD Laser Diffractometer; Myrenne, Röttgen, Germany (www.myrenne.com). Price indication (2006): Euro 14.450, US$ 19,000 (Figure 7). The measuring system of this instrument is based on plateplate geometry. It operates at room temperature and analyzes the diffraction patterns with a four-diode sensor. Measurements are made at various shear rates utilizing a very small blood sample diluted in a viscous medium [28].

Figure 7. Rheodyn SSD for measurement of RBC deformability

. x

RheoScan-D Microfluidic Ektacytometer, Sewon Meditech,Inc, Seoul, Korea: (www.rheoscan.com). Price indication (2006): US$ 29,000; disposable chip: US$ 1.2 (Figure 8). In this instrument the laser diffraction technique has been combined with micro fluidic rheometry using a disposable element which is in contact with the blood sample [29]. The measuring principle is based on vacuum driven shearing of a thin layer of a RBC suspension through a micro channel, with the RBC deformation analyzed as a function of shear stress (i.e., pressure gradient) during passage through the channel.

252


Figure 8: RheoScan-D for measurement of RBC deformability

3.2.5. Comparison of RBC Deformability Measurement Techniques Only relatively few direct comparative studies regarding different methods for measurement of RBC deformation have been published. Deformability values of RBC from 525 patients with various diseases measured by both ektacytometry (LORCA) and by filtration (Cell Transit Analyzer) showed a rather poor correlation (r=0.036, p<0.5) between the techniques [26]. In contrast, RBC deformability measurements by laserdiffraction technique and cone-plate Rheoscope were found in close agreement (r=0.965) over a shear stress range of 0.6 – 24.6 Pa [30]. Wang, et al. [31] compared RBC deformability measurements using two ektacytometers (i.e., LORCA and Rheodyne SSD) and found that both instruments had similar reproducibility but showed different values for the same blood samples. They concluded that EI as calculated by LORCA via ellipse-fit of diffraction pattern was closer to the actual cell deformation than EI as calculated by Rheodyn SSD that analyzes light intensity on 4 equidistant points of the diffraction pattern. Baskurt, et al. [32] demonstrated the importance of temperature control: detection of a given impairment in RBC deformability induced by experimental sepsis in rats was only observed by ektacytometry measurements at 37 °C. 3.3. Erythrocyte Shape Red blood cell shape is most commonly evaluated by light microscopy, using brightfield, phase contrast or interference contrast optics. Cells are suspended at a low hematocrit (§ 1%) in the desired medium and placed between a slide and cover slip or into a “wet mount” well that provides some separation between the slide and cover slip; this separation avoids the disc to spiked sphere “glass slide” shape change artifact that can occur when only a slide and coverslip are used. A rapid flow cytometric technique to assess the distribution of RBC shapes and its association with various membrane components has been described by Piagnerelli, et al. [33]. 3.4. Erythrocyte Shape Recovery RBC shape change in response to deforming forces is a reversible process, with the cell regaining its original bi-concave shape when the forces are removed. This mechanical


253

aspect of the cell can be studied using micropipettes (i.e., pull and release the edge of the cell opposite to the edge attached to a flat surface), but as mentioned above it is a labor-intensive method; its advantage lies in its ability to obtain data for specific types of cells in a population. RBC shape recovery has also been studied after abrupt cessation of shear in a Couette system either by analyzing the time course of laser light reflection or by serial measurements of video elongation indexes from laser diffraction patterns (ektacytometry) [34]. Software available for the LORCA system allows the RBC relaxation time constant to be calculated using reflected laser light [35]. 3.5. Erythrocyte Orientation in Flow During shear flow, the membrane of normal RBC moves around the cell interior in a so-called “tank tread” motion [36], resulting in orientation of the deformed cells with the flow vector, like a school of fishes (Figure 9). This behavior of RBC is essential for the uniform ellipsoid diffraction pattern obtained during ektacytometry. Without a tank treading membrane, elongated RBC will not orientate but rather tumble, resulting in a distorted diffraction pattern and zero or negative EI values. Orientation of RBC in a shear field has been demonstrated and quantified using rheoscope methods, in which a microscope is used to observe cells exposed to fluid shear, see below (5. New developments).

Figure 9. Orientation of erythrocytes in viscometric flow under high shear conditions [37]

3.6. Erythrocyte aggregation In contrast to blood platelets, human RBC aggregate spontaneously in resting whole blood; under normal circumstances RBC aggregates resemble stacks of coins and are termed rouleaux. At stasis or under very low shear rates the stacks can grow, forming end-to-end, side-to-side and side-to-end configurations, finally forming 3D-networks. Since the presence of such RBC aggregates has a considerable effect on the flow behavior of blood (see chapter II 3.a) the study of RBC aggregation has attracted great interest in the field of hemorheology. Although the phenomenon of reversible RBC aggregation was first demonstrated in the 17th century [2], it was only in 1958 when Zijlstra first described the technique for quantification of this process by measuring the decrease in light backscatter and proposed the name “syllectometry”, derived from the Greek word “ıȣȜȜİȖȠμĮȚ”

254


meaning “to gather” [38]. A stylized syllectogram for laser backscatter is shown in Figure 10. Note that the term is also used when referring to light transmission through an aggregating RBC suspension, showing the reverse, i.e., increased intensity during aggregation. An essential requirement for measurement of the RBC aggregation process by this method is that RBC aggregates are fully dispersed immediately prior to the abrupt cessation of shear. For normal human blood this state can be achieved at a shear rate of about 500 s-1. However, since incomplete dispersion of existing RBC aggregates leads to underestimation of aggregation, blood from patients in a hyperaggregating state needs a much higher initial shear rate for complete dispersion of the strong aggregates [39-42]. This same requirement is also relevant to the study of blood from strong aggregating species like horses or for cells suspended in solutions of strongly aggregating polymers.

Figure 10: Stylized Syllectogram [i.e., laser backscatter intensity (I) versus time (t)] before and following a sudden stop of a sheared blood sample at time zero. Disaggregated RBC lose their ellipsoid shape at the moment the shear stress abruptly stops and recover their resting biconcave morphology. The enlarged surface facing the laser beam results initially in increased back scatter (upstroke) immediately followed by a typical decrease of intensity due to formation of linear rouleaux and subsequent 3D-aggregates. au = arbitrary units

Apart from specially dedicated aggregometers (see below), several procedures have been used to quantify RBC aggregation: x Low-Shear Viscometry (on any suitable blood viscometer) x Erythrocyte Sedimentation Rate (ESR) x Zeta Sedimentation Ratio (ZSR, previously sold by Coulter Electronics Ltd.) In all of these methods, RBC aggregation is an essential but not the sole determining factor. It has been previously shown that alterations in other factors such as high plasma viscosity and decreased RBC deformability can affect low shear viscometry, and high plasma viscosity and density can affect sedimentation. Therefore, low shear viscosity and related indexes may not always reflect changes of RBC aggregation, and it has


x

255

been concluded that this parameter should not be the sole technique to assess RBC aggregation [43]. The same is true for erythrocyte sedimentation behavior. Ultrasound backscatter In addition to the backscatter of light, the use of ultrasound has also been investigated as a method to measure RBC aggregation [44]. An advantage of this technique is that it has the potential for non-invasive in vivo aggregation measurements. Note that for evaluating RBC aggregation kinetics, results obtained by measuring ultrasonic backscatter using a specially adapted echo probe as inner cylinder in a Couette shearing system are consistent with those obtained by syllectometry [45].

3.6.1. Aggregometers x

Myrenne MA-1, GmbH, Roetgen, Germany (www.myrenne.com). Approximate price (2006): Euro 7.400, US$ 10,000 (Figure 11). The MA1 is a compact, micro processor controlled aggregometer with a coneplate shearing system and the ability to measure infra red light transmitted through a RBC suspension. It operates at room temperature and requires a 35 μl sample. As normally supplied, RBC aggregates are dispersed using a shear rate of 600 s-1 following which light transmission is integrated for 5 or 10 seconds and there is an option to integrate the light at stasis or at a low shear rate of 3 s-1. Note that the Myrenne Aggregometer can be computer controlled via third-party software [46] in order to provide kinetic data, the minimal shear rate needed to disperse aggregates, and the Flow-to-Stasis Aggregation Ratio (FSAR).

Figure 11. Myrenne aggregometer MA-1. Cone-plate shearing system is in the compartment at the right; push button keys allow for initiating the system (A) and for measuring aggregation at stasis (M) or at low shear (M1).

x

Laser-assisted Optical Rotational Cell Analyzer (LORCA); produced under the name LoRRca by R&R Mechatronics, Hoorn, the Netherlands. (www.mechatronics.nl), Approximate price (2006): Euro 35.600, US$ 46,000, (Figures 6, 12 and 13). The aggregation mode of this instrument has adjustable shear rates from 0-3500 s-1, with high shear useful for dispersing aggregates from patients with greatly enhanced aggregation. Measurements are based on changes in laser backscatter intensity from whole blood or RBC suspensions (Figure 12). This device provides both kinetic and static aggregation indices, minimal shear rate to prevent aggregation (i.e., aggregation strength) and Flow-to-Stasis

256


Aggregation Ratio (FSAR); it can operate at room temperature and above and requires § 1.0 ml of sample.

Figure 12: Principle of laser backscatter intensity measurement for RBC aggregation. Laser light reflects from the RBC suspension and is measured by diodes embedded in the inner cylinder.

I (t )

I r e t / Tr I f e Shape recovery (upstroke)

t / T f

Rouleaux formation

I s e t / Ts I 0 3D aggregate formation

Fig. 13: RBC aggregation parameters obtained using a syllectogram as determined by the LORCA system. I= laser back scatter intensity (au = arbitrary units); Amp = amplitude; t1/2 = aggregation half time. Aggregation Index (AI) calculated as noted in upper right of figure. Tri-exponential syllectogram curve fit results in time constants for shape recovery, (fast) rouleaux, and (slow) 3D aggregate formation.


257

3.6.2. Comparison of RBC aggregometers Three optical methods to study RBC aggregation, (i.e., Myrenne MA-1, LORCA, and SEFAM Erythroaggregometer which is no longer produced) have been evaluated with regard to their sensitivity in detecting the aggregation of RBC suspended in different normal and hyper-aggregating media [47]. It was concluded that the aggregation parameters obtained by these instruments were stable and exhibited good reproducibility. The Erythroaggregometer had the lowest CV, and a calculated overall half-time for aggregation was similar for all three instruments. Regarding the suitability for clinical applications, there are results indicating that in situations of hyperaggregation (e.g., multiple myeloma), the Myrenne Aggregation Index value was paradoxically decreased compared to values found by other instruments [47]. Similar studies have been performed comparing Myrenne and LORCA as to their sensitivity towards different fibrinogen concentrations [48]: higher correlation values (p<.001) between RBC aggregation and fibrinogen concentration were obtained with the laser back scatter system. Baskurt, et al [32] have shown the importance of measurement temperature in detecting differences in RBC aggregation index and recommended the use of a temperature controlled device, set at 37 °C.

4. Special Hemorheological Laboratory Techniques 4.1. RBC Sub-Populations Obtained by Density (Age) Separation 1. Venous blood samples are centrifuged at 1,000 x g for ten minutes and the plasma removed. The RBC pellet, having a hematocrit of about 80%, is than centrifuged at 10,000 x g for 15 min. Following centrifugation the top and bottom 10% or less of the RBC column are removed and designated as young and old RBC, respectively. This technique is a modification of the method initially published by Murphy [49]. 2. A more sophisticated method that provides cell populations of defined density and is useful for pathologies such as sickle cell disease, employs density gradient centrifugation. Different dilutions of various media such as Percoll (www.sigmaaldrich.com) and Nycodenz manufactured by Nycomed Pharma, Oslo, Norway (www.technoclone.com, www.accurate chemical.com) have been used for preparation of discontinuous and continuous isotonic gradients A ready-to-use solution of the latter (OptiPrep) has the additional advantage of being able to form continuous gradients in situ by centrifugation. An example of discontinuous gradient densities suitable for separation of RBC subpopulations of sickle cell samples is: 1.08, 1.085, 1.09, 1.095, 1.10, 1.11, 1.12, 1.14 and 1.16 g/mL [50]. 4.2. Preparation of Blood Samples for Various Treatment Modalities Venous blood samples, anti-coagulated with EDTA (or heparin), are centrifuged at 1.400 x g for 10 minutes. After separation the plasma is kept for later re-suspension and the cells are washed in 10 mM isotonic Phosphate-Buffered Saline (PBS, pH 7.4). After centrifugation the washed cells are then re-suspended in PBS or plasma or other media at the required hematocrit. Aliquots of this suspension can be used for the various treatment modalities mentioned below.

258


4.3. Heat Treatment: Subtle Decrease in RBC Deformability Heat treatment is known to increase mainly the shear elastic modulus of the membrane. Washed RBC in PBS at a hematocrit of 5% are incubated at 48 °C for 5, 10, 15 or 20 minutes. Following this treatment the suspension is centrifuged again and the cells are separated and re-suspended in plasma or other media at the required hematocrit for further testing. 4.4. Glutaraldehyde Treatment for Decreased RBC Deformability Washed RBC are re-suspended in PBS at a hematocrit of 5% and incubated at 25 °C for 30 minutes in the presence of glutaraldehyde at concentrations appropriate to the desired degree of rigidification. Levels of 0.5% to 1% yield very rigid “fixed” cells whereas lower concentrations up to about 0.1% can generate cells with graded degrees of decreased deformability. Following this treatment the cells are washed three times in PBS and re-suspended in plasma or other media at the required hematocrit for further testing. 4.5. Ghost Preparation Several techniques for the preparation of resealed RBC-ghosts (i.e., discoid membranes obtained after the removal of hemoglobin) have been described. The cytoplasmic composition depends on the preparatory method, and can yield white ghosts which are nearly hemoglobin free, or red resealed ghosts containing more residual hemoglobin. The following procedure is from Yip et al [51]. Fresh blood is washed three times in a solution containing 5 mM Tris-HCl (pH 7.4) and 140 mM NaCl (isotonic Tris-saline). Ghosts are prepared by hypotonic lysis in 20 volumes of ice-cold 5 mM Tris-HCl, 7 mM NaCl (pH 7.4) and are washed once with the same lysing buffer. After centrifugation and removal of supernatant, the pellet is re-suspended in 10 volumes of isotonic Tris-saline and incubated at 37 0C for 1 hr to promote resealing, then re-suspended to the desired hematocrit. For ektacytometric studies of ghosts and the generation of a well-defined laser-diffraction pattern it is recommended to use Stractan as the suspending medium since it has a higher refractive index than currently used media prepared with PVP or dextran [52]. 4.6. Correction of Viscosity to Specific Hematocrit A formula for calculating blood viscosity at a specific hematocrit (e.g., 45%) based on knowing plasma viscosity, the original hematocrit of the blood tested, and viscosity of the blood sample at original hematocrit has been described by Matrai, et al. [53]:

Șrel (45) = {Șrel (Hct) } 45/ Hct where Șrel is the viscosity of blood relative to that of plasma, Hct is the original hematocrit (%) of the blood sample, and 45 the specific hematocrit (%) selected. Note that this equation employs a semi-log relation for viscosity versus hematocrit, and generates a line based upon two data points: relative viscosity of unity (i.e., plasma viscosity) at zero hematocrit and known relative viscosity at sample hematocrit.


259

4.7. Hematocrit Adjustment to Specific Value A simple calculation procedure for preparation of blood samples with a specific hematocrit, using centrifugation followed by concentration and/or dilution with plasma, can be based upon the following approach: Let A= starting volume (ml), B= starting hematocrit (%) and C= desired final hematocrit (%). When B < C, the volume (ml) of plasma that has to be added can be calculated as A (1 - B/C), and when B > C, the volume (ml) of plasma that has to be removed is A (B/C - 1).

5. New Developments x

x x

x

Automated Rheoscope and Cell Analyzer (ARCA), (www.mechatronics. nl) (under development). This instrument combines the technique of ektacytometry with microscopy (Figure 14). A dilute RBC suspension is subjected to shear stress between two counter-rotating glass plates, and the elongated cells are observed with an inverted microscope using stroboscopic illumination [54]. An image analyzing computer program selects individual, non-overlapping RBC that are in focus and calculates the Elongation Index obtained at a given shear stress by ellipse fitting. Automation allows the measurement of a statistically significant number of cells within a reasonable time limit: 1000 images can be acquired in about 15 min; the operator then needs about one hour for automated analysis and visual validation of these images to produce a deformability distribution histogram. Finally, polar plots indicating the orientation of elongated RBC relative to the direction of shear flow can be obtained (Figure 15). As has been discussed in section 4.5 of this chapter, RBC are normally orientated to the flow vector. Abnormal orientation (i.e., perpendicular to the flow vector) has been reported to occur in certain diseases (e.g., sickle cell anemia [55] and elliptocytosis [56]). Similar behavior has been demonstrated in blood from the Lama Vicuna (Cameloid species) having circulating rigid but ellipsoidal RBC [57]. ProRheo LS-300, the successor of Contraves LS 30, available in 2007; (www.proRheo.de). An oscillating viscometer for bedside measurements, recently described [58]; ([email protected]). The viscometer consists of an oscillating resonator probe mounted directly into a disposable vacutainer tube for blood withdrawal. Results were generally higher than by Couette viscometry. A pressure-scanning capillary viscometer [59]; ([email protected]). A single measurement of pressure variation with time replaces the flow rate and pressure drop measurements that are usually required for the operation of a capillary tube viscometer.

260


Figure 14: Schematic drawing of the Automated Rheoscope and Cell Analyzer (ARCA). The upper plate is mounted on a holder which serves as a water bath for controlling the suspension temperature. The two plates rotate in opposite directions (CW=clockwise, CCW=counter clockwise) [54].

Figure 15. RBC orientation relative to shear vector (ĳ=0).; a) healthy control at 3 Pa; b), c) and d) elliptocytosis at respectively 1, 3 and 10 Pa [54].


261

6. Standardization and Reference Values 6.1. Effect of Anticoagulant No effects related to the type of anti-coagulant used for hemorheological measurements have been reported, except the effect of dilution associated with the use of citrate solutions that decrease viscosity and RBC aggregation. Currently, EDTA (1.5 mg/ml) is primarily used as anti-coagulant in hemorheological studies. 6.2. Maximal In Vitro Time Since blood is a biological material, possible changes occurring following removal from its physiological milieu must be considered. Time and temperature conditions are of paramount concern since it is often necessary to transport a blood sample to either a local clinical laboratory or to a more-distant laboratory or colleague. According to the ICSH Guidelines [4], hemorheological measurements should be performed within four hours after the sample was collected: if the sample is kept at room temperature, measurements should be carried out within 4-8 hours [60, 61], and up to a 12 hour period is suggested for samples kept at 4 °C [62]. However, a more recent study of blood viscosity versus storage times and conditions indicates an absence of major changes for up to eight hours at room temperature, up to four days at 4 °C and up to six hours at 37 °C [14]. For mice, rats, guinea pigs and dogs even shorter times apply (i.e., 0.5, 1, 3, and 7 hrs, respectively) [63]. All studies indicate that storage changes are most significant for measurements performed at low shear rates [14, 16]. 6.3. Temperature Control Since under normal circumstances the viscosity of aqueous liquids decreases with increasing temperature, it is important that measurements be performed at a constant and well defined temperature, usually 37 °C. 6.4. Reference Standards Unfortunately, reference standards are lacking for most hemorheological laboratory measurements although viscometers can be calibrated with commercially available Newtonian standards having appropriate viscosity levels (www.paragonsci.com/htm/p-vs-med.htm; www.canoninstrument.com). A possible solution for a standard in RBC rheology measurements may be found in the use of frozen and thawed RBC and plasma samples [64]. 6.5. Normal Values Blood or RBC suspension viscosity values are really only useful when given in combination with the hematocrit, temperature and shear rate at which they were measured. Two approaches to the “correct” hematocrit for viscosity studies exist in the literature: 1) Use of native, as-drawn hematocrit provides data relevant to in vivo conditions for the donor at the time of venipuncture. However, even if mathematically corrected to a standard hematocrit [53], one can not infer with certainty changes of

262


RBC aggregation tendency or deformability; 2) Correcting hematocrit to a standard level (e.g., 40 or 45%) by appropriate combination of RBC and suspending medium (e.g., plasma, polymer solutions). This approach provides little or no information about the in vivo conditions for the donor at the time of venipuncture, but does allow donorto-donor and temporal studies. Clearly these two approaches are not mutually exclusive and both are used when other means for assessing RBC aggregation and deformability are not available. Measurements should be conducted over a shear rate range appropriate for the needs of a study and within the useful range of the viscometer; a wide range of shear rates is preferred (e.g., about 1 s-1 to 1,000 s-1) to allow studies under aggregating and non-aggregating conditions. Clinically-related studies are usually carried out at 37 qC whereas basic rheological measurements tend to use either 37 qC or a stated room temperature. Viscosity-shear rate data for blood from normal adult donors are presented below: Table 1 Viscosity values for norman blood* SHEAR RATE, s-1 VISCOSITY, mPa.s 0.05 60.1 ± 10.4 0.50 25.7 ± 3.05 70 5.14 ± 0.31 *Data are mean ± SD for normal blood from 15 subjects adjusted to a hematocrit of 45% and measured at 37 qC [65]. Table 2 Viscosity values for normal blood at three hematocrits* VISCOSITY, mPa.s VISCOSITY, mPa.s VISCOSITY, mPa.s (Hematocrit 28%) (Hematocrit 41%) (Hematocrit 56 %) 1.3 7.15 ± 1.1 18.6 ± 3.9 39.7 ± 4.2 3.2 5.30 ± 0.6 12.0 ± 2.1 23.5 ± 2.2 8.1 4.29 ± 0.4 8.57 ± 1.3 15.6 ± 1.4 15 3.86 ± 0.3 7.22 ± 1.0 12.5 ± 1.1 51 3.36 ± 0.2 5.65 ± 0.7 9.07 ± 0.8 150 3.12 ± 0.2 4.46 ± 0.5 6.86 ± 0.7 750 2.67 ± 0.2 3.63 ± 0.4 5.36 ± 0.7 1,500 2.57 ± 0.2 3.44 ± 0.4 5.03 ± 0.7 *Data are mean ± SD for normal blood at reduced (28%, n=8), normal (41%, n=44) and elevated (56%, n=8) hematocrits measured at 37 qC [14]. SHEAR RATE, s-1

Rosenson, et al. [66] have presented values for the Newtonian viscosity of plasma and serum from normal subjects (mean ± SD, n=126, 37 qC) as follows: Plasma Viscosity (mPa.s) 1.39 ± 0.08 Serum Viscosity (mPa.s) 1.27 ± 0.06

7. Contemplation: Clinical Relevance of In Vitro Blood Viscosity Measurements The rheological behavior of blood is determined by intrinsic factors (e.g., hematocrit, plasma viscosity, RBC deformability and RBC aggregation) which, in turn, can be influenced by various extrinsic factors (e.g., shear rate, osmolality, pH, etc.). Since both intrinsic and extrinsic factors may vary in different parts of the body, the flow and composition of blood is not constant throughout the body. For example, hemorheological parameters have been found to differ in arterial versus venous blood in patients [67 - 69], in healthy subjects [70], in an animal model [71] and within an organ [72]. Even within a blood vessel the suspension is not homogenous since RBC


263

are concentrated in the center while plasma and platelets are found mainly in the near vessel wall region (Fåhraeus-Lindquist effect, see chapters II.1 and II.3.a). Hemorheological measurements are usually performed on blood obtained from the antecubital vein in the forearm, and it is thus likely that local reversible changes in blood viscosity or its determinants are missed by studying blood taken from this location. Examples of such local factors are hematocrit which is lower in the microcirculation compared to large vessels, the hyperosmotic challenge to RBC in the renal micro circulation, and extreme low pH-values in certain parts of the spleen leading to sequestration of old and/or rigid RBC. Given that hemorheological measurements performed on blood from the antecubital vein may not exactly reflect local circulatory conditions, and that some viscometers may not be sensitive to a sub-population of poorly deformable cells [73], one could question the clinical or physiological relevance of blood rheology studies. While it is not yet possible to provide a single answer to this question, it would appear that both macro-rheological data (e.g., blood viscosity) and micro-rheological data (e.g., RBC deformability) are of value for the understanding of blood flow in the circulation: 1. Measurements of viscosity-shear rate relations for blood and other RBC suspensions provides useful information relevant to blood flow in larger vessels and geometries. Further, if the hematocrit in a micro-vessel is known, it is possible to calculate pressure-flow relations for small vessels based upon viscometric data (see chapter II 3.a). As described above, there are empirical methods for correcting blood viscosity at a given hematocrit to the expected value at another hematocrit [53]. Further, Newtonian calibration fluids having various viscosity values are available, thus allowing standardization between laboratories and over long periods of time. Nevertheless, it is possible that measuring some determinants of blood viscosity (e.g., hematocrit, plasma viscosity and RBC indices) may be sufficient for many clinical applications [73]. 2. Micro-rheological parameters such as RBC deformability and RBC aggregation indices provide greater insight into factors affecting both blood viscosity and microvascular blood flow. There are now several literature reports related to relations between micro-rheological parameters and physiological findings, including the effects of RBC aggregation and rigidity on vascular resistance, vascular responsiveness, and hematocrit distribution in tissues and organs (see chapters III.2 and III.3). In some situations (e.g., malaria tropica, sickle cell disease or other hemoglobinopathies) it is useful to obtain information on the distribution of cellular deformability within a given RBC population and hence the automated cell analyzer ARCA (see 5. New Developments) would be of value. Note that unlike viscometers, calibration and standardization of micro-rheological instruments is still problematic, in that there are not yet well-defined cell suspensions with known degrees of rigidity or aggregation. 3. Lastly, it is important to note that for both macro- and micro-rheological types of measurements, the normally functioning spleen removes abnormal and/or very rigid cells from the circulation [74]. As a consequence, grossly rigid RBC are not found in the circulation. Thus, for example, even with a device designed to provide the distribution of cell rigidities within a population, the RBC being studied are only those capable of remaining in the circulation: to properly study circulating cells

264


it is therefore necessary to have a measuring system with good reproducibility and with a sensitivity that exceeds that of the normal spleen [75].

References [1]

A. van Leeuwenhoek, Microscopical observations concerning blood, milk, bones, the brain, spittle and cuticula, Philosoph. Trans. Roy. Soc. London 9 (1674), 121 – 128. [2] A. van Leeuwenhoek, Concerning the circulation and stagnation of the blood in tadpoles, Philosoph. Trans. Roy. Soc. London 22 (1702), 447 – 455. [3] R. Fahraeus and T. Lindqvist, The viscosity of blood in narrow capillary tubes, Am. J. Physiol. 96 (1931), 562 – 568. [4] International Committee for Standardization in Hematology, Guidelines for measurement of blood viscosity and erythrocyte deformability, Clin. Hemorheol. 6 (1986), 439 – 453. [5] H. Barbee, The effect of temperature on the relative viscosity of human blood, Biorheology. 10 (1973), 1-5. [6] International Committee for Standardization in Haematology, Recommendations for a selected method for the measurement of plasma viscosity, J. Clin. Pathol. 37 (1984), 1147–1152. [7] J.F. Stoltz, S. Gaillard and A. Larcan, Evaluation expérimentale en hémorhéologie d’un viscosimètre a faibles vitesses de cisaillement, J. Mal. Vasc. 4 (1979), 145-149. [8] E. Ernst and A. Matrai, Rotationsviskometer in der Hämorheologie, Ärztl. Lab. 28 (1982), 33–38. [9] A. Matrai, R.B. Whittington and R. Skalak, Biophysics, In: Clinical Hemorheology, S. Chien et al, Eds., Martinus Nijhoff Publishers, Dordrecht, Boston, 1987, pp. 9-71. [10] R.E. Wells, R. Denton and E. Merrill, Measurement of viscosity of biological fluids by cone plate viscometer, J. Lab. Clin. Med. 57 (1961), 646-656. [11] S. Wang, A.H. Boss, K R. Kensey and R S. Rosenson, Variations of whole blood viscosity using Rheolog™ a new scanning capillary viscometer, Clin. Chim. Acta 332 (2003), 79 – 82. [12] R. Rosencranz, S.A. Bogen. Clinical laboratory measurement of serum, plasma and blood viscosity. Am. J. Clin. Pathol 125 (2006), S78-S86. [13] R.S. Rosenson and D. Wolff,. Comparison of two whole blood viscometers: implications for cardiovascular and thrombosis research, Thromb. Haemost. 92 (2004), 225-226. [14 ] T. Alexy, R.B. Wenby, E. Pais, L.J. Goldstein and W. Hogenauer and H.J. Meiselman, An automated tube-type blood viscometer: Validation studies, Biorheology 42 (2005), 237-247. [15] E. Evans, N. Mohandas and A. Leung, Static and dynamic rigidities of normal and sickled erythrocytes, J Clin. Invest. 73 (1984), 477-488. [16] G.R. Cokelet and H.J. Meiselman, Rheological comparison of hemoglobin solutions and erythrocyte suspensions, Science 162 (1968), 275-277. [17] M.W. Lande, R.L. Andrew, M.R. Clark, N.V. Braham, D.M. Black, S.H. Embury and W.C. Mentzer, The incidence of painful crises in homozygous sickle cell disease; correlation with red cell deformability, Blood 72 (1988), 2056 – 2059. [18] H.A Cranston, C.W. Boylan, G.L. Carroll, S.P. Sutera, J.R. Williamson, I.Y. Gluzman and D.J. Krogstad, Plasmodium Falciparum maturation abolishes physiological red cell deformability, Science 223 (1983), 400 – 402. [19] J.A. Chasis and N. Mohandas, Erythrocyte membrane deformability and stability: two distinct membrane properties that are independently regulated by skeletal protein associations, J. Cell. Biol. 103 (1986), 343 – 350. [20] H.L. Reid, A.J. Barnes, P.J. Lock, J.A. Dormandy and T.L. Dormandy, A simple method for measuring erythrocyte deformability, J. Clin. Pharmacol. 29 (1976), 855 – 858. [21] M. Paulitschke and G.B. Nash, Micropipette methods for analysing blood cell rheology and their application to clinical research, Clin. Hemorheol. 11 (1993), 407– 434. [22] P.J.H. Bronkhorst, E.J. Nijhof and J.J. Sixma, Parametrization of the deformation curve as a tool for standardization and interpretation of ektacytomeytric measurements, Clin. Hemorheol. 15 (1995), 803816. [23] G.J. Streekstra, J.J. Zwaginga, E.J. Nijhof, J.J. Sixma and R.M. Heethaar, The relation between blood viscosity and plasma viscosity in the nephrotic syndrome, Clin. Hemorheol. 14 (1994), 769 – 778. [24] O.K. Baskurt and H.J. Meiselman, Analyzing shear stress-elongation index curves: comparison of two approaches to simplify data presentation, Clin. Hemorheol. Microcirc. 31 (2004), 23 – 30.


265

[25] M.R. Hardeman, P.T. Goedhart, J.G.G. Dobbe and K.P. Lettinga, Rotational Cell Analyzer (L.O.R.C.A.); I. A new instrument for Laser-assisted Optical measurement of various structural hemorheological parameters, Clin. Hemorheol. 14 (1994), 605 – 618. [26] M.R. Hardeman, P.T. Goedhart and N.H. Schut, Laser-assisted Optical Rotational Cell Analyzer (L.O.R.C.A.); II. Red blood cell deformability: elongation index versus cell transit time, Clin. Hemorheol. 14 (1994), 619 – 630. [27] S.B. Zaets, T.L. Berezina, C. Morgan, M. Kamiyama, Z. Spolarics, D-Z Xu, E.A. Deitch and G.W. Machiedo, Effect of trauma-hemorrhagic shock on red blood cell deformability and shape, Shock 19 (2003), 268-273. [28] H. Schmid-Schönbein, P. Ruef and O. Linderkamp, The shear stress diffractometer Rheodyn SSD for determination of erythrocyte deformability. I. Principles of operation and reproducibility, Clin. Hemorheol. 16 (1996), 745 – 748. [29] S. Shin, Y.H. Ku, M.S. Park, S.Y. Moon, J.H. Jang and J.S. Suh, Laser-diffraction slit rheometer to measure red blood cell deformability, Rev. Sci. Inst. 75 (2004), 559 – 561. [30] M.R. Hardeman, R.M. Bauersachs and H.J. Meiselman, RBC laser diffractometry and RBC aggregometry with a rotational viscometer: comparison with rheoscope and myrenne aggregometer,. Clin. Hemorheol. 8 (1988), 581-593. [31] X. Wang, H. Zhao, F.Y. Zhuang and J.F. Stoltz, Measurement of erythrocyte deformability by two laser diffraction methods, Clin. Hemorheol. Microcirc. 21 (1999), 291 - 295. [32] O.K. Baskurt and F. Mat, Importance of measurement temperature in detecting the alterations of red blood cell aggregation and deformability studied by ektacytometry: A study on experimental sepsis in rats, Clin. Hemorheol. Microcirc. 23 (2000), 34 – 49. [33] M. Piagnerelli, K.Z. Boudjeltia, D. Brohee, P. Piro, E. Carlier, J.L. Vincent, P. Lejeune and M. Vanhaeverbeek, Alterations of red blood cell shape and sialic acid membrane content in septic patients, Crit. Care Med. 31 (2003), 2156-2162. [34] O.K. Baskurt and H.J. Meiselman, Determination of red blood cell shape recovery time constant in a Couette system by analysis of light reflectance and ektacytometry, Biorheology 32 (1996), 489-503. [35] J.G.G. Dobbe, G.J. Streekstra, J. Strackee, M.C.M. Rutten, J.M.A. Stijnen and C.A. Grimbergen, Syllectometry: Effect of aggregometer geometry in the assessment of red blood-cell shape recovery and aggregation, IEEE Trans. Biomed. Eng. 50 (2003), 97-106. [36] H. Schmid-Schönbein, Erythrocyte rheology and the optimization of mass transport in the microcirculation, Blood Cells 1 (1975), 285 – 306. [37] T. Fisher and H. Schmid-Schönbein, Tank tread motion of red cell membranes in viscometric flow, In: Red Cell Rheology, M. Bessis, S.B. Shohet, N. Mohandas, Eds., Springer Verlag, Berlin, Heidelberg, New York, 1978, p. 347 -358. [38] W.G. Zijlstra, Syllectometry, a new method for studying rouleaux formation of red blood cells, Acta Phys. Pharm. Neerl. 7 (1958), 153 -154. [39] H.R. Arntz, L. Hennig and H.J. von Kaiserlingk, Die Wertigkeit haemorheologischer Parameter beim Plasmozytom, In: Haemorheologie und operative Medizin, L. Heilmann, A.M. Ehrly, Münchner Wissenschaftl Publ., Munich,1988, pp. 439 – 445. [40] T. Böhler and O. Linderkamp, Effects of neuraminidase and trypsin on surface charge and aggregation of red blood cells, Clin. Hemorheol. 13 (1993), 775 – 778. [41] H. Schmid-Schönbein, Letter to the editors-in chief, Clin. Hemorheol. 13 (1993), 803 – 805. [42] F. Nguyen, L. Drouet, M. Boisseau, P. Leger and H. Juchet, Erythrocyte hype-raggregation and thrombogenic dysfibrinogenemia, Clin. Hemorheol. Microcirc. 18 (1998), 235 – 243. [43] O.K.Baskurt and H.J. Meiselman, Cellular determinants of low shear blood viscosity, Biorheology 34 (1997), 235 – 247. [44] G.Cloutier and Z. Qin, Ultrasound backscattering from non-aggregating and aggregating erythrpocytes – a review, Biorheology 34 (1997), 443 – 470. [45] H.Bot, W.E.M. Kok, M.R.Hardeman and P.T.Goedhart, Comparison of high frequency ultrasound and laser backscatter during erythrocyte aggregation, Biorheology 27 (1995), 239. [46] T.C. Fisher and H.J. Meiselman, personnel communication, Department of Physiology and Biophysics, Keck School of Medicine, Los Angeles, CA90033, USA. [47] H. Zhao, X.Wang and J.F. Stoltz, Comparison of three optical methods to study erythrocyte aggregation, Clin. Hemorheol. Microcirc. 21 (1999), 297 – 302. [48] Zs. Marton, G. Kesmarky, J. Vekasi, A.Cser, R. Russai, B. Horvath and K.Toth, Red blood cell aggregation measurements in whole blood and in fibrinogen soluitions by different methods, Clin. Hemorheol. Microcirc. 24 (2001), 75-83. [49] J.R. Murphy, Influence of temperature and method of centrifugation on the separation of erythrocytes, J. Lab. Clin. Med. 82 (1973), 334-341.

266


[50] R. Wenby, personnel communication, Department of Physiology and Biophysics, Keck School of Medicine, Los Angeles, CA90033, USA. [51] R. Yip, N. Mohandas, M.R. Clark, S. Jain, S.B. Shohet and P. Dallman, Red cell membrane stiffness in iron deficiency, Blood 62 (1983), 99 – 106. [52] B.P. Heath, N. Mohandas, J.L. Wyatt and S.B. Shohet, Deformability of isolated red blood cell membranes, Biochim. Biophys. Acta 691 (1982), 211 – 219. [53] A.Matrai, R.B. Whittington and E. Ernst, A simple method of estimating whole blood viscosity at standardized hematocrit, Clin. Hemorheol. 7 (1987), 261-265. [54] J.G.G. Dobbe, G.J. Streekstra, M.R. Hardeman, C. Ince and C.A. Grimbergen, The measurement of the distribution of red blood-cell deformability using an automated rheoscope, Clin. Cytometry 50 (2002), 313 – 325. [55] M. Bessis and N. Mohandas, Laser diffraction patterns of sickle cells in fluid shear fields, In: Red Cell Rheology, M. Bessis, S.B. Shohet, N. Mohandas, Eds., Springer Verlag, Berlin, Heidelberg, New York, 1978, p. 225 – 232. [56] J.G.G. Dobbe, M.R. Hardeman, G.J. Streekstra, J. Strackee, C. Ince and C.A. Grimbergen, Analyzing red blood-cell deformability distributions, Blood Cells Mol. Dis. 28 (2002), 373 – 384. [57] M. R. Hardeman and P. Klaver, Rheological and hematological aspects of red blood cells (RBC) from Lama Vicugna: similarity to human sickle cells, Clin. Hemorheol. 13 (1993), 319. [58] M. Mark, K. Häusler, J. Dual and W.H. Reinhart, Oscillating viscometer – Evaluation of a new bedside test, Biorheology 43 (2006), 133-146. [59] S. Shin, Y. Ku, M-S Park and J-S Suh, Measurement of blood viscosity using a pressure-scanning capillary viscometer, Clin. Hemorheol. Microcirc 30 (2004) 467-470. [60] J.P. Barras, Blood rheology – General review, Bibliot. Haematol. 33 (1969), 277-297. [61] W. Zingg , J.C. Sulev and C.D. Morgan, Study of possible sources of error in clinical blood viscosity determinations with the Wells-Brookfield viscometer. Effects of fasting, food intake and sample handling, Biorheology 10 (1973), 509-515. [62] W.I. Rosenblum and E.W. Warren, Elevation of blood viscosity produced by shearing in a rotational viscometer and its inhibition by refrigeration, Biorheology 10 (1973,) 43-49. [63] J. Zhang, X. Zhang, N. Wang, Y. Fan, H. Ju, J. Yang, J. Wen and X. Qu, What is the maximum duration to perform the hemorheological measurement for the human and mammals, Clin. Hemorheol. Microcirc. 31 (2004), 157-160. [64] M.R. Hardeman and J.W.M. Lagerberg, Rheology of 10 year-old frozen and thawed blood: Possible use of frozen RBC as standard reference material in hemorheology, Biorheology 39 (2002), 650-651. [65] M.R. Hardeman, unpublished observations. [66] R.S. Rosenson, A. McCormick and E.F. Uretz, Distribution of blood viscosity values and biochemical correlates in healthy adults, Clin. Chem. 42 (1996), 1189-1195. [67] S. Forconi, M. Guerrini, P. Ravelli, C. Rossi, C. Ferrozzi, S. Pecchi and G. Biasi, Arterial and venous blood viscosity in ischemic lower limbsin patients affected by peripheral obliterative arterial disease, J. Cardiovasc. Surg. 20 (1979), 379-384. [68] H. Tsuchida, H. Yamaguchi, S. Ischimaru and K. Furukawa, Local changes in red cell deformability in peripheral arterial disease, In: Microcirculation, an update, vol. 1, M. Tsuchiya, Ed., Elsevier Science Publisher, New York, 1987, 491-492. [69] F.Ch. Mokken, F.J.M. van der Waart, Ch.P. Henny, P.T. Goedhart and A.W. Gelb, Differences in peripheral arterial and venous hemorheological parameters, Ann. Hematol. 73 (1996), 135-137. [70] L.N.W. Daae, S. Halvorsen, P.M. Mathisen and K.A. Mironska, A comparison between haematological parameters in “capillary” and venous blood from healthy adults, Scand. J. Clin. Lab. Invest. 48 (1988), 723-726. [71] H.H. Lipowski, S. Usami and S. Chien, In vivo measurement of “apparent viscosity” and microvessel hematocrit in the mesentery of the cat, Microvasc. Res. 19 (1980), 297-319. [72] J.R. Pappenheimer and W.B. Kinter, Hematocrit ratio of blood within mammalian kidney and its significance for renal hemodynamics, Am. J. Physiol. 185 (1956), 377-390. [73] N.H. Schut, M.R. Hardeman, P.T. Goedhart, H.J.G. Bilo and J.M. Wilmink, Blood viscosity measurements are not sensitive enough to detect changes in erythrocyte deformability in cyclosporintreated patients and its subsequent reversal with fish and corn oil, Clin. Hemorheol. 13 (1993), 465-472. [74] O.K. Baskurt, The role of the spleen in suppressing the rheological alterations in circulating blood, Clin. Hemorheol. Microcirc. 20 (1999), 181 – 188. [75] M.R. Hardeman and C. Ince, Clinical potential of in vitro measured red cell deformability, a myth? Clin. Hemorheol. Microcirc. 21 (1999), 277 – 284.


267

Comparative Hemorheology Ursula WINDBERGERa,1 and Oguz K. BASKURTb Decentralized Biomedical Facilities, Core Unit for Biomedical Research, Medical University of Vienna, Austria and bDepartment of Physiology, Akdeniz University Faculty of Medicine, Antalya, Turkey

a

Introduction Hemorheological values vary widely among the animal species. To understand the structure-function relationships of red blood cells (RBC) together with associated physiological mechanisms, comparative studies are still a classical approach. Some animal species show extraordinary values of blood viscosity and RBC indices which would be pathologic for man. These differences can reflect an adaptation process to a specific environment or way of life. Different species also use different mechanisms to maintain blood flow, and in some cases these differences might indicate which variables limit the demand for oxygen delivery in a species or under certain circumstances. Nevertheless, when comparing values to man, it should be kept in mind that every hemorheologic “disturbance” in a healthy animal reflects a physiologic circumstance. It is necessary to know that the hemorheological profile of an animal cannot be judged by a single rheological value, but must be considered as part of the cardiovascular system in which the blood is flowing. That is, the hemorheological profile of an animal species is a conglomerate of properties which has to be considered in evaluating its cardiovascular relevance in a species-specific manner. These speciesspecific differences look chaotic at first sight, and no clear guiding principle has been found to combine these profiles to a universal “logic” at present. Therefore changes in parameters during disease, after interventions, or during environmental or associated changes need to be related to their species-specific reference values; describing an animal’s clinical or physiological condition using comparisons to values from other species is most likely not a valid approach.

1. General Aspects RBC of the different classes of vertebrates have different sizes and shapes, variations which are sometimes huge due to the nucleus within non-mammalian RBC. This variation and the diversities of RBC cytoplasm and membrane constituents, together with variations in blood cell count and blood plasma composition between the species markedly influences bulk blood flow in different parts of the vasculature. A list of

1 Corresponding author: Decentralized Biomedical Facilities, Medical University of Vienna, Borschkegasse 8a, 1090 Vienna, Austria; E mail: [email protected]

268

U. Windberger and O.K. Baskurt / Comparative Hemorheology

hematological parameters is given by Hawkey et al. [1] or can be found in textbooks [2, 3]. The first group of parameters which have to be considered in hemorheology are cell size and geometry. RBC size differs widely among the species – and to a lesser extent also between individuals within a species - with fish having the highest and mammals the lowest values although exceptions exist. For example, fish RBC volume can be more than 100-fold higher than those of a mammal [4]. Diversity in mammalian RBC volume alone is much smaller, ranging from a mean value of 18 fl in goat to 160 fl in giant anteater [5]. The size of the mammal has no association with its RBC volume; large species like horse may have smaller RBC than smaller species like rabbit. Mean cell volume (MCV) correlates strongly and inversely with RBC count [1, 6]. Besides hydrodynamic forces and RBC membrane mechanical properties, RBC size is most important in determining the effectiveness of tissue perfusion since it is easier for RBC to enter and transit capillaries if the cell size matches the vessel caliber and if vessels are more or less straight. Capillary geometry differs between various tissues (e.g., brain vs. kidney or skeletal muscle), with capillary density also affected by tissue metabolic demand which is measurable by capillary density per mitochondrial number. In case of skeletal muscle for instance, capillary geometry depends on the activity of the animal or on the environment the animal is adapted to [7]. Differences in rheological behavior of blood may also help to optimize O2 supply to tissues in various species [8-11]. Mean capillary diameters in species reach values from 3 μm in pigeon and rat, up to 20 μm in frog with tuna, hummingbird, bat, and rabbit lying in between [12-14]. In a small number of species, capillaries have been found to be smaller than 3 μm [15], but in most animals corrosion cast measurements show capillary diameter values below 5 μm [15-17], which is indeed smaller than the diameter of most animal RBC. As part of an in vivo study of capillary diameter and RBC geometry, frog RBC were found to travel through the capillaries with their major axis predominantly parallel to the flow direction [13]. In the presence of a greater mismatch of capillary caliber and RBC diameter as in fish, in vivo capillary transit time might be increased as suggested by results of in vitro filtration tests [4, 18]. However, an increased capillary transit time should enhance the time for O2 delivery, which might be an advantage in animals adapted to cold temperature since it would compensate for the left shift of the O2 dissociation curve. On the other hand, mammalian RBC are much smaller, which should facilitate O2 diffusion across the RBC membrane by providing a greater membrane surface per unit volume of blood; the smaller size also reduces the energy required for RBC to enter and pass the capillaries. Due to the inverse correlation of RBC size and number, a greater amount of small RBC may be associated with the increased metabolic capacity in animals with constant body temperature. A second important parameter for RBC function is cytoplasmic viscosity, reflected by mean cellular hemoglobin concentration (MCHC). MCHC is different among the species. Very high MCHC has been observed in seals, possibly to optimize O2 storage during long-term dives [19]. The cost of this is an increased cytoplasmic viscosity which may affect RBC deformability, although the RBC function of carrying O2binding proteins to guarantee adequate O2 supply seems to be more important under such circumstances. MCHC is increased by about 20% in birds compared to mammals due to the presence of a nucleus which displaces part of the cytoplasm [1, 20]. Besides size and cytoplasmic hemoglobin concentration, RBC also differ in specific structural characteristics. Only mammals have RBC without a nucleus.


269

Additionally, in animals other than mammals, a complex tubulin structure is present which anchors to the plasma membrane and affects the specific shape of RBC [21-26]. Both nucleus and tubulin filaments increase cell stability against shear forces and lead to a specific orientation of RBC in the flow field. In contrast, mammals including man, have biconcave disc shaped RBC, and shear forces can induce greater modifications in RBC shape (e.g., folding) than in other animals. Differences in the extent of shape changes during various levels of shear stresses do exist among mammals, allowing reduction of whole blood viscosity through this mechanism. As an exception, camelids have a more intermediate status among the mammalian species. They do not have a nucleus, however their RBC are spindle shaped and some authors report the presence of a marginal band [27]. Such basic morphologic diversity between mammals and other zoological classes are among the reasons for the pronounced differences found in RBC deformability and aggregation. These principal species differences influence both single RBC behavior and bulk blood flow. However, the prediction of blood flow phenomena in a living animal using in vitro viscometry or measurements of single cell properties is difficult because of the impact of the vascular function and geometry on the organization or stabilization of flow [14].

2. Specific Aspects 2.1. Mammals Mammals are the most widely investigated zoological class due to their agricultural and experimental use. Many mammals have been tested. However a list of hemorheological values may never be complete, and although relevant differences might exist even among relatives, common characteristics include the following principles: mammalian RBC are non-nucleated hemoglobin-carrying biconcave disks of various sizes; hematocrit varies from 30 to 50%; RBC aggregation and deformability varies strongly among mammals. 2.1.1. Plasma and Whole Blood Viscosity Plasma viscosity (PV) depends primarily on plasma protein concentration, indicating that PV can vary in disease [28-30]. The highest PV within a group of 9 species was found in cattle (mean: 1.72 mPa.s) and the lowest in rabbit (mean: 1.30 mPa.s) with horse, cat, dog, pig, rat, mouse, and sheep being between 1.3 and 1.7 mPa.s. These values are higher than those obtained for man by the same investigator [31]. Although PV contributes to the value of whole blood viscosity (WBV), PV and WBV values do not correlate between species. A species with low physiologic PV, compared to a “standard” of human values, can have high values of WBV. For example, horse has an increased WBV, however, its PV is less than cattle due to the high plasma fibrinogen concentration in the latter species. This high PV in cattle, however, is accompanied by a low WBV at low shear rates [31]. For further description of plasma protein concentration and PV in animal species see Chapter II.2.

270

U. Windberger and O.K. Baskurt / Comparative Hemorheology Table 1. References for hemorheological values in healthy animals Mammalian Species

Non-Mammalian

Reference

Species

Reference

Birds Horse

[31, 32, 37-50]

Chicken

[51-54]

Rat

[6, 31, 39, 47, 55-60]

Turkey

[4, 60, 61]

Cow

[31, 32, 44, 45, 47-50, 59]

Pigeon

[62]

Sheep

[4, 31, 32, 34, 47, 49, 50, 55, 63, 64]

Penguin

[65, 66]

Goat

[4, 34, 45, 47, 64]

Antarctic birds

[66]

Rabbit

[6, 31, 32, 45, 55-58, 60]

Reptiles

Pig

[31, 36, 47, 49, 50, 57, 63]

Turtle

[4, 60, 62, 67-69]

Mouse

[6, 31, 47, 55]

Snake

[60]

Hamster

[6, 47, 57]

Lizard,Crocodile

[70, 71]

Guinea pig

[6, 55, 56]

Amphibians

Gerbil

[58]

Bull frog

[4, 60, 67, 72-75]

Dog

[4, 6, 31, 34, 47, 55, 58, 63, 64, 76]

Toad

[77]

Cat

[31, 57, 58]

Fish

Elephant

[4, 32, 34, 47, 64, 78]

Trout

[18, 79-83]

Camel, Llama

[32, 47, 60]

Carp

[84, 85]

Seal, Whale

[19, 35, 36, 47, 66, 86-89]

Arctic, Antarctic fishes

[90-92]

Zoo Animals

[32, 45, 47]

Congo eel

[4, 11, 60]

Primates

[32, 47, 93]

Other fishes

[94, 95]

271

U. Windberger and O.K. Baskurt / Comparative Hemorheology Table 2: Whole blood viscosity (mean and median values) of various mammalian species Low Shear Viscosity (mPa.s)

High Shear Viscosity (mPa.s)

Low/High Shear Rate (s-1)

Hct (%)

T1 (°C)

Device

Ref.

Cat (median)

30.2

4.4

0.7 / 94

40

37

LS302

[31]

Cattle (mean)

6.7

5.8

1 / 150

30

20

OCR-D3

[50]

Cattle (mean)

5.4

3.6

0.277 / 128

35

37

LS30

[32]

Cattle (median)

6.6

4.8

0.7 / 94

40

37

LS30

[31]

Deer (mean)

45

4.7

0.277 / 128

42

37

LS30

[32]

Dog (mean)

10.4

5.7

1 / 150

37

20

OCR-D

[50]

Dog (median)

62

5.3

0.277 / 128

50

37

LS30

[76]

Dog (median)

22.9

5.6

0.7 / 94

40

37

LS30

[31]

28.3

5.4

0.7 / 94

37

37

LS30

[78]

53

5.6

0.277 / 128

43

37

LS30

[32]

Horse (mean)

93

4.4

0.277 / 128

41

37

LS30

[32]

Horse (mean)

12.8

6.5

1 / 150

44

20

OCR-D

[50]

Horse (median)

38.2

5.2

0.7 / 94

40

37

LS30

[31]

Horse (mean)

8.3

4.0

5.75 / 230

40

37

Brookfield DV-II4

[37]

Goat (mean)

5.1

3.4

0.277 / 128

31

37

LS30

[32]

Mouse (median)

13.4

4.9

0.7 / 94

40

37

LS30

[31]

Pig (mean)

14.3

7.2

1 / 150

45

20

OCR-D

[50]

Pig (median)

24.7

4.9

0.7 / 94

40

37

LS30

[31]

Rabbit (mean)

12.4

3.8

0.277 / 128

40

37

LS30

[32]

Rabbit (median)

8.3

4.0

0.7 / 94

40

37

LS30

[31]

Rat (median)

35.4

6.3

0.7 / 94

40

37

LS30

[31]

Sheep (mean)

4.4

3.9

1 / 150

31

20

OCR-D

[50]

Sheep (mean)

5.0

3.3

0.277 / 128

36

37

LS30

[32]

Sheep (median)

6.6

4.4

0.7 / 94

40

37

LS30

[31]

Human (median)

33.5

6.0

0.7 / 94

40

37

LS30

[31]

Species

African elephant (median) Asian elephant (mean)

1

Measurement temperature LS-30 Couette viscometer (Contraves GmbH, Zurich, Switzerland) 3 Oscillating Capillary Rheometer and Densitometer OCR-D (Paar GmbH, Graz, Austria) 4 Brookfield DV-II cone-plate viscometer (Brookfield Engineering Laboratories Inc., Stoughton, MA) 2

272

U. Windberger and O.K. Baskurt / Comparative Hemorheology Table 3. Plasma viscosity (mean and median values) of various mammalian species

1

Plasma Viscosity (mPa.s)

Device

T1 (°C)

Reference

Cat (median)

1.7

OCR-D

21

[31]

Cattle (mean)

2.5

OCR-D

20

[50]

Cattle (mean)

1.9

Capillary2

20

[41]

Cattle (median)

1.7

OCR-D

21

[31]

Deer (mean)

1.3

LS30

37

[32]

Dog (single value)

1.1

Coaxial2

37

[4]

Dog (median)

1.6

OCR-D

21

[31]

Dog (mean)

1.6

OCR-D

25

[96]

African elephant (median)

1.9

OCR-D

21

[78]

Asian elephant (mean)

1.6

LS30

37

[32]

Horse (mean)

1.5

LS30

37

[32]

2

Horse (mean)

1.5

Capillary

20

[41]

Horse (median)

1.7

OCR-D

21

[31]

Horse (mean)

1.6

OCR-D

21

[97]

Goat (mean)

1.6

LS30

37

[32]

Mouse (median)

1.3

OCR-D

21

[31]

Pig (median)

1.6

OCR-D

21

[31]

Rabbit (mean)

1.1

LS30

37

[32]

Rabbit (median)

1.3

OCR-D

21

[31]

3

Rat (mean)

1.0

Harkness

37

[98]

Rat (median)

1.6

OCR-D

21

[31]

Sheep (mean)

1.8

OCR-D

20

[50]

Sheep (median)

1.5

OCR-D

21

[31]

Sheep (mean)

1.6

OCR-D

25

[96]

Human (median)

1.2

OCR-D

21

[31]

Measurement temperature Custom made device [41] 3 Harkness viscometer (Coulter Electronics Ltd., England, UK) 2


273

Viscometric data from a wide range of animals have been presented by Johnn, et al. [32], with other authors describing smaller numbers of species (see Table 1). Whole blood viscosity is strongly and directly related to hematocrit and is inversely related to temperature in species, but relations between species vary. For example, in some animals living in a cold environment (e.g., bowhead whales), WBV does not increase with decreasing temperature to the extent that would be expected for human blood [33]. WBV increases with hematocrit but the degree of viscosity increase with hematocrit is species specific, especially at low shear rates [34]. Apparent whole blood viscosity of mammalian blood shows shear thinning behavior, but the effect is low in animals with low aggregation such as goat, sheep, and cow. In vitro WBV at low shear rate (0.7 s-1) is in the range of 6-7 mPa.s (40% hematocrit, 37oC) in these species. In contrast, horse, donkey, and zebra show pronounced shear thinning and increased low shear blood viscosity due to their “hyper-aggregating” blood. Their WBV at low shear rate is in the range of 38-40 mPa.s [31, 32]. Other extreme diversities in whole blood viscosity include phocid seals [35, 36] (see below). Most other mammalian species have WBV between the values for horse and goat. A ranking from data presently known can be as follows: equines> felines> chimpanzee> elephant> rat> deer> bison> mouse> rabbit> cattle> sheep> goat> camel (For references see table 1). Please note that the absolute values can sometimes not be directly compared due to the different devices used for the measurement by different laboratories. Table 2 and 3 give insight into species differences of whole blood and plasma viscosity values in mammals. 2.1.2. RBC Deformability RBC deformability data from a wide range of animals have been presented by Smith, et al. [47]. However, due to the variability of RBC size and shape between species, values for the term “deformability” are difficult to compare because these size and shape influence the measurement outcome. Using small (§ 1 μm) micropipettes it is possible to measure membrane mechanical properties (e.g., shear elastic modulus, apparent viscosity), while larger (§ 3-5 μm) micropipettes allow aspiration of the entire cell and hence a measure of cellular deformability (e.g., entry time at a given aspiration pressure). It is of interest to note that membrane measurements made using small pipettes are insensitive to cell size, whereas cellular deformability measurements via complete cell aspiration depend strongly on the sizes of the micropipette and the cell. There are other methods to measure deformability, including cell elongation at various shear stresses (e.g., laser diffractometry) [47, 78, 96] or RBC capillary entry and transit time (e.g., automated filtration method) [55, 57, 64, 76]. For example, with laser diffractometry, RBC elongation rises with the deforming force and is characterized by a specific elongation index (EI) – shear stress relationship for each species; increased elongation at a given shear stress means increased deformability (Figure 1). Figure 1 demonstrates that five species (rabbit, mouse, hamster, rat and pig) have more deformable RBC than the other species included in this study. It is also clear that horse and elephant RBC are the least deformable in this group, with dog RBC having an intermediate degree of deformability. Sheep RBC exhibit unique behavior with significantly lower EI values at the higher shear stress range; note that at such high shear stress levels, elongation indexes appear to be reaching a maximum for most species. Such a behavior may reflect the exceptionally small size of sheep RBC compared to other species included in this study.

274


0.7 Elephant

0.6

Horse

0.5

Dog

EI

0.4

Rabbit

0.3

Mouse

0.2

Hamster

0.1

Rat Pig

0.0 1

-0.1

10

100

Sheep

SS (Pa)

Figure 1. RBC elongation indexs (EI) versus shear stress (SS) data obtained via ektacytometry for nine species. EI is calculated from the diffraction image as the (length-width)/(length+width) of the image.

Table 4. RBC elongation indexes (EI) measured at a shear stress of 5.38 Pa, calculated shear stress needed for one-half maximal deformation (SS1/2), and mean cell volume (MCV) values for nine species. Data should be regarded as preliminary since sample size varies among species. Species

EI at 5.38 Pa

SS1/2 (Pa)

MCV (fL)

Hamster

0.448

2.46

47

Mouse

0.447

1.91

46

Rabbit

0.447

1.81

57

Rat

0.446

2.82

47

Pig

0.419

3.52

51

Dog

0.381

5.90

63

Horse

0.279

12.7

42

Elephant

0.238

7.71

138

Sheep

0.293

3.68

32

Since RBC elongation index-shear stress curves actually represent several pairs of EI-stress data, (see Chapter II.6), it is difficult to provide a single index for deformability. However, various methods have been described to derive such a parameter by calculating the shear stress for half-maximal deformation (SS1/2) of the cell population [100]. SS1/2 values for the EI-shear stress curves shown in Figure 1 are presented in Table 4. It is also obvious from these values that, compared to the other species, elephant and horse RBC require higher shear stresses to reach half of their maximum deformation; this shear stress was 3- to 6-fold higher for elephant and horse RBC. RBC deformability of some animals (e.g., camel, llama, bank vole Myodes glareolus) cannot easily be measured using ektacytometry since RBC orientation is not


275

always aligned with the shear field. Rather, these cells tend to tumble and/or become perpendicular to the lines of flow (unpublished observation). The reason for these observations has not yet been investigated in detail, but most likely relates to the nonbiconcave shape of these cells. To investigate factors involved in RBC deformability, some membrane components have been tested in various species. Modifications of RBC membrane content or the function of band 4.2 and band 3 proteins have been associated with RBC shape changes [101, 102]. Percentages of band 4.2 protein have been shown in a large number of species by Guerra-Shinohara and de O’Barretto [103]. Band 4.2 protein was found to be absent in horse, guinea pig and two rodents of this study. The deficiency of band 4.2 in horse was associated with the high appearance of echinocytes in this species [39]. Band 3 has also been investigated in camelids and the rotational and lateral mobility of the membrane domain of the protein has been found to be decreased, and a tight and close connection of the cytoplasmic domain to ankyrin has been observed [104, 105]. Membrane phospholipids [63] and glycosphingolipids [106] differ among some species, but the role of dietary fatty acid intake on RBC membrane lipid content in this context is not conclusive. The quantity of unsaturated fatty acids in the RBC membrane of sheep has been observed to be twofold higher than in pig and horse [107]. This finding is surprising since sheep RBC have a lower elongation at high shear stress and a high concentration of unsaturated fatty acids of membrane phospholipids which should make the membrane more deformable. This discrepancy most likely reflects the several factors that affect RBC deformation in a shear field: membrane mechanical properties that are mainly determined by the membrane skeletal network and not by membrane lipid composition, surface area to cell volume ratio, and cytoplasmic viscosity. In another study in guinea pigs, RBC deformability remained unchanged in spite of a 50% decrease of membrane cholesterol after statin treatment [108]. Such studies indicate that differences in membrane lipid structures do not explain species specific differences in RBC deformability satisfactorily. A better insight would be provided by a systematic investigation of membrane skeletal proteins and their linkage to membrane proteins. 2.1.3. RBC Aggregation RBC aggregation depends on cell deformability since it requires the formation of parallel membrane surfaces; poorly deformable RBC such as those containing a nucleus do not aggregate and thus almost all studies have been conducted using mammalian cells [58]. Values of aggregation range from a “hyperaggregating” group of animals (e.g., equine, rhinoceros), to a medium type (e.g., primates, pig, carnivora, many ruminants, elephants, rabbit), and to a “no” aggregation type group (e.g., cow, sheep, goat, camelids, rodents). However, as expected, exceptions will be found in future (see Table 1). Phocid seals show very unusual aggregation and will be discussed later. Interestingly, domestic bovines have very low, essentially non-measurable RBC aggregation in contrast to some wildlife relatives [32]. The diverse aggregation of cow, sheep, and goat against horse, donkey and zebra has been known for many decades [2]; aggregation in horse is so intense that even echinocytes are included into rouleaux [40]. Reasons for this diverse aggregation behaviour include variations in RBC deformability and membrane surface properties [109-111] as well as quantitative and qualitative differences in the plasma protein

276


concentrations [37, 112]. A comparative study on cow, horse, and man recently showed a reduction of depletion layer thickness in cow, which could, according to the depletion model of aggregation (see II.4.b), explain the low aggregation of this species [41]. In comparing two species that reflect mammalian extremes of aggregation – horse and goat – one can postulate that in goat the distribution of RBC among a transverse section of the vessel should be more homogenous [113, 114]. Goat blood is more susceptible to turbulent flow due to the lower WBV which increases the Reynolds number [115]. This may be important for engineering studies where high Reynolds numbers are achieved, for instance during studies with mechanical devices for cardiac assist or for experimental vascular surgery. Formation of a marginal cell-free layer is decreased in non-aggregating blood and near-wall shear stress should be higher for this reason. However, one should be careful in applying results obtained in glass capillaries to in vivo conditions, since glass capillaries are over-simplified models of the circulatory system [116]. In contrast, in horse blood the parabolic velocity profile is expected to be blunted and the rate of axial migration is increased in vertical tubes. The readjustment of axial migration after junctions and branching should be quicker depending on the aggregation kinetics and tube geometrics; flow partitioning at bifurcations might be increased. There have been suggestions in the literature that athletic species are characterized by higher RBC aggregation [121]. This is true for certain species such as horse, dog, cat and antelope. However, although rodents are amongst the most active species, they have low RBC aggregation compared to larger athletic species. Again, one can speculate that this lower aggregation in rodents is consistent with shorter blood vessel lengths (i.e., body size) which prevent full development of RBC aggregates within a given vessel segment, thereby countering any hemodynamic advantage expected from RBC aggregation-related phenomena. In contrast, it is not possible to understand the physiological meaning of the absence of RBC aggregation in domestic bovine. The animal kingdom is full of extreme variability in RBC aggregation, even within a given species. For example, ringed seals blood exhibits no measurable aggregation while Weddell seals have very high RBC aggregation [35]. It follows from the above discussion that neither athletic capacity nor body size variations may fully explain the variations in RBC aggregation among species. 2.1.4. Unusual Mammalian Species 2.1.4.1. Camelids Camel, dromedary, and llama have a low hematocrit resulting in a low whole blood viscosity [32], and their RBC aggregation is low (unpublished observation). Camelid RBC withstand high fluctuations of osmotic pressures during dehydration and rapid rehydration as a result of their adaptation to desert environments [122]. They also are resistant to metabolic depletion [104]. Camelid RBC are small, elliptical discs [47, 123, 124] that have an increased thickness in the central region in contrast to other mammals which show central depressions. Hemoglobin is homogenously distributed inside the cytoplasm with binding sites on band 3 protein, or, alternatively, is concentrated in crystals inside llama RBC [124]. Studies on camelid RBC membranes [125] in order to determine their principal properties have been carried out for some time. Key findings include: 1) RBC membranes of llama have increased amounts of sialic acid, and their glycoproteins differ qualitatively from human RBC membranes [104]; 2) Spectrin and


277

band 4.2 concentrations are equivalent to man [103], however, band 3 protein is increased threefold [104]; 3) The rotational and lateral mobility of band 3 is decreased even if parts of the membrane cytoskeleton are removed; 4) Band 3 protein has a tight and close connection to ankyrin and peripheral membrane proteins [105]; 5) Parts of a marginal band have been found [27]; 6) RBC deformability is different from other mammals due to structural reasons. However, further studies are still necessary to clarify the mechanism by which membrane rigidity is elevated in camel RBC. 2.1.4.2. Marine Mammals Most knowledge in the field of hemorheology of this group of species has been obtained from seals. During submergence, characterized by apneic exercise with increased environmental pressure and decreased temperature, the cardiovascular adjustment and redistribution of blood flow in seals involves increased peripheral resistance and reduced cardiac output [126], termed the dive response. As an adaptation to long-term apnea, resting hematocrit, MCHC, MCV, and total blood volume are increased [88, 127]. To overcome the body’s O2 demand during apnea, hematocrit increases further and can reach to values of 70% when the animal is diving. Although the spleen contributes to RBC storage effectively, time constants for hematocrit increase and splenic contraction favor the assistance of the hepatic sinuses for RBC delivery into the venous circulation [128]. The physiological polycythemia and the enhanced RBC hemoglobin concentration increase WBV and RBC density. The value of WBV increases with the hematocrit, however, the rise is lower than expected from terrestrial animals. O2 transport capacity is improved for this reason [35, 88, 89]. During a dive, shear forces are reduced which affects WBV remarkably if RBC aggregation is present. Indeed, RBC aggregation reaches high values in Weddell seals, sometimes exceeding even those obtained for horse [19]. This high aggregation develops shortly after birth of pups [35]. RBC aggregation varies extremely between the species and can be ranked as follows: ringed seal < elephant seal < Weddell seal, with ringed seals showing no detectable RBC aggregation or blood sedimentation. A potential benefit of a high RBC aggregation as in Weddell seal or of “no” RBC aggregation like in the ringed seal cannot be elucidated as yet and need further investigations. The role of the more unsaturated fatty acid content in RBC membrane phospholipids in elephant and fur seals compared to man might indicate some trend toward increased RBC deformability [129]. Blood viscosity and RBC aggregation were also investigated in bowhead whales. These marine mammals also have high RBC aggregation both in autologous plasma and standard aggregation media [35]. In contrast, WBV was found to be lower than human samples when the hematocrit was adjusted to 50% for both species [33]. Additionally, bowhead whale blood exhibited significantly less temperature dependence compared to human blood, with smaller increments in WBV as the measurement temperature decreased to 5 °C [33]. 2.2. Hemorheology of Other Vertebrates Non-mammalian RBC differ structurally from mammalian cells. They contain a nucleus and most authors found a microtubular bundle connected to a membraneassociated cytoskeleton called a marginal band [3]. RBC are larger, elliptical, resistant to bending, and blood cell count is decreased. Nucleated RBC are believed to exhibit no aggregation [4, 58] although viscometric data show some degree of shear-thinning

278


[62, 68, 69] which might be explained by some deformability and orientation of RBC under flow conditions. Studies with duck blood in glass tubes showed non-alignment between RBC and the axis of the tube [130]. This deviation was reduced at smaller tube diameters indicating better alignment with the streamlines. The authors concluded that the instability of cell orientation during blood flow may lead to an increase of viscous resistance. RBC membrane rigidity is considerably greater [18, 60, 99], although it is not clear which structures of the cell membrane are actually resisting deformation. Additionally, the nucleus is a relevant factor causing poor RBC deformability. Temperature of the environment and some forms of stress following adrenergic stimulation or hypoxia may decrease RBC deformability [3, 131]. The marginal band has been described in more detail in birds; it is found at the equator, near to the plasma membrane, in one plane only [21]. No microtubules were observed between the nucleus and the position of the marginal band in RBC of adult animals. The microtubules, whose main proteins are tubulins, curve with the profile of the cell. The number of microtubules in a RBC differ significantly among species, and has been found to vary with the size of the RBC [22]. Since polymerized actin colocalizes with the marginal band, an interaction between actin filaments and microtubules may exist [23, 24]. The response of intact marginal bands to bending, stretching and microtubule polarity has been tested in some amphibia [25, 26]. Blood viscosity in poikilotherm species is not a fixed value, but varies with the animal’s body temperature [67]. Therefore, the temperature dependence of whole blood and plasma viscosity, as well as other hemorheological values, are of interest in species adapted to low environmental temperatures. Since warming of such blood may damage red cells [18] by RBC swelling [83], comparative measurements should be made at a lower temperature. For a summary of literature containing hemorheologic data from non-mammalian vertebrates see table 1. 2.2.1. Birds Avian RBC are spherical, lenticular, and somewhat flattened to produce an equator. They are able to deform in a specific way by folding along their major axis [130] and to orient in a shear field. The nucleus contains approximately 20% of the cytoplasmic volume; the cytoplasmic hemoglobin concentration is about 15-20% greater [1, 132] when compared with mammalian RBC and hence there is an increase of RBC density [20]. Hematocrit is lower than most mammals and hence blood viscosity is reduced; the reduction in WBV with increasing shear rate is very low in chicken blood compared to mammals. During disease [51, 133], and during development [54], blood and plasma viscosity values change as a result of variations of the animal’s hematocrit and plasma protein concentration. In broilers, a diurnal variation of whole blood viscosity was observed, influenced by the variation of environmental temperature and food and water intake [134]. Deformability of avian RBC is decreased compared to mammals [52, 130, 132], and differences in RBC rigidity is likely to exist between avian species since a difference has been found between layer and broiler chickens [53]. Despite relevant functional differences between avian and mammalian RBC, an adequate tissue O2 supply has to be maintained in birds, since they have high metabolic rates which often exceed those of mammals.


279

2.2.2. Reptiles Reptiles are poikilotherm animals, except for some metabolic intermediate species; tissue perfusion varies therefore in association with changes in ambient temperature. Blood viscosity may have a special effect on heat exchange by modulating skin blood flow in reptiles [67]. Hematocrit and mean cellular hemoglobin (MCH) are lowered, possibly associated with a reduced metabolic need of this group [1]. Whole blood viscosity is therefore lower than in mammals. Viscosity values vary with temperature [67, 70], however, a drop in temperature does not elevate viscosity as much as in mammals [62]. WBV in the painted turtle (Chrysema picta) is a function of body temperature: 3.5 mPa.s at 10 oC and 1.5 mPa.s at 45 oC [67]. In other turtles, plasma viscosity was found to be 1.32 mPa.s in Mauremys leprosa [62]. Others found blood viscosity values of Trachemys scripta at 20% hematocrit and 5 oC of 9-12 mPa.s [68]. Blood viscosity is also decreased in Chelonia mydas hatchlings despite similar hematocrit values [69]. RBC deformability is decreased compared to mammals and values of RBC membrane shear elastic modulus are presented by Waugh [60]. Turtles are interesting species in view of their cold adaptation during hibernation. In Trachemys scripta, hematocrit is essentially maintained during cool room adaptation, leading to an increase in blood viscosity. However, hematocrit standardized values of blood viscosity were lower in cold adapted than in room adapted animals[68]. During hibernation of Sternotherus odoratus submerged in water at 5 oC for five months, hematocrit increased but WBV did not as might be expected based on the hematocrit increase [135], thus indicating a possible adaptation to lower temperature. 2.2.3. Amphibians Blood viscosity of frog and toad show hematocrit (r=0.93) and temperature (r=0.96) dependency as in other species [67, 75, 77], with the temperature dependency of blood viscosity in the bullfrog reported to resemble that in mammals [73]. Blood viscosity was reported to vary between 2.5 (toad at 27oC) [72] and 4 mPa.s (bullfrog at 5oC and 150 s-1) [74]. In vivo hematocrit varies with the environmental temperature, leading to a seasonal variation in blood viscosity. Experimental data on American bullfrog showed that hematocrit and RBC count increased when the temperature of the animals´ environment was decreased to 5oC, whereas MCV decreased under these circumstances. In this circumstance hematocrit and blood viscosity did not vary significantly, however, mean viscosity values, measured at native hematocrit, increased slightly [74]. The relevance of blood viscosity for blood flow has been investigated for toads: systemic blood flow decreased as WBV increased. [77] However, general anesthesia during such studies might modify regulatory processes that maintain blood flow. 2.2.4. Fish Fish RBC are elliptical cells which bulge in the region of their nucleus. The largest RBC volumes can be found in this class [4]. An excess of surface area is present which enables RBC to enter small capillaries of 3μm diameter. However, the shear elastic modulus was reported to be significantly higher for fish RBC and capillary entry time and transit time is increased significantly compared to man [18]. Hematocrit in fishes drops with water temperature, but it is believed that the O2 demand is satisfied due to the increased O2 concentration in cold water and the reduced metabolic need of the

280


animal in a cold environment. This hematocrit dependency on temperature should lead to diversities in blood viscosity of fishes living at different regions. Some studies focused on RBC deformability, and showed that membrane rigidity in trout was dependent on the temperature [18]. In temperature-acclimated trout, however, pipette entry time of RBC membranes has been shown to be maintained [92]. A special focus in fish hemorheology is the blood and plasma viscosity of ice fishes. These animals live at an environmental temperature down to -2 qC which should result in high viscosity values. However, some of these fish have blood with a very low cellular content which thus lowers viscosity. Antarctic ice fish show an increase of whole blood viscosity with a temperature drop from 5 to -1oC, with the thermal sensitivity of WBV in fish being reduced compared to man. Blood of white-blooded (i.e., almost no red blood cells) fish (Chaenocephalus kathleenae) showed nearly Newtonian behavior [92], and the plasma viscosity of C. aceratus was comparable to human plasma at low temperatures. Plasma viscosity of another Antarctic ice fish with low hematocrit (7.5%) and low MCHC, Notothenia sp., was significantly lower than the value of Chaenocephalids; however, its blood viscosity was increased due to the presence of blood cells [90].

References [1] [2] [3] [4] [5] [6]

[7] [8] [9]

[10] [11]

[12] [13] [14] [15]

C.M. Hawkey, P.M. Bennett, S.C. Gascoyne, M.G. Hart and J.K. Kirkwood, Erythrocyte size, number and haemoglobin content in vertebrates, Br. J. Haematol. 77 (1991), 392-397. B.F Feldman, J.G. Zinkl and N.C. Jain. Schalm's Veterinary Hematology. Lippincott, Hagerstown, MD, 2000. M.Nikinmaa. Vertebrate red blood cells. Springer, Berlin, 1990. S. Chien, S. Usami, R.J. Dellenback and C.A. Bryant, Comparative hemorheology - hematological implications of species differences in blood viscosity, Biorheology 8 (1971), 35-57. S.C. Gascoyne and C.M. Hawkey, Patterns of variation in vertebrate haematology, Clin. Hemorheol. 12 (1992), 627-637. L.N. Katyukhin, A.M. Kazenow, M.N. Maslove and Y.A. Matskevich, Rheologic properties of mammalian erythrocytes: relationship to transport ATPases, Comp. Biochem. Physiol. B 120 (1998), 493-498. S. Egginton, Temperature and angiogenesis: the possible role of mechanical factors in capillary growth, Comp. Biochem. Physiol. A 132 (2002), 773-787. G. Casotti and E.J. Braun, Structure of the glomerular capillaries of the domestic chicken and desert quail, J. Morphol. 224 (2005), 57-63. S. Cecon, B. Minnich and A. Lametschwandtner, Vascularisation of the brains of the Atlantic and Pacific hagfishes, Myxine glutinosa and Eptatretus stouti: a scanning electron microscopic study of vascular corrosion casts, J. Morphol. 253 (2002), 51-63. O. Mathieu-Costello, Morphometry of the size of the capillary-to-fiber interface in muscles, Adv. Exp. Med. Biol. 345 (1994), 661-668. S. Egginton and I.A. Johnston, An estimate of capillary anisotropy and determination of surface and volume densities of capillaries in skeletal muscles of the conger eel (Conger conger L), Q. J. Exp. Physiol. 68 (1983), 603-617. A.G. Koutsaris, Volume flow estimation in the precapillary mesenteric microvacsculature in vivo and the princple of constant pressure gradient, Biorheology 42 (2005), 479-491. R.G.A. Safranyos, C.G. Ellis, K. Tyml and A.C. Groom, Heterogeneity of capillary diameters in skeletal muscle of the frog, Microvasc. Res. 26 (1983), 151-156. J.H. Jeong, Y. Sugii, M. Minamiyama and K. Okamoto, Measurement of RBC deformation and velocity in capillaries in vivo, Microvasc. Res. 71 (2006), 212-217. R.F. Potter and A.C. Groom, Capillary diameter and geometry in cardiac and skeletal muscle studied by means of corrosion casts, Microvasc. Res. 25 (1983), 68-84.


281

[16] B. Stöttinger, M. Klein, B. Minnich and A. Lametschwandtner, Design of cerebellar and nontegmental rhombencephalic microvascular bed in the sterlet, Acipenser ruthenus: a scanning electron microscope and 3D morphometry study of vascular corrosion casts, Microsc. Microanal. 12 (2006), 1-14. [17] P.B. Canham, R.F. Potter and D. Woo, Geometric accomodation between the dimensions of erythrocytes and the calibre of heart and muscle capillaries in the rat, J. Physiol. 347 (1984), 697-712. [18] G.B. Nash and S. Egginton, Comparative rheology of human and trout red blood cells, J. Exp. Biol. 174 (1993), 109-122. [19] H.J. Meiselman, M.A. Castellini and R. Elsner, Hemorheological behavior of seal blood, Clin. Hemorheol. 12 (1992), 657-675. [20] D.P. Nguyen, K. Yamaguchi, P. Scheid and J. Piiper, Kinetics of oxygen uptake and release by red blood cells of chicken and duck, J. Exp. Biol. 125 (1986), 15-27. [21] O. Behnke, A comparative study of microtubules of disk shaped blood cells, J. Ultrastruct. Res. 31 (1970), 61-75. [22] L. Goniakowska-Witalinska and W. Witalinski, Evidence for a correlation between the number of marginal band microtubules and the size of vertebrate erythrocytes, J. Cell. Sci. 22 (1976), 397-401. [23] S. Kim, M. Magendantz, W. Katz and F. Solomon, Development of a differentiated microtubule structure: formation of the chicken erythrocyte marginal band in vivo, J. Cell. Biol. 104 (1987), 51-59. [24] E. Birgbauer and F. Solomon, A marginal band-associated protein has properties of both microtubule and microfilament associated proteins, J. Cell. Biol. 109 (1989), 1609-1620. [25] R.E. Waugh and G. Erwin, Flexural rigidity of marginal bands isolated from erythrocytes of the newt, J. Cell. Biol. 108 (1989), 1711-1716. [26] U. Euteneuer, H. Ris and G.G. Borisy, Polarity of marginal-band microtubules in vertebrate erythrocytes, Eur. J. Cell. Biol. 37 (1985), 149-155. [27] W.D. Cohen and N.B. Terwilliger, Marginal bands in camel erythrocytes, J. Cell. Sci. 36 (1979), 97107. [28] J. Linke, Untersuchungen zur plasmaviskositat beim hund, Tierarztl. Prax. 24 (1996), 524-528. [29] U. Windberger and A. Bartholovitsch, Hemorheology in spontaneous animal endocrinopathies, Clin. Hemorheol. Microcirc. 31 (2004), 207-215. [30] A.A. Kaymaz, S. Tamer, I. Albeniz, K. Cefle, S. Palanduz, S. Ozturk and N. Salmanyeli, Alterations in rheological properties and erythrocyte membrane proteins in cats with diabetes mellitus, Clin. Hemorheol. Microcirc. 33 (2005), 81-88. [31] U. Windberger, A. Bartholovitsch, R. Plasenzotti, K.J. Korak and G. Heinke, Whole blood viscosity, plasma viscosity and erythrocyte aggregation in nine mammalian species: reference values and comparison of data, Exp. Physiol. 88 (2003), 431-440. [32] H. Johnn, C. Phipps, S.C. Gascoyne, C. Hawkey and M.W. Rampling, A comparison of the viscometric properties of the blood from a wide range of mammals, Clin. Hemorheol. 12 (1992), 639-647. [33] R. Elsner, H.J. Meiselman and O.K. Baskurt, Temperature-viscosity relations of bowhead whale blood: a possible mechanism for maintaining cold blood flow, Marine Mammal Sci. 20 (2004), 339-344. [34] S. Usami, S. Chien and M.I. Gregersen, Viscometric characteristics of blood of the elephant, man, dog, sheep and goat, Am. J. Physiol. 217 (1969), 884-890. [35] M. Castellini, R. Elsner, O.K. Baskurt, R.B. Wenby and H.J. Meiselman, Blood rheology of Weddell seals and bowhead whales, Biorheology 43 (2006), 57-69. [36] L.L. Wickham, R.M. Bauersachs, R.B. Wenby, S. Sowemimo-Coker, H.J. Meiselman and R. Elsner, Red cell aggregation and viscoelasticity of blood from seals, swine and man, Biorheology 27 (1990), 191-204. [37] F.A. Andrews, N.L. Korenek, W.L. Sanders and R.L. Hamlin, Viscosity and rheologic properties of blood from clinically normal horses, Am. J. Vet. Res. 53 (1991), 966-971. [38] R.K. Archer and B. Allen, The viscosity of equine blood plasma: a new non-specific test, Vet. Record 86 (1970), 360-363. [39] O.K. Baskurt, R.A. Farley and H.J. Meiselman, Erythrocyte aggregation tendency and cellular properties in horse, human, and rat: a comparative study, Am. J. Physiol 273 (1997), H2604-H2612. [40] O.K. Baskurt and H.J. Meiselman, Susceptibility of equine erythrocytes to oxidant-induced rheologic alterations, Am. J. Vet. Res. 60 (1999), 1301-1306. [41] H. Baumler, B. Neu, R. Mitlohner, R. Georgieva, H.J. Meiselman and H. Kiesewetter, Electrophoretic and aggregation behavior of bovine, horse and human red blood cells in plasma and in polymer solutions, Biorheology 38 (2001), 39-51. [42] L. Dintenfass and L. Fu-Lung, Plasma and blood viscosities and aggregation of red cells in racehorses, Clin. Phys. Physiol. Measures 3 (1982), 293-301. [43] M.R. Fedde and S.C. Wood, Rheological characteristics of horse blood: significance during exercise, Respir. Physiol. 94 (1983), 323-335.

282


[44] M. Kaibara, Rheological behaviours of bovine blood forming artificial rouleaux, Biorheology 20 (1983), 583-592. [45] M. Kumaravel and M. Singh, Sequential analysis of aggregation process of erythrocytes of human, buffalo, cow, horse, goat and rabbit, Clin. Hemorheol. 15 (1995), 291-304. [46] A.S. Popel, P.C. Johnson, M.V. Kameneva and M.A. Wild, Capacity for red blood cell aggregation is higher in athletic mammalian species than in sedentary species, J. Appl. Physiol. 77 (1994), 1790-1794. [47] J.E. Smith, N. Mohandas and S.B. Shohet, Variability in erythrocyte deformability among various mammals, Am. J. Physiol 5 (1979), H725-H730. [48] M.I. Spengler and M. Rasia, Influence of plasma proteins on erythrocyte aggregation in three mammalian species, Vet. Res. Commun. 25 (2001), 591-599. [49] X. Weng, G. Cloutier, P. Pibarot and L.G. Durand, Comparison and simulation of different levels of erythrocyte aggregation with pig, horse, sheep, calf and normal human blood, Biorheology 33 (1996), 365-377. [50] U. Windberger, V. Ribitsch, K.L. Resch and U. Losert, The viscoelasticity of blood and plasma in pig, horse, dog, ox, and sheep, J. Exp. Anim. Sci. 36 (1994), 89-95. [51] M.H. Maxwell, G.W. Robertson and C.C. McCorquodale, Whole blood and plasma viscosity in normal and ascetic broiler chickens, Brit. Poultr. Sci. 33 (1992), 871-877. [52] S.M. Mirsalimi, P.J. O'Brien and R.J. Julian, Changes in erythrocytes deformability in NaCl-induced right sided cardiac failure in broiler chickens, Am. J. Vet. Res. 53 (1992), 2359-2362. [53] S.M. Mirsalimi and R.J. Julian, Reduced erythrocyte deformability as a possible contributing factor to pulmonary hypertension and ascites in broiler chickens, Avian Dis. 35 (1992), 374-379. [54] G.W. Robertson and M.H. Maxwell, Plasma viscosity values and the relationship with age and sex in normal commercial broiler and layer strains of chickens, Br. Poultr. Sci. 37 (1996), 309-316. [55] O.K. Baskurt, Deformability of red blood cells from different species studied by resistive pulse shape analysis technique, Biorheology 33 (1996), 169-179. [56] O.K. Baskurt, M. Bor-Kucukatay, O. Yalcin and H.J. Meiselman, Aggregation behavior and electrophoretic mobility of red blood cells in various mammalian species, Biorheology 37 (2000), 417428. [57] T.S. Hakim and A.S. Macek, Effect of hypoxia on erythrocyte deformability in different species, Biorheology 25 (1988), 857-868. [58] K. Ohta, F. Gotoh, M. Tomita, N. Tanahashi, M. Kobari, T. Shinohara, Y. Tereyama, B. Mihara and H. Takeda, Animal species differences in erythrocyte aggregability, Am. J. Physiol. 262 (1992), H1009H1012. [59] S.O. Sowemimo-Coker, P. Whittingstall, L. Pietsch, R.M. Bauersachs, R.B. Wenby and H.J. Meiselman, Effects of cellular factors on the aggregation behavior of human, rat and bovine erythrocytes, Clin. Hemorheol. 9 (1989), 723-737. [60] R.E. Waugh, Red cell deformability in different vertebrate animals, Clin. Hemorheol. 12 (1992), 649656. [61] S. Usami, V. Magazinovic, S. Chien and M.I. Gregersen, Viscosity of turkey blood: rheology of nucleated erythrocytes, Microvasc. Res. 2 (1970), 489-499. [62] G. Viscor, J.R. Torrella, V. Fouces and T. Pages, Hemorheology and oxygen transport in vertebrates. A role in thermoregulation? J. Physiol. Biochem. 59 (2003), 277-286. [63] R.L. Engen and C.L. Clark, High performance liquid chromatography determination of erythrocyte membrane phospholipid composition in several animal species, Am. J. Vet. Res. 51 (1990), 577-580. [64] M.I. Gregersen, C.A. Bryant, S. Chien, R.J. Dellenback, V. Magazinovic and S. Usami, Species differences in the flexibility and deformation of erythrocytes (RBC), Bibl. Anat. 10 (1968), 104-108. [65] J. Clarke and S. Nicol, Blood viscosity of the little penguin, Eudyptula minor, and the Adelie penguin, Pygoscelis adeliae: effects of temperature and shear rate, Physiol. Zool. 66 (1993), 720-731. [66] C.L. Guard and D.E. Murrish, Effects of temperature on the viscous behavior of blood from antarctic birds and mammals, Comp. Biochem. Physiol. A 52 (1975), 287-290. [67] B.L. Langille and B. Crisp, Temperature dependence of blood viscosity in frogs and turtles: effect on heat exchange with environment, Am. J. Physiol 239 (1980), R248-R253. [68] D.K. Saunders and K.H. Patel, Comparison of blood viscosity in red-eared sliders (Trachemys scripta) adapted to cold and room temperature, J. Exp. Zool. 281 (1998), 157-163. [69] R. Wells and J. Baldwin, Oxygen transport in marine green turtle (Chelonia mydas) hatchlings: blood viscosity and control of haemoglobin oxygen-affinity, J. Exp. Biol. 188 (1994), 103-114. [70] G.K. Snyder, Influence of temperature and hematocrit on blood viscosity, Am. J. Physiol 220 (1971), 1667-1672. [71] R.M.G. Wells, L.A. Beard and G.C. Grigg, Blood viscosity and hematocrit in the Estuarine crocodile, Crocodylus porosus, Comp. Biochem. Physiol. A 99 (1991), 411-414.


283

[72] A. Biswas, Body fluid and haematogical changes in a poikilothermic animal on cold exposure: some physical changes, Biorheology 3 (1976), 385-387. [73] N.M. Palenske and D.K. Saunders, Comparisons of blood viscosity between amphibians and mammals at 3oC and 38oC, J. Thermal Biol. 27 (2002), 479-484. [74] N.M. Palenske and D.K. Saunders, Blood viscosity and haematology of American bullfrog (Rana catesbeiana) at low temperature, J. Thermal Biol. 28 (2003), 271-277. [75] W.W. Weathers, Influence of temperature on the optimal hematocrit of the bullfrog (Rana catesbeiana), J. Comp. Physiol. 105 (1976), 173-184. [76] A.R. Bodey and M.W. Rampling, A comparative study of the haemorheology of various breeds of dog, Clin. Hemorheol. Microcirc. 18 (1998), 291-298. [77] S.S. Hillman, P.C. Withers, M.S. Hedrick and P.B. Kimmel, The effects of erythrocythemia on blood viscosity, maximal systemic oxygen transport capacity and maximal rates of oxygen consumption in an amphibian, J. Comp. Physiol. B 155 (1985), 577-581. [78] U. Windberger, R. Plasenzotti and T. Voracek, The fluidity of blood in African elephants (Loxodonta Africana), Clin. Hemorheol. Microcirc. 33 (2005), 321-326. [79] G. Chioccia and R. Mozais, Effect of catecholamines on deformability of red cells from trout: relative roles of cyclic AMP and cell volume, J. Physiol. 412 (1989), 321-332. [80] G.L. Fletcher and R.T. Haedrich, Rheological properties of rainbow trout blood, Can. J. Zool. 65 (1987), 879-883. [81] G.M. Hughes, Y. Kikuchi and J. Barrington, Physiological salines and the mechanical of trout red blood cells, J. Fish. Biol. 29 (1986), 393-402. [82] T. Lecklin, G.B. Nash and S. Egginton, Do fish acclimated to low temperature improve microcirculatory perfusion by adapting red cell rheology? J. Exp. Biol. 198 (1995), 1801-1808. [83] R.M.G. Wells and R.E. Weber, Is there an optimal hematocrit for rainbow trout, Oncorhynchus mykiss (Walbaum)? An interpretation of recent data on blood viscosity measurement, J. Fish. Biol. 38 (1991), 53-65. [84] G.M. Hughes and C. Albers, Use of filtration on methods in evaluation of the condition of fish red blood cells, J. Exp. Biol. 138 (1988), 523-527. [85] Y. Kikuchi, G.M. Hughes and C. Albers, Temperature dependence of the deformability of carp red blood cells, Experentia 38 (1982), 823. [86] R. Elsner and H.J. Meiselman, Splenic oxygen storage and blood viscosity in seals, Marine Mam. Sci. 11 (1995), 93-96. [87] M.S. Hedrick, D.A. Duffield and L.H. Cornell, Blood viscosity and optimal hematocrit in a deep-diving mammal, the northern elephant seal (Mirounga angustirostris) , Can. J. Zool. 64 (1986), 2081-2085. [88] L.L. Wickham, R. Elsner, F.C. White and L.H. Cornell, Blood viscosity in phocid seals: possible adaptations to diving, J. Comp. Physiol. B 159 (1989), 153-158. [89] L.L. Wickham, Blood rheology of captive and free ranging northern elephant seals and sea otters, Can. J. Zool. 68 (1990), 375-380. [90] S. Egginton, Blood rheology of Antarctic fishes: viscosity adaptations at very low temperatures, J. Fish. Biol. 48 (1996), 513-521. [91] M.S. Graham, G.I. Flecther and R.T. Haedrich, Blood viscosity in Arctic fishes, J. Exp. Zool. 234 (1985), 157-160. [92] R.M.G. Wells, J.A. MacDonald and G. Diprisco, Thin blooded antarctic fishes: a rheological comparison of the haemoglobin-free icefishes Chionodraco kathleenae and Cryodraco antarcticus with a red-blooded nototheniid, Pagothenia bernacchii, J. Fish. Biol. 36 (1990), 595-609. [93] M.I. Talan and B.T. Engel, Morning increase in whole blood viscosity: a consequence of a homeostatic nocturnal haemodynamic pattern, Acta. Physiol. Scand. 147 (1993), 179-183. [94] G.M. Hughes, Y. Kikuchi and H. Watari, A study of the deformability of red blood cells of a teleost fish, the yellow-tail (Seriola quinqueradiata), and a comparison with human erythrocytes, J. Exp. Biol. 96 (1982), 209-220. [95] R.M.G. Wells and M.E. Forster, Dependence of blood viscosity on haematocrit and shear rate in a primitive vertebrate, J. Exp. Biol. 145 (1989), 483-487. [96] R. Plasenzotti, B. Stoiber, M. Posch and U. Windberger, Red blood cell deformability and aggregation behaviour in different animal species, Clin. Hemorheol. Microcirc. 31 (2004), 105-111. [97] B. Stoiber, C. Zach, B. Izay and U. Windberger, Whole blood, plasma viscosity, and erythrocyte aggregation as a determining factor of competitiveness in standard bred trotters, Clin. Hemorheol. Microcirc. 32 (2005), 31-41. [98] S. Toplan, N. Dariyerli, S. Ozdemir, M.C. Akyolcu, H. Hatemi and G. Yigit, The effect of experimental hypothyroidism on hemorheology and plasma fibrinogen concentration, Endocrine 28 (2005), 153-156. [99] R. Waugh and E.A. Evans, Viscoelastic properties of erythrocyte membranes of different vertebrate animals, Microvasc. Res. 12 (1976), 291-304.

284


[100] O.K. Baskurt and H.J. Meiselman, Analyzing shear stress-elongation index curves: comparison of two approaches to simplify data presentation, Clin. Hemorheol. Microcirc. 31 (2004), 23-30. [101] P. Wong, A basis of echinocytosis and stomatocytosis in the disc-sphere transformations of the erythrocyte, J. Theor. Biol. 196 (1999), 343-361. [102] D.N. Wang, Band 3 Protein: structure, flexibility and function, FEBS Lett. 346 (1994), 26-31. [103] E.M. Guerra-Shinohara and O.C. de O Baretto, The erythrocyte cytoskeleton protein 4.2 is not demonstrable in several mammalian species, Braz. J. Med. Biol. Res. 32 (1999), 683-687. [104] J.K. Khohadad and R.S. Weinstein, The band 3-rich membrane of Llama erythrocytes: studies on cell shape and the organization of membrane proteins, J. Membrane Biol. 72 (1983), 161-171. [105] R.A. McPherson, W.H. Sawyer and L. Tilley, Band 3 mobility in camelid elliptocytes: implications for erythrocyte shape, Biochemistry 32 (1993), 6696-6702. [106] J. Oulevey, E. Bodden and O.W. Thiele, Quantitative determination of glycosphingolipids illustrated by using erythrocyte membranes of various mammalian species, Eur. J. Biochem. 79 (1977), 265-267. [107] R. Plasenzotti, U. Windberger, F. Ulberth, W. Osterode and U. Losert, Influence of fatty acid composition in mammalian erythrocytes on cellular aggregation, Clin. Hemorheol. Microcirc. Submitted (2007). [108] M. Uyuklu, H.J. Meiselman and O.K. Baskurt, Effect of decreased plasma cholesterol by atorvastatin treatment on erythrocyte mechanical properties., Clin. Hemorheol. Microcirc. 36 (2007), 25-33. [109] E.H. Eylar, M.A. Madoff, O.V. Brody and J.L. Oncley, The contribution of sialic acid to the surface charge of the erythrocyte, J. Biol. Chem. 237 (1962), 1992-2000. [110] M.W. Rampling, H.J. Meiselman, B. Neu and O.K. Baskurt, Influence of cell-specific factors on red blood cell aggregation, Biorheology 41 (2004), 91-112. [111] H.J. Meiselman, Red blood cell role in RBC aggregation: 1963-1993 and beyond, Clin. Hemorheol. 13 (1993), 575-592. [112] W.Kraft and U.M.Dürr. Klinische Labordiagnostik in der Tiermedizin. Schattauer, Stuttgart-New York, 1997. [113] C. Mahony, J. Ferguson and P.L.C. Fischer, Red cell aggregation and the echogenicity of whole blood, Ultrasound Med. Biol. 18 (1992), 579-586. [114] D.G. Paeng, R.Y. Chiao and K.K. Shung, Echogenicity variations from porchine blood I: the "bright collapsing ring" under pulsatile flow, Ultrasound Med. Biol. 30 (2004), 45-55. [115] H.Schlichting and K.Gersten. Grenzschicht-Theorie. Springer, Berlin, Heidelberg, 1997. [116] Y. Suzuki, N. Tateishi, M. Soutani and N. Maeda, Deformation of erythrocytes in microvessels and glass capillaries: effect of erythrocyte deformability, Microcirculation 3 (1996), 49-57. [117] J.L. Sebastian, S.M. San Martin, M. Rancho and J.M. Miranda, Erythrocyte rouleaux formation under polarized electromagnetic fields, Physical Rew. E 72 (2005), 031913. [118] M. Cabel, H.J. Meiselman, A.S. Popel and P.C. Johnson, Contribution of red blood cell aggregation to venous vascular resistance in skeletal muscle, Am. J. Physiol. 272 (1997), H1020-H1032. [119] R.J. Geor, D.J. Weiss and C.M. Smith, Hemorheologic alterations induced by incremental treadmill exercise in Thoroughbreds, Am. J. Vet. Res. 55 (1994), 854-861. [120] C.S. Sommardahl, F.M. Andrews, A.M. Saxton, D.R. Geiser and P.L. Maykuth, Alterations in blood viscosity in horses competing in cross country jumping, Am. J. Vet. Res. 55 (1994), 389-394. [121] A.S. Popel, P.C. Johnson, M.V. Kameneva and M.A. Wild, Capacity for red blood cell aggregation is higher in athletic mammalian species than in sedentary species, J. Appl. Physiol. 77 (1994), 1790-1794. [122] B. Schmidt-Nielsen, K. Schmidt-Nielsen, T.R. Houpt and S.A. Jarnum, Water balance of the camel, Am. J. Physiol 185 (1956), 185-194. [123] K. Yamaguchi, K.D. Jurgens, H. Bartels and J. Piiper, Oxygen transfer properties and dimensions of red blood cells in high-altitude camelids, dromedary camel and goat, J. Comp. Physiol. 157 (1987), 1-9. [124] D. van Houten, M.G. Weiser, L. Johnson and F. Garry, Reference hematologic values and morphologic features of blood cells in healthy adult Llamas, Am. J. Vet. Res. 53 (1992), 1773-1775. [125] A. Eitam, B. Aloni and A. Livne, Unique properties of the camel erythrocyte membrane II. Organisation of membrane proteins, Biochem. Biophys. Acta 426 (1976), 647-653. [126] S.J. Thornton and P.W. Hochanchka, Oxygen and the diving seal, Undersea Hyperbaric Med. Soc. 31 (2004), 81-95. [127] F. Boily, S. Beaudoin and L.N. Measures, Hematology and serum chemistry of harp (Phoca groenlandica) and hooded seals (Cystophora cristata) during breeding season, in the Gulf of St. Lawrence, Canada, J. Wildlife Dis. 42 (2006), 115-132. [128] R. Elsner. Cardiovascular adjustments to diving. In: The biology of Marine Mammals, H.T.Anderson, Ed. Academic Press, New York , 1969, pp. 117-145. [129] C. Fayolle, C. Leray, P. Ohlmann, G. Gutbier, J.P. Cazenave, C. Gachet and R. Groscolas, Lipid composition of blood platelets and erythrocytes of southern elephant seal (Mirounga leonina) and Antarctic fur seal (Arctocephalus gazelle) , Comp. Biochem. Physiol. B 126 (2000), 39-47.


285

[130] P. Gaehtgens, F. Schmidt and G. Will, Comparative rheology of nucleated and non-nucleated red blood cells I. Microrheology of avian erythrocytes during capillary flow, Pflügers Arch. 390 (1981), 278-282. [131] M. Singh and J.F. Stoltz, Influence of temperature variation from 5oC to 37oC on aggregation and deformability of erythrocytes, Clin. Hemorheol. Microcirc. 26 (2002), 1-7. [132] P. Gaehtgens, G. Will and F. Schmidt, Comparative rheology of nucleated and non-nucleated red blood cells. II. Rheological properties of avian red cells suspensions in narrow capillaries, Pflügers Arch. 390 (1981), 283-287. [133] S.F. Fontes, R. Hernandes, M. Macari and F.M. Bernal, Blood viscosity as diagnostic parameter for ascites in broiler chickens strains of different susceptibility, Rev. Bras. Cienc. Avic. 2 (2000), 45-51. [134] W.T. Zhou, M. Fujita and S. Yamamoto, Effect of food and water withdrawal and high temperature exposure on diurnal variation in blood viscosity of broiler chickens, Br. Poultr. Sci. 39 (1998), 156160. [135] D.K. Saunders, A.C. Roberts and G.R. Ultsch, Blood viscosity and haematological changes during prolonged submerge in normoxic water of northern and southern musk turtles (Sternotherus odoratus), J. Exp. Zool. 287 (2000), 459-466.


III. Hemodynamics



289

Basic Principles of Hemodynamics Timothy W. SECOMBa,1 and Axel R. PRIESb Department of Physiology, University of Arizona, Tucson, AZ 85724, USA and b Charité - Universitätsmedizin Berlin, Campus Benjamin Franklin, Dept. of Physiology, and Deutsches Herzzentrum Berlin, Berlin, Germany a

Introduction The circulatory system consists of a pump (the heart) and an extensive, highly branching system of tubes (blood vessels) containing a fluid (blood) with specialized capabilities for the transport of oxygen, nutrients, many other substances and heat. The rates of blood flow through the blood vessels depend on many physical factors, including the diameters, lengths and other geometric features of the vessels, their mechanical properties, the structures of networks that they form, the pressure generated by the heart to drive the flow, and the rheological properties of the blood. All of these factors are themselves subject to variation according to a number of short-term and long-term biological control mechanisms. In order to understand this system, it is helpful to start by considering the mechanics of fluid flow through a single tube with a uniform cylindrical cross-section. Under appropriate conditions, the relationship between driving pressure and flow rate can be described by the equation generally known as Poiseuille’s law. In this chapter, a derivation of this equation is presented, and its restrictions and limitations are discussed. This provides a basis for consideration of a range of more complex fluid dynamical phenomena occurring in the circulatory system. More detailed discussions of many of the topics mentioned here can be found in the several books [1-5].

1. Blood Flow in Arteries and Veins 1.1. Steady Laminar Flow in a Uniform Tube In the mid-nineteenth century, the relationship between the fluid flow rate Q in a tube of diameter D and length L and the driving pressure 'p was explored by J.L.M. Poiseuille [6], who sought to understand the physical factors governing blood flow. From his experimental observations, he established the fourth-power relationship Q = K'pD4/L, where the factor K depended on the type and temperature of the fluid. Subsequent theoretical analysis [7] led to the relationship now commonly known as Poiseuille’s law: S 'pD 4 (1) Q 128 LP 1

Corresponding author: Department of Physiology, University of Arizona, Tucson, AZ 85724, USA; E mail: [email protected]

290

T.W. Secomb and A.R. Pries / Basic Principles of Hemodynamics

where μ is the fluid viscosity. The history of Poiseuille’s law is discussed by Sutera and Skalak [8]. The relationship Eq. (1) can be assumed to hold only under special conditions, as follows: x the tube is a uniform circular cylinder; x the tube is rigid; x the fluid is Newtonian; x the flow is steady, i.e., constant in time; x the flow is laminar, i.e., not turbulent; x the flow is not subject to entrance effects, i.e., non-uniformities associated with the entrance of fluid into the tube. In many situations of practical interest, including blood flow in the circulation, one or more of these conditions is not satisfied, as discussed in subsequent sections. Even so, Poiseuille’s law is central to understanding of the hemodynamics of the circulation. The following derivation gives insight into the fluid forces generated by Poiseuille flow in a tube. We consider a tube of radius a = D/2 and length L, filled with a fluid of fixed viscosity μ. Further, we consider the forces acting on a narrower cylindrical region of fluid, with radius r < a, concentric with the tube (Figure 1). The only stresses acting on this cylindrical region that generate net resultant forces acting along the tube are the hydrostatic pressures p1 and p2 acting on the ends of the region and the wall shear stress W acting on the curved surface. Under the assumptions listed above, each of these stresses is uniform over the surface on which it acts, and the resultant forces can be obtained by multiplying the stress by the corresponding surface area. If effects of gravity are neglected, the resultant forces must sum to zero since the flow is steady and the acceleration is zero, i.e. Sr2p2 í Sr2p1 í 2SrLW = 0 and so W = r'p /2L

(2)

where 'p = p2 í p1. In the presence of gravity g, the same result holds with the pressure p replaced by p' = p + ȡgz, where ȡ is the fluid density and z is the vertical coordinate. The shear stress therefore varies linearly with radial position, from zero on the centerline to a maximum of Ww = a'p/2L at the wall of the tube, where Ww is the wall shear stress.

a

r L

Sr2 p2

Sr2 p1 2SrLW

Figure 1. Definition of geometry and forces used in derivation of Poiseuille’s law

291


It is important to note that the result in Eq. (2) makes no assumption about the viscosity of the fluid, or whether it is a Newtonian fluid. It can be applied in situations where the viscosity of the fluid varies with radial position. This may occur because the viscosity of blood depends on the shear rate, as discussed in Chapter II.3.a, or because the distribution of red blood cells within the tube varies with radial position, as discussed in Chapters III.2 and III.3 If the fluid is Newtonian, the shear stress equals the product of viscosity and velocity gradient:

W P

du dr

(3)

where u(r) is the fluid velocity. A minus sign appears here because W is defined in the negative direction (Figure 1). Combining Eqs. (2) and (3) gives du dr

W P

r'p 2 LP

(4)

Viscous fluids must satisfy the no-slip condition, that the fluid velocity adjacent to a solid surface matches that of the surface. Eq. (4) can be integrated with the boundary condition u(a) = 0, giving 'p 2 (a r 2 ) 4 LP

u (r )

(5)

The velocity profile, i.e., a plot showing velocity as a function of position, is then a parabola with its vertex on the center-line of the vessel, as shown in Figure 2. The peak velocity, which occurs on the center-line, is given by 'pa 2 4 LP

u max

(6)

r u(r)

Figure 2. Parabolic velocity profile in Poiseuille flow

The volume flow rate Q in the tube is then calculated by integrating the velocity across the circular cross-section of the tube, Q

³

a

0

u ( r ) 2Sr dr

S 'pa 4 8 LP

(7)

which is equivalent to Eq. (1) with D = 2a. The mean velocity of flow in the tube can be computed as

292


u mean

Q Sa 2

'pa 2 8 LP

(8)

Comparison with Eq. (6) shows that the mean velocity in Poiseuille flow is exactly half of the center-line velocity. 1.2. Flow Resistance A helpful analogy can be drawn between a network of blood vessels and a network of electrical resistors. In this analogy, pressure corresponds to electrical potential and volume flow rate to electrical current. Then flow resistance R is defined as the ratio of pressure difference to flow rate, i.e. R

'p Q

128 LP S D4

(9)

This relationship is of central importance in understanding how blood flow is controlled and distributed in the circulatory system. In particular, it shows that the resistance to blood flow is a sensitive function of the diameter D, being proportional to 1/D4. Therefore, blood flow can be modulated over a wide range by moderate changes in the diameter of the blood vessels responsible for most of the flow resistance in the circulatory system, namely the small arteries and arterioles. Local regulation of blood flow is achieved by active contraction and dilation of these vessels. Flow in networks containing many vessels can be analyzed using the concept of flow resistance. For example, if N identical vessels are connected in parallel, their overall flow resistance is given by R

'p Q

1 128 LP N S D4

(10)

The concept of flow resistance can be extended to the whole circulatory system. In this case, the driving pressure is the difference between mean arterial pressure (MAP) and central venous pressure (CVP), and the volume flow rate is the cardiac output (CO). Thus the total peripheral resistance (TPR) is defined by: TPR = (MAP í CVP) / CO

(11)

The total peripheral resistance is defined by the geometric properties of the vascular system at any moment, including effects of vascular tone on vessel diameter, and by the rheological properties of blood. Through Eq. (11), it is evident how the flow properties of blood directly influence the workload on the heart, i.e., the pressure it must generate to produce a given cardiac output. 1.3. Governing Equations for General Fluid Motions Analysis of blood flow in more complex geometries requires the solution of the NavierStokes equations, which are the general equations of motion for a viscous fluid. The derivation of these equations is found in text books of fluid mechanics [9]. For most purposes, blood can be regarded as incompressible with constant density, U, and conservation of mass gives the incompressibility condition

293


. u = 0

(12)

where u(x,t) describes the velocity as a vector function of position x. Conservation of momentum gives U (wu/wt + u .u) = íp + P 2u + F

(13)

where F is a body force (per unit volume) acting on the fluid, typically gravity. Generally the incompressible Navier-Stokes Eqs. (12) and (13) must be solved for the unknown variables p and u to predict the fluid motion in a given situation. Solutions of these equations encompass a remarkably wide range of behaviors, depending on the flow domain and other boundary conditions. The equations are nonlinear in the velocity as a result of the term u .u, which describes the acceleration experienced by a fluid as it moves through a non-uniform flow field. This nonlinearity leads to complexity in solution methods. Under some circumstances, one or more of the terms in Eq. (13) can be neglected, permitting the use of simplified solution methods. For example, in very slow or highly viscous flows, the inertial terms on the left hand side of Eq. (13) may be negligible, whereas the viscous term P2u may be neglected under some circumstances if velocities are high and effects of viscosity are slight. 1.4. Reynolds Number and Turbulence Insight into the nature of fluid flows can be obtained by consideration of dimensionless parameters associated with the flows. The most important of these is the Reynolds number, which is derived by considering the relative orders of magnitude of the viscous and inertial terms in steady flow, for which wu/wt = 0. We assume that the flow is characterized by a typical velocity U and a typical length L. The magnitude of the inertial and viscous terms may then be estimated as U u .u ~ UU2/L and P 2u ~ PU/L2

(14)

where ‘~’ means ‘is of the same order of magnitude as’. The Reynolds number is defined as the ratio of the inertial term to the viscous term, i.e. Re = (UU2/L)/(PU/L2) = UUL/P

(15)

When Re is much less than 1, the inertial terms in (13) are generally negligible, and can be neglected. This case is known as Stokes flow. When Re is much larger than 1, inertial effects are dominant, but viscosity may still exert a strong influence on the flow, particularly with regard to the occurrence of laminar and turbulent flow. In laminar flow, the fluid motion is smooth and ordered. Although the flow may be unsteady (i.e., time-varying), the variation with time is predictable, at least in principle, under a given set of conditions. Conversely, turbulent flow involves unpredictable fluctuations of velocity. It results from the uncontrolled growth of random small disturbances in the velocity field. Viscosity (i.e., viscous damping) tends to inhibit the growth of such disturbances. With increasing Re, the effect of viscosity diminishes, and turbulence can develop if the Re is high enough. For example, if the flow rate is gradually increased in a long straight cylindrical tube, the flow remains laminar up to a Reynolds number (Re = UUD/P where D is tube diameter) of about 2400, beyond which turbulence is observed [2]. In the presence of turbulence, viscous dissipation in the fluid is increased. The pressure drop along a tube with a given rate of

294


turbulent flow is substantially higher than would be predicted based on Poiseuille’s law, Eq. (1). The fact that the transition to turbulent flow occurs at this Re shows that even a relatively small amount of viscous damping can prevent the development of turbulence. No complete theory for turbulence exists, and flows in turbulent regimes are predicted based on a combination of theory and experimental information. It is instructive to consider the values of Reynolds number in the circulatory system. The density U of blood is approximately 1 g/cm3 and its viscosity P is about 0.05 dyn.s/cm2 (Poise) or 0.005 Pa.s. In a large human artery, typical values are U = 100 cm/s and D = 1 cm, giving Re = 2000. In a capillary, corresponding values are U = 0.1 cm/s, D = 0.001 cm, giving Re = 0.002. The Reynolds number therefore varies over at least six orders of magnitude in the circulation. Inertial effects are negligible in capillaries, and the Stokes flow regime applies. In large arteries, the Reynolds number is large but still generally within the laminar range for a uniform tube. However, the Reynolds number at which turbulence may occur is dependent on the geometry of the flow domain. At any location where fluid emerges from a narrow opening into a larger region, forming a jet, the flow is more likely to be unstable and turbulence can develop at lower Reynolds numbers. Examples are a stenosed artery or heart valve. The sounds resulting from turbulence generated at such locations are useful for the diagnosis of vascular or cardiac abnormalities. 1.5. Entrance Effects When fluid enters a tube, the initial velocity profile depends on conditions upstream of the vessel and the parabolic profile characteristic of Poiseuille flow is not established immediately. Entrance effects may persist a considerable distance downstream, depending on the Reynolds number. The situation is illustrated in Figure 3, for the case when the initial velocity profile is almost uniform across the tube cross-section. The tube is considered to be uniform and the flow to be steady in time. According to the no-slip condition, fluid immediately adjacent to the wall must have zero velocity. A zone of reduced velocity increases with distance downstream, as the effects of viscosity act to spread the initially sharp gradient in velocity at the wall. This zone is known as a ‘boundary layer.’ As the thickness of this layer increases, fluid in the central region of the tube must speed up to maintain the same total flow rate at each point along the tube. Eventually, the layer spreads to fill the whole tube and the velocity profile approaches the Poiseuille profile. The centerline velocity is then double the initial uniform velocity, for reasons already discussed. In the entrance region of the flow, the kinetic energy of the fluid increases [10], and viscous dissipation is higher than in the fully developed flow further downstream. Therefore, an additional pressure drop beyond that computed using Poiseuille’s law is required to drive the flow. In this case, the ‘entrance length’ may be defined as the distance downstream at which the velocity profile closely approaches the parabolic shape. The rate at which the boundary layer spreads depends on the fluid viscosity, and therefore on the Reynolds number. As Re increases, the boundary layer spreads more slowly, and the entrance length is increased. For Re in the range of about 10 to 2000, such that flow is laminar, the entrance length is proportional to Re and is given approximately by Le ~ 0.03 Re D. At lower Reynolds numbers, the entrance length approaches a lower limit of about one vessel diameter. In the case of a large artery, with D = 1 cm and Re = 2000, Le = 60 cm. Therefore, the whole length of such an artery typically lies within the entry length, and fully developed Poiseuille flow is not attained in the artery.


295

Boundary layer Figure 3. Development of boundary layer in fluid entering a tube

1.6. Flow in Curved and Helical Tubes The arteries do not follow straight paths through the body. Most strikingly, the thoracic aorta has a strongly curved trajectory, ascending from the heart and then descending to the abdomen. The change in momentum of blood as it traverses such bends implies the presence of pressure gradients across the vessel. The higher-velocity fluid in the central region tends to move toward the outside of the curve, and a compensating secondary flow towards the inside of the curve is set up in the slower moving fluid near the wall (Figure 4). This behavior occurs in a tube whose center-line lies in a plane. However, the aorta is non-planar and follows a helical path. This property, together with the swirl of the blood flow generated within the left ventricle, generates a helical flow in the aorta. The interaction of the effects of curvature and helicity results in complex flow patterns [11]. It has been argued that the presence of helical, swirling flows in the heart and major arteries confers advantages in terms of flow stability and reduced energy dissipation [12].

Outer wall A

B

Figure 4. Flow in a curved tube. A. Distortion of axial velocity profile. B. Stream lines of secondary flow in tube cross-section. After [2].

1.7. Oscillatory Flow The discussion so far has been of steady flow, defined as a flow velocity that is constant in time at every location. However, a high degree of unsteadiness is characteristic of blood flow in the circulation, particularly in the arteries, as a result of the pulsatile pumping action of the heart. Velocity profiles in pulsatile flow may differ greatly from the parabolic profile of steady Poiseuille flow. If laminar flow in a uniform tube is assumed and entrance effects are neglected, the inertial effects are represented by the term Uwu/wt in the Navier-Stokes Eqs. (13). The order of magnitude

296


of this term may be estimated by assuming that the flow is characterized by a typical velocity U and a typical length T. The ratio of the inertial term to the viscous term in (13) is then of magnitude (UU/T)/(PU/L2) = UL2/PT. If the flow in a tube with radius a is periodic with angular frequency Z (= 2ʌf where f is frequency), then we may let L = a and T = 1/Z and define a non-dimensional parameter D = (UL2/PT)1/2 = a(UZ/P)1/2, known as the unsteadiness parameter or as the Womersley number. When D is small, the effect of pulsatility on the flow profile is slight, and the flow is quasi-steady, and at each instant the flow is approximately the same as the steady flow that would be obtained with the pressure gradient prevailing at that instant. Conversely, the velocity profiles are strongly influenced by pulsatility when D is large. The flow waveform generated by the heart is complex, but useful insight may be obtained by considering sinusoidally varying flow in a uniform cylindrical tube. In that case, the resulting flow profiles may be computed exactly in terms of Bessel functions [4]. If the pressure gradient is given by dp/dx= íRe[p' eiȦt] then the velocity is given by u(r,t) = Re[u' eiȦt] where uc

p ca 2 ª J 0 (Di 3 / 2 r / a) º » «1 iPD 2 ¬ J 0 ( Di 3 / 2 ) ¼

(16)

where J0 is a Bessel function of the first kind. Examples of resulting velocity profiles are shown in Figure 5 for several values of D. When D = 1, the velocity profiles are nearly parabolic and the velocity varies with time in phase with the pressure gradient. Wall shear stress, as indicated by the velocity gradient near the wall, varies in phase with pressure gradient. This corresponds to quasi-steady behavior. At larger D values, velocity profiles are blunted in the interior of the tube and fluid velocity has large variations in amplitude and phase near the wall. The variations with time of velocity in the central region and of wall shear stress both lag the variation of pressure gradient. When D is large, the phase lags of the interior velocity and the wall shear stress are 90° and 45° respectively, relative to the pressure gradient. D=1

D=2

D=5

D = 10

Figure 5. Sequences of velocity profiles in a tube with a sinusoidally varying pressure gradient, for indicated values of unsteadiness parameter D. Velocity profiles represent one complete cycle of the oscillation. Leftmost profile corresponds to moment of maximum pressure gradient.

297


The flow in a uniform tube driven by a pressure with a more complex timedependence, as generated by the heart, may be computed by expressing the pressure gradient waveform as a Fourier series (i.e., sum of steady component, fundamental frequency and harmonics) and superimposing the corresponding velocity profiles. Such an analysis makes clear that the instantaneous relationships between pressure gradient, flow rate and wall shear stress in arteries may differ greatly from the Poiseuille relationships, Eqs. (1) and (2). 1.8. Pulse Propagation in Arteries The walls of arteries are compliant, containing layers of smooth muscle, elastin and collagen, and their diameters vary in response to the time-varying pressures generated by the heart. Consequently, the pressure pulse delivered by the heart is not transmitted immediately to all parts of the arterial system, but is propagated along the vessels in the form of a traveling wave. The speed of this wave depends primarily on the ratio between the elastic resistance of the wall to deformation and the inertia of the fluid. Although this process involves a complicated interaction between several phenomena, including the fluid mechanical effects already discussed, a useful approximation can be obtained using a one dimensional model, in which the fluid velocity along the artery is expressed as u(x,t), and radial variations are neglected. Effects of blood viscosity and the no-slip condition are neglected in this analysis. The fundamental principles of conservation of mass and momentum then give

wA w ( Au ) 0 wt wx

wu 1 wp wu u wx U wx wt

(17)

0

(18)

where A = Sa2 is the cross-section area of the artery and a is the radius. The mechanical behavior of the wall is represented approximately by assuming that the radius depends on pressure in the tube relative to the external pressure (i.e., a = a(p)). A further simplification, using conditions not satisfied in reality, can be made by assuming that the tube radius has small variations about a value a0, and that the fluid velocity is small. Under these assumptions, a linearized analysis of Eqs. (17) and (18) yields:

J

wp a0 wu wt 2 wx

wu 1 wp wt U wx

0

(19) (20)

0

where J = 2 da/dp is the compliance and a0 is the mean radius. These may be combined to give a wave equation for pressure: w2 p w2 p c2 2 2 wt wx

0

where c = (a0/JU)1/2 is the wave speed. The general solution is

(21)

298


p = f(x í ct) + g(x + ct)

(22)

where f and g are unknown functions. This solution represents a superposition of waves of arbitrary shape traveling in both directions along the tube. In arteries, the dominant wave travels away from the heart, but a reflected component is generated due to non-uniformities in the artery, including branch points and vessel taper. For a tube with a thin wall of thickness h, composed of a material with Young's modulus E, the compliance is given by J = 2a02/(Eh), and the wave speed is c = (Eh/2Ua0)1/2

(23)

This relationship was first derived in 1809 by Thomas Young, and is known as the Moens-Korteweg formula [2]. Estimates for the canine aorta of the quantities in this equation are E = 4.8 u 105 N/m2, h/2a0 = 0.07 and U = 103 kg/m3, giving c = 5.8 m/s [2]. This simplified analysis neglects a number of factors that influence the propagation of the arterial pressure pulse. (i) The analysis assumes that the compliance J is a constant. In reality, the elastic properties of the wall are highly nonlinear, so that da/dp is a function of pressure. The amplitude of the pulse wave is not small, and the pressure varies over a substantial range. As a consequence, the effective compliance of the wall is lower at the peak of pressure (systole) and higher at the trough of pressure (systole). The wave velocity is therefore higher for the peak than for the trough. This causes steepening of the leading edge of the pressure pulse wave. (ii) Arterial diameters decrease with distance in the downstream direction, causing amplification of the pulse wave. (iii) The assumption of small velocity means, in effect, that the fluid velocity is assumed to be small compared to the wave velocity. In fact, the fluid velocity may be about 1 m/s, which is not negligible relative to a pulse velocity in the range 5-10 m/s. (iv) Deformation of the wall and surrounding tissue is not a purely elastic process. The viscous component of the response to deformation results in energy dissipation. (v) Fluid viscosity creates radial variations in velocity, as already discussed, which affects the momentum of the blood flow and also results in energy dissipation. These aspects of the hemodynamics of pulse wave propagation are discussed in more detail by Fung [3]. 1.9. Flow Parameters in the Circulatory System The circulatory system consists of an immense number of vessels, connected together in a branched network. Data on the number and geometry of the vessels making up the circulatory system of the dog are represented graphically in Figure 6. The wide range of lengths and diameters is immediately evident. A less obvious but highly significant feature is that venous vessels are larger than corresponding arterial vessels, typically with about twice the diameter. The flow rates and lengths of corresponding arterial and venous vessels are necessarily virtually equivalent. Therefore, according to Eq. (9), the flow resistance is much larger in the arterial side of circulation than in the venous side, and thus the pressure drop is concentrated on the arterial side, specifically in the small arteries and arterioles. This arterio-venous asymmetry has the important consequence that capillary pressure is much lower than the mean of arterial and venous pressures, which would be the capillary pressure if the system were symmetric. High capillary pressures would lead to increased fluid leakage from capillaries. Also, the fact that most of the flow resistance is concentrated in the small arteries and arterioles permits


299

these vessels to control blood flow over a wide range by active variation of their diameters.

109

Total cross-sectional 2 area (cm )

Number 108

103 102 3

Volume (cm )

101 100

Length (cm) 10-1 Diameter (mm) 10-2 es es r ta les ies ules eins ches eins cava ies v v n Ao arter anch anch terio pillar n a r r r Ve inal a b b bra arge Ven A C y r ge al L rm ous e La arter rmin T n ve in Te in Ma Ma

Figure 6. Geometry of the vascular system in the dog, based on observations of the mesenteric vascular bed by F. Mall [1].

A further significant feature of the data shown in Figure 6 is that the number of vessels varies inversely with the cube of vessel diameter. On the arterial side, for instance, the number of vessels increases by a factor of 109 as the diameter declines by a factor of 103. From Eqs. (1) and (2), the wall shear stress in a cylindrical tube is given by 32P Q (24) Ww S D3 Therefore, the wall shear stress remains roughly constant throughout the arterial tree, down to the capillaries. Much evidence shows that blood vessels sense and respond to levels of wall shear stress, and tend to adjust their diameters in order to maintain it at a roughly constant level [13]. The level maintained varies with size of mammalian species, from almost 100 dyn/cm2 (10 Pa) in mice and rats to less than 10 dyn/cm2 (i.e., 1 Pa) in dogs and humans [14]. 1.10. Venous Hemodynamics As noted above, the veins are larger in diameter than the corresponding arteries, and contain the majority of the blood in the systemic circulation. According to Eqs. (9) and (24), pressure drops and shear stresses are substantially lower on the venous side than on the arterial side of the circulatory system. The lower shear stresses and shear rates

300


in the veins and venules increase the likelihood of significant nonlinear rheological effects as a result of aggregation of red blood cells [15]. The walls of veins support relatively low internal pressures, and are thinner than those of corresponding arteries. Valves that prevent backflow of blood are present in many veins. The strong pulsatility of the blood flow in the arteries is strongly damped as the flow passes through the microcirculation due to the compliance of the vessels and the strong viscous damping occurring at low Re. However, smaller venules were shown to exhibit pulsatility with respect to pressure and flow velocity [16, 17]. Furthermore, several circumstances can lead to unsteady venous flow. A pulsatile component of pressure is transmitted some distance upstream from the right atrium into the superior vena cava as a result of the oscillatory intake into the right atrium of the heart. Veins embedded in skeletal muscle are subject to external squeezing during muscle contraction. Periodic muscle contraction in the presence of one-way flow valves causes pumping of blood along the vein. Thin-walled tubes can show unstable oscillations under certain combinations of upstream, downstream and external pressures. This phenomenon has not been demonstrated in veins under normal in-vivo conditions, but contributes to the generation of Korotkoff sounds in the brachial artery during the measurement of blood pressure using an inflated cuff on the arm. These and other aspects of venous hemodynamics are discussed by Fung [3].

2. Blood Flow in the Microcirculation The microcirculation is generally considered to include all vessels with diameters less than about 300 μm. The Reynolds number of flow in these vessels is generally small, and the flow is laminar and governed by the equations of Stokes flow. The dynamics of blood flow in microvessels are strongly influenced by the suspension characteristics of blood. In vessels with diameters much larger than the dimensions of individual blood cells, blood can be considered as a continuum with nonlinear rheological properties, as discussed in Chapter II.3.b. In microvessels, however, the finite size of the suspended elements gives rise to non-continuum behavior, including reduction of hematocrit within microvessels, variation of the apparent viscosity of blood with tube diameter, and unequal partition of hematocrit between branches of diverging microvascular bifurcations. These phenomena are briefly reviewed here, with more detailed discussion of certain aspects presented in Chapter III.2. 2.1. The Fåhraeus Effect Observations of blood flowing in narrow tubes show that the concentration of red blood cells within the tube (i.e., tube hematocrit, HT) is less than the concentration in the blood entering the tube and leaving the tube (i.e., discharge hematocrit, HD). This difference is known as the Fåhraeus effect [18] and arises from the fact that red blood cells travel faster than plasma on average and therefore have shorter transit times. The two hematocrit values are related by HT HD

Vbulk Vrbc

(25)

301


where Vrbc is the mean velocity of red blood cells within the tube, and Vbulk is overall the mean flow velocity [19]. Based on observations of blood flow in glass tubes, the following empirical relationship for the dependence of the Fåhraeus effect on tube diameter and hematocrit was obtained [20]: HT HD

H D (1 H D ) (1 1.7 e 0.35 D 0.6 e 0.01D )

(26)

The dependence of this function on diameter for discharge hematocrit HD = 0.45 is shown in Figure 7. 1.0

HT/HD

0.8 0.6

In vitro Two-phase model experiments

0.4 0.2 0.0 3

10

30

100

300

1000

Diameter (Pm)

Figure 7. Variation of Fåhraeus effect with tube diameter for hematocrit HD = 0.45. The solid curve represents an empirical fit to experimental in-vitro data [20]. The dashed curve corresponds to a two-phase model with cell-free layer width 1.8 μm, as discussed in the text.

The underlying cause of the reduction in tube hematocrit of blood in narrow tubes is the formation of a layer of cell-free or cell-depleted plasma near the tube wall, such that red blood cells travel preferentially in the central region of the tube where the velocity is higher. The impact of such a layer on flow resistance can be seen by considering a simple two-phase model of blood flow, in which a cylindrical core region of viscosity μcore centered on the tube axis is surrounded by a cell-free layer of viscosity μp. An analysis similar to that used in Section 1.1 can be used to calculate the resulting velocity profile in the tube and to derive the following equation for the Fåhraeus effect [21]: HT HD

1 O4 (1 P p / P core ) 2 O2 (2 P p / P core )

(27)

where Ȝ is the ratio of the core radius to the tube radius, and Ȝ = 1 í į/a where į is the width of the cell-free layer and a is the tube radius. A relatively narrow cell-free layer with width į = 1.8 μm leads to a significant reduction tube hematocrit HT (Figure 7). 2.2. The Fåhraeus-Lindqvist Effect The resistance to blood flow in narrow tubes is conveniently discussed in terms of the apparent viscosity, which is defined by inverting Eq. (1) to give

302


P app

S 'pD 4 128 LQ

(28)

The relative apparent viscosity is defined by μrel = μapp/μp where μp is the viscosity of the plasma or other suspending medium. Experiments using glass tubes show that relative apparent viscosity of blood declines substantially with tube diameter at diameter below 300 μm, a phenomenon known as the Fåhraeus-Lindqvist effect [20, 22, 23]. This effect is reversed in very small tubes, near the minimum that allows passage of red blood cells which is about 3 μm. The variation of μrel with diameter and hematocrit can be described by the following empirical equations [20]: P rel

1 (P 45 1)

(1 H D )C 1 (1 0.45)C 1

(29)

where the relative apparent blood viscosity for a fixed discharge hematocrit of 0.45, is given by μ45 = 220 · exp(í1.3D) + 3.2 í 2.44 · exp(í0.06D0.645),

(30)

D is the lumen diameter of the tube in μm and the shape of the viscosity dependence on hematocrit is governed by the factor C: C

(0.8 e 0.075 D ) (1

1 1 10

11

12

D

)

1 1 10

11

D12

.

(31)

The dependence of apparent viscosity on diameter when HD = 0.45 is shown in Figure 8. In these relationships and throughout, all diameters are given in μm. The main cause of the reduction in apparent viscosity of blood in narrow tubes is again the formation of a layer of cell-free or cell-depleted plasma near the tube wall. The simple two-phase model of blood flow discussed above can be used to predict the relative apparent viscosity for this configuration [25]

P rel

1 1 O (1 P p / P core ) 4

(32)

where Ȝ is defined as before. If it assumed that the width į of the cell-free layer is independent of tube diameter, a good fit to the empirical curve is obtained with į = 1.8 μm for diameters 30 μm and above (Figure 8). A reduction of viscosity in this relatively narrow layer near the wall thus has a substantial impact on flow resistance. The mechanics of blood flow in smaller tubes, particularly in the diameter range of capillaries, can be analyzed in terms of the mechanics of individual red blood cells [21].

303


Relative apparent viscosity

5

4

In vivo experiments

3

In vitro experiments

2

Two-phase model 1 3

10

30

100

300

1000

Diameter (Pm)

Figure 8. Variation of apparent viscosity with tube diameter for hematocrit HD = 0.45. The lower solid curve represents an empirical fit to experimental in-vitro data [20]. The upper solid curve represents the dependence deduced from in-vivo experiments [24]. The dashed curve corresponds to a two-phase model with cell-free layer width 1.8 μm, as discussed in the text.

The in-vitro results described above were long assumed to describe microvascular flow resistance in vivo. However, experimental observations of the distributions of flow and hematocrit in microvascular networks indicated that flow resistance in living microvessels is substantially higher than in glass tubes with corresponding diameters [26]. Based on data obtained in the rat mesentery, Pries et al. [24] proposed the following modified in-vivo empirical relationship: P rel

2 2 ª (1 H D ) C 1 § D · º § D · ¨ «1 (P 45 1) ¸ » ¨ ¸ C (1 0.45) 1 © D 1.1 ¹ »¼ © D 1.1 ¹ «¬

(33)

where the relative apparent blood viscosity for HD = 0.45, is given by μ45 = 6 · exp(í0.085D) + 3.2 í 2.44 · exp(í0.06D0.645)

(34)

and the quantity C is given by equation (31) as before. The resulting variation of apparent viscosity HD = 0.45 is shown in Figure 8. Subsequent studies have led to the conclusion that the principal cause of the difference between in-vitro and in-vivo blood viscosity in microvessels is the presence of a relatively thick endothelial surface layer (ESL, also called glycocalyx) bound to the inner surface of endothelial cells, with width of order 1 μm. Based on this result, a more recent analysis [27] was carried out to deduce the variation of ESL thickness with microvessel diameter and hematocrit, and thereby to derive new relationships for the dependence of Fåhraeus effect and apparent viscosity on diameter and hematocrit. 2.3. Phase Separation Effect in Bifurcations A further consequence of the non-continuum behavior of blood in microvessels becomes evident at diverging microvascular bifurcations. The partition of red blood cells into the two downstream branches does not generally correspond to the partition of total blood flow, with the result that different hematocrits are found in the downstream branches. Generally, the higher-flow branch receives a higher hematocrit.

304


This phenomenon was studied in the rat mesentery by Pries et al. [28], and a set of empirical relationships were developed to describe the dependence of phase separation on the vessel diameters and on the hematocrit of the parent vessel. These equations were subsequently modified slightly to render their predictions more robust for extreme combinations of input hematocrit and vessel diameter [27] as follows. The fractional flow of erythrocytes into one daughter branch FQE is calculated from the respective fractional blood flow FQB using logit FQE = A + B logit [(FQB í X0)/(1 í X0)]

(35)

where logit x = ln[x/(1íx)] and the parameters A, B and X0 defining the phase separation characteristics of the bifurcation were obtained from linear fits to experimental data [28]: A = í13.29 [(DĮ2/Dȕ2 í 1)/(DĮ2/Dȕ2 + 1)] (1 í HD)/DF

(36)

B = 1 + 6.98 (1 í HD)/DF

(37)

X0 = 0.964 (1 í HD)/DF .

(38)

Here, DĮ, Dȕ and DF are the diameters of the daughter branches and the mother vessel and HD is the discharge hematocrit in the mother vessel (i.e. the flow fraction of red blood cells in the blood passing through the vessel). The relationships presented above, in combination with the equations governing distribution of flow in networks [26, 29], provide a basis for simulating the hemodynamics of flow in microvascular networks, and for predicting the distributions of velocity, flow rate, wall shear stress, pressure and hematocrit.

3. Conclusions Poiseuille’s law, Eq. (1), remains a cornerstone in the study of hemodynamics. It expresses a basic relationship between the geometry of blood vessels, the rheological behavior of blood, and functionally critical variables such as blood pressure and flow. However, as shown in this chapter, Poiseuille’s law must be regarded as an approximation at best, and in many cases it leads to incorrect conclusions; for example, with regard to the distributions of velocity and stress under pulsatile flow conditions. Therefore, the study of hemodynamics requires consideration of the phenomena that are discussed above, and others not discussed here. Many of the main developments in research in this field occurred some time ago, and the period 1970-1980 was particularly productive with regard to the study of arterial flow dynamics. Hemodynamics remains an active area of research, and many recent publications apply advanced measurement and computational techniques to study detailed patterns of blood flow in specific vascular geometries. The importance of hemodynamic factors in determining biological behaviors of the cells and other components of the walls of blood vessels is well recognized. For instance, wall shear stress and intravascular pressure play critical roles in the acute regulation of blood flow [30] and in the long-term structural adaptation of blood vessels [31]. The development of atherosclerosis and the formation of thromboses are sensitively dependent on the


305

patterns of velocity and stress in arteries [32]. The biological and medical importance of such phenomena provides strong motivation for continued work in hemodynamics.

Acknowledgements This work was supported by NIH grant HL034555 and by the Deutsche Forschungsgemeinschaft.

References 1. A.C. Burton, Physiology and Biophysics of the Circulation. Year Book Medical Publishers, Chicago, 1972. 2. C.G. Caro, T.J. Pedley, R.C. Schroter and W.A. Seed, The Mechanics of the Circulation. Oxford University Press, Oxford, 1978. 3. Y.C. Fung, Biomechanics: Circulation. Second edition. Springer-Verlag, New York, 1997. 4. D.A. McDonald, Blood Flow in Arteries. Edward Arnold, London, 1974. 5. T.J. Pedley, The Fluid Mechanics of Large Blood Vessels. Cambridge University Press, 1980. 6. J.L.M. Poiseuille, Recherches expérimentales sur le mouvement des liquides dans les tubes de trèspetits diamêtres. Mémoires presentés par divers savants à l'Académie Royale des Sciences de l'Institut de France IX (2007), 433-544. 7. E. Hagenbach, Uber die Bestimmung der Zähigkeit einer Flüssigkeit durch den Ausfluss aus Röhren. Poggendorf's Annalen der Physik und Chemie 108 (1860), 385-426. 8. S.P. Sutera and R. Skalak, The History of Poiseuille's Law, Ann. Rev. Fluid Mech. 25 (1993), 1-19. 9. G.K. Batchelor, An Introduction to Fluid Mechanics. Cambridge University Press, 1967. 10. G. Hagen, Über die Bewegung des Wassers in engen cylindrischen Röhren, Ann. Phys. Chem. Poggendorf 46 (1839), 423-442. 11. M.G. Blyth, A.J. Mestel and L. Zabielski, Arterial bends: the development and decay of helical flows, Biorheology 39 (2002), 345-350. 12. P.J. Kilner, G.Z. Yang and D.N. Firmin, Morphodynamics of flow through sinuous curvatures of the heart, Biorheology 39 (2002), 409-417. 13. S. Rodbard, Vascular caliber, Cardiology 60 (1975), 4-49. 14. J.M. Greve, A.S. Les, B.T. Tang, M.T. Draney Blomme, N.M. Wilson, R.L. Dalman, N.J. Pelc and C.A. Taylor, Allometric scaling of wall shear stress from mice to humans: quantification using cine phasecontrast MRI and computational fluid dynamics, Am. J. Physiol. 291 (2006), H1700-H1708. 15. S. Kim, A.S. Popel, M. Intaglietta and P.C. Johnson, Effect of erythrocyte aggregation at normal human levels on functional capillary density in rat spinotrapezius muscle. Am. J. Physiol. 290 (2006), H941H947. 16. B.W. Zweifach, Quantitative studies of microcirculatory structure and function. II. Direct measurement of capillary pressure in splanchnic mesenteric vessels, Circ. Res. 34 (1974), 858-866. 17. J. Lindert, J. Werner, M. Redlin, H. Kuppe, H. Habazettl and A.R. Pries, OPS imaging of human microcirculation: a short technical report, J. Vasc. Res. 39 (2002), 368-372. 18. R. Fahraeus, Die Strömungsverhältnisse und die Verteilung der Blutzellen im GefäEsystem. Zur Frage der Bedeutung der intravasculären Erythrocytenaggregation, Klin. Wochenschr. 7 (1928), 100-106. 19. S.P. Sutera, V. Seshadri, P.A. Croce, and R.M. Hochmuth, Capillary blood flow. II. Deformable model cells in tube flow, Microvasc. Res. 2 (1970), 420-433. 20. A.R. Pries, D. Neuhaus, and P. Gaehtgens, Blood viscosity in tube flow: dependence on diameter and hematocrit, Am. J. Physiol. 263 (1992), H1770-H1778. 21. Secomb, T.W. 2003. Mechanics of red blood cells and blood flow in narrow tubes. In Hydrodynamics of Capsules and Cells. C. Pozrikidis, Ed., Chapman & Hall/CRC, Boca Raton, Florida. 163-96. 22. R. Fahraeus and T. Lindqvist, The viscosity of the blood in narrow capillary tubes, Am. J. Physiol. 96 (1931), 562-568. 23. P. Martini, A. Pierach and E. Schreyer, Die strömung des blutes in engen gefäEen. Eine abweichung vom Poiseuille'schen gesetz, Dt. Arch. Klin. Med. 169 (1930), 212-222. 24. A.R. Pries, T.W. Secomb, T. Gessner, M.B. Sperandio, J.F. Gross and P. Gaehtgens, Resistance to blood flow in microvessels in vivo, Circ. Res. 75 (1994), 904-915.

306


25. V. Vand, Viscosity of solutions and suspensions. I. Theory, J. Phys. Colloid Chem. 52 (1948), 277299. 26. A.R. Pries, T.W. Secomb, P. Gaehtgens and J.F. Gross, Blood flow in microvascular networks. Experiments and simulation, Circ. Res. 67 (1990), 826-834. 27. A.R. Pries and T.W. Secomb, Microvascular blood viscosity in vivo and the endothelial surface layer, Am. J. Physiol. 289 (2005), H2657-H2664. 28. A.R. Pries, K. Ley, M. Claassen and P. Gaehtgens, Red cell distribution at microvascular bifurcations, Microvasc. Res. 38 (1989), 81-101. 29. H.H. Lipowsky and B.W. Zweifach, Network analysis of microcirculation of cat mesentery, Microvasc. Res. 7 (1974), 73-83. 30. T.W. Secomb and A.R. Pries, Information transfer in microvascular networks, Microcirculation 9 (2002), 377-387. 31. A.R. Pries, B. Reglin and T.W. Secomb, Remodeling of blood vessels: responses of diameter and wall thickness to hemodynamic and metabolic stimuli, Hypertension 46 (2005), 725-731. 32. D.M. Wootton and D.N. Ku, Fluid mechanics of vascular systems, diseases, and thrombosis, Annu. Rev. Biomed. Eng. 1 (1999), 299-329.


307

Blood Rheology Aspects of the Microcirculation Herbert H. LIPOWSKY1 Department of Bioengineering The Pennsylvania State University University Park, PA 16802 USA

Introduction To elucidate the basis for the drop in pressure from large artery to vein, Jean Léonard Marie Poiseuille (1797-1869) undertook a series of seminal studies that defined the laws of fluid flow in tubes of uniform cross-section and that identified the microcirculation as the major site of the resistance to flow [1, 2]. While best known for his meticulous experimental studies of viscous flow in glass tubes [2], he also made numerous observations in the mesentery of the frog and other microvascular preparations; these observations served to highlight the dynamics of red cell distribution, the presence of the annular plasma layer and the “skimming” of plasma by capillaries, and the adhesion of white cells to the endothelium of post-capillary venules [3]. Yet it wasn’t until almost a century later that physiologists began to methodically explore the resistance to flow within the microcirculation proper using techniques of intravital microscopy. Landis [4] pioneered many quantitative methods for describing microvascular structure and function. Using finely drawn pipettes inserted into microvessels of frog mesentery, and calculating the velocity of bolus infusions of dyes through successive microvascular divisions, Landis attempted to calculate the resistance to flow within the capillary network and concluded that Poiseulle’s law cannot be applied to the flow of blood through the capillary network except in a very limited sense. However, with the advent of sophisticated instruments for measurement of capillary pressure [5, 6] and red blood cell velocity [7], relatively precise quantitative measurements of pressure drops [8] and flow rates [9, 10] could be obtained; such measurements thus provided direct in situ flow resistance data for the hierarchy of microvessels from arteriole to venule [11]. As shown in Figure 1, measurements of pressure drops and flows in single unbranched microvessels [11] reveal that over the broad range of diameters (D) within the microcirculation proper, Poiseuille’s fourth power relationship (i.e., resistance varies inversely with D4) far overshadows other determinants of resistance in the normal flow state. The significance of this relationship cannot be understated in light of the fact that control and regulation of microvascular blood flow is manifest by the ability of the vascular system to alter resistance in response to vasomotor adjustments. However, striking departures from Poiseuille’s relationship may dictate the outcome of 1 Corresponding author: Department of Bioengineering, The Pennsylvania State Universit, University Park, PA 16802 USA; E mail: [email protected]

308

H.H. Lipowsky / Blood Rheology Aspects of the Microcirculation

pathological flow states, with some examples including the low flow state, inflammation, and blood cell disorders. While the resistance to flow spans nearly five decades as blood courses its way from arteriole to venule, the large scatter in the experimental data may reflect significant departures from the flow of a Newtonian fluid through a smooth walled tube of circular cross-section. The effects of irregularities in geometry, broad variations in microvessel hematocrit and shear rates, blood cell deformability, red cell aggregation and blood cell adhesion to the endothelium are reviewed in the following.

R/L (mmHg/mm 3/s/Pm)

1000

100

10

1

.1

.01 60

50

40

ARTERIOLES

30

20

10 10

20

30

40

CAPS

50

60

VENULES

VESSEL DIAMETER ( Pm) Figure 1. Distribution of resistance per unit of vessel length (R/L) calculated for microvessels from direct in situ measurements of pressure drops and flows in intestinal mesentery. Power-law 6 -4.04 ) for arterioles and R / regressions (i.e., curved lines through data) reveal R / L = (1.02 x 10 )(D 6 -3.94 ) for venules. Redrawn from Lipowsky et al. [11]. L = (1.07 x 10 )(D

1. In Vitro Foundations Measurements of blood viscosity by bulk viscometry have emphasized the relative roles of shear rate, hematocrit, red cell aggregation and deformability in affecting blood flow in large vessels as well as establishing a framework for evaluating microcirculatory behavior [12]. With the use of tube, Couette and cone-plate viscometers and the calculation of an apparent viscosity as the ratio of shear stress divided by shear rate, in vitro studies have revealed that blood viscosity falls about 75% as shear rates ( J ) rise from § 0.1 to 1000 s-1. A comparison of this “shear thinning” of blood in the presence and absence of aggregating agents suggests that


309

about 75% of the decrease is a result of the disruption of red cell aggregates, and 25% is due to red cell deformation in response to increased shear stresses. At a given shear rate, blood viscosity rises exponentially with increasing red blood cell (RBC) volume fraction in the suspension (i.e., hematocrit), with the sensitivity dependent upon the prevailing J . The viscosity of the suspending medium in blood, termed plasma, has been shown to be Newtonian (i.e., invariant with J ) and is dependent mainly upon protein content and temperature. To relate bulk viscometry to blood behavior in tubes comparable in size to vessels of the microcirculation, numerous studies of flow in small bore tubes have provided a basis for understanding in vivo resistance to flow. The classical studies of Fåhraeus [13] demonstrating large reductions in tube hematocrit with diminishing tube diameter, and those of Fåhraeus and Lindqvist [14] showing concomitant reductions in the apparent viscosity of blood, have spawned many studies of blood viscosity in small bore tubes. In large part, these studies have been consistent with the relatively few direct in vivo measurements of apparent viscosity within single microvessels. To illustrate, Figure 2 shows measurements of wall shear stress vs. shear rate obtained in vitro by Barbee and Cokelet [15] for a glass tube with nominal diameter of 29 μm (dashed line) and in vivo for a 34 Pm diameter arteriole (solid line and data). In vivo values of wall shear stress ( W WALL ) were calculated from measurements of pressure drop ( 'P ),vessel length ( l ) and diameter ( D ) [8], under the assumption of a vessel of uniform diameter and circular cross-section, and application of the principle of static equilibrium such that W WALL 'P D / 4l . In vivo and in vitro measurements are consistent with one another, although the in vitro measurements were made at substantially lower wall shear rates. In vivo measurements of shear stress at shear rates below 300 s-1 are typically much higher that those obtained in vitro, thus suggesting a greater increase in apparent viscosity with reductions in shear rate in vivo [11]. These in vitro studies also demonstrated that shear stress-shear rate relations for glass tubes of microvessel dimensions could be described regardless of tube size provided that the tube hematocrit was correctly specified. It was thus hypothesized that given the correct tube hematocrit, the relationship between blood viscosity and shear rates in microvessels could be uniquely specified for tubes representative of arterioles and venules. Because of the complexities of red cell distribution and heterogeneity of hematocrit within the microvascular network, this hypothesis has not been validated in vivo. However, at the low levels of microvessel hematocrit typical of the microcirculation, a nearly linear relationship between viscosity and tube hematocrit has been suggested by bulk viscometric measurements [16]. Direct measurements of pressure drops and flows in small arterioles (24 – 47μm diameter) have revealed a linear relationship between apparent viscosity and hematocrit when microvessel hematocrits were varied by intentional hemodilution with cell free plasma over a range of hematocrits from 3 – 35% [17]. This relationship was identical to that obtained by in vitro viscometry using a Weissenberg rheogoniometer at comparable shear rates of 2000 s-1.

310


2

10

10

29

wall

(dyn/cm2 )

ARTERIOLE 34 m

m GLASS TUBE Hct = 8.8%

1 1

10

2

.

3

10

10 -1

(SEC

4

10

)

Figure 2. Comparison of wall shear stress (WWALL) vs shear rate (Ȗ) for measurements in a 29 μm glass tube [15] with in vivo measurements for a small 34 μm arteriole [11].

With diminishing vessel diameter, the particulate nature of blood dominates the resistance to flow. In vitro simulation of capillary flows using polycarbonate sieves with 5 μm pores was pioneered by Gregersen, Chien and Usami [18] and emphasized the dominant contribution of red cell deformability to perfusion of capillaries. Through such experimental simulations it is now recognized that the initial deformation that red blood cells (RBC) and white blood cells (WBC) incur upon entry to a capillary contributes significantly to the pressure drop across individual capillaries [19-21]. Thus, the contribution of the intrinsic properties of blood to hemodynamic resistance changes markedly throughout the succession of vessels from arterioles to capillaries as diameter changes between divisions.

2. In Vivo Determinants of Blood Viscosity and Resistance 2.1. Blood Cell Deformability The contribution of blood cell deformability to hemodynamic resistance varies markedly throughout the succession of vessels from arterioles to capillaries as diameter varies between divisions. Blood cell deformability affects the entrance of blood cells into capillaries and RBC with reduced deformability in pathological disorders (e.g., sickle cell disease) may be sequestered at the capillary entrance. Stiffening of the red cell membrane or elevations in hemoglobin viscosity may impede RBC transit through capillaries [22-24]. Studies of the flow of RBC with impaired deformability have been

311


performed by infusing hardened cells into the subject animal and observing flow effects by intravital microscopy. For example, by hardening the RBC membrane with diamide and then infusing these cells into a rat, it was found that flow velocities in mesenteric capillaries were reduced by 30% under normal perfusion pressures and vascular stasis ensued following reductions in perfusion pressure[25]. Similar studies using RBC partially hardened with glutaraldehyde found that cells were preferentially sequestered in non-tube-like, sinusoidal vessels with preference for sequestration in bone marrow, liver, lung and spleen[26] .The normally stiffer WBC traverse the capillary network through larger thoroughfare channels [27], and WBC may become trapped at the capillary entrance or incur a prolonged transit time following their stiffening with activation during inflammation[28, 29]. 5

TTrbc/TTpl

4 ȕE >$10 10

3 10 ȕE<< 10

2

1

0 0

2

4

6

8

E

10

12

14

16

18

20

Figure 3. Mean Transit Time (TT) of hardened RBC relative to plasma for the transit from terminal arteriole to collecting venule across the capillary bed of cremaster muscle [30]. Prior to the infusion of bolus injections RBC deformability was measured by the steady state filtration method; the filtration parameter ȕ was computed from the ratio of pressure drop across 5 Pm filters for cells and suspending medium according to the method of Skalak et al. [31].

Use of the in vitro steady flow micropore filtration technique [18] to derive a quantitative assessment of cell deformability has placed in vivo observations on a firmer foundation. For example, using the deformability parameter, ȕ, defined as the ratio of resistance to flow through a filter pore in the presence of red cells to that for suspending medium alone, it has been possible to understand how cell deformability affects RBC [30] and WBC [27] transit through the capillary bed. For example, shown in Figure 3 is the variation of the ratio of mean transit time of red cells through the capillary network in cremaster muscle, normalized to the mean transit time of plasma through the same vascular segment [30]. Transit time (TT) was computed from analysis of intensity-time curves following bolus injections of fluorescently labeled RBC or dextran dissolved in plasma. Intensity-time curves in terminal arterioles and collecting venules were recorded following bolus infusions of RBC hardened to various degrees with glutaraldehyde, with RBC deformability characterized by filtration. As shown, with increases in ȕ from the norm of 2.4 up to 10.0, RBC traverse the capillary network with little hindrance and are able to recruit additional pathways through the capillary

312


network as needed. However, as ȕ increases further, the transit time increases dramatically due to: (a) reductions in red cell velocity with increased rigidity; (b) reductions in the number of available parallel pathways due to sequestration of cells in the smallest diameter capillaries of the segment. 12

1.0 (A)

(B)

Fraction Trapped

FLOW (ml/min)

10 8 6 4 2 0

0.6

0.4

0.2

0.0 0

2

4

6

0

2

4

6

0.4

1.0

(C)

(D)

0.8

Frequency

Cumulative Probability

0.8

0.6 Log-Normal Fit 0.4

0.3

Mean = .25 ± 1.72 SD

0.2

0.1 0.2 0.0

0.0 0

2

4

Pressure (cmH2O)

6

0

1

2

3

4

Pressure (cmH2O)

Figure 4. Transient filtration of leukocytes through 5 μm pore polycarbonate filters reveal the heterogeneity of cell deformability [35]. (A) A Bolus of WBC was initially trapped in the majority of pores in the filter and the flow steadily increased. (B) As pressure drop across the filter increased with flow, the fraction of cells trapped in pores decreased. (C) The cumulative probability of a cell passing through the filter at a given pressure drop was calculated as 1 – Fraction Trapped, and fit with a log-normal distribution. (D) Frequency distribution of yield pressures required to force a WBC through the pores.

While the steady state filtration technique provides a quantitative measure of cell deformability, it does not model the relationship between cell entrapment and the heterogeneity in size and deformability of the cell population. It has been shown that as flow rate increased through 5 μm pores of polycarbonate filters, the retention of WBC decreases exponentially with total flow through the filter [32]. In vivo studies of the sequestration of WBC in the pulmonary capillaries have shown that infused radiolabelled WBC exhibited an exponential washout with time that could be prolonged by hardening the WBC with glutaraldehyde prior to infusion [33]. Biophysical studies of the entry of WBC into micropipettes reveal that there exists a critical yield pressure (Py) which must be exceeded for a WBC of a given diameter to enter a micropipette of a specific size and that the independent geometric variable is the ratio of cell to pipette diameter [34]. To illustrate the effects of heterogeneity in cell deformability and diameter for a given population of WBC, and the heterogeneity of pore diameter in a polycarbonate filter, the transient filtration test was analyzed to determine a mean filtration pressure [35]. Shown in Figure 4A is the variation of total flow through a filter vs. pressure drop following the initial entrapment of a population of cells approximately equal in number to the total number of pores in the filter. As the

313


pressure drop increases from zero, flow initially increases non-linearly. The fraction of cells trapped at a given pressure (Figure 4B) was computed from the total number of pores and the resistance to flow ('P/Q) measured during the washout of cells. Using this relationship to define the probability that a WBC remains trapped (pr) for all pressures below the value on the abscissa, the probability that a cell will pass through the filter was then taken as 1-pr and fit with a log-normal distribution (panel 4C). The probability density function that represents the frequency distribution of yield pressures, Py, was then computed (panel 4D). Under these idealized assumptions, it is evident that while the mean yield pressure is a relatively low value (i.e., mean = 0.25 cm H2O), the standard deviation is relatively large and may represent the summated effects of heterogeneity in deformability as well as the ratio of cell to pore diameter.

Yield Pressure (cm H2O)

0.8

0.6

0.4

0.2

0.0 1.2

1.3

1.4

1.5

1.6

1.7

OCell Dia/Pore Dia Figure 5. Mean yield pressure determined for a population of rat WBC as measured by the transient filtration method. The filter pores were enlarged by etching the filter in concentrated NaOH. The ratio of cell to pore diameter was calculated for the measured mean diameter of cells and filter pores; mean diameters determined by orifice-type electronic sizing and pore diameter via hydrodynamic analysis.

To assess the effect of varying pore diameter, polycarbonate micropore filters were etched with NaOH to enlarge the pores. The mean pore diameter (DPORE) was then calculated from the measured pressure drop and flow of buffer through the filter, assuming Poiseuille’s law holds for each of the measured total number of pores in the filter for a filter of known thickness (i.e., pore length). The mean cell diameter (DCELL) of WBC in the suspension was determined using a Coulter counter calibrated against microspheres of known size. The resulting trends of mean yield pressure as a function of Ȝ= DCELL/DPORE are shown in Figure 5. Over the range of 1.3 Ȝ 1.5 there appears to be a relative plateau in mean yield pressure. This plateau appears to arise from the deformation of the WBC with greater pressure drops and flow rates through the filter as well as cells being swept through the pathways of least resistance (i.e., larger filter pores) at a given pressure drop. Such behavior has been noted in vivo as bolus infusions of WBC traverse the capillary network in cremaster muscle [27]. As shown in these muscle studies, WBC tend to take more centralized pathways through the

314


tissue via thoroughfare channels or the capillaries with the largest diameter. As Ȝ increases above 1,5, a dramatic rise in the mean filtration pressure occurs, which is indicative of the propensity for cell sequestration with stiffened WBC during activation during the inflammatory process [27, 29, 36]. 2.2. Red Cell Aggregation The role of red cell aggregation (RCA) in affecting the resistance to flow in microvessels has not been fully clarified. Although in vitro viscometry demonstrates increases in blood viscosity with reductions in shear rate ( J ) [37], the resistance to flow in small-bore tubes does not necessarily exhibit the same correlation. In vitro studies with glass tubes comparable in size to microvessels [38] have suggested that the apparent viscosity of blood may decrease with onset of RCA, presumably because of exclusion of aggregates from the tube entrance or their radial migration toward the axial core of the tube. In contrast, direct in vivo hemodynamic measurements in normal [39] and low-flow states [40] reveal a dramatic rise in apparent viscosity and increased resistance with reductions in shear rate. It has been suggested that the behavior of aggregates at the entrance to successive arteriolar bifurcations and venular confluences may be a source of increased resistance to explain the basis for this disparity. It has also been postulated that the increased viscosity that arises from RCA may be negated by the migration of red cell aggregates to the central axial core of small vessels [41]. However, in vivo studies of the trajectories of red cell aggregates in venous vessels reveal that this effect is small and that a blunting of the velocity profile may be the dominant effect [41]. Analysis of the in vivo effect of RCA on resistance has been impeded by the inability to quantitatively define the degree of aggregation and the variability of aggregate strength obtained by different means. The use of high molecular weight dextrans (70–150 kDa) to induce RCA has been frequently used to induce RCA for studies of changes in flow and resistance [40-44]. In vitro and theoretical studies have provided a biophysical foundation for understanding the relationship between shearing forces and the strength of red cell aggregation. It has been shown that as the strength of aggregation increases, RBC form rouleaux and then clumps [45]. In vivo studies have demonstrated that the rouleaux are much more easily disrupted at bifurcations whereas clumps may become lodged at the entrance to capillaries [46]. A major problem with elucidating the effects of aggregation by direct intravital microscopy of the microcirculation has been that, as aggregation increases (e.g., dextran infusion), red cells become trapped in organs outside the field of view and hematocrit falls dramatically. Measurements of pressure drop from feeding arteriole (PA) to collecting venule (PV) in the modular network of rabbit omentum have shown that resistance, calculated from 'PAV/QARTERIOLE, first rises by about 10% as the circulating concentration of 500 kDa dextran increases to 1 gm%, then falls to 50% of normal as the concentration reaches 2 gm% [47]. This entrapment of cells in tissues outside the field of view suggests that the greatest influence of RCA is not its ability to increase blood viscosity, but rather its effect on cell sequestration at branch points in the microcirculation. To illustrate the variability of hematocrit and aggregate formation observed in the microvasculature, Figures 6A and 6B show red cells flowing at reduced flow rate in post-capillary venules of rat mesentery. Induction of RCA by infusing fibrinogen into

315


the systemic circulation of the rat, up to its maximum solubility of around 0.7 gm% which is the upper limit observed in clinical disorders, results in rouleaux formation in post-capillary venules during an induced low flow state. In contrast, the formation of clump-type RBC aggregates by infusion of 500 kDa dextran results in dramatic reductions in microvessel hematocrit. On average, microvessel hematocrit changed insignificantly due to fibrinogen but in response to dextran the packed cell fraction declined by almost 40% with a circulating concentration of 3 gm% [46, 48]. 4

WBC

FLOW

(B) Dextran FLOW

G, Aggregation Index

(A) Fibrinogen

(C) *† Control: 0.07 ± 0.07 g% Fb (n=21)

3

0.74 ± 0.24 g% Fb (n=9)

*†

3.0 ± 0.5 g% Dx500 (n=12)

2 * *

*

1 WBC

50 μm 0

100

200 .

300

400

500

J, Shear Rate (s

-1

600

)

Figure 6. (A) Rouleaux formation of red cells in a post-capillary venule observed in the mesentery of the rat during reductions in flow achieved by upstream occlusion with a blunted micropipette for estimated wall shear rates < 50 s-1. (B) Formation of clumped aggregates of red cells following infusion of 500 kDa dextran. (C) Aggregation index computed from light scattering measurements in post-capillary venules as a function of estimated wall shear rate, J , calculated from measured red cell velocity (VMEAN) and vessel diameter (D) as 8VMEAN/D). Shear rates were reduced by a proximal occlusion of the feeding arterioles with a blunted micropipette. From [46] with permission.

Numerous techniques have been devised to quantitate red cell aggregation in a flow field, with applications to either in vitro flow models or direct microvascular measurements: estimates of aggregate size have been extracted from direct measurement of clump or aggregate size [44, 49], Fourier analysis of spatial variations in light intensity [50], analysis of transmitted and reflected light [51], and measurement of the light scattering properties of blood [48]. To illustrate the variability of RCA with shear rates in microvessels, Figure 6C presents an index of aggregation (G) computed from the light scattering of red blood cells [46]. The parameter G was computed from measurements of the scattering component of optical density, ODSCAT.. Assuming that the effective size of a particle could be deduced from theoretical considerations of the relationship between ODSCAT and the surface area to volume ratio of an aggregate, G then corresponds to the number of particles (i.e., individual red cells) per aggregate. With reductions in shear rates by occlusion of proximal microvessels, it appears that the dextran-induced aggregation is greater and more sensitive to shear rate compared with fibrinogen-induced aggregation. Observations of the disruption of aggregates at arteriolar to capillary bifurcations suggest that the rouleaux formed with fibrinogen are much more easily dispersed compared with the

316


clumps resulting from the 500 kDa dextran. In terms of hydrodynamics, the greatest difference between the two types of aggregation (i.e., fibrinogen vs. 500 kDa dextran) appears to be the effects on hematocrit and WBC margination. Although the fibrinogen-induced aggregation is much less in magnitude compared to dextran, the maintenance of a higher tube hematocrit with fibrinogen-induced RCA results in an enhanced radial migration of WBC and increased WBC flux along the venular wall [46]. In the case of dextran, the formation of large plasma gaps (Figure 6B) between aggregates tends to trap WBC and lessen their frequency of contact with the endothelium. The effects of the intensity of aggregation and the size of aggregates on the resistance to flow remains to be explored. It has been postulated that disruption of aggregates as blood flows through microvessels having an irregular cross-section that varies with vessel length may cause an increase in energy dissipation [50]. This scenario has been implicated previously as a possible cause of the rise in measured effective viscosity of blood in arterioles and venules in response to reductions in shear rates [11]. Although aggregation measurements were not made in these studies, a fourfold increase in effective viscosity of blood was found in arterioles and venules with reductions in mean flow velocity from 4 to 0.2 mm/s. These large increases of in vivo viscosity were also attributed to the presence of leukocyte-endothelium adhesion, which affected the resistance to flow by obstructing the lumen. It was subsequently shown that when as few as 10 WBC adhere to the endothelium per 100 μm of vessel length in post-capillary venules, the resistance to flow may increase 2-fold [52]. The potential for enhanced WBC flux due to RCA may thus contribute to increased resistance to flow due to elevated WBC-endothelial cell adhesion. 2.5 mean ± SE

EAGG/ERBC

2.0

1.5

1.0

0.5

0.0 0

2

4

6

8

10

% Dextran 150 Figure 7. Steady state filtration test of red cell flow through 5Pm pore polycarbonate filters. The filtration parameter ȕ represents the ratio of the resistance to flow through a pore with red cells present to that with buffer alone [31] and is normalized to its value in a non-aggregating suspension (i.e., RBC in Ringers solution). Red cells from rat were suspended in Ringers solution containing the indicated concentration of 150 kDa dextran. The curve is a least squares quadratic fit of the data.

The effect of RCA on the ability of RBC to enter capillaries is also difficult to measure in vivo. However, in vitro tests using the steady state filtration method with 5


317

μm pore polycarbonate filters may provide new insights into abnormal increases in resistance attendant to RCA. Illustrated in Figure 7 are preliminary data for rat red cells resuspended in Ringers solution containing 150 kDa dextran. As shown, the steady state filtration parameter ȕ (i.e., ratio of resistance to flow within a pore in the presence of RBC to resistance with buffer alone) increases by about 50% to a maximum at a concentration of dextran on the order of 3 gm%. If these relations apply in vivo, one would need a 50% increase of pressure gradient at the capillary level to maintain flow in capillaries. 2.3. Vascular Geometry It must be emphasized that all rheological and flow measurements within the microvasculature are subject to the uncertainties associated with an imprecise knowledge of the vascular geometry. For example, when computing the effective viscosity of blood from measurements of pressure drop and flow it is assumed that the vessel is a smooth-walled cylinder of circular cross-section. Histological evidence clearly suggests otherwise. For example, many venules appear to be elliptical in crosssection, and this departure from the ideal is ignored when measuring venular diameter from observations of the width of a venule in the focal plane of a living tissue. Furthermore, normal vascular tone may change and the shape of the lumen may change dramatically with increased smooth muscle tone [53]. In previous studies measuring the in vivo apparent viscosity of blood in the microcirculation [11], it was found that the viscosity was 43% greater in venules compared to arterioles (5.15 vs 3.59 mPa.s, respectively). This disparity was attributed, in part, to irregularities in the microvessel walls, greater red cell aggregation due to the lower shear rates in venules, low levels of WBC-EC adhesion in the postcapillary venules and possibly unequal hematocrits. To determine the errors incurred by neglecting irregularities in vascular diameter, studies have been performed to view the cross-section of venules in mesenteric tissue by using techniques of optical sectioning microscopy. By digitizing a stack of up to 50 images taken in 0.5 μm increments along the optical axis (i.e., z- direction), the luminal shape may be computed normal to the longitudinal axis of the vessel [54]. Shown in Figure 8 is the resulting cross-section of a post-capillary venule in a rat mesentery taken under fluorescence microscopy. The elliptical shape and other departures from a circular cross-section are clearly noted. One can account for non-circular shapes in the measurement of flow resistance from Poiseuille’s law by using the hydraulic diameter, DH. The hydraulic diameter is defined as the diameter of an equivalent tube of circular cross-section that would have the same pressure drop for the same level of wall shear stress acting on tube or vessel walls having a wetted area S, and is defined by the relation: DH = 4A/S, where A is the cross-sectional area of the vessel. For six venules ranging in width in the x-y focal plane from 9.0 to 23.0 Pm with an average width of 17.5 ± 4.8 Pm (mean ± SD), the average DH was 15.9 ± 4.0 μm. Thus, due to the fourth power relationship between diameter and resistance given by Poiseuille's Law, the apparent 10% over estimation of luminal diameter using the width of the microvessel could result in a 40% overestimate of the apparent viscosity in vivo. The significance of these findings lies in the fact that earlier calculations of in vivo apparent viscosity [11] revealed a 50% greater apparent viscosity in postcapillary venules compared with arterioles in mesentery of the cat. Hence, given that arterioles are generally much more circular than venules in their nonvasoconstricted states [53], it is likely that irregularities in diameter may have played a

318


role in elevating the venular apparent viscosity. It would appear that a systematic analysis of the cross-sectional shape of small blood vessels is needed to fully understand resistance changes in normal and pathological flow states.

10P m Figure 8. Computer reconstruction of the cross-section of a post-capillary venule in the mesentery of the rat. The venular endothelium was stained with the fluorescent dye acridine orange and a stack of 50 images digitized in the optical axis z- direction. The cross-section was computed by deconvolving the stack of x-y images using the techniques of optical sectioning microscopy [54]. The lumen of the vessel is outlined by the white line.

In addition to these departures from the ideal, the presence of the macromolecular surface layer on the endothelium [55] has also been implicated as being hemodynamically significant [56]. The presence of a 0.3–1.0 Pm thick layer (i.e., the glycocalyx) composed of proteoglycans and polysaccharides has been implicated as a significant source of increased resistance to flow in capillaries [56, 57]. Numerical simulations of blood flow through capillaries [58] suggest that red cell movement through a capillary may be hindered by the glycocalyx. Experimental measurements of the velocity profile within small venules suggest the presence of a significant displacement thickness due to the glycocalyx on the endothelium [59]. It has also been shown that the composition of the glycocalyx, and presumably its thickness, may be altered due to changes in wall shear stresses and/or enzymatic degradation during the inflammatory process [60]. The importance of these observations in affecting the resistance to flow in normal and pathological flow states remains to be explored.

3. Conclusions The evolution of knowledge of the rheological behavior of blood flow in the microcirculation has provided a solid foundation for interpreting in vivo observations in normal and pathological flow states. The development of new instrumentation and approaches has paralleled the acquisition of rheological data and provided many new tools to bear on the problem. Many problems remain to be explored on both the cellular and molecular levels. The general approach to date has been to make observations at the cellular and/or molecular level and interpret them within the framework of a continuum (e.g., Poiseuille’s law). However, with the advent of major advances in computational power, and new techniques of microscopy such as multiphoton spectroscopy or other fluorescence techniques, there is greater potential for addressing rheological questions at the molecular level and more completely


319

understanding the mechanics of microvascular perfusion. Such studies would facilitate the integration of the rheological properties of blood with mechanical forces that affect cell signaling events which, in turn, affect microvascular function in health and disease.

Acknowledgements Supported by NIH-NHLBI Research Grant R01 HL39286.

References [1] [2] [3] [4] [5] [6] [7] [8]

[9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]

J.R. Pappenheimer. Contributions to microvascular research of Jean Leonard Marie Poiseuille. In: APS Handbook on Microcirculation, Vol I., E.M.Renkin and C.C.Michel, Eds. 1984, pp. 1-10. S. Sutera and R. Skalak, The History of Poiseuille's Law, Annu. Rev. Fluid Mech. 25 (1993), 1-19. J.L.M. Poiseuille, Recherches sur les causes du mouvement du sang dans les vaisseaux capillaries., C. R. Acad Sci. 6 (1835), 554-560. E.M. Landis, Poiseulle's law and the capillary circulation., Am. J. Physiol. 103 (1933), 432-443. M. Intaglietta, R.F. Pawula and W.R. Tompkins, Pressure measurements in the mammalian microvasculature, Microvasc. Res. 2 (1970), 212-220. C.A. Wiederhielm, J.W. Woodburry, S. Kirk and R.F. Rushmer , Pulsatile pressures in the microcirculation of frog's mesentery, Am. J. Physiol. 207 (1964), 173-176. H. Wayland and P.C. Johnson, Erythrocyte velocity measurement in microvessels by a correlation method, Bibl. Anat. 9 (1967), 160-163. H.H. Lipowsky and B.W. Zweifach, Methods for the simultaneous measurement of pressure differentials and flow in single unbranched vessels of the microcirculation for rheological studies, Microvasc. Res. 14 (1977), 345-361. M. Baker and H. Wayland, On-line volume flow rate and velocity profile measurement for blood in microvessels, Microvasc. Res. 7 (1974), 131-143. H.H. Lipowsky and B.W. Zweifach, Application of the "two-slit" photometric technique to the measurement of microvascular volumetric flow rates, Microvasc. Res. 15 (1978), 93-101. H.H. Lipowsky, S. Kovalcheck and B.W. Zweifach, The distribution of blood rheological parameters in the microvasculature of cat mesentery, Circ. Res. 43 (1978), 738-749. S. Chien, S. Usami, J. Dellenback and V. Magazinovic, Blood rheology after hemorrhage and endotoxin, Adv. Exp. Med. Biol. 33 (1972), 75-93. R. Fahraeus, The suspension stability of blood, Physiol. Rev. 9 (1929), 241-274. R. Fahraeus and T. Lindqvist, The viscosity of blood in narrow capillary tubes., Am. J. Physiol. 96 (1931), 562-568. J.H. Barbee and G.R. Cokelet, Prediction of blood flow in tubes with diameters as small as 29 microns, Microvasc. Res. 3 (1971), 17-21. S. Usami, S. Chien and M.I. Gregersen, Viscometric characteristics of blood of the elephant, man, dog, sheep, and goat, Am. J. Physiol. 217 (1969), 884-890. H.H. Lipowsky, S. Usami and S. Chien, In vivo measurements of "apparent viscosity" and microvessel hematocrit in the mesentery of the cat, Microvasc. Res. 19 (1980), 297-319. M.I. Gregersen, C.A. Bryant, W.E. Hammerle, S. Usami and S. Chien, Flow Characteristics of Human Erythrocytes through Polycarbonate Sieves, Science 157 (1967), 825-827. R. Skalak, L. Soslowsky, E. Schmalzer, T. Impelluso and S. Chien, Theory of filtration of mixed blood suspensions, Biorheology 24 (1987), 35-52. P.S. Lingard, Capillary pore rheology of erythrocytes. 1. Hydroelastic behaviour of human erythrocytes, Microvasc. Res. 8 (1974), 53-63. G.W. Schmid-Schonbein, R. Skalak, S. Usami and S. Chien, Cell distribution in capillary networks, Microvasc. Res. 19 (1980), 18-44. T.S. Hakim, Effect of erythrocyte heat treatment on pulmonary vascular resistance, Microvasc. Res. 48 (1994), 13-25. D.K. Kaul and M.E. Fabry, In vivo studies of sickle red blood cells, Microcirculation 11 (2004), 153165. K. Parthasarathi and H.H. Lipowsky, Capillary recruitment in response to tissue hypoxia and its dependence on red blood cell deformability, Am. J. Physiol. 277 (1999), H2145-H2157.

320


[25] G.K. Driessen, T.M. Fischer, C.W. Haest, W. Inhoffen and H. Schmid-Schonbein, Flow behaviour of rigid red blood cells in the microcirculation, Int. J. Microcirc. Clin. Exp. 3 (1984), 197-210. [26] S. Simchon, K.M. Jan and S. Chien, Influence of Reduced Red-Cell Deformability on Regional BloodFlow, Am. J. Physiol. 253 (1987), H898-H903. [27] M.J. Eppihimer and H.H. Lipowsky, Leukocyte sequestration in the microvasculature in normal and low flow states, Am. J. Physiol 267 (1994), H1122-H1134. [28] M. Bathe, A. Shirai, C.M. Doerschuk and R.D. Kamm, Neutrophil transit times through pulmonary capillaries: the effects of capillary geometry and fMLP-stimulation, Biophys. J. 83 (2002), 1917-1933. [29] A.G. Harris and T.C. Skalak, Effects of leukocyte capillary plugging in skeletal muscle ischemiareperfusion injury, Am. J. Physiol 271 (1996), H2653-H2660. [30] H.H. Lipowsky, L.E. Cram, W. Justice and M.J. Eppihimer, Effect of erythrocyte deformability on in vivo red cell transit time and hematocrit and their correlation with in vitro filterability, Microvasc. Res. 46 (1993), 43-64. [31] R. Skalak, T. Impelluso, E.A. Schmalzer and S. Chien, Theoretical modeling of filtration of blood cell suspensions, Biorheology 20 (1983), 41-56. [32] G.P. Downey and G.S. Worthen, Neutrophil retention in model capillaries: deformability, geometry, and hydrodynamic forces, J Appl. Physiol 65 (1988), 1861-1871. [33] G.P. Downey, D.E. Doherty, B. Schwab, III, E.L. Elson, P.M. Henson and G.S. Worthen, Retention of leukocytes in capillaries: role of cell size and deformability, J Appl. Physiol 69 (1990), 1767-1778. [34] E. Evans and A. Yeung, Apparent viscosity and cortical tension of blood granulocytes determined by micropipet aspiration, Biophys. J. 56 (1989), 151-160. [35] M.J. Eppihimer and H.H. Lipowsky, The mean filtration pressure of leukocyte suspensions and its relation to the passage of leukocytes through nuclepore filters and capillary networks, Microcirculation 1 (1994), 237-250. [36] M.J. Eppihimer and H.H. Lipowsky, Effects of leukocyte-capillary plugging on the resistance to flow in the microvasculature of cremaster muscle for normal and activated leukocytes, Microvasc. Res. 51 (1996), 187-201. [37] Chien S. Biophysical behavior of red cells in suspensions. In: The Red Blood Cell, D.MacN Surgenor, Ed. New York: Academic Press, 1975, pp. 1031-1133. [38] W. Reinke, P. Gaehtgens and P.C. Johnson, Blood viscosity in small tubes: effect of shear rate, aggregation, and sedimentation, Am. J. Physiol 253 (1987), H540-H547. [39] G. Mchedlishvili, L. Gobejishvili and N. Beritashvili, Effect of intensified red blood cell aggregability on arterial pressure and mesenteric microcirculation, Microvasc. Res. 45 (1993), 233-242. [40] M. Cabel, H.J. Meiselman, A.S. Popel and P.C. Johnson, Contribution of red blood cell aggregation to venous vascular resistance in skeletal muscle, Am. J. Physiol 272 (1997), H1020-H1032. [41] P.C. Johnson, J.J. Bishop, S. Popel and M. Intaglietta, Effects of red cell aggregation on the venous microcirculation, Biorheology 36 (1999), 457-460. [42] O.K. Baskurt and H.J. Meiselman, Blood rheology and hemodynamics, Semin. Thromb. Hemost. 29 (2003), 435-450. [43] J.J. Bishop, A.S. Popel, M. Intaglietta and P.C. Johnson, Rheological effects of red blood cell aggregation in the venous network: a review of recent studies, Biorheology 38 (2001), 263-274. [44] J.J. Bishop, P.R. Nance, A.S. Popel, M. Intaglietta and P.C. Johnson, Relationship between erythrocyte aggregate size and flow rate in skeletal muscle venules, Am. J. Physiol. 286 (2004), H113-H120. [45] R. Skalak, P.R. Zarda, K.M. Jan and S. Chien, Mechanics of Rouleau formation, Biophys. J. 35 (1981), 771-781. [46] M.J. Pearson and H.H. Lipowsky, Effect of fibrinogen on leukocyte margination and adhesion in postcapillary venules, Microcirculation 11 (2004), 295-306. [47] H.H. Lipowsky, Microvascular rheology and hemodynamics, Microcirculation 12 (2005), 5-15. [48] M.J. Pearson and H.H. Lipowsky, Influence of erythrocyte aggregation on leukocyte margination in postcapillary venules of rat mesentery, Am. J. Physiol. 279 (2000), H1460-H1471. [49] S. Berliner, R. Ben Ami, D. Samocha-Bonet, S. Abu-Abeid, V. Schechner, Y. Beigel, I. Shapira, S. Yedgar and G. Barshtein, The degree of red blood cell aggregation on peripheral blood glass slides corresponds to inter-erythrocyte cohesive forces in laminar flow, Thromb. Res. 114 (2004), 37-44. [50] K. Osterloh, P. Gaehtgens and A.R. Pries, Determination of microvascular flow pattern formation in vivo, Am. J. Physiol. 278 (2000), H1142-H1152. [51] A. Gaspar-Rosas and G.B. Thurston, Erythrocyte aggregate rheology by transmitted and reflected light, Biorheology 25 (1988), 471-487. [52] S.D. House and H.H. Lipowsky, Leukocyte-endothelium adhesion: microhemodynamics in mesentery of the cat, Microvasc. Res. 34 (1987), 363-379. [53] J.E. Greensmith and B.R. Duling, Morphology of the constricted arteriolar wall: physiological implications, Am. J. Physiol 247 (1984), H687-H698.


321

[54] Z. Shen and H.H. Lipowsky, Image enhancement of the in vivo leukocyte-endothelium contact zone using optical sectioning microscopy, Ann. Biomed. Eng 25 (1997), 521-535. [55] A.R. Pries, T.W. Secomb and P. Gaehtgens, The endothelial surface layer, Pflugers Arch. 440 (2000), 653-666. [56] A.R. Pries, T.W. Secomb, H. Jacobs, M. Sperandio, K. Osterloh and P. Gaehtgens, Microvascular blood flow resistance: role of endothelial surface layer, Am. J. Physiol 273 (1997), H2272-H2279. [57] A.R. Pries and T.W. Secomb, Rheology of the microcirculation, Clin. Hemorheol. Microcirc. 29 (2003), 143-148. [58] T.W. Secomb, R. Hsu and A.R. Pries, Motion of red blood cells in a capillary with an endothelial surface layer: effect of flow velocity, Am. J. Physiol. 281 (2001), H629-H636. [59] E.R. Damiano, The effect of the endothelial-cell glycocalyx on the motion of red blood cells through capillaries, Microvasc. Res. 55 (1998), 77-91. [60] A.W. Mulivor and H.H. Lipowsky, Inflammation- and ischemia-induced shedding of venular glycocalyx, Am. J. Physiol. 286 (2004), H1672-H1680.

322


In Vivo Hemorheology Oguz K. BASKURTa,1 and Herbert J. MEISELMANb Department of Physiology, Faculty of Medicine, Akdeniz University, Antalya, Turkey and bDepartment of Physiology and Biophysics, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA a

Introduction Relations describing fluid flow in cylindrical tubes were formulated by Poiseuille approximately 150 years ago, resulting in the well-known Poiseuille Equation. The experiments underlying this formulation were conducted using simple fluids and the viscosity concept was introduced in the equation as a constant, being directly proportional to flow resistance. This was an oversimplification for blood flow and experimental work in the early 1900’s revealed that blood viscosity could not be represented with a constant, but rather depended on flow conditions (i.e., tube diameter, flow rate). Blood viscosity expressed as apparent viscosity (i.e., the empirical ratio of the volumes of water and blood which would flow in a given time under the same specified conditions) was reported to take values between 5-100, depending on the velocity of flow and diameter of the tube [1]. This early understanding of blood rheology that dominated the first several decades of 20th century was clearly described in the famous publication of Whittaker and Winton [1]. This publication is not primarily famous for this description of the shear rate dependence of blood viscosity, but rather because it pointed out that measurements of blood viscosity in cylindrical tubes could not be used to predict its effects on in vivo blood flow [1]. Whittaker and Winton compared the apparent viscosity of blood under a constant pressure difference as determined by simultaneously measuring flow through a dog hind limb preparation and a glass viscometer arranged in parallel. The results clearly indicated that the apparent viscosity of blood determined using the flow rate through the hind limb was lower than the value obtained using the glass viscometer (Figure 1). The differences between apparent viscosity values measured in vivo and ex vivo were more prominent at higher hematocrit values [1]. This approach for calculating apparent viscosity using pressure drop and flow rate data measured in vivo has been used by other investigators under various conditions and in different experimental settings [2, 3]. These investigations were usually done using isolated organs [2] or specially designed arterio-venous shunts to allow the measurements of pressure and flow through a calibrated tube [3]. Additionally, the role of hemorheological parameters in determining pressure-flow relationship in various organs [4-8] or under various pathophysiological conditions [9] have been investigated.

1 Corresponding author: Department of Physiology, Akdeniz University Faculty of Medicine, Antalya, Turkey; E Mail: [email protected].

O.K. Baskurt and H.J. Meiselman / In Vivo Hemorheology

323

Apparent viscosity

15 Glass viscometer

12 9 6

Hind limb 3 0 0

20 40 60 Hematocrit (%)

80

Figure 1. Apparent viscosity of blood with various hematocrit values determined simultaneously during flow through a glass viscometer and a dog hind limb preparation. The curves represent mean ± 95% confidence interval. Redrawn from Whittaker and Winton [1].

1. Pressure-Flow Relationships In Vivo As mentioned above, comparisons of pressure-flow curves obtained in vivo and ex vivo indicated lower apparent viscosity of blood when studied in its natural in vivo flow environment. Additionally, it was demonstrated that the impact of alterations of the factors that determine blood viscosity may be specific to the organ or tissue under investigation. An obvious reason for such differences is that the response to a given hemorheological alteration should be related to vascular geometry, and thus specific geometric features of the vasculature of various organs and tissue are the underlying factors for these differences. The term “vascular hindrance” has been used to differentiate between the geometric and rheological components of flow resistance [4]. It should also be kept in mind when analyzing in vivo pressure-flow data that vascular hindrance is a variable, not a constant, and variations of vascular hindrance are responses in an attempt to compensate for alterations of blood rheology and perfusion pressure in order to maintain adequate blood flow [10]. This aspect is further discussed in Section 3 of this chapter. The influences of alterations of hematocrit, plasma viscosity and red blood cell (RBC) properties on pressure-flow relationship in vivo have been experimentally studied, with salient findings presented below. 1.1. Effect of Hematocrit Hematocrit (i.e., the volume fraction of RBC in blood) is the major determinant of blood viscosity when measured using a viscometer of any type. However, as first demonstrated by Whittaker and Winton (Figure 1) [1], the impact of hematocrit on blood fluidity in vivo is significantly blunted. Their study was done in isolated dog

324


hind limb, a frequently used preparation for the study of in vivo hemodynamics. Further studies extended the observations to other organs [4, 5]. Fan et al. studied pressure-flow relationships in intact organs of dogs in experiments in which hematocrit was altered from 10 to 78 % [4]; arterial pressure remained relatively constant within this range of hematocrit. Blood flow was estimated by the infusion of labeled microspheres. It was demonstrated that the patterns of altered blood flow in response to hematocrit changes were significantly different among the six organs investigated (i.e., heart, brain, intestine, liver, kidney and spleen). Specifically, myocardium and brain tissues responded to hematocrit alterations by extensive vasomotion in both directions (i.e., increased or decreased vascular hindrance in response to decreased or increased hematocrit respectively) and maintained a relatively constant oxygen supply, while kidney and spleen responded by vasoconstriction to both increments and decrements of hematocrit [4]. The difference in vasomotor reactivity in response to hematocrit changes has also been demonstrated to affect the optimum hematocrit range for maximum oxygen transport in the coronary circulation: the coronary circulation is characterized by a wider range of hematocrit (20 to 60 %) compared to the systemic circulation (40 to 60 %) [5]. Furthermore, it has been demonstrated that the coronary circulation has a significant autoregulatory reserve that can effectively compensate for hematocrit increments [11]. In dogs with normal coronary arteries, coronary blood flow resistance increased by only about 8% after hematocrit increases of 57%. However, after inducing a critical stenosis that exhausted the autoregulatory reserve in the coronary artery, the increase of flow resistance was 22%. This significant difference between the alterations in flow resistance was proposed to be an indicator of the effectiveness of vascular control mechanisms in compensating for hemorheological alterations (i.e., hematocrit increment) in hearts with normal vascular geometry [11]. However, it should be noted that the 57% increase of hematocrit only elevated flow resistance by 22% even in critically stenosed coronaries, whereas high-shear whole blood viscosity increased 106% for the same hematocrit change [11]. Therefore, although vascular control mechanisms need to be considered in any discussion of mechanisms underlying blunted effects of hematocrit alterations in vivo, other factors must also be considered. 1.2. Effect of Plasma Viscosity Gustafsson et al. investigated the effects of altered plasma viscosity on the in vivo flow behavior of blood in a series of experimental studies [12, 13] using an isolated dog hind limb, and calculated apparent blood viscosity by comparing pressure-flow relationships during perfusion with blood samples and reference solutions [12]. They reported that the in vivo apparent viscosity of blood was significantly affected by the suspending phase viscosity which was modified by adding high or low molecular weight dextrans [12]. This effect of suspending phase viscosity was more pronounced with suspensions having higher hematocrit values. Since these observations can be interpreted in terms of the ex vivo flow behavior of blood, the significant effect of suspending phase (i.e., plasma) viscosity is not unexpected. Gustafsson thus suggested that the in vivo effect of altered plasma viscosity can be predicted based on ex vivo measurements [13]. Chen et al. demonstrated in a dog model that there was a significant variation in regional blood flow changes after a four-fold increment in plasma viscosity achieved by exchange transfusions of dextran solutions with plasma [14]. They observed that blood flow was decreased in the small intestine, spleen and thyroid gland while it


325

remained unchanged in most other tissues and organs, and thus concluded that this difference in the response of various organs was related to the vasomotor control [14]. Recent studies have drawn attention to another aspect of the influence of plasma viscosity on in vivo blood flow resistance [15-17]. These studies were based on the fact that plasma is the main interface between vascular endothelium and blood: the mechanical influence of blood flow is thus mediated through plasma and therefore the rheological properties of plasma are among the determinants of this interaction. It has been demonstrated that plasma viscosity plays a significant role in regulating vascular tone and maintaining functional capillary density [15, 17], with increased plasma viscosity shown to increase endothelial nitric oxide (NO) synthesis and decrease vascular resistance [16]. Therefore, the expected effect of enhanced plasma viscosity on blood flow resistance is mitigated by the decreased vascular tone, thereby providing another explanation for the discrepancy between in vivo and ex vivo measurements of blood apparent viscosity. This issue is further detailed in Section 3 of this chapter. 1.3. Effect of RBC Deformability RBC deformability is an important determinant of ex vivo blood viscosity as determined in viscometers having geometries much larger than the RBC diameter. Additionally, it should be obvious that cell deformability is the primary determinant of flow resistance in blood vessels with dimensions similar to the RBC size. Entrance of RBC into true capillaries is characterized by a dramatic increase in flow resistance [18], with the degree of alteration of flow resistance determined by RBC deformability. Experimental data clearly indicate that RBC transits times through the microcirculation are prolonged during perfusion with suspensions of RBC rigidified by glutaraldehyde treatment [19]. Impaired RBC deformability has also been demonstrated to affect capillary recruitment in response to hypoxia [20]. It is generally accepted that the effects of RBC deformability alterations are primarily at the microcirculatory level [19, 21], although the presence of rigid RBC may cause additional hemodynamic abnormalities in larger vessels of the circulatory system [18]. Simchon et al. studied regional blood flow resistance in rats before and after exchange transfusions of minimally-hardened RBC [22]. Their results showed that alterations of blood flow resistance due to impaired RBC deformability ranged between more than a four-fold increment in the spleen to almost no change in myocardium and kidney [22]; the degree of flow resistance changes were significantly correlated with the trapping of hardened RBC in these organs. Such differences could again be related to the differences in vascular geometry and vascular autoregulatory reserve in the various organs and tissues. Further studies by the same group extended these observations using RBC with reduced surface charge consequent to neuraminidase treatment [23]. Treated RBC were mainly sequestered in the spleen and liver, and blood flow measurements again revealed that the increment in flow resistance was closely related to the degree of RBC trapping in a given organ [23]. The important role of vascular control mechanisms and autoregulatory reserve in determining the magnitude of the response to a given alteration of RBC deformability can be demonstrated even more directly. RBC minimally hardened by treatment with increasing levels (i.e., 0.0005% - 0.002%) of glutaraldehyde were suspended in autologous plasma and used for perfusing the isolated hind limb of rats [10]. The impairment of RBC deformability at the highest glutaraldehyde concentration caused a 16% decrement in deformability as measured by ektacytometer elongation indexes,

326


Change in flow resistance (%)

with this degree of impairment of deformability corresponding to a 78% increment in flow resistance (Figure 2). However, after eliminating vascular smooth muscle tone via perfusion with 10-4 M papaverin, the increase of flow resistance in response to the same degree of deformability impairment was 250% (Figure 2). Such experiments thus clearly demonstrate the role of vascular autoregulatory reserve in determining the response to a given change of RBC deformability [10]. 300

Before papaverin

*

After papaverin 225

* 150

* 75

0 0

5

10

15

20

Change in RBC deformability (%) Figure 2. Flow resistance alterations in response to RBC deformability impairment in isolatedperfused rat hind limbs. RBC deformability was progressively reduced by treating cells with increasing levels of glutaraldehyde ( 0.0005% - 0.002%). Change in flow resistance is expressed as the percentage of the resistance measured during perfusion with RBC suspensions with normal deformability. Redrawn from Baskurt et al. [10].

1.4. Effect of RBC Aggregation Even a brief examination of the current literature indicates that the influence of RBC aggregation on in vivo blood flow has attracted more attention than the effects of RBC deformability. Interestingly, while there is general agreement regarding the adverse effects of impaired deformability on in vivo blood flow, the role of RBC aggregation in in vivo blood flow dynamics is still controversial [24]. RBC aggregation is the major determinant of low-shear blood viscosity and thus it would be expected to affect blood flow, at least in the low-shear zones of circulatory system. However, early studies by Fåhraeus of blood flow in glass tubes began to challenge this expectation [25]. He observed that: 1) the accumulation of RBC in the central zone of tube flow was more pronounced with higher degrees of aggregation; 2) this accumulation of RBC in the central zone resulted in a decrement in flow resistance due to lower energy dissipation at the tube wall [25]. These observations were later confirmed by several other investigators [26-31], with their experimental results providing further details regarding the influence of RBC axial migration on flow resistance in cylindrical tubes. Important features of the aggregation effects include: 1) Decreased flow resistance due to enhanced RBC aggregation occurred in vertical tubes,


327

but not in horizontal tubes [27, 29]. Rather, enhanced RBC aggregation in horizontal tubes results in increased flow resistance due to sedimentation of RBC toward the dependant side of the tube [26]; 2) The effects of RBC aggregation on flow resistance can only be detected if the flow rate is below a critical level, with this critical flow rate inversely related to tube diameter [29]; 3) The effects of RBC aggregation are time dependent and thus reflect the kinetics of axial migration [26, 28]. Although these in vitro observations provide important information regarding the effects of RBC aggregation in cylindrical tubes, they are not directly applicable to the in vivo flow environment. The geometry of the vascular system is extremely complex [32], and branching of blood vessels limits the length of the vascular zone available for the development of effective axial migration and related phase separation [27]. Further, the in vivo orientation of blood vessels with respect to gravity is also complex, and thus the actual result of altered RBC aggregation may strongly depend on the primary orientation of blood vessels in a given organ or tissue. Therefore, only in vivo experiments provide the correct source of data for interpreting the influence of RBC aggregation on flow resistance in vascular beds. In vivo experimental research designed to assess the effect of RBC aggregation on blood flow resistance employ either intravital microscopy techniques or measurements of blood flow in an isolated whole organ or a given blood vessel. In a study using intravital microscopy in rat mesentery and cremaster muscle, flow resistance was measured after the infusion of 500 kDa dextran solutions [33]. Since high molecular weight dextran infusions enhance both RBC aggregation and plasma viscosity, RBC aggregation in some studies was inhibited by an anti-aggregating pharmacological agent (troxeuritine) to differentiate between the effects of enhanced aggregation and increased plasma viscosity. The investigators concluded that enhanced RBC aggregation increased microvascular resistance, with this effect being independent from the increased plasma viscosity [33]. Dextran 70 (MW: 70 kDa) was used in another study to increase in vivo RBC aggregation in rabbit mesentery; microvascular flow resistance was again reported to be increased with increasing dextran 70 concentrations, which in turn correlate with the level of RBC aggregation [34]. Other studies using intravital microscopy mostly agree that higher levels of RBC aggregation result in increased flow resistance [35, 36]. In contrast to findings via intravital microscopy, in vivo flow resistance in wholeorgan perfusion studies was found to be increased [37], decreased [38] or un-altered [39] in response to enhanced RBC aggregation. Charansonney et al. studied blood flow resistance in isolated-perfused rat hearts [38] using two different levels of RBC aggregation induced by either 1% or 2% dextran 70. They report that perfusion with RBC suspensions with a moderate increase of aggregation (i.e., 1% dextran 70) was characterized by a decrease of flow resistance compared to perfusion with a nonaggregating RBC suspension. Conversely, a higher degree of aggregation by using 2% dextran 70 resulted in enhanced flow resistance, thus suggesting an effect determined by the degree of RBC aggregation. An obvious difference between intravital studies and whole-organ perfusion studies is the orientation of the blood vessels under investigation [40]. Tissues for intravital studies of the microcirculation are usually spread flat to improve microscopic observation, and thus the blood vessels are oriented horizontally with respect to gravity. In this case the expected effect of enhanced RBC aggregation would be increased flow resistance given the observations in horizontal glass tubes [27, 29]. However, in whole organ studies, blood vessels are oriented in a large variety of

328


directions and the net result is determined by the dominant orientation: obviously, observations using perfused whole organs correspond more closely to physiological conditions. It is clear from the above that most in vivo experimental studies employ a common approach for modifying RBC aggregation: introduction of high-molecular weight polymers (e.g., dextrans) or proteins (e.g., fibrinogen) into the circulation. These proteins or polymers induce RBC aggregation in a dose dependent manner, but also increase the suspending phase viscosity [41]. Dextrans are preferred in such studies since they are available in a wide variety of molecular weights that have different abilities to alter RBC aggregation. Low molecular weight dextrans (i.e., less than 40 kD) are not effective in inducing aggregation; above this threshold effectiveness in inducing aggregation increases with molecular weight. Dextrans have been used in studies designed to compare the effects of increased plasma viscosity and increased RBC aggregation [12, 14]. It has been reported that the in vivo apparent viscosity of blood or RBC suspensions is mainly determined by the level of plasma viscosity, with RBC aggregation having no significant effects [12]. These reports thus strongly suggest that the impact of RBC aggregation can easily be masked due to other phenomena related to the introduction of high molecular weight dextrans, especially if “opposing” effects are expected. Additionally, as discussed in Section 1.2 and detailed in Section 3 of this chapter, altered plasma viscosity may interfere with vascular control mechanisms [16], further complicating the overall picture. Recently, a new method has been developed to modify RBC aggregation without altering the suspending phase composition or viscosity. This method is based on the attachment of specific polymers (i.e., poloxamers), which have self-association properties and can form micelles, onto the RBC membrane [42]. RBC aggregation can be enhanced or inhibited by selecting the appropriate molecular size of the polymer, and graded alterations can be achieved by modifying the concentration of the polymer during the covalent attachment process [42]. RBC coated with specific polymers can be suspended in autologous plasma and exchange transfused into intact animals or used in perfusing isolated organs [43-46]. The poloxamer coating technique to modify RBC aggregation has been used to investigate the effects of various degrees of aggregation on in vivo blood flow resistance in the guinea pig hind limb (Figure 3) [44]. RBC aggregation in plasma was increased in five steps between 63–200% of control by using poloxamer F98 concentrations of 0.0125 to 0.5 mg/ml during the coating process. Experimental results demonstrated that a moderate increase of RBC aggregation significantly increased flow resistance compared to the resistance for normal rat blood (Figure 3). Flow resistance returned to control with a further increment of RBC aggregation, then finally a second phase of significant elevation of blood flow resistance followed with the highest degree of aggregation [44]. These results thus tend to support the findings of Charansonney mentioned above [38], suggesting that the nature of the impact of RBC aggregation on in vivo blood flow is determined by the degree of aggregation. Note that the results shown in Figure 3 are not complicated by enhanced suspending phase viscosity in addition to increased aggregation, and thus are not affected by changes of suspending phase viscosity as occurred in previous studies. Additionally, it should be mentioned that the significant effects of aggregation could only be observed if the isolatedperfused hind limb was pre-treated with papaverin to inhibit smooth muscle tone; no

329


statistically significant changes were observed in preparations with intact vascular tone [44].

Change in PRU (%)

200

*

150

*

100

50 0

50

100

150

200

250

Increase of RBC aggregation (%) Figure 3. Effect of enhanced RBC aggregation on blood flow in an isolated-perfused guinea pig hind-limb. The change in blood flow resistance (PRU) is shown as percentage of resistance during perfusion with unmodified guinea pig blood (i.e., control level= 100), with the change in aggregation shown as percentage of aggregation for unmodified guinea pig blood. Aggregation was modified by covalent attachment of Pluronic F98 to the RBC surface. The preparation was pre-treated by 10-4 M papaverin to inhibit vascular smooth muscle tonus. * indicates significant difference from flow resistance during perfusion with unmodified blood (p<0.05). Redrawn from Baskurt and Meiselman [40].

The influence of RBC aggregation on in vivo blood flow resistance is also determined by hemodynamic factors: the effect of increased RBC aggregation on blood flow resistance in an isolated-perfused guinea pig heart strongly depends on the perfusion pressure [45]. Significant differences between flow resistance for RBC suspensions with increased and normal aggregation were only observed if the perfusion pressure was below 50 mmHg, suggesting an explanation based upon the shear rate dependence of red cell aggregation. The dependence of RBC aggregation effects on perfusion pressure also provides an explanation for the greater importance of aggregation in determining blood flow on the venous side of the circulatory system versus the arterial side. The formation of RBC aggregates in the venous circulation has been clearly demonstrated [47, 48], and the role of RBC aggregation in venous blood flow is discussed in detail by Bishop et al. [49]. Cabel et al. demonstrated an inverse relationship between venous conductance and blood flow in cat lateral gastrocnemius muscle preparations perfused with “normally” aggregating cat blood [50]. This “normal” relationship was altered for either non-aggregating suspensions (i.e., RBC in 40 kDa dextran solutions) or for RBC suspensions with higher aggregation induced by infusions of 250 kDa dextran. The observed effects of increased aggregation were pressure-dependent, with a drop of perfusion pressure from 100 mmHg to 40 mmHg resulting in up to a 300% increase of

330


venous vascular resistance [51]. It has been proposed that most of the increment in flow resistance in venous vessels can be explained by increased RBC aggregation under diminished shear forces [50, 51], indicating that RBC aggregation plays an important role in venous vascular function by significantly contributing to venous vascular resistance [50]. Further, Bishop et al. demonstrated in rat spinotrapezius muscle that enhanced RBC aggregation results in significant blunting of velocity profiles in venous vessels under low-flow conditions [51]: such alterations of velocity profiles results in increased energy loss [51] and may contribute to increased venous flow resistance. Functional capillary density (i.e., number of capillaries with RBC flow in a selected microcirculatory field) is another parameter that has been used to investigate the role of perfusion pressure in determining the impact of RBC aggregation [52]. Kim et al. demonstrated that increased RBC aggregation due to 500 kDa dextran infusions had only a slight effect on functional capillary density in rat spinotrapezius muscle at normal and 50 mmHg arterial pressures, but density was significantly reduced if arterial pressure was reduced to 25 mmHg [52]. In overview, it is safe to say that the effects of RBC aggregation on in vivo hemodynamics are much more complex when compared to other hemorheological parameters such as hematocrit, plasma viscosity or RBC deformability: the reason for this complexity being the multiple mechanisms, with often opposing effects on flow resistance, that are influenced by altered aggregation.

2. Mechanisms Underlying “Different” In Vivo Blood Fluidity It should be obvious from Section 1 of this chapter that the discrepancy between in vivo and ex vivo pressure-flow relationships can be explained, at least in part, by the compensatory action of vascular control mechanisms which are effective in vivo [53]. However, the difference between in vivo and ex vivo flow properties still exists even after totally eliminating vascular autoregulatory reserve [11] or vascular tone [44], thereby indicating the involvement of other in vivo mechanisms [54]. The mammalian vascular system is characterized by a very complex geometry, with blood vessel sizes that vary over a very wide range from a few microns to a few centimeters in diameter. Decreased vessel diameter is accompanied by an increased number of individual vessels due to extensive branching, and since total cross-sectional area increases, blood flow velocity progressively decreases as it approaches the capillaries [32]. The major resistance to blood flow in the systemic circulation occurs in the microcirculation, and thus the large pressure gradient that exists across this region augments shear forces in this zone [32]. Further, there are specific and effective mechanisms in blood vessels with diameters below 1000 microns that may affect the in vivo fluidity of blood at this circulatory level. Inasmuch as these mechanisms are discussed in detail in Chapters III.1 and III.2, they are only briefly mentioned below. 2.1. Axial Migration of Red Blood Cells The formation of a flow-dependent cell-poor or cell-free layer near blood vessel walls in living organisms was observed in the 18th century and analyzed in detail in the 20th century [25, 55, 56]. Formation of this marginal flow zone with reduced RBC concentration was attributed to the accumulation of RBC in the axial region; such redistribution of suspended particles is a well-understood phenomenon in fluid


331

mechanics and is not limited to blood flow [53]. While a detailed description of the physical mechanism underlying RBC axial migration is beyond the scope of this chapter, it is possible to indicate that RBC axial migration is affected by: 1) RBC rheological properties (i.e., deformability and aggregation); 2) hemodynamic conditions since it has been demonstrated that the thickness of the marginal cell-free zone depends on the shear forces [32, 56]. The physiological consequences of RBC axial migration during flow in blood vessels at the microcirculatory level are the Fåhraeus and Fåhraeus-Lindqvist Effects and plasma skimming. 2.2. Fåhraeus and Fåhraeus-Lindqvist Effects Concepts based on the early work of Fåhraeus and co-workers [57], and further detailed by others [58, 59], have provided an elegant explanation for the differences between in vivo and in vitro flow behavior of blood. The reader is referred to the excellent review by Goldsmith, Cokelet and Gaehtgens for a detailed analysis of these concepts [53]. Briefly, Robin Fåhraeus observed that RBC accumulate in the axial region during flow in blood vessels with diameters below 300 microns, and therefore RBC mean velocity is higher than that of whole blood. This difference in mean velocities of the components of blood (i.e., RBC and plasma) results in a reduction of hematocrit within a tube (tube hematocrit, HT) compared to the hematocrit of blood presented to the tube (feed hematocrit, HF). This reduction of tube hematocrit is known as Fåhraeus Effect. Obviously, the velocity difference only exists during blood flow, with HT inversely related to flow rate [53]. Another factor affecting the HT/HF ratio is tube diameter, with the ratio becoming smaller with decreasing tube diameter [53]; the reduction in HT/HF continues until the tube radius approaches ~15 microns [54]. It should be noted that the hematocrit of the blood discharged from the tube should equal HF; if not, there is RBC screening at the entrance to the tube. An obvious outcome of hematocrit reduction as blood flows in smaller vessels is the reduction of apparent viscosity: the apparent viscosity of blood decreases with vessel diameter [57] and approaches a minimum at ~7 microns [54]. The reduction in viscosity with reducing tube diameter is known as the Fåhraeus-Lindqvist Effect. Cokelet has compared data from a number of separate studies investigating the Fåhraeus and Fåhraeus-Lindqvist effects and has concluded that the viscosity reduction can be explained by Fahråeus Effect for the tube sizes above 20 micron [57]. In vessels with diameters below 20 microns, blood cannot be regarded as a continuous fluid and RBC rheological properties (i.e., deformability and aggregation) and plasma viscosity dominate separately [24, 54]. An alternative or additional explanation for Fåhraeus-Lindqvist Effect relates to the alignment of RBC in small-diameter (~10 micron) tubes with a lubricating layer of plasma between the cells and the tube wall, thereby resulting in minimal energy dissipation during flow [60]. Conversely, in larger diameter tubes, a more irregular movement of RBC dominates, with more friction between RBC and between RBC and the tube or vessel wall, increased energy dissipation and increased apparent viscosity. It should be noted that the Fåhraeus-Lindqvist Effect results in marked reduction of blood viscosity, with apparent viscosity only 10-15% greater than plasma viscosity in 8-10 micron capillaries [24]; plasma viscosity thus plays a very significant role in determining flow resistance in the microcirculation [24].

332


2.3. Plasma Skimming Axial migration of RBC has another consequence that also results in decreased hematocrit as blood approaches the micro vessels. The phase separation due to axial migration affects the cellular content of blood flowing into the side branches of blood vessels. A branch originating from a vessel of higher order is fed mainly by the marginal stream of the higher order vessel and receives blood with a lower hematocrit. Further, even at bifurcations significant heterogeneity of hematocrit can occur, especially if there is a difference between the sizes of the daughter branches [60]: larger branches with higher flow rates receive relatively more RBC and therefore have a higher hematocrit blood compared to smaller branches. 2.4. Microvascular Hematocrit It has been demonstrated in a variety of tissues that hematocrit in the microcirculation is significantly lower than large vessel hematocrit [61], with the hematocrit in microvessels as low as 20% of that in large vessels [61]. Both the Fåhraeus Effect and plasma skimming contribute to the difference between hematocrit values at various levels of the circulation [59]. Intravital microscopy and image analysis techniques are frequently used to estimate microvascular hematocrit [62, 63]. Alternatively, the mean hematocrit value of blood circulating in vessels of all sizes in a given tissue can be calculated using tracers that separately label RBC and plasma [64, 65]. The values by either approach are called tissue hematocrit [65, 66], and have been found to be between microvascular and large vessel hematocrits, usually between 75-90% of the latter [61]. It is obvious from this argument that the hematocrit measured in venous or arterial blood samples does not represent the blood composition at all levels of circulation. Thus, locationspecific hematocrit provides yet another explanation for the discrepancy between in vivo apparent viscosity of blood and ex vivo determinations performed using samples with large vessel hematocrit values. Microvascular and tissue hematocrits have been demonstrated to be influenced by both hemodynamic conditions and the properties of blood elements. While experimental data generally indicates that there is a positive correlation between microvascular or tissue hematocrit and blood flow to that tissue [67-71], there are conflicting reports [72]. Vicaut and Levy demonstrated that tissue hematocrit was different at various depths of left ventricular myocardium [65]; this alteration of tissue hematocrit was attributed to different hemodynamic conditions (e.g., perfusion pressures) at various tissue depths. The effect of RBC deformability on microvascular hematocrit has been investigated in rat cremaster muscle and shown to be increased with impaired deformability [19]. It has also been demonstrated that the myocardial hematocrit gradient first demonstrated to exist by Vicaut and Levy [65] is influenced by alterations of both RBC deformability [73] and aggregation [46, 74]. 2.5. Role of RBC Aggregation in Axial Migration As discussed above, axial migration of RBC and related mechanisms play an important role in determining the in vivo apparent viscosity of blood, and, for glass tubes, the degree of RBC aggregation has been identified as a determinant of axial migration and


333

apparent viscosity [25, 27-29]. However, additional factors need to be considered when evaluating the influence of RBC aggregation on in vivo axial migration and related mechanisms. For example, it has been shown that effective axial migration requires a critical time to develop in a given vessel segment, and hence there is a minimal time needed to develop the migration-dependent phenomena discussed above [49]. Therefore, the transit time of blood in a given vessel segment should be longer than this critical time in order to have an impact of RBC axial migration on vascular hemodynamics. In turn, transit times depend on the unbranched length of the vessel segment and flow rate; frequent branching, especially in venules, usually prevents the formation of a well-developed cell-poor layer near the vessel wall. This requirement for an appropriate time to develop the layer is especially important in venous vessels, since with each confluent branch a new stream of RBC is introduced into the peripheral zone of the venous vessel, abolishing the phase separation [50]. It has been demonstrated that shear rates should be lower than 5 s-1 for the formation of an effective cell-poor marginal layer in cat skeletal muscle, and that this low level of shear can not be achieved with physiological perfusion pressures [49]. The kinetics of RBC aggregation can have a strong impact on the balance between the critical time and transit time by influencing the rate of axial migration and hence the time required for axial migration to develop. Aggregation time constants are normally measured ex vivo, using shearing systems of various geometries, and reported to be on the order of seconds with variations dependent on the specific geometry of the system being used [75-77]. However, it has been shown that aggregation time constants in post-capillary venules can be at least an order of magnitude shorter than those measured ex vivo [48]. This difference is most probably due to the smaller scale of the system in which aggregation takes place (e.g., venules of 10-15 microns in diameter). Additionally, it has been observed that the formation of RBC aggregates occurs within 15-30 microns after RBC entrance into a venule [48]. These dynamic considerations regarding the effects of RBC aggregation on axial migration and related mechanisms may aid in explaining the experimental findings summarized in Section 1.4 of this chapter. The impact of RBC aggregation on in vivo hemodynamics has multiple components that can be classified into two groups which oppose each other in their effects on flow resistance. The first group tends to decrease flow resistance in response to enhanced aggregation: 1) Enhanced aggregation increases axial migration of RBC and promotes the formation of a marginal cell-free or cell-poor layer, thereby decreasing energy dissipation due to lower frictional resistance with the vessel wall; 2) Increased axial migration also promotes the Fåhraeus effect and plasma skimming, thereby decreasing microvascular hematocrit. The second group includes factors that tend to increase flow resistance in response to enhanced aggregation: 1) RBC aggregation increases the low shear viscosity of blood when measured in a viscometer and may affect resistance in large blood vessels when shear forces are sufficiently low; 2) RBC aggregates must be dispersed in order for cells to enter microcirculatory vessels [47], with the dis-aggregation energy reflected by increased flow resistance; 3) Enhanced RBC aggregation may influence vasomotor tone as briefly described in Section 3, thereby increasing vascular tone and flow resistance. The net overall effect of a given change of RBC aggregation will therefore be determined by the interplay between the components of each group.

334


2.6. Effects of Inertial Pressure Loses on In Vivo Flow Dynamics Almost all data discussed above for in vivo versus ex vivo flow behavior differences support explanations based on mechanisms that involve RBC. However, Benis et al. repeated the experiments of Whittaker and Winton in an isolated dog hind limb using either RBC suspensions or cell-free Newtonian fluids [2]. They confirmed the findings of Whittaker and Winton regarding the discrepancy between in vivo and ex vivo apparent viscosities, yet found that the difference existed in experiments either with RBC suspensions or with cell-free fluids. They concluded that mechanisms based on RBC behavior (e.g., Fåhraeus-Lindqvist Effect) could not explain the difference observed with cell-free fluids, and indicated that inertial pressure loses in larger blood vessels must be considered when evaluating in vivo versus ex vivo differences [2]. Interestingly, a later study could not confirm the results of Benis et al [2], and rather ascribed any apparent inertial effects to changes of vascular geometry [78].

3. Hemorheology and Vascular Control Mechanisms It is generally accepted that vascular control mechanisms can play an important role in the hemodynamic consequences of hemorheological alterations [10]. That is, a significant hemorheological alteration may easily be compensated for by vasomotion, thereby attenuating or totally eliminating the expected change in hemodynamic resistance [10]. It is thus possible that this normally effective compensation mechanism is one reason for some misunderstanding of the importance of blood rheology alterations in clinical medicine [53]. However, it should be kept in mind that the effectiveness of compensation mechanisms can be significantly reduced if vascular geometry is altered due to a disease process (e.g., atherosclerosis). In such a situation, vascular autoregulatory reserve might be exhausted and a hemorheological alteration that could be easily compensated for in healthy individuals might have serious consequences [11]. Experimental data also indicate that hemorheological alterations may have a direct influence on vascular regulatory mechanisms. Nitric oxide (NO) generation in endothelial cells and diffusion to vascular smooth muscle is an important factor in vasomotor control [79]. NO output of endothelial cells is controlled by shear stress acting on the endothelial cells [80, 81] as well as other physiological mechanisms [82]. Hence, wall shear stress plays a role in the control of NO synthesizing mechanisms, and is effective for both acute activation of NO production and for the expression of endothelial nitric oxide synthase (eNOS) [82, 83]: elevated wall shear stress in a vascular bed leads to increased NO release and vascular relaxation, thereby decreasing vascular hindrance. Since wall shear stress is determined by fluid viscosity at the vessel wall and by local fluid velocity [84, 85], any hemorheological or hemodynamic alteration that results in changes of local viscosity near the vessel wall may affect NO generation by endothelial cells and thus modulate vascular hindrance. As a result of RBC axial migration, plasma viscosity is an important parameter determining wall shear stress as it is the interface between blood flow and endothelial cells [15]. Tsai et al. demonstrated that elevated plasma viscosity increases perivascular NO concentrations in a hamster dorsal skin model, and also observed that eNOS expression was enhanced in aortic endothelial cells in animals with elevated plasma viscosity [16]. Up-regulation of NO synthesizing mechanisms in endothelium and the


335

accompanying increase of NO release most likely has beneficial effects on microvascular perfusion [15-17]. RBC aggregation is another factor that may affect wall shear stress in blood vessels since enhanced aggregation promotes axial migration of RBC [26, 29, 49, 86]. Enhanced RBC aggregation may therefore lead to down-regulation of NO synthesis that is mediated by decreased wall shear stress due to decreased local fluid viscosity at the vessel wall. It has been demonstrated in a rat model that chronically enhanced RBC aggregation results in blunted flow-mediated dilation in arterioles, downregulation of eNOS expression in skeletal muscles [43], and the progressive development of hypertension during a four day period following infusion of strongly aggregating RBC suspensions. These findings thus support the abovementioned hypothesis that chronically enhanced RBC aggregation results in diminished wall-shear stress, most probably due to enhanced axial accumulation of RBC.

4. Effect of Leukocytes on In Vivo Blood Flow The contribution of leukocytes to the cellular population of blood is very small vis-àvis RBC, and therefore their impact on flow resistance in most parts of the circulatory system is negligible. However, leukocytes generate significant resistance at the microvascular level, with their impact on flow resistance determined by cellular mechanical properties that can undergo significant alterations during activation. This aspect of in vivo hemorheology is covered in detail in Chapter II.4.c. Radial distribution of leukocytes is also affected by the phase separation discussed in Section 2. Axial accumulation of RBC displaces other elements of blood towards the peripheral flow region and promotes margination of leukocytes and platelets [87]. This phenomenon has been termed the “inverse Fåhraeus Effect” for leukocytes and platelets, as these cellular elements are accumulated in the marginal zone and their concentration within the tube is higher compared to the feed or discharge value [88]. Enhanced RBC aggregation significantly influences platelet function [87] and leukocyte margination [89], with the enhanced leukocyte margination and rolling on the vessel wall increasing flow resistance, especially at microcirculatory level [90].

References [1] [2] [3] [4] [5] [6] [7]

S.R.F. Whittaker and F.R. Winton, The apparent viscosity of blood flowing in the isolated hindlimb of the dog and its variation with corpuscular concentration, J. Physiol. Lond. 78 (1933), 339-368. A.M. Benis, S. Chien, S. Usami and K.M. Jan, Inertial pressure losses in perfused hindlimb: a reinterpretation of the results of Whittaker and Winton, J. Appl. Physiol. 34 (1973), 383-389. W.G. Frasher, H. Wayland and H.J. Meiselman, Viscometry of circulating blood in dogs. I. Heparin injection; II. Platelet removal, J. Appl. Physiol. 25 (1968), 751-760. F. Fan, R.Y.Z. Chen, G.B. Schuessler and S. Chien, Effects of hematocrit variations on regional hemodynamics and oxygen transport in the dog, Am. J. Physiol. 238 (1980), H545-H552. K.M. Jan and S. Chien, Effect of hematocrit variations on coronary hemodynamics and oxygen utilization, Am. J. Physiol. 233 (1977), H106-H113. S. Shlomoh, K.M. Jan and S. Chien, Influence of reduced red cell deformability on regional blood flow, Am. J. Physiol. 253 (2005), H898-H903. O. Yalcin, H.J. Meiselman, J.K. Armstrong and O.K. Baskurt, Effect of enhanced red blood cell aggregation on blood flow resistance in an isolated-perfused guinea pig heart preparation, Biorheology 42 (2005), 511-520.

336 [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35]

O.K. Baskurt and H.J. Meiselman / In Vivo Hemorheology O. Yalcin, M. Uyuklu, J.K. Armstrong, H.J. Meiselman and O.K. Baskurt, Graded alterations of red blood cell aggregation influence in vivo blood flow resistance, Am. J. Physiol. 287 (2004), H2644H2650. P.M. Scholz, J.H. Karis, F.E. Gump, J.M. Kinney and S. Chien, Correlation of blood rheology with vascular resistance in critically ill patients, J. Appl. Physiol. 39 (1975), 1008-1011. O.K. Baskurt, O. Yalcin and H.J. Meiselman, Hemorheology and vascular control mechanisms, Clin. Hemorheol. Microcirc. 30 (2004), 169-178. O.K. Baskurt, E. Levi, S. Caglayan, N. Dikmenoglu, O. Ucer, R. Guner and S. Yorukan, The role of hemorheological factors in the coronary circulation, Clin. Hemorheol. 11 (1991), 121-127. L. Gustafsson, L. Appelgren and H.E. Myrvold, Effects of increased plasma viscosity and red blood cell aggregation on blood viscosity in vivo, Am. J. Physiol. 241 (1981), H513-H518. L. Gustafsson, Effects of plasma hyperviscosity on skeletal muscle blood flow and blood viscosity in vivo, Int. J. Microcirc. Clin. Exp. 6 (1987), 169-177. R.Y. Chen, R.D. Carlin, S. Simchon, K.M. Jan and S. Chien, Effects of dextran-induced hyperviscosity on regional blood flow and hemodynamics in dogs, Am. J. Physiol. 256 (1989), H898-H905. P. Cabrales and A.G. Tsai, Plasma viscosity regulates systemic and microvascular perfusion during acute extreme anemic conditions, Am. J. Physiol. 291 (2006), H2445-H2452. A.G. Tsai, C. Acero, P.R. Nance, P. Cabrales, J.A. Frangos, D.G. Buerk and M. Intaglietta, Elevated plasma viscosity in extreme hemodilution increases perivascular nitric oxide concentration and microvascular perfusion, Am. J. Physiol. 288 (2005), H1730-H1739. A.G. Tsai, B. Friesenecker, M. McCarthy, H. Sakai and M. Intaglietta, Plasma viscosity regulates capillary perfusion during extreme hemodilution in hamster skinfold model, Am. J. Physiol. 44 (1998), H2170-H2180. H.H. Lipowsky, Rheology of blood flow in the microcirculation, In: Microvascular Research, D. Shepro, Ed., Elsevier, Amsterdam, 2006, pp. 233-238. H.H. Lipowsky, L.E. Cram, W. Justice and M.J. Eppihimer, Effect of erythrocyte deformability on in vivo red cell transit time and hematocrit and their correlation with in vitro filterability, Microvasc. Res. 46 (1993), 43-64. K. Parthasarathi and H.H. Lipowsky, Capillary recruitment in response to tissue hypoxia and its dependence on red blood cell deformability, Am. J. Physiol. 277 (1999), H2145-H2157. T.W. Secomb and R. Hsu, Resistance to blood flow in nonuniform capillaries, Microcirculation 4 (1997), 421-427. S. Simchon, K.M. Jan and S. Chien, Influence of reduced red cell deformability on regional blood flow, Am. J. Physiol. 253 (1987), H898-H903. S. Simchon, K.M. Jan and S. Chien, Studies on sequestration of neuraminidase-treated red blood cells, Am. J. Physiol. 254 (1988), H1167-H1171. H.J. Meiselman and O.K. Baskurt, Hemorheology and hemodynamics: Dove andare?, Clin. Hemorheol. Microcirc. 35 (2006), 37-43. R. Fahraeus, The influence of the rouleaux formation of the erythrocytes on the rheology of the blood, Acta Med. Scand. 161 (1958), 151-165. C. Alonso, A.R. Pries and P. Gaehtgens, Time-dependent rheological behavior of blood flow at low shear in narrow horizontal tubes, Biorheology 26 (1989), 229-246. C. Alonso, A.R. Pries and P. Gaehtgens, Time-dependent rheological behavior of blood at low shear in narrow vertical tubes, Am. J. Physiol. 265 (1993), H553-H561. C. Alonso, A.R. Pries, K.D. Lerche and P. Gaehtgens, Transient rheological behavior of blood in lowshear tube flow: velocity profiles and effective viscosity., Am. J. Physiol. 268 (1995), H25-H32. G.R. Cokelet and H.L. Goldsmith, Decreased hydrodynamic resistance in the two-phase flow of blood through small vertical tubes at low flow rates, Circ. Res. 68 (1991), 1-17. T. Murata, Effects of sedimentation of small red blood cell aggregates on blood flow in narrow horizontal tubes, Biorheology 33 (1987), 267-283. W. Reinke, P. Gaehtgens and P.C. Johnson, Blood viscosity in small tubes: effect of shear rate, aggregation, and sedimentation, Am. J. Physiol. 253 (1987), H540-H547. A.S. Popel and P.C. Johnson, Microcirculation and hemorheology, Ann. Rev. Fluid Mech. 37 (2005), 43-69. J.J. Durussel, M.F. Berthault, G. Guiffant and J. Dufaux, Effects of red blood cell hyperaggregation on the rat microcirculation blood flow, Acta Physiol. Scand. 163 (1998), 25-32. M. Soutani, Y. Suzuki, N. Tateishi and N. Maeda, Quantificative evaluation of flow dynamics of erythrocytes in microvessels: influence of erythrocyte aggregation., Am. J. Physiol. 268 (1995), H1959H1965. G. Mchedlishvili, L. Gobejishvili and N. Beritashvili, Effect of intensified red blood cell aggregability on arterial pressure and mesenteric microcirculation, Microvacs. Res. 45 (1993), 233-242.


337

[36] E. Vicaut, X. Hou, L. Decuypere, A. Taccoen and M. Duvelleroy, Red blood cell aggregation and microcirculation in rat cremaster muscle, Int. J. Microcirc. Clin. Exp. 14 (1994), 14-21. [37] H. Rogausch, The apparent viscosity of aggregating and non-aggregating erythrocyte suspensions in the isolated perfused liver, Biorheology 24 (1987), 163-171. [38] O. Charansonney, S. Mouren, J. Dufaux, M. Duvelleroy and E. Vicaut, Red blood cell aggregation and blood viscosity in an isolated heart preparation, Biorheology 30 (1993), 75-84. [39] C.M. Verkeste, P.F. Boekkooi, P.R. Saxena and L.L. Peeters, Increased red cell aggregation does not reduce uteroplacental blood flow in the awake, hemoconcentrated, late-pregnant guinea pig, Pediatr. Res. 31 (1992), 91-93. [40] O.K. Baskurt and H.J. Meiselman, Hemodynamic effects of red blood cell aggregation, Ind. J. Exp. Biol. 45 (2007), 18-24. [41] C. Tanford, Physical chemistry of macromolecules. Wiley, New York, NY, 1961. [42] J.K. Armstrong, H.J. Meiselman, R. Wenby and T.C. Fisher, Modulation of red blood cell aggregation and blood viscosity by the covalent attachment of Pluronic copolymers, Biorheology 38 (2005), 239247. [43] O.K. Baskurt, O. Yalcin, S. Ozdem, J.K. Armstrong and H.J. Meiselman, Modulation of endothelial nitric oxide synthase expression by red blood cell aggregation, Am. J. Physiol. 286 (2004), H222-H229. [44] O. Yalcin, M. Uyuklu, J.K. Armstrong, H.J. Meiselman and O.K. Baskurt, Graded alterations of RBC aggregation influence in vivo blood flow resistance, Am. J. Physiol. 287 (2004), H2644-H2650. [45] O. Yalcin, H.J. Meiselman, J.K. Armstrong and O.K. Baskurt, Effect of enhanced red blood cell aggregation on blood flow resistance in an isolated-perfused guinea pig heart preparation, Biorheology 42 (2005), 511-520. [46] O. Yalcin, F. Aydin, P. Ulker, M. Uyuklu, F. Gungor, J.K. Armstrong, H.J. Meiselman and O.K. Baskurt, Effects of red blood cell aggregation on myocardial hematocrit gradient using two approaches to increase aggregation, Am. J. Physiol. 290 (2006), H765-H771. [47] J.J. Bishop, P.R. Nance, A.S. Popel, M. Intaglietta and P.C. Johnson, Relationship between erythrocyte aggregate size and flow rate in skeletal muscle venules., Am. J. Physiol. 286 (2004), H113-H120. [48] S. Kim, A.S. Popel, M. Intaglietta and P.C. Johnson, Aggregate formation of erythrocytes in postcapillary venules, Am. J. Physiol. 288 (2005), 584-590. [49] J.J. Bishop, A.S. Popel, M. Intaglietta and P.C. Johnson, Rheological effects of red blood cell aggregation in the venous network: a review of recent studies, Biorheology 38 (2001), 263-274. [50] M. Cabel, H.J. Meiselman, A.S. Popel and P.C. Johnson, Contribution of red blood cell aggregation to venous vascular resistance in skeletal muscle, Am. J. Physiol. 272 (1997), H1020-H1032. [51] J.J. Bishop, P.R. Nance, A.S. Popel, M. Intaglietta and P.C. Johnson, Effect of erythrocyte aggregation on velocity profiles in venules, Am. J. Physiol. 280 (2001), H222-H236. [52] S. Kim, A.S. Popel, M. Intaglietta and P.C. Johnson, Effect of erythrocyte aggregation at normal human levels on functional capillary density in rat spinotrapezius muscle, Am. J. Physiol. 290 (2006), H941H947. [53] H.L. Goldsmith, G.R. Cokelet and P. Gaehtgens, Robin Fahraeus: evolution of his concepts in cardiovascular physiology, Am. J. Physiol. 257 (1989), H1005-H1015. [54] G.R. Cokelet, Viscometric, in vitro and in vivo blood viscosity relationships: how are they related? Biorheology 36 (1999), 343-358. [55] L.E. Byliss, The axial drift of the red cells when blood flows in a narrow tube, J. Physiol. 149 (1959), 593-613. [56] L.E. Byliss, The flow of suspensions of red blood cells in capillary tubes. Changes in the "cell-free" marginal sheath with changes in the shearing stress, J. Physiol. 179 (1965), 1-25. [57] R. Fahraeus and T. Lindqvist, The viscosity of the blood in narrow capillary tubes, Am. J. Physiol. 96 (1931), 562-568. [58] G.R. Cokelet, Speculations on the cause of low vessel hematocrit in the microcirculation, Microcirculation 2 (1982), 1-18. [59] P. Gaehtgens, Flow of blood through narrow capillaries: rheological mechanisms determining capillary hematocrit and apparent viscosity, Biorheology 17 (1980), 183-189. [60] A.R. Pries and T.W. Secomb, Rheology of the microcirculation, Clin. Hemorheol. Microcirc. 29 (2003), 143-148. [61] O.K. Baskurt, O. Yalcin, F. Gungor and H.J. Meiselman, Hemorheological parameters as determinants of myocardial tissue hematocrit values, Clin. Hemorheol. Microcirc. 35 (2006), 45-50. [62] H.H. Lipowsky, S. Usami and S. Chien, Invivo Measurements of Apparent Viscosity and Microvessel Hematocrit in the Mesentery of the Cat, Microvasc. Res. 19 (1980), 297-319. [63] I.H. Sarelius and B.R. Duling, Direct measurement of microvessel hematocrit, red cell flux, velocity, and transit time, Am. J. Physiol. 243 (1982), H1018-H1026.

338


[64] J.C. Frisbee, Striated muscle microvascular hematocrit: the increase from rest to contraction, Microvacs. Res. 55 (1998), 184-186. [65] E. Vicaut and B.I. Levy, Transmural hematocrit gratient in left ventricular myocardia of rats, Am. J. Physiol. 259 (1990), H403-H408. [66] M.M. Todd, J.B. Weeks and D.S. Warner, Cerebral blood-flow, blood volume, and brain-tissue hematocrit during isovolemic hemodilution with hetastarch in rats, Am. J. Physiol. 263 (1992), H75H82. [67] D.M. Brizel, B. Klitzman, M. Cook, J. Edwards, G. Rosner and M.W. Dewhirst, A comparison of tumor and normal tissue microvascular hematocrits and red cell fluxes in a rat window chamber model, Int. J. Radiat. Oncol. Biol. Phys. 25 (1993), 269-276. [68] J.C. Frisbee and J.K. Barclay, Microvascular hematocrit and permeability- surface area product in contracting canine skeletal muscle in situ, Microvacs. Res. 55 (1998), 153-164. [69] V.E. Hjortdal, E.S. Hansen, T.B. Heniksen, D. Kjolseth, K. Soballe and J.C. Djurhuus, The microcirculation of myocutaneous island flaps in pigs studied with radioactive blood volume tracers and microspheres of different sizes, Plas. Reconst. Surg. 89 (1992), 116-122. [70] B. Klitzman and B.R. Duling, Microvascular hematocrit and red cell flow in resting and contracting striated muscle, Am. J. Physiol. 237 (1979), H481-H490. [71] B. Klitzman and P.C. Johnson, Capillary network geometry and red cell distribution in hamster cremaster muscle, Am. J. Physiol. 242 (1982), H211-H219. [72] Y. Goncalez, M.C.C. Machado, E.A. Andre, P.L. Aguirrecosta, S.N. Sampietri, F. Goncalez and H.W. Pinotti, Modifications of tissue blood-flow in acute pancreatitis, Brazillian J. Med. Biol. Res. 24 (1991), 741-746. [73] O.K. Baskurt, M. Edremitlioglu and A. Temiz, Effect of erythrocyte deformability on myocardial hematocrit gradient, Am. J. Physiol 268 (1995), H260-H264. [74] O.K. Baskurt and M. Edremitlioglu, Myocardial tissue hematocrit: existence of a transmural gradient and alterations after fibrinogen infusions, Clin. Hemorheol. 15 (1995), 97-105. [75] S. Chen, G. Barshtein, B. Gavish, Y. Mahler and S. Yedgar, Monitoring of red blood cell aggregability in a flow-chamber by computerized image analysis, Clin. Hemorheol. 14 (1994), 497-508. [76] S. Chen, B. Gavish, S. Zhang, Y. Mahler and S. Yedgar, Monitoring of erythrocyte aggregate morphology under flow by computerized image analysis, Biorheology 32 (1995), 487-496. [77] H. Schmid-Schonbein, P. Gaehtgens and H. Hirsch, On the shear rate dependence of red cell aggregation in vitro, J. Clin. Invest. 47 (1968), 1447-1454. [78] E. Eliassen, B. Folkow and B. Oberg, Are there significant inertial losses in the vascular bed? Acta. Physiol. Scand. 87 (1973), 567-569. [79] L.J. Ignarro, G. Cirino, A. Casini and C. Napoli, Nitric oxide as a signaling molecule in the vascular system: an overview., J. Cardiovasc. Pharmacol. 34 (1999), 879-886. [80] O. Traub and B.C. Berk, Laminar shear stress: mechanisms by which endothelial cells transduce an atheroprotective force, Arterioscler. Thromb. Vasc. Biol. 18 (1998), 677-685. [81] T. Ziegler, P. Silacci, V.J. Harrison and D. Hayoz, Nitric oxide synthase expression in endothelial cells exposed to mechanical forces, Hypertension 32 (1998), 351-355. [82] I. Fleming and R. Busse, Molecular mechanisms involved in the regulation of the endothelial nitric oxide synthase, Am. J. Physiol. 284 (2003), R1-12. [83] O. Traub and B.C. Berk, Laminar Shear Stress : Mechanisms by Which Endothelial Cells Transduce an Atheroprotective Force, Arterioscler. Thromb. Vasc. Biol. 18 (1998), 677-685. [84] R.M. Nerem, R.W. Alexander, D.C. Chappell, R.M. Medford, S.E. Vagner and W.R. Taylor, The study of the influence of flow on vascular endothelial biology, Am. J. Med. Sci. 316 (1998), 169-175. [85] R.S. Reneman, T. Arts and A.P.G. Hoeks, Wall shear stress - an important determinant of endothelial cell function and structure in the arterial system in vivo, J. Vasc. Res. 43 (2006), 251-269. [86] J.J. Bishop, A.S. Popel, M. Intaglietta and P.C. Johnson, Effects of erythrocyte aggregation and venous network geometry on red blood cell axial migration, Am. J. Physiol. 281 (2001), H939-H950. [87] H.L. Goldsmith, D.N. Bell, S. Spain and F.A. McIntosh, Effect of red blood cells and their aggregates on platelets and white cells in flowing blood, Biorheology 36 (1999), 461-468. [88] W.S.J. Uijttewaal, E.J. Nijhof, P.J.H. Bronkhorst, E. Denhartog and R.M. Heethaar, Near-Wall Excess of Platelets Induced by Lateral Migration of Erythrocytes in Flowing Blood, Am. J. Physiol. 264 (1993), H1239-H1244. [89] M.J. Pearson and H.H. Lipowsky, Influence of erythrocyte aggregation on leukocyte margination in postcapillary venules of rat mesentery, Am. J. Physiol. 279 (2000), H1460-H1471. [90] B. Das, P.C. Johnson and A.S. Popel, Computational fluid dynamic studies of leukocyte adhesion effects on non-Newtonian blood flow through microvessels, Biorheology 37 (2000), 239-258.


339

Endothelium and Hemorheology Tommaso GORI1 and Sandro FORCONI Dipartimento di Medicina Interna, Cardiovascolare e Geriatrica, Università degli Studi di Siena, Siena, Italy

Introduction Vascular endothelium is a monocellular layer positioned between the muscular media, or the adventitia in capillaries, and the circulating blood [1]. While it has long been recognized that this tissue acts as a selective sieve to facilitate bi-directional passage of macromolecules and blood gases between tissues and blood, the strategic importance of the endothelium in regulating vascular homeostasis, as well as its protective role, have been more recently described. The critical location of this tissue allows it to sense changes in hemodynamic forces and blood-borne signals. In turn, these stimuli trigger a response that is mediated by the release of a number of autocrine and paracrine substances. For example, myogenic or adrenergic tone are endothelium-independent, yet vascular homeostasis is controlled by a balanced release of endothelium-derived bioactive factors. The loss of the structural and/or functional integrity of the endothelium (i.e. endothelial dysfunction) disrupts this balance, thereby predisposing the vessel wall to vasoconstriction, leukocyte adhesion, platelet activation, mitogenesis, peroxidation, thrombosis, impaired coagulation, vascular inflammation, and ultimately, atherogenesis [2] (Figure 1). The following paragraphs describe how hemorheology interferes with the production of endothelial autacoids, how the endothelium functions, and how it influences vascular flow and hemorheology.

1. Anatomy, Physiology and Pathophysiology of the Vascular Endothelium As described above, the endothelium is a metabolically complex tissue with a very simple anatomical structure represented by a single layer of mesenchymal cells. The luminal surface of the endothelium is covered by a well-developed glycocalyx, a thick layer composed of macromolecules (i.e., proteins, glycolipids, glycoproteins and proteoglycans) bound to endothelial membranes. Its molecular domains allow cellular adhesion (i.e., selectins and integrins, that are involved in immune reactions and inflammation) [3, 4], modulate coagulation and fibrinolysis by activating and/or inhibiting thrombin, tissue factor and plasminogen [5], control fluid and metabolite transport, and, most importantly, represent an important mechanical transducer for physical stimuli that regulate endothelial biochemical function (Figure 2). The physicochemical composition of the endothelial glycocalyx favors the dynamic adhesion of a number of plasma proteins, determining a complex tri-dimensional 1

Corresponding author: Tommaso Gori, MD PhD, Department of Internal, Cardiovascular and Geriatric Medicine, University of Siena. Phone: +39 0577 585350; E mail: [email protected]

340

T. Gori and S. Forconi / Endothelium and Hemorheology

structure that, with its thickness of approximately 100 nm, significantly reduces the lumen of capillaries [6]. This structure, whose important functions are still being clarified, is termed the endothelial surface layer (ESL). As described below, the ESL has a critical role in modulating the endothelial biochemical response to its principal agonist, i.e. shear stress. As well, the ESL modifies the properties of the flowing blood: many of the components of the glycocalyx are electrically charged, which favours interactions with blood cells and proteins.

Figure 1: Endothelial “function” (i.e., the production of protective autacoids by the vascular endothelium) and “dysfunction” (i.e., the involvement of the endothelium in vascular pathology).

When measured at the level of microvessels where the ratio of ESL thickness to vascular lumen is highest, it can be shown that vascular resistance is much higher compared to the resistance that is measured in a glass capillary with the same diameter [7]. This important observation is due to the ESL reducing the functional lumen, that is the space where blood is allowed to flow without spatial restriction or interference from electrically charged structures. The interaction between blood cells and the ESL also contributes importantly to shear stress, with the modifications imposed by this physical force on the ESL and the intracellular cytoskeleton determining the activation of molecular pathways, in particular the synthesis of nitric oxide (NO) [8]. In sum, the mechanical force imposed on the ESL by the flowing blood acts to trigger structural modifications in this complex structure, which in turn functions as a transducer to trigger biochemical reactions in the endothelial cellular layer. In response to shear stress, endothelial cells change their shape into one that is elongated in the direction of the flow, with an eccentric nucleus and increase in thickness. Very quickly after increases in shear stress, rapid changes in ionic conductance, inositol triphosphate production as well as in the cytosolic Ca2+ concentrations can be observed in the endothelial cell. Opening of K+ channels facilitates membrane hyperpolarization, which provides an electrochemical gradient for Ca2+ entry. The plasma membrane thickens and begins to form invaginations that are named caveolae [9]. It is at the level of these sub-cellular structures that the synthesis of


341

NO occurs. NO is a highly reactive free radical that is formed through a two-step oxidation of L-arginine [10]. Among its many effects, NO is a potent modulator of hemorheology [11]: it increases red blood cell and platelet deformability, reduces platelet adhesion and aggregation [12], reduces leukocyte adhesion [13, 14] and, most importantly, causes vasodilation [15]. If hemodynamic forces that determine endothelial activation are maintained over prolonged periods, genomic activation can be observed, probably mediated by activation of the nuclear factor kB (NFkB) transcriptional factor. Activation of this system has been proposed to mediate gene regulatory responses to flow and shear stress changes. Because NFkB binding sites are found in the promoter regions of a variety of genes, this system might be of particular importance in controlling gene expression in response to variations in shear stress.

Figure 2: Two of the roles of the endothelial surface layer: a transducer for mechanical stress and a receptor favouring the interaction with blood cells.

To date, a number of studies have demonstrated that prolonged exposure to increased shear stress not only increases NO production and release, but also increases the expression of endothelial NO synthase, thus providing a molecular basis for this increased production [16]. It is believed that these changes in NO production triggered by increases in shear stress explain the beneficial effects of physical exercise in cardiovascular patients [17]. Taken together, these mechanisms confirm the original observation by the German physiologist Schretzenmayr who, at the beginning of the last century, observed a dilation of large arteries upon exposure to increased flow [18]. Since this observation, it has become clear that blood vessels are not passive tubes, but metabolically active tissues able to react to forces exerted by the circulating blood. In addition to those involved in the regulation of vascular tone, endothelial genes that are either transiently or permanently up regulated by shear stress include transcription factors, growth factors, adhesion molecules and enzymes [19].

342


Interestingly, the redox state of endothelial cells plays a critical role in modulating these processes. The endothelial cell redox state has been found to be dependent on the type of shear stress applied, with steady laminar shear stress causing induction of superoxide-producing NADH oxidase [20]. The importance of reactive oxygen species production as the common pathway of vascular pathobiology cannot be overstated, and it has been discussed in detail elsewhere [21 ,22]. Finally, pulsatile shear stress downregulates the expression of the gene encoding for endothelin-1, a potent vasoconstrictor and a trigger in a feed-forward mechanism of oxygen free radical formation [23]. Oxygen free radicals are important factors of cellular metabolism and play a role in intercellular signalling and defence mechanisms. At the same time, high concentrations of free radicals are very toxic to cellular structures due to their capacity to oxidize and damage or inactivate a variety of cellular structures. Taken together, these phenomena provide a background rationale as to why atherosclerotic lesions preferentially originate in areas of disturbed flow associated with low shear stress [24]. For example, it was observed a few decades ago that changes in either direction from normal levels of shear stress are the most likely explanation for atherosclerosis to develop preferentially at vascular bifurcations and in other areas where anatomical structures cause a deviation from unidirectional laminar blood flow. Since these first observations, evidence supporting the concept that an abnormal, low or high shear stress is a plaque-modulating factor has accumulated [25, 26]. Potential mechanisms that have been proposed for this association include preferential accumulation of LDL particles near the vessel wall, shear-dependent accumulation of inflammatory cells as a process that is promoted by the increased expression of ligands such as ICAM and VCAM, as well as shear stress-dependent gene expression. In regards to this latter possibility, physiological levels of intermittent shear stress have been demonstrated to induce, in vitro, atheroprotective endothelial gene expression patterns, while a low-grade shear stress level is associated with the expression of an atherogenic phenotype [27].

2. Determinants of Shear Stress – Flow Properties of Blood The mechanical forces imposed by vascular hemodynamics upon the vessel walls result in two different types of mechanical stress: one that is circumferential, due as it is to the pressure in the vascular lumen, and shear stress which is longitudinal. The direction of the vector of this second force is determined by the direction of the blood flow at the vessel wall since shear stress results from the blood interacting with the vessel wall. The opposite force, friction, which is applied to the blood by the vascular wall, has a direction opposite to shear stress. Blood flow in arteries, arterioles and capillaries causes shear stresses that are in the range of 0-5 Pa [28]. This force varies in magnitude, frequency and direction during the cardiac cycle, and is influenced by local factors such as the presence of bifurcations or aneurysms and the tortuosity of the vessel. As discussed above, shear stress is applied mainly to the luminal layer of the arterial wall, the vascular endothelium, and is transduced by the glycocalyx to induce a variety of biochemical responses in the endothelial cell. The resultant of circumferential stresses and shear stresses determine vascular diameter via mechanisms that are endothelium-dependent (i.e., mostly NO-mediated dilation) as well as NO-independent (i.e., vascular wall elasticity or distensibility as well as myogenic tone). In turn, shear stress can be estimated as the product of wall shear rate and blood viscosity.


343

In more detail, shear rate is normally expressed as the rate at which adjacent layers of fluid move with respect to each other, usually expressed as inverse seconds or seconds-1. Based on the fundamental assumption of fluid mechanics that the velocity of a fluid at a surface is zero, shear rate expresses the velocity gradient of the fluid between the vascular wall and its maximal flow velocity located at some distance from the wall. On the other hand, viscosity is an intrinsic property of a fluid that essentially describes its capacity to offer resistance to flow. While the mathematics and physics of these characteristics of fluids have been described elsewhere in this book, it has to be emphasized here that the interaction between shear rate and blood viscosity is a critical modulator of endothelial function, and, consequently, of vascular homeostasis. For instance, studies employing blood substitutes have clearly shown that an elevated viscosity elicits a vasodilatory response due to increased shear stresses. In turn, these vasodilatory responses are able to enhance flow within capillaries [29]. The interdependence of these parameters is also very important: Lipowsky et al investigated how blood hematocrit, viscosity and shear rate interact, and showed that blood viscosity is slightly modified by shear rate changes at low levels of hematocrit, while the effect of changes in shear rate on blood viscosity becomes larger when hematocrit increases. At low shear rates, such as those observed in the microcirculation and in veins, variations in hematocrit induce larger changes in blood viscosity. At the same time, the Fåhraeus phenomenon governs changes in local hematocrit that are dependent on the vessel size: because red blood cells tend to travel closer to the centre of the vessels leaving a cellpoor layer at the wall, the local hematocrit within the vessel decreases with decreasing vessel diameter, resulting in a reduction in blood viscosity. In sum, shear rate, hematocrit and viscosity together determine shear stress and, through the endothelial cell’s biochemical apparatus, regulate vascular homeostasis. The next paragraph will discuss how changes in viscosity alter this equilibrium.

3. Blood Hyperviscosity and its Effects on Endothelial Function According to the 1970 Wells’ classification, hyperviscosity syndromes are divided into three types: x Polycythemic syndromes, which follow an increase in the number of circulating blood cells as evidenced by increases of hematocrit. x Sclerocythemic syndromes, where the decreased fluidity of the blood is determined by an altered deformability of red blood cells which are normal in number. x Syndromes associated with an increased serum viscosity where no abnormality can be observed in the number and deformability of red blood cells, but where an altered concentration and/or specific properties of an abnormally produced plasma protein increase blood viscosity. An example of this latter class of hyperviscosity syndromes are paraproteinemias. The clinical conditions associated with an increased viscosity, as well as the symptoms induced by these changes in hemorheology, have been described in detail elsewhere in this book. Briefly, some of the syndromes associated with “primary hyperviscosity” are the following: both primary and secondary forms of polycythemia, contracted plasma volume syndromes, acute and chronic leukemias, reactive leukocytosis, thrombocytosis, thrombocythemia and platelet hyperactivity. Obviously,

344


these forms are associated with an increased blood viscosity due to abnormalities in the cellular component of blood. On the other hand, plasma hyperviscosity is caused by monoclonal immunoglobulins or cryoglobulins as well as in the presence of hyperfibrinogenemia and cryofibrinogenemia. Several lines of evidence demonstrate that these conditions are associated with a worsened endothelial function and patient prognosis. For instance, in the case of sickle cell disease, vasoocclusive crises are a characteristic manifestation and a very common cause of morbidity and mortality. Generally speaking, the accepted concept is that interactions between abnormal blood viscosity and the endothelium may contribute to the pathogenesis of these crises. Sickle erythrocytes, clusters of red or white blood cells, and/or the precipitation of paraproteins may occlude microvessels by adhering directly to the endothelium or by impairing its capacity to react to vasodilator stimuli. In this hypothesis, enhanced adhesion of blood cells to the vascular endothelium on one hand and abnormal vasomotor tone regulation on the other may cooperate, thereby leading to the occurrence of vaso-occlusion. In recent studies, patients with sickle cell disease were shown to have a 50% increase in cardiac output and brachial artery flow [30, 31]. However, despite the increased wall shear stress due to the increase in flow and viscosity, and despite a normal responsiveness to specific endothelial stimuli such as acetylcholine and L-NMMA, these patients had normal resting brachial artery diameters and a markedly blunted flow-mediated dilation. In sum, the relationship between endothelial regulation of vascular tone and red blood cell rheology seems to play a key role in the pathophysiology of these syndromes. One possible explanation to the apparently paradoxical observation that, in conditions of increased viscosity, the NOS and endothelial functions are not upregulated may be that the composition and therefore the viscosity of blood is not uniform throughout the cross-section of the vessel [32]. The reason for such nonuniform distribution is the axial migration of red blood cells, a phenomenon that is further emphasized by increased RBC aggregation. This accumulation of red cells near the center of the vessel causes the hematocrit and blood viscosity to be lower near the vessel wall, therefore resulting in less stimulation of the endothelium. Thus, conditions of increased blood aggregation may paradoxically be associated with a decreased wall shear stress, providing a lesser stimulus to the endothelial production of NO and causing an increase in vascular resistance. Taken together, these data seem to point out that in the course of primary hyperviscosity syndromes the mechanisms responsible for the transduction of the endothelium-dependent vasodilator signal are compromised, and thus there is impaired endothelium-dependent arterial diameter adjustment to the chronically increased wall shear stress. In other words, there is now evidence to support the concept that endothelial dysfunction in patients with syndromes associated with hyperviscosity is due to a specific impairment of shear stress-mediated vasodilation [30]. These abnormalities might have important clinical implications. Failure of vessels to accommodate their lumen diameter to changes in shear stress may indeed favor interactions between blood cells and the vascular walls, triggering vasoocclusive events. Similarly, a reduction in the endothelial NO production, by stimulating platelet adhesion and aggregation and the expression of adhesion molecules such as VCAM-1, may have non-hemodynamic implications that are relevant to vascular disease and atherogenesis. In this scenario, blood viscosity and its major determinants (i.e. hematocrit and plasma viscosity), show a strong association with surrogate markers of systemic atherogenesis such as intimamedia thickness, at least in men [33]. This association remains significant even when


345

major traditional cardiovascular risk factors such as total cholesterol, blood pressure and smoking habits were taken into account. In sum, it is generally accepted that an elevated blood viscosity may promote atherosclerotic development by increasing platelet adhesion to the subendothelium, protein infiltration into the arterial wall and by altering local shear forces. Obviously, these effects result in increased vascular risks due to increased viscosity which increases peripheral resistance and decreases tissue perfusion and, in certain cases, increased blood cell aggregation and/or reduced deformability, leading to vasoocclusive crises.

4. Secondary Blood Hyperviscosity Syndromes Besides the conditions described above which have been associated with increased blood viscosity and are traditionally known as primary hyperviscosity syndromes, it has been widely accepted - since the early 1980’s - that several other conditions are associated with increased plasma viscosity. It has been consistently demonstrated, for instance, that cardiac, peripheral and cerebral ischemia are associated with an increased blood viscosity [34, 35]. Classical models that have led to such observations include patients with Raynaud’s syndrome where the viscosity of the blood leaving ischemic territories is higher than from non-ischemic regions, [36], peripheral arterial disease where blood viscosity appears to be linearly correlated with Fontaine stage, carotid atherosclerosis [37], and cardiac ischemia where we showed that blood viscosity increases in patients who develop ischemia during exercise testing and during atrial pacing [38]. These observations led us to propose a further classification of hyperviscosity syndromes, dividing those conditions where hyperviscosity is the primary mechanism of disease leading to abnormalities in blood flow and ischemia (i.e., Wells’ classification), from those where hyperviscosity is actually caused by, or associated with, ischemia. Since the introduction of this concept of “secondary hyperviscosity syndromes” [39], many advances have been made, and it is now known that blood stasis and activation of the ischemic endothelium leads to a series of molecular events that cause changes in blood viscosity [40]. Consequent to ischemia, significant damage to the microcirculation occurs due to: 1) embolism of thrombotic material from the ruptured plaque; 2) reperfusion damage caused by a rapid increase in oxygen free radical levels that provoke oxidative membrane damage. Structurally, these vascular areas present endothelial protrusions, fibrin clots, platelet aggregates, and adhesion of platelets, red blood cells and leukocytes, and, most importantly, modifications of the deformability of these cells. While these modifications can be observed following a true ischemic injury, abnormalities in cellular aggregability and deformability can also be observed in the absence of true ischemic damage. For example, normalized whole blood viscosity was observed to be significantly increased in the morning hours in patients with risk factors for and/or chronic cardiovascular disease, even in the absence of ongoing ischemia [41, 42, 43]. Several lines of evidence confirm that, during ischemia, viscosity and blood cell deformability are affected. In a recent report, Tozzi-Ciancarelli, et al demonstrated that in a group of patients with myocardial infarction, decreased erythrocyte filtration and increased blood viscosity in the absence of any significant change in plasma viscosity can be observed; these changes were accompanied by an increased rigidity of the erythrocyte membrane [44]. Similar changes induced by ischemia in an animal model have been shown to be associated with an increased production of oxygen free radicals

346


by membrane NADH oxidases [45]. In particular, RBCs appear to be particularly sensitive to oxidative damage due to their richness in iron, a transition metal that catalyzes the formation of reactive species. In the setting of impaired defense mechanisms (e.g., reduced antioxidant enzymes such as superoxide dismutase and reduced glutathione) as found in the presence of cardiovascular risk factors, free radical damage to red blood cells causes a significant increase in membrane rigidity and aggregation and reduced cellular deformability [46]. These data confirm the critical importance of oxygen free radicals in vascular pathophysiology. In another study, low shear rate blood viscosity was shown to be above control levels in patients with acute myocardial infarction, and scaled according to the incidence of risk factors and the presence of coronary artery disease [47]. In this study, the hemorheological pattern at the three month follow up visit was not significantly different from that seen at the first evaluation of the patient. Collectively, these results demonstrate that in young subjects, a secondary hyperviscosity syndrome can be observed in patients with cardiovascular disease. This condition does not seem to be associated with the acute ischemic event per se, but rather it appears to depend on vascular oxygen free radical production. In sum, there is now sufficient evidence to suggest that in these conditions, the hyperviscosity of the effluent blood is most likely a consequence and not a cause of the vascular disease. On the other hand, beyond a certain point these abnormalities in blood viscosity and cellular deformability combine to increase vascular resistance to a point were perfusion is further impaired, thus leading to incremental ischemia. In this scenario, activation of platelets and leukocytes is a critical determinant of the release of cytotoxic oxidants, vasoconstrictor leukotrenes and proteolytic enzymes as well as of complement activation [48]. The mechanism(s) of these changes have been extensively investigated, but nevertheless secondary hyperviscosity syndromes are not as easily explainable as primary ones where specific abnormalities in cellular number or physical properties or abnormalities in specific plasma components can be demonstrated. However, even in secondary syndromes, important reductions in red blood cell and leukocyte deformability [49], as well as an increased expression of binding molecules at the level of both blood and endothelial cells, have been demonstrated. In sum, the increased viscosity in these syndromes is associated with changes in both the physical and the biological properties of these cells. Further, upon activation, leukocytes release a number of lytic enzymes, as well as reactive oxygen species that exert a potent disruptive effect on the endothelial surface layer (ESL). Obviously, a consequence is the exposure of binding domains on the surface of endothelial cells which were previously protected by the ESL. In vitro and in vivo models of ischemia and reperfusion have shown that a reduction in the thickness of the ESL is directly proportional to the impairment in endothelium-mediated vasodilation. The ESL appears to be sensitive to ROS and oxidized LDL [50]. However, the glycosaminoglycans of the ESL are able to bind antioxidant enzymes such as superoxide dismutase and catalase to protect themselves and the underlying endothelial cell [50]. Collectively, these mechanisms are responsible for the hyperviscosity observed in the setting of ischemia.

5. Clinical Implications Nitric oxide (NO), being the key endothelium-derived relaxing factor, has a critical role in maintaining vascular tone and metabolic activity [51]. NO inhibits the action of


347

potent endothelium-derived factors such as angiotensin II and endothelin-1 which favor atherogenesis, and also serves to inhibit platelet and white cell activation and to maintain the vascular smooth muscle in a non-proliferative state. On the other hand, cardiac risk factors in general lead to an increase in oxidative stress, which can result in reduced net NO bioavailability due to direct quenching [22] as well as to an increased expression of both angiotensin II and endothelin-1. It is now well established that the loss of functional integrity of the vascular endothelium is probably the most important risk factor for the progression of atherosclerosis and for plaque rupture (i.e. for acute coronary events such as myocardial infarction) [52]. Recently, studies have demonstrated that endothelial dysfunction has negative prognostic implications for cardiovascular morbidity and mortality [53], and that oxidant injury to the endothelium is a critical determinant of these phenomena [54, 55]. A corollary of our description of the mechanisms coupling shear stress to NO production is that the integrity of the ESL maintains the biochemical activity of the endothelium. When the integrity of this structure is damaged, blood cells become activated, less deformable and start adhering to the endothelial surface. In this scenario, it has been recognized that shear stress has a crucial role in regulating the progression of atherosclerosis. While being a systemic disease, atherosclerosis affects focal areas of the circulation, in particular those areas that are exposed to low shear stress levels. These areas are characterized by slow blood flow whose direction changes during the cardiac cycle. In contrast, those vascular beds with a steady, homogeneous, high-level shear stress are more often disease-free.

Figure 3: An hypothesis regarding the clinical significance of primary and secondary hyperviscosity syndromes: in the former, an excessive mechanical load damages the ESL and endothelial production of NO. In the latter, the increased shear stress might initially serve as a compensatory mechanism to maintain vascular homeostasis.

348


Because shear stress is determined by shear rate and blood viscosity, interpretation of the effects of changes in these parameters might not be as obvious as commonly believed. The increased viscosity observed in coronary artery disease and/or peripheral arteriopathy has been traditionally interpreted as an epiphenomenon, or a contributing factor, of ischemia. However, changes in the hemorheology of cardiovascular patients might, at the beginning, act to increase “functional” viscosity in the endothelial microenvironment. As discussed above, this might serve as a stimulus for endothelial biochemical activity and subsequent NO release, maintaining the vascular environment in an anti-atherogenetic phenotype characterized by increased expression of vasodilators, growth inhibitors, fibrinolytics, antiplatelet factors, antioxidants, and a decreased production of vasconstrictors, growth factors, mediators of inflammation, adhesion molecules and oxidant agents. Additionally, reduced deformability of red blood cells might increase their residence time within microvessels, favoring oxygen extraction and tissue perfusion. In other words, hemorheological changes of secondary syndromes might, at least initially, represent important compensatory mechanisms that tend to normalize vascular homeostasis by enhancing NO synthesis. In agreement with this point of view, a recent paper demonstrates that both acute severe hyperglycemia and infusion of osmotic agents, which cause increased viscosity, are associated with a reduced necrotic area after experimental induction of a myocardial infarction [56]. Failure of these compensatory mechanisms to return to normal conditions might produce the opposite effects, as persistent hyperviscosity will lead to impaired perfusion and further ischemia. In conclusion, the interaction between the ESL, the hemorheological properties of blood, and the biochemical apparatus of the endothelium maintain vascular homeostasis in a complex and coordinated manner. Disruption of this equilibrium is the likely cause of secondary hyperviscosity syndromes which are commonly seen as important predictors of and mechanisms for cardiovascular disease. On the other hand, the modifications leading to blood viscosity might, within certain limits and for short periods of time, actually be the expression of protective mechanisms. The fact that protective mechanisms going awry are further sources of disease is a common finding in nature and in cardiovascular medicine: vascular hypertrophy in hypertension, cardiac dilation in heart failure, and, in general, the immune system disorders are known examples. Secondary hyperviscosity syndromes might be another expression of the same concept.

References [1] [2] [3] [4] [5] [6] [7]

H.F. Galley and N.R. Webster, Physiology of the endothelium, Br. J. Anaesth. 93 (2004), 105-113. S. Verma, M.R. Buchanan and T.J. Anderson, Endothelial function testing as a biomarker of vascular disease, Circulation 108 (2003), 2054-2059. K. Ley, Molecular mechanisms of leukocyte recruitment in the inflammatory process, Cardiovasc. Res. 32 (1996), 733-742. T.A. Springer, Traffic signals on endothelium for lymphocyte recirculation and leukocyte emigration, Annu. Rev. Physiol. 57 (1995), 827-872. L.V. Rao and U.R. Pendurthi, Tissue factor on cells, Blood Coagul. Fibrinolysis. 9 (1998), S27-S35. A.R. Pries, T.W. Secomb and P. Gaehtgens, The endothelial surface layer, Pflugers Arch. 440 (2000), 653-666. A.R. Pries and T.W. Secomb, Microvascular blood viscosity in vivo and the endothelial surface layer, Am. J. Physiol. 289 (2005), H2657-H2664.

T. Gori and S. Forconi / Endothelium and Hemorheology [8] [9] [10] [11] [12] [13] [14] [15] [16] [17]

[18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33]

349

S. Weinbaum, X. Zhang, Y. Han, H. Vink and S.C. Cowin, Mechanotransduction and flow across the endothelial glycocalyx, Proc. Natl. Acad. Sci. U.S.A. 100 (2003), 7988-7995. H. Masuda, K. Kawamura, H. Nanjo, E. Sho, M. Komatsu, T. Sugiyama, A. Sugita, Y. Asari, M. Kobayashi, T. Ebina, N. Hoshi, T.M. Singh, C. Xu and C.K. Zarins, Ultrastructure of endothelial cells under flow alteration, Microsc. Res. Tech. 60 (2003), 2-12. W.K. Alderton, C.E. Cooper and R.G. Knowles, Nitric oxide synthases: structure, function and inhibition, Biochem. J. 357 (2001), 593-615. D.S. Celermajer, Endothelial dysfunction: does it matter? Is it reversible? J. Am. Coll. Cardiol. 30 (1997), 325-333. M.W. Radomski, P. Vallance, G. Whitley, N. Foxwell and S. Moncada, Platelet adhesion to human vascular endothelium is modulated by constitutive and cytokine induced nitric oxide, Cardiovasc. Res. 27 (1993), 1380-1382. C. Godin, A. Caprani, J. Dufaux and P. Flaud, Interactions between neutrophils and endothelial cells, J. Cell Sci. 106 (1993), 441-451. K.M. Naseem, The role of nitric oxide in cardiovascular diseases, Mol. Aspects Med. 26 (2005), 33-65. R.F. Furchgott, M.H. Carvalho, M.T. Khan and K. Matsunaga, Evidence for endothelium-dependent vasodilation of resistance vessels by acetylcholine, Blood Vessels 24 (1987), 145-149. M. Uematsu, Y. Ohara, J.P. Navas, K. Nishida, T.J. Murphy, R.W. Alexander, R.M. Nerem and D.G. Harrison, Regulation of endothelial cell nitric oxide synthase mRNA expression by shear stress, Am. J. Physiol. 269 (1995), C1371-C1378. R. Hambrecht, A. Wolf, S. Gielen, A. Linke, J. Hofer, S. Erbs, N. Schoene and G. Schuler, Effect of exercise on coronary endothelial function in patients with coronary artery disease, N. Engl. J. Med. 342 (2000), 454-460. A. Schretzenmayr, Uber kreislaufregulatorische Vorgange an den grossen Arterien bei der Muskelarbeit, Pflugers Arch. 232 (1933), 743-748. J.N. Topper and M.A., Jr. Gimbrone, Blood flow and vascular gene expression: fluid shear stress as a modulator of endothelial phenotype, Mol. Med. Today. 5 (1999), 40-46. G.W. De Keulenaer, D.C. Chappell, N. Ishizaka, R.M. Nerem, R.W. Alexander and K.K. Griendling, Oscillatory and steady laminar shear stress differentially affect human endothelial redox state: role of a superoxide-producing NADH oxidase, Circ. Res. 82 (1998), 1094-1101. T. Gori and S. Forconi, The role of reactive free radicals in ischemic preconditioning--clinical and evolutionary implications, Clin. Hemorheol. Microcirc. 33 (2005), 19-28. G. Kojda and D. Harrison, Interactions between NO and reactive oxygen species: pathophysiological importance in atherosclerosis, hypertension, diabetes and heart failure, Cardiovasc. Res. 43 (1999), 562571. A.M. Malek, A.L. Greene and S. Izumo, Regulation of endothelin 1 gene by fluid shear stress is transcriptionally mediated and independent of protein kinase C and Camp, Proc. Natl. Acad. Sci. U.S.A. 90 (1993), 5999-6003. M.H. Friedman, C.B. Bargeron, O.J. Deters, G.M. Hutchins and F.F. Mark, Correlation between wall shear and intimal thickness at a coronary artery branch, Atherosclerosis 68 (1987), 27-33. A. Gnasso, C. Irace, C. Carallo, M.S. De Franceschi, C. Motti, P.L. Mattioli and A. Pujia, In vivo association between low wall shear stress and plaque in subjects with asymmetrical carotid atherosclerosis, Stroke 28 (1997), 993-998. D.P. Giddens, C.K. Zarins and S. Glagov, The role of fluid mechanics in the localization and detection of atherosclerosis, J. Biomech. Eng. 115 (1993), 588-594. A.M. Malek, S.L. Alper and S. Izumo, Hemodynamic shear stress and its role in atherosclerosis, JAMA 282 (1999), 2035-2042. D.N. Ku and D.P. Giddens, Pulsatile flow in a model carotid bifurcation, Arteriosclerosis 3 (1983), 3139. A.G. Tsai, P. Cabrales and M. Intaglietta, Oxygen-carrying blood substitutes: a microvascular perspective, Expert. Opin. Biol. Ther. 4 (2004), 1147-1157. L. Belhassen, G. Pelle, S. Sediame, D. Bachir, C. Carville, C. Bucherer, C. Lacombe, F. Galacteros and S. Adnot, Endothelial dysfunction in patients with sickle cell disease is related to selective impairment of shear stress-mediated vasodilation, Blood 97 (2001), 1584-1589. W. Covitz, M. Espeland, D. Gallagher, W. Hellenbrand, S. Leff and N. Talner, The heart in sickle cell anemia. The Cooperative Study of Sickle Cell Disease (CSSCD), Chest, 108 (1995), 1214-1219. O.K. Baskurt, O. Yalcin and H.J. Meiselman, Hemorheology and vascular control mechanisms, Clin. Hemorheol. Microcirc. 30 (2004), 169-178. A.J. Lee, P.I. Mowbray, G.D. Lowe, A. Rumley, F.G. Fowkes and P.L. Allan, Blood viscosity and elevated carotid intima-media thickness in men and women: the Edinburgh Artery Study, Circulation 97 (1998), 1467-1473.

350


[34] T. Di Perri, M. Guerrini, F.L. Pasini, A. Acciavatti, D. Pieragalli, C. Galigani, P.L. Capecchi, A. Orrico, M. Franchi and P. Blardi, Hemorheological factors in the pathophysiology of acute and chronic cerebrovascular disease, Cephalalgia 5 Suppl 2 (1985), 71-77. [35] J.A. Dormandy, C.J. Yates and G.A. Berent, Clinical relevance of blood viscosity and red cell deformability including newer therapeutic aspects, Angiology 32 (1981), 236-242. [36] S. Forconi, M. Guerrini, D. Agnusdei, F.L. Pasini and T. Di Perri, Letter: Abnormal blood viscosity in Raynaud's phenomenon, Lancet 2 (1976), 586. [37] C. Carallo, A. Pujia, C. Irace, M.S. De Franceschi, C. Motti and A. Gnasso, Whole blood viscosity and haematocrit are associated with internal carotid atherosclerosis in men, Coron. Artery Dis. 9 (1998), 113117. [38] S. Forconi, M. Guerrini, D. Pieragalli, A. Acciavatti, C. Del Bigo, C. Galigani and T. Di Perri, [Hemorrheological changes in ischemic heart disease], Ric. Clin. Lab. 13 Suppl 3 (1983), 195-208. [39] S.Forconi, D. Pieragalli, M. Guerrini, C. Galigani and R. Cappelli, Primary and secondary blood hyperviscosity syndromes, and syndromes associated with blood hyperviscosity, Drugs 33 Suppl 2 (1987), 19-26. [40] S. Forconi, From hyperviscosity to endothelial dysfunction: a return trip? Clin. Hemorheol. Microcirc. 30 (2004), 155-165. [41] L. Nobili, G. Schiavi, E. Bozano, F. De Carli, F. Ferrillo and F, Nobili, Morning increase of whole blood viscosity in obstructive sleep apnea syndrome, Clin. Hemorheol. Microcirc. 22 (2000), 21-27. [42] N. Antonova and I. Velcheva, Hemorheological disturbances and characteristic parameters in patients with cerebrovascular disease, Clin. Hemorheol. Microcirc. 21 (1999), 405-408. [43] M. Mares, C. Bertolo, V. Terribile and A. Girolami, Hemorheological study in patients with coronary artery disease, Cardiology 78 (1991), 111-116. [44] M.G. Tozzi-Ciancarelli, C. Di Massimo, A. Mascioli, E. Tozzi, P. Gallo, F. Fedele and A. Dagianti, Rheological features of erythrocytes in acute myocardial infarction, Cardioscience 4 (1993), 231-234. [45] N. Nemeth, T. Lesznyak, M. Szokoly, I. Furka and I. Miko, Allopurinol prevents erythrocyte deformability impairing but not the hematological alterations after limb ischemia-reperfusion in rats, J. Invest. Surg. 19 (2006), 47-56. [46] O.K. Baskurt, A. Temiz and H.J. Meiselman, Effect of superoxide anions on red blood cell rheologic properties, Free Rad. Biol. Med. 24 (1998), 102-110. [47] G. Caimi, E. Hoffmann, M. Montana, B. Canino, F. Dispensa, A. Catania and R. Lo Presti, Haemorheological pattern in young adults with acute myocardial infarction, Clin. Hemorheol. Microcirc. 29 (2003), 11-18. [48] R.L. Engler, G.W. Schmid-Schonbein and R.S. Pavelec, Leukocyte capillary plugging in myocardial ischemia and reperfusion in the dog, Am. J. Pathol. 111 (1983), 98-111. [49] J. Dormandy, E. Ernst and D. Bennett, Erythrocyte deformability in the pathophysiology of the microcirculation, Ric. Clin. Lab. 11 Suppl 1 (1981), 35-38. [50] H. Vink, A.A. Constantinescu and J.A. Spaan, Oxidized lipoproteins degrade the endothelial surface layer : implications for platelet-endothelial cell adhesion, Circulation 101 (2000), 1500-1502. [51] R.F. Furchgott and J.V. Zawadzki, The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine, Nature 288 (1980), 373-376. [52] C. Pepine, Endothelial dysfunction and its role in the cycle of cardiovascular disease, Can. J. Cardiol. 14 suppl D (1998), 5D-7D. [53] F. Perticone, R. Ceravolo, A. Pujia, G. Ventura, S. Iacopino, A. Scozzafava, A. Ferraro, M. Chello, P. Mastroroberto, P. Verdecchia and G. Schillaci, Prognostic significance of endothelial dysfunction in hypertensive patients, Circulation 104 (2001), 191-196. [54] P.O. Bonetti, L.O. Lerman and A. Lerman, Endothelial dysfunction: a marker of atherosclerotic risk, Arterioscler. Thromb. Vasc. Biol. 23 (2003), 168-175. [55] T. Heitzer, T. Schlinzig, K. Krohn, T. Meinertz and T. Munzel, Endothelial dysfunction, oxidative stress, and risk of cardiovascular events in patients with coronary artery disease, Circulation 104 (2001), 26732678. [56] H. Chen, W.L. Shen, X.H. Wang, H.Z. Chen, J.Z. Gu, J. Fu, Y.F. Ni, P.J. Gao, D.L. Zhu and H. Higashino, Paradoxically enhanced heart tolerance to ischaemia in type 1 diabetes and role of increased osmolarity, Clin. Exp. Pharmacol. Physiol. 33 (2006), 910-916.


351

Methods in Hemodynamics Sehyun SHINa, Hideyuki NIIMIb, Max R. HARDEMANc,1, Peter.T. GOEDHARTc a Department of Mechanical Engineering, Korea University, Seoul, Korea b National Cardiovascular center Research Institute, Osaka, Japan & Division of Research Affairs, Faculty of Medicine, Chulalongkorn University, Bangkok, Thialand c Department of Physiology, Academic Medical Center, Amsterdam, the Netherlands

Introduction Over the last 30 years significant progress has been made in the fields of hemodynamics and hemorheology, spurred on by innovative developments in measurement techniques [1] and instrumentation [2-3]. Measurements obtained via these methods have been used for monitoring hemodynamic phenomena and diagnosing circulatory disorders, thus providing a deeper understanding of hemodynamic-related diseases in humans [4]. Understanding of hemodynamics was greatly enhanced by Poiseuille (1799-1869), who developed a relation between flow rate and pressure (Poiseuille’s law). The most frequently measured parameters in hemodynamics are blood flow rate and blood pressure, and thus one can calculate resistance to flow as the ratio of pressure to flow. Of course, this resistance concept is a “black box” approach in which all of the parameters that affect resistance are lumped together. For in vitro studies of flow in single tubes or networks, it is rather straightforward to define and measure such parameters (e.g., diameter, viscosity, length) and to understand their interrelations. Conversely, it is difficult to measure the in-vivo hemodynamic parameters. Recently, however, developments in optics and electronics have resulted in various techniques for measuring these hemodynamic parameters, with some techniques commercially available. The objective of this chapter is to introduce the principles and application of conventional and new techniques for laboratory and clinical measurements of hemodynamic parameters.

1. Blood Pressure Blood pressure is the most routinely measured hemodynamic parameter since it is a good index to assess the state of the circulatory system. The routine measurement of blood pressure should be a non-invasive, painless, simple, quick, and indirect method. It is hard to imagine thatany pressure sensors would require insertion into an artery for making routine pressure measurements. Instead, clinicians determine systolic and diastolic pressure by indirect methods such as the auscultatory and palpatory methods. However, there are also direct methods to measure arterial pressure. Blood pressure is 1

Corresponding author: Department of Physiology, Academic Medical Center, Amsterdam, the Netherlands; E mail: [email protected]

352

S. Shin et al. / Methods in Hemodynamics

defined as the normal force exerted by blood on a unit area of the vessel wall, and is measured in terms of the height of a column of fluid, usually as millimeters of mercury (mmHg) because mercury has a suitable density [4]. Heights of columns of other fluids can be used (e.g., water) since conversion of units only requires knowledge of fluid densities; pressure can also be expressed as fundamental units (e.g., dynes/cm2, N/m2). 1.1. Auscultatory Method The most commonly used method for measuring blood pressure is the auscultatory (Riva-Rocci) method, which employs a sphygmomanometer and a stethoscope (Figure 1)[3]. When pulsatile blood flow passes though large and smooth arteries, the blood flow is laminar and any flow-induced sound is inaudible by a stethoscope. However, when the pulsatile blood flow passes though an externally compressed artery, the flow becomes turbulent and creates flow-induced low-frequency sounds. These sounds, the so called Korotkoff sounds, occur during the period between systole and diastole and are audible with a standard stethoscope or with specialized electronic microphones. Thus, one can identify systolic and diastolic pressure using the Korotkoff sounds with a sphygmomanometer. However, this auscultatory method does not easily allow measurement of diastolic pressure because no specific sound enables an exact identification of diastole; systolic pressure is taken as the start of sounds whereas diastolic has a very “soft” endpoint. Further, there can be many poorly audible sounds prior to systole which can cause confusion. Korotkoff sound

(mmHg)

Cuff pressure Systolic pressure

120 80 40

Mercury manometer

Diastolic pressure

0

Figure 1: Ausculatory method of measuring systolic and diastolic arterial pressure

1.2. Palpatory Method This method is simple and requires a blood pressure cuff, an aneroid gauge or a mercury manometer, and palpable brachial or radial pulses. This method is highly subjective and is of limited value for serial measurements that involve numerous examiners. During slow cuff deflation, the point of return of a palpable pulse in an

353


artery distal to the cuff is taken as the initial return of blood flow. The simultaneously measured pressure value is then assumed to be the systolic pressure. 1.3. Systolic Pressure Manual Methods Oximeters that are currently used for monitoring arterial oxygen saturation also serve as pulse detectors for estimating systolic pressure by pulse reappearance during cuff deflation. This method involves the measurement of light absorption from the photoelectric device placed against the skin on a finger. Changes in absorption of infrared light are produced by phase changes in blood volume associated with pulsatile blood flow. A pulse wave identical in appearance to an intra-arterial waveform is displayed. 1.4. Automated Pressure Measurements Automated indirect blood pressure measuring devices provide measurements of systolic, diastolic, and mean arterial pressures without a stethoscope, and without manual inflation and deflation of the cuff. These automatic devices utilize various technologies (Figure 2) [5]: 1. Oscillometric detection of pulsatile arterial wall vibration by an automatically inflating/deflating cuff coupled to electronic instrumentation; 2. Infrasonde detection, by microphone, of the pulsatile vibrations of the arterial wall underlying the automatically deflating cuff; 3. Ultrasonic detection (using Doppler principle) of the speed and direction of blood flow by determination of the changes in pitch in the sound of flowing blood as it gets nearer to and then suddenly “downlifts” as it passes a sensing device housed within the encircling cuff; 4. Arterial tonometry, which uses a pressure sensor, positioned over and partly flattening the radial artery against the underlying bone. The pressure sensor detects the pulsatile forces exerted by the artery. The beat-to-beat forces are then converted to an electrical signal, which is computer analyzed and continuously displayed.

Figure 2: Recent developments of automated pressure measuring systems

354


1.5. Direct (Invasive) Pressure Measurement Direct pressure measurement is an invasive method requiring puncture of a vessel and insertion of a catheter. The most common puncture sites are brachial and radial arteries although other sites can be used (e.g., femoral artery). Direct pressure measurements can be divided into extra-vascular and intra-vascular techniques. Extra-vascular pressure can be measured using a catheter equipped with a three-way valve stopcock and a pressure sensor. The catheter is inserted into a vessel by means of percutaneous puncture using a needle or guide wire. The catheter is positioned in an area of interest and blood pressure is transmitted to the external pressure sensor. Intra-vascular pressure measurements are sometimes required to collect information from specific locations, and can be measured by inserting a miniature sensor into a blood vessel. Intra-vascular pressure measurement eliminates the undesired pressure loss through a connecting tube or catheter, thus allowing high-frequency pressure data to be obtained. Different types of pressure transducers have been used for direct measurements of blood pressure, including a strain gauge, linear-variable differential transformer, conductive polymer sensors, fiber-optic pressure sensors, piezoelectric materials, and semiconductor devices [6].

2. Blood Flow Blood flow is an important index to assess the function of the heart and circulatory system, and is defined as the volume of blood that passes a given point in a given period. Blood flow is frequently used as a reference value to describe cardiac output, which is defined as the volume of blood leaving the heart in a given period. Blood flow rate can be expressed as milliliters per second (ml/s) or liters per minute (l/min), and can be measured by various methods and devices termed flow meters. 2.1. Indicator Dilution Method The so-called Fick principle, named after the German physicist Adolph Fick, has been developed to measure blood flow using the so-called “indicator dilution method” [7]. While this dilution concept is frequently utilized to measure a total volume (i.e., dilution of a known amount of material in an unknown volume), it can also be applied to determine flow rate. A substance m is injected into the blood stream at some rate, and the flow rate of blood through a given vessel can be determined if the injection rate of the indicator and the change in concentration of the indicator in the blood are known: Flow rate (Q) =

dV dt

dm / dt 'C

where V is the blood volume, m is the quantity of indicator substance injected, dm/dt is the injection rate of m, and C is the concentration of m in blood. ˂C is the change at a single point at two different times or the instantaneous difference in concentration between two different locations.


355

2.1.1. Fick Method Cardiac output measurement in-vivo was introduced in 1870 based on the Fick principle, which relates the cardiac output (Q) to oxygen consumption and to the arterial and venous concentrations of oxygen: Q

dV dt

dm / dt 'C

dm / dt C a Cv

where Q is the cardiac output, dm/dt is the consumption of O2, and Ca and Cv are the arterial and venous concentrations of oxygen. In this case, oxygen is used as the indicator to determine cardiac output, and a device called a spirometer is used to measure oxygen consumption. Simultaneously, blood samples are taken, preferentially from the right ventricle or pulmonary artery and from a systemic artery [8]. 2.1.2. Dye Dilution method Indocyanine green, a dye which has been proven to be clinically safe, is used as an injected indicator to measure blood flow. This method can measure not only cardiac output, but also local cerebral blood flow. It is based on the continuous measurement of the concentration of a dye in the blood downstream of the injection site where the dye is injected as a bolus (Figure 3). This method generates a dye concentration versus time curve. The area under the curve represents the known amount of dye injected. The flow rate can be determined by the following equation: q

C (t )Q(t 2 t1 ) where C

³

t2

t1

c dt t 2 t1

Figure 3: Schematic for measuring blood flow with the dye dilution principle.

2.1.3. Thermodilution Method The thermodilution method utilizes temperature as an indicator for measuring blood flow. Cold saline is injected into the right atrium and a thermistor is placed in the pulmonary artery to measure blood temperature. Then, flow rate can be determined by the following equation:

356


q

Q

t

U b c b ³ ' T b ( t ) dt 0

where q is the heat content of injected saline (J), Ub is the density of blood, cb is the specific heat of blood, and ˂T is temperature change. 2.2. Electromagnetic Flowmeter When a conducting material passes through a magnetic field, an electromotive force (EMF) is generated in the conducting material [9] and the EMF results in a flow of current. This is well known principle for production of electricity by an electrical generator. Based on this principle, an electromagnetic flowmeter can measure flow rate of a conducting fluid which passes through a steady magnetic field. Figure 4 shows that the same principle applies for generation of EMF in blood when it moves through a magnetic field [4]. The blood vessel is placed between the poles of a permanent magnet, and electrodes are connected on the two sides of the vessel perpendicular to the magnetic lines of force. When blood passes through the vessel, the generated EMF between the two electrodes is proportional to the mean (i.e., average) velocity of blood in the vessel. The EMF generated between the electrodes placed on either side of vessel is given in the following equation. + o -

N

+

S

N

S

(a)

(b)

Figure. 4: Principle of the electromagnetic flowmeter [4]. (a) EMF generation in a wire when a conducting material passes through a magnetic field; (b) EMF generation in electrodes on a vessel containing flowing blood.

e &

³

L

0

& & & u u B dL

&

&

where u is the instantaneous blood velocity, B is the magnetic flux density, and L is the distance between electrodes. For a uniform magnetic flux density B and a uniform velocity profile (U), the above equation can be simplified into e = BLU. The electromagnetic flowmeter has several advantages in measuring blood flow rate [9]. The main advantage is its ability to measure blood flow without opening the vessel so that it may be used in clinical environments. Another advantage is that it can


357

record changes in flow occurring in less than 0.01 second, allowing accurate recording of pulsatile changes in flow as well as steady flow. 2.3. Ultrasonic Doppler Flowmeter Another type of non-invasive flowmeter is the ultrasonic flowmeter, which has many of the same advantages as the electromagnetic flowmeter [10]. The ultrasonic flowmeter is based on the Doppler Effect (Figure 5): when a target recedes from a fixed sound transmitter, the frequency becomes lower. For an ultrasonic flowmeter, a small piezoelectric transducer functions as the sound transmitter as well as sound detector, while the blood cells are the moving targets. A piezoelectric transducer mounted on the surface of the device generates and transmits sound at a frequency of several MHz along the flowing blood. A part of the transmitted sound is reflected by blood cells in the blood stream, and the reflected sound wave travels backward from the cells toward the piezoelectric transducer. The reflected waves have a lower frequency than the original transmitted wave since the red cells are moving away from the piezoelectric transducer. For relatively small changes of frequency, the following equation is valid: Fd / Fo u / c . Here Fd is the Doppler shift frequency, Fo is the source frequency, u is the target velocity, and c is the velocity of sound. Like the electromagnetic flowmeter, the ultrasonic flowmeter is capable of measuring rapid, pulsatile changes of blood flow.

Re ce Em ive it

u RBCs Figure 5: Schematic of ultrasonic Doppler flowmeter [7]

2.4. Laser Doppler Flowmeter The same physical principle of frequency shift as used in the ultrasound Doppler flowmeter is used in a laser Doppler flowmeter termed the O2C (Oxygen to See, LEA, Giessen, Germany). In this apparatus, the laser Doppler flow measurement is combined with a hemoglobin saturation measurement (Figure 6) [11].

3. Velocity Measurement 3.1. Particle Tracing Methods When blood flows steadily through microvessels, a single blood cell (red, white or platelet) or a group of cells have a streamline or trajectory along the axis of vessel. Tracing their trajectories enables one to evaluate velocities of the cells using particle

358


tracing methods. There are a number of techniques used to visualize the trajectory, including injection of dye or fluorescence tracers into blood. Instantaneous velocities of cells can be calculated by measuring the length of their movement during consecutive frames based on video-microscopic recordings. If multiple positions of cells are available in a single video frame, velocity profile of the cells can also be calculated; multiple positions can be obtained using strobe-illumination or photo-excitation synchronized with video framing.

Figure 6: Overview of the laser Doppler flowmeter [11]

3.2. Particle Image Velocimetry Particle image velocimetry (PIV) has potential for quantitatively measuring pulsatile or non-pulsatile blood flow through microvessels when used in combination with highspeed video images. In principle, the PIV technique detects particle velocities using a cross-correlation technique on different, consecutive microscopic images. In order to overcome optical limitations related to resolving power and out-of-focus images, a novel PIV technique has been developed to improve measurement accuracy by using microparticles or a spatio-temporal derivative technique (e.g., sub-pixel analysis). For measurement of pulsatile blood flow in a microvessel, a few seconds of videomicroscopic images are used together with a simultaneous recording of blood pressure. Velocity vectors of red blood cells are analyzed in frames corresponding to one cardiac cycle. Three types of average values (i.e., time-, moving- and ensemble-average) are calculated to allow fuller understanding of the velocity fields. Figure 7 shows an example of blood flow at a bifurcation of arterioles in the rat mesenteric microvasculature [12].

359


Figure 7: Blood flow visualization using PIV [12]. (a) An instantaneous video image of red cell flow at bifurcation of rat mesenteric arterioles using a high-speed digital camera (1000 frames/s with 1/2000 shutter). Vessel diameters are 28 μm (parent) and 20 and 18 μm in #1 and #2 daughter vessels. Six cross-sections are chosen for the analysis of velocity profiles, indicated by lines 1-1’ to 6-6’. (b) Instantaneous calculated twodimensional velocity vectors. (c) Time-averaged velocity profiles at cross-sections 1-1’ to 6-6’. Dash-dotted lines are centerlines of vessels, dotted lines are extended centerlines of daughter vessels.

3.3. Optical Coherence Tomography Optical coherence tomography (OCT) is a novel technique to detect the Doppler frequency shift due to red blood cells flowing through microvessels. It can provide the cross-sectional distribution of red cell velocity and hence the velocity profile can be rapidly measured. For measurement of blood flow velocity in a microvessel, Doppler frequency shift due to light scattering by red blood cells is calculated using a short-time Fourier transform of the original un-filtered OCT signal. The OCT signal is sampled during repeated OCT scans across the vessel, and its signal is analyzed off-line to yield a spectrogram. To obtain the absolute velocity of red blood cells, the angle between the optical and vessel axes is measured as follows: two consecutive cross-sectional images of the tissue, including the target vessel, are taken with OCT along the vessel axis, and the angle is determined from 3-dimensional reconstruction of the vessel.

(mmHg)

(mm/s)

30 20 10 0 100

50

0

0

1

2

3

4

Phase (S) Figure 8: Red cell velocity along the vertical diameter of an arteriole (diameter: 62.4μm) in rat cerebral cortex measured by Doppler OCT [13]. Temporal changes of red cell velocity (Vrbc) in every OCT scan and femoral arterial pressure (Pfemo) measured simultaneously.

360


Figure 8 shows an example of a velocity profile in a pial microvessel using Doppler OCT [13]. The scan speed is 10 mm/s and the scan length is 0.5mm; the centerline velocity is obtained once every 50 ms. The phasic change of the velocity with cardiac pulsation is determined by rearranging the centerline velocities in the order of cardiac phase as determined using simultaneously recorded arterial pressure. 3.4. Ultrasound PIV with Flow Ttracer The ultrasound PIV method, termed echo PIV, is an innovative technique that is capable of measuring a time-resolved, two-dimensional flow field. The echo PIV method is a combination of echo imaging and the PIV velocity measurement technique. In fact, the echo imaging technique enables clinicians to visualize flow in spite of the opaque nature of blood, while the PIV technique provides instantaneous twodimensional velocity fields. In the echo PIV technique, radio frequency-formatted data from sequential B (i.e., brightness)-mode scan images of flow, seeded with small gasfilled microbubbles acting as the ultrasound contrast agent, are analyzed using particle image velocimetry (PIV) techniques (Figure 9). The echo PIV method consists of identifying and tracking a flow tracer (e.g., microbubbles) within the flow field, then computing local velocity vectors using a velocimetry algorithm (Figure 10) [14].

Figure 9: Use of non-linear backscattering to identify the microbubbles

RF-formatted data 1.0

Flow direction

FFT-based cross correlation Correlation based error correction

0.5

D

r/R

Ultrasound beam

Decrease window size with offsetting

0.0

Probe

0

« optical PIV

-0.5

-0.02

c echo PIV (D=90q) echo PIV (D=75q)

-0.04

Velocity vector result Y (m)

-0.06

-1.0

-0.08

0.0

-0.1

Scan converted vector

0.25

0.5

0.75

1.0

U/Uc

-0.12 -0.14 -0.06

-0.04

-0.02

0 X (m)

0.02

0.04

0.06

Figure10: Flow chart for echo PIV algorithm and comparison of echo, optical and analytic velocity profiles of laminar steady tube flow (Courtesy of Prof. Jean Hertzberg, U. Colorado, USA)

361


An advantage of the echo PIV is the excellent measurement of shear rate as well as the instantaneous velocity of pulsatile blood flow when compared to optical PIV. Despite somewhat underestimating pulsatile wall shear, echo PIV measurement of shear rate should be useful for examining flow in relatively superficial arteries and veins such as the carotid, brachial, and femoral vessels, and in pediatric patients. Another advantage of the echo PIV is that this technique is capable of measuring blood flow non-invasively and overcomes the angle dependence of the Doppler technique. 3.5. X-Ray PIV The x-ray PIV method is an innovative technique which is capable of measuring a time-resolved, two-dimensional flow field surrounded by opaque blood vessels and tissues [15]. This method is a combination of synchrotron x-ray imaging and the PIV velocity measurement technique (Figure 11). The method can visualize in vivo blood flow patterns by enhancing the phase-contrast and interference characteristics of blood cells without using any contrast media or seeding particles. The enhanced x-ray images are achieved by optimizing the sample-to-scintillator distance, the sample thickness, and hematocrit. The quantitative velocity fields of blood flow inside blood vessels are obtained by applying a two-frame PIV algorithm to the x-ray images of flow; the measured velocity field data show good agreement with the analytical and optical PIV results. In the x-ray PIV method, speckle patterns of blood flow are illuminated with coherent synchrotron x-rays which induce classic Fresnel edge diffraction in the radiological images. The speckles in the captured x-ray images make it easy to discern the edges of specimens such as RBC. By optimizing the sample-to-scintillator distance, clear x-ray images of blood flow can be obtained using the propagation-based phasecontrast method. When an x-ray beam passes through a thick blood sample, the x-rays pass through many layers of blood cells causing a number of individual speckle patterns to be superimposed on the x-ray image. Interestingly, in x-ray imaging of blood, the superimposed pattern of a blood sample is clearly distinguishable and has a nearly uniform size, similar to that of RBC.

Sample

Mechanical shutter

H

d

Scintillator Mirror

x-ray BNC B

C

D

B

C

D

EXT/GATE

Sample holder

BNC

Delay Generator

A

B

C

D

A

B

C

D

EXT/GATE

CCD camera

Objective lens

Attenuator

A A

Shutter Controller

visible light

Slit set

Data acquisition system

Figure 11: Schematic diagram of x-ray PIV system (Courtesy of Prof S.J. Lee, POSTECH, Korea)

362


4. Vascular and Tissue Imaging 4.1. Intravital Microscopy Intravital microscopy is a technique to visualize the in vivo microcirculation [16, 17] and provides detailed pictures of this portion of the circulation. It uses microscope methods to visualize the microcirculation, which can be a drawback since the sophisticated microscopes that are required are not suitable for bedside measurements. Further, in some techniques, intravital microscopy is an invasive technique. 4.2. OPS Imaging The Orthogonal Polarization System (OPS) uses a small, handheld video microscope for non-invasive imaging of the microcirculation [18, 19]. In a normal microscope placed on the nail fold, on the skin, on the sublingual area of the mouth, or on any other organ, most of the light is directly reflected back from the surface, and hence blurs the image of the underlying vessel structures. OPS imaging uses green light that is mostly absorbed by erythrocytes and not by tissue. The incoming light is polarized in one direction; the reflected light from the surface is still polarized and is trapped by a second orthogonally placed polarization filter. The only light that reaches the video camera comes from deeper tissues and is no longer polarized due to multiple reflections in the tissue (Figure 12).

Figure 12: Schematic of OPS imaging system and a view of the sublingual microcirculation [18]. The image shows the flow of blood against a light background; it is usually possible to visualize individual red and white cells.

4.3. Sidestream Dark Field (SDF) Sidestream Dark Field (SDF) imaging [20, 21] produces views of the microcirculation similar to OPS imaging; the device is termed the MicroScan Video Microscope System (Microvision Medical, Amsterdam). The difference between this technique and OPS imaging is that the light provided by flashing LED’s around the optical tip of the apparatus is synchronous with the video frame rate (Figure 13). This eliminates scattered light from reaching the camera and the internal optics. The flashing LEDs also eliminate some of the motion blur that occurs when there is a fast flow of erythrocytes.

363


CCD camera Magnifying lens Green LED

Figure 13: Schematic of SDF imaging system and SDF image of vessels [20].

4.4. Optical Coherence Tomography

Optical coherence tomography (OCT) can provide cross-sectional images of biological tissues to a few millimeters depth with high spatial resolution [22, 23]. The OCT system is, in essence, composed of a fiber optic interferometer using a lowcoherence light source; the center frequency is 813 nm and the spectral width is 21 nm. The light intensity reflected from the specified depth within the target sample is extracted using coherence gating. The spatial resolution in the longitudinal direction is as small as the coherence length of the light source, which is on the order of 10 Pm. For example, in cortical tissue, the longitudinal resolution is around 11μm while the lateral resolution is around 14 μm. Figure 14 shows an example of OCT images of rat cerebral cortex [23]. (mm) 0

pial arteriole

0

scan line

0.5

1 1 (mm)

Figure 14: In vivo OCT image of a rat cerebral cortex. A pial arteriole (diameter: 83μm), the dura, arachnoid membrane, and cortical tissue are clearly observed [23].

4.5. Fluorescence Visualization The microcirculation in some organs and tissue, such as ear and brain, can be visualized using fluorescence microscopy by labeling the cell nucleus, membrane or matrix with fluorescent substances. There are several fluorescent substances used for

364


visualization: 1) Fluorescein-isothiocyanate (FITC)-labeled red blood cells are used to visualize the motion of cells flowing through microvessels [24]; 2) Rhodamine-B isothiocyanate (RITC)-labeled dextran (MW 150 kDa) has been used to visualize the lumen of microvessels; 3) Rhodamine 6G is used to stain the mitochondria of leukocytes in order to study the behavior of leukocytes in microvessels and to analyze leukocyte-endothelial interactions. The fluorescent tracers are injected intravenously through the femoral vein or other venous site; in vivo observations are made via a fluorescence microscope equipped with a sensitive video-camera having various filters, and are recorded on video tape or disc. A microvascular network can be reconstructed from fluorescence images obtained by scanning a wide area of the cortex; Figure 15 presents an example of a cerebral microvascular network reconstruction [25].

Figure 15: A microvascular network in the cerebral cortex using fluorescence visualization [25]. The network is reconstructed from Rhodamine-dextran video images scanned over a wide area of the cerebrocortical surface areas (2-5 mm2) in the cat left marginal gyrus (A: arterioles, V: venules).

4.6. Diffuse Optical Tomography Diffuse optical tomography (DOT) is an emerging medical technique for functional imaging of biological tissues through large depths (e.g., greater than several centimeters) [26]. The basic principle of DOT imaging is to illuminate the tissue with an array of light sources and to measure the light leaving the tissue with an array of detectors. For instance, when a flashlight is shined onto a hand, one can clearly see that light can travel through centimeters of tissue and still be detected. Figure 16 describes how photons may travel in tissue. Similarly, when tissue is illuminated by near infra red (NIR) light from an array of sources, the scattered light can be detected with an array of detectors. The localized optical properties of the illuminated tissue can be inferred using a model of the propagation physics. For each source location, one records an image of the light reaching each detector from that particular source. A model of the propagation of light in tissue has been developed and parameterized in terms of the unknown scattering and/or absorption as a function of position in the tissue. Then, using the model together with the ensemble of images over all the sources, one can attempt to “invert” the propagation model to recover the parameters of interest, or, in other words, to estimate the scattering and/or absorption parameters from the data using the model.


365

Figure 16: Schematic of photon trajectories in tissues

Fortunately, the primary absorbers such as water, oxy-hemoglobin (HbO) and deoxy-hemoglobin (HHb), have relatively weak absorption at the wavelengths of NIR(near infrared) light; this fortuitous phenomenon provides a spectral range which allows one to see the HbO and HHb forms of hemoglobin. Moreover, various tissue types can be differentiated due to their distinct scattering properties. Thus, in addition to tissue scattering characteristics, Diffuse Optical Tomography (DOT) can image three-dimensional spatial variations in blood parameters, particularly hemoglobin concentration and oxygen saturation, and thus metabolic factors which these concentrations reflect. While the DOT method can never compete with anatomical imaging techniques (e.g., x-radiography, ultrasound and MRI) in terms of the spatial resolution, it has several distinct advantages in sensitivity to functional changes, safety, cost-effectiveness and ease of use at the bedside [27]. Currently, the most important applications of DOT are in the screening, diagnosis, and basic research of breast cancer, and the study of the brain, including stroke, hemorrhage, and brain function. The basic concept of using DOT to detect breast cancer is based upon the greater blood supply of tumors compared to surrounding tissue, and thus absorption inhomogeneity. A similar idea allows one to image bleeding in the brain, as well as the association between cerebral activity and increased oxygen supply. 4.7. Laser Speckle Imaging Laser speckle imaging (LSI) is an optical technique which can measure local dynamic properties in scattering media at both high temporal and spatial resolution [28]. The basic principle of LSI is equivalent to that of laser Doppler velocimetry (LDV). Laser speckle is a random interference effect that gives a grainy appearance to objects illuminated by laser light (Figure 17). If the object consists of individual moving scatterers such as blood cells, scattered light forms a fluctuating speckle pattern and shows a frequency shift which is directly proportional to velocity (i.e., the Doppler Effect). Since these fluctuations provide information about the velocity distribution of the scatterers, LSI records temporal fluctuations of the scattered light in order to map the dynamics of a moving object rather than merely analyzing Doppler frequency shifts.

366


Laser Cuvette

Speckle

Figure 17: Principle of laser speckling [28]

In fact, the frequency and time domain in quasi-elastic light scattering are completely interchangeable by using a simple Fourier transformation. Thus, if the distribution of temporal fluctuations is known, the spectrum of frequency shifts can be calculated and vice versa. The time-varying laser speckle method has been applied to the monitoring of blood flow, mainly in the retina and capillaries. Time-varying speckle imaging is an emerging technique that can be used noninvasively to measure capillary blood flow in the skin. This method works by analyzing the intensity fluctuations in scattered laser light. Laser speckle contrast imaging has the advantage of being truly real-time and based upon easily available components. It is also capable of producing video-rate images of changing perfusion maps. In brief, laser speckle imaging can achieve real-time operation at low cost and can provide images of dynamic changes in blood flow; it is anticipated that ongoing developments will make laser speckle contrast imaging a competitive alternative to laser Doppler imaging.

References J.M. Hasenkam, In vivo hemodynamics measurements, In: Advances in hemodynamics and hemorheology, T.V. Ho, Ed., Jai Press, Inc., New York, 1996, pp. 325-372. [2] R. Mootanah, Pressure sensors, In: Wiley Encyclopedia of Biomedical Engineering, M. Akay, Ed., Wiley-Interscience, New York, 2006, pp. 2836-2852. [3] J.G. Webster, Blood pressure measurement, In: Encyclopedia of Medical Devices and Instrumentation, Vol 1, J G Webster, Wiley, New York, 1990, pp.467–482. [4] A.C. Guyton and J. E. Hall, Textbook of Medical Physiology, 9th ed., W.B. Saunders Co., Philadelphia, 1996, pp.161-181. [5] G.M. Drzewiecki, Noninvasive assessment of arterial blood pressure and mechanics, In: The Biomedical Engineering Handbook, J D Bronzino ed., CRC, Boca Raton, FL, 1995, pp.1196–211. [6] R.A. Peura, Blood pressure and sound, In: Medical Instrumentation-Application and Design, 3rd edition, J.G. Webster, Ed., New York: John Wiley and Sons, 1998, pp.287-331. [7] L. Waite, Biofluid Mechanics in Cardiovascular Systems, McGraw-Hill, 2006, pp. 155-160. [8] A. Gedeon, P. Krill and B. Österlund, Pulmonary blood flow (cardiac output) and the effective lung volume determined from a short breath hold using the differential fick method, J. Clin. Monit. Comput. 17 (2002), 313-321. [9] W. Nichols and M. O’Rourke, MacDonalds’s Blood Flow in Arteries, In: Theoretical, Experimental and Clinical Principles, 3rd ed., E. Arnold, Ed., London, UK, 1990. [10] H. Feigenbaum, Echocardiography, 5th edition, Lea & Febiger, Philadelphia, PA, 1994. [11] K. Wolf, E. Höcherl, T. Derfuß and A. Krug, Laser-doppler-flowmetry and absorption-tissuespectrometry of the transposed groin flap - A comprehensive and independent analysis of microcirculation, Appl. Cardiopulm. Pathophysiol 11 (2003), [12] A. Nakano, Y. Sugii, M. Minamiyama and H. Niimi, Measurement of red cell velocity in microvessels using particle image velocity (PIV), Clin. Hemorheol. Microcirc. 29 (2003), 445-455. [1]


367

[13] J. Seki, Y. Satomura, Y. Ooi, T. Yanagida, and A. Seiyama, Velocity profiles in rat brain by optical coherence tomography, Clin. Hemorheol. Microcirc. 34 (2006), 233-239. [14] H.B. Kim, J. Hertzberg, C. Lanning and R. Shandas, Noninvasive measurement of steady and pulsating velocity profiles and shear rates in arteries using echo piv: in vitro validation studies, Ann. Biomed. Eng. 32 (2004), 1067–1076. [15] G.B. Kim and S.J. Lee, X-ray PIV measurements of blood flows without tracer particles, Exp. Fluids 41 (2006), 195–200. [16] C. Lawler, W.A. Suk, B.R. Pitt, C.M. St. Croix, and S.C. Watkins, Multimodal optical imaging, Am. J. Physiol. 285 (2003), L269-L280. [17] P.J. Matheson and R.N. Garrison, Intravital intestinal videomicroscopy: techniques and experiences, Microsurg. 25 (2005), 247-257. [18] V. Cerny, Z. Turek and R. Parizkova, Orthogonal polarization spectral imaging: A review, Physiol. Res. 56 (2007), 141-147. [19] J. Lindert, J. Werner, M. Redlin, H. Kuppe, H. Habazettl and AR. Pries, OPS imaging of human microcirculation: a short technical report, J. Vasc. Res. 39 (2002), 368-372. [20] C. Ince, Sidestream dark field imaging: an improved technique to observe sublingual microcirculation, Crit. Care 9 (2005),72-79. [21] D.M.J. Milstein, J.A.H. Lindeboom and C. Ince, Sidestream dark-field imaging and image analysis of oral microcirculation under clinical conditions, In: Anesthesia Pain Intensive Care and Emergency Medicine, A. Gullo, Ed., Springer Verlag, Berlin, 2005, pp. 79-88.. [22] D. Huang, E. A. Swanson, C. P. Lin, J. S. Schuman, W. G. Stinson, W. Chang, M. R. Hee, T. Flotte, K. Gregory, C. A. Puliafito, and J. G. Fujimoto, Optical coherence tomography, Science 254 (1991), 11781181. [23] Y. Satomura, J. Seki, Y. Ooi, T. Yanagida, and A. Seiyama, In vivo imaging of the rat cerebral microvessels with optical coherence tomography, Clin. Hemorheol. Microcirc. 31 (2004), 31-40. [24] S. Yamaguchi, T. Yamakawa, and H. Niimi, Red cell velocity and microvascular diameter measurement by a two fluorescent tracer method under epifluorescence microscopy; application to cerebral microvessels of cats, Int. J. Microcirc. Clin. Exp. 11 (1992), 404-416. [25] H. Niimi, Y. Komai, S. Yamaguchi, and J. Seki, Microembolic flow disturbances in the cerebral microvasculature with an arcadal network: a numerical simulation, Clin. Hemorheol. Microcirc. 34 (2006), 247-255. [26] A.D. Boas, D.H. Brooks, E.L. Miller, C.A. DiMarzio, M. Kilmer, R.J. Gaudette and Q. Zhang, Imaging the body with diffuse optical tomography, IEEE Signal Processing Magazine 18 (2001), 57-75. [27] A. P. Gibson, J. C. Hebden and S. R. Arridge, Recent advances in diffuse optical imaging, Phys. Med. Biol. 50 (2005), R1–R43. [28] J D. Briers, Laser Doppler, speckle and related techniques for blood perfusion mapping and imaging, Physiol. Meas. 22 (2001), R35–R66.


IV. Clinical Aspects of Hemorheology



371

Hyperviscosity: Clinical Disorders James P. ISBISTER1 Department of Haematology and Transfusion Medicine, Royal North Shore Hospital of Sydney, Faculty of Medicine, University of Sydney, St Leonards, New South Wales. Australia 2065

1. General Concepts of Hyperviscosity Maintenance of blood fluidity is central to the maintenance of life and health. Disorders of hemorheology and microcirculatory blood flow play an important role in the pathophysiology and clinical manifestations of a wide range of disease states [1-4]. Tissue perfusion is proportional to perfusion pressure, radius of the vessel and the viscosity of blood, including deformability of the individual cellular components. Microcirculatory failure may result from inadequate perfusion pressure, vessel narrowing or reduced blood fluidity. There is thus a wide range of causes of microvascular failure; this chapter will address those in which the principal pathophysiology focuses around the fluidity of blood. Most of these disorders are hematological and immunological in which there are quantitative or qualitative alterations in the cellular and or plasma components of the blood. The rheological abnormalities usually manifest their pathophysiological effects in the microcirculation, but they may interact with large vessel disease or disorders in which there is low perfusion pressure, and are thus of relevance to many subspecialties of medicine. The term hyperviscosity may be used in a general sense, implying reduction in blood fluidity, which may or may not be measured in the laboratory as increased whole blood or plasma viscosity. These measurable abnormalities are generally due to a quantitative increase in one or more of the blood components. However, significant hyperviscosity may be present in the blood due to interactions of cellular and plasma components, increased rigidity and/or adhesion of individual cells or particular physico-chemical characteristics of an abnormal plasma protein (e.g., cryoglobulins that may not be detectable by standard in vitro blood viscosity measurements). Whatever the cause, microcirculatory failure or insufficiency is the end result of blood hyperviscosity. 1.1. Blood and Plasma Volume, RBC Mass, Hematocrit Interrelations and Regulation The maintenance of adequate microcirculatory function requires circulation of blood of the appropriate composition by an effective cardiac output which, in turn, is dependent on an adequate intravascular blood volume and its return to the heart. A brief review of the physiology of blood volume, red cell mass and plasma volume regulation and the interrelationships with blood fluidity, hematocrit in particular, is appropriate [5]. Factors regulating the relationship between red cell mass and plasma volume in 1

Corresponding author: Department of Haematology and Transfusion Medicine, Royal North Shore Hospital of Sydney, Faculty of Medicine, University of Sydney, St Leonards, New South Wales. Australia; E mail: [email protected]

372

J.P. Isbister / Hyperviscosity: Clinical Disorders

determining the hematocrit remain unclear. Most scientific literature relating to total blood volume is found in relation to shock and resuscitation, the plasma volume literature relates to salt and water homeostasis and hypertension, and red cell mass literature relates to hematological disorders. The casual observer could be excused for thinking that red cell mass and plasma volume are independent variables, but this is clearly not the case. The hematocrit of blood in the macrocirculation is higher than that calculated from measurements of the red cell mass and plasma volume (i.e., body hematocrit). In a normal stable state the ratio of body hematocrit to venous hematocrit is approximately 0.9, with this ratio reflecting variations in red cell mass distribution within the vascular space [6]. The hematocrit in the microcirculation is lower than in the macrocirculation. The splanchnic circulation has a central role as the cardiovascular volume "buffer" to maintain a functional relationship between the total circulating blood volume and the size of the vascular compartment [7-9]. When blood volume is centralized as a result of venoconstriction, the pattern of neurohumoral responses resembles those occurring with hypertransfusion. Unless the venoconstriction is relieved, re-establishing an appropriate relationship between the size of the vascular compartment and blood volume can only be achieved by reducing plasma volume. This plasma volume contraction is achieved by salt and water shift into the interstitial space and the lymphatic system and by diuresis. This transcapillary efflux, in which atrial natriuretic peptide and endothelin have a role, occurs in the systemic circulation with protection of the pulmonary system, to avoid excess lung water [10]. The spleen has a role in plasma volume reduction by rapidly moving plasma into the lymphatic system which has its own capacitance [11, 12]. Mechanisms by which venous compliance may be reduced via increased sympathetic nervous system activity include: hypoxia, exercise, cold exposure and mental/physical stress [13, 14]. Sympathetic activation or the infusion of alpha receptor agonists, such as noradrenaline, lead to a reduction in the total circulating blood volume as a result of plasma volume contraction, hemoconcentration and an increase in blood viscosity. Sympathetic antagonists and other vasodilators result in redistribution of blood volume, expansion of plasma volume by interstitial fluid influx, thereby resulting in hemodilution. Sudden requirements for increases or decreases in total intravascular blood volume can only be effectively achieved by alterations in plasma volume as the red cell mass cannot be acutely altered. This is in contrast to when chronic stresses are placed on the system: the red cell mass component of the intravascular blood volume can be appropriately regulated by increased or decreased erythropoiesis under the influence of erythropoietin. It is thus evident that acute maintenance of an appropriate relationship between the volume of the intravascular compartment (“the vessel”) and the absolute intravascular blood volume (“the fluid”) results in acute changes in hematocrit as plasma volume adapts. In some circumstances, these changes in hematocrit due to hemoconcentration or hemodilution may be appropriate to the stimulus, whereas in other circumstances, the changes in hematocrit are "unavoidable" by-products to permit a "higher priority" acute volume adaptation to occur. The ultimate determinants of total blood volume and hematocrit remain unclear. In the short-term, the maintenance of adequate cardiac filling and output is obvious. However, in the overall picture, total blood volume is probably predominantly determined by the requirement for a reserve to respond to various stresses. In several physiological situations, there are substantial changes in total blood volume and, in most cases, hematocrit.


373

1.2. The optimal hematocrit The reference range for hematocrit is relatively wide and there is difficulty in defining a "normal" population since fitness status, smoking habits, alcohol intake, body mass index and stress may all influence hematocrit. Within the reference range for hematocrit a person’s red cell mass correlates with aerobic capacity rather than the circulating oxygen carrying capacity. 1.3. Sensors There is much debate and controversy regarding the nature of the central mechanisms operative in achieving optimal tissue oxygen delivery. There are many mechanisms that interact with varying priorities depending on the specific stressor(s) involved. The neuroendocrine system, via the autonomic nervous system, renin/angiotensin system, vasopressin, atrial natriuretic factor and endothelin have important roles in determining vascular tone, venous compliance and capillary permeability. Much controversy surrounds to question of the presence or absence of viscoreceptors. As the kidney receives a blood supply in excess of its oxygen requirements, oxygen extraction from hemoglobin as reflected in the arteriovenous oxygen difference is small compared with most other organs. The kidney seems to be an appropriate organ in which to locate early sensor mechanisms for detection of tissue hypoxia. The kidney, under the higher influence of the autonomic nervous system, is at the hub of sensor and effector activity in response to acute and chronic adaptive regulation of oxygen transport. However, the regulation of the relationship between the red cell mass and plasma volume has still not been explained. A viscoreceptor has been postulated for many years, and logic would support the concept that there should be a sensor for registering blood fluidity. In this respect there are many interacting determinants of blood fluidity at various shear rates, including; erythrocytes, platelets, leukocytes and plasma proteins and interactions between these blood components (see chapter II.4.a). In relation to hematocrit, it is the author's view that there probably is a viscoreceptor, perhaps in the renal circulation: this sensor would function maximally in the high blood viscosity range, but it is likely that such a receptor would be relatively insensitive within the physiological range of blood viscosity. The concept of a viscoreceptor is supported by the clinical observation that patients presenting with severe primary or secondary polycythemia and a marked increase in red cell mass have associated intravascular hypervolemia. When venesecting such patients the hematocrit does not immediately fall, and it is not until the patient has been venesected several times before the hematocrit falls to the extent expected on the basis of the volume of blood removed. In this "latent" phase, the clinical features of intravascular hypervolemia resolve; subsequently, the hematocrit falls in proportion to the degree of venesection. It is likely that in the resting state, with adequate oxygen availability, volume receptors have priority in circulatory control. Within the normal hematocrit range, it is not possible to identify a correlation between a person's venous hematocrit level and any measurable parameter except the body mass index. Further evidence for a viscoreceptor is supported by the correlation of erythropoietin responses to plasma viscosity. Erythropoietin formation in response to anemia in patients with plasma hyperviscosity (e.g., macroglobulinemia) has been shown to be inversely related to plasma viscosity. There a linear relationship existing

374


between circulating red cell mass and hemoglobin concentration in subjects with a low or normal red cell mass, and in this range blood volume has priority and is closely controlled. However, this does not appear to be the case as the hemoglobin rises above normal or plasma viscosity is elevated. It would thus appear that, within the reference range for hematocrit and in the context of a normal plasma viscosity, an individual's set point for hematocrit relates to body habitus, aerobic fitness and altitude, and it is only when the hematocrit is abnormally high or low that a correlation with erythropoietin or viscosity occurs. 1.4. The Spleen in Blood Volume Regulation The spleen does not contain a large volume of red cells and cannot be seen as an organ of clinically significant autotransfusion during hemorrhage. The sinusoidal nature of the spleen allows plasma skimming and hemoconcentration of red cells, but the exact role of this process in splenic function has been debated. There is evidence that the spleen may be a hemoconcentrating organ that is capable of rapidly reducing plasma volume to increase oxygen carrying capacity: it can rapidly move skimmed plasma into the lymphatics and thus acutely reduce plasma volume. The spleen is highly sensitive to the action of atrial natriuretic factor and thus probably plays an important role in hemoconcentration. On the other hand, hypervolemia causes an increase in intrasplenic filtration of cell-free fluid out of the circulation. The spleen thus appears to have role in blood volume regulation via its effects on plasma volume rather than red cell mass [15]. Questions remain about the role of the spleen in acute stresses on oxygen transport. The anatomical characteristics of the spleen raise some tantalising questions. The spleen is sometimes referred to as the "lymph node of the blood" as it has no afferent lymphatics but has rich efferent lymphatics. It is also intriguing to consider why the spleen is in the portal circulation. That is, since the spleen has a role as an organ of stress in relation to blood volume, it would seem logical to place the splenic vein directly into the splanchnic circulation which is the major venous capacitance reservoir of the circulation. Any changes in splanchnic pressures would be rapidly reflected in splenic circulation and plasma volume could be acutely altered. The spleen receives up to 5% of the total cardiac output, suggesting that its role is not insignificant; the fact that in the resting state splenic venous return is 75% of splenic artery inflow would support this concept. 1.5. Adaptations of blood volume and hematocrit 1.5.1. Aerobic Versus Anaerobic Stress Subjects involved in anaerobic sports such as weightlifting and sprinting have blood volumes within the normal range, but "highish" hematocrits that correlate with body mass index [16]. There is an advantage to these higher hematocrits since there is a "pre-existing" high content of oxygen circulating in the blood. In contrast, people who are chronically adapted to intense aerobic activity have an expanded total blood volume due to an increase in red cell mass and plasma volume. The plasma volume increase is greater than that of the red cell mass, resulting in hemodilution. These adaptations provide a greater blood volume and red cell mass reserve for the aerobic athlete when exercising [17, 18]. As plasma volume contracts with exercise and as the blood hemoconcentrates, the hematocrit will move into the normal range but the


375

individual will not become polycythaemic, thus avoiding the potential risk of hyperviscosity as has been seen with blood doping. 1.5.2. Hypoxia The most classical form of stress polycythaemia in which plasma volume contracts is seen in acute hypoxic stress of altitude [19]. The initial response to acute hypoxia is an increase in sympathoadrenal activity manifest as increased cardiac output, venoconstriction, centralization of the blood volume and subsequent hemoconcentration. The net effect of plasma volume contraction is to increase oxygen carrying capacity of the blood, albeit by reducing the total blood volume. This is the short-term response resulting in the circulation being in a potentially precarious position to respond to further stresses or volume loss. The longer-term response to hypoxic stress is an increase in red cell mass and return of the blood volume to normal. 1.5.3. Pregnancy During pregnancy there is an increase in total blood volume but the mechanisms of this remain sub judice. It is postulated that low maternal aldosterone levels and reduced levels of vasodilator substances, such as prostacyclin and kallikrein, may have a causal role [20]. It has also been demonstrated that there is an alteration in the setting of volume receptors and the sensitivity to atrial natriuretic factor which allow expansion of the blood volume. If these changes do not occur placental development may be impeded, the hemodilution of pregnancy does not occur, and foetal growth retardation may be a consequence [21-25]. It has been hypothesised that the development of the placenta may be a central factor in further initiating the changes in maternal blood volume and hemodilution [26]. Similar changes of blood volume and hematocrit occur in hypersplenism and bone marrow expansion; like the placenta, these organs are arteriovenous shunts with an interposed sinusoidal stroma which permits the skimming off of plasma-rich blood. 1.5.4. Ageing In the elderly there are significant alterations in blood volume regulation mechanisms due to alterations in the autonomic nervous system and thirst mechanisms [27]. Total blood volume decreases with age and the ability of the adrenergic system of older subjects to respond and adapt to environmental challenges is blunted [28].

2. The Acute Hematological Stress Response When an organism is exposed to the stresses outlined above, the changes described are sometimes referred to as the hematological stress response. This term has not been well defined as there are probably a variety of hematological stress responses depending on the stimulus. This author prefers to reserve the term hematological stress response for the acute changes described above in relation to adrenergic activation. The more delayed acute phase response follows release of cytokines and is a manifestation of post-insult host defense and healing responses. It is important that these two stress responses are delineated as they are probably activated relatively independently, but sequentially in some cases, by different stimuli and different temporal sequences.

376


The acute phase response which may go on to a more chronic phase (i.e., anemia of chronic disease) is related to the inflammatory and healing response [29]. The acute hematological stress response relates to the acute adrenergically mediated "fight-flight" reaction [30]. During the acute hematological stress response, hemoconcentration and priming of the haemostatic system (i.e., platelets and coagulation factors) occur in response to a real or potential insult [31, 32]. The fluidity of blood is pre-emptively reduced in anticipation of potential blood loss. Following injury the acute phase response is associated with hemodilution, ensuring the maintenance of blood fluidity and an increase in microcirculatory flow, presumably to ensure delivery of the inflammatory and healing response. The early phase of this has been referred to as the anemia of injury and with the passage of time the anemia of chronic disease [33].

3. Clinical Presentation of Hyperviscosity Whatever the underlying pathophysiology, the net effect of hyperviscocity is the reduction or cessation of tissue perfusion and ischemia or infarction, with resultant impacts on end organ function. The clinical presentation (i.e., history and end organ impacts) may be strong pointers to the underlying pathology. The oxygen transport system, being a complex multi-linkage chain starting with the inspiration of oxygen and ultimate delivery to mitochondria in the tissues, is only as strong as it weakest link. On this basis the hyperviscosity syndromes may present in several different ways as described below. 3.1. The "Classical" Hyperviscosity Syndrome This syndrome is typically seen in association with conditions such as Waldenstrom's macroglobulinemia and polycythemia rubra vera where there is a combination of hyperviscosity and hypervolemia. The following clinical features may be seen in this syndrome: visual or auditory disturbances, headache, neurological dysfunction and hypervolemia. There may also be clinical findings suggestive of congestive cardiac failure. The chest X-ray may show pulmonary congestion with a normal cardiac shadow, jugular, venous pressure may be elevated, the patient looks and feels congested in the head, may have haemostatic failure such as epistaxis, and there may be bruising and purpura due to a combination of hypervolemia and platelet dysfunction. Examination of the retinal fundus reveals venous engorgement with "cattle trucking" on the veins, hemorrhages and, in severe cases, papilloedema. In this setting the patient has a precarious microcirculation, and anything which may aggravate the hyperviscosity such as dehydration, diuretics, hypotension and blood transfusion should be avoided. 3.2. Primary Microcirculatory Failure Syndromes In some disorders there is primary microcirculatory obstruction with end organ failure in the presence of normal large vessels. This may manifest as digital ischemia with normal peripheral pulses or multi-system organ failure in the presence of normal blood pressure and macrocirculation. Such a presentation is typically seen in vasculitis,


377

disseminated intravascular coagulation, heparin induced thrombosis thrombocytopenia syndrome and thrombotic thrombocytopenic purpura. 3.3. Hypertension Hypertension, especially when the diastolic pressure is disproportionately elevated, may be indicative of hyperviscosity. As flow depends on the viscosity, radius and the perfusion pressure, flow will only be effectively maintained in the presence of hyperviscosity if the perfusion pressure rises. A corollary to this is reduction of perfusion pressure in the presence of hyperviscosity may precipitate a perfusion crisis, a fact which should be carefully considered before reducing the blood pressure in any patient with primary vascular disease. 3.4. Combination of Large Vessel and Small Vessel Hypoperfusion or Occlusion Similar to anemia, hyperviscosity may present with what appears, at first sight, to be solely large vessel disease such as coronary artery disease, cerebrovascular disease or peripheral vascular disease. Atheroma is common in the elderly and they are thus at risk from superimposed hyperviscosity. Large vessel occlusion and hyperviscosity are a bad combination, and in these circumstances lesser degrees of hyperviscosity that may not otherwise be clinically manifest are likely to be symptomatic. The normal hemodilution mechanisms of the microcirculation are inactivated beyond a large vessel block. This lack of hemodilution in the presence of a low perfusion pressure results in the inevitable “sludging” of blood (i.e., RBC aggregate formation) and loss of fluidity in the microcirculation. Primary hyperviscosity disorders of the blood will compound this pathophysiology: a presentation such as this, especially when the heart or brain is affected, constitutes a medical emergency requiring careful analysis of the pathophysiology involved. 3.5. The Unexpected Laboratory Finding Full blood counts and biochemical profiles are most frequent blood tests performed and incidental findings are common. Such findings can include polycythemia leukocytosis thrombocytosis, elevated erythrocyte sedimentation rate, elevated total protein, red cell abnormalities (e.g., spherocytosis, sickle cells, malaria), cryoglobulins, cold agglutinins and cryofibrinogen

4. Laboratory Features of Hyperviscosity There are few other disorders where in vitro or in vivo laboratory studies could be potentially more complex, yet have so much valuable information available from two of the cheapest and most frequently performed tests (i.e., the full blood count and the biochemical profile). Although specific measurements of blood fluidity may be helpful in the diagnosis and management of hyperviscosity and microcirculatory disease, most information can be gained from simple investigations and the clinical effects inferred when interpreted in the light of the clinical findings

378


4.1. Full Blood Examination Simple examination of the peripheral blood will detect most cellular causes of hyperviscosity and at the same time indicate the underlying diagnosis (e.g., leukemias, sickle cell disease, microangiopathic red cell changes). Elevation in plasma proteins and cold agglutinins will be reflected in the erythrocyte sedimentation rate or background staining on the blood film. Ironically, the ESR may be misleading in extreme plasma hyperviscosity or in the presence of a cryoglobulin where the plasma may be so viscous as to impede RBC sedimentation. 4.2. Biochemical Analyses A multiparameter biochemical screen will usually raise the suspicion of the presence of an abnormal monoclonal protein if it is present in large enough amounts to be responsible for a hyperviscosity syndrome. 4.3. Immunological and Other Protein Investigations Serum electropheresis, cryoglobulins, cold agglutinins and examination for cryofibrinogen will usually exclude most causes of plasma hyperviscosity. Antinuclear antibodies, rheumatoid factor, anti-neutrophil cytoplasmic antibody and complement levels may all help in identifying autoimmune diseases which may be complicated by small vessel vasculitis. 4.4. Test of Blood Fluidity Methods to measure blood rheology are discussed in detail elsewhere (chapters II.1, II.3.a, II 6). In general, the tests of blood fluidity have not established a role in most routine clinical laboratories. Plasma viscosity relative to water is a relatively simple measurement, but its interpretation in terms of hyperviscosity is difficult as it does not take into account any of the cellular components of blood. Whole blood viscosity is a truer measurement of what is likely to be occurring in vivo. High and low shear rates are used in the viscometer to simulate the flow in larger arterial vessels versus the flow characteristics of the microcirculation and venous systems. Unfortunately, the normal range is wide and highly hematocrit dependent. As a general principle the basic hematological, biochemical and immunological investigations in the clinical context provide the best information for decision making in the diagnosis and management of patients with hyperviscosity syndromes. There are an increasing range of tests available that may assist in assessing microvascular end-organ dysfunction. Microvascular disease classically causes diffuse fluctuation in end-organ function, and sophisticated methods are now available for the study of most organs of the body. The following investigations may be of assistance in specific circumstances: cytoscanning capillary blood flow, retinal fundus examination, EEG, renal blood flow scan, urine examination for evidence of glomerular disease, pulmonary, mesenteric and renal angiography and, more recently, cerebral blood flow studies.


379

4.5. Tests for an Underlying Cause Detailed tests for underlying causes are beyond the scope of this chapter and the reader is referred to hematology texts and reviews [34-36].

5. General Principles of Management of Impaired Blood Fluidity 5.1. Avoidance of aggravating factors x x x x x x x x x x

Dehydration Diuretics Hypotension Radiological contrast media Infection Blood transfusion Cold exposure and other vasoconstrictive influences Hypoxia Stasis Hypoglycemia

5.2. Correct other weak links in oxygen transport x x x x

Hypoxemia Cardiac failure Salt, water and intravascular volume depletion Factors increasing hemoglobin affinity: alkalosis, carbon monoxide, hypothermia

5.3. Initial therapy aimed at improving blood fluidity Although long-term management of a hyperviscosity syndrome requires definitive therapy directed at the basic cause, and most definitive therapy requires days to weeks to become effective. However, immediate temporising therapy may reduce mortality and morbidity. The initial therapy will depend on the specific alteration in blood to reduced blood fluidity. A summary of causes and general approaches to initial therapy is presented in Table 1.

6. Major Disturbances of Blood Rheology and Microcirculatory Flow 6.1. Polycythemia The term polycythemia has been given several different meanings. Strictly, the term should be used in its broadest sense to mean excess red cells per unit volume of blood, irrespective of the underlying cause. Some clinicians have restricted the term polycythemia to conditions where there is a demonstrable increase in the red cell mass

380


and used the term relative polycythemia for all disorders where plasma volume contraction is responsible [37]. As a result of this approach there has been a tendency to neglect the hyperviscosity implications of polycythemia secondary to plasma volume contraction. Many hematologists lose interest in the further investigation of polycythemia as soon as contraction of plasma volume is identified as the cause. This is a restrictive approach and may be detrimental to the patient by ignoring a significant blood fluidity problem. The investigation and management of polycythemia is far beyond the scope of this chapter, but a general approach to the nature, diagnosis and management of this common clinical problem is addressed. Table 1. Summary of causes and general approaches to initial therapy Pathophysiology Erythrocytes Polycythemia Defects in red cell fluidity Leukocyte disorders Leukostasis syndrome Neutrophil aggregation syndromes Platelet disorders Thrombocytosis

Initial therapy Phlebotomy and hemodilution Hemodilution

Leukapheresis Corticosteroids

Antiplatelet therapy

Thrombocythemia

Antiplatelet therapy + / - thrombopheresis

Platelet aggregation syndromes

Antiplatelet therapy

Plasma disorders Paraproteinemia IgM>IgA>IgG

Plasma exchange

Cryoglobulinemia

Plasma exchange

Hyperfibrinogenemia

Defibrination therapy + / -

Cryofibrinogenemia

Plasma exchange

Cold Agglutinins

Warming + / - plasma exchange

Vessel wall Vasculitis Microangiopathies

Corticosteroids +/ - plasma exchange Depends on the underlying pathophysiology and remains controversial in many cases

Arterial obstruction

Correct compounding hemorheological factors as and address vessel blockage

Venous thrombosis

Anticoagulants, and correction of any hemorheological factors as

In the initial assessment of polycythemia there are certain key questions which need to be addressed from the outset. x At what level of hemoglobin should a patient be investigated? There is continuing debate in relation to the reference range for the hemoglobin level and there is increasing agreement that hemoglobin above 170 g/l in a male and 160 g/l in a female requires explanation and


x

x

x

381

should be considered as a possible risk factor associated with hyperviscosity problems. Is the patient acutely hyperviscous and at risk of a vascular event? If there is already evidence of vascular compromise or there is a significant risk present, initial therapy may take precedence over diagnosis. Is the underlying cause obvious from the clinical history? If the cause is obvious, investigations can be more specifically directed or no further testing may be necessary, particularly when the polycythemia is due to acute plasma volume contraction Is it necessary to perform red cell mass and plasma volume studies? These parameters will usually need to be measured to clearly establish whether the polycythemia is due to a true increase in the red cell mass or a contraction in the plasma volume. Until this has been established there is little point in proceeding with extensive investigations unless there are clinical and/or laboratory features indicating the cause. There are well standardized and accurate radioisotopic methods available for the measurement of these parameters and both measurements should be performed in patients with unexplained polycythemia. The main problems relate to reference ranges, which are unacceptably wide in order to allow for variations in body dimensions. Red cell mass and plasma volume are not independent variables; a patient may have significant polycythemia due to a combination of plasma volume contraction and increased red cell mass yet each parameter may lie within the reference range. For these reasons, the diagnostic sensitivity in any individual patient may be limited and, without previous laboratory measurements, it is not possible for an individual patient's "set point" for hematocrit to be considered when interpreting red cell mass and plasma volume measurements.

6.2. Polycythemia Due to Increased Red Cell Mass If the red cell mass and plasma volume studies indicate an increased red cell mass further investigation is directed towards finding the cause. As a general rule, the cause will be apparent from the outset in that the clinical findings or initial laboratory findings are diagnostic. Under these circumstances any further tests are usually confirmatory or of a baseline nature. On the other hand, the cause may not be obvious in which case underlying causes need to be excluded in a systematic fashion. In most circumstances the cause becomes apparent, but occasionally the clinician is left with the rather unsatisfactory diagnosis of idiopathic erythrocytosis. However, classical polycythemia rubra vera, severe hypoxic lung disease, intracardiac shunt and smoker's polycythemia are relatively easily diagnosed in the majority of patients. More recently, better markers of a clonal disease for polycythemia rubra vera have been identified, and there are new insights into cases of previously categorized idiopathic erythrocytosis [35, 38-44]. There are numerous investigations, some complex and time consuming; available for the investigation of polycythemia, but in most circumstances it is only necessary to perform specific tests to confirm a provisional diagnosis made from the initial clinical and laboratory assessment. The initial management of polycythemia secondary to increased red cell mass should focus on assessing the clinical significance and, if necessary, the initial control

382


of hyperviscosity. Polycythemia rubra vera has always been the classical disease for which venesection has been the mainstay of therapy. As the red cell mass may be markedly increased, large volumes of blood may need to be removed before the hematocrit is reduced to an acceptable level. The importance of adequate intravascular volume replacement to maintain normovolemia, especially in elderly patients, is crucial. Any sudden reduction in perfusion pressure may have additive detrimental effects on microcirculatory flow in the presence of hyperviscosity or large vessel stenosis. Thrombocytosis commonly occurs following venesection. This is not a reactive thrombocytosis as the disease process is autonomous, and it is more likely that the increase in circulating platelets is an indication of platelet mobilization from the microcirculation as sludging is relieved by the hemodilution. Antiplatelet therapy is desirable if the platelet count is elevated. 6.2.1. Polycythemia Rubra Vera Polycythemia rubra vera is a myeloproliferative disease in which there is autonomously increased activity of erythropoiesis with various degrees of excess granulopoietic and megakaryocytic proliferation. It has a variable natural history, gradually progressing to a "spent" phase with myelofibrosis and myeloid metaplasia. The disease is clonal in nature and has the potential for acute leukemic transformation. Polycythemia rubra vera typically occurs in patients over the age of 40 years; clinical features include hyperviscosity and hypervolemia, haemostatic impairment, pruritus, thrombotic events and gout. Patients are typically plethoric with signs of hyperviscosity, hypertension, splenomegaly and hepatomegaly. The reader is referred to more comprehensive reviews [45-47]. 6.2.2. Polycythemia Secondary to Hypoxemia Any condition causing intermittent or chronic reduction in arterial oxygen saturation will stimulate the erythropoietin system, resulting in a reactive erythroid hyperplasia in the bone marrow. It is not essential to demonstrate hypoxemia on a random blood sample to suspect this type of secondary polycythemia. The maximum hypoxic stimulus may occur at other times, especially at night and sleep studies may be necessary. The possible causes of hypoxemic polycythemia include; low barometric pressure (i.e., altitude dwellers), cyanotic heart disease and other right to left shunts, primary pulmonary disease and hypoventilation syndromes. 6.2.3. Polycythemia Secondary to Tissue Hypoxia There are several conditions where there is defective release of oxygen to the tissues and in which erythropoietin production is stimulated. This can occur in chronic carbon monoxide poisoning either from smoking or environmental exposure. Not only does carbon monoxide make some hemoglobin unavailable for oxygen transport, but the remaining functional hemoglobin has increased oxygen affinity. Carboxyhemoglobin measurements should be made at the times of peak exposure. An uncommon cause of secondary polycythemia is seen in families with congenitally abnormal high affinity hemoglobins where the hemoglobin dissociation curve can be markedly left-shifted. These are sometimes called Llama hemoglobins as their oxygen dissociation curve is similar to that of these high living altitude mammals in the Andes [48].


383

6.2.4. Polycythemia Secondary to Inappropriate Erythropoietin Production On rare occasions polycythemia may be the presentation of an underlying erythropoietin producing tumour or relate to renal pathology. Renal ischemia, cysts and hydronephrosis as well as hypernephroma may be responsible. Other non-renal erythropoietin producing tumours include hepatomas, uterine myomas and cerebellar hemangiomas. 6.3. Polycythemia due to Plasma Volume Contraction: Relative Polycythemia It is becoming increasing recognized that in the majority of patients with mild polycythemia the underlying mechanism is contraction of the plasma volume, either due to a salt and water deficit, capillary leak of protein rich fluid or to a contraction of the venous capacitance volume with associated compensatory reduction in plasma volume to maintain normal central cardiac filling pressures and volumes [5]. Plasma volume contraction secondary to water and/or salt depletion is usually due to excessive losses with inadequate intake, but disorders of thirst control may be responsible. Therapeutic dehydration is commonly used in order to treat or prevent edema formation in vital organs, especially the central nervous system and lung. Such therapy will severely contract the plasma volume and hemoconcentrate the blood. A fine line may be walked with such therapy and the whole blood hyperviscosity induced by such therapy cannot be ignored. In contrast to the elderly, young patients can tolerate quite extreme hyperviscosity with few detrimental effects. Constant maintenance of adequate peripheral perfusion pressure is essential in the presence of severe hemoconcentration. Defects in the body's thirst mechanism classically occur in hypothalamic disease, and its rarity would normally allow only passing mention. However, there is increasing interest in disorders of osmoregulation and the thirst mechanism in the elderly that may cause severe dehydration, plasma volume contraction and polycythemia [49]. The syndrome particularly occurs in elderly patients, who may or may not have recognizable neurological disease, resulting in chronic hypodypsia and dehydration. These patients have intact osmoreceptors and appropriate vasopressin responses, but the message is not communicated to the thirst centre, so the patients will unknowingly dehydrate themselves despite ready availability of water. Acute and chronic diuretic therapy may cause both plasma volume contraction and polycythemia. The administration of diuretics to patients for the first time is particularly likely to result in sudden plasma volume reduction and associated hemodynamic consequences. Rapid ascent to high altitude may be associated with the development of a syndrome of headache, insomnia, nausea, irritability, oliguria and sometimes lifethreatening pulmonary edema and coma. This acute mountain sickness, although initiated by a low inspired oxygen tension, has a complex pathophysiology of which an important component is polycythemia with hyperviscosity and microcirculatory failure. As discussed above, the rise in hematocrit on ascending to high altitude is initially due to acute plasma volume contraction and hemoconcentration. Acute hypoxic states from other causes, such as acute respiratory failure or carbon monoxide poisoning, produce varying degrees of plasma volume contraction. Any condition in which there is sudden release of large amounts of vasoactive amines will result in a sudden leakage of plasma into the interstitial space and acute hemoconcentration. The combination of acute hypovolemia in conjunction with

384


hyerviscosity may lead to severe microcirculatory failure. Cases of acute capillary leak, as may occur in allergic reactions, anaphylaxis, cold urticaria, monoclonal immunoglobulins and envenomation, may result in acute plasma volume contraction, hypovolaemia and polycythemia [50]. Some endocrine disorders may be complicated by plasma volume contraction and secondary polycythemia. Acute diabetic hyperosmolar coma presents with severe hyperglycemia, dehydration, hyperosmolality and hemoconcentration, with minimal evidence of ketoacidosis. This syndrome occurs in elderly, maturity onset non insulindependent diabetics. The profound osmotic diuresis precipitated by severe hyperglycaemia results in plasma volume depletion. The patients commonly present with cerebrovascular occlusive episodes. With early diagnosis of this syndrome followed by swift plasma volume expansion and more gradual correction of hyperosmolality the patients commonly improve dramatically and are left with surprisingly little residual neurological deficit. 6.3.1. Stress Polycythemia It is not strictly correct to give this syndrome the status of a disease as many different disorders have been included under a multitude of different synonyms. This syndrome was originally described by Gaisbock in a group of patients with polycythemia and hypertension (i.e., polycythemia hypertonica with splenomegaly). Some of these patients may have had polycythemia rubra vera. With the introduction of red cell mass and plasma volume measurements it became clear that there were patients with relative polycythemia due to plasma volume contraction. Various terms were introduced to cover this entity including chronic relative erythrocytosis, benign erthrocytosis, spurious polycythemia, pseudopolycythemia and stress polycythemia. Note that in addition to the known association of psychological stress with hypertension, other sources of “stress” have been added, including obesity, vascular occlusive disease, excess alcohol intake and cigarette smoking. With all these common associations it is natural that there will be difficulties in elucidating causal relationships. Several theories have been proposed relating polycythemia to stress, with the most tenable one at present being the increased venous tone theory: catecholamines are known to cause venoconstriction and centralise the blood volumes, and so in order to maintain normal cardiac filling pressures and volumes the plasma volume must contract [37]. 6.3.2. Smoker’s Polycythemia The association between smoking and polycythemia has been recognized for many years, but it is only in recent years that there has been a clearer understanding of the pathophysiology. Until recently it was assumed that pulmonary disease with arterial hypoxia was necessary for polycythemia to develop. Red cell mass and plasma volume studies have revealed several different patterns. Some patients have clearly elevated total red cell mass, but the majority have a significant plasma volume contraction with minimal or no increase in red cell mass [51]. In some patients both parameters are within the normal range, but the plasma volume is at the lower limit and the red cell mass is at the upper limit. When there is plasma volume contraction present, the effects of smoking on hematocrit are relatively quickly reversible (personal observations); cessation of smoking may lead to the hemoglobin falling in a matter of 24 hours. It is a common observation in the coronary care unit that many patients are mildly polycythemic on


385

admission with the hematocrit returning to normal in the days following admission. This is more common in smokers, but recent diuretic therapy, dehydration or stress may also be incriminated. 6.4. Disorders of Red Cell Deformability There are many disorders in which abnormalities of red shape and deformability may be responsible for impairment of microcirculatory function and oxygen delivery to the tissues [52]. Detailed discussion of these disorders is beyond the scope of this chapter and the reader is referred to other chapters and relevant reviews. Disorders of red cell deformability include: sickle cell disease [53], hereditary spherocytosis [54], microangiopathies [55], thrombotic thrombocytopenic purpura [55-57], hemolytic uremic syndrome [58], malaria [59-62], sepsis [63, 64], trauma [65], shock [66, 67] and red cell storage [68-70]. 6.5. Leucostasis Syndromes The hemorheological significance of extreme leukocytosis is variable. Leucocytes are less deformable than red cells, and high white cell counts may result in blood hyperviscosity affecting the microcirculation. The clinical importance of this probably relates to cell size and type, with cells of the granulocytic series more likely to be associated with clinical evidence of hyperviscosity and microcirculatory failure. This occurs in its most dramatic form as the fulminant leucostasis syndrome in which there appears to be a "chain" reaction set up: the high count of primitive granulocytic cells leads to obstruction of the microcirculation, breakdown of lysosomes and local activation of the coagulation system. This results in sudden peripheral circulatory obstruction, particularly manifest in the cerebral and pulmonary circulations and seen histologically as leucocyte-fibrin thrombi. This fulminant syndrome is usually irreversible and rapidly fatal, and may be seen as the agonal event in untreated acute leukemia, especially those of myeloid origin, but may occasionally occur in extreme reactive leukocytosis [71]. There is a more common and less severe form of leucostasis in which the clinical presentation is similar to the hyperviscosity associated with polycythemia and plasma protein abnormalities [72]; this situation may be relieved by intensive leucapheresis. These patients may develop, or be precipitated into, the fulminant leukostasis syndrome by blood transfusion, dehydration or sepsis. 6.6. Thrombocytosis, Thrombocythemia, Thrombocytopenia and Platelet Hyperactivity Given the central role of platelets in host defense and haemostatic mechanisms, it is not surprising that they are important in the pathophysiology of a variety of diseases [73]. Excessive numbers or overactivity of platelets may threaten blood fluidity and thus microcirculatory flow and hence are an increasing target for therapy in many vascular disorders [74, 75]. The spectrum of disease states resulting in part or solely from platelet pathology range from the gradual evolution of atheromatous disease to acute disseminated platelet aggregation; more recently the role of platelet microparticles has received increasing attention [76, 77].

386


6.6.1. Essential Thrombocythemia In essential thrombocythemia there is autonomous overproduction of platelets, and the characteristic symptoms of microvascular obstruction may occur (e.g., peripheral ischemic lesions and occasionally gangrenous digits). Impairment of the circulation in other vascular beds such as mesenteric, splenic, hepatic or cerebral may occur [78]. Headache, parasthesia and transient cerebral ischemic attacks are common. The peripheral blood shows a thrombocytosis with giant platelets, some with abnormal intracellular morphology. The bone marrow typically shows a marked increase in megakaryocytes with abnormal dysplastic forms present. Therapy is aimed at reducing platelet aggregability with antiplatelet therapy and reducing the platelet numbers, with the long-term therapy of essential thrombocythemia usually involving marrow suppressive agents [79]. 6.6.2. Reactive Thrombocytosis Reactive thrombocytosis is seen a part of the acute phase reaction and is probably of minimal rheological significance unless there are other factors affecting blood flow [80]. Occasionally the thrombocytosis may be extreme (e.g., 1,000 x 109/l), especially in asplenic individuals and thus one must be concerned about the rheological significance [81]. Differentiation of reactive thrombocytosis from essential thrombocythemia is generally not a problem [82]. Antiplatelet therapy is the most appropriate therapy until the stimulus has resolved. Causes of marked reactive thrombocytosis, often termed "platelet millionaires", include: post-splenectomy, malignancy, bacterial sepsis, inflammatory diseases such as rheumatoid arthritis; in some situations thrombocytosis may be a marker of disease activity such as in rheumatoid arthritis. 6.6.3. Heparin Induced Thrombosis Thrombocytopenia Syndrome (HITS) This complication of heparin therapy warrants specific mention due to its clinical importance. In this complication an immune reaction occurs to heparin after 7-10 days of therapy, at which time the patient's platelets aggregate and thrombocytopenia develops [83, 84]. In contrast to other causes of drug induced thrombocytopenia there may be arterial, microvascular or venous thrombosis associated with the platelet aggregation. The best way to avoid this potentially devastating complication is to keep heparin therapy as brief as possible and always be alert to the diagnostic possibility in any patient who develops thromboctopenia or any arterial or venous thrombotic episode; heparin antibodies can usually be detected in the patient’s plasma [85]. 6.7. Coagulopathies Associated with Impaired Blood Fluidity Any inappropriate activation of the haemostatic system may threaten blood fluidity in the macro and/or microcirculation. Hyperactivity of platelets has been discussed above, but excessive activation of the coagulation cascade as occurs in venous thromboembolism and disseminated intravascular coagulation may have serious clinical consequences. The reader is referred to recent reviews of this important and common clinical disorders [86-93].


387

6.8. Plasma Hyperviscosity Hyperviscosity of the blood may result from several different plasma abnormalities [94]. In some, the presence of an abnormal protein or normal protein in excess is responsible, in others the particular characteristics of the protein leads to microcirculatory disease. In most patients plasma exchange is an effective mode of therapy, usually followed up with definitive therapy for the underlying disease [95-97]. 6.8.1. Monoclonal Immunoglobulins Monoclonal IgM, one of the features of Waldenstrom's macroglobulinemia (i.e., lymphoplasacytic lymphoma), may be responsible for the hyperviscosity [98]. Less commonly, a similar picture may be seen with IgA multiple myeloma or occasionally with IgG. Lymphoplasmacytic lymphoma comprises approximately 5% to 10% of the indolent disseminated lymphomas/leukemias, and occurs in older adults presenting with bone marrow involvement and a monoclonal serum IgM paraprotein. Patients presenting in this manner may have the classical features of Waldenstrom's macroglobulinemia in which the circulating IgM level may be sufficient to cause the hyperviscosity syndrome, hypervolemia and impairment of hemostasis. Initial therapy in classical Waldenstrom's macroglobulinemia is directed towards reducing the IgM level, which may require plasma exchange. Adequate hydration and the avoidance of blood transfusion is important if there is hypervolemia and hyperviscosity [99]. 6.8.2. Cryoglobulins The various forms of cryoglobulinemia may manifest clinically as microcirculatory impairment with or without vasculitis [100]. The cryoglobulin may consist entirely of monoclonal immunoglobulin produced by a malignant (e.g., myeloma) or benign (e.g., benign monoclonal gammopathy) immunoproliferative disease [101]. In other patients it may be a mixed cryoglobulinemia in which one of the components may be a monoclonal immunoglobulin. Whatever the type of cryoglobulin, rapid removal by plasma exchange is a matter of urgency for some patients. This therapy may present technical difficulties as the cryoglobulin tends to precipitate in the extracorporeal circuit and obstruct flow. It is essential that the patient be kept warm at all times, and that the plasma exchange procedure be carried out at a high ambient temperature with pre-warming of the cell separator and the infusion fluids. 6.8.3. Cold Agglutinins Many patients with cold agglutinins are found to have a monoclonal IgM immunoglobulin in their plasma [102]. If this is of high titre with a narrow thermal range, the clinical manifestations are those of microcirculatory obstruction in the acral areas in relation to cold exposure. However, if the autoantibody has a wide thermal range and reacts above 30 °C, hemolysis is usually the clinical problem. Some patients with cold autoagglutinins may have an underlying immunoproliferative disease which will direct therapy.

388


6.8.4. Hyperfibrinogenemia and Cryofibrinogenemia Fibrinogen, being an acute and chronic phase reactant protein, will be elevated in a range of reactive states. The development of mild anemia with the reaction usually offsets any hyperviscosity effects. However, there is increasing interest in the hematological stress syndrome and its relation to vascular disease. Any patient with vascular disease should have the basic hemorheological parameters measured and a specialist opinion sought if significant abnormalities are found [103]. Cryofibrinogenemia is uncommon and rarely causes major clinical problems [104, 105]. If present in large amounts, the clinician should suspect an underlying malignancy or autoimmune disease. Hyperfibrinogenemia usually settles when the stimulus resolves, but in chronic state drugs such as metformin, arabolin or stanazol may have a role in therapy.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]

G.L. Buchele, G.A. Ospina-Tascon and D. De Backer, How microcirculation data have changed my clinical practice, Curr. Opin. Crit. Care 13 (2007), 324-31. E.H. Serne, R.T. Jongh, E.C. Eringa, R.G. Ijzerman, M.P. deBoer and C.D. Stehouvver, Microvascular dysfunction: causative role in the association between hypertension, insulin resistance and the metabolic syndrome? Essays Biochem. 42 (2006) 163-76. V. Turchetti, L. Boschi, G. Donati, L. Trabalzini and S. Forconi, Impact of hemorheological and endothelial factors on microcirculation, Adv. Exp. Med. Biol. 578 (2006), 107-12. S.A. Evans, Haemorheology--science and medicine, J. R. Soc. Med. 88 (1995), 52P-53P. J.P. Isbister, Physiology and pathophysiology of blood volume regulation, Transfus. Sci. 18 (1997), 409-23. J.P. Isbister, The F-cell ratio: a clinically important parameter or just fine tuning? Infusionsther. Transfusionsmed. 22 (1995), 69-70. J.N. Cohn, Relationship of plasma volume changes to resistance and capacitance vessel effects of sympathomimetic amines and angiotensin in man, Clin. Sci. 30 (1966), 267-78. C. Risoe, W. Tan and O.A. Smiseth, Effect of carotid sinus baroreceptor reflex on hepatic and splenic vascular capacitance in vagotomized dogs, Am. J. Physiol. 266 (1994), H1528-H1533. C.F. Rothe, Reflex control of veins and vascular capacitance, Physiol. Rev. 63 (1983), 1281-1342. R.S. Zimmerman, A.J. Martinez, M. Maymind and R.W. Barbee, Effect of endothelin on plasma volume and albumin escape, Circ. Res. 70 (1992), 1027-1034. I.B. Stewart and D.C. McKenzie, The human spleen during physiological stress, Sports Med. 32 (2002), 361-369. M. Laub, K. Hvid-Jacobsen, P. Hovind and I.L. Kanstrup, Spleen emptying and venous hematocrit in humans during exercise, J. Appl. Physiol. 74 (1993), 1024-1026. J.A. Neubauer, Invited review: Physiological and pathophysiological responses to intermittent hypoxia, J. Appl. Physiol. 90 (2001), 1593-1599. S.M. Patterson, D.S. Krantz, J.S. Gottdiener, G. Hecht, S. Vargot and D.S. Goldstein, Prothrombotic effects of environmental stress: changes in platelet function, hematocrit, and total plasma protein, Psychosom. Med. 57 (1995), 592-599. M. Laub, K. Hvid-Jacobsen, P. Hovind and I.L. Kanstrup, Spleen emptying and venous hematocrit in humans during exercise, J. Appl. Physiol. 74 (1993), 1024-1026. R.D. Telford and R.B. Cunningham, Sex, sport, and body-size dependency of hematology in highly trained athletes, Med. Sci. Sports Exerc. 23 (1991), 788-794. E.T. Stevenson, K.P. Davy and D.R. Seals, Maximal aerobic capacity and total blood volume in highly trained middle-aged and older female endurance athletes, J. Appl. Physiol. 77 (1994), 1691-1696. L.M. Weight, M. Klein, T.D. Noakes and P. Jacobs, 'Sports anemia'--a real or apparent phenomenon in endurance-trained athletes? Int. J. Sports Med. 13 (1992), 344-347. T.D. Poulsen, T. Klausen, J.P. Richalet, I.L. Kanstrup, N. Fogh-Andersen and N.V. Olsen, Plasma volume in acute hypoxia: comparison of a carbon monoxide rebreathing method and dye dilution with Evans' blue, Eur. J. Appl. Physiol. Occup. Physiol. 77 (1998), 457-461.


389

[20] S. Kaufman, Control of intravascular volume during pregnancy, Clin. Exp. Pharmacol. Physiol. 22 (1995), 157-163. [21] S.P. Salas, P. Rosso, R. Espinoza, J.A. Robert, G. Valdez and E. Donoso, Maternal plasma volume expansion and hormonal changes in women with idiopathic fetal growth retardation, Obstet. Gynecol. 81 (1993), 1029-1033. [22] P. Rosso, E. Donoso, S. Braun, R. Espinoza, C. Fernandez and S.P. Salas, Maternal hemodynamic adjustments in idiopathic fetal growth retardation, Gynecol. Obstet. Invest. 35 (1993), 162-165. [23] J.J. Duvekot, E.C. Cheriex, F.A. Pieters, P.P. Menheere, H.J. Schouten and L.L. Peeters, Maternal volume homeostasis in early pregnancy in relation to fetal growth restriction, Obstet. Gynecol. 85 (1995), 361-367. [24] G.D. Michailidis, R.W. Morris, A. Mamopoulos, P. Papageorgiou and D.L. Economides, The influence of maternal hematocrit on placental development from the first to the second trimesters of pregnancy, Ultrasound Obstet. Gynecol. 20 (2002), 351-355. [25] M.L. Blankson, R.L. Goldenberg, G. Cutter and S.P. Cliver, The relationship between maternal hematocrit and pregnancy outcome: black-white differences, J. Natl. Med. Assoc. 85 (1993), 130-134. [26] V. Jonsson, J.E. Bock and J.B. Nielsen, Significance of plasma skimming and plasma volume expansion, J. Appl. Physiol. 72 (1992), 2047-2051. [27] T. Ito, A. Takamata, K. Yaegashi, T. Itoh, T. Yoshida, T. Kawabata, M. Kimura and T. Morimoto, Role of blood volume in the age-associated decline in peak oxygen uptake in humans, Jpn. J. Physiol. 51 ( 2001), 607-612. [28] K.P. Davy and D.R. Seals, Total blood volume in healthy young and older men, J. Appl. Physiol. 76 (1994), 2059-2062. [29] F. Ceciliani, A. Giordano and V. Spagnolo, The systemic reaction during inflammation: the acute-phase proteins, Protein Pept. Lett. 9 (2002), 211-223. [30] H.S. Bracha, Freeze, flight, fight, fright, faint: adaptationist perspectives on the acute stress response spectrum, CNS Spectr. 9 (2004), 679-85. [31] M.T. Allen and S.M. Patterson, Hemoconcentration and stress: a review of physiological mechanisms and relevance for cardiovascular disease risk, Biol. Psychol. 51 (1995), 1-27. [32] L. Zgraggen, J.E. Fischer, K. Mischler, D. Preckel, B.M. Kudielka and R. von Kanel, Relationship between hemoconcentration and blood coagulation responses to acute mental stress, Thromb. Res. 115 (2005), 175-183. [33] G. Weiss and L.T. Goodnough, Anemia of chronic disease, N. Engl. J. Med. 352 (2005), 1011-1023. [34] J.T. Marsden, Erythropoietin-- measurement and clinical applications, Ann. Clin. Biochem. 43 (2006), 97-104. [35] M.F. McMullin, D. Bareford, P. Campbell, A.R. Green, C. Harrison, B. Hunt, D. Oscier, M.I. Polkey, J.T. Reilly, E. Rosenthal, K. Ryan, T.C. Pearson and B. Wilkins, Guidelines for the diagnosis, investigation and management of polycythaemia/erythrocytosis, Br. J. Haematol. 130 (2005), 174-195. [36] R. Hoffman, E. Benz and S. Shattil, Haematology: Principles and Practice, 4th Edition, Elsevier Health Sciences, 2005. [37] J.P. Isbister, The contracted plasma volume syndromes (relative polycythaemias) and their haemorheological significance, Baillieres Clin. Haematol. 1 (1987), 665-693. [38] M.L. Percy, Genetically heterogeneous origins of idiopathic erythrocytosis, Hematology 12 (2007), 131-139. [39] L.M. Scott, W. Tong, R.L. Levine, M.A. Scott, P.A. Beer, M.R. Stratton, P.A. Futreal, W.N. Erber, M.F. McMullin, C.N. Harrison, A.J. Warren, D.G. Gilliland, H.F. Lodish and A.R. Green, JAK2 exon 12 mutations in polycythemia vera and idiopathic erythrocytosis, N. Engl. J. Med. 356 (2007), 459-468. [40] G. Finazzi, X.T. Gregg, T. Barbui and J.T. Prchal, Idiopathic erythrocytosis and other non-clonal polycythemias, Best Pract. Res. Clin. Haematol. 19 (2006), 471-482. [41] F.S. Lee, M.J. Percy and M.F. McMullin, Oxygen sensing: recent insights from idiopathic erythrocytosis, Cell Cycle 5 (2006), 941-945. [42] M. Usman, F. Bilwani, G.N. Kakepoto, S.N. Adil and R. Sajid, Polycythemia vera and idiopathic erythrocytosis: comparison of clinical and laboratory parameters, J. Pak. Med. Assoc. 54 (2004), 249251. [43] H.A. Blacklock and G.A. Royle, Idiopathic erythrocytosis--a declining entity, Br. J. Haematol. 115 (2001), 774-781. [44] T.C. Pearson and M. Messinezy, Idiopathic erythrocytosis, diagnosis and clinical management, Pathol. Biol. (Paris) 49 (2001), 170-177. [45] A. Tefferi, Polycythemia vera: a comprehensive review and clinical recommendations. Mayo Clin. Proc. 78 (2003), 174-94. [46] M.A. Elliott and A. Tefferi, Thrombosis and haemorrhage in polycythaemia vera and essential thrombocythaemia, Br. J. Haematol. 128 (2005), 275-90.

390


[47] J.J. Michiels, Z.N. Berneman, W. Schroyens and H.H. van Vliet, Pathophysiology and treatment of platelet-mediated microvascular disturbances, major thrombosis and bleeding complications in essential thrombocythaemia and polycythaemia vera, Platelets 15 (2004), 67-84. [48] H. Wajcman and F. Galacteros, Hemoglobins with high oxygen affinity leading to erythrocytosis. New variants and new concepts, Hemoglobin 29 (2005), 91-106. [49] G.W. Mack, C.A. Weseman, G.W. Langhans, H. Scherzer, C.M. Gillen and E.R. Nadel, Body fluid balance in dehydrated healthy older men: thirst and renal osmoregulation, J. Appl. Physiol. 76 (1994), 1615-1623. [50] M. Doubek, Y. Brychtova, M. Tomiska and J. Mayer, Idiopathic systemic capillary leak syndrome misdiagnosed and treated as polycythemia vera, Acta Haematol. 113 (2005), 150-151. [51] J.R. Smith and S.A. Landaw, Smokers' polycythemia, N. Engl. J. Med. 298 (1978), 6-10. [52] M. Puig-de-Morales-Marinkovic, K.T. Turner, J.P. Butler, J.J. Fredberg and S. Suresh, Viscoelasticity of the human red blood cell, Am. J. Physiol. Cell Physiol. (2007), (In press). [53] P.S. Frenette and G.F. Atweh, Sickle cell disease: old discoveries, new concepts, and future promise, J. Clin. Invest. 117 (2007), 850-858. [54] J. Delaunay, The molecular basis of hereditary red cell membrane disorders, Blood Rev. 21 (2007), 1-20. [55] M. Franchini, Thrombotic microangiopathies: an update, Hematology 11 (2006), 139-146. [56] J.N. George, Clinical practice. Thrombotic thrombocytopenic purpura, N. Engl. J. Med. 354 (2006), 1927-1935. [57] R.J. Murrin and J.A. Murray, Thrombotic thrombocytopenic purpura: aetiology, pathophysiology and treatment, Blood Rev. 20 (2006), 51-60. [58] I. Amirlak and B. Amirlak, Haemolytic uraemic syndrome: an overview, Nephrology (Carlton) 11 (2006), 213-218. [59] B.M. Cooke, N. Mohandas and R.L. Coppel, Malaria and the red blood cell membrane, Semin. Hematol. 41 (2004), 173-188. [60] A.M. Dondorp, F. Omodeo-Sale, K. Chotivanich, D. Taramelli and N.J. White, Oxidative stress and rheology in severe malaria, Redox Rep. 8 (2003), 292-294. [61] B.M. Cooke, N. Mohandas and R.L. Coppel, The malaria-infected red blood cell: structural and functional changes, Adv. Parasitol. 50 (2001), 1-86. [62] A.M. Dondorp, P.A. Kager, J. Vreeken and N.J. White, Abnormal blood flow and red blood cell deformability in severe malaria, Parasitol. Today. 16 (2000), 228-232. [63] M. Piagnerelli, K.Z. Boudieltia, M. Vanhaeverbeek and J.L. Vincent, Red blood cell rheology in sepsis, Intensive Care Med. 29 (2003), 1052-1061. [64] O.K. Baskurt, A. Temiz, and H.J. Meiselman, Red blood cell aggregation in experimental sepsis, J. Lab. Clin. Med. 130 (1997), 183-190. [65] T.L. Berezina, S.B. Zaets and G.W. Machiedo, Alterations of red blood cell shape in patients with severe trauma, J. Trauma 57 (2004), 82-87. [66] S.B. Zaets, T.L. Berezina, C. Morgan, M. Kamiyama, Z. Spolarics, D.Z. Xu, E.A. Deitch and G.W. Machiedo, Effect of trauma-hemorrhagic shock on red blood cell deformability and shape, Shock 19 (2003), 268-273. [67] J. Tatarishvili, T. Sordia and G. Mchedlishvili, Comparison of blood rheological changes in the microcirculation during experimental hemorrhagic and traumatic shock, Clin. Hemorheol. Microcirc. 35 (2006), 217-221. [68] C. Godin and A. Caprani, Effect of blood storage on erythrocyte/wall interactions: implications for surface charge and rigidity, Eur. Biophys. J. 26 (1997), 175-182. [69] R.T. Card, Red cell membrane changes during storage, Transfus. Med. Rev. 2 (1988), 40-47. [70] L.C. Wolfe, The membrane and the lesions of storage in preserved red cells, Transfusion 25 (1985), 185-203. [71] P. Porcu, et al., Hyperleukocytic leukemias and leukostasis: a review of pathophysiology, clinical presentation and management, Leuk. Lymphoma 39 (2000), 1-18. [72] M. Zarkovic and H.C. Kwaan, Correction of hyperviscosity by apheresis, Semin. Thromb. Hemost. 29 (2003), 535-542. [73] P. Harrison and D. Keeling, Platelet hyperactivity and risk of recurrent thrombosis, J. Thromb. Haemost. 4 (2006), 2544-2546. [74] L. Shore-Lesserson, Platelet inhibitors and monitoring platelet function: implications for bleeding, Hematol. Oncol. Clin. North Am. 21 (2007), 51-63. [75] R.A. Chaer, J.A. Graham and L. Mureebe, Platelet function and pharmacologic inhibition, Vasc. Endovasc. Surg. 40 (2006), 261-267. [76] K.T. Tan and G.Y. Lip, Platelet microparticles and platelet adhesion: therapeutic implications for the prevention and treatment of stroke, Curr. Treat. Opt. Cardiovasc. Med. 8 (2006), 251-258.


391

[77] R. Flaumenhaft, Formation and fate of platelet microparticles, Blood Cells Mol. Dis. 36 (2006), 182187. [78] C.N. Harrison and A.R. Green, Essential thrombocythaemia, Best Pract. Res. Clin. Haematol. 19 (2006), 439-453. [79] E.I. Penninga and O.W. Bjerrum, Polycythaemia vera and essential thrombocythaemia: current treatment strategies, Drugs 66 (2006), 2173-2187. [80] A.I. Schafer, Thrombocytosis, N. Engl. J. Med. 350 (2004), 1211-1219. [81] D.H. Buss, A.W. Cashell, M.L. O'Connor, F.2nd Richards and L.D. Case, Occurrence, etiology, and clinical significance of extreme thrombocytosis: a study of 280 cases, Am. J. Med. 96 (1994), 247-253. [82] J. Kutti and H. Wadenvik, Diagnostic and differential criteria of essential thrombocythemia and reactive thrombocytosis, Leuk. Lymphoma 22 Suppl 1 (1996), 41-45. [83] H.L. Daneschvar and H. Daw, Heparin-induced thrombocytopenia (an overview), Int. J. Clin. Pract. 61 (2007), 130-137. [84] B. Girolami and A. Girolami, Heparin-induced thrombocytopenia: a review, Semin. Thromb. Hemost. 32 (2006), 803-809. [85] T.E. Warkentin and J.A. Sheppard, Testing for heparin-induced thrombocytopenia antibodies, Transfus. Med. Rev. 20 (2006), 259-272. [86] C.H. Toh and W.K. Hoots, The scoring system of the Scientific and Standardisation Committee on Disseminated Intravascular Coagulation of the International Society on Thrombosis and Haemostasis: a 5-year overview, J. Thromb. Haemost. 5 (2007), 604-606. [87] P. Cauchie, C. Cauchie, K.Z. Boudjeltia, E. Carlier, N. Deschepper, D. Govaerts, M. Migaud-Fressart, B. Woodhams and D. Brohee, Diagnosis and prognosis of overt disseminated intravascular coagulation in a general hospital -meaning of the ISTH score system, fibrin monomers, and lipoprotein-C-reactive protein complex formation, Am. J. Hematol. 81 (2006), 414-419. [88] M. Franchini, G. Lippi and F. Manzato, Recent acquisitions in the pathophysiology, diagnosis and treatment of disseminated intravascular coagulation, Thromb. J. 4 (2006), 4. [89] M. Levi, E. de Jonge and T. van der Poll, Plasma and plasma components in the management of disseminated intravascular coagulation, Best Pract. Res. Clin. Haematol. 19 (2006), 127-142. [90] Y.K. Agrawal, H. Vaidya, H. Bhatt, K. Manna and P. Brahmkshatriya, Recent advances in the treatment of thromboembolic diseases: Venous thromboembolism. Med. Res. Rev. (2007), (In press). [91] L.A. Gorski, Venous thromboembolism: a common and preventable condition. Implications for the home care nurse, Home Healthc. Nurse 25 (2007), 94-100. [92] J.B. Segal, M.B. Streiff, L.V. Hofmann, K. Thornton and E.B. Bass, Management of venous thromboembolism: a systematic review for a practice guideline, Ann. Intern. Med. 146 (2007), 211-222. [93] F. Rogers, J.A. Rebuck and R.F. Sing, Venous thromboembolism in trauma: an update for the intensive care unit practitioner, J. Intensive Care. Med. 22 (2007), 26-37. [94] J. Mehta and S. Singhal, Hyperviscosity syndrome in plasma cell dyscrasias, Semin. Thromb. Hemost. 29 (2003), 467-471. [95] H.G. Hoffkes, B.M. Heemann, C. Teschendorf, M. Uppenkamp and T. Philipp, Hyperviscosity syndrome: efficacy and comparison of plasma exchange by plasma separation and cascade filtration in patients with immunocytoma of Waldenstrom's type, Clin. Nephrol. 43 (1995), 335-338. [96] J.R. Beck, B.M. Quinn, F.A. Meier and H.M. Rawnsley, Hyperviscosity syndrome in paraproteinemia. Managed by plasma exchange; monitored by serum tests, Transfusion 22 (1982), 51-53. [97] J.P. Isbister, Plasma exchange in the management of hyperviscosity syndromes, Bibl. Haematol. 47 (1981), 228-241. [98] A. Vijay and M.A. Gertz, Waldenstrom Macroglobulinemia, Blood (2007), (In press). [99] S.A. Johnson, J. Birchall, C. Luckie, D.G. Oscier and R.G. Owen, Guidelines on the management of Waldenstrom macroglobulinaemia, Br. J. Haematol. 132 (2006), 683-697. [100]A. Tedeschi, C. Barate, E. Minola and E. Morra, Cryoglobulinemia, Blood Rev. (2007), (In press). [101]Z.K. Shihabi, Cryoglobulins: an important but neglected clinical test, Ann. Clin. Lab. Sci. 36 (2006), 395-408. [102]F.P. McNicholl, Clinical syndromes associated with cold agglutinins, Transfus. Sci. 22 (2000), 125-133. [103]L. Doweik, T. Maca, M. Schillinger, A. Budinsky, S. Sabeti and E. Minar, Fibrinogen predicts mortality in high risk patients with peripheral artery disease, Eur. J. Vasc. Endovasc. Surg. 26 (2003), 381-386. [104]H. Blain, P. Cacoub, L. Musset, N. Costedoat-Chalumeau, C. Silberstein, O. Chosidow, P. Godeau, C. Frances and J.C. Piette, Cryofibrinogenaemia: a study of 49 patients, Clin. Exp. Immunol. 120 (2000), 253-260. [105]T.D. Amdo and J.A. Welker, An approach to the diagnosis and treatment of cryofibrinogenemia, Am. J. Med. 116 (2004), 332-337.

392


Clinical Significance of Hemorheological Alterations 1

Kalman TOTH1, Gabor KESMARKY1* and Tamas ALEXY1, 2 First Department of Medicine, University of Pecs, Pecs, Hungary; 2 Department of Physiology and Biophysics, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA

Introduction The preceding sections of this handbook have described many aspects of the macroand micro-rheological behavior of mammalian blood, and both the fundamental and applied areas of hemorheology have been considered. Several aspects of hemodynamics have also been addressed. Thus, up to this point, the careful and patient reader has hopefully gained insight into both hemorheology and hemodynamics. In the material below, the significance of hemorheological alterations in several clinical states is discussed. Vascular disease is covered in some detail, with subsequent chapters dedicated to such topics as pulmonary diseases, hematological and oncological disorders, ophthalmology and exercise. Each section provides a brief overview of the specific area, and then proceeds to a more detailed discussion of relevant hemorheological changes associated with the topic. It should be noted that the authorship and structure of this section differ somewhat from other chapters: 1) Essentially all authors have current or previous association with the Medical School of the University of Pecs, Hungary, and thus interact frequently. Such interactions allow easy exchange of information and comments and, to some degree, tend to foster a common approach to the material; 2) As might be expected, many literature citations are common to each clinical area presented. Therefore, to avoid confusion and repetition, a single reference list is provided at the end of the entire section. 1. Vascular Diseases – Atherosclerosis Contributing authors: Beata HORVATH, Gabor KESMARKY, Kalman TOTH, First Department of Medicine, University of Pecs, Hungary Vascular diseases are the leading cause of death and disability among the adult and elderly population in developed countries as well as in many developing nations. Atherosclerosis, a progressive process affecting the entire vascular system, serves as the basis for the development of vascular diseases [1]. Atherosclerosis starts in early

* Corresponding author: First Department of Medicine, University of Pecs, Pecs, Hungary; E mail: [email protected]

K. Toth et al. / Clinical Significance of Hemorheological Alterations

393

childhood and progresses asymptomatically until it is manifest as ischemic coronary syndromes, stroke, transient ischemic attack or peripheral artery disease later in life [2]. 1.1. Hemodynamic Factors Despite the systemic nature of risk factors, (e.g., hypertension, elevated lipid and lipoprotein levels, smoking, diabetes, physical inactivity, obesity, male gender, age), atherosclerotic lesions do not develop randomly throughout the vascular system. Instead, plaques tend to occur at the outer wall of vessel bifurcations, the inlet of branches and the inner, distal wall of arterial curvatures. The specific localization of atherosclerotic lesions and the close association between mean arterial blood pressure and the extent of the atherosclerotic process suggest that hemodynamic factors (e.g., altered lateral pressures, altered shear stresses, turbulence and flow separation) may also be important in plaque formation [1]. Blood vessels are constantly subject to mechanical forces such as wall stretch (i.e., both cyclic mechanical strain owing to the pulsatile nature of blood flow) and shear stress. Blood pressure is the major determinant of vessel stretch: it creates radial and tangential forces affecting all cell types in the vessel wall. Acute changes in vascular stretch or shear stress are compensated for via the transient adjustment of vessel diameter mediated by the release of vasoactive agents or by changes in the myogenic tone, while altered mechanical forces frequently initiate important adaptive alterations in the vessel wall shape and composition, a process termed vascular remodeling [3]. Disturbed blood flow leads to high shear forces at the inner vessel wall and turbulent, low shear stress at the outer wall and thus at the preferential sites for atherogenesis. Any increase in the vessel cross-section area initiates the local decrease of shear, in particular at the borderline zone. The ensuing disturbed flow with flow curves perpendicular to the vessel wall, and the periodic reversal of flow stream, can disrupt endothelial cell junctions and chronically over-stimulate cell metabolism. Thus, oscillating shear forces mechanically alter endothelial permeability, open cellular junctions and lead to denudation [1]. Decreased flow velocity leads to the sluggish transport of plasma components, thereby allowing a prolonged contact time between blood constituents and the endothelial cell surface [4]. In addition, axial migration of red blood cells results in a slower moving erythrocyte-poor layer along the vessel wall [5]: such low shear rates favor increased vascular permeability, white blood cell adherence and emigration [6]. As a consequence, trans-endothelial transport of lipoproteins and several other plasma macromolecules increases, leading to their local accumulation. This implies an alteration of the endothelial cell layer, which normally represents a selective barrier and regulates nutritive exchange between the blood and organ tissues [4, 7]. In contrast, endothelial cells are extended in the direction of the flow at the opposite vessel wall where they are exposed to high shear forces. Notably, structural abnormalities are less frequent at these sites [1]. In areas of flow separation, fluid circulates in an eddy or vortex between the mainstream and vessel wall, resulting in stagnation points downstream of arterial branches and stenotic regions. Red blood cells and platelets in vortices following a vessel expansion travel at different velocities and tend to migrate spirally outward into the mainstream. However, aggregates of red cells and platelets can centralize and grow within the vortices prior to joining the mainstream. Red cells adhere in stagnation regions near the flow reattachment points, whereas platelets adhere upstream. Consequently, it appears that areas of flow separation promote interaction of red cells,

394


platelets and the vessel wall and that such interactions may encourage platelet deposition and its atherogenic consequences [1]. 1.2. Hemorheological Parameters Based on current experimental data, there is a strong correlation between hematocrit and the extent of coronary and cerebral atherosclerosis. These associations could conceivably result from the rheological effect of red blood cells on atherogenesis, mediated by their interactions with the endothelium and/or platelets [1]. Decreased blood flow velocity subsequently allows RBC to begin to regain their resting, discoid shape. Under these conditions, tank-treading behavior of red cells (i.e., cell membrane rotation around the cell contents owing to tangential shear forces, a mechanism believed to facilitate erythrocyte flow) does not occur. Consequently, red cell aggregation is facilitated, in particular in the presence of high fibrinogen concentrations and low shear rates, thereby inducing a further deterioration of blood flow [4]. Plasma fibrinogen concentration has also been shown to be associated with the extent of coronary atherosclerosis. This correlation may reflect the effects of fibrinogen on red blood cell and platelet aggregation, fibrin formation and plasma viscosity. Low density lipoprotein (LDL) can either absorb on to the endothelial lining partly composed of fibrinogen/fibrin, or be taken up by a layered thrombus of fibrinogen aggregates [1]. Platelet deposition is further facilitated by the separation of blood flow at vessel bifurcations and bends [4]. Under normal circumstances, there is little if any adhesion of platelets to vascular endothelium. However, in case of vessel wall injury, subendothelial collagen fibrils are exposed to flowing blood, allowing rapid platelet adherence. As a consequence, a platelet rich thrombus forms that may contribute to the atherosclerotic process and may also embolize depending on local flow conditions, plug geometry and the mechanical properties of the thrombus. Furthermore, platelet adhesion and endothelial injury lead to the release of growth factors derived from platelets, endothelial and other cells into the arterial sub-endothelium. Proliferation of smooth muscle cells is stimulated by these growth factors; such stimulated cells can migrate from the media to the intima, synthesize collagen and accumulate lipid to form atheromatous plaques [1]. Polymorphonuclear leukocytes are activated by structurally-altered cells and have a tendency to adhere to injured endothelial cells [4]. They are also a major source for various biological mediators, with injured endothelial cells acting as “amplifiers” and “transmitters” by losing their antithrombogenic character and expressing tissue factor and secreting plasminogen activator inhibitor [5]. In overview, separation of blood flow favors the interaction of red blood cells, platelets, neutrophils and plasma macromolecules with the endothelium. This leads to vessel wall alterations via different mechanisms that can be considered as a pre-atherogenic state [4, 11]. 1.3. Molecular Mechanisms Receptors present on the endothelial cell (EC) surface and vascular smooth muscle cells (VSMC) allow vessels to adapt to changes of their physical environment. Various mechano-transduction cascades can be activated, depending on the nature of the stimulus received. Within vascular cells cytoskeletal proteins transmit and modulate the


395

tension between focal adhesion sites, integrins and the extracellular matrix. In addition to the above noted structural effects, signal transduction pathways might also initiate intracellular ionic changes mediated by ion channels, stimulate various membrane receptors, and thus promote complex biochemical responses. Many intracellular pathways, such as the mitogen-activated protein (MAP) kinase cascade and the nuclear factor-țB (NF-țB) pathway, are activated by flow or stretch and can initiate the activation of transcription factors and subsequent gene expression via sequential phosphorylations. Thus, blood vessels are capable of autoregulation, allowing them to adapt to their ever changing mechanical environment [3, 6]. Prolonged shear stress reduces MAP kinase induction stimulated by inflammatory mediators, thereby attenuating tumor necrosis factor (TNF)-stimulated expression of vascular cell adhesion molecule (VCAM). Conversely, oscillatory shear stress induces the expression of endothelial adhesion molecules, chemokines and growth factors that are important for leukocyte recruitment and extravasation, and is also associated with a shift in the balance between the levels of nitrogen oxide (NO) and reactive oxygen species (ROS), favoring the latter [3]. These changes in gene expression are mediated by the activation of NF-țB that occurs predominantly under low shear conditions [8] and in vascular regions with a high probability for atherosclerotic plaque formation [3]. VCAM-1 is also up-regulated in response to hypercholesterolemia [8]. In areas of high wall shear stress, NO synthesis appears to be a key mediator of atheroprotection by reducing endothelial permeability, migration of leukocytes and VSMC proliferation whilst concurrently promoting EC survival [3]. Low shear induced over-expression of endothelial cell NO synthase (eNOS) inhibits NF-țB activation. Laminar shear stress promotes eNOS synthesis in a dose-dependent manner, whereas oscillatory shear stress can down regulate eNOS. The production of NO and the attenuation of endothelium-dependent arterial relaxation are also some of the earliest responses to hypercholesterolemia [8]. There is increasing evidence that ROS may serve as strong pro-atherogenic mediators, with NADPH oxidase being the most important source for reactive oxygen species formation. This enzyme is regulated differentially by shear stress, with oscillatory shear stress considered to be the most important factor responsible for its activation. Other members of the reactive oxygen species system are also influenced by flow conditions, including heme-oxygenase and Cu/Zn superoxide dismutase. Consequently, the critical balance between the pro-oxidant and anti-oxidant factors may determine the likelihood of developing atherosclerosis, with this balance influenced by the magnitude and degree of oscillation of shear stress [8]. 1.4. Humoral Factors Fluid shear stress may play an additional role in the process of atherosclerosis by modulation of synthesis and secretion of humoral factors released by vascular endothelium that mediate platelet-wall interaction. von Willebrand factor (vWF), a multimeric glycoprotein, is a key molecule involved in the acute occlusion of atherosclerotic vessels. Under physiologic conditions, endothelial-derived vWF mediates white cell adhesion to the vessel wall and promotes platelet adhesion to the subendothelium if there is endothelial cell detachment. Early events that determine platelet-vWF interaction depend on the platelet glycoprotein (GP) Ib-IX-V receptor complex, whereas subsequent events are mediated by the GP IIb-IIIa receptor complex exposed on the surface of activated platelets. Laminar shear stress greater than about

396


1 Pa acting on a cell monolayer induces a significant increase in the amount of vWF released in culture media, with the amount of release dependent on the level of shear. Conditions associated with vascular dysfunction (i.e., hypercholesterolemia, diabetes, hypertension, smoking) are characterized by an increase in circulating vWF levels, suggesting an important role of vWF in mediating thrombotic complications of atherosclerosis [9]. 1.5. Hemorheological Parameters and Conventional Cardiovascular Risk Factors Patients with hyperlipoproteinemia type II have elevated Į-2 antiplasmin levels and high plasma fibrinogen concentrations. Blood viscosity measured at different shear rates, plasma viscosity and hematocrit are also increased, whereas red cell deformability is not significantly different from control values [4]. A large body of evidence supports the association between smoking and elevated hemoglobin and fibrinogen levels, plasma viscosity, red cell aggregation, platelet aggregation and leukocyte count [4]. Smoking also increases erythrocyte mass, thus increasing whole blood viscosity [1]. Several hemorheological parameters are altered in arterial hypertension: whole blood and plasma viscosity, hematocrit, fibrinogen and plasma protein concentration, red cell aggregation and red cell rigidity are all increased [4]. The heart, when facing elevated peripheral resistance due to hyperviscous blood, can only maintain cardiac output by increasing the perfusion pressure. The ensuing high blood pressure can compensate, at least in part, for the increased blood viscosity [1]. Thus, rheological mechanisms might be involved in hypertensive microangiopathy; studies have shown increased vascular resistance at the level of the myocardial microcirculation [4]. Diabetes is a disorder associated with excessive cardiovascular risk, in particular when present with other known risk factors. Diabetic patients are known to have several rheological alterations; the magnitude of these disturbances correlates with the extent of the metabolic derangement and can be improved by normalizing plasma glucose levels. The predictive power of decreased blood fluidity for diabetic microcirculatory complications has been reported [4]. In addition to the above noted disorders, several other conditions are known to increase cardiovascular risk and to have hemorheological disturbances. These include, but are not limited to obesity, low socioeconomic class, stress, sedentary lifestyle and the use of oral contraceptives [4]. It is still not yet clear whether rheological alterations are contributing factors or are simply present as the consequence of these disorders. However, the fact that rheological alterations appear prior to the development of vascular lesions suggests that they might have a significant role in the pathogenesis of atherosclerosis [10]. 1.6. Complications of Atherosclerotic Plaques The formation of thrombi on underlying coronary atherosclerotic plaques is believed to be a common feature of acute coronary events. Thrombembolism leading to transient ischemia or major infarction of the brain or limbs usually arises from a cardiac source or from stenotic and/or ulcerated atheromatous lesions in the nutritive arteries. Ulceration of atherosclerotic plaques exposes blood to the sub-endothelium which is rich in collagen, lipids and cell debris, all of which might lead to platelet activation and coagulation. However, the exact pathomechanism of arterial plaque rupture and


397

ulceration is poorly understood to date. High shear stress is expected to develop above a critical arterial stenosis, with this high shear forces promoting the mechanical ulceration of the vessel wall and/or causing platelet activation. High shear stress might continuously liberate ADP from red blood cells at stenotic sites, especially during plaque rupture, promoting platelet activation and thrombus formation [1]. On the other hand, post-stenotic deceleration of the blood stream induces flow separation and the development of recirculation zones (i.e., vortices), allowing prolonged blood cellvessel wall contact. Erythrocyte aggregation is favored in these areas, in particular in regions with high local hematocrit. These phenomena, in cooperation with reduced shear forces, lead to increased local blood viscosity. Fibrinogen seems to play a significant role in this process: it induces platelet and red cell aggregation and increases whole blood and plasma viscosity [4]. Decreased shear forces and granulocyte activation at the site of endothelial injury lead to prolonged endothelial cell-blood cell contact at the microcirculatory level, thereby leading to temporary capillary occlusion [4]. Several epidemiological studies have shown that high-normal levels of hematocrit, plasma fibrinogen, and white cell count are primary risk factors for sudden coronary death, myocardial infarction, unstable angina and stroke: such associations might result from the rheological effects of risk factors [1]. 1.7. Conclusion Experimental, clinical and prospective epidemiological studies provide convincing evidence that atherosclerotic and thrombotic states are associated with alterations in hemorheological and hemostatic profiles [4]. There are at least three possibilities by which hemorheological factors might promote thrombogenesis: reduction of blood flow owing to unfavorable rheological effects, predisposition to thrombosis via a hypercoagulable state, or an enhancement of atherosclerosis by fibrinogen and its metabolites. A better understanding of the mechanisms involved in atherosclerosis could lead to a further improvement in the prevention and treatment of vascular diseases [4].

2. Vascular Diseases - Cardiovascular Contributing authors: Gabor KESMARKY, Kalman TOTH, First Department of Medicine, University of Pecs, Hungary Cardiovascular diseases are the most frequent cause of mortality world wide, yet despite extended research in this field, their exact pathomechanisms remain unclear. Investigations have explored the possible pathophysiological role of the vascular wall, cellular interactions and many of the underlying biochemical processes. However, much less attention has been paid to the properties of the circulating blood. 2.1. Epidemiological Data In most cases, ischemic heart disease is caused by coronary artery disease (i.e., the stenotic lesion of one or more coronary branches). Both hemodynamic and hemorheologic factors are of crucial importance to consider in the coronary circulation. Several studies have shown that hemorheological parameters are primary

398


cardiovascular risk factors: 1) the Framingham Study, Puerto Rico Heart Health Program and Honolulu Heart Program demonstrated hematocrit to be a cardiovascular risk factor; 2) the Framingham and Northwick Park Studies showed the same for elevated plasma fibrinogen concentration [12, 13]; 3) the Monica Project showed viscosity to be a risk factor [14]; 4) the Edinburgh Artery Study and Caerphilly and Speedwell Collaborative Heart Disease Study provided further confirmation regarding the role of several hemorheological variables in the development of cardiovascular diseases [15, 16, 17]. As shown in the Physicians’ Health Study, elevated baseline plasma fibrinogen concentration was associated with an increased risk for future myocardial infarction during a five-year follow-up period, with this finding independent of any other risk factors [18]. Carotid intima-media thickness is an accepted marker for the early, sub-clinical stage of atherosclerosis. The Edinburgh Artery Study revealed a strong association between selected hemorheological factors and the carotid intima-media thickness [19]; this association was independent of other cardiovascular risk factors, including total cholesterol, blood pressure and smoking. Fibrinogen, one of several hemorheological variables, may promote atherosclerosis via increasing platelet aggregation, fibrin formation, and blood viscosity while diminishing fibrinolysis. Elevated blood viscosity may, in turn, promote platelet adhesion to the endothelium, increase protein filtration into the arterial wall, and alter local shear forces at sites of atherogenesis. The Scottish group explained the greater susceptibility of males to elevated blood viscosity based on gender differences of vascular geometry and wall shear forces. Toth, et al. showed that an evaluation of hemorheological variables, along with the other diagnostic procedures, can help clinical decision-making in persons with suspected ischemic heart disease [20]. Neumann, et al. described an association between impaired blood fluidity and chronic coronary artery disease, with the finding being independent of the severity of coronary atherosclerosis. The same group found that plasma viscosity and red blood cell aggregation were higher in subjects with unstable angina, and that the marked elevation of these parameters identified a patient subgroup with unstable angina who were at high risk for acute myocardial infarction [21, 22]. Junker, et al. showed plasma viscosity to be related to the severity of coronary artery disease, and suggested plasma viscosity as a linking mechanism between other cardiovascular risk factors and coronary heart disease [23]. Lowe, et al. proposed that blood viscosity is linked to the extent of coronary heart disease, but they did not find such an association regarding plasma viscosity [24]. Rainer, et al. discovered abnormal hemorheological variables in patients with angiographically proven coronary artery disease although no association with the extent of the disease could be established [25]. Kesmarky, et al. have proposed that the severity of coronary artery disease, as demonstrated by coronary angiography, was associated with more abnormal hemorheological variables compared to those with mild vessel disease [26]. Fibrinogen has been shown to be an independent risk factor for cardiovascular diseases; this protein plays a central role in platelet and erythrocyte aggregation and is one of the major determinants of plasma and whole blood viscosity. In recent studies, elevated fibrinogen and von Willebrand factor levels were proposed to be risk factors for re-stenosis following percutaneous coronary interventions (PCI). Another acute phase marker, C-reactive protein, was also found to be increased in acute ischemic coronary syndromes. Acute phase reactions occur following an organic injury as part of a systemic inflammatory response. Plasma fibrinogen concentration is known to be elevated following unstable angina and acute myocardial infarction, with higher values


399

associated with poorer outcome. Several mechanisms may be responsible for the inflammatory response induced by coronary interventions: plaque rupture, hemorrhage, arterial wall damage, the release of inflammatory and chemoattractant molecules from activated platelets and leukocytes, and repeated ischemia and reperfusion owing to several balloon inflations and deflations. Elevation in fibrinogen after PCI might increase the risk for recurrent ischemic coronary events [27, 28, 29]. Inflammation is an established risk factor for ischemic heart disease (IHD). However, the underlying pathomechanisms are still debated: 1) it might increase the incidence of clinical events via plaque destabilization and making blood hypercoagulable, 2) it might reflect more extensive coronary atherosclerosis. Besides inflammation, the rheological properties of blood also contribute to the pathogenesis of ischemic heart disease. The major components of blood viscosity are blood cell mass (i.e., hematocrit), plasma viscosity, red cell aggregation and red cell deformability. Red cell aggregation can be estimated via measuring the erythrocyte sedimentation rate (ESR), which also reflects the plasma concentration of acute phase response proteins. Higher ESR values have been found to be independently associated with the severity of angiographically documented coronary atherosclerosis. Although statistically ESR “explained” only a small fraction of coronary atherosclerosis, the same was true regarding the classic cardiovascular risk factors. Moreover, ESR was a predictive marker for cardiac death that proved to be more powerful than gender, smoking or hypercholesterolemia [30]. Abnormal coronary blood flow can be associated with hyperlipidemia even in the absence of structural changes in the coronary arteries. This has been partly attributed to endothelial dysfunction, yet on the other hand, increased plasma and whole blood viscosity were observed in patients with hypercholesterolemia. Lipid lowering drugs can normalize abnormal coronary blood flow reserve without affecting coronary morphology [31]. However, the simplest evidence showing the importance of hemorheological changes was provided by a study examining the effects of water consumption (i.e., two or fewer vs. five or more glasses of water a day). Low water intake was associated with a higher incidence of cardiovascular morbidity and mortality [32]. Smoking, obesity, sedentary lifestyle and diabetes mellitus are known as classic risk factors, and their accompanying hemorheological disturbances may serve as the common basis for the development of cardiovascular diseases. Blood flow characteristics may also serve as the link between geographical and seasonal variations of ischemic heart diseases. 2.2. Myocardial Ischemia The coronary vessel system is a special part of the circulation since there is a continuous change in blood flow, perfusion pressure and shear rate during the cardiac cycle. In addition, the narrowest capillaries of the body can be found in the myocardium, with their diameter as small as 3-5 μm and thus several micrometers below the 8 μm resting diameter of RBC. Therefore, the role of rheological alterations may be of greater importance in the cardiac circulation than in other parts of the vascular system. Under normal circumstances, coronary blood flow is primarily determined by hemodynamic factors, yet under critical conditions, the effects of blood rheology become increasingly important [1, 33, 34]. Abnormal rheological parameters might play a role in the deterioration of blood flow in small vessels and might also contribute to the delayed restoration of coronary circulation following a successful

400


coronary intervention. At rest, maximal resistance to coronary blood flow is offered by the coronary arterioles with only a minimal contribution by the capillaries and venules. However, under specific circumstances (e.g., during hyperemia), coronary arterial and venous resistance decreases significantly with capillary resistance remaining unaffected. Thus, capillary resistance determines the maximal possible increase in hyperemic coronary blood flow. Further, resistance to flow is greater for equal hematocrit samples in capillaries than in a glass tube with the same luminal diameter, most likely due to interactions between irregular vessel endothelium and blood cells. Large vessel dilation is induced by changes in microvascular flow and not vice versa, and thus increased capillary resistance caused by increased blood viscosity can explain the lack of large vessel dilation in response to vasodilators or the no-reflow phenomenon following thrombolysis or coronary balloon dilation [31]. In chronic ischemic heart disease, all hemorheological factors are in the abnormal range, including plasma fibrinogen, hematocrit, red cell filterability and plasma and whole blood viscosity. Whether these changes are primary or secondary to the ischemia is still uncertain, but their effects on the flow properties of blood might further worsen myocardial ischemia. Changes in flow properties when myocardial perfusion pressure is decreased are expected to be critical in areas with reduced reserve for reflex vasodilation (i.e., vascular “reserve”). Adverse hemorheological changes can be suspected as a primary cause when patients have clinical and electrocardiographic evidence of myocardial ischemia in the absence of any demonstrable coronary artery defects. Hemorheological changes have been found to be more pronounced in cases of unstable angina or preceding an acute infarction; hemorheological abnormalities can facilitate arterial thrombotic processes, thereby turning them into an acute event [1]. In acute myocardial infarction, there is a series of vascular and myocardial alterations, that include rheological changes [1, 35, 36] such as primary hemoconcentration, increased plasma fibrinogen level, red blood cell aggregation, and decreased red cell filterability. These alterations lead to an increase in blood viscosity, although their time scales differ. High hematocrit has been documented immediately after acute myocardial infarction with decreasing values during the first week. The elevated hematocrit might be explained by fluid restriction and increased perspiration during the acute event, with the subsequent decline attributed to the body’s autoregulation, intravenous fluid administration, and frequent blood samplings. Epidemiological data suggest that a higher hematocrit, even within the normal range, is a risk factor for myocardial infarction, with this acute event being rare in patients with anemia. Thus, lower hematocrit values might serve as a possible explanation for the more favorable morbidity rates observed in women. Plasma fibrinogen concentration increases as part of the acute phase reaction following myocardial infarction, with increased fibrinogen levels first observed one or two days after the acute event and reaching a peak on the 3-5th day. Plasma viscosity increases in parallel with fibrinogen, and the concentration of other plasma proteins (e.g., Į2-macroglobulins) increases similarly, although they have a less prominent effect on plasma viscosity. Increased fibrinogen facilitates red blood cell aggregation that is responsible for the elevated blood viscosity values at low flow rates. Blood filterability has been demonstrated to decrease (i.e., slower flow through filter) after myocardial infarction; the lowest values were found twelve hours after the acute event with a return to a near-normal range after twenty-four hours. The unfavorable effects of white blood cells are thought to be responsible, at least in part, for the observed alterations since in vitro red cell filterability tests indicate slower filter flow in the presence of contaminating white blood cells. In microscopic studies, white


401

blood cells can often be seen obstructing capillaries with trains of red cells piling up behind them, and thus alterations of leukocyte deformability can also play a significant role in the development and progression of myocardial damage. All of the factors noted above affect whole blood viscosity. Although methodological difficulties make direct comparisons difficult, numerous studies are in agreement that blood viscosity is impaired in acute myocardial infarction. We can therefore assume that deterioration of blood flow properties plays a significant role in the pathological and clinical outcome of the acute event. Studies on experimental animals and clinical evidence show that early rheological changes relate to the size of myocardial necrosis and to the clinical prognosis. The severity of clinical symptoms and complications (e.g., arrhythmia, shock, death) can also be correlated to elevated fibrinogen, plasma and whole blood viscosity, and to reduced erythrocyte filterability. A vicious cycle can be imagined where the local metabolic and vascular changes cause hemorheological deteriorations which, in turn, increase the size of the ischemic area. 2.3. Hypertension Hypertension is the most frequent cardiovascular disease and is one of the most important risk factors for the development of other vascular disorders and organ damage. Arterial pressure is determined by the peripheral resistance and cardiac output; peripheral resistance is composed of arterial vascular resistance (i.e., vascular hindrance) and the viscous resistance of blood. When peripheral resistance increases owing to hyperviscosity, the heart can only maintain cardiac output by increasing perfusion pressure; both systolic and diastolic blood pressures correlate with hematocrit, plasma fibrinogen concentration, plasma and whole blood viscosity. Blood pressure remains associated with blood viscosity over a wide range of hematocrit, indicating that red cell aggregation and deformability also contribute to hypertension. Red blood cell rigidity can be affected by cellular volume and thus may be affected by the genetically disturbed function of Na+/K+ ATP-ases that is also responsible for sodium retention in hypertensive patients. Adrenergic hyperactivity may contribute to red blood cell rigidity as well as to vascular tone, and abnormal hemorheological parameters may contribute to the organic complications of hypertension, especially in left ventricular hypertrophy and hypertensive retinopathy [1, 34, 37]. 2.4. Heart Failure Rheological abnormalities most likely contribute to disease progression in heart failure and cardiomyopathies. Coronary reserve capacity is severely impaired in these conditions owing to hemodynamic changes, and hemorheological disturbances can further deteriorate tissue perfusion. In experimental animal studies, myocardial microcirculatory disturbances were suggested to cause dilated cardiomyopathy. Decreased red cell deformability, elevated plasma fibrinogen and whole blood viscosity values found in patients with dilated cardiomyopathy may have contributed to reduced coronary blood flow reserve. Hypoxemia observed in severe heart failure causes increased red cell production and hematocrit in an effort to enhance oxygen transport, yet when the hematocrit increase is excessive it can lead to hyperviscosity with a concomitant increase in tissue hypoxia, thus leading to a vicious circle aggravating heart failure [38].

402


3. Vascular Diseases - Cerebrovascular Contributing author: Laszlo SZAPARY, Department of Neurology, University of Pecs, Hungary Stroke is a highly prevalent disorder worldwide: it is the third main cause of death following coronary heart diseases and cancer, and is the leading cause of severe disability. Strokes are generally categorized into two groups, with recent data showing that 72-86% of strokes are ischemic with the remaining events due to intracranial or subarachnoidal hemorrhages. Arterial obstruction and microcirculatory stasis are the main factors in the pathomechanism of brain ischemia. While the pathophysiological importance of certain risk factors (e.g., smoking, hypertension, diabetes, etc.) for the development of brain ischemia has been known for a long time, the importance of hemorheological parameters has only been recognized for the last three decades. Several studies have shown that elevated hematocrit (HCT), plasma fibrinogen concentration (PFC), plasma viscosity (PV), whole blood viscosity (WBV) and increased red blood cell (RBC) aggregation are important risk factors for cardio- and cerebrovascular diseases. Not surprisingly, the incidence of stroke is elevated in diseases with abnormal blood flow properties. The Framingham study showed that the incidence of cerebral infarction doubled when the hemoglobin exceeded 150 g/l in men and 140 g/l in women. As a possible explanation, whole blood viscosity increases with hematocrit and leads to reduced cerebral blood flow, and elevated PFC is reported to be a significant risk factor for ischemic stroke. Previous studies have demonstrated hemorheological alterations in an asymptomatic group balanced for stroke risk factors as well as in patients with transient ischemic attack. These results support the hypothesis that blood fluidity parameters are impaired before an ischemic stroke and may become important factors in the development of brain ischemia. 3.1. Hemorheological Parameters and Cerebral Blood Flow A number of factors that lead to decreased blood cerebral blood flow (CBF) following acute ischemic stroke have been identified [39, 40]. In many cerebrovascular patients, the diffuse reduction of CBF cannot always be explained by vascular obstruction; following the removal of a significant carotid stenosis, CBF does not always improve. Blood fluidity has a physiological influence on CBF, and rheological parameters may promote thrombogenesis, atherosclerosis and may become important factors in the regulation of cerebral blood flow. Previous data have shown that elevated WBV correlates with decreased CBF and that there is a statistically significant association between CBF and both serum fibrinogen level and hematocrit [40]. Other authors have found the same relationship for elevated fibrinogen concentration, HCT and PV. It is possible that in patients with acute ischemic stroke, increased whole blood viscosity leads to a further reduction of CBF in regions of low flow. Erythrocyte aggregation is closely associated with cerebral blood flow, and a significant negative correlation was found between CBF and mean RBC aggregation index in subjects over the age of 45, whereas no such association was found for subjects less than 45 years old. The influence of blood fluidity on cerebral blood flow might be more pronounced when autoregulation is impaired (e.g., in severe cerebrovascular diseases).


403

3.2. Hemorheological Disturbances in Acute Ischemic Stroke The basis for blood hyperviscosity in patients with acute stroke is uncertain. An acute phase response develops following various infections, inflammation or tissue necrosis but the specific reaction occurring in cerebral infarction is poorly defined to date. Previous data suggest that the acute phase response developing after cerebral infarction is less common and weaker than that following myocardial infarction and that the extent of the response is associated with preceding bacterial infections or with an infectious complication [41]. Increased concentrations of acute phase proteins, such as plasma fibrinogen and C-reactive protein, have been reported in acute ischemic stroke and are associated with an unfavorable outcome. One can thus suspect that the general stress response and the associated increase of acute phase proteins have a negative effect on blood fluidity. However, it remains controversial whether rheological abnormalities are causally related to acute brain ischemia, or simply reflect the presence of an acute phase response. An etiological role for impaired rheological parameters is supported by several investigators who found that abnormalities of blood fluidity are present before the acute event. A transient ischemic attack (TIA), especially with severe carotid artery stenosis, represents a precursor of stroke, and previous studies have reported significantly elevated plasma viscosity, erythrocyte aggregation and impaired RBC deformability in TIA patients [42]. In focally ischemic regions, the gradient of shear rates within a vessel is reduced as the velocity of blood flow decreases distal to the narrowed or occluded segment of the cerebral artery. A vicious cycle emerges as the elevated blood viscosity within the region of slow blood flow further compromises perfusion and enhances ischemia. Several studies have evaluated hemorheological parameters in patients with acute ischemic stroke, with acute cerebral ischemia related to increased hematocrit and fibrinogen levels, high WBV and enhanced RBC aggregation [39, 43, 44]; an association between elevated HCT and cerebral infarct size has also been reported. Several investigators have shown that rheological abnormalities play an essential role in the pathomechanism of microvascular diseases including lacunar brain infarcts. Cerebral microcirculation begins where 30-70 μm arterioles penetrate into the brain substance, and blood flow characteristics are markedly more complex in this microvasculature than in the macrocirculation [39]. At the microvascular level of the cerebral circulation, PV, RBC aggregation and deformability are the main determinants of blood flow resistance, with platelet aggregation being a minor contributor. During acute brain ischemia, RBC aggregation and deformability might emerge as dominant factors controlling the distribution of blood flow in the microvasculature. Reports also suggest that RBC aggregation is elevated in patients with acute lacunar brain infarcts and contributes to microcirculatory disturbances. RBC aggregation in the three main subtypes of acute cerebral infarction (i.e., atherothrombotic, lacunar, embolic) has been studied, with results indicating that aggregation is enhanced in acute ischemic stroke but that no differences exist between the subtypes. These results suggest that all ischemic strokes are associated with significant rheological abnormalities, independent of the subtype [44]. Temporal studies of blood rheology have been reported, with results indicating that hemorheological disturbances in patients with acute ischemic stroke improved gradually but persisted for several weeks after onset of cerebral infarction; normalization of hemorheological parameters paralleled clinical recovery.

404


3.3. Hemorheological Disturbances in Chronic Cerebrovascular Patients Among the 80-85% of patients who survive a stroke, the risk of a recurrent stroke is somewhere between 5 and 15% during the first year, and 30-40% within 5 years. Impairment of blood fluidity is not solely a consequence of the acute event, since in many patients this deficit persists for at least a year, predisposing them to repeated cerebro- or cardiovascular events: more than 40% of subjects have hemorheological abnormalities following an acute ischemic stroke. Chronic ischemic cerebrovascular disorders are characterized by chronic hyperviscosity, persistently increased WBV and elevated fibrinogen levels in 48% of patients. Other studies have documented increased RBC aggregation and significantly reduced deformability values following cerebral ischemia [44]. Rheological abnormalities may become predominant even when other vascular abnormalities are also present. A hemodynamically significant carotid stenosis changes blood velocity and shear forces in the vicinity of the luminal narrowing and the post-stenotic region. Several studies have demonstrated a positive correlation between increased plasma fibrinogen concentration and the degree of carotid artery stenosis, implying that high fibrinogen levels can accelerate atherosclerosis. Hemorheological factors may also be of prognostic relevance in cerebrovascular disorders inasmuch as patients with a subsequent stroke were found to have elevated plasma fibrinogen concentration, PV, WBV, and red cell aggregation. The importance of serious hemorheological deficits in chronic cerebrovascular diseases is emphasized by the observation that altered rheological parameters persist during secondary prevention despite vigorous medical treatment [45]. Thus the presence of impaired hemorheological conditions should be considered in the therapeutic approach to vascular disorders: correction of disturbed blood fluidity could be helpful in the prevention of acute ischemic attacks and in the reduction of their incidence in chronic ischemic stroke patients.

4. Vascular Diseases - Peripheral Contributing authors: Peter KENYERES1, Zsolt PECSVARADY2, 1First Department of Medicine, University of Pecs, Hungary and 2Department of Medicine-AngiologyCardiology, Flor Ferenc Hospital, Kistarcsa, Hungary 4.1. Peripheral Arterial Disease Peripheral arterial disease (PAD) is a manifestation of atherosclerosis in which large or medium-sized arteries other than the coronaries, carotid and cerebral arteries are affected, thereby leading to a wide variety of diseases and symptoms. Atherosclerosis is a systemic disease affecting the entire vascular system, and vascular branch points and curvatures of arteries where laminar flow becomes turbulent are especially prone to vessel injury. However, these predilection sites seem to show different susceptibility for various risk factors. The PDAY study (Pathological Determination of Atherosclerosis in Youth) examined atherosclerotic predilection sites in young people who died from non-cardiovascular cause. Smoking was associated with plaques in the abdominal aorta, while impaired glucose tolerance affected primarily the right coronary


405

artery. Diabetes mellitus promoted the development of PAD in the lower legs to a significantly greater extent than at other sites [54]. The pathophysiology of atherosclerosis has been described in previous sections. These mechanisms lead to endothelial dysfunction and plaque formation, with plaque progression associated with stenotic regions that maintain turbulent blood flow. A number of risk factors that accelerate atherosclerosis have been introduced, and numerous studies have demonstrated that hemorheological parameters are impaired in atherosclerosis: cause and effect relations are not yet certain. Koenig, et al. described a positive association between PAD and increased plasma viscosity and also found a tendency for higher hemoglobin values in 4022 subjects [47]. The Edinburgh Artery Study involved 1592 men and women at age of 55-74 years [48], and several relevant conclusions based upon the resulting data have been published: 1) Woodburn, et al. found fibrinogen levels, von Willebrand factor, plasminogen activator inhibitor and fibrin turnover to be strongly associated with the presence of symptomatic PAD and an elevated white cell count was noted [49]; 2) Smith, et al. found that elevated fibrinogen levels, plasma and whole blood viscosity in patients with uncomplicated intermittent claudication were risk factors for a later vascular intervention during a 6year follow-up [50]; 3) In contrast, Tzoulaki et al. found only C-reactive protein, fibrinogen levels, lipoprotein(a) and hematocrit to be significantly associated with PAD after adjusting for cardiovascular risk factors and baseline cardiovascular disease. The roles of interleukin 6 (IL-6), intercellular adhesion molecule-1, d-dimer, tissue plasminogen activator antigen, plasma and whole blood viscosity values were considered to be minor [51]; In a study by Koksal et al., non-diabetic, normotensive patients with critical limb ischemia were found to have fibrinogen and plasma viscosity values similar to those of healthy controls, while blood viscosity and erythrocyte rigidity were even lower than the controls [52]. Possible biophysical pathways by which stenoses might lead to direct or indirect erythrocyte damage and thus impaired red cell deformability have been presented [53]. A number of authors found reduced blood filterability in PAD patients, yet filter transit times of washed red blood cells did not differ significantly from normal controls, suggesting that leukocytes might be blocking the filter pores. Thus, while the deformability of erythrocytes may be unaltered, the rigidity of white blood cells may be increased, with these rigid leukocytes causing severe microcirculatory disturbances [1]. Venous blood samples obtained from a patient’s leg during acute ischemia showed rapid, irreversible changes in blood viscosity, with observation only found in individuals with significant PAD [1]. Since vascular resistance is directly proportional to the fourth power of vessel diameter, an atherosclerotic plaque might cause critical stenosis such that vasodilation distal to the stenosis is unable to maintain flow and perfusion. Under these conditions the effect of blood viscosity on flow resistance becomes a substantial importance: in dilated arterioles blood flow velocity is slower than normal, and thus viscosity is higher than it would be at physiologic shear rates. Thus, blood viscosity is an important factor in arteries and arterioles whereas capillary resistance is primarily determined by plasma viscosity and erythrocyte deformability. 4.2. Raynaud’s Phenomenon A transient disorder of the microcirculation can be experienced in Raynaud’s phenomenon, and while vascular and neurological mechanisms have been considered in its pathogenesis, blood rheology undoubtedly also plays an important role. Skin and

406


peripheral regions are colder than core temperature, leading to increased blood and plasma viscosity; the increases are greater if cryoproteins or excessive amounts of fibrinogen are present. Baseline viscosity parameters and fibrinogen level of Raynaud’s phenomenon patients are usually increased, and while these increases alone do not trigger attacks, they make the patients more susceptible to them. Attacks are triggered by vasoconstriction, which turns the region pale and almost white; blood flow slows, its viscosity further increases, and when shear forces drop below the yield stress of blood, flow completely stops. Erythrocyte deformability is often impaired in these patients, further affecting microcirculatory blood flow. The static blood loses oxygen giving rise to cyanotic coloring with prolonged ischemia leading to tissue damage; platelet activation and impaired fibrinolysis during attacks have also been suggested. The attacks end with the restoration of vascular tone and reactive hyperemia [1, 55]. 4.3. Chronic Venous Insufficiency Chronic venous insufficiency of the lower limb, as well as crural ulcer, are common diseases. Chronic venous insufficiency leads to the development of crural ulcer by increasing venous pressure in the affected area. Insufficiency of either the deep, superficial or perforator veins plays a role in ulcer development as a consequence of reflux, occlusion or defective muscle-pump function [56]. Such an abnormality is unknown in animals, suggesting that standing in bipeds plays an essential role in its formation. In veins, one-way valves protect lower legions from hydrostatic forces. Centrifugal tension and venous hypoxia, however, slowly lead to vessel wall damage and remodeling. Altered shear conditions around valves create stagnant zones where white cells adhere and migrate. The vessel wall remodels and dilates, valves become insufficient and intravenous pressure further increase [6]. Devehat et al. reported hemorheological changes in patients with chronic venous insufficiency and found that they had higher fibrinogen levels, and thus increased plasma viscosity and erythrocyte aggregation. These changes were even more profound for samples drawn from leg veins: during stasis these parameters and hematocrit increased while erythrocyte deformability decreased, with no such changes seen in controls [57]. Chronic venous stasis and increased venous blood pressure serve as the basis for the development of crural ulcer. Capillary permeability increases due to oxygen deficiency caused by reduced blood flow and increased venous pressure, leading to extravasation of protein- and fibrinogen-rich fluid from the capillaries. High fibrinogen concentration leads to fibrin cuff formation around the veins, impeding the diffusion of nutrients. The microcirculation becomes impaired, and without intervention, necrosis is inevitable. Activated leukocytes, predominantly neutrophil granulocytes, play the principal role in the development of crural ulcer. Although neutrophil granulocytes are present in the peripheral circulation in only relatively low numbers, they are 2-3 times larger and their deformability is approximately 1000-times lower than red blood cells; they may thus slow down or even block capillary circulation. Further neutrophil activation leads to free oxygen radical generation and also to the activation of various proteolytic enzymes. These alterations, in addition to poor circulation and nutrient supply, contribute significantly to tissue necrosis and ulcer formation. Endothelial function also is greatly impaired: there are decreases in the amounts of endothelium-related protective agents (e.g., NO, endothelium derived hyperpolarizing factor, prostacyclin), while cell adhesion molecules are over-expressed. Local thrombus formation also contributes to this process through platelet activation and the


407

coagulation system. Tissue plasminogen activator (t-PA), an important regulator of the fibrinolytic system produced by endothelial cells, plays a key role in the control of thrombus formation by hindering the development of peri-capillary fibrin cuffs. However, the production of t-PA is significantly reduced in chronic venous insufficiency owing to damaged endothelial cells. Increased capillary permeability leads to an increase in blood viscosity that may further impair microcirculatory blood flow [58].

5. Vascular Diseases - Diabetes Mellitus Contributing authors: Katalin KOLTAI1, Zsolt PECSVARADY2, 1First Department of Medicine, University of Pecs, Hungary and 2Department of Medicine-AngiologyCardiology, Flor Ferenc Hospital, Kistarcsa, Hungary Diabetes mellitus is a clinical term denoting a group of metabolic impairments which affect glucose utilization and lead to hyperglycemia. The importance of diabetes is reflected by the fact that the worldwide prevalence of the disease is projected to double to 300 million by 2025. Type 1 diabetes is characterized by the complete absence of insulin, while type 2 diabetes is characterized by hyperinsulinemia and insulin resistance which precedes the development of hyperglycemia. The latter form of diabetes accounts for at least 90% of diabetic cases worldwide and it is associated with modern lifestyle characterized by abundant nutrient supply and reduced physical activity. As vascular disease represents the main etiology for death, and accounts for a great percent of morbidity in diabetic patients, poorly controlled diabetes is considered to be a vascular disease. Hyperglycemia may influence hemorheological parameters through enhanced advanced glycation end product (AGE) formation: reducing sugars may react non-enzymatically with the amino groups in proteins or lipids, ultimately leading to the formation of stable covalent adducts. AGE can bind to biological membranes in a nonspecific manner. They also induce specific cellular responses, including the release of pro-fibrogenic and pro-inflammatory cytokines by interacting with RAGE (receptor for AGE), a cellular surface receptor that binds AGE-modified proteins with high affinity. Several other receptors and cell surface molecules that are capable of binding AGE-modified proteins have been identified recently. The consequence of AGE-RAGE interaction is the generation of reactive oxygen species (ROS), partially caused by the activation of NADPH oxidase. AGE-RAGE interaction induces the expression of vascular endothelial growth factor (VEGF) in endothelial cells. AGEs also inhibit prostacyclin production and stimulate plasminogen activator inhibitor-1 (PAI-1) synthesis by endothelial cells. AGE thus stimulate the growth of microvascular endothelial cells, leading to angiogenesis on one hand and to a prothrombotic state on the other hand [59, 60, 61, 62]. Obesity is both a cause and a consequence of type-2 diabetes mellitus. The adipose tissue synthesizes and releases numerous bioactive substances, several of which are known to have an effect on blood rheology. PAI-1 and interleukin-6, important regulators of plasma fibrinogen levels, are also produced in significant amounts by adipocytes. Low-grade inflammation is present in adiposity as reflected by elevated levels of C-reactive protein, interleukin-6 and tumor necrosis factor alpha in obese individuals. The low-grade, chronic inflammatory state may contribute to insulin resistance and endothelial dysfunction. Adipose tissue also releases leptin, a cytokine

408


that circulates at high concentrations in the plasma of obese, insulin resistant patients. This hormone has been recently shown to stimulate proliferation and migration of myocytes in the vessel wall, and thus has been suggested to be a possible pathogenic factor for atherosclerosis. Positive correlations have been reported between leptin and both plasma viscosity and RBC partial disaggregation threshold measured by laser backscattering [63]. Accelerated atherosclerosis in type-2 diabetic patients is multifactorial and includes very complex interactions, including hyperglycemia, hyperinsulinemia, hyperproinsulinemia, dyslipidemia, oxidative stress, accelerated aging and platelet hyperreactivity as well as alterations in coagulation and fibrinolysis and changes in hemorheological parameters. The initial lesion of atherosclerosis is manifested by changes of endothelial cell function. Endothelial dysfunction has been shown in patients with type-2 diabetes as well as in type-1 diabetic patients early in the course of the disease, especially when microalbuminuria is present. Moreover, endothelial dysfunction has been shown to be present in insulin resistant patients with impaired glucose tolerance (IGT) and in individuals with former gestational diabetes [64]. There is a growing body of evidence showing that poly(ADP) ribose polymerase (PARP) activation has an important role in the pathophysiology of endothelial dysfunction in diabetes. In animal models, destruction of pancreatic islet cells with streptozotocin leads to PARP activation and endothelial dysfunction. Studies on wild-type and PARP deficient mice showed that elevated blood glucose levels represented a very strong stimulus for PARP activation. These studies led to the conclusion that PARP activation due to hyperglycemia is influenced by the genesis of superoxide species; O2 free radicals may be produced in endothelial cells exposed to hyperglycemia. 5.1. Hemorheological Parameters Glycemic control seems to be a major factor for determining the hemorheological consequences of diabetes. Type-1 diabetic patients with poor glycemic control exhibit increased plasma and whole blood viscosity when compared to normoglycemic individuals; blood viscosity is also negatively correlated with insulin sensitivity. Treatment of insulin resistance by exercise training specifically improves plasma viscosity due to the close association between plasma viscosity and insulin sensitivity. Positive associations have also been found between parameters of glycemic control (HbA1C, fructoseamine), fibrinogen levels and red blood cell aggregation; fibrinogen levels are also closely correlated to insulin resistance. A single hyperglycemic spike increases red blood cell aggregation in both type-1 and type-2 diabetic patients, and alters fibrinogen concentration and activity in type-1 diabetic patients. Several studies have reported that insulin improves hemorheological abnormalities in diabetes, and when studied in vitro, incubation of red cells obtained from diabetic patients with insulin results in improved cellular deformability as measured by micropore filtration. Interestingly, this observation could not be reproduced when utilizing washed red blood cells. These ex vivo results suggest a direct effect of insulin on red cell membrane fluidity; although beneficial hemorheological effects related to insulin treatment in vivo may be indirectly mediated by metabolic improvements. Nevertheless, it now seems that insulin affects red cell rheology via direct effects on the membrane, including alterations of the lipid membrane bilayer composition and microviscosity and changes in membrane Na+/K+ ATP-ase function. It is interesting to


409

note that supra-physiological levels of insulin can have adverse effects on red cell deformability, with very high in vitro levels decreasing red blood cell deformability. Both type-1 and type-2 diabetes are associated with diabetic thrombocytopathy, a condition related to increased platelet adhesiveness and aggregability. Enhanced platelet aggregation is present in diabetes early in the course of the disease, well before the development of diabetic vascular lesions. Several biochemical abnormalities play a role in diabetic platelet hyperreactivity: 1) membrane fluidity is reduced due to changes in the lipid composition of the membrane or glycation of membrane proteins; 2) increased intracellular Ca2+ levels and decreased magnesium concentrations reduce membrane fluidity and increase platelet adhesiveness; 3) platelets from diabetic individuals produce less NO and prostacyclin, agents which normally promote endothelium-mediated vasodilation and inhibit platelet-endothelium interactions; 4) increased platelet arachidonic acid metabolism leads to enhanced TXA2 production, a possible underlying cause for platelet hyperreactivity; 5) platelets from diabetic subjects contain reduced levels of antioxidant molecules which might also contribute to their hyperaggregability. The binding of fibrinogen to the GP IIb-IIIa receptor is increased in diabetic patients, and they also have a higher ratio of platelets expressing activation-dependent adhesion molecules such as activated GP IIb-IIIa, lysosomal Gp53, thrombospondin and P-selectin; plasma fibrinogen levels are also increased in both types of diabetes. Platelets may interact with glycosylated low density lipoproteins, von Willebrand factor or immune complexes, and platelet turnover may be shortened in diabetes, thereby contributing to the observation that antiplatelet agents such as aspirin and clopidogrel have a diminished effect in these patients. Chronically poor metabolic control is associated with increased platelet activation and aggregation. Hyperglycemia has been described to result in enhanced platelet aggregation, with reduced sensitivity to aspirin observed in type-2 diabetic patients with poor metabolic control. Soluble P-selectin is a widely accepted marker for platelet activation: elevated plasma levels have been proved to be significantly associated with diabetes mellitus [65, 66, 67]. Polymorphonuclear leucocytes derived from diabetic patients were found to be more rigid as measured by filtration techniques, suggesting an activated stage of these cells. Leucocytes are larger and much more rigid than erythrocytes, thus they can strongly influence microvascular blood flow. Moreover, polymorphonuclear leucocytes (PMN), the largest fraction of leucocytes, are capable of causing microvascular damage by the release of proteases and toxic oxygen radicals. Comparing PMN of normoglycemic and diabetic patients, the rigidity of leukocytes obtained from individuals with diabetes is significantly elevated both in their basal stage and following their activation with the bacterial polypeptide fMLP. Diabetic and hypertensive individuals have less deformable PMN than patients with diabetes and normal blood pressure [68, 69, 70].

6. Pulmonary Diseases Contributing authors: Laszlo CZOPF1, Lajos BOGAR2, 1First Department of Medicine and 2Department of Anaesthesia and Intensive Care, University of Pecs, Hungary The pulmonary circulation normally represents a system with low-resistance (i.e., approximately one-tenth of that of the systemic circulation), low-pressure (i.e., mean

410


pulmonary artery pressure about 15 mmHg, total pressure drop along the pulmonary circuit about 10 mmHg) and high-flow (i.e., 6 l/min). Although the normal pulmonary vascular resistance is extraordinarily small, it has a remarkable facility for becoming even smaller as the vascular pressure rises. Recruitment of previously unperfused vessels and caliber distension are the main mechanisms responsible for the resistance decrease. Lung volume is another important determinant of pulmonary vascular resistance: at large lung volumes, stretching of the capillaries and increased alveolar pressure leads to decreased resistance, while the decreased diameter of extra-alveolar vessels at low lung volumes leads to higher resistance and elevated critical opening pressure of the vascular bed [71]. Although passive factors dominate vascular resistance of the pulmonary circulation under normal conditions, an active vasoconstriction occurs upon reduced alveolar oxygen tension or at low blood pH [72]. Obstructive and restrictive pulmonary disorders are frequently accompanied by hypoxia, resulting in secondary polycythemia via an erythropoietin mediated compensatory process. Erythropoietin is a 46 kDa glycoprotein produced by the tubular capillaries of the kidneys (80%) and by the liver (20%). Polycythemia increases oxygen transport capacity to a certain level, but above a critical hematocrit of 48-50%, elevated whole blood viscosity may lead to microcirculatory problems thus causing parenchymal organ dysfunction. Elevated whole blood viscosity due to increased hematocrit levels increases pulmonary arterial pressure and leads to impaired alveolar gas exchange. Microcirculatory failure in the general circulation decreases cardiac output and arteriovenous oxygen difference, thus diminishing oxygen delivery to the tissues. Physiological polycythemia occurs at high altitudes, where the decreased O2 partial pressure of inhaled air results in hypoxia and thus erythropoietin overproduction leading to increased erythrocyte mass and elevated hematocrit [72]. Clinical signs and symptoms of these secondary polycythemias are similar to those of polycythemia vera: dizziness, fainting, vertigo, nausea, weakness, sweating, vision disturbances, tinnitus, paresthesias and numbness of acral areas (e.g., fingers, toes, nose, ears), cyanosis, dysregulation of motor or sensory functions, thromboembolic complications, transient cerebral ischemia, stroke and myocardial ischemia resulting in acute coronary syndromes. Therapeutic options include treatment of the underlying pulmonary or vascular pathology, correction of the congenital defect and/or optimizing whole blood viscosity via reducing hematocrit using isovolemic hemodilution. Oxygen treatment may lead to decreased erythropoietin levels and may prevent further deterioration of rheological status, thereby disrupting the vicious circle of hypoxia polycythemia - hyperviscosity - hypoxia [1,73]. In vitro rheological studies have clearly established that whole blood viscosity and yield stress are elevated in patients with erythrocytosis. However, a number of factors ensure that these patients, under physiological conditions, do not show the clinical features observed in other hyperviscosity states. These factors include red cell axial migration in flowing blood and plug flow in the largest vessels. In addition, a small increase in vessel diameter leads to decreased resistance and large increases of blood flow, and generally high blood flows produce the lowest blood viscosity values. The elevated hemoglobin levels and the increased oxygen-carrying capacity at high hematocrit values compensate for the tissue hypoxia. In non-hypoxemic erythrocytosis (e.g., polycythemia vera, idiopathic and apparent erythrocytosis), there is an increased incidence of vascular occlusions in untreated patients. The reasons for this observation include reduced peripheral blood flow, increased platelet-vessel wall interactions, and hyperviscosity due to abnormally low in vivo blood flow in pathological conditions. In


411

the erythrocytosis of hypoxemic lung disease and its associated hypoxemia, pulmonary vasoconstriction particularly enhances susceptibility to hyperviscosity. Moreover, the vasoconstriction caused by hypoxemia prevents the normal adaptive changes of increased vessel diameter [73]. Sickle cell disease, a hemoglobinopathy affecting the beta units of hemoglobin, is characterized by paroxysmal microcirculatory failure and capillary clogging due to abnormally shaped and more rigid erythrocytes when sickle hemoglobin (HbS) is in the deoxygenated state. One of the most serious complications of sickle cell disease, acute chest syndrome characterized by new pulmonary infiltrate and respiratory symptoms, represents the leading cause of mortality in this patient population; individuals with repeated episodes are at high risk for the development of chronic lung disease and pulmonary hypertension [74].

7. Hematological and Oncological Disorders Contributing author: Tamas ALEXY, First Department of Medicine, University of Pecs, Hungary and Department of Physiology and Biophysics, Keck School of Medicine, Los Angeles, CA, USA 7.1. Sickle Cell Disease Sickle cell disease (SCD) is the most striking example of a disorder with diminished red blood cell (RBC) deformability causing marked rheological alterations in the microcirculation and ischemic manifestations. It is a common genetic disorder, with about 1 in 600 African and African American newborns affected. The abnormality is due to an amino acid substitution (i.e., valine for glutamic acid) in the sixth position of the ȕ-globin chain leading to the formation of hemoglobin S (HbSS). This hemoglobin variant has a strong tendency to polymerize under low oxygen tensions, thereby causing the RBC to become characteristically distorted, elongated and rigid with a high internal viscosity. From the hemorheological aspect, the low hematocrit (HCT) observed in SCD patients compensates, at least in part, for the rigidity of sickle cells, leaving the patient with a close to normal blood viscosity when tested under reduced oxygen tensions. In contrast, when evaluated at a normal HCT (e.g., 40 to 45%), the viscosity of deoxygenated HbSS blood is greatly above control values for normal HbAA blood, with the difference evident but less prominent under oxygenated conditions. The latter observation is explained by the presence of irreversibly sickled cells (ISC) compromising 1-25% of circulating erythrocytes. ISC are the most rheologically and metabolically compromised cells and do not reverse to their normal, discoid shape upon oxygenation. Their formation results from repeated, alternate exposure to oxygenated and deoxygenated conditions in the circulation and leads to several cellular abnormalities: 1) polymerization of HbSS even at high oxygen tensions; 2) membrane damage with the loss of cytoplasmic potassium and water leading to an increase in mean corpuscular hemoglobin concentration (MCHC) and cell density; 3) intracellular calcium accumulation; 4) attachment of HbSS to the inner leaflet of the cell membrane decreasing its fluidity. Thus the ISC subpopulation has a constant contribution to the impaired blood rheology in SCD, independent of oxygen tension. The number of circulating ISC has been shown to be closely correlated with

412


the extent of hemolysis, but not to correlate well with the clinical severity of the disease [75]. Upon decreasing the oxygen tension in vitro, HbSS RBC suspensions show progressively reduced filterability through 5 μm pores owing to the formation of reversibly sickled cells (RSC). These erythrocytes possess a normal discoid shape when oxygenated, with up to 50% of them undergoing sickling at oxygen tensions similar to that in postcapillary venules. Due to the remarkable rigidity of RSC upon deoxygenation, whole blood viscosity (WBV) increases significantly, with up to a 10-fold greater viscosity than that measured at equal shear rates and HCT under oxygenated circumstances. A pathophysiological hallmark of SCD is the episodic occurrence of vasoocclusive events, termed crises, often triggered by hypoxia, low temperatures, acidosis, dehydration or infections. The ensuing RBC sickling promotes stasis, capillary obstruction and tissue ischemia, clinically manifested as episodic painful infarctions often superimposed on silently developing organ failure syndromes. On the basis of hemorheological alterations, it is not yet clear whether the crises are initiated by ISC obstructing microvessels and thus causing ischemia, or by the greatly increased impedance of blood flow in the postcapillary venules promoting capillary stasis and sickling. Additional cellular abnormalities have been identified that might contribute to the pathophysiology of SCD and the abrupt manifestation of sickle crisis. These findings include abnormal RBC membrane phosphorylation, displacement of procoagulant molecules from the inner to the outer leaflet of the membrane bilayer, increased free radical generation by HbSS RBC, abnormal thiol oxidation and enhanced adhesion of ISC to reticuloendothelial cells and the vascular endothelium. Modest polymorphonuclear leukocytosis, thrombocytosis and elevated serum IgA concentrations are also frequent findings. It is not yet known which of these abnormalities are of rheological significance, but additional studies are underway to explore their exact pathophysiological role in SCD. No rational therapy exists to date to cure sickle cell disease. Despite the recent advances in the area of stem cell transplantation and gene therapy, these therapeutic interventions are not likely to be widely available in the near future. Hydroxyurea has been the only drug shown to reduce the need for repeated blood transfusions, the frequency of painful episodes and the mortality rate in SCD [76]. The development of a reliable and simple high throughput screening method will thus be necessary in order to identify novel drug candidates with potent antisickling properties. Transfusion therapy has been shown to reduce the incidence of severe SCD complications: guidelines typically define a target post-transfusion HCT of 35% to enhance the oxygen carrying capacity of blood while avoiding hyperviscosity, yet the benefits of transfusion may vary depending on local flow conditions and organ specific hemodynamics. 7.2. Thalassemia Syndromes Thalassemias are frequent among the Mediterranean populations, in the Middle East and throughout South Asia. Defects in the synthesis of Į- or ȕ-globin leads to defective hemoglobin formation and the precipitation of the unpaired chains as tetramers in the cytoplasm of RBC precursors. In homozygous ȕ thalassemia, the inclusion bodies formed by the excess Į chains tend to associate with the inner leaflet of the reticulocyte cytoplasmic membrane. The ensuing increase in cellular rigidity may be responsible


413

for the local, intramedullary destruction of reticulocytes, leading to the characteristic ineffective erythropoiesis observed in ȕ thalassemia. The anemia is further escalated as the rigid erythrocytes are destroyed prematurely when trying to pass through the microvasculature of an enlarged spleen. Owing to the transfusions administered every 6 to 8 weeks, hemorheological experiments are difficult to perform in subjects with homozygous ȕ thalassemia. The limited number of studies available on non-splenectomized ȕ thalassemia individuals confirm decreased cellular deformability as indexed by reduced filterability of RBC through 5 μm polycarbonate filter pores [77]. Since splenectomy is often performed in these patients after five years of age, studies in these individuals might uncover previously sequestered erythrocyte subpopulations with the most severe rheological abnormalities, including those with extreme membrane rigidity. In addition to the inclusion body formation, a variety of cellular alterations have been described in ȕ thalassemia that may further impair rheological parameters; most of these are membrane-related abnormalities and are direct consequences of excess free radical formation and damage. In heterozygous ȕ thalassemia, the excess production of Į-globin chains is less than in homozygotes, with red cell cytoplasmic concentration of Į-chains further reduced by the proteolytic enzymes of erythroid precursors. Despite normal WBV values, sensitive filtration techniques have revealed impaired RBC deformability in this patient population as well. In Į thalassemia patients, the defective synthesis of Į-globin chains leads to an excess of free Ȗ chains in the fetus and an excessive ȕ chain production in adults. Soluble ȕ4 tetramers are formed that do not precipitate to a significant degree in erythroid precursors, and thus ineffective erythropoiesis is not a feature of Į thalassemia. However, in some individuals, ȕ4 tetramers may precipitate and form inclusion bodies within aging erythrocytes. Although hemorheological abnormalities of Į thalassemia have not been fully studied to date, it seems likely that they have alterations similar to those observed in the mild variant of ȕ thalassemia. 7.3. Polycythemia Polycythemia is defined as the increase of HCT or packed cell volume (PCV) above the normal range; in adults tested at sea level, HCT values exceeding 52% in males and 48% in females indicate this disorder. Polycythemic patients can be separated into two groups based on clinical and laboratory evaluations: 1) those with absolute polycythemia; 2) those with relative polycythemia. Subjects with absolute polycythemia may be further categorized based upon the pathophysiology of the increased HCT: x Polycythemia rubra vera (i.e., primary proliferative polycythemia) is a malignant myeloproliferative disease characterized by the uncontrolled, clonal proliferation of a pluripotent stem cell in the bone marrow. Despite the suppressed erythropoietin synthesis, RBC counts may exceed 8 to 9x109/l, leading to HCT values above 75%. White blood cell (WBC) and platelet counts are also elevated in most patients. x Secondary polycythemias in which renal erythropoietin synthesis is increased, in turn stimulating RBC production and maturation to enhance the oxygen delivery of blood. Secondary polycythemia serves an important physiological adaptation mechanism in individuals living at

414


high altitudes but is also found under pathological conditions associated with low oxygen saturation of arterial blood, including hypoxic lung disease, congenital cyanotic heart disease and the presence of certain hemoglobin variants with high oxygen affinity. Secondary polycythemia may also develop in individuals with erythropoietin secreting malignancies (e.g., renal, liver, cerebral or uterine tumors) despite normal oxygen saturation of arterial blood. x Idiopathic erythrocytosis is diagnosed in patients with absolute polycythemia without detectable underlying cause. Cerebrovascular events have been shown to be more frequent in this population. In relative polycythemia an increase of HCT is the typical laboratory finding. However, in this situation reduced plasma volume rather than increased red cell mass is the underlying cause. Relative polycythemia develops as a consequence of unopposed intravascular fluid loss, including that which occurs in burns, dehydration, hypoalbuminemia and stress polycythemia. It is important to note that in vitro studies have established that HCT is the single most important determinant of WBV, with an exponential relationship between the two parameters (see chapter II.3.a). Thus, when tested at a given shear rate, a one percent change in HCT within the physiological range has a lesser impact on blood viscosity than the same change at higher, pathological values. This finding is due to the crowded conditions in high hematocrit blood, and hence the inability of RBC to reach optimal deformation for streamlining and minimal flow resistance; the unfavorable rheological effects of high HCT become more prominent when viscosity is tested at lower rates of shear. Although WBC and platelet counts are usually elevated in polycythemia vera, these cells have a negligible effect on WBV owing to their insignificant volume fraction compared to that of RBC. No pathological alteration of erythrocyte aggregation, deformability or plasma viscosity (PV) has been identified in polycythemic patients. Thus, the principal consequence of erythrocytosis is the HCTdependent elevation of WBV which is primarily responsible for clinical abnormalities such as: 1) cerebrovascular manifestations (e.g., headache, vertigo, tinnitus, chorea and blurred vision) in the early stages of the disease that can progress to a cerebrovascular accident over time owing to the inverse relationship between cerebral blood flow and venous HCT; 2) coronary artery disease since elevated PCV leads to a reduced width of the plasmatic zone at the vessel wall favoring platelet activation and aggregation; 3) superficial thrombophlebitis and deep vein thrombosis complicated with pulmonary embolism. Venous thrombi are most common in the lower limbs where stasis is frequent, shear rates are lower and hyperviscosity is more pronounced. However, involvement of the deep abdominal veins has also been reported. In addition to the unfavorable rheological profile, qualitative platelet defects can also be present in patients with primary proliferative polycythemia (e.g., the absence of PGD2 receptors, increased expression of Fc receptors); these abnormalities tend to further promote thrombus formation. Reduction of HCT and thus WBV is the primary therapeutic aim in polycythemia, with phlebotomy and/or hemodilution representing the most effective remedies; small doses of chemotherapeutic drugs may also be beneficial for selected patients. Maintaining PCV in the low-normal range has been shown to be the most efficient in preventing vaso-occlusive episodes.


415

7.4. Iron Deficiency Anemia Iron deficiency anemia is one of the most common chronic diseases worldwide. It may develop as a consequence of inadequate dietary iron intake, malabsorption, chronic blood loss or the increased utilization of iron during pregnancy and lactation. Hemoglobin synthesis is reduced leading to microcytic anemia. Several studies using viscometry of RBC suspensions have shown reduced erythrocyte deformability in these patients owing to the unfavorable change in the normal RBC surface to volume ratio [78]. However, some investigators have suggested that the observed alterations of blood viscosity are caused by the inappropriate determination of packed cell volume, and that no significant difference exists between the deformability of microcytic and normocytic erythrocytes if the HCT is adjusted correctly prior to testing [79]. Clearly, further studies are required to better understand hemorheological alterations in iron deficiency anemia. 7.5. Megaloblastic Anemias Megaloblastic anemias represent a group of disorders caused by the deficiency of folate or vitamin B12. Underlying pathophysiological mechanisms include inadequate dietary intake or increased requirements of these nutritional supplements, intrinsic factor deficiency and malabsorption. Macrocytic erythrocytes are formed (MCV>95 fl) owing to the impaired DNA synthesis along with unaffected hemoglobin production in RBC progenitors. Intravascular hemolysis and ineffective erythropoiesis with normal reticulocyte counts are prominent features of megaloblastic anemia. Hemorheological abnormalities are consistent with the severity of the disease and are primarily attributable to the increased mean corpuscular hemoglobin (MCH) that leads to RBC rigidity. Erythrocyte membrane protein pattern abnormalities have also been described and shown to increase membrane viscosity; these alterations are additive and lead to reduced RBC deformability. 7.6. Hyperleukocytic Leukemias The concept of hyperleukocytic syndrome (HLS) was proposed by Lichtman and Rowe [80] and describes the circulatory and rheological abnormalities observed in the late, accelerated phase of certain hyperleukocytic leukemias. It is most common in children with chronic myeloid leukemia (CML), usually with WBC counts exceeding 300x109/L. HLS is less frequent in acute myelocytic leukemia (AML) or acute lymphoblastic leukemia (ALL) and is rare in patients with chronic lymphoblastic leukemia (CLL). WBC can have a significant influence on WBV depending on the relative volume they occupy in the suspension (i.e., the leukocrit). Blood viscosity increases exponentially with increasing leukocrit, with the relationship even more pronounced and steep if the observed leukocrit exceeds 15%; due to plasma trapping during centrifugation a correction is required (e.g., an observed leukocrit of 15% equals a true hematocrit of 10%) [81]. The basis for this correction is the finding that due to their nucleus and viscous cytoplasm, normal WBC are approximately 10-fold less deformable than RBC; leukemic blasts are even more rigid than normal WBC. It is important to note that, despite the extremely high WBC count, WBV measured ex vivo is rarely elevated in leukemias and, when noted, is usually confined to CML. This is explained by the anemia invariably associated with hyperleukocytic leukemias, thereby

416


serving as a hemorheological compensatory mechanism to reduce viscosity and maintain adequate oxygen delivery. However, when compared to controls with equal HCT or hemoglobin levels, WBV is markedly elevated in patients with CML. As detailed in III.2, cellular deformability is an important determinant of microcirculatory blood flow. In HLS, the dynamic distribution of RBC within the microvessels can be significantly altered by the large number of rigid, leukemic leukocytes that are virtually unable to pass through capillaries with a diameter below 5 μm. As a consequence, columns of WBC aggregates without fibrin strands are formed in these nutritive microvessels, leading to vessel occlusion and the development of potentially fatal clinical manifestations including: 1) fever and tachycardia with no evidence of infection; 2) cardio-respiratory symptoms, most frequently tachypnea, dyspnea, hypoxia and myocardial ischemia; 3) neurological features including tinnitus, dizziness, ataxia, confusion, delirium and coma; 4) ocular symptoms. Emergency leukopheresis rapidly alleviates clinical symptoms via reducing WBC count and WBV. Centrifugal methods are preferred over filtration techniques because blast cells tend to occlude the narrow filter pores. Cytotoxic drugs (e.g., hydroxyurea) administered concomitantly with leukopheresis are obligatory to maintain low leukocyte counts. Allopurinol and certain supportive measures may be necessary to minimize the manifestation and extent of “cell lysis syndrome” initiated by chemotherapy. However, simple blood transfusions aiming to correct anemia should be avoided to prevent the associated increase of blood viscosity that might further deteriorate microcirculatory blood flow; exchange transfusions may be utilized in children if leukopheresis is not available or difficult to perform. 7.7. Plasma Hyperviscosity Syndromes Plasma hyperviscosity syndrome (HVS) was first described by Fahey and co-workers as a group of clinical symptoms associated with high concentrations of plasma proteins capable of increasing PV [82]. Occasionally, these proteins might be of polyclonal origin (e.g., Sjögren’s syndrome, rheumatoid arthritis), but most often HVS is associated with paraproteinemias. Paraproteinemias are a group of disorders characterized by the excess presence of an abnormal immunoglobulin secreted by malignant B-lymphoid cells of monoclonal origin. These immunoglobulins exclusively bear a kappa or a lambda light chain and can be categorized as IgG, IgA, IgM, IgD or IgE via immunological methods; free light or heavy chain fragments can be secreted as well. Although multiple myeloma (MM) and Waldenström’s macroglobulinemia (WM) are the most common disorders complicated with paraproteinemia, it may also be present in certain lymphomas and CLL. HVS is caused by the excess presence of these abnormal, high molecular weight immunoglobulins with extraordinary characteristics. An approximately exponential relationship has been described between the paraprotein concentration and PV. However, the slope of this association is modified by the intrinsic viscosity of the particular paraprotein which is determined by the shape and size of the molecule with IgM>IgA>IgG, as well as the natural tendency of IgA and IgG3 to polymerize into high molecular weight complexes. Measured at 25°C, PV frequently exceeds 2.5 mPa.s in this patient population but values as high as 5.0 mPa.s have been reported; the reference range for PV is 1.40 - 1.80 mPa.s at 25°C. Besides their effect on PV, abnormal proteins also promote intense RBC aggregation. In the clinical setting, this rheological phenomenon might be demonstrated by the extremely elevated erythrocyte


417

sedimentation rate (ESR), high aggregation indices and also via microscopic examination of the peripheral blood smear: large RBC aggregates form rapidly and are resistant to dispersion even at high rates of shear. When compared to controls at a given HCT, blood viscosity is markedly elevated in subjects with paraproteinemia, especially at low rates of shear. However, anemia is a prominent finding in HVS: an inverse correlation has been shown between HCT and PV values. Owing to this compensatory drop in HCT, microcirculatory perfusion and the oxygen transport efficacy of blood are much less compromised, thereby reducing the incidence of symptomatic hyperviscosity. Not surprisingly, WBV measured at low shear rates has been documented to be a better predictor of clinical complications than PV alone [83]. HVS produces clinical symptoms in approximately 50% of patients with WM, where monoclonal plasma cells produce IgM. The incidence is 8-10% on average in MM and depends on the immunoglobulin isotype: due to its tendency to form dimers, trimers and tetramers, secreted IgA is associated with greater risk than IgG. Clinical features of HVS include: 1) weakness, dyspnea, anorexia and fatigue; 2) ocular abnormalities; 3) bleeding diathesis owing to the interaction of paraproteins with coagulation factors or to hypoxic microvessel damage; 4) neurological symptoms related to cerebral ischemia; 5) high output cardiac failure owing to the anemia and hypervolemia since a linear relationship exists between PV and plasma volume. The therapeutic goal in HVS is to reduce PV and hypervolemia below the symptomatic threshold of the individual. Plasmapheresis is the fastest and most effective method to reduce paraprotein concentration, and is followed by an exponential decrease of PV and WBV. Especially in patients with WM, single plasma exchange might be of value because it significantly reduces IgM concentration. Following such procedures, blood transfusions may be instituted to improve the oxygen carrying capacity of blood. Parallel with the primary intervention, disease-specific chemotherapy is initiated to control the proliferative activity of malignant clones. In addition, all therapeutic regimens aggravating hyperviscosity (e.g., diuretics) should be avoided or administered with care. 7.8. Malignancy Hemorheological abnormalities are frequently documented in malignancies (e.g., breast, lung, ovarian, pulmonary cancers) and include modest to marked increases of plasma fibrinogen concentration, PV, RBC aggregation and a significant reduction of erythrocyte deformability. These alterations promote hyperviscosity that is compensated for, at least in part, by the marked anemia observed in patients with malignancy. Although the majority of the rheological changes noted above are attributable to non-specific mechanisms, the scope of these abnormalities, as reviewed by von Tempelhoff, et al., has been suggested to correlate well with the clinical stage of several cancers, their prognosis and the associated thrombotic risk [84]. Hypoxic microenvironments, characterized by low oxygen tension, low pH, glucose deficiency, high interstitial fluid pressure and increased lactate concentration, are characteristic features of neoplasms. The local synthesis of vascular endothelial growth factor (VEGF) is enhanced in response to tissue hypoxia, thus promoting neovascularisation. However, a chaotic network of dilated, torturous microvessels is formed with an apparently abnormal anatomical architecture favoring fluid extravasation and the perpetuation of hypoxia. In addition, the normal axial flow of RBC (i.e., the Fåhraeus-Lindqvist Effect), WBC and circulating tumor cells is

418


disrupted in the tumor area, displacing the cells towards the vessel wall. As a consequence, HCT and WBV increase, with blood flow decreasing in the local microcirculation leading to periods of transient stasis in the capillaries. While one would assume that these hemorheological alterations are consequences of tumor spread, Dintenfass has suggested that they might promote the arrest of viable tumor cells in the microvessels and favor metastasis formation [85], contribute to thrombus formation and limit the efficacy of chemotherapeutic drugs. Thus, favorable modification of hemorheological parameters (e.g., reduction of elevated WBV and PV) may represent a supplemental therapeutic regimen to chemotherapy in cancer treatment.

8. Ophthalmology Contributing authors: Judit VEKASI1, Zsolt MARTON2, 1Department of Ophthalmology and 2First Department of Medicine, University of Pecs, Hungary In the human body, two major factors regulate blood flow to various organ systems and tissues: 1) perfusion pressure; 2) peripheral vascular resistance as determined by blood vessel tone and the hemorheological properties of blood. Owing to the significant differences in nutritional requirements, these factors play an important role in regulating local blood supply. Since retinal circulation has only a limited capacity to adapt to perfusion pressure changes, hemorheological parameters are of crucial importance in the maintenance of normal retinal blood flow. Thus a thorough hemorheological examination may help to clarify the etiology of various ophthalmologic diseases, especially retinopathy. 8.1. Hypertensive Retinopathy Hypertension is one of the most frequent cardiovascular diseases leading to severe ophthalmic complications, including macro- and microangiopathies. Hypertensive retinopathy is one of the most common devastating outcomes of hypertension, leading to visual loss in the most severe cases. Posterior segment changes are widely observed and represent a severe complication in hypertension. The gross appearance of the fundus in hypertensive retinopathy is primarily determined by the severity of hypertension and the condition of the retinal arterioles. Mild angiopathy (i.e., arterial narrowing, minor caliber changes) is characteristic for moderately elevated blood pressures, while in more advanced stages, retinal alterations become more evident and vessel tortuousity may appear; ensuing capillary endothelial damage may lead to hemorrhages. Cotton-wool (C-W) spots represent focal infarcts of the retinal nerve fiber layer caused by the occlusion of precapillary arterioles. In addition, microaneurysms usually form at the periphery of the C-W spots owing to the weakness of the capillary walls. As a result of increased vessel permeability, chronic retinal edema develops leading to the deposition of hard exudates around the fovea in the Henle layer, assuming a macular star configuration. Optic nerve head swelling is the hallmark of malignant hypertension: initially the periphery of the papilla becomes blurred initially, and later, parallel with expansions of the edema, the prominence becomes increasingly evident [86, 87].


419

Several studies have shown that hemorheological parameters (e.g., hematocrit, fibrinogen level, plasma and whole blood viscosity) of hypertensive patients are significantly abnormal compared to healthy control individuals [88, 89]. In addition, the severity of the fundus alteration was found to be proportional to the hemorheological disturbances. Thus in patients with hypertensive retinopathy and abnormal hemodynamics, it is important to gauge the negative effects of abnormal rheological factors via evaluating perfusion [37, 90]. 8.2. Diabetic Retinopathy Diabetes mellitus develops as the consequence of absolute or relative insulin deficiency, leading to severe metabolic disturbances, and has emerged as one of the most common causes of blindness affecting young adults throughout the world. Retinopathy is the consequence of severe diabetic microangiopathy and appears 10-15 years following the initial disease development, even in patients with well-controlled diabetes mellitus. While the exact pathophysiological mechanisms for diabetic retinopathy have not yet been not fully elucidated, the role of factors other than metabolic disturbances is suspected. Several studies have confirmed the presence of hemorheological disturbances in patients with diabetes mellitus. Elevated levels of fibrinogen and various plasma proteins lead to increased plasma viscosity, RBC aggregation and whole blood viscosity. Ditzel, in 1955, was the first to observe that the properties of blood flow are altered in patients with diabetes mellitus; he documented increased RBC aggregation in the conjunctival vessels of these subjects. In 1961, Cogan showed elevated plasma viscosity and five years later Skovborg reported elevated total blood viscosity in diabetic individuals. In 1972, Kwaan noted enhanced platelet aggregation that correlated well with the extent of diabetic complications, while Volger described a decrease in RBC filterability in 1974. Later, in 1991, Isogai reported a close association between the frequency of retinopathy, age and the duration of diabetic metabolic disturbances; a close correlation between the severity of microangiopathy and the pathological increase in blood viscoelasticity was also described. In 1996, Giasanti documented an increase of RBC aggregation and whole blood viscosity parallel with a decrease in arteriole diameter and conjunctival blood flow velocity in diabetic patients. With greater RBC aggregation, not only is the extent of erythrocyte aggregation enhanced creating more and larger aggregates, but also aggregate formation is faster and higher shear rates are required to disrupt RBC aggregates [91, 92, 93]. While the presence of hemorheological alterations has been extensively documented in patients with diabetes mellitus, it is still unclear whether these changes have a primary role in the development of microvascular complications or are secondary changes further complicating the pre-existing angiopathy [61]. At the initial stage of metabolic abnormalities, the local retinal blood supply deteriorates and reversible vein dilation develops. This, in turn, leads to a compensatory arteriolar vasoconstriction owing to circulatory autoregulation. If the autoregulatory process becomes ineffective, irreversible microvascular changes develop with degenerative vessel wall changes (e.g., thickening of the capillary basal membranes, endothelial dysfunction). Hemorheological changes extend vessel wall damage, further increasing flow resistance, and accelerated fibrinogen and platelet turnover with increased platelet aggregability promote microthrombi development. Owing to the increased RBC

420


aggregation, blood flow is primarily affected in areas of low shear such as capillaries and post-capillary venules, further promoting stasis and microthrombi development in these vessels. The ensuing circulatory deficit leads to tissue hypoxia and the proliferative diabetic changes of the retina. Hemorheological alterations thus cause microcirculatory disturbances and ischemia of the retina that, in turn, promote new capillary formation leading to retinopathy. However, it is important to note that the retinal alterations detailed above are not specific for diabetes, and similar changes can be observed in various hematological and vascular diseases that cause insufficient retinal perfusion (i.e., myeloma multiplex, sickle cell anemia, retinal occlusion). 8.3. Retinal Vascular Occlusions It is now well accepted that alterations of blood components, pathologic changes of the vessel wall, and stasis all favor vascular occlusion. One of the most dramatic diseases in ophthalmology is central retinal vein thrombosis; it is the second most common vascular lesion following diabetic retinopathy. Depending on the severity and manifestation of the disease, it can assume central retinal vein occlusion (CRVO) or branch retinal vein occlusion (BRVO). Systemic and local factors (i.e., physical occlusion at the level of the lamina cribrosa, hemodynamical and hemorheological alterations) are equally important in the development of vascular blockage; the effects of these mechanisms are additive in leading to central retinal vein thrombosis. BRVO develops, almost inevitably, at arteriovenous intersections: histological studies show that retinal arterioles and their corresponding veins share a common adventitial sheath and thus a thickened arteriole in the anterior position might compress the vein passing beneath. The ensuing secondary alterations include venous endothelial cell loss, thrombus formation and occlusion. Similarly, the central retinal vein and artery share a common adventitial sheath posterior to the lamina cribrosa, and thus atherosclerotic thickening of the artery may compress the vein and precipitate CRVO. Diseases affecting the venular walls (e.g., phlebitis) may also precipitate vascular blockage [94]. Many authors have confirmed the presence of hemorheological disturbances in patients with retinal vein occlusion; these alterations not only promote disease development but might also contribute to its progression. The presence of altered rheological factors in occlusive retinal vessel disease was first published by McGrath, et al. Later, in 1976, the results of Ring, et al. confirmed that whole blood viscosity, plasma viscosity and fibrinogen levels are elevated in all patients with venous occlusion. These authors also found that higher blood viscosity values increase the probability of capillary venous stasis that, in concert with elevated plasma fibrinogen and globulin concentrations, promotes red blood cell aggregation. Thus elevated blood viscosity confers an increased risk for retinal ischemia and neovascularisation that may lead to further complications. Data from Peduzzi in the early 1980s documented increased whole blood and plasma viscosity, fibrinogen concentration and hematocrit in patients with retinal venous occlusion. In 1990, Piermarocchi described the pathophysiological importance of hemorheological factors in retinal vascular diseases. During the past decades, several studies have confirmed the results referenced above, and have further demonstrated the pathogenetic role of both macro- and microrheological parameters in the development of retinal vascular occlusions [95, 96].


421

8.4. Anterior Ischemic Optic Neuropathy (AION) AION is caused by the occlusion of the posterior ciliary arteries in the retrolaminar cribrosa area where the optic nerve capillaries are relatively less abundant. Thus acute visual loss owing to ischemic optic neuropathy should be associated with ophthalmoscopic evidence of disk edema within 24 hours after the onset of visual symptoms (e.g., decrease in color vision, visual fields defects that are often altitudinal). These neuropathies can be divided into two groups: arteritic and the non-arteritic (idiopathic) AION. The basis for the arteritic-type ischemic optic neuropathy is giant cell arteritis developing as a result of the blockage of the posterior ciliary arteries. To establish a diagnosis, a biopsy of the temporal artery and certain lab tests (e.g., erythrocyte sedimentation rate) are necessary. There is an approximately 40% probability that, within a few years, the contralateral eye will be affected as well. Atherosclerotic plaques developed in vessels supplying the papillae, mechanical factors, and rheological abnormalities are responsible for the development of idiopathic or non-arteritic AION. It is important to note that the exact cause of the disease can not always be established; 40% of patients have hypertension and 20% have diabetes mellitus, and thus the existence of small vessel circulatory disorders as well as abnormal hemorheological parameters seem probable. The pathophysiological importance of hemorheological factors is further emphasized by observations that the normalization of hemorheological parameters can help to reduce the frequency of AION developing in the contralateral eye from 40% to 5% [97, 98]. 8.5. Primary Open-Angle Glaucoma Glaucoma is an optic neuropathy with a characteristic appearance of the optic disc and a specific pattern of visual field defects associated frequently with increased intraocular pressure (IOP). Elevated IOP in primary open-angle glaucoma is caused by the increased resistance to aqueous outflow in the trabecular meshwork; retinal ganglion cell death occurs predominantly via apoptosis. Multiple factors influence the rate of cell damage, with current opinions favoring either ischemic or mechanical etiologies. Some authors emphasize that in open-angle glaucoma, there is also a primary degenerative disorder of the optic nerve due to mechanical distortion of the lamina cribrosa, to vascular insufficiency, or to both. This view is supported by the observation that loss of function sometimes continues to progress even after the intraocular pressure has been normalized by medical therapy or surgery. Also, patients with systemic disease (e.g., diabetes, arteriosclerosis) are more likely than others to suffer optic nerve damage as a result of ocular hypertension. Based upon current knowledge, the cause of the visual field defects of patients with simplex glaucoma who are being treated to maintain normal ocular pressure is partly vascular in origin. The blood supply of the papillae is determined by the diameter of the supplying small vessels, the perfusion pressure prevailing in them, vascular resistance and the autoregulation capacity of the vessels. As the autoregulation becomes insufficient, damage of the papillae and the subsequent visual field defects may be due to altered blood rheology. Literature data and our own results indicate that among the patients with primary open-angle glaucoma, hemorheological factors are in the pathological range. In “low-tension” glaucoma subjects, as well as among patients with simplex glaucoma suffering from visual field defects in spite of the well-adjusted ocular pressure, levels of fibrinogen, hematocrit and Į-2 globulin are higher than

422


among patients with non-progressive visual field simplex glaucoma. These factors emphasize the importance of regular control of patients with simplex glaucoma, and indicate that in addition to measuring ocular pressure, a visual field examination every six months, fundus examination, and frequent observation of hemorheological factors supplemented with the adequate treatment will markedly help in slowing the progression of the disorder [99, 100].

9. Exercise and Hemorheology Contributing authors: Zsolt MARTON, Kalman TOTH, First Department of Medicine, University of Pecs, Hungary It is well known that regular exercise may lead to a reduction in cardiovascular risk and mortality. Generally, the protective effect of physical exercise is associated with metabolic improvements (e.g., lipid parameters, carbohydrate status), but it is important to note that in addition to metabolic changes, exercise has multiple effects on body functions including rheological factors. The hemorheological effects of exercise are triphasic: 1) short-term effects (i.e., hyperviscosity mostly due to fluid shifts and alterations of erythrocyte rigidity and aggregation); 2) middle-term effects (i.e., the reversal of acute effects due to plasma volume expansion termed auto-hemodilution; 3) long-term effects that further improve blood fluidity and parallel the classical traininginduced hormonal and metabolic alterations. 9.1. Short-Term Effect of Exercise on Hemorheology Exercise, regardless of its intensity, induces acute changes in the rheologic properties of blood [20, 101, 102, 103, 104]. The most common hemorheological alteration that can be observed during physical training is a significant increase in whole blood viscosity. Whole blood viscosity is determined by hematocrit, plasma viscosity and the rheological behavior of erythrocytes. Although all of these determinants have been reported to be acutely changed during exercise, the most, constant finding is that both hematocrit and plasma viscosity are increased. The increase in these parameters is a quite complex process and includes different mechanisms: redistribution of red cells in the vascular system; splenic contraction that increases the number of circulating erythrocytes; increasing plasma protein concentration presumably due to protein from the lymphatics; loss of water in sweat for thermoregulation, and entrapment of water into muscle cells [102, 105]. Several mechanisms for the structural changes of blood cells during exercise have been postulated. One of the main causes of these changes is the increase of lactate concentration in blood; its accumulation into blood reflects the level of exercise and is a consequence of anaerobic processing of carbohydrates. Lactate shrinks red cells and decreases their flexibility. Correlations between lactate levels during exercise and erythrocyte rigidity have been reported, supporting the concept that lactate, at least when it rises above a 4 mmol/l threshold, impairs red cell deformability during exercise. Even a moderate lactate increase during low-intensity exercise periods is associated with a transient increase in erythrocyte rigidity [102, 106]. Another mechanism that may impair blood rheology, especially erythrocyte deformability, during exercise is oxidative stress induced by increased free radical


423

production. The marked increase in oxygen utilization that occurs during exercise results in production of free radicals by several sources, including the mitochondria and the white cells. Transient tissue hypoxia due to a rapid acceleration of oxygen consumption in exercising muscles and to inadequate oxygen supply at the pulmonary level may also lead to free radical formation. As a consequence of oxidative damages, the malondialdehyde content of the red cell membrane, and hence lipid peroxidation, has been shown to increase during prolonged exercise. Permeability to anions also increases in the cell membrane after exercise [107]. The fluid status of the body also has a major influence on erythrocyte rheology during exercise: red blood cells contain more than 60% water, mostly associated with the cytoplasm. The percentage of this associated or “bound” water in the cells determines their deformability, oxygen transport, and other properties. During acute exercise the total amount of water does not change or even slightly decreases in erythrocytes, but the percentage of free water increases, resulting in the decrease of associated water leading to a decrease of deformability [102, 108]. There are also acute changes in the extent and strength of erythrocyte aggregation during exercise, both of which cause an increase in viscosity at low shear rates. Elevation of plasma proteins, especially fibrinogen, during exercise results in an augmented aggregation process. Alterations in red blood cell aggregation can also be associated with metabolic changes such as increased lactate and oxygen free radical levels. Traumatic damage of red cells during running due to compression in the plantar circulatory bed has been demonstrated and may lead to changes in red blood cell deformability and aggregation [109]. Compared to red blood cells, less is known about the role of white blood cells in the changes of rheological properties of blood during exercise. Granulocyte activation may play a role in the hemorheological effects of strenuous exercise. Previous studies support the role of an inflammatory response to strenuous exercise as a factor contributing to hemorheological deterioration, with the most likely cause of the inflammatory response after strenuous exercise being generalized muscle damage. A gradual change in white blood cell count has been demonstrated during exercise, which at least partially may be a consequence of an inflammatory response. In addition to this mechanism, de-margination plays an important role in the increase of white blood cell count. About half of the leukocytes are in a marginal pool loosely adherent to the vascular endothelium or trapped in the microcirculation, and with exercise, these cells are released into the circulating pool. While after a short training period there is no significant alteration of count, increasing the time and the level of exercise results in considerable changes: after a marathon race white blood cell count can be elevated almost 5-fold. In addition to their increased number, white cell activation can occur; such activated granulocytes are more rigid and may directly influence microcirculatory hemodynamics and block microcirculatory circuits. Further, activated leucocytes can interact with red cells through oxygen free radicals and inflammatory cytokines and decrease erythrocyte deformability. Given their increased number, rigidity and their effect on red cells, leukocytes may be among the important factors that contribute to post-exercise circulatory problems [102, 110]. Platelet count rises after a few minutes of vigorous exercise, usually by 30%. The main source of platelets is considered to be the spleen but similar responses can be seen in subjects with no spleen, suggesting other possibilities including increased release from the bone marrow, from megakaryocytes in the lungs, or de-margination due to increased flow resulting in cells previously at the periphery of the blood stream being

424


brought into the more-rapidly flowing blood. Changes in platelet function have also been reported, with short-term exercise leading to increased aggregation that may be explicable on the basis of a rapid rise in catecholamines [101]. Most hormones affect blood rheology, some of them also being involved in exercise physiology. Glucagon, norepinephrine, leukotriene B4 and leukotriene C4 decrease red cell deformability, while atrial natriuretic peptide increases it. 9.2. Middle-Term Effect of Exercise on Hemorheology Following exercise rheological parameters exhibit changes that are opposite in direction from those seen during the exercise period, with these changes mostly due to plasma volume expansion. During the hours following exercise, there is an increase in plasma volume, termed "autohemodilution”, which reverses the acute hyperviscosity described above since this autohemodilution lowers hematocrit and plasma viscosity. Other potential mechanisms for the post-exercise lowering of hematocrit have been hypothesized, including intramuscular red cell damage or traumatic destruction of erythrocytes in the foot circulation during running. Note that autohemodilution not only increases plasma volume, but also decreases plasma protein concentration and protein composition: a significant reduction of plasma fibrinogen can be observed [111]. Post-exercise decreases of lactate level have an effect on the structural properties of red blood cells, thereby restoring their deformability. However, the mechanical impairment of erythrocytes continues long after blood lactate concentrations return to normal values, suggesting the involvement of other factors in exercise-induced rigidity of cells. One of these may be the increased oxidative stress during exercise, with repair of free radical induced structural damages requiring longer periods. Restoring the ratio of “free” to “bound” water via an increase in "bound" water is also linked to the increase of erythrocyte deformability [102, 108]. Significant decrements in red blood cell aggregation in both autologous plasma and in a standard dextran aggregating medium can also be observed for up to 8-12 hours. Decreased aggregation in plasma may result from alterations of both plasma composition (e.g., decreased fibrinogen concentration related to plasma volume expansion) and red blood cell properties. The red blood cell aggregation measurements in a standard aggregating medium suggest that cellular alterations also play role in decreased aggregation after heavy exercise. Impaired red blood cell deformability would be expected to reduce red blood cell aggregability since rigid cells exhibit lower levels of aggregation. Alternatively, red blood cell surface properties that play a role in aggregation may be altered by oxidation and/or proteolytic enzymes released by activated granulocytes during exercise [101, 102]. Changes in white blood cell count observed after exercise are partially due to plasma volume expansion, but margination of leukocytes and reduced inflammatory processes also play an important role in the decrease of white blood cell count; platelet count and activation also decreases after exercise. 9.3. Long-Term Effect of Exercise on Hemorheology Both cross-sectional and longitudinal studies suggest that the rheological properties of blood correlate with regular training; sportsmen have a lower blood viscosity and regular training improves hemorheological parameters. The hemorheological


425

adaptations to long-term exercise occur with aerobic or endurance exercise such as middle and long distance running, cycling, swimming and jogging. The improvement of blood rheology parallels the correction of metabolic and body composition disturbances, and close correlations between insulin resistance, body fat and abnormalities of blood rheology have been demonstrated [112]. A fall in hemoglobin and hematocrit of about 5% occurs after regular exercise in non-trained individuals. The most important factor for the reduced hemoglobin is an increase in plasma volume of 5-10% with no change or even a small increase in red cell mass. This phenomenon is referred to as athletes' anemia, but this term is a misnomer as the total body hemoglobin is normal or increased. Decreased plasma viscosity can also be observed with regular training, with both volume expansion and a marked reduction in plasma fibrinogen likely explanations; the decreased plasma fibrinogen levels may explain the decreased erythrocyte aggregation found in trained individuals. Regular training increases red cell filterability and prevents its decrease during training. Studies of long-term changes due to exercise have shown reduced fibrinogen levels and increased resting fibrinolysis, and there have been many reports of reduced platelet activity and aggregability with training [101, 102]. In addition to the above-mentioned processes, nuclear magnetic resonance measurements show that exercise training increases the water content of red cells with a proportional decrease in "free" water and increase in "bound" water, while red cell volume does not change. This percentage of "bound" water in the red cells seems to be linked to cell deformability, oxygen transport and physical capacity. 9.4. Overtraining The overtraining syndrome is frequent in athletes who are training for a competition beyond the body's capabilities; inappropriate training regimens may actually decrease performance. Overtraining correlates with blood viscosity, leading to increased plasma viscosity and hematocrit. Therefore, the early signs of overtraining in elite sportsmen are associated with a hemorheological pattern that suggests some degree of reversal of the fitness-associated “autohemodilution”. In fact, there are earlier reports of mild dehydration after exhaustive endurance training, with the increased hematocrit proposed to be due to an excess plasma water loss [102]. 9.5. Effect of Hemorheological Parameters on Aerobic Capacity An association between blood fluidity and physical condition has been reported previously. Red cell flexibility correlated with adductor isometric strength, and correlations of blood fluidity with aerobic working capacity, time of endurance until exhaustion and blood lactate response have been demonstrated. In addition to hemodynamic parameters, improvements in hemorheological parameters are likely to aid better performance. Theoretically, increased blood fluidity may improve oxygen delivery to muscles during exercise in well-trained individuals. However, if during exercise the autoregulatory reserve of the coronary circulation is fully utilized (i.e., resistance vessels are maximally dilated due to metabolic vasodilators), any increment in blood viscosity may reduce blood flow [102].

426


10. Miscellaneous Contributing author: Lajos BOGAR, Department of Anaesthesia and Intensive Care, University of Pecs, Hungary 10.1. Septic Shock Sepsis is defined as systemic inflammatory response to infection, and without resolution, sepsis progresses to so-called severe sepsis characterized by a single organ system dysfunction. Septic shock is defined as arterial hypotension with systolic blood pressure less than 90 mmHg or mean arterial pressure below 70 mm Hg despite adequate intravenous fluid resuscitation, together with evidence of perfusion abnormalities (e.g. lactic acidosis, oliguria, mental impairment). Septic progression is associated with profound hemodynamic and microcirculatory alterations characterized by mal-distribution of blood flow within organs and tissues, loss of perfused capillary density, subsequent increase of blood flow heterogeneity and capillary water leakage [113]. Bacterial products such as lipopolysaccharide (LPS), and reactive endogenous cytokines like tumor necrosis factor-alpha (TNF-alpha) and interleukin (IL) molecules , elicit the innate, non-specific inflammatory response in monocytes and polymorphonuclear (PMN) leukocytes, thrombocytes, endothelial and red blood cells (RBC), all of which have been implicated in the development of microvascular damage in sepsis [114]. Due to cell damage caused by high shear forces of the hyperkinetic septic circulation and oxygen radicals, RBC are removed from the blood at a higher rate than in a normal subject. This process results in a slowly progressing anemia, and since hematocrit is the main determinant of whole blood viscosity, the low hematocrit of septic patients prevents the development of hyperviscosity syndrome and macrocirculatory hemodynamic pathology. However, all cellular components of blood undergo profound changes that compromise microcirculatory flow conditions in septic patients. 10.1.1. RBC Deformability, Aggregation and Adhesion Several studies in septic animals and patients have demonstrated decreased RBC deformability [113, 115, 116], with cell membrane lipid peroxidation as the result of oxidant load usually considered the leading factor for cell damage [115, 117]. Elevated cytosolic calcium content may also contribute to RBC mechanical impairment and decreased deformability [118]. Most likely, oxidative stress originates from activated white blood cells (WBC) causing lipid peroxidation and spectrin-hemoglobin crosslinking [115, 117]. It is generally accepted that nitric oxide (NO) production is enhanced by the inducible NO synthase in endothelial cells. The process is triggered by LPS, TNF-alpha and interferon-gamma leading to relaxation of vascular smooth muscle cells and subsequent vasodilation, increased bulk blood flow to the tissues characterized by arterial hypotension, and maldistribution of oxygen supply due to shunt openings. Studies have shown that LPS promotes adhesion of human RBC to endothelial cells in vitro, probably by decreasing sialic acid ligands on the outer surface of cell membrane [116]. Elevated level of fibrinogen as part of the acute phase reaction contributes to enhanced RBC aggregation [119], and since RBC aggregation is


427

increased at low flow states, it seems very likely that this increase adversely affects the microcirculation in septic patients and thus contributes to the development of tissue hypoxia. Lactate accumulation in tissues and blood is one of the direct consequences of cellular oxygen debt in flow states such as septic shock, and serum lactate concentration correlates significantly with the severity and mortality risk of shock states. Pronounced lactate acidosis impairs blood flow properties by increasing whole blood viscosity at both low and high shear rates [120]. Acidosis results in significant RBC swelling which contributes to decreased cellular deformability and consequent elevation of bulk viscosity measured at high shear rates. Recently, Piagnerelli and co-workers have demonstrated that RBC are more spherical in patients with sepsis than in healthy volunteers [121]. 10.1.2. WBC Rheology Severe acute infections cause an increase in fibrinogen concentration and trigger a cytokine cascade involving primary activation of innate cellular immune response, with monocytes and PMN playing a central role in this process. It has been postulated that activation of these cells decreases effective capillary blood flow and eventually contributes to microvascular occlusion [122]; several studies have demonstrated decreased PMN deformability in severe sepsis and/or septic shock [114, 122, 123, 124]. In the microcirculation, the dominant location of RBC aggregation can be found in the postcapillary venules that are characterized by low shear conditions which promote WBC margination. Increased viscosity in postcapillary venules tends to increase the capillary pressure resulting in an increased fluid loss [125]. Increased PMN adhesion and aggregation, and decreased cellular deformability, lead to sequestration of PMN especially in lung capillaries [114]. Accumulation of leukocytes in the pulmonary microcirculation is regarded as one of the key elements in the development of acute lung injury and, in its more advanced stage, acute respiratory distress syndrome. 10.2. Pregnancy Cardiac output and circulating blood volume increase about 30% by 12 weeks of pregnancy and this level is maintained until term; total peripheral resistance falls by about the same extent. Whole blood viscosity decreases in the first trimester without further changes in the second and a slight increase in the third trimester. This is mainly due to slowly progressing anemia and partially counteracting continuous increase of fibrinogen level lasting to the beginning of the third trimester. Preeclampsia, the most frequent severe complication of pregnancy, is an endothelial disorder leading to increased capillary permeability, platelet microvascular thrombosis and increased vascular tone resulting in elevated systolic and/or diastolic blood pressure (140 and/or 90 mmHg, respectively) and proteinuria of at least 0.3 grams per 24 hours [126]. In preeclampsia, utero-placental hypoperfusion has been detected with subsequent RBC membrane damage causing decreased cell deformability [127]. Elevation of RBC aggregation has also been shown in preeclampsia due to either conformational changes of the membrane occurring during arterial hypertension or a redistribution of the ionic changes on the two surfaces of the membrane [128]. However, recent publications on RBC deformability and aggregation in preeclampsia

428


seem to be in conflict: 1) Some authors report decreased deformability while others did not observe any significant differences between preeclamptic and normal pregnancies [129]; 2) Similarly, RBC aggregation has been found to be elevated [128] or unaltered [129] in preeclamptic pregnancies.

11. References 1 2 3 4 5 6 7 8 9

10 11 12 13 14

15

16

17

18 19

20

21

S. Chien, J. Dormandy, E. Ernst and A. Matrai, Clinical Hemorheology, Martinus Nijhoff Publishers, Dordrecht, 1987. J.F. Viles Gonzales, S.X. Anand, C. Valdiviezo, U. Zafar, R. Hutter, J. Sanz, T. Rius, M. Poon, V. Fuster and J.J. Badimon, Update in atherothrombotic disease, Mt. Sinai J. Med. 71 (2004), 197-208. S. Lehoux, Y. Castier and A. Tedgui, Molecular mechanisms of the vascular responses to haemodynamic forces, J. Int. Med. 259 (2006), 381-392. W. Koenig and E. Ernst, The possible role of hemorheology in atherothrombogenesis, Atherosclerosis 94 (1992), 93-107. E. Ernst and W. Koenig, Hemorheology, thrombogenesis, and atherosclerosis, Semin. Thromb. Hemost. 19 (1993), 99-103. M.R. Boisseau, Roles of mechanical blood forces in vascular diseases. A clinical overview, Clin. Hemorheol. Microcirc. 33 (2005), 201-207. K.R. Kensey, The mechanistic relationship between hemorheological characteristics and cardiovascular disease, Curr. Med. Res. Opin. 19 (2003), 587-596. P.A. Vanderlaan, C.A. Reardon and G.S. Getz, Site specificity of atherosclerosis site-selective responses to atherosclerotic modulators, Arterioscler. Thromb. Vasc. Biol. 24 (2004), 12-22. M. Galbusera, C. Zoja, R. Donadelli, S. Paris, M. Morigi, A. Benigni, M. Figliuzzi, G. Remuzzi and A. Remuzzi, Fluid shear stress modulates von Willebrand factor release from human vascular endothelium, Blood 90 (1997), 1558-1564. A. Vaya, M. Martinez, J. Dalmau, M. Labios and J. Aznar, Hemorheological profile in patients with cardiovascular risk factors, Haemostasis 26 (1996), 1666-1670. G. Kesmarky, G. Feher, K. Koltai, B. Horvath and K. Toth, Viscosity, hemostasis and inflammation in atherosclerotic heart diseases, Clin. Hemorheol. Microcirc. 35 (2006), 67-73. W.B. Kannel, R.B D'Agostino and A.J. Belanger, Fibrinogen, cigarette smoking, and risk of cardiovascular disease: insights from the Framingham study, Am. Heart J. 113 (1987), 1006-1010. C. Carter, D. McGee, D. Reed, K. Yano and G. Stemmermann, Hematocrit and the risk of coronary heart disease: The Honolulu heart program, Am. Heart J. 105 (1983), 674-679. G.D.O. Lowe, W.C.S. Smith, H.D. Tunstall-Pedoe, I.K. Crombie, S.E. Lennie, J. Anderson and J.C. Barbenel, Cardiovascular risk and haemorheology - results from the Scottish heart health study and the MONICA project, Glasgow, Clin. Hemorheol. 8 (1988), 517-524. G.D.O. Lowe, F.G.R. Fowkes, J. Dawes, P.T. Donnan, S.E. Lennie and E. Housley, Blood viscosity, fibrinogen, and activation of coagulation and leukocytes in peripheral arterial disease and the normal population in the Edinburgh artery study, Circulation 87 (1993), 1915-1920. P.M. Sweetnam, H.F. Thomas, J.W.G. Yarnell, A.D. Beswick, I.A. Baker and P.C. Elwood, Fibrinogen, viscosity and the 10-year incidence of ischemic heart disease: The Caerphilly and Speedwell studies, Eur. Heart J. 17 (1996), 1814-1820. J.W. Yarnell, I.A. Baker, P.M. Sweetnam, D. Bainton, J.R. O-Brien, P.J. Whitehead and P.C. Elwood, Fibrinogen, viscosity, and white blood cell count are major risk factors for ischemic heart disease. The Caerphilly and Speedwell collaborative heart disease studies, Circulation 83 (1991), 836-844. J. Ma, C.H. Hennekens, P.M. Ridker and M.J Stampfer, A prospective study of fibrinogen and risk of myocardial infarction in the physicians’ health study, J. Am. Coll. Cardiol. 33 (1999), 1347-1352. A.J. Lee, P.I. Mowbray, G.D.O. Lowe, A. Rumley, F.G.R. Fowkes and P.L. Allan, Blood viscosity and elevated carotid intima-media thickness in men and women. The Edinburgh artery study, Circulation 97 (1998), 1467-1473. K. Toth, T. Habon, I. Horvath, B. Mezey, I. Juricskay and Gy. Mozsik, Hemorheological and hemodynamical parameters in patients with ischemic heart disease at rest and at peak exercise, Clin. Hemorheol. 14 (1994), 329-338. F.J. Neumann, H. Tillmanns, P. Roebruck, R. Zimmermann, H.M. Maupt and W. Kubler, Haemorheological abnormalities in unstable angina pectoris: a relation independent of risk factor profile and angiographic severity, Br. Heart J. 62 (1989), 421-428.

K. Toth et al. / Clinical Significance of Hemorheological Alterations 22

23

24

25

26 27 28 29 30 31

32 33 34 35 36

37 38 39 40 41 42 43 44 45

46 47

429

F.J. Neumann, H.A. Katus, E. Hoberg, P. Roebruck, M. Braun, H.M. Haupt, H. Tillmanns and W. Kubler, Increased plasma viscosity and erythrocyte aggregation - indicators of an unfavorable clinical outcome in patients with unstable angina pectoris, Br. Heart J. 66 (1991), 425-430. R. Junker, J. Heinrich, H. Ulbrich, H. Schulte, R. Schönfeld, E. Köhler and G. Assmann, Relationship between plasma viscosity and the severity of coronary heart disease, Arterioscler. Thromb. Vasc. Biol. 18 (1998), 870-875. G.D.O Lowe, M.M. Drummond, A.R. Lorimer, I. Hutton, C.D. Hutton, C.R.M. Prentice and J.C. Barbenel, Relation between extent of coronary heart disease and blood viscosity, Br. Med. J. 280 (1980), 673-674. C. Rainer, D.T. Kawanishi, P.A.N. Chandraratna, R.M. Bauersachs, C.L. Reid, S.H. Rahimtoola and H.J. Meiselman, Changes in blood rheology in patients with stable angina pectoris as a result of coronary artery disease, Circulation 76 (1987), 15-20. G. Kesmarky, K. Toth, L. Habon, G. Vajda and I. Juricskay, Hemorheological parameters in coronary artery disease, Clin. Hemorheol. Microcirc. 18 (1998), 245-251. E. Ernst, Plasma fibrinogen - an independent cardiovascular risk factor, J. Int. Med. 227 (1990), 365372. G. Montalescot, A. Ankri, E. Vicaut, G. Drobinski, Y. Grosgogeat and D. Thomas, Fibrinogen after coronary angioplasty as a risk factor for restenosis, Circulation 92 (1995), 31-38. G. Montalescot, J.P. Collet, R. Choussat and D. Thomas, Fibrinogen as a risk factor for coronary heart disease, Eur. Heart J. 19 (1998), H11-H17. A. Natali, A. L’Abbate and E. Ferrannini, Erythrocyte sedimentation rate, coronary atherosclerosis, and cardiac mortality, Eur. Heart J. 24 (2003), 639-648. S.J. Rim, H. Leong-Poi, J.R. Lindner, K. Wei, N.G. Fisher and S. Kaul, Decrease in coronary blood flow reserve during hyperlipidemia is secondary to an increase in blood viscosity, Circulation 104 (2001), 2704-2709. J. Chan, S.F. Knutsen, G.G. Blix, J.W. Lee and G.E. Fraser, Water, other fluids, and fatal coronary heart disease - the adventist health study, Am. J. Epidemiol. 155 (2002) 827-833. L. Dintenfass, Blood rheology in pathogenesis of the coronary heart diseases, Am. Heart J. 77 (1969), 139-147. S. Chien, Hemorheology in clinical medicine, Clin. Hemorheol. 2 (1982), 137-142. K. Toth, B. Mezey, I. Juricskay and T. Javor, Hemorheological changes during the hospital phase of acute myocardial infarction: sex differences? Med. Sci. Res. 17 (1989), 841-844. Zs. Marton, B. Horvath, T Alexy, G. Kesmarky, L. Czopf, T. Habon, L. Kovacs, E. Papp, R. Halmosi, B. Mezey, E. Roth and K. Toth, Follow-up of hemorheological parameters and platelet aggregation in patients with acute coronary syndromes, Clin. Hemorheol. Microcirc. 29 (2003), 81-94. J. Vekasi, K. Toth, I. Juricskay and B. Kovacs, The role of hemorheological factors in hypertensive retinopathy, Clin. Hemorheol. 16 (1996), 187-192. C.G. Gustavsson, S.U. Persson, H. Larsson and S. Persson, Changed blood rheology in patients with idiopathic dilated cardiomyopathy, Angiology 45 (1994), 107-111. J.H. Wood and D.B. Jr. Kee, Hemorheology of the cerebral circulation in stroke, Stroke 16 (1985), 765-722. J. Grotta, R. Ackerman, J. Correia, G. Fallick and J. Chang, Whole blood viscosity parameters and cerebral blood flow, Stroke 13 (1982), 296-301. J. Syrjänen, A.M. Teppo, V.V. Valtonen, M. Iivanainen and C.P.J. Maury, Acute phase response in cerebral infarction, J. Clin. Pathol. 42 (1989), 63-68. E. Ernst, A. Matrai and M. Marshall, Blood rheology in patients with transient ischemic attacks, Stroke 19 (1988), 634-636. M. Fisher and H.J. Meiselman, Hemorheological factors in cerebral ischemia, Stroke 22 (1991), 11641169. N. Tanahashi, F. Gotoh, M. Tomita, T. Shinohara, Y. Terayama, B. Mihara, K. Ohta and M. Nara, Enhanced erythrocyte aggregability in occlusive cerebrovascular disease, Stroke 20 (1989), 1202-1207. L. Szapary, B. Horvath, Zs. Marton, T. Alexy, N. Demeter, M. Szots, A. Klabuzai, G. Kesmarky, I. Juricskay, V. Gaal, J. Czopf and K. Toth, Hemorheological disturbances in patients with chronic cerebrovascular diseases, Clin. Hemorheol. Microcirc. 31 (2004), 1-9. E. Ernst and A. Matrai, Intermittent claudication, exercise, and blood rheology, Circulation 76 (1987), 1110-1114. W. Koenig, E. Ernst and A. Matrai, Blood rheology associated with cardiovascular risk factors and chronic cardiovascular diseases: results of an epidemiologic cross-sectional study, Angiology 39 (1988), 986-995.

430 48

49

50

51

52 53 54 55 56 57 58 59 60

61 62

63 64

65

66 67 68 69 70

71 72 73 74 75

K. Toth et al. / Clinical Significance of Hemorheological Alterations F.G. Fowkes, E. Housley, E.H. Cawood, C.C. Macintyre, C.V. Ruckley and R.J. Prescott, Edinburgh Artery Study: prevalence of asymptomatic and symptomatic peripheral arterial disease in the general population, Int. J. Epidemiol. 20 (1991), 384-392. K.R. Woodburn, G.D. Lowe, A. Rumley, J. Love and J.G. Pollock, Relation of haemostatic, fibrinolytic, and rheological variables to the angiographic extent of peripheral arterial occlusive disease, Int. Angiol. 14 (1995), 346-352. F.B. Smith, G.D. Lowe, A.J. Lee, A. Rumley, G.C. Leng and F.G. Fowkes, Smoking, hemorheologic factors, and progression of peripheral arterial disease in patients with claudication, J. Vasc. Surg. 28 (1998), 129-135. I. Tzoulaki, G.D. Murray, A.J. Lee, A. Rumley, G.D. Lowe and F.G. Fowkes, Inflammatory, haemostatic, and rheological markers for incident peripheral arterial disease: Edinburgh Artery Study, Eur. Heart J. 28 (2007), 354-362. C. Koksal, M. Ercan and A.K. Bozkurt, Hemorheological variables in critical limb ischemia, Int. Angiol. 21 (2002), 355-359. S.A. Jones, The effects of arterial stenosis on red cells, J. Heart Valve Dis. 7 (1998), 396-399. P.A. VanderLaan, C.A. Reardon and G.S. Getz, Site specificity of atherosclerosis: site-selective responses to atherosclerotic modulators, Arterioscler. Thromb. Vasc. Biol. 24 (2004), 12-22. A.L. Herrick, Pathogenesis of Raynaud’s phenomenon, Rheumatology 44 (2005), 587-596. S. Sanjeev and J.H. Scurr, Chronic venous insufficiency, In: Textbook of Vascular Medicine, J.E. Tooke and G.D.O. Lowe, Eds., Arnold, 1996, pp. 580-595. C. Le Devehat, M.R. Boisseau, M. Vimuex, G. Bondoux and A. Bertrand, Hemorheological factors in the pathophysiology of venous disease, Clin. Hemorheol. 9 (1989), 861-870. P.D. Coleridge Smith, P. Thomas, J.H. Scurr and J. Dormandy, Causes of venous ulceration: a new hypothesis, Br. Med. J. 296 (1988), 1726-1727. C. Le Devehat, M. Vimeux and T. Khodabandehlou, Blood rheology in patients with diabetes mellitus, Clin. Hemorheol. Microcirc. 30 (2004), 297-300. A.C. Mellinghoff, A.J. Reininger, L.J. Wurzinger, R. Landgraf and K.D. Hepp, Influence of glycemic control on viscosity and density of plasma and whole blood in type-1 diabetic patients, Diabetes Res. Clin. Pract. 33 (1996), 75-82. J. Vekasi, Zs. Marton, G. Kesmarky, A. Cser, R. Russai and B. Horvath, Hemorheological alterations in patients with diabetic retinopathy, Clin. Hemorheol. Microcirc. 24 (2001), 59-64. H. Yonekura, Y. Yamamoto, S. Sakurai, T. Watanabe and H. Yamamoto, Roles of the receptor for advanced glycation endproducts in diabetes-induced vascular injury, J. Pharmacol. Sci. 97 (2005) 305311. J.F. Brun, Hormones, metabolism and body composition as major determinants of blood rheology: potential pathophysiological meaning, Clin. Hemorheol. Microcirc. 26 (2002), 63-79. J.F. Brun, I. Aloulou and E. Varlet-Marie, Hemorheological aspects of the metabolic syndrome: markers of insulin resistance, obesity or hyperinsulinemia? Clin. Hemorheol. Microcirc. 30 (2004), 203-209. K. Koltai, G. Feher, G. Kesmarky, Zs. Keszthelyi, L. Czopf and K. Toth, The effects of blood glucose levels on hemorheological parameters, platelet aggregation and activation in oral glucose tolerance tests, Clin. Hemorheol. Microcirc. 35 (2006), 517-525. J.A. Colwell and R.W. Nesto, The platelet in diabetes: focus on prevention of ischemic events, Diabetes Care 26 (2003), 2181-2188. J.G. Mabley and F.G. Soriano, Role of nitrosative stress and poly(ADP-ribose) polymerase activation in diabetic vascular dysfunction, Curr. Vasc. Pharmacol. 3 (2005), 247-252. E. Ernst and A. Matrai, Altered red and white blood cell rheology in type II diabetes, Diabetes 35 (1986), 1412-1415. Z. Pecsvarady, T.C. Fischer, C.H. Darwin, A. Fabok, T.S. Maqueda, M. Saad and H.J. Meiselman, Decreased polymorphonuclear leucocyte deformability in NIDDM, Diabetes Care 17 (1994), 57-63. G. Caimi, L. Lo Presti, M. Montana, C. Carollo, G. Amodeo, A. Romano and B. Canino, Polymorphonuclear leukocyte: rheology, metabolism and integrin pattern in vascular atherosclerotic disease and in type 2 diabetes mellitus, Clin. Hemorheol. Microcirc. 30 (2004), 229-235. J.B. West, Respiratory Physiology 3rd edition, Williams&Wilkins, Baltimore, 1984, pp. 8-9 and 31-48. J.B. West, Pulmonary Pathophysiology 2nd edition, Williams&Wilkins, Baltimore, 1984, pp. 112-133. T.C. Pearson, Hemorheology in the erythrocytoses, Mt Sinai J. Med. 68 (2001), 182-191. S. Claster and E. Vichinsky, Acute chest syndrome in sickle cell disease: Pathophysiology and management, J. Intens. Care Med. 15 (2000), 159-166. J. Stuart, Sickle cell disease and vascular occlusion - rheological aspects, Clin. Hemorheol. 4 (1984), 193-207.

K. Toth et al. / Clinical Significance of Hemorheological Alterations 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90

91 92 93 94 95 96 97 98 99 100 101 102

103 104

105

431

S. Charache, M. Terrin and R. Moore, Effect of hydroxyurea on the frequency of painful crisis in sickle cell disease, N. Engl. J. Med. 322 (1995), 1317-1322. W. Tillmann and W. Schröter, Rheological properties of erythrocytes in heterozygous and homozygous ȕ-thalassemia, Br. J. Hematol. 43 (1979), 401-411. R. Yip, N. Mohandas, M.R. Clark, S. Jain, S.B. Shohet and P.R. Dallman, Red cell membrane stiffness in iron deficiency, Blood 62 (1983), 99-106. T.C. Pearson, D.L. Guthrie, N.G. Slater and G. Wetherley-Mein, Method of PCV measurement and the effect on whole blood viscosity in polycythemia, Br. J. Hematol. 52 (1982), 166-169. M.A. Lichtman and J.M. Rowe, Hyperleukocytic leukemia: rheological, clinical and therapeutic considerations, Blood 60 (1982), 279-283. M.A. Lichtman, The relationship of excessive white cell accumulation to vascular insufficiency in patients with leukemia, Kroc Found. Ser. 16 (1984), 295-306. J.L. Fahey, W.F. Barth and A. Solomon, Serum hyperviscosity syndrome, JAMA 192 (1965), 464-467. M.A. McGrath and R. Penny, Paraproteinemia, blood hyperviscosity and clinical manifestations, J. Clin. Invest. 58 (1976), 1155-1162. G.F. von Tempelhoff, L. Heilmann, G. Hommel and K. Pollow, Impact of rheological variables in cancer, Semin. Thromb. Hemost. 29 (2003), 499-513. L. Dintenfass, Hemorheology of cancer metastases: an example of malignant melanoma. survival times and abnormality of blood viscosity factors, Clin. Hemorheol. 2 (1982), 259-271. S.S. Hayreh, G.E. Servais and P.S. Virdi, Fundus lesions in malignant hypertension. III. Arterial blood pressure, biochemical, and fundus changes, Ophthalmology 93 (1986), 45-59. S.S. Hayreh, G.E. Servais and P.S. Virdi, Fundus lesions in malignant hypertension. IV. Focal intraretinal periarteriolar transudates, Ophthalmology 93 (1986), 60-73. L. Dintenfass, Viscosity factors in hypertensive and cardiovascular diseases, Cardiovasc. Med. 2 (1977), 337-353. E. Longhini, R. Agosti, P. Cherubini, A. Clivati, P. Farini and L. Marazzini, Hemorheology in hypertension, Clin. Hemorheol. 6 (1986), 567-576. K. Toth, G. Kesmarky, J. Vekasi, J. Nemes, L. Czopf, P. Kapronczay, R. Halmosi and I. Juricskay, Hemorheological and hemodynamic parameters in patients with essential hypertension and their modification by alpha-1 inhibitor drug treatment, Clin. Hemorheol. Microcirc. 21 (1999) 209-216. A.J. Barnes, J. Oughton and E.M. Kohner, Blood rheology and the progression of diabetic retinopathy: a prospective study, Clin. Hemorheol. 7 (1987), 460. E.M. Hoare, A.J. Barnes and J.A. Dormandy, Abnormal blood viscosity in diabetes mellitus and retinopathy, Biorheology 13 (1976), 21-25. Y. Isogai, A. Iida, K. Michizuki and M. Abe, Hemorheological studies on the pathogenesis of diabetic microangiopathy, Thromb. Res. 8 (1976), 17-24. The Eye Disease Case-Control Study Group, Risk factors for central retinal vein occlusion, Arch. Ophthalmol. 114 (1996), 545-554. M. Peduzzi, A. Debbia and F. Guerrieri, Abnormal blood rheology and retinal vein occlusion. a preliminary report, Graefes Arch. Clin. Exp. Ophthalmol. 224 (1986), 83-85. O. Arend, A. Remky, F. Jung, H. Kiesewetter, M. Reim and S. Wolf, Role of rheologic factors in patients with acute central retinal vein occlusion, Ophthalmology 103 (1996), 80-86. M.B. Beri, M.R. Klugman, J.A. Kohler and S.S. Hayreh, Anterior ischemic optic neuropathy. vii. incidence of bilaterality and various influencing factors, Ophthalmology 94 (1987), 1020-1028. S.S. Hayreh, Acute ischemia of the optic nerve, In: Amaurosis Fugax, E.F. Bernstein, Springer, Berlin, 1988, pp. 93-113. P. Gasser, Ocular vasospasmus: A risk factor in the pathogenesis of low tension glaucoma, Int. Ophthalmol. 13 (1989), 281-290. J.H.J. Klaver, E.L. Greve and H. Goslinga, Blood and plasma viscosity measurements in patients with glaucoma, Br. J. Ophthalmol. 69 (1985), 765-770. E. Watts, Haematological adaptation to exercise, Clin. Hemorheol. 12 (1992), 499-503. J.F. Brun, S. Khaled, E. Raynaud, D. Bouix, J.P. Micallef and A. Orsetti, The triphasic effects of exercise on blood rheology: which relevance to physiology and pathophysiology? Clin. Hemorheol. Microcirc. 19 (1998), 89-104. E. Ernst and A. Matrai, Regular physical exercise increases blood fluidity, Rev. Port. Hemorreol. 1 (1987), 33-40. O. Yalcin, A. Erman, S. Muratli, M. Bor-Kucukatay and O.K. Baskurt, Time course of hemorheological alterations after heavy anaerobic exercise in untrained human subjects, J. Appl. Physiol. 94 (2003), 997-1002. M.S. El-Sayed, Effects of exercise and training on blood rheology, Sports Med. 26 (1998), 281-292.

432


106 J.F. Brun, C. Fons, I. Supparo, C. Mallard and A. Orsetti, Could exercise-induced increase in blood viscosity at high shear rate be entirely explained by hematocrit and plasma viscosity changes? Clin. Hemorheol. 13 (1993), 187-199. 107 N. Gurcan, D. Erbas, E. Ergen, A. Bilgehan, S. Dundar, A. Aricioglu and N. Dikmenoglu, Changes in blood haemorheological parameters after submaximal exercise in trained and untrained subjects, Physiol. Res. 47 (1998), 23-27. 108 V.A. Muravyov, V.N. Levin, M.L. Berkovsky, E.P. Suloev and V.I. Boldina, The red cell rheology. its cytoplasmic water balance and O2-transport under exercise, Clin. Hemorheol. 13 (1993), 352. 109 J.F. Brun, J.F. Monnier, E. Raynaud, J.P. Micallef and A. Orsetti, Erythrocyte disaggregability is reduced during submaximal exercise, Haemostasis 26 (1996), S3. 110 E. Ernst and M. Marshall, Reduced leukocyte filterability after acute physical stress, Clin. Hemorheol. 11 (1991), 129-132. 111 E. Ernst, L. Daburger and T. Saradeth, The kinetics of blood rheology during and after prolonged standardized exercise, Clin. Hemorheol. 11 (1991), 429-439. 112 A. Häieggen, E. Fossum, A. Moan, E. Enger and S.E. Kjeldsen, Whole blood viscosity and the insulinresistance syndrome, J. Hypertens. 16 (1998), 203-210. 113 M. Piagnerelli, K. Zouaoui Boudjeltia, M. Vanhaeverbeek and J-L Vincent, Red blood cell rheology in sepsis, Intens. Care Med. 29 (2003), 1052-1061. 114 E.M. Drost, G. Kassabian, H.J. Meiselman, D. Gelmont and T.C. Fisher, Increased rigidity and priming of polymorphonuclear leukocytes in sepsis, Am. J. Respir. Crit. Care Med. 159 (1999), 1696-1702. 115 O.K. Baskurt, D. Gelmont and H.J. Meiselman, Red blood cell deformability in sepsis, Am. J. Respir. Crit. Care Med. 157 (1998), 421-427. 116 O. Eichelbronner, A. Sielenkamper, G. Cepinskas, W.J. Sibbald, I.H. Chin-Yee, Endotoxin promotes adhesion of human erythrocytes to human vascular endothelial cells under conditions of flow, Crit. Care Med. 28 (2000) 1865-1870. 117 R.J. Powell, G.W. Machiedo, B.F. Rush and G. Dikdan, Oxygen free radicals: effect on red cell deformability in sepsis, Crit. Care Med. 19 (1991), 732-735. 118 J.C. Todd and D.L. Mollitt, Effect of sepsis on erythrocyte intracellular calcium homeostasis, Crit. Care Med. 23 (1995), 459-465. 119 E. Alt, B.R. Amann-Vesti, C. Madl, G. Funk and R. Koppensteiner, Platelet aggregation and blood rheology in severe sepsis/septic shock: relation to the sepsis-related organ failure assessment (SOFA) score, Clin. Hemorheol. Microcirc. 30 (2004), 107-115. 120 W.H. Reinhart, R. Gaudenz and R. Walter, Acidosis induced by lactate, pyruvate, or HCl increases blood viscosity, J. Crit. Care. 17 (2002), 68-73. 121 M. Piagnerelly, K. Zouaoui Boudjeltia, D. Brohée, A. Vereerstraeten, P. Piro, J-L. Vincent and M. Vanhaeverbeek, Assessment of erythrocyte shape by flow cytometry techniques, J. Clin. Pathol. doi:10.1136/jcp.2006.037523 (epub.) 122 P.C. Yodice, M.E. Astiz, B.M. Kurian, R.Y. Lin and E.C. Rackow, Neutrophil rheologic changes in septic shock, Am. J. Respir. Crit. Care Med. 155 (1997), 38-42. 123 M. Nishino, H. Tanaka, H. Ogura, Y. Inoue, T. Koh, K. Fujita and H. Sugimoto, Serial changes in leukocyte deformability and whole blood rheology in patients with sepsis or trauma, J. Trauma 59 (2005), 1425-1431. 124 Y. Inoue, H. Tanaka, H. Ogura, I. Ukai, K. Fujita, H. Hosotsubo, T. Shimazu and H. Sugimoto, A neutrophil elastase inhibitor, sivelestat, improves leukocyte deformability in patients with acute lung injury, J. Trauma 60 (2006), 936-943. 125 G.D.O. Lowe and C.D. Forbes, Rheology of Cardiovascular Disease, In: Clinical Blood Rheology Vol. II, G.D.O. Lowe, Ed., CRC Press Inc., Boca Raton, Florida, USA, 1988, pp. 113-141. 126 L. Heilmann, W. Rath and K. Pollow, Hemorheological changes in women with severe preeclampsia, Clin. Hemorheol. Microcirc. 31 (2004), 49-58. 127 B. Schauf, B. Mannschreck, S. Becker, K. Dietz, D. Wallwiener and B. Aydeniz, Evaluation of red blood cell deformability and uterine blood flow in pregnant women with preeclampsia or IUGR, and reduced uterine blood flow following the intravenous application of magnesium, Hypertens. Pregn. 23 (2004), 331-343. 128 A.L. Tranquilli, G.G. Garzetti, G. De Tommaso, M. Boemi, E. Luciano, P. Fumelli and C. Romanini, Nifedipine treatment in preeclampsia reverts the increased erythrocyte aggregation to normal, Am. J. Obstet. Gynecol. 167 (1992), 942-945. 129 D.J. Pepple, M.R. Hardeman, A.M. Mullings and H.L. Reid, Erythrocyte deformability and erythrocyte aggregation in preeclampsia, Clin. Hemorheol. Microcirc. 24 (2001), 43-48.


433

Treatment in Clinical Hemorheology: A Current Overview Michel R. BOISSEAUa,1, Katalin KOLTAIb, Zsolt PECSVARADYc, Kalman TOTHb a Pharmacological Department, Université de Bordeaux 2, Bordeaux cedex, France, b First Department of Medicine, University of Pecs, Hungary and cFlor Ferenc Hospital, Kistarcsa, Hungary

Introduction Although the concept of hemorheological therapy has been discussed for over four decades, the current status of this topic remains somewhat unclear. Further, the issue of an individual suffering from “abnormal blood rheology” has yet to be fully resolved [1]. Some have even questioned the merit of studying the rheology of blood obtained from convenient vessels (e.g., brachial vein) when, in fact, it is difficult to know how such studies reflect the microvascular effects of a given therapy. There is also the “chicken versus the egg” question, in which it is unclear whether the altered rheology caused the problems or the problems resulted in altered rheological properties [2]. Indeed, the quality of blood flow depends on numerous factors, including the vasomotor activity of arterial vessels, and hence can be affected by local conditions and neural input as well as by blood rheology. This complexity has often led to confusion between the terms “vasoactive drugs” and “rheological drugs”; some agents may have multiple effects and it can be difficult to specifically define their mechanism(s) of action. In order to evaluate the efficiency of any hemorheological treatments, it is necessary to clarify their definition, consider the pathological fields where they are really active, and avoid those with strongly adverse side effects. Also, one should consider how a specific therapy relates to the possible treatments available. Finally, of course, the reliability of tests which are used for the evaluation of hemorheological treatment should also be considered, particularly since it has been asserted that some tests are merely evaluating parameters that only reflect “epiphenomena” [2]. Given such stringent evaluation criteria, it is possible that they could lead to a dramatic reduction, at least temporarily, in the number of treatments in clinical hemorheology. Currently (i.e., mid-2007), relatively few new hemorheological agents have been proposed, and hence there are less scientific sessions devoted to this topic and consequently less involvement by medical practitioners. However, the authors believe that careful basic science and clinical research, combined with an expanded emphasis on oral and printed presentations, will bring new ideas and new therapeutic approaches to the field of clinical hemorheology. 1 Corresponding author: Pharmacological Department, Université de Bordeaux 2, 146 rue Léo Saignat, 33076, Bordeaux cedex, France ; E mail: [email protected]

434

M.R. Boisseau et al. / Treatment in Clinical Hemorheology: A Current Overview

1. How to Define Hemorheological Treatment? The definition of treatments aimed at correcting hemorheological disorders has been a matter of discussion for some time, and terms such as “rheotherapy” or “therapeutic hemorheology” have been proposed yet are not really sufficient. Hence a more specific definition has been suggested: “Medical intervention to increase the fluidity of blood to prevent or improve ischemia.” [1] While ischemia is important and such a definition is helpful, it is solely focused on ischemia: some disorders are not linked to ischemia, at least during the first part of the disease, but rather to other factors. For example, ocular abnormalities in hyperglobulinemias are related more to mechanical impairment of flow rather than to hypoxia. A different approach for selecting a definition of hemorheological treatment is to focus on primary pathological events that only take into account treatments aiming at initial pathological factors. The rationale for this approach is that identification of secondary rheological disorders is difficult and their role somewhat uncertain. For example, in post stenosis or post thrombosis vascular syndromes, one cannot exactly know what is due to hemorheology (e.g., blood viscosity) and what is dependent on thrombotic processes (e.g., platelet aggregation, activated coagulation factors, inflammation, hypoxia). Thus, hemorheological treatments should be defined as therapeutics aimed at the correction of primary events leading to alterations of blood flow and interactions of blood cells with vascular endothelium, and which are the main causes for progression of the disease.

2. Treatment of Primary Hemorheological Disorders The involvement of hemorheological treatment in the course of diseases where abnormal blood flow is the main pathological factor is shown in Table 1. Note that the list separates disorders associated with altered amounts of blood or plasma and with altered vasoreactivity (i.e., primary blood mass disorders) from those associated with disorders of the formed elements in blood (i.e., primary blood cell disorders). 2.1. Primary Disorders of Blood/Plasma Volumes 2.1.1. Polycythemia The progressive increase of the number of red cells in this condition leads to a typical rheological disease: primary polycythemia or Vasquez’s disease. Visual symptoms, vertigo and highly colored mucosa and skin indicate the increased levels of red cells within microvessels; clinical evidence and hematocrit are sufficient for diagnosis. Initially, whole blood is removed from the cardiovascular system in order to prevent a cerebral-vascular accident; subsequent therapy includes a selective program of red cell removal (i.e., erythropheresis) which preserves plasma and blood iron, followed by chemotherapy. 2.1.2. Alterations of Plasma Composition A high concentration of macroglobulins also constitutes a model of a primary rheological disorder. In Waldenström’s disease, increased levels of IgM molecules in


435

the blood increases plasma viscosity and red cell aggregation, leading to obstruction of microvessels due to large aggregates in post-capillary venules. There is also a concomitant functional abnormality of platelets (i.e., acquired thrombopathy). Therapy initially involves a program of plasmapheresis in order to decrease IgM levels followed by chemotherapy.

Table 1. Therapeutic targets in primary hemorheological disorders PRIMARY BLOOD/PLASMA VOLUME DISORDERS Pathological Factor A. Red cell mass Increased amount of red cells

Rheological Clues

Treatment

Polycythemia

Blood letting, Hemodilution

B. Plasma composition Increased plasma Clinical evidence, macroglobulins Plasma viscosity

Waldenström, Others

Plasmapheresis

C. Dysregulation of vessel physiology Dysregulation Clinical evidence, of vascular tone Blood volume

Shock, Vasoplegia

Infusions of plasma substitutes, Hemodilution

Stress polycythemia, Pearson’s disease, Hypertension (some cases)

Treatment of risk factors: smoking, obesity, HTA alcohol, stress, Hydration

Plasma volume contraction

Clinical evidence, Hematocrit, Blood volume study

Disease

Clinical context, Plasma viscosity, Fibrinogen level

PRIMARY BLOOD CELL DISORDERS Pathological Factor A. Red cells Erythrocyte rigidity

Rheological Clues

Disease

Treatment

Clinical evidence

Sickle cell disease

Oxygen therapy, Drugs

B. Leucocytes White cell adhesion to arteries

Plasma markers (VCAM)

Atherogenesis (covert states)

Prevention: diet, life style hypo-lipemics

White cell plugs in micro vessels

None

Primary ischemia/ reperfusion (I/R)*

Deoxygenated blood, Hemodilution

White cell adhesion in venous valves

Plasma markers (VCAM)

Venous disease** White cell trapping

Prevention, veno-active drugs

* But not secondary ischemic/reperfusion **Chronic but not venous thrombosis

2.1.3. Dysregulation of Vessel Physiology Arteries and microvessels on the arterial side of the microcirculation are innervated by a rich network of sympathetic nerves that affect vascular tone. Smooth muscle in the vessel wall responds to neural stimuli, thereby altering vessel diameter; changes in vessel diameter can affect the apparent viscosity of blood in that vessel as well as vascular resistance (see chapter II.3.a). In the case of vasoplegia, blood volumes are

436


changed and blood flow subsequently altered. A dilution of blood can occur, or conversely, hemoconcentration and hypovolemia may result; clinical shock is a typical example of the latter. Given altered perfusion there is a decrease of capillary filtration pressure and of filtration resulting in fluid movement into the vascular space and thus hemodilution. This situation is, however, transitory and is usually followed by loss of fluid and hemoconcentration due to lesions of endothelial cells as seen in burns and septic shock. Treatment includes infusion of plasma substitutes using dextran and gelatin solutions per current protocols [3]. Curiously, such therapy involves rheological principles, yet is seldom referred to as hemorheology. Contracted plasma volume (CPV) is a primary rheological disease in which there is a decrease in plasma volume, elevated hematocrit and high fibrinogen levels that serve as specific diagnostic findings [4]; sometimes, however, the disease looks deceptively like polycythemia (i.e., false polycythemia of Pearson). CPV is also observed in some cases of hypertension and frequently in older people; in the elderly, hematocrit is usually low due to renal insufficiency yet there can be a small increase of hematocrit due to CPV. To a varying degree, CPV is observed in subjects having cardiovascular risk factors (e.g., smoking, obesity, alcohol, fatty diets). Treatment of CPV is initially based on increasing hydration via oral intake of fluids followed by infusions and attempts to change the life style of patients. It is interesting that this frequent syndrome, primary but covert, is often hidden due to the general medical advice “to frequently drink water”. 2.2. Primary Blood Cell Disorders 2.2.1. Red Cell Rheological Abnormalities Alteration of red cell deformability can be considered to be a “classic” finding in a hemorheological disease. The most well-known model is homozygous sickle cell (HbSS) disease that leads to anemia and to acute painful crisis. The mechanism for the decrease in deformability is well known: polymerization of HbS at low oxygen tensions and acidic conditions resulting in deformed rigid cell with altered membrane and cellular mechanical properties. Sickle cell crisis is thought to occur due to obstruction of microvessels by rigid red cells and adhesion of erythrocytes to the wall of post capillary venules; white blood cells (i.e., PMN) may also contribute to the obstruction. Treatment is symptomatic and includes oxygen therapy, fluids, analgesics and antibiotic therapy; long term treatments are limited to bone marrow transplant and hydroxyurea to promote fetal hemoglobin production. Drugs aimed at improving cellular behavior usually focus on prevention of cellular dehydration since the rate of polymerization is very sensitive to intracellular hemoglobin concentration, or on agents that increase fetal hemoglobin levels and hence retard polymerization. Some physicians indicate that they use pentoxifylline or almitine, although the latter product is no longer commercially available and the former scarcely indicated nowadays [5]. Monitoring the effects of such drugs is somewhat difficult inasmuch as red cell deformability should be measured at arterial and venous oxygen tensions; both micropore filtration and ektacytometry (i.e., diffraction of laser beam by cells under fluid shear) have been employed.


437

2.2.2. Rheological Disorders of White Cells White cell adhesion to the endothelial cells of the arterial wall has been recently recognized as the initial process of atherosclerosis [6]. Leucocytes adhere to areas such as branching of collaterals or bifurcations where shear stresses are low. This allows the cells to “roll” slowly at the surface of the endothelial cells, then to migrate across the sub-endothelium and reach the media. White cells which migrate are monocytes, macrophages and mast cells; some of them with a high fat content are known as spumous cells. The frequent presence of CPV (i.e., contracted plasma volume) during atherogenesis causes blood hyperviscosity, thereby changing the shearing profile within a vessel and promoting white cell movement towards the vessel wall. Increased plasma markers released from the endothelium (e.g., vascular cell adhesion molecule, VCAM) can be used as an index of this very frequent but covert phenomenon. Treatment of atherosclerosis involves reducing risk factors via diet and exercise, and the use of hypolipemics, clofibrate and/or statins. Unfortunately, there presently do not appear to be drugs which act directly on white cell adhesive behavior, perhaps due to concerns about secondary leucopenia. Note that while white cells are involved in the formation and growth of atherosclerotic plaques, they are not involved in thrombosis; the latter phenomenon is due to platelets which aggregate at the surface of fissured plaques allowing the thrombus to form and retard or stop flow. 2.2.2.1. White Cell Adhesion in the Microcirculation Ischemic reperfusion syndrome (I/R) is due to the plugging of microvessels with leucocytes during ischemia and the failure of these cells to be removed during reperfusion, thus rendering downstream tissues anoxic. Most I/R syndromes are secondary (see below). During cardiac surgery, when blood flow is to be restored after extracorporeal circulation, the I/R syndrome in the ischemic myocardium is prevented by the transfusion of deoxygenated blood at low hematocrit. Drugs to counter the I/R syndrome include allopurinol or pentoxifylline, but are used sparingly for fear of postoperative leucopenia and infection. 2.2.2.2. White Cell Adhesion to the Venous Wall Adhesion of leucocytes to venous valves has been recently reported to occur very early in the development of chronic venous disease [7]; this is a process which mimics that occurring in atherogenesis and which, as for arteries, does not cause thrombosis. Rather, deep venous thrombosis is due to the coagulation processes (e.g., thrombin generation, fibrin deposition, elevated D-dimer levels) which often develop in valves during stasis. Note that this situation is different from the development of venous disease due to monocytes and mast cells, since such cells are able to enter the wall inside the cups of valves where the endothelium is activated. Neutrophils can adhere to the wall and participate in the activation of the surrounding milieu, but do not enter it. Nevertheless, they can be “trapped” in venules of the skin during standing for long periods (i.e., “white cell trapping”) in patients with venous disease and varicosities. Treatment aiming at decreasing adhesion of white cells is mainly based on prevention: physical exercise, improvement of life style, elastic compression devices at work and when traveling, and by veno-active drugs [8] (see below).

438


3. Hemorheological Treatments in Secondary Alterations of Blood Flow Cerebral ischemia and cardiovascular diseases are especially prone to secondary hemorheological disorders. Hypoxia and inflammation due to vessel stenosis induce hyperviscosity (e.g., increased fibrinogen, CPV, high red cell aggregation), reduced red cell deformability, and activation of leukocytes and the endothelium: these pathological factors tend to worsen the status of the circulation. Since both the patient population and the potential for profit are large, several drugs for treatment have been suggested, especially in the 1980’s and 1990’s. Unfortunately, little success has been achieved: the proposed medications were vasoactive agents having little effect in acute and chronic situations, and there is now competition from anti-thrombotic drugs such as clopidogrel. Hemodilution is not universally used in critical ischemia owing to concerns about side effects; Iloprost is used in specific indications. Further, advancements in angiology and vascular surgery have led to a reduction in chronic states due the development of prosthesis, stents and prevention of risk factors. 3.1. Drugs and Therapies Affecting Hemorheological Parameters 3.1.1. Hemodilution 3.1.1.1. Basic Considerations The treatment of cardiovascular diseases, predominantly those caused by atherosclerosis, is possible through numerous methods. The defect of blood supply and function of the damaged target organ should be treated in addition to preventing or slowing the development and the progression of the pathological process. Improving the nutrient supply capacity of blood at the microcirculatory level plays an essential role within this complex treatment strategy, inasmuch as the capacity of the arterial system decreases due to the progression of atherosclerosis. Up to a point, decreased blood supply to post-stenotic areas can be compensated for via local vasodilation. However, the reserves of this compensatory mechanism are limited, and subsequently blood viscosity becomes a direct determinant of capillary perfusion. Numerous studies have shown that a decrease in blood viscosity may result in significant improvement of tissue perfusion; the most effective way to reduce blood viscosity is hemodilution. A hematocrit between 30-45% is most favorable from the point of view of oxygen transporting capacity, based on mathematical calculations using a hematocrit/viscosity ratio or on in vitro measurements. According to most reports, a hematocrit over 45% should prompt consideration of reducing it in order to improve tissue perfusion. However, hemodilution should be performed only if the dilution of coagulation factors does not raise the risk of bleeding; adverse interactions with other chronic diseases or agents (e.g., peptic or duodenal ulcer, medications such as heparin derivatives, vitamin K antagonists, anti-platelet agents) should also be avoided. 3.1.1.2. Peripheral Obliterative Vascular Diseases The most numerous and most informative studies concerning hemodilution have been conducted in the context of treating peripheral obliterative diseases >9, 10@. Numerous methods and agents have been tested (e.g., crystalloids, volume supplements, vasodilators) with and without blood removal. The results indicated that isovolemic hemodilution is favored compared to hyper- and hypo-volemic treatments; hydroxyethyl-starch (HES) and dextran solutions proved to be most suitable, with the former used more often. These solutions have also been shown to have a beneficial


439

effect on leukocyte-endothelial cell interactions which have a pivotal role in the pathomechanism of atherosclerosis. The Trans Atlantic Inter-Society Consensus (TASC) Working Group declaration from 2000 recommends hemodilution in chronic limb ischemia (level B recommendation), whereas the ACC/AHA guidelines from 2005 do not discuss it. Its use as a periodic chronic therapy is recommended beginning at Fontaine II/b stage when the Doppler index is below 0.69 (level A recommendation). 3.1.1.3. Acute Cerebrovascular Diseases In the case of acute ischemic stroke, a routinely performed iso- or hypervolemic hemodilution does not result in significant improvement of neurological status. However, in the case of carefully-selected subjects, it contributes to the prevention of deep vein thrombosis and pulmonary embolism without adverse side effects, and significantly reduces mortality compared to the control group >11@. The international guidelines do not recommend hemodilution treatment in acute ischemic stroke (level A recommendation). Yet according to the results of several randomized trials in patients without exsiccosis, an individualized isovolemic hemodilution is not contraindicated if there are no other contraindications and hematocrit is 50% either because of polycythemia vera or secondary polyglobulia, Isovolemic hemodilution may be used for the secondary prevention of ischemic stroke in the case of pathological levels of whole blood and plasma viscosity and associated hyperfibrinogenemia (i.e., >4.0 g/l, level B recommendation). In the early period following surgery for subarachnoid bleeding, the ”triple H therapy” (i.e., hypervolemia, hemodilution, hypertension) is employed in neurosurgical intensive care units in order to maintain the patient’s hemodynamic parameters at an optimal level and to improve cerebral microcirculation by increasing cardiac output. An optimal hematocrit is between 30-35%, and with the above treatment approach, ischemic stroke may be prevented, or if already present, the area of cerebral penumbra may be decreased (level A recommendation). 3.1.1.4. Chronic Obstructive Pulmonary Disease In a study by Borst et al., decreased whole blood viscosity resulting from repeated hemodilution improved pulmonary gas exchange parameters in chronic obstructive pulmonary disease (COPD) >12@. In these studies, a 6% hydroxy ethyl starch solution with a molecular weight of 40 kDa was exchanged with 5-6 blood draws within three months. Cardiac output and arterial oxygen tension increased significantly, while mean pulmonary artery pressure and pulmonary vascular resistance index decreased (level C recommendation). 3.1.1.5. Ischemic Heart Disease Hemorheological parameters are known to play a role in the pathogenesis of ischemic heart disease. In unstable angina pectoris, plasma viscosity is significantly higher than in stable angina, and it is significantly higher in patients with acute coronary syndrome leading to an acute myocardial infarct compared to those who did not suffer myocardial necrosis. Ischemic heart disease patients scheduled for coronary bypass graft surgery tolerate isovolemic hemodilution well, and their cardiac output and oxygen extraction ratio increase significantly. Isovolemic hemodilution performed in coronary artery stenosis patients decreased hematocrit from 46 % to 39 % on average, yet did not significantly decrease the frequency of angina nor did it improve hemodynamic parameters >13, 14, 15@. At present there are only level B and C recommendations concerning the long-term efficacy of hemodilution in the practice of cardiology.

440


3.1.1.6. Hematological Diseases Polycythemia vera is a myeloproliferative disease, has a prevalence of 0.7 to 1.6 per 100,000, and a total red blood cell volume measured with 51Cr of > 35 ml/kg (male) or > 32 ml/kg (female). Myeloproliferation extends to all three cell lines (i.e., leukocyte, erythrocyte, platelet), but in the early stage of the disease only one or two lines may be affected. It results in increased risk of thrombosis that can be decreased with the reduction of erythrocyte concentration. Secondary polyglobulia is especially frequent in certain diseases: in peripheral obliterative angiopathy its prevalence is 18-25%, in sleep apnea 20-22%, in ischemic heart disease 18-25%, in obesity and metabolic syndrome 28-34%, and in chronic obstructive pulmonary disease 10-26%. Interestingly, it does not lead to elevated leukocyte count or splenomegaly. Clinical trials have shown that bleeding time increases, platelet aggregation decreases, and there are less thrombo-embolic events in parallel with the reduction of hematocrit. Pre-operative hemodilution increases the stay-open ratio after reconstructive surgery in lower limb obliterative vascular diseases, whereas hemoglobin levels raised by transfusion result in an increase of cardiovascular events and cardiovascular mortality. In the diseases mentioned in the preceding paragraph, normalization of pathological parameters via blood-draw and adequate volume replacement is a level A recommendation. 3.1.2. Indications for Hemodilution Hemodilution may be the therapy of choice in relative or absolute hyperviscosity syndromes, with hyperviscosity recognized if whole blood- and plasma viscosity levels exceed the values shown in Table 2. Table 2: Whole blood and plasma viscosities in hyperviscosity syndrome Whole blood viscosity low-shear ( 10 s-1) medium shear (50-100 s-1) high shear ( 150 s-1) Plasma viscosity

> 8.0-9.0 mPa.s > 5.0 mPa.s > 4.5 mPa.s > 1.35 mPa.s

Special indications for hemodilution include peripheral arterial occlusive disease for Fontaine stage II/b (dysbasic distance < 200 m) in severity and/or with a Doppler index below 0.69, in chronic obstructive pulmonary disease when paO2 < 70 mm Hg, and/or paCO2 > 60 mm Hg, polycytemia, secondary polyglobulia, occlusion of the central retinal artery, hypacusis, tinnitus, vertigo or sudden idiopathic hearing loss. Laboratory findings that suggest considering hemodilution in the above pathologies are shown in Table 3. Table 3: Criteria for hemodilution Hematocrit Hemoglobin Fibrinogen:

Males > 45 % > 160 g/l > 3.5 g/l

Females > 42 % > 150 g/l > 3.5 g/l


441

3.1.3. Individualized Approach to Hemodilution The objective of hemodilution is to improve the perfusion of tissue or organs with inadequate blood flow, and hence to alleviate clinical symptoms and ameliorate the patient’s complaints. Determining the degree of hemodilution should include consideration of co-existing diseases (e.g., coronary or pulmonary disease) and the compensatory capacity of the vascular system. It remains uncertain whether hemodilution is indicated in order to reach a level below the optimal hematocrit range (i.e., 30-45%). In order to decide if hemodilution is appropriate, whole blood and plasma viscosity measurements should be performed, and a complete blood count and parameters of hemostasis should be determined (i.e., bleeding and clotting time, prothrombin value, activated partial thromboplastin time, thrombin time, fibrinogen concentration). Several types of infusions may be used during hemodilution, including crystalloid solutions, albumin, gelatin derivatives, dextran, hydroxyethyl starch. Based on our present knowledge, hydroxyethyl starch seems to be most beneficial based on its sideeffect profile, availability, and duration of effect; crystalloid infusion is proposed in cases when drugs are given with the infusion. 3.1.4. Drugs Purported to Affect Erythrocyte Deformability During the past two decades numerous products have been claimed to be useful for improving red cell deformability in cardio vascular diseases, and an exhaustive list and related trials have been published [2]. Actually, most of these products are vasoactive drugs, with the background for their clinical development based upon improved walking distances in patients with peripheral arterial disease. Pentoxifylline was the most popular agent based upon several trials, but eventually the drug failed to be accepted as a primary therapy; similar outcomes occurred for isoxsuprine, naphtidrofuryl, bencyclane, piracetam, ginkgo biloba extract and nicergolin. Presently, such products are only developed and promoted based upon their original specificity (e.g., Į blocker, ȕ blocker). Some products in the field of venous diseases have been focused on reducing red cell aggregation: troxerutin showed positive results based on both erythrocyte aggregation and reduction of PAI-1 [16]; rutosides exhibited the same effects, yet neither troxerutin nor rutosides were commercially successful.

3.1.5. White Cell Active Drugs 3.1.5.1. Arterial Disease One of the most active drugs for white cells is iloprost which affects the generation of free radicals and leukocyte adhesion. Its usage has decreased due to the rapid progress in vascular surgery during the last 2-3 decades; it continues to be used in special cases of critical ischemia but only in specialized vascular surgical centers. The use of pentoxifylline and numerous “anti oxidants” including vitamin E and PAF-inhibitors has also been suggested. Products such as dipyramidol and buflomedil that cause an increased production of adenosine seem logical choices since they should affect adhesion; buflomedil increases plasma adenosine via an action on endothelial A2A receptors, leading to short term improvement of the microcirculation during acute states of peripheral arterial disease [17]. All of these products demonstrate activity, and show some benefit but only after long periods of treatment; they have recently

442


been superseded by anti-platelet agents such as aspirin and clopidogrel although buflomedil remains in use for peripheral vascular disease. 3.1.5.2. Venous Disease The basis for this pathology is a coagulation disorder (i.e., generation of thrombin in valves) that requires heparin and coumadin therapy and physical exercise. In the course of chronic venous disease, which is now well documented, the role of white cells is well described [7]. Two main mechanisms related to the progression of chronic venous disease are targets for pharmacological intervention: those consisting mainly of alterations of the venous wall at the macrocirculatory level and those dealing with microcirculatory disorders, and hence the mode of action of an agent depends on the “target”. Chronic venous disease is related to the failure of venous valves that are affected by inflammation. Activated leukocytes can migrate into the endothelium of proximal surfaces of the valves as well as into proximal vein walls, thus promoting destruction of supporting structures and remodeling of the valves with consequent valvular insufficiency [18]. Immunohistochemical studies using monoclonal antibody specific for monocytes and macrophages have demonstrated monocyte/macrophage infiltration, and antibody studies have shown greater leukocyte infiltration into the base of the valves and the proximal venous wall. Since venous valves exist in regions of low shear stress and disturbed flow [19], it may be that these phenomena explain leukocyte deposition in these regions [7]. It is not known what initiates the inflammatory events in venous valves and walls, although altered shear stress may be involved in several ways. Prolonged pooling of blood causes distension of lower limb veins, distortion of venous valves, and leakage through valves that exposes endothelial cells to flow reversal. Venous stasis, even in the absence of reflux, produces regions of low or zero shear stress, and subsequent structural changes may induce regions of disturbed flow. It thus appears that inflammatory processes involving leukocyte–endothelial interactions and triggered largely in response to abnormal venous flow are important in causing the adverse changes in venous valves and vein walls. Currently available drugs have been shown to attenuate various elements of the inflammatory cascade, particularly the leukocyte-endothelial interactions [20, 21]. A micronized purified flavonoid fraction (MPFF) suppresses damage to the valve structures, and the use of phlebotropic drugs such as horse chestnut seed extract and hydroxyrutosides result in a reduction of edema and symptoms of chronic venous insufficiency, but have failed to demonstrate superiority over compression. Recently much attention has been focused on the involvement of growth factors [22] and leukocytes in the development of venous ulceration [23], thus opening new areas of investigation. Among the many mechanisms at work in the pathogenesis of venous ulceration, the mechanism involving leukocyte activation and interaction with the endothelium seems to be the one most responsive to pharmacological treatment [24]. Plasma levels of some adhesion molecules at the surface of both leukocytes and endothelium are elevated in patients with chronic venous disease and changes in their levels may be feasible markers for response to therapy. MPFF taken daily over two months reduced several indices of inflammation in the microcirculation [25, 26]: levels of ICAM-1, VCAM and leukocyte adhesion molecules (monocyte or neutrophil CD62L) decreased significantly from baseline (p <0.001), thus mirroring the ability of MPFF to inhibit the binding of neutrophils and monocytes to endothelial cells.


443

Given the above, it seems important to differentiate between mechanisms involving hemorheological factors and venous stasis or hypertension. One is the generation of thrombin at the bottom of vein valves, mainly in deep veins, leading to deep vein thrombosis. Here the treatment should be the anti-thrombin: heparin, coumadin and, especially in young subjects, fibrinolytic treatment with rt-PA. The second situation, which is more frequently observed in superficial veins, is the development of chronic venous disease (e.g., varicose veins). Here the process is presumably linked to white cell infiltration of valves and some parts of the wall, and should thus prompt development of “anti-leukocyte” drugs.

Acnowledgements Some parts of a previous article of MR Boisseau were here included with the agreement of the Editor Servier France Laboratory. The author is most grateful to Françoise Pitsch PhD for her suggestions and advice.

References [1] [2] [3] [4] [5] [6] [7] [8]

>9@

>10@ >11@

>12@

>13@

>14@

E. Ernst, Hemorheological treatment,. in Clinical Hemorheology, S. Chien, J. Dormandy, E. Ernst and A. Matrai, Eds., Martinus Nijhoff Publ., Dordrecht, 1987, pp. 329- 373. A.M. Ehrly, Therapeutic Hemorheology, Springer Verlag, Berlin, 1991. Advances in Hemorheology on Hemodilition, proceedings from a Symposium held in Graz, Clin. Hemorheol. 12 (Supl 1) (1992), 1-93. J.P. Isbister, The stress polycythemia syndromes and their hemorheological significance, Clin. Hemorheol. 7 ( 1987), 159-179. J. Moussoki, General Manager, Association pour la prévention de la drépanocytose et les hémoglobinopathies, personal communication, 2007. M.R. Boisseau, Effect of shear stress in vascular endothelial changes. Phlebolymphology 40 (2000), 143-155. J.J. Bergan, G.W. Schmid-Shönbein, P.D. Coleridge–Smith, A.N. Nicolaides, M.R. Boisseau and B. Eklof, Mechanisms of disease: chronic venous disease. N. Engl. J. Med. 335 (2006), 488-498. A.A. Ramelet, M.R. Boisseau, C. Allegra, A. Nicolaides, K. Jaeger, P. Carpentier, R. Cappelli, and S. Forconi, Veno-active drugs in the management of chronic venous diseases. An international consensus statement: current medical position, prospective views and final resolution, Clin. Hemorheol. Microcirc. 33 ( 2005), 309-320. H.G. Höffkes, A.M. Ehrly, A. Franke and H. Landgraf, Optimal hematocrit in patients with peripheral occlusive arterial disease (POAD): Exercise induced muscle tissue pO2 after isovolemic hemodilution, Clin. Hemorheol. 16 (1996), 321-327. E. Ernst A. Matrai and L. Kollar, Placebo-controlled, double-blind study of haemodilution in peripheral arterial disease, Lancet 27 (1987), 1449-1451. F.T. Aichner, F. Fazekas, M. Brainin, W. Polz, B. Mamoli and K. Zeiler, Hypervolemic hemodilution in acute ischemic stroke. The Multicenter Austrian Hemodilution Stroke Trial (MAHST), Stroke 29 (1998), 743-749. M.M. Borst, M. Leschke, U. König and H. Worth, Repetitive hemodilution in chronic obstructive pulmonary disease and pulmonary hypertension: effects on pulmonary hemodynamics, gas exchange, and exercise capacity, Respiration 66 (1999), 225-232. L. Herregods, A. Moerman, L. Faubert, N. Den Blauwen, E. Mortier, J. Poelaert and M. Struys, Limited intentional normovolemic hemodilution: ST-segment changes and use of homologous blood products in patients with left main coronary artery stenosis, J. Cardiothorac. Vasc. Anesth. 11 (1997), 18-23. M. Leschke W. Motz and B.E. Strauer, Hemorheologic therapy application in coronary heart disease, Wien Med. Wschr. 136 (1986), 17-24.

444


>15@ L. Bogar, I. Juricskay, G. Kesmarky, G. Feher, P. Kenyeres and K. Toth, Gender differences in hemorheological parameters of coronary artery disease patients, Clin. Hemorheol. Microcirc. 35 (2006), 99-103. [16] M.R. Boisseau, A. Taccoen, C. Garreau. Fibrinolysis and haemorheology in chronic venous insufficiency: a double blind study of troxerutine efficiency, J. Cardiovasc. Surg. 36 (1995), 369-374. [17] P. Cappechi, F. Laghi-Pasini, N. Sodi, M. Chiavetta, S. Sensi, A. Da Lalla, L. Volpi and T. Di Perri, Increase of plasma adenosine and adenine nucleotides after intravenous infusion of buflomedil in humans, J Cardiovasc Pharmacol 1 (1995), 35-39. [18] T. Ono, J.J. Bergan, G.W. Schmid-Shönbein and S. Takase, Monocyte infiltration into venous valves. J Vasc Surg 27 (1998), 158-166. [19] F. Lurie, R.L. Kistner, B. Eklof and D. Kessler, Mechanism of venous closure and role of the valve in circulation, a new concept, J. Vasc. Surg. 38 (2003), 955-961. [20] C. Boougelet, I.H. Roland, N. Ninane, T. Arnould, J. Remacle and C. Michiels, Effect of aescine on hypoxia-induced neutrophil adherence to umbilical vein endothelium, Eur. J. Pharmacol. 345 (1968), 89-95. [21] S. Takase, L. Pascarella, J.J. Bergan, G.W. Schmid-Shönbein, Venous hypertension, inflammation and valve remodeling, Eur. J. Vasc. Endovasc. Surg. 28 ( 2004), 484-493. [22] D.D. Wright, P.J. Franks, S.D. Blair, C.M. Backhouse, C. Moffatt and C.N. Mc Collum, Oxerutins in the prevention of recurrence in chronic venous ulceration: randomised controlled trial, Br. J. Surg. 78 (1991), 1269-1270. [23] A. Jull, A. Waters and B. Arroll, Pentoxifylline for treatment of venous leg ulcers: a systematic review, Lancet 359 (2002), 1550. [24] S.S. Shoab, J.B. Porter and P.D. Coleridge-Smith PD, Effect of oral micronised flavonoid fraction treatment of adhesive adhesion molecule expression in patients with chronic venous disease, a pilot study, J. Vasc. Surg. 31 (2000), 456-461. [25] P.D. Coleridge-Smith, C. Lok and A.A. Ramelet, Venous leg ulcer: a meta-analysis of adjunctive therapy with micronised purified flavonoid fraction, Eur. J. Vasc. Endovasc. Surg. 30 (2005), 198-208. [26] M.P. Jacob, M. Cazaubon, A. Scemama, D. Prié, F. Blanchet, M.C. Guillin and J.B. Michel, Plasma matrix metalloproteinase 9 as a marker of blood stasis in varicose vein, Circulation 106 (2002), 535538.


445

A Note on the Editors Oguz Baskurt received an MD degree from Hacettepe University (Ankara, Turkey) in 1982, and a PhD in physiology in 1988 from the same institution. He joined the Hacettepe University Department of Physiology as an Assistant Professor in 1989, and in 1990 moved to the Department of Physiology of Akdeniz University Faculty of Medicine (Antalya, Turkey) where he is presently Professor and Chairman. His current research interests are mainly focused on the hemodynamic effects of blood rheology alterations, especially alterations of red blood cell aggregation; he also has an active interest in comparative hemorheology. Dr. Baskurt received the Lafon Hemorheology and Microcirculation Award given by the International Society for Clinical Hemorheology in 1995. Max R. Hardeman obtained a Master’s degree in Biochemistry at the University of Amsterdam in 1968. After fulfillment of a military duty obligation he continued his scientific career at the Central Laboratory of the Dutch Red Cross Blood Transfusion Service, completing a doctoral thesis on platelet preservation in 1974. Registration as a Clinical Biochemist occurred in 1977 following a training period at the Academic Medical Center of the University of Amsterdam. This position was held until retirement in 2005; a continuing appointment at the Academic Medical Center enables him to continue with research activities. His research interests range from fundamental clinical hemorheological to the development of instrumentation for measurement of red cell blood deformability and aggregation. Mike Rampling received a BSc in Physics from Imperial College, London and started his professional career at the UK Atomic Energy Authority. He moved to St. Mary’s Hospital Medical School, London in 1967 where he completed a doctoral thesis on the erythrocyte sickling phenomenon; he continued as faculty member until transferring to Imperial College in 1998. He retired as Reader in Biophysics in 2005, but given his continuing scientific interests he retains an honorary position with Imperial College. His research interests have concentrated on basic and clinical aspects of hemorheology. He was awarded the Fahraeus Medal at the 2007 Conference of the European Society for Clinical Hemorheology and Microcirculation. Herbert J. Meiselman, ScD received a BS in Chemical Engineering from the Michigan Technological University and a doctoral degree (ScD) in chemical engineering and hemorheology at the Massachusetts Institute of Technology; post-doctoral work related to the rheology of the in vivo microcirculation was done at the California Institute of Technology. Since 1972 he has been a faculty member in the Department of Physiology and Biophysics, Keck School of Medicine, University of Southern California, where he is presently Professor and Vice-Chair. His research interests continue to be focused on the basic science and clinical rheological aspects of blood and its formed elements. He received the Fåhraeus Award in 1993 at the Eighth European Conference on Clinical Hemorheology, Vienna, Austria.


447


Subject Index Acute ischemic stroke 403 Acute lung injury Transfusion related, TRALI 234 Acute phase response 376 Adhesion 153 Leukocytes 153 Platelets 153 Adhesion molecules Collagen 159 Integrin 158, 159 Platelet glycoproteins 159 Selectin 158 von Willebrand factor 159 Adult respiratory distress syndrome Blood transfusion 233 Aerobic capacity 425 Ageing Blood volume and haematocrit 375 Aggregability 118, 183 Clinical conditions 124 Donor effects (adults) 118–121 Donor effects (neonates) 121 In vivo cell age 121, 123 Theoretical interpretation 130 Aggregation 314 Aggregation models 115 Bridging 115 Depletion 115 Depletion concept 115, 116 Aggregometers LORCA 255 Myrenne 255 SEFAM 257 Anemia of chronic disease 376 Angular distortion 24, 25 Animal RBC 127 Anterior ischemic optic neuropathy 421 Aorta 295 Apparent viscosity 314 Apparent viscosity of blood 322 Arterial 295 Arterioles 316

Artificial blood 210 Artificial hearts 208 Artificial kidneys 209 Artificial liver 210 Artificial lungs 209 Atherosclerosis 164, 392–397 Cardiovascular risk factors in 396 Complications of 396 Hemodynamic factors in 393 Hemorheologic parameters in 394 Humoral factors in 395 Molecular mechanisms in 394 Atomic force microscopy 98, 103, 140 Atrial natriuetic factor 372 Auscultatory method 352 Autoregulatory reserve 325, 330 Axial migration 326, 331, 332 Aggregation time constant 333 Role of erythrocyte aggregationt 332 Bifurcations 303 Biohybrids 210 Blood cell deformability 310 Blood flow 354 Blood pressure 351 Blood salvage devices 212 Blood storage 230 Blood storage lesion 231 Clinical significance 233 Blood transfusion Age of stored blood 236 Efficacy 235 Fresh blood 232 Massive transfusions 231 Blood velocity 357 Blood viscoelasticity 80–82 Boundary layer effects 84 Clinical applications 87 Deformability 79, 80 Frequency 77 Geometry effects 84 Hematocrit 83 Measurement methods 72, 73

448

Rheological definitions 73–76 Steady flow vs. oscillatory flow 77 Temperature 82, 83 Thixotropy 83 Viscoelasticity profile 76–78 Blood viscosity 308 Behavior 243 Clinical relevance 262 Normal values 261 Blood volume 371 Sensors and effectors in blood volume control 373 Bulk cell filtration 101 Capillary network 312 Cardiac output 292, 355 Cardio pulmonary bypass 208 Cardiopulmonary support 209 Central venous pressure 292 Cerebral blood flow 402 Chemokine 158 Chronic venous insufficiency 406 Cigarette smoking 384 Circulation In fetus 194, 195 In neonate 194, 195 Clumps 314 Coagulopathies 386 Cold agglutinins 387 Comparative hemorheology Amphibians 279 Birds 278 Camelids 276 Fish 279 Mammals 269–277 Marine mammals 276, 277 Reptiles 279 Comparative studies Red cell aggregometers 257 Red cell deformability measurement techniques 252 Viscometers 247 Constitutive equations 26, 54 Bingham 26 Casson equation 52, 53, 55 Newtonian 26 Power law 26 Quemada equation 55 Continuum model 21

Size limitation 22 Coordinate systems 22 Cylindrical 25 Rectangular 22, 23 Spherical 25 Couette system 249 Critical yield pressure 312 Cryofibrinogenemia 388 Cryoglobulinemia 173 Cryoglobulins 387 Curvature 295 Deformability 313, 341, 346 Leukocyte 137 Deformation curve parameterization 249 Depletion layer 127 Interaction energies 128 Thickness 128 Dextran 315, 327, 328 Diabetic retinopathy 419 Diffuse Optical Tomography (DOT) 364 Doppler effect 357 Drugs purported to affect erythrocyte deformability 441 Dysregulation of vessel physiology 435 Echo PIV 360 Effective viscosity 317 Ektacytometers LORCA 250 Rheodyn-SSD 251 Rheoscan 251 Ektacytometry 249 Elastic modulus 93, 95, 96, 100–104 Electromagnetic flowmeter 356 Electron dense knobs 98 Electrophoretic mobility 116 Electrostatic interactions 129 Ellipsometry. See Ektacytometry Endothelial nitric oxide synthase 325, 334, 335 Endothelial Surface Layer (ESL) 303, 340, 346 Endothelin 372 Endothelium 339 Energy 75 Entrance effects 294 Entrance length 294

449

Erythrocyte. See Red blood cell Erythrocytes 36, 37 Blood viscosity 37 Hematocrit 36 Microcirculation 37 Packed cell volume 36 Exercise 422–425 Long-term effects 424 Middle-term effects 424 Short-term effects 422 Exported malaria proteins 99 Extracorporeal membrane oxygenation 209 Fåhraeus effect 193, 300, 331, 333 Fahraeus 309 Fåhraeus-Lindqvist effect 193, 302, 331 Fetal Hematocrit 191 Plasma viscosity 192 Fibrinogen 114, 180, 315 Fick principle 355 Filters 230 Filtration 312 Flow 307 Flow chamber 156 Flow resistance 292 Flow-induced hemolysis 207 Fetus 38 Hematological values 38 Glycocalyx 127 Polymer penetration 129 Glycocalyx 318 Hematocrit Body hematocrit 372 Hemoconcentration and hemodilution 372 Optimal hematocrit 373 Regulation of hematocrit and responses to stressful stimuli 374 Relationship to red cell mass and plasma volume 371 Hematological stress response acute 375 Hemoconcentration 372 Hemoglobinopathies 96 Hemolytic uremic syndrome 385 Hemorrhage Blood transfusion 235

Resuscitation 235 Haptoglobin 213 Hardened cells 311 Heart failure 401 Heart valves 210 Hematocrit 309 Hematocrit alterations 171 Hematocrit effects 50, 51 Hematological and oncological disorders 411–418 Hemodialysis 209, 210 Hemodilution 220, 438–441 In acute cerebrovascular diseases 439 In chronic obstructive pulmonary diseases 439 In hematological diseases 440 In ischemic heart diseases 439 In neonatal polycythemia 200 In peripheral obliterative vascular diseases 438 Indications for 440 Hemoglobinopathies 174 Hemoglobinuria 213 Hemolysis 208, 210, 212, 213, 220 Hemolysis rate index 214 Hemorheolgical treatment 434 Hemorheological treatments in secondary alterations of blood flow 438–443 Hemorheological vicious cycle 184 Hemorheology In fetus 194, 195 In neonate 194, 195 Heparin induced thrombosis thrombocytopenia syndrome 386 Hereditary elliptocytosis 95 Hereditary spherocytosis 95 HRI. See Hemolysis Rate Index Hydraulic diameter 317 Hyperfibrinogenaemia 388 Hyperleukocytic leukemias 415 Hypertension 377, 401 Hypertensive retinopathy 418 Hyperviscosity 343 Classical hyperviscosity syndrome 376 Clinical presentations 376

450

General principles of management 379 Hyperviscosity investigations Biochemical analyses 378 Blood examination 378 Immunological tests 378 Test of blood fluidity 378 Hypoxia 375 Impedance of blood flow 85 In vivo blood flow Effect of leukocytes 335 In vivo flow 66 In vivo pressure-flow relationship 323–330 In vivo versus ex vivo Apparent viscosity of blood 322 Pressure-flow relationship 323 Incompressibility condition 292 Index of aggregation 315 Indicator dilution method 354 Inertial pressure lose 334 Infants of diabetic mothers Hemorheology of 200 Inflammation 164 Integrin 158 Interspecies differences 268 Capillary diameter 268 Cytoplasmic hemoglobin concentration 268 Mean cell volume 268 Mean corpuscular hemoglobin concentration 268 Red blood cell morphology 269 Intravital microscopy 307, 332 Intrinsic viscosity 84 Inverse Fåhraeus effect 335 Iron deficiency anemia 415 KAHRP 99, 100, 107 Korotkoff sound 300, 352 Laboratory investigations Hyperviscosity 377 Laboratory techniques, red cells 257 Density gradient centrifugation 257 Ghost preparation 258 Glutaraldehyde treatment 258 Heat treatment 258 Hematocrit adjustment 258 Sub population separation 257 Laminar flow 293

Laminar shear flow systems 101 Laminar shear rates 25 Concentric cylinder 25 Cone-plate 26 Parallel plate 25 Tube 25 Laser backscatter 253 Laser diffraction 249 Laser Speckle Imaging (SPI) 365 Laser tweezers 101, 103, 104, 106 L-carnitine 174, 210, 220 Leucodepletion Pre-storage leucodepletion 236 Leucoreduction Stored blood 234 Leukocyte 137, 153 Activation 147, 184 Adhesion 153, 341 Cytoskeleton 146 Deformability 137 Filterability 140 Microcirculation 148 Pathology 148 Viscoelastic models 141 Leukocyte-induced alteration of red blood cell rheology 184 Leukocytes 37, 38 Leukostasis 172, 385 Low shear viscosity 53 RBC aggregation 53 RBC rigidity 54 Macro rheology – general 49 RBC aggregation 49 RBC deformability 49 Shear rate effects 49 Tk index 50 Malaria 97, 109, 385 Malignancy 417 Mammals 41 Breed variation 43 Erythrocytes 42 Plasma viscosity 269, 272 Red blood cell aggregation 275, 276 Red blood cell deformability 273, 275 Red blood cell elongation indexes 274 274 SS1/2

451

Whole blood viscosity 269, 271 Marginal zone 335 Margination 161 Margination of leukocytes 335 Mean arterial pressure 292 Mean filtration pressure 314 Measurement difficulties 45 Sedimentation 47 Surface films 46 Torque-time effects 48 Mechanical Fragility Index (MFI) 219 Mechanical fragility 218 Mechanical stress and RBC aggregation 216–218 Mechanical stress and RBC deformability 215 Mechanically induced hemolysis 212–214 Mechanisms of blood damage 206, 207 Megaloblastic anemias 415 Membrane proteins 93 Membrane rigidity 346 Membrane skeleton 94 Micro channels MC FAN 248 Micro pipette 248 Micro rheology – general 55 Continuum model 5 Fahraeus effect 55, 56 Fahraeus-Lindqvist effect 57, 58 Mooney-RabinowitzWeissenberg equation 58 Microaggregates 230 Microangiopathies 385 Micro-cavitation 207 Microcirculation 300, 307 Microcirculatory networks 67 Geometric data 68 Models 68 Whittaker-Winton 68 Microfluidics 104 Micropipette aspiration 102, 138 Micropore filtration technique 311 Microvascular hematocrit 332 Modified RBC properties 124 Aldehyde treatment 126 Enzyme treatment 125 Heat treatment 126 Phospholipids 126

Monoclonal hypergammaglobulinemia 173 Monoclonal immunoglobulins 387 Morphological and rheological characteristics Blood storage 230 Multiorgan failure Blood transfusion 233 Myocardial ischemia 399 Navier-Stokes equations 292 Near Infra Red (NIR) 364 Neonatal Accelerated aging of RBC 196 Hematocrit 191 Leukocytes 196, 197 MHCH 196 Plasma viscosity 192 RBC aggregation 192 RBC deformability 192 Neonatal septicemia Hemorheology in 200 Network flow 60 Bifurcations 61 Flow distribution 62 Flow fraction 62 Flux fraction 62 Non-axial flow 64 Vessel hematocrit 60, 61 NIH. See Normalized index of hemolysis Nitric oxide 334, 335, 340 Normal values 261 Normalized index of hemolysis 213, 214 Ophthalmology 418–422 Optical Coherence Tomography (OCT) 359, 363 Optical traps 103 Optimal hematocrit 172, 373 Orthogonal Polarization System (OPS) 362 Oscillatory flow 295 Overtraining 425 Oxidative damage 346 Oxygen Target oxygen therapy 235 Oxygenators 213 P. falciparum 96–100,103–107,111 P. vivax 97, 107

452

Palpatory method 352 Papaverin 326, 328 Particle image velocimetry 358 Pathological aggregation 183 Pathology 40, 41 Anaemia 40 Erythrocytes 40 Leukaemias 41 Leukocytes 41 Plasma proteins 40 Polycythaemia 41 Perfluorochemical-based blood substitutes 210 Perinatal hypoxia and acidosis 198 Peripheral arterial disease 404 PfEMP1 98–100, 107 Phase separation 303 Placental transfusion 198 Plasma 35, 36 Fibrinogen 36 Ions 35 Metabolic molecules 35 Proteins 35 Viscosity 36, 334 Measurement 243 Plasma composition alterations 172 Plasma free hemoglobin 213, 214 Plasma Hyperviscosity 387 Plasma hyperviscosity syndromes 416 Plasma skimming 331–333 Plasma volume 371 Platelet Hyperactivity 385 Platelet 153 Adhesion 153, 341 Aggregation 160 Platelets 38 Poiseuille 307 Poiseuille flow 292 Poiseuille’s law 289, 307, 322 Poloxamer 328 Polycythemia 172, 413 Due inappropriate erythropoietin production 383 Due to hypoxia 382 Due to increased red cell mass 381 Due to plasma volume contraction 383 Polycythemia rubra vera 379 Smoker’s polycythemia 384

Polycythemia in the neonate 197 Polymer-induced aggregation 120, 121 Pregnancy 375, 427 Blood volume and haematocrit changes 375 Pressure drop 307 Pressure-flow relationship Effect of erythrocyte aggregation 326 Effect of erythrocyte deformability 325 Effect of hematocrit 323 Effect of plasma viscosity 324 Primary open-angle glaucoma 421 Protection of RBC from mechanical trauma 219 Pulmonary diseases 409–411 Pulsatile flow 72 Pulsatility 296 Pulse propagation 297 Raynaud’s phenomenon 405 Reactive oxygen species 342 Red blood cell aggregation 180–183 Alterations in red blood cell properties 181, 182 Fibrinogen 180 Plasma composition alterations 180 Red blood cells 91, 310 Red blood cell deformability Alterations 174–180 Effect of adrenaline 180 Effect of cellular metabolism 176, 177 Effect of insulin 180 Effect of mechanical forces 179 Effect of nitric oxide 179 Effect of osmotic pressure 178 Effect of pH 179 Lipid composition 176 Local metabolic changes 178 Membrane alterations 175 Microenvironmental influences 178 Oxidative damage 177, 178 Red cell aggregates 314 Red cell aggregation 253, 314 Erythrocyte sedimentation rate 254 Hyperaggregation 254 Kinetics 256 Low shear viscometry 254

453

Rouleaux 253 Ultrasound backscatter 255 Red cell deformability 247 Clinical significance and clinical disorders 385 Deformation curve 249 Filterability 248 Flow cytometry 252 Laser diffraction pattern 249 Malaria tropica 247 Micro channels 248 Micropipette 248 Relaxation time constant 253 Shape 252 Sickle cell anemia 247 Red cell mass 371 Red cell membrane tank tread motion 253 Red cell orientation 253, 259 Red cell shape recovery 252 Time constant 256 Redox 342 Regulation of blood flow 292 Relaxation time, Maxwell 75, 77 Resistance 308, 310 Retinal vascular occlusions 420 Reynolds number 293 Rheological drugs 433 Rheoscope ARCA 259 Rheoscopy 101 Rosetting 100 Rotational viscometer Cone-plate geometry 244 Couette system 244 Mooney chamber 246 Plate-plate geometry 244 Rouleaux formation 253 Time constant 256 Rouleaux 36, 114, 314 Fibrinogen 36 Secondary blood hyperviscosity 345 Selectin 158 Septic shock 426, 427 Serum viscosity Measurement 243 Shear rate 314 Shear rate level 243 Shear stress 309, 342

Shear thinning 308 Sickle cell disease 411 Sidestream Dark Field (SDF) 362 South East Asian (Melanesian) ovalocytosis 95 Spherocytosis 385 Spleen Role in blood volume control 374 Role in blood volume regulation 372 Sports anemia 179 Standardization 261 Anticoagulant 261 In vitro time 261 Reference standards 261 Temperature 261 Steric interactions 128 Stored red cells 228 Microcirculation 232 Morphology 231 Stored blood lesion Clinical significance 233 Stress polycythaemia 384 Sub-lethal RBC damage 214–218 Suspension viscosity 26, 27 Casson equation 27 Einstein equation 27 Quemada equation 27 Syllectometry 253 Thalassemia syndromes 412 Thalassemias 174 Theoretical and computational models 105 Therapeutics for RBC rheological diseases 106 Thermodilution method 355 Thixotropy 83 Thrombocythaemia 385 Thrombocytopenia 385 Thrombocytosis 385 Thrombosis 164 Thrombotic thrombocytopenic purpura 385 Tissue hematocrit 332 Tissue imaging 362 Tonometry 353 Total peripheral resistance 292 TRALI. See Acute lung injury Transfusion hazards 228

454

Transfusion reactions mechanisms 229 Transfusion Related Immunomodulation (TRIM) 229 Transit time 311 Treatment of Polycthemia 434 Primary hemorheological disorders 434–437 Red cell rheological abnormalities 436 Rheological disorders of white cells 437 White cell adhesion disorders 437 Troxeuritine 327 Tube flow, oscillatory 86, 87 Tube hematocrit 309 Turbulent flow 293 Ultrasonic flowmeter 357 Unsteadiness parameter 296 Vascular control mechanisms 325, 334 Vascular disease 392–409 Cardiovascular 397–401 Cerebrovascular 402–404 Diabetes mellitus 407–409 Hyperviscosity 377 Peripheral disease 404–407 Vasoactive drugs 433 Vasodilation 341 Veins Hemodynamics 299 Velocity profile 291 Venoconstriction 372 Ventricular assist devices 208 Venules 316 Vertical versus horizontal tubes 326 Viscoelasticity 245

Viscoelasticity of blood. See Blood viscoelasticity Viscometer type Capillary 244 Couette system 244 Falling ball 245 Oscillatory flow 245 Rolling ball 245 Rotational 244 Viscometers 27, 244 Concentric cylinder 30, 31 Cone-plate 32 Contraves Prorheo LS-300 246 Rheolog 246 Tube 27–30 Vilastic bioprofiler 246 ViscoLab 450 247 Wells-Brookfield 246 Viscometry 100, 101 Viscoreceptor 373 Viscosity 343 Volume flow rate 291 Waldenström’s macroglobulinemia 173 Wall shear stress 299 Wave speed 297 Arterial 298 White blood cells 310 White cell active drugs 441–443 Whittaker and Winton 322 Whole blood viscosity. See Blood viscosity Womersley number 296 X-ray PIV 361 Yield Shear Stress 52 Casson equation 52, 53 Yield Stress 76

455


Author Index Alexy, T. Antaki, J.F. Baskurt, O.K. Bogar, L. Boisseau, M.R. Chien, S. Cokelet, G.R. Cooke, B.M. Cranmer, S.L. Czopf, L. Forconi, S. Goedhart, P.T. Goldsmith, H.L. Gori, T. Hardeman, M.R. Henderson, N.M. Horvath, B. Isbister, J.P. Kameneva, M.V. Kenyeres, P. Kesmarky, G.

392, 411 206 vii, 170, 267, 322 409 433 v 21, 45 91 153 409 339 242, 351 v 339 vii, 242, 351 72 392 228, 371 206 404 392

Koltai, K. 433 Lim, C.T. 91 Linderkamp, O. 191 Lipowsky, H.H. 307 Marton, Z. 418 Meiselman, H.J. vii, 21, 45, 114, 322 Nash, G.B. 137, 153 Neu, B. 114 Niimi, H. 351 Pecsvarady, Z. 404, 433 Pries, A.R. 289 Rampling, M.W. vii, 3, 34 Secomb, T.W. 289 Shin, S. 242, 351 Szapary, L. 402 Thurston, G.B. 72 Toth, K. 392, 433 Tran-Son-Tay, R. 137 Vekasi, J. 418 Windberger, U. 267




Handbook of Hemorheology and Hemodynamics

Recommend Documents