Practical Transfusion Medicine, 2nd edition

Practical Transfusion Medicine EDITED BY

Michael F. Murphy MD, FRCP, FRCPath Professor of Blood Transfusion Medicine University of Oxford Consultant Haematologist National Blood Service and Department of Haematology The John Radcliffe Hospital, Oxford

Derwood H. Pamphilon MD, MRCPCH, FRCP, FRCPath Consultant Haematologist Institute for Transfusion Sciences National Blood Service Bristol

FOREWORD BY

D.J. Weatherall Second edition

Practical Transfusion Medicine

Practical Transfusion Medicine EDITED BY

Michael F. Murphy MD, FRCP, FRCPath Professor of Blood Transfusion Medicine University of Oxford Consultant Haematologist National Blood Service and Department of Haematology The John Radcliffe Hospital, Oxford

Derwood H. Pamphilon MD, MRCPCH, FRCP, FRCPath Consultant Haematologist Institute for Transfusion Sciences National Blood Service Bristol

FOREWORD BY

D.J. Weatherall Second edition

© 2005 by Blackwell Publishing Ltd Blackwell Publishing, Inc., 350 Main Street, Malden, Massachusetts 02148-5020, USA Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK Blackwell Publishing Asia Pty Ltd, 550 Swanston Street, Carlton, Victoria 3053, Australia The right of the Author to be identified as the Author of this Work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. First published 2001 Second edition 2005 Library of Congress Cataloging-in-Publication Data Practical transfusion medicine / edited by Michael F. Murphy, Derwood H. Pamphilon. — 2nd ed. p. ; cm. Includes bibliographical references and index. ISBN 1-4051-1844-X 1. Blood— Transfusion. [DNLM: 1. Blood Transfusion. 2. Blood Grouping and Crossmatching. 3. Communicable Disease Control. 4. Specimen Handling. WB 356 P8957 2005] I. Murphy, Michael F. (Michael Furber) II. Pamphilon, Derwood H. RM171.P727 2005 615¢.39— dc22 2004016676 ISBN-13: 978-1-4051-184-46 ISBN-10: 1-4051-184-4X A catalogue record for this title is available from the British Library Set in 9.5 on 12 pt Sabon by SNP Best-set Typesetter Ltd., Hong Kong Printed and bound in India by Gopsons Paper Ltd, Noida Commissioning Editor: Maria Khan Production Editor: Rebecca Huxley Production Controller: Kate Charman Project Manager: Richard Lawrence For further information on Blackwell Publishing, visit our website: http://www.blackwellpublishing.com The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards.

Contents

List of Contributors, vii Foreword, xi Preface to the second edition, xiii Preface to the first edition, xv

Part 1 Basic principles of transfusion 1 Introduction Ian M. Franklin, 3 2 Essential immunology for transfusion medicine Willem H. Ouwehand and Tim B. Wallington, 13 3 Human blood group systems Geoff Daniels, 24 4 Human leucocyte antigens Cristina V. Navarrete, 34 5 Platelet and neutrophil antigens David L. Allen, Geoffrey F. Lucas, Willem H. Ouwehand and Michael F. Murphy, 50

Part 3 Complications of transfusion 13 Haemolytic transfusion reactions Sue Knowles and Geoff Poole, 161 14 Febrile reactions and transfusion-related acute lung injury Michael F. Murphy and Sheila MacLennan, 171 15 Urticarial and anaphylactic reactions David J. Unsworth, 179 16 Bacterial contamination Patricia E. Hewitt, 184 17 Post-transfusion purpura Michael F. Murphy, 191 18 Immunomodulation and graft-versus host disease Lorna M. Williamson and Cristina V. Navarrete, 195 19 Transfusion-transmitted infections Alan D. Kitchen and John A.J. Barbara, 208 20 Variant Creutzfeldt–Jakob disease Marc L. Turner, 229

Part 2 Clinical transfusion practice 6 The effective and safe use of blood components Brian McClelland and Tim Walsh, 67 7 Bleeding associated with trauma and surgery Beverley J. Hunt, 86 8 Prenatal and childhood transfusions Irene Roberts, 97 9 Haematological disease Michael F. Murphy and Simon J. Stanworth, 119 10 Transfusion strategies in organ transplant patients Derwood H. Pamphilon, 132 11 Inherited and acquired coagulation disorders Joanne E. Joseph and Samuel J. Machin, 138 12 Uses of intravenous immunoglobulin David J. Unsworth and Tim B. Wallington, 151

Part 4 Practice in blood centres and hospitals 21 Donors and blood collection Liz Caffrey and Moji Gesinde, 241 22 Blood donation testing and the safety of the blood supply David Wenham and Simon J. Stanworth, 250 23 Production and storage of blood components Lorna M. Williamson and Rebecca Cardigan, 259 24 Medicolegal aspects Patricia E. Hewitt, 274 25 Blood transfusion in hospitals Sue Knowles and Geoff Poole, 280 26 Autologous transfusion Dafydd Thomas, 298 27 Tissue banking Deirdre Fehily and Ruth M. Warwick, 309 v

Contents

28 Cord blood banking Ruth M. Warwick, Sue Armitage and Deidre Fehily, 320 29 Therapeutic apheresis Tim B. Wallington and David J. Unsworth, 328

Part 5 Developments in transfusion medicine 30 Blood substitutes Chris V. Prowse and David J. Roberts, 341 31 Cytokines in transfusion practice Derwood H. Pamphilon, 350 32 Haemopoietic stem cell processing and storage David H. McKenna and Mary E. Clay, 357 33 Haemopoietic stem cell transplantation and immunotherapy Ian M. Franklin, 369 34 Gene therapy Colin G. Steward and Marina Cavazzana-Calvo, 390

vi

35 Recombinant antibodies and other proteins Marion Scott, 403 36 Blood transfusion in a global context David Roberts, Jean-Pierre Allain, Alan Kitchen, Stephen Field and Imelda Bates, 415 37 The design of interventional trials in transfusion medicine Paul Hébert, Alan Tinmouth and Dean Fergusson, 424 38 Getting the most out of the evidence for transfusion medicine Simon J. Stanworth, Susan J. Brunskill and Chris J. Hyde, 436 39 The future of transfusion medicine Walter Sunny Dzik, 445 Index, 457 Colour plates are found between pp. 304 and 305

List of Contributors

Jean-Pierre Allain

Rebecca Cardigan

Dean Fergusson

Division of Transfusion Medicine National Blood Service University of Cambridge Long Road Cambridge, CB2 2PT

National Blood Service Cresent Drive Brentwood Essex CM15 8DP

Centre for Transfusion Research University of Ottawa Ottawa, Ontario K1M 8L6 Canada

David L. Allen

Marina Cavazzano-Calvo

National Blood Service John Radcliffe Hospital Headley Way Headington Oxford, OX3 9DU

Director of Biotherapy Hopital Necker-Enfants Malades 149 Rue des Sevres Cedex 15 Paris 75743 France

Susan Armitage

Mary E. Clay

National Blood Service Cord Blood Bank Deansbrook Road Edgeware Middlesex, HA8 9DB

Department of Laboratory Medicine & Pathology MMC 198 Room D-251 Mayo Building University of Minnesota Medical School 420 Delaware Street SE Minneapolis, MN 55455 USA

Stephen Field National Blood Service North London Colindale Avenue Colindale London, NW9 5BG

Ian M. Franklin

John A.J. Barbara National Blood Service North London Colindale Avenue Colindale London, NW9 5BG

Geoff Daniels Bristol Institute for Transfusion Sciences National Blood Service Southmead Road Bristol BS10 5ND

Imelda Bates Liverpool School of Tropical Medicine Pembroke Place Liverpool L3 5QA

Susan Brunskill National Blood Service John Radcliffe Hospital Headley Way Headington Oxford, OX3 9BQ

Walter Sunny Dzik Massachusetts General Hospital Blood Transfusion Service, J-224 55 Fruit Street Boston, MA 02446 USA

Deirdre Fehily National Transplant Centre Via Giano della Bella 00161 Rome Italy

Department of Medicine University of Glasgow Royal Infirmary Glasgow G31 2ER

Moji Gesinde National Blood Service Leeds Blood Centre Bridle Path Leeds LS15 7TW

Paul Hébert Department of Medicine The Ottawa Hospital/General Campus 501 Smyth Road Room 1812H, Box 201 Ottawa, ON K1H 8L6 Canada

Patricia E. Hewitt National Blood Service North London Colindale Avenue Colindale London, NW9 5BG

Elizabeth Caffrey

Beverly J. Hunt

National Blood Service University of Cambridge Long Road Cambridge CB2 2PT

Department of Haematology & Rheumatology Guy’s and St Thomas’ Foundation Trust Lambeth Palace Road London SE1 7EH

vii

List of Contributors

Christopher J. Hyde

David H. McKenna

David J. Roberts

National Blood Service John Radcliffe Hospital Headley Way Headington Oxford, OX3 9BQ

Clinical Cell Therapy Laboratory University of Minnesota Medical School 1900 Fitch Avenue St Paul, MN 55108 USA

National Blood Service John Radcliffe Hospital Headley Way Headington Oxford, OX3 9BQ

Joanne E. Joseph

Sheila MacLennan

Irene A.G. Roberts

Department of Haematology and Stem Cell Transplantation St Vincent’s Hospital Victoria Street Darlinghurst, NSW 2010 Australia

National Blood Service Leeds Blood Centre Bridle Path Leeds LS15 7TW

Department of Haematology Commonwealth Building, 4th Floor Imperial College Hammersmith Campus Du Cane Road London, W12 0NN

Michael F. Murphy

Marion Scott

National Blood Service John Radcliffe Hospital Headley Way Headington Oxford, OX3 9BQ

Bristol Institute for Transfusion Sciences National Blood Service Southmead Road Bristol BS10 5ND

Cristina V. Navarrete

Simon J. Stanworth

National Blood Service North London Colindale Avenue Colindale London, NW9 5BG

National Blood Service John Radcliffe Hospital Headley Way Headington Oxford, OX3 9BQ

Willem H. Ouwehand

Colin G. Steward

National Blood Service Southmead Road Bristol, BS10 5ND

National Blood Service University of Cambridge Long Road Cambridge CB2 2PT

BMT Unit Royal Hospital for Sick Children Upper Maudlin Street Bristol BS2 8BJ

Samuel J Machin

Derwood H. Pamphilon

Dafydd Thomas

Department of Haematology University College London Medical School 3rd Floor Cecil Fleming House Grafton Way London, WC1E 6DB

Bristol Institute for Transfusion Sciences National Blood Service Southmead Road Bristol BS10 5ND

Morriston Intensive Care Unit Swansea NHS Trust Morriston Hospital Swansea

Brian McClelland

Geoff Poole

Edinburgh and SE Scotland Blood Transfusion Centre National Science Laboratory 21 Ellen’s Glen Road Edinburgh, EH17 7QT

National Blood Service Southmead Road Bristol BS10 5ND

Alan D. Kitchen National Blood Service North London Colindale Avenue Colindale London, NW9 5BG

Susan Knowles Department of Haematology Epson & St Helier University Hospitals NHS Trust Wrythe Lane Carshalton Surrey, SM5 1AA

Geoffrey F. Lucas

Alan Tinmouth

Christopher V. Prowse Scottish National Blood Transfusion Service National Science Laboratory 21 Ellen’s Glen Road Edinburgh EH17 7QT

viii

Centre for Transfusion Research University of Ottawa Ottawa, Ontario K1H 8L6 Canada

List of Contributors

Marc L. Turner

Tim Walsh

David Wenham

Edinburgh and SE Scotland Blood Transfusion Centre Royal Infirmary of Edinburgh 51 Little France Crescent Old Dalkeith Road Edinburgh, EH16 4SA Scotland

Department of Anaesthetics Critical Care and Pain Medicine New Royal Infirmary of Edinburgh Little France Edinburgh EH16 4SU

National Blood Service North London Colindale Avenue Colindale London, NW9 5BG

Ruth M. Warwick

Lorna M. Williamson

National Blood Service Tissue Services Deansbrook Road Edgeware Middlesex, HA8 9DB

Division of Transfusion Medicine National Blood Service University of Cambridge Long Road Cambridge, CB2 2PT

David J. Unsworth National Blood Service Southmead Road Bristol BS10 5ND

Tim B. Wallington National Blood Service Southmead Road Bristol BS10 5ND

ix

Foreword

Although we now take blood transfusion very much for granted, and it is an integral part of clinical practice, the early days of its development were anything but smooth. Indeed, it seems likely that it spawned two of the earliest documented cases of scientific fraud and medical malpractice. In 1654 a Florentine physician, Francesco Folli, claimed that he had invented blood transfusion and even published a book many years later to illustrate the complex equipment which he had used. He subsequently confessed that he had not yet done the experiment and, as far as is known, he never did! The first well-documented blood transfusions were carried out in 1667, in Oxford and Paris. The Oxford experiments were the work of the physician, Richard Lower, who was stimulated by the studies of the architect, astronomer and polymath Christopher Wren, who had invented a series of cannulas for injecting drugs into the veins of animals. Lower’s first successful transfusion was from the cervical artery of one dog into the jugular vein of another, previously exsanguinated. Perhaps stimulated by news of the first transfusion involving a human being in Paris in the same year, two years later Lower injected a small amount of sheep blood into a mildly deranged clergyman before an admiring audience at the Royal Society. The patient survived and claimed to feel better. In the same year a French physician, Jean-Baptiste Denis, began a series of experiments in which he transfused varying amounts of animal blood into patients with mental illnesses. Things seemed to go well until he gave repeated injections of the blood of ‘a gentle calf’ to a lunatic; the patient had a typical transfusion reaction and, although he recovered temporarily, died two months later. Denis’s enemies persuaded the patient’s wife to bring a legal action against him but, in the event, the defence was successful in proving that the man had been poisoned with arsenic by his wife! Readers of this book might wish to remind their counsels of this possibility next time they are

facing legal proceedings for a mismatched transfusion. Although over the next two centuries there was considerable progress in developing better ways of transferring blood from one individual to another, it was the beautifully elegant studies of Karl Landsteiner, carried out just over 100 years ago, that formed the basis for modern immunohaematology and the successful development of blood transfusion. It is a remarkable fact that some of the most important discoveries that have changed medical practice have been based on extremely small-scale and simple experiments; the demonstration of the efficacy of penicillin required only eight mice, four treated and four controls. Landsteiner described blood groups A, B and C (later called O), using serum and red cells from six healthy males; the results were confirmed with sera from 16 other healthy individuals. Group AB was discovered a year after Landsteiner’s classical experiment, the M, N and P groups were reported by Landsteiner and Levine in 1927, and the rhesus system was characterized by Levine and Weiner in the early 1940s. More than 250 red cell antigens have now been described and most belong to one of 29 systems. Many of the genes that regulate these systems have been identified and much is known about the structure of the blood group antigens and the molecular basis for their diversity. Yet despite all this sophisticated knowledge, and equally tantalising information from studies of varying susceptibility of individuals with different blood groups to a wide variety of diseases, we still don’t know why we have blood groups and in many cases have little understanding of their biological function. Since the Second World War blood transfusion medicine has changed dramatically and is likely to undergo even more dramatic developments as the new millennium evolves. Rapid progress towards the definition of subpopulations of stem cells, an increasing ability to alter the properties of cell popxi

Foreword

ulations by recombinant DNA technology and the vista, if distant, of specific organ therapy based on work on human embryonic stem cells all point to an extremely exciting future for the field. There seems little doubt that, given their expertise in handling and storing cells, blood transfusion specialists will play an increasing role in the practical applications of these new advances in cell biology. Any young person who enters the field over the next few years can be guaranteed an exciting future. The second edition of this fine book provides a comprehensive and practical account of modern blood transfusion practice. While encompassing descriptions of some of the scientific developments

which are occurring at the fringes of the speciality, it focuses mainly on the problems that are encountered daily in transfusion medicine. Because this field abuts on almost every aspect of clinical practice it should be of value to a wide range of clinicians as well as to students and practitioners of transfusion medicine. Whether by accident or intent the first edition of this book appeared exactly 100 years after Karl Landsteiner made his seminal observations. What better tribute could there have been to one of the most beautifully simple and clinically important experiments in the history of medicine, a view confirmed by the early appearance of this second edition. I wish it all the success it deserves. D.J. Weatherall Oxford, July 2004

xii

Preface to the second edition

This second edition has become necessary because of rapid changes in transfusion medicine over the last three years. The pace of change seems likely to increase with new scientific and technological developments, the challenge of ‘emerging’ pathogens, and renewed efforts to improve clinical transfusion practice. In the UK, the implications of the probable transmission of variant Creutzfeldt–Jakob disease by blood transfusion are wide-ranging across the whole transfusion chain from donor to patient. The primary aim of the second edition remains the same as the first, that is to provide a comprehensive guide to transfusion medicine. The book includes information in more depth than contained within handbooks of transfusion medicine, and is presented in a more concise and ‘userfriendly’ manner than standard reference texts. The feedback we received on the first edition from reviews and colleagues was that this objective was achieved, and that we had provided a consistent style and format throughout the book. We have strived to maintain this to provide a text that will be useful to the many clinical and scientific staff, both established practitioners and trainees, who are involved in some aspect of transfusion medicine and who require an accessible text. The book is again divided into five sections which systematically take the reader through the

principles of transfusion medicine, the use of transfusion in specific clinical areas, its practical aspects in blood centres and hospitals, the complications of transfusion and potential advances. This latter section in the first edition was particularly well received, and it is expanded in the second edition with new chapters on stem cell processing, recombinant antibodies/proteins, transfusion in the tropics, design of clinical trials in transfusion medicine, and a final chapter reviewing advances since 1995 and ‘horizon scanning’ about likely future developments up to 2010 and beyond. We are very grateful to the colleagues who have contributed to this book at a time of continuing change. Although, as with the first edition, most authors work in the blood services in the UK, contributors for the second edition include those in full-time clinical practice, and colleagues from outside the UK to provide a broader perspective. We acknowlege the contribution to Practical Transfusion Medicine of two colleagues, Cynthia Beatty and Gail Williams, who were unable to update their chapters because of new commitments. We are grateful to Janet Birchall and Simon Stanworth for providing critical comments on several chapters, and to Helen Williams for her invaluable assistance. We have again received enormous support from our publishers, particularly Maria Khan, Rebecca Huxley and Claire Bonnett. Michael Murphy Derwood Pamphilon 2004

xiii

Preface to the first edition

Blood transfusion continues to enjoy an ever increasing public profile. This has occurred in part because of the emergence of new pathogens which have posed a significant threat to the safety of the blood supply, and also due to major scientific developments. In the new millennium advances in technology have facilitated the provision of highquality blood components and a range of sophisticated diagnostic and specialist services within modern blood centres. There has been enormous progress in transfusion medicine which has developed into a specialist area of its own in the last decade. It now encompasses many important areas of medicine including haematology, immunology, transplantation science, microbiology, epidemiology, clinical practice and research and development. In this book we have aimed to provide a comprehensive guide to transfusion medicine. This includes information in more depth than contained within handbooks of transfusion medicine, but at the same time presented in a more concise and ‘user-friendly’ manner than standard reference texts. Ably assisted by many expert colleagues, we have compiled a text which should prove invaluable to haematologists in training as well as consultants in established practice. We have also aimed to provide useful information to oncologists, surgeons, anaesthetists and other clinicians, nursing staff in general and specialist units and scientific and technical staff in haematology and blood transfusion.

We have endeavoured to provide information that defines practical approaches to the problems that are encountered in transfusion medicine. To this end we have used a consistent format to make access to information easy, irrespective of whether the book is read cover to cover by haematologists updating or revising for exams, or used as a reference book by clinical or laboratory staff faced with specific problems. To facilitate this approach the book is divided into five sections which systematically take the reader through the principles of transfusion medicine, the use of transfusion in specific clinical areas, its practical aspects in blood centres and hospitals, the complications of transfusion and potential advances, some of which are already with us and some of which will continue to impact significantly on transfusion services in the future. We are grateful to the colleagues who have contributed to this book at a time of rapid development and considerable organizational change in healthcare as a whole but specifically within blood services in the UK. We are indebted to Bridget Hunt and Susan Sugden for their patience and forbearance; without their invaluable assistance in compiling the text this book would not have been possible. We have received enormous support from our publishers, particularly Andrew Robinson, who gave us considerable assistance at a time when this book was at its early conceptual stages, and Marcela Holmes whose wisdom and expertise have been invaluable in its completion. Michael Murphy Derwood Pamphilon 2001

xv

Part 1

Basic principles of transfusion

Chapter 1

Introduction Ian M. Franklin

Nearly 4 years have elapsed since the first edition of this book. Has anything occurred in the world of transfusion medicine to alter the concepts that were important then? In the past 12 months, a number of crucial events have occurred that have again acted to increase global anxiety about the safety of blood transfusion. The arrival of severe acute respiratory syndrome (SARS) and its prompt recognition as a novel coronavirus in early 2003 focused attention on the difficulties of maintaining blood safety in the face of an unknown emerging infection. In the absence of any knowledge of the epidemiology of the infection, it had to be assumed that there was the potential for SARS to be transmitted by blood. This remains an unresolved issue that will have to await a better understanding of the virus, which perhaps may be obtained in any new outbreak in 2004 or later. Anxieties over SARS were followed quickly by the expected US summer epidemic of West Nile virus (WNV), known to be a transfusion-transmitted infection, and for which precautions in the USA and Europe were urgent and needed to be robust. These included the use of nucleic acid testing for WNV genome in all donations in the USA, Canada and Mexico. In Europe, recent visitors to North America were not accepted as donors for 4 weeks after return. Most recently, after a few years in which the expected major epidemic failed to materialize, the possibility that variant Creutzfeldt–Jakob disease (vCJD) may well be a transfusion-transmitted infection in humans became more likely, following worrying results in sheep transfusion studies some years ago. A patient, one of only 48 known to be at risk through receiving a labile blood component from a donor who later developed vCJD, devel-

oped and died of vCJD in 2003, 7 years after receiving the blood. The donor was healthy at the time of donation in 1996, but became unwell and died of vCJD in 2000. In the absence of a blood test for vCJD, and with no way of confirming that the two patients had the same or different ‘strains’ of vCJD, this is not conclusive evidence for transmission. On the balance of probabilities, however, it seems likely that the transfusion recipient acquired vCJD from the blood transfusion. This event has triggered a further round of new initiatives in the UK to protect blood safety and retain confidence in the transfusion of blood. In addition to leucocyte depletion of blood components and importing both plasma for fractionation and fresh frozen plasma (FFP) for those born after 31 December 1995, it appears likely at present (January 2004) that the exclusion of donors who have received a transfusion in the UK since 1980 will be added to this list. Renewed efforts to reduce inappropriate transfusion because of concerns about the impact of this new measure on the sufficiency of the blood supply is also probable. Perhaps the one cause for optimism comes from the failure of a massive epidemic of vCJD to develop in the UK, at least to date, and most estimates of the ultimate size of the epidemic have been reduced considerably. This makes it even more important to minimize secondary cases acquired from blood transfusion. As the UK blood services prepare for additional precautions to prevent vCJD, through deferral of transfused persons as donors, fears over ‘chicken flu’ are beginning to dominate the headlines and once more pictures from Asia show citizens wearing masks as they go about their daily lives. Although this is currently topical, it may appear 3

Chapter 1

out of date later in the year and over the next 2–3 years. Other crucial events have included the relentless march of two new technologies aimed at improving blood safety. The first, nucleic acid testing (NAT) for viral pathogens, is already established, although concerns over cost–benefit analyses, at least where NAT is a second-line test to a highly effective antibody detection system, may lead to review. The second, pathogen inactivation (PI) also appeared to be heading for implementation, and one system, Intercept, had been licensed for treatment of plasma in the EU. However, after a few patients developed antibodies to aspects of the agents of the system, a delay in further trials is inevitable until the safety profile can be assessed further. Another, different, PI system has developed similar problems with neoantigen formation. Although other PI systems are continuing to be developed, all of these work by using a chemical agent to prevent nucleic acid replication, and so each must have a potential for antigenicity that will require extensive study before any such system could be introduced for large-scale use. The above events continue to make safety and supply the main priorities of blood services and this has changed little from where they were 5 or even 20 years ago. Therefore, the four key areas considered in the previous edition still appear to be as relevant now as then and are listed below. The four principal areas to be considered are: • blood safety; • the appropriate and effective use of blood and blood products; • donor recruitment and retention; and • informing patients about blood transfusion. The opinions expressed in this introduction are those of the author alone. Blood has been assumed to have mystical qualities from the early days of transfusion experiments in the seventeenth century by Lower in England and Denis in France. A number of predictable disasters caused the subject to fall into disrepute, and progress in transfusion had to wait until there was adequate understanding of blood groups to enable safe transfusions between individuals. The imperatives of the Second World War were also important in emphasizing the need for transfusion services 4

and for providing the clear logistical base from which they might be organized. The early transfusion services, certainly in the UK, were often related to military practice and modern practitioners might be forgiven for believing that the sole objectives were collection, process and supply of (at that time) bottled blood and plasma. There also seems little doubt in retrospect that there was great profligacy in the use of blood and in particular plasma. Some of this stemmed, no doubt, from inadequacies in surgical practice and an equivalent lack of understanding of blood coagulation, but the failure to collect even the most basic evidence of any benefits of blood or plasma transfusions has bedevilled the field ever since. Following a consistent increase from the 1950s, blood usage in the USA has shown a downward trend in the past decade from a peak in 1986, and demand has been decreasing for the last 3 years in the UK. The reasons for this are probably multifactorial, but include improved surgical techniques as well as concerns about blood safety. Despite this, there is evidence for disparities in blood usage between surgeons and between hospitals, for similar activities. There is also wide variation in the use of blood avoidance strategies such as autologous transfusion using cell salvage and preoperative deposit. There is, in the UK, little use of preoperative clinics to enable haemoglobin correction with iron, other haematinics or erythropoietin. In the UK, these issues will be addressed over the next few years by ‘Better Blood Transfusion’ initiatives.

Blood safety Trends in transfusion practice in the past two decades, since the identification of acquired immunodeficiency syndrome (AIDS), have been in the general direction of enhanced safety of plasma products and cellular components, as well as improved purity. With pooled fractionated plasma products there was a shift from low, then to intermediate and eventually to highly purified factor VIII, for example, which provided many benefits in safety and specificity of treatment. Together with the development of the necessary technology, these advances led to the realization that recombinant

Introduction

products, ideally free of any added human or animal proteins such as albumin, were the ultimate expression of the drive towards total safety and absolute purity. This success in improving the safety of plasma products by eliminating donor-derived material seemed to have encouraged the view that the goal of zero risk from transfusion was to be required by regulators and governments. Although there have been no specific statements to change this, a trend seems to be emerging in favour of a ‘balance of risk’ approach. In the Netherlands, the health minister has made clear that optimal, not maximal, safety is the goal. Although it is not clear what this means precisely, the inference is that some form of cost–benefit judgement must be included in the equation for achieving blood safety. The European Union (EU) Commissioner for Health and Consumer Protection, David Byrne, who has a portfolio that includes food and blood safety, stated in a speech entitled ‘Irrational Fears or Legitimate Concerns’ on 3 December 2003 that zero risk cannot be achieved. And in the UK, ministers have begun to question how much must be spent on the safety of the railways before this becomes excessive. The inference is that other areas of public life must achieve a balance between delivering an effective service without crippling costs arising from chasing absolute safety. Prior to these public statements, it appeared that the provision of blood by national blood services was almost unique in the political imperative that required total safety, at whatever cost. This obsession with reducing risks to zero led to there being a perception that there are problems with the safety of blood. Some of this came about because of later criticism of earlier decisions, in particular in the UK over delays in implementing hepatitis C virus (HCV) testing (discussed in detail in the first edition). The failure to introduce the firstgeneration test for HCV antibody led to a delay in the effective testing for this known transfusiontransmitted virus, and there is no question that some patients acquired HCV during this period. This delay was strongly criticized in the judgement in the English courts by Justice Burton, who considered that testing should have been introduced in January 1991 and not September as happened.

One obstacle to early implementation was the desire to have the whole of the UK introducing the test at the same time, so that there would be no difference in quality of component anywhere. Although in an ideal world all parts of an individual blood service would implement testing of a new agent at the same time, this seems inappropriate when a significant delay is introduced thereby. It would seem to be preferable for larger blood services to begin testing as soon as possible in some parts of the service, even if others are not yet ready. At least in this way some donations would be protected. In countries where there is no single authority managing blood services, this already happens. In the UK, leucocyte depletion of all labile blood components was implemented to prevent the then ‘theoretical’ risk that vCJD might be transmitted by blood transfusion. Leucocyte depletion was phased in as soon as it was possible operationally – there was no ‘big bang’ before which components were not leucocyte depleted and after which they were. New virus or other tests and safety measures should be managed similarly in future. The other obstacle to new tests or safety measures is the impact on supply, i.e. on donors. There is no doubt that blood donors are the essential cornerstones of a transfusion service. Nevertheless patients expect and believe that the transfusions or tissues they receive will be as safe as possible, and that those donors who may pose an additional risk to safety should not be accepted. The range of risks for which donors are deferred continues to increase, and significant numbers are turned away because of recent tattoos or body piercings, or travel to areas where there are concerns about old or emerging agents, such as WNV, malaria or Trypanosoma cruzi exposure. Testing is becoming more complex and extensive and the prospect that PI might achieve the same result as more testing, in a single manufacturing process, is most attractive. On the face of it, PI of cellular components, e.g. of platelet concentrates using the psoralen S-59 and ultraviolet A light, holds great promise by preventing virus and bacterial replication. Removing the risk of transfusion-transmitted graft-versus-host disease by preventing T-cell replication would be an added bonus. Unfortunately, the occurrence of antibodies to blood cells produced by two of these 5

Chapter 1

systems may well prove a major, if not fatal, blow to this approach for the time being. Regulation of blood services

Blood service regulation developed following a number of episodes of transfusion-transmitted infections that occurred in the 1960s and 1970s. In the USA the responsible body is the Food and Drug Administration (FDA), through its Center for Biologics and Research (CBER) division. In the UK the regulator has been the Medicines Control Agency, mainly for the production of pharmaceuticals from plasma, as cellular components were not considered to meet the requirement for a ‘product’. This latter nicety was dealt with by Justice Burton in his HCV judgement, which confirmed that cellular blood components were indeed products. The Medicines Control Agency merged with the Medical Devices Agency in 2003 to form the Medicines and Healthcare products Regulatory Agency (MHRA). For Europe, there is an overarching medicines safety body, the European Agency for the Evaluation of Medicinal Products (EMEA), but no unifying EU legislation until the EU Blood Directive entered EU law in January 2003. This will be implemented by the end of 2004 by EU member states, and requires defined standards for all aspects of the blood supply chain. For the first time in the UK there will be a legal requirement to trace blood donations to the recipient, which will take regulators into hospitals for the first time. There are still no legal requirements to consider transfusion alternatives or to implement optimal blood-use programmes. Risk management

Awareness of the importance of protecting patients from potential risk following transfusion has taken a much higher profile recently. Ten years ago there were delays in introducing tests that would clearly have impacted significantly at the 1 in 1000 or 1 in 10 000 level for HCV transmission. Now, blood services in Europe and the USA are implementing tests using nucleic acid amplification by polymerase chain reaction (PCR) where it is possible to detect events with an incidence of 6

between 1 in 300 000 and 1 in 2 000 000. These risks are now perceived to be politically and economically worth preventing. Sir Kenneth Calman, the former Chief Medical Officer of the Department of Health in the UK, addressed issues of risk in a series of articles. He provided examples of activities associated with moderate risk, such as smoking 10 cigarettes a day (1 in 200 chance of death in any one year), to infinitesimal risks, such as being struck by lightning. Schreiber and colleagues, writing on behalf of the US Retrovirus Epidemiology Donor Study, estimated the risk for transfusion-transmitted virus infections at between 1 in 63 000 for hepatitis B virus (Calman risk level, very low) and 1 in 493 000 for human immunodeficiency virus (HIV) (Calman risk level, minimal) (Table 1.1). There are no equivalent figures for the UK, although an estimate for HIV can be made from evidence of only two known transmissions of HIV since the introduction of testing in 1985. During that time some 30 million donations have been transfused. In the first year of testing for HCV RNA using PCR, only one true PCR-positive, HCV antibody-negative donation has been detected, during a period when about 3 million donations were tested. Although PCR testing for HCV RNA was initially introduced for testing plasma donations, it has been a mandatory release criterion for cellular components since 2000, in order to remove a risk of around 1 in 2 000 000 or less. The number of NAT-positive, HCV antibody-negative donations has been very small since then, and the cost of each transfusiontransmitted case avoided has been immense. Why should such minimal or even infinitesimal risks be unacceptable in blood transfusion? There is no doubt that the appalling stigmatization of individuals that occurred during the development of the AIDS epidemic in the USA and Europe has some part to play. Descriptions of transfusiontransmitted infections in the media invariably use words such as ‘tainted’ and ‘contaminated’ in relation to the blood supply. The invasion of the body by an unseen, unknown and unwelcome virus or other agent may explain some of the psychological revulsion. Commissioner Byrne alluded to this issue in his 3 December speech, and suggested that the control that individuals can exert over a risk is

Introduction Table 1.1 Descriptions of risk in relation to the risk of an individual dying (D) in any one year or developing an adverse

response (A). (From Calman 1996 with permission.) Term used

Risk range

Example

Risk estimate

High

> 1 : 100

Moderate

1 : 100–1 : 1000

Low

1 : 1000–1 : 10 000

Very low

1 : 10 000–1 : 100 000

Minimal

1 : 100 000–1 : 1 000 000

Negligible

< 1 : 1 000 000

(A) Transmission to susceptible household contacts of measles and chickenpox (A) Transmission of HIV from mother to child (Europe) (A) Gastrointestinal effects of antibiotics (D) Smoking 10 cigarettes a day (D) All natural causes, age 40 (D) All kinds of violence and poisoning (D) Influenza (D) Accident on road (D) Leukaemia (D) Playing soccer (D) Accident at home (D) Accident at work (D) Homicide (D) Accident on railway (A) Vaccination-associated polio (D) Hit by lightning (D) Release of radiation by nuclear power station

1 : 1–1 : 2 1:6 1 : 10–1 : 20 1 : 200 1 : 850 1 : 3300 1 : 5000 1 : 8000 1 : 12 000 1 : 25 000 1 : 26 000 1 : 43 000 1 : 100 000 1 : 500 000 1 : 1 000 000 1 : 10 000 000 1 : 10 000 000

crucial in the acceptance of that risk. So accepting a lift in a car enables a person to decide whether the car seems roadworthy, the driver sober and likely to be a safe choice. Similarly perhaps with issues such as home versus hospital child birth, where a significant minority of women choose home delivery despite evidence of greater risk. With a blood transfusion, or a food additive, no such choice is possible. The recipient or consumer must accept the safety of the blood or food at face value. If that acceptance is later found to have led to a transfusion-transmitted infection, then anger, compensation claims and litigation are the common responses. Unfortunately, there is no arena in which a dispassionate discussion about blood safety can be held. It would surely be helpful to discuss these important issues in a forum in which patient groups, transfusion professionals, clinical users of blood and those responsible for funding new developments could consider the issues outside the blame culture that follows a perceived transfusion problem. Involvement of ‘stakeholders’ in deciding the principles by which new, expensive and perhaps only moderately effective measures should

be introduced might be interesting and educational for all concerned. Issues to be considered when deciding whether to implement new testing or other safety measures for a transfusion-transmitted infection include the following. 1 Nature of agent being tested for, and the disease it causes. 2 Is there effective treatment? 3 How much does that treatment cost? 4 Is there perceived stigmatization or implications for subsequent lifestyle, e.g. sexually transmissible. 5 What compensation might be payable if no testing is implemented? 6 What is the potential loss of reputation to the blood service? 7 How much does the test or intervention cost? 8 How effective is the test or intervention at preventing future transmission? vCJD precautions

The publication of the results of experiments in sheep which showed that vCJD could be transmitted by whole blood transfusion suggested that it 7

Chapter 1

would be only a matter of time before vCJD was transmitted by blood transfusion between humans. Although the case reported in December 2003 remains only ‘possible’, the acquisition of vCJD by one of only 48 patients being followed who were known to be at transfusion risk is highly suggestive. The extension of the risk reduction measures already introduced (importation of plasma from countries with no or low vCJD, universal leucocyte depletion of fresh components and importation of FFP for children born after 31 December 1995) to include the deferral of all previously transfused donors is imminent. This will put pressure on supplies at a time when donor attendance seems to be falling. As was the case 4 years ago, a more appropriate move in terms of addressing concerns about safety would be a more rigorous process of thought for each and every transfusion, especially in those individuals who are likely to have a long survival after it. This would include all children and those adults who do not have life-threatening diseases, such as candidates for replacement hip surgery. The very large sums of money allocated to vCJD prevention in the UK (£70–100 million per year) might have been better spent, for example, by investing in an educational programme for hospital workers at all levels. The introduction of hospital transfusion teams to ensure that patients get the right blood and are not excessively or unnecessarily transfused would have been another approach that should have been considered.

Appropriate and effective use of blood and blood products Recent studies indicate that the most important effect on the effective use of blood within a hospital or group of hospitals seems to be its culture of transfusion. It has been known for some years that blood transfusion activity is based more on local custom and practice than on evidence. Clear differences exist in transfusion practice and blood usage between individuals and between hospitals. Although there has long been an assumption that blood must be a good thing, recent evidence suggests that even moderate transfusion practices may 8

in fact carry risks. A 1999 randomized trial of red cell transfusion thresholds in the setting of intensive care suggested that less was best, and a systematic review of albumin use in critically ill patients strongly suggested an adverse outcome in those patients who received albumin rather than crystalloids. The vested interests of those on either side of the albumin controversy demonstrated the difficulty of both collecting evidence that would be believed universally and in the acceptance by clinicians of the possibility that they may have been wrong all along. It has been well known for some time that individuals who reject blood transfusion for religious reasons, such as Jehovah’s Witnesses, can undergo open heart surgery with a reasonably high degree of safety. This in itself might suggest that for many years there has been a greatly excessive use of blood (as perioperative red cell transfusions). This is not to say that blood transfusion has not enabled new and innovative surgical procedures to be initiated. Blood remains essential for many cardiac surgery operations and for liver surgery, to cite but two, and of course many patients with malignancy could not receive chemotherapy without the use of blood components to support them. Even in situations where blood transfusion is life-saving, risks remain from errors in the transfusion process, leading to the ‘wrong blood [being] given’, issues highlighted in the UK Serious Hazards of Transfusion (SHOT) reports. In the face of an increasing body of evidence suggesting that blood transfusion carries both known and unknown risks, surely we should be seeking to eliminate unnecessary transfusion. Evidence of the clear benefits of red cell transfusion from good randomized trials is lacking, although there is now evidence that patients with cardiac decompensation tolerate anaemia badly and do benefit from transfusion to higher haemoglobin levels. It is therefore incumbent upon clinicians to think once, twice and three times before transfusing patients and serious consideration must be given to involving patients in these decisions (see below). One thing that does seem clear is that now vCJD seems likely to be transmissible through blood transfusion then there will be an interest in each and every transfusion received by a person who contracts

Introduction

vCJD (or even tests ‘positive’ for it if and when there is a test). Clinicians responsible for prescribing blood must be able to justify each transfusion. The appropriate and inappropriate indications for transfusion are covered elsewhere in this book. However, the evidence continues to accumulate that there are still hazards of blood transfusion that it is not possible to avoid, and that blood transfusion will never be zero risk. The time is overdue for a concerted effort to reduce the use of allogeneic blood to those situations where it is essential to saving or prolonging life, or the quality of life. Since the previous edition of this book, the four UK Chief Medical Officers convened a further seminar in September 2001 to consider the issue of ‘Better Blood Transfusion’. This was followed in 2002 by a further Health Service Circular (HSC) to the chief executives of hospitals in the UK, setting out an agenda for hospitals to follow. This second seminar was held partly because of a generally disappointing response to the first in 1998! Although the second HSC provides a clear toolkit for implementation of better transfusion practice, hospitals have many competing priorities, and it is still difficult to maintain blood transfusion at a high enough level of urgency for hospitals to respond in a consistent way. The cynic might be forgiven for believing that only if there is a blood shortage, sufficient to impact on surgical activity, will hospitals really tackle the issues of best transfusion practice. Reducing wastage

A discussion about the disparity between the demands for blood placed on transfusion services by clinicians and the true needs of the patients being treated is beyond the scope of this chapter. However, one good first step towards ensuring that there is always sufficient blood would be to check that no blood donation is wasted. Unfortunately, this is far from the case and figures of between 5 and 40% are quoted informally for different regions, hospitals or blood groups. The loss of potential donations begins as soon as a prospective donor arrives to offer a donation. An increasing proportion of individuals who come forward offering themselves as donors are unsuitable for reasons of low haemoglobin, lifestyle issues

known to be associated with a higher risk (e.g. transmissible infectious disease) and other temporary reasons for deferral such as body piercing, tattooing and international travel. Technical difficulties in the process of donation may also impair the percentage of units going forward for patients, such as low volume donations, long donation time and technical problems with leucocyte filtration, to give some examples. Where donors would find wastage unacceptable would be if they were aware that their donation might simply go out of date because nobody had used it or because it had been left carelessly out of a blood refrigerator. Improvements in crossmatch to transfusion ratios are continuing all the time but much more needs to be done because it is imperative that blood is not ‘tied up’ waiting for patients who are very unlikely to need it, and so unavailable to those who do. In this way so-called ‘electronic crossmatching’ holds out much hope and is already implemented safely in many parts of the world. Innate conservatism and lack of investment seem to have inhibited its more widespread acceptance. Many of these measures can be implemented if only there was a sufficient will to do so.

Donor recruitment After the end of the Second World War there was a strong sense of community, and in addition many people worked in large industrial settings with a strong sense of identity. This made blood collection easy, since workplace sessions readily recruited large numbers of willing donors. Gradually, many of the large industries have disappeared, and in their place service sector jobs more widely dispersed geographically have arisen. Competition, the changes in the place of women in society (most now work) and a perception that everyone now has less free time have provided challenges to which blood services have had to adjust. Sometimes these responses have been slow. For too long the premise seemed to be that individuals would tolerate a wait of many hours to donate, and the whole process was very centred towards the blood service collection system rather than donor requirements. Only recently has this been fully 9

Chapter 1

acknowledged as inappropriate and moves towards donation by appointment, improving the processing of donors through the session (‘donor flow’) and an increased emphasis on the professionalism of donor staff have all helped to maintain the donor base. It is essential that transfusion services continue to make it easier and more convenient for individuals to donate. There is now a more mature and active relationship developing between donors and the blood services, and this process should continue since it appears that donors are not solely motivated by general altruism – a non-specific wish to do a good thing – but are aware of specific issues. More might be done to strengthen this, perhaps by using advertising more targeted to providing information about the uses of blood and how it makes a specific difference, over and above the general exhortations such as ‘we can’t operate without you’. Is there really a reduction in altruism in the UK, as has been suggested? There may be a change, particularly in young people who perhaps appear rather more self-obsessed than previous generations. The lack of major conflicts such as wars and other common adverse circumstances, while most welcome, tends to reduce the opportunities for building community spirit. However, on reflection and reviewing some of the literature in the area, it is more likely that it is the change in society in terms of longer working hours and more commitment to careers in early adulthood causing less time to consider or attend for donation that is important. It is up to transfusion services, the healthcare industry in general and government in addition to generate and maintain interest in and awareness of the need for blood donation.

Informing patients about blood transfusion In many countries it has been a specific requirement that informed consent is obtained from each patient prior to blood or plasma transfusion. Difficulties in defining what constitutes informed consent and what information must be imparted are considerable. In the USA, a legal decision has meant that recipients must be given information about the alternatives to allogeneic blood transfu10

sion as part of the consent process. In the UK the consent issue has been considered repeatedly over the past few years and is one area where medical care is lagging well behind what is likely to be considered acceptable in the event of a legal challenge. The biggest difficulty appears to be dissecting the need to obtain informed consent and the resources required to provide staff with the time and expertise to discuss the issues. In the absence of any significant momentum to obtain consent as a matter of good practice, perhaps concerns over shortages of blood, the need to consider alternatives, and potential litigation might encourage some form of dialogue between recipient and the healthcare team. In an era of potential blood shortage, blood conservation measures might achieve importance and preadmission clinics, which would need to be a minimum of 3 weeks prior to surgery, might be one way for this to occur. Discussion of alternatives to transfusion such as correction of anaemia, perioperative salvage or predeposit donation all need time and could be combined with a formal agreement by the patient to receive allogeneic blood if that proved necessary. Certainly the current situation where most patients receive little or no pretransfusion information or advice cannot be allowed to continue for much longer without a real risk of litigation in the future. Also, the lack of information makes it impossible to discover the true opinion of individuals about to have a transfusion, or the likely interest in alternative strategies such as autologous transfusion or other blood conservation approaches, or to deliver them nationally with equity. At present, well-informed individuals in major cities probably have a chance of accessing an autologous blood programme, but certainly not the great majority of potential recipients. The challenge for the transfusion services is to convince themselves and colleagues that delivering information about transfusion really is an imperative. What else might be done in the interim? Literature for patients already exists about blood transfusion and its risks, but these do not always reach the parts of the health service that most need them, i.e. the medical and surgical wards and clinics. Perhaps, rather like package inserts for pharmaceutical products that must contain a patient infor-

Introduction

mation leaflet, a leaflet should be issued with each unit of blood, plasma or platelets and handed to the recipient. This might become tedious for blood and marrow transplant units with recipients of multiple transfusions but may be useful for the majority of patients, or their relatives, who receive blood for major surgery as a single event.

Conclusion The past two decades have seen blood transfusion services in developed nations trying desperately to minimize the risk of the next transfusion-transmitted infection, one of which seems to appear every 5 years or so. Douglas Starr, in his book Blood: an Epic History of Medicine and Commerce, spells out most forcibly the errors of omission and commission made over the years. These were more usually due to a combination of denial and naivety rather than gross negligence. Five years on, and it still makes compulsory reading for anyone working in a senior position in a blood service (see Further reading). Attempts to educate the public about risk will fail as long as blood transfusion mishaps are newsworthy, even where they occur by chance in an otherwise effectively functioning system. The only realistic way forward is to engage all participants in the blood transfusion process in active discussion. The most obvious way to begin such a dialogue would be through a pretransfusion interview that would bring physician/surgeon together with the patient to discuss blood safety, and as an obvious prerequisite would require the blood services to provide training and information for colleagues in hospitals. Such an innovation might just pave the way for a realistic debate about the wisdom of further attempts to reduce the risks of transmission of known viruses by blood transfusion to an unattainable singularity of zero risk. Much of this introduction has focused on the problems and challenges that face blood services as we enter the new millennium. That there are plenty of opportunities as well as threats is certain, and the very dependence of blood services on good manufacturing practice and good laboratory practice is opening doors for crucial collaborations in the related fields of cellular immunotherapy, gene

transfer, tissue engineering and tissue and organ banking. Exciting developments in virus inactivation and in blood cell substitutes continue to provide research opportunities at the clinical interface, and improving the education of donors and patients will provide great opportunities for those in donor and patient care services. Transfusion medicine will continue to be a little like walking through a tropical rainforest, where the known paths are clear but still require careful navigation, and new and unseen threats may still lurk around the next corner to trap the unwary. But just as the rainforest contains a huge biodiversity to keep the most jaded traveller interested, so the field of transfusion medicine can never be anything other than a fascinating and rewarding area in which to work.

Further reading Bird SM. Recipients of blood or blood products ‘at vCJD risk’. Br Med J 2004; 328: 118–19. Calman KC. Cancer: science and society and the communication of risk. Br Med J 1996; 313: 799–802. Calman KC, Royston G. Personal paper: risk language and dialects. Br Med J 1997; 313: 939–42. Chalmers I. Human albumin administration in critically ill patients. I would not want an albumin transfusion (letter, comment). Br Med J 1998; 317: 885. Cochrane IGAR. Human albumin administration in critically ill patients: systematic review of randomised controlled trials (see comments). Br Med J 1998; 317: 235–40. Goodnough LT, Shander A, Brecher MA. Lancet 2003; 361: 161–9. Hebert PC, Wells G, Blajchman MA et al. A multicentre, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion requirements in Critical Care Investigators, Canadian Critical Care Trials Group (see comments). N Engl J Med 1999; 340: 409–17. Hunter N. Scrapie and experimental BSE in sheep. Br Med Bull 2003; 66: 171–83. Llewelyn CA, Hewitt P, Knight RSG et al. Possible transmission of variant Creutzfeldt–Jakob disease by blood transfusion. Lancet 2004; 363: 417–21. McClelland B. Albumin: don’t confuse us with the facts. Rather than fulminating, seek to answer the questions raised (editorial, comment). Br Med J 1998; 317: 829–30.

11

Chapter 1 Schreiber GB, Busch MP, Kleinman SH, Korelitz JJ. The risk of transfusion-transmitted virus infections. N Engl J Med 1996; 334: 1685–90. Serious Hazards of Transfusion. Annual Report 2001–2002. www.shotuk.org Sharp Research Ltd. Altruism and Blood Donation. Qualitative Research Report. London: Research

12

Division, Central Office of Information, November 1998, Report No. SR0015. Starr D. Blood: An Epic History of Medicine and Commerce. London: Little, Brown, 1999. Wu YY, Snyder EL. Safety of the blood supply. Role of pathogen reduction. Blood Reviews 2003; 17: 111–22.

Chapter 2

Essential immunology for transfusion medicine Willem H. Ouwehand and Tim B. Wallington

The immune system is a sophisticated and multilayered defence against infection. It is based on the recognition of non-self in any potential pathogen. Since donor organs are not self, their transplantation from one individual to another is only possible if the obstacles inherent in the recipient’s immune system can be overcome. The transfer of blood components (either as therapy or during the course of pregnancy) is a form of transplantation. This discussion concentrates on the immunobiology of clinical problems that are encountered as a result. There are excellent texts which discuss human immunobiology in detail (see Further reading). The immune system has evolved distinct mechanisms for coping with extracellular pathogens, such as bacteria, based on the production of antibody and with intracellular pathogens, such as viruses, based on the activity of effector T cells. Two essential layers of defence are utilized: • innate immunity, which is primitive in evolution and not single pathogen specific (e.g. mannan-binding protein which binds microbial cell-wall saccharides and Toll-like receptors which can be directly involved in macrophage activation); and • adaptive immunity in higher animals, which adding to innate mechanisms brings specificity and memory to immune responses (e.g. antibody formation in defence against bacteria). Most problems encountered in transfusion medicine are antibody based, i.e. the humoral immune response, and this will be considered in greater detail below.

Cellular basis of the immune response The key effector cells are T cells, B cells and natural killer (NK) cells. The progenitors of T cells, B cells and NK cells are derived from the same haematopoietic stem cells (HSC) that give rise to other types of blood cell. Cells of the monocyte–macrophage series, including Langerhans’ cells and dendritic cells, process and present antigen to both T and B cells. Progenitor cells migrate from the circulation into the epithelial thymus to become T cells. There they interact with the stromal cells and their soluble products to undergo cell division, clonal selection and maturation. In addition they acquire their antigen receptor (T-cell receptor or TCR) and other surface molecules which will determine their function, CD8 on cytotoxic T cells, CD4 on helper T cells. Immature T cells initially express both CD4 and CD8 molecules, which interact respectively with major histocompatibility complex (MHC) class II or class I molecules on thymic stromal cells to influence their maturation into CD4 or CD8 T cells. Through this process self-reactive T cells are removed. Later, when they migrate to the periphery, T cells may undergo selective clonal activation triggered by antigen, which leads to proliferation and maturation. Subsequent function depends on whether the cells carry the CD4 or CD8 antigens. B-cell development is a multifocal process that is concentrated in fetal liver before bone marrow becomes the major haematopoietic organ. Progenitor cells receiving signals from local stromal cells begin to divide and begin the process that will provide an antigen receptor, in this case surface immunoglobulin (SIg). Like T cells, immature B cells are easily tolerated or killed by 13

Chapter 2

premature stimulation via their antigen receptors to prevent damage to self. After migrating from the bone marrow, B cells mature, express SIg antigen receptors, and respond to antigens together with T-cell help from CD4 cells by undergoing proliferation and plasma cell differentiation. NK cells are non-T, non-B lymphoid cells capable of killing virus-infected cells either specifically targeted by the presence of antibody on their surface (antibody-dependent cell-mediated cytotoxicity, ADCC) or through the recognition of changes in the infected cells surface that allow NK cell attack. This mechanism is greatly enhanced by the cytokine interferon (IFN)-g, illustrating the fact that these key effector cells usually act in concert in the defence against infection.

Humoral immune response Antibody

The specific effector molecule is an antibody which is secreted into the extracellular space from plasma cells. It is a unique tetramer made up of two identical heavy and two identical light chains (Fig. 2.1). These combine variability of amino acid sequence and thus variability of tertiary structure (Fab) with N terminal

VH Hinge region CH3

CH1 VL

CH2

S C terminal

CL

S

VL Fc

Fab

VH

Fig. 2.1 Basic structure of an immunoglobulin molecule.

Domains are held in shape by disulphide bonds, though only one is shown. CH1–3, constant domain of an H chain; CL, constant domain of a light chain; VH, variable domain of an H chain; VL, variable domain of a light chain.

14

a constant region (Fc), which allows the molecule to bind target antigen via Fab and trigger effector functions through the Fc portion. These molecules are more generally called immunoglobulins. They also serve as antigen receptors on B cells. Antibody effector functions

The constant regions of the heavy (H) chain of immunoglobulins are responsible for the triggering of an effector pathway. This occurs either: • by binding to appropriate Fc receptors on effector cells, such as leucocytes and mast cells; or • by activation of the complement cascade. There are five immunoglobulin isotypes based on different genes for the C domains of the H chain (Table 2.1). Immunoglobulin G (IgG) and IgA have four and two subclasses, respectively. The immunoglobulin isotypes and subtypes differ significantly in their ability to recruit effector functions (see Table 2.1). This is of clinical significance in transfusion as the ability of antibodies to bring about erythrocyte or platelet destruction varies according to their isotype and IgG subclass. Basis of antibody variability

The molecular biology of this is complex. The genes for the five heavy chains and the l and k light chains are found on separate chromosomes, at band 14q32, 22q11 and 2p11, respectively. Each chain is separately synthesized before being assembled into an antibody molecule. On chromosome 14, which carries the H-chain genes, there are three clusters: approximately 50 variable (V) region genes, which encode the first 95 amino acids of the V portion, more than 20 diversity (D) region genes and six joining (J) region genes. Together these genes encode for H-chain V regions. Like letters of the alphabet they can be joined at random into three-letter ‘words’, thus providing much variability in the receptor portion of the H chain. Similarly, 22q11 and 2p11 have two clusters of genes for the V portion of the k and l light chains, respectively, which can recombine in this way. The incredible diversity of antibody specificity, found even at the level of the germline, is the result of these events, coming together in

Essential immunology Table 2.1 Immunoglobulin classes and their functions.

Structure

Function

Isotype

Heavy chain

Light chain

Configuration*

Complement fixation†

Reaction with FcR

Placental passage

IgM IgG1 IgG2 IgG3 IgG4 IgA1 IgA2 IgD IgE

m g1 g2 g3 g4 a1 a2 d e

k, l k, l k, l k, l k, l k, l k, l k, l k, l

Pentamer Monomer Monomer Monomer Monomer Monomer Dimer in secretion Monomer Monomer

+++ +++ + +++ – – – – –

L M, N, P, L, E P, L M, N, P, L, E N, L, P — — — B, E, L

– ++ +/– ++ + – – – –

so

me

* Pentamer, five basic tetrameric units (in vitro good agglutination); dimer, two basic units; monomer, one basic unit.Two or more basic units are held together by a J chain. † Classical pathway. B, basophils/mast cells; E, eosinophils; L, lymphocytes; M, macrophages; N, neutrophils; P, platelets.

Ch

ro

mo

'Silent' area = intron (V)n

D

J

Cm

Cd Cg3 Cg1 Ca1 Cg2 Cg4 Ce Ca2

14 H The product is VHCm, i.e. an IgM heavy chain with a particular variable region

VDJCm (V)n

J

Ck

2 k

Final product IgMk or IgMl

VJCk (V)n

J

Cl

22 l Fig. 2.2 Genes encoding antibodies (see

text for explanation).

the tertiary structure of the Fab2 portion of the immunoglobulin molecule (Fig. 2.2). The variability of the receptor for antigen on T lymphocytes (TCR) is the product of similar mechanisms. Somatic mutation

Antibody function is further refined in the sequential production of immunoglobulin isotypes as the adaptive immune response matures, as follows.

VJCl

• There is switching in the immunoglobulin isotype mix to molecules that are more effective in neutralizing the wide variety of pathogen types that may be encountered; this process is controlled and driven by helper T lymphocytes. • Mutation at hotspots where the V, D and J genes join and similarly within the V portion of light chains, which refine the shape of the receptor area and thus the specificity of the antibodies produced. 15

Chapter 2

Blood cell antibodies illustrating the above principles

These mechanisms at work are well illustrated by the behaviour of commonly found blood cell antibodies. In a typical T-cell-dependent antiprotein (e.g. anti-D) immune response, the switching of the immunoglobulin isotype from IgM to IgG is associated with an increase in the affinity of the antibody. In the early phase of the response the V region is encoded by genes from the germline and antibody affinity is low. As a result, in the early phase of an anti-D response, a panreactive antibody, reactive with several blood groups, might be detected in panel studies. This reaction is most likely caused by low-affinity IgM antibodies which, although responding to RhD, are able, particularly at 4°C, to react with other blood groups because of an avidity effect. Maturation of the immune response with the selection of B and T cells with greater specificity for RhD results in improved antibody affinity and the disappearance of cross-reactivity. This is the consequence of somatic mutation of the rearranged V gene. In high-affinity anti-D antibodies, up to 20 of the 90 codons encoding the V region have changed from germline. Temperature dependency of antibody– erythrocyte interactions is used as a surrogate marker for antibody affinity. • Low-affinity antibodies generally do not bind sufficiently strongly at 37°C to be detected by agglutination but they do at lower temperatures. They are also generally of no clinical significance. • Antibodies of intermediate and high affinity do remain bound and are detected in the antiglobulin test. Pretreatment of red cells with proteases is also an effective method to reveal the presence of lowaffinity antibodies against, for example, RhD as this reduces the strength of interaction between antibody and cell required for agglutination. Other circumstances can also favour the binding of low-affinity antibodies. Some of the isotypes able to activate complement are detected in haemolysin tests if the antibody is present in excess and the activation of complement is facilitated by lowering the pH of the reaction medium. 16

Antigen in the adaptive humoral immune response

The immune response is driven by antigens which select the lymphocytes that are able to participate. Therefore selective use of V genes in antibody production against a certain antigen might be expected. Studies on the V-gene use of blood cell antibodies support this and have thrown light on certain serological anomalies. Most evidence has been acquired by studies on the molecular structure of the V domains of monoclonal antibodies against the carbohydrate antigen I and the protein antigen RhD on the red cell membrane. These studies suggest that: • there is preferential but not exclusive use of certain V genes in the generation of these specificities; • there is a significant overlap in the amino acid sequence of the V domains of cold agglutinins against the lactosylceramide I and anti-D antibodies; • pathological anti-I cold agglutinins, as observed in the majority of patients with cold haemagglutinin disease, uniquely use the VH gene segment DP63 (V-4.34); • postinfectious polyclonal anti-I antibodies seem to make use of the same VH gene segment, while cold agglutinins with other specificities do not; and • in over 50% of monoclonal IgM anti-D antibodies the VH domain is encoded by the DP63 VH gene, the same as that encountered in pathological anti-I cold agglutinins. It is attractive to speculate that RhD-specific B cells evolve from B cells with anti-I specificity. This suggests that in the germline these cells provide an SIg receptor which is best fit at that stage in the immune response for the tertiary structure presented by D. In this scenario the drift in antibody specificity from anticarbohydrate (anti-I) to antiprotein (anti-D) is best explained by minor changes in the amino acid sequence of the V domains of anti-I antibodies brought about by somatic mutation. Ultimately the low-affinity binding for I is lost. This is also influenced by the switch from IgM to IgG. In serology these structural observations are supported by the functional observations on low-affinity interactions mentioned earlier and by the fact that certain IgM

Essential immunology

monoclonal antibodies used for D typing show reactivity at 4°C with protease-treated RhDnegative red cells. The isotype and subclass of blood cell antibodies are at least partly determined by the chemical nature of the antigen which had stimulated their production. • Blood cell antibodies against carbohydrate antigens are generally IgM or IgG2 and IgG4 or a combination of these. • Antibodies against protein blood group antigens are typically of the IgG class, predominantly IgG1 and IgG3, although autoantibodies can be of the IgA type. This suggests a direct involvement of the antigen. The source of antigen might not of course be red cells if the same structure is shared, so-called crossreactivity. An increase in the titre of anti-I antibodies occurs after infection with Mycoplasma pneumoniae. Some preparations of the vaccine TAB (typhoid, paratyphoid A and paratyphoid B) stimulate anti-A and anti-B and cause not only a rise in agglutinin titre but also the development of immune characteristics. Many of the low-affinity reactions seen in red cell serology reflect part of the response to bacterial antigens, usually carbohydrate. When the affinity and concentration of such antibodies increases above certain thresholds complement-mediated haemolysis can occur, making the phenomenon of clinical significance. T-cell-independent antibody formation

As we have seen, the formation of antibodies by B cells is dependent on interaction with helper T cells. However, some antigens can stimulate certain subsets of B cells (B1 and marginal zone B cells) directly, independent of T-cell help. These cells provide a first line of defence to bacteria by producing antibody specific to bacterial polysaccharide. This route to antibody production is also important in the response to certain red cell antigens. The presence of naturally occurring IgM antibodies against A and B is an excellent example of T-cell-independent antibody formation. In the presence of an intact immune system isoagglutinins to the missing A or B antigens are always found, although there has been no exposure to red

cells carrying these antigens. The response is essentially limited to IgM because T cells are not involved and are not available to induce isotype switching, although there is evidence that some switching can occur in the absence of T cells. The repetitive carbohydrate structures (epitopes) presented by the A- and B-determining portions of the relevant cell surface glycolipids and glycoproteins are structurally the same as bacterial polysaccharide and indeed it is antigens from gastrointestinal bacteria that trigger isoagglutinin production. The isoagglutinins are present from the first months of life. IgG anti-A or anti-B antibodies can also be formed to this stimulus, usually of the IgG2 or IgG4 type. T-cell-dependent antibody formation

Unlike glycolipids and glycoproteins, the formation of antibodies against blood cell membrane proteins is always dependent on interaction with T-helper cells. The immune response to RhD is an example. • RhD is the most immunogenic red cell membrane protein antigen. RhD is a 30-kDa nonglycosylated membrane protein. • Analysis suggests that only short peptide loops, part of the molecule, are displayed on the cell surface. • Ample evidence indicates that anti-D antibodies recognize discontinuous amino acid sequences derived from several of the extracellular RhD loops. • These discontinuous residues come together in the tertiary structure of the RhD protein. Therefore isolation of RhD from the membrane disrupts the majority of B-cell epitopes and reactivity with anti-D antibodies. This is not a repetitive structure as with ABO and the B-cell response requires T-cell help. The response of T-helper cells, like B cells, is antigen specific but triggered in a totally different way. T-helper cells recognize short linear segments of amino acids derived by intracellular digestion from the RhD protein and presented to the helper T cell by the human leucocyte antigen (HLA) class II molecule, which is a cell surface molecule. This is achieved most effectively by professional antigen17

Chapter 2

presenting cells (APCs). APCs belong to a family of cells with diverse anatomical locations and of diverse ontogeny: • Langerhans’ cells in the skin; • interdigitating, follicular and germinal centre dendritic cells in lymph nodes and spleen; and • B cells and macrophages. The antibody response to a red cell antigen like RhD involves the collaboration of at least three cell types and of antigen in two forms, both intact and processed. This enables two important features of the immune response: • an efficient mechanism for tolerance to self antigens enables prevention of the failure to discriminate self from non-self that is the basis of autoimmunity; and • the maturation of the antibody response, through isotype switching, is facilitated. This is illustrated in detail by the specific example from transfusion medicine described below. Human platelet antigen-1a presentation via the HLA class II route

If a human platelet antigen (HPA-1a)-negative mother is carrying an HPA-1a-positive fetus, platelets may enter the maternal circulation and immunize her against HPA-1a. This is the end result of quite complex events and can have disastrous consequences for the fetus. Fetal HPA-1a may be ingested by maternal APCs and degraded by endosomal enzymes like cathepsin G. Short fragments of 12–15 amino acids will be produced in endosomes, which in the trans-Golgi network fuse with HLA class II-containing vesicles. The fusion results in a downwards pH shift, which results in the removal of the invariant chain from the HLA class II molecule (this chain prevents the premature loading of the cleft in the molecule that is used for presentation of the digested antigen). A plethora of peptides will be bound in the HLA class II groove, of which some will have been derived from the fetal GPIIIa. Once migrated to the surface of the APC, specific helper T cells then recognize the change in the HLA class II molecule produced by the peptides to which they are specific and T-cell proliferation will ensue. Cytokines produced by expanding clones of reactive T-helper 18

cells will drive the expansion of HPA-1a-specific B cells responding to intact HPA-1. With time the process of antigen take-up, processing and presentation will pass from the classical APCs to HPA1a-specific B cells. This helps to bring together the complex interaction of cells needed for a mature antibody response, as both the surface molecules needed for antigen presentation to helper T cells and for interaction with intact HPA-1 are present on the same cell set. HLA class II restriction of antibody response

We have seen that there must be a genetic element to an individual’s immune response in that the first encounter with antigen is dependent on the germline V-region genes, which show differences between individuals. Whether or not processed peptide from a particular alloantigen can interact with a particular HLA class II molecule to trigger a T cell is also dependent on genetic variability. In the immune response this is important to the immunogenicity of antigens in individuals. Sometimes the peptide is presented exclusively by a certain HLA class II molecule and a linkage between HLA class II type and antibody response can be observed. The HLA DRB3*0101 restricted response against HPA-1a (GPIIIa, leucine 33) is the best example of an HLA class II restricted response in humans. The very much lower immunogenicity of the antithetical antigen HPA-1b (HPA-1b, proline 33) is most likely explained by a less good fit of the peptide containing the proline at position 33 when compared with the one containing leucine at that position. Antibody-mediated blood cell destruction

Most red blood cell alloantibodies and autoantibodies of the IgG isotype bring about destruction in the extravascular compartment via the interaction of the IgG constant domain with Fcg receptors on cells of the mononuclear phagocytic system. Several receptor types are described. • FcgRI is the most important in blood cell destruction. This is a relatively high-affinity receptor found predominantly on monocytes. The consequence of adherence of IgG-coated red cells to

Essential immunology

FcgRI-positive cells is phagocytosis and lysis. This is usually extravascular and takes place in the spleen. The lysis can be demonstrated in vitro as ADCC. • FcgRII is a lower affinity receptor found on monocytes, neutrophils, eosinophils, platelets and B cells. • FcgRIII is also relatively low affinity and found on macrophages, neutrophils, eosinophils and NK cells. It is responsible for the ADCC demonstrable in vitro with NK cells. • There is also an FcgR on the placenta of a different molecular family which mediates the transfer of IgG into the fetus. The severity of red blood cell sequestration by IgG antibodies is determined by the concentration of antibody, its affinity for the antigen, antigen density and the IgG subclass. IgG2 and IgG4 antibodies are generally unable to reduce red cell survival, while IgG1 and IgG3 are. There is ample evidence in patients with warm-type autoimmune haemolytic anaemia that IgG1 and IgG3 are more effective in causing red cell destruction than IgG2 and IgG4. The level of IgG1 coating of red cells needs to exceed a threshold of approximately 1000 molecules per red cell to cause cell destruction. For a long time it has been speculated that polymorphisms in the genes of the family of Fcg receptors might be significant in causing differences of severity of blood cell destruction observed between patients with apparently similar levels of IgG coating. So far, firm evidence for such polymorphisms has been lacking, although a single amino acid polymorphism of the FcgRIIa receptor dramatically alters the affinity for human IgG2 and additional polymorphisms might have an effect on the interaction with IgG1 and IgG3.

Complement system The complement system, either working alone or in concert with antibody, is important for effective immunity to many extracellular pathogens. It also often plays an important part in immune red cell destruction and can be the reason for important systemic complications of haemolysis. Naturally occurring IgM antibodies against A and B are

often of low affinity and do not bind to red cells at 37°C. However, when ABO blood group antibodies do bind at 37°C, there will be rapid complemant-mediated destruction of incompatible red cells where there is a major A to O or B to O mismatch. This may result from a transfusion error, and remains an important cause of transfusionrelated mortality and morbidity. Blood cell antibodies that can activate complement are more effective in achieving cell destruction than non-complement-activating antibodies. In contrast to extravascular FcgR-mediated destruction, complement-mediated lysis occurs in the intravascular compartment. The ensuing release of anaphylatoxins such as C3a and C5a contributes to the acute systemic effects that occur. IgM, IgG1 and IgG3 antibodies are the most effective isotypes in binding C1q and initiating activation of the complement cascade via the classical pathway. However, they are dependent on aggregation for a sufficiently high antibody density to trigger C1q and overcome the regulators of complement activation that are present. The concentration of antibody may be too low to achieve the density necessary. The antigen topography (e.g. of RhD) can prevent the binding and activation of the C1q molecule. Complement is a complex system of plasma proteins, both part of innate immunity and vital to the effector functions of complement-fixing immunoglobulin isotypes. Central to complement’s function is the activation of C3 as this leads to the opsonization of bacteria (Fig. 2.3). C3 can be activated by three routes: the classical pathway, the alternate pathway and lectin binding. Lysis is dependent on activation, downstream from C3, of components of the membrane attack pathway. The classical complement pathway consists of: • four numbered components (C1–C4); and • two regulatory proteins (C1 inhibitor, C4binding protein). The first component (C1) comprises three subcomponents, C1q, C1r and C1s. It is the interaction between C1q and aggregated IgG or IgM bound to antigen that initiates activation of the classical complement sequence. The fixation of C1q activates C1r and C1s. C1s cleaves C4 and C2, whose 19

Chapter 2 Classical

Alternate

Clqrs

C3b MBL - MASP +

C4b

Factor D

C2b

Factor B C3 Opsonization

C3bBb

C3b4b2b Properdin

C5 Convertases C3b4b2a—classical C3bBbP—alternate

C5 C6 Lysis

Key enzymes C3 Convertases C4b2a—classical C3bBb—alternate

Mediators of inflammation

C7 C8 C9

Final lytic

active fragments C4b and C2a form the classical pathway C3 convertase. The alternative pathway to C3 activation consists of: • C3b, factor B and factor D; and • the regulatory proteins, properdin and factors H and I. Factor B binds to a cleavage fragment of C3, C3b, to form C3bB. Factor D cleaves the bound factor B to form the alternative pathway C3 convertase (C3bBb). It activates C3 in a fashion similar to the C3 convertase of the classical pathway, C4b2a. Properdin acts to stabilize this alternative pathway C3 convertase, as do carbohydrate-rich cell surfaces, by partially shielding the convertase from inhibitors. Activation via the alternative pathway would otherwise be unchecked if it were not inhibited, as it requires no specific stimulus. The lectin pathway is initiated by mannanbinding lectin. This is structurally related to C1q and binds avidly to carbohydrate on the surface of microorganisms. It activates C4 through a serine protease, which is similar to C1r and C1s with the same outcome. The attack pathway is dependent on the formation of the trimolecular complex of C4b2a3b or 20

Ba C2 Kinin C3a C5a

Fig. 2.3 The different pathways for

C567

complement activation. MBL, mannanbinding lectin; MASP, MBL-associated serine protease.

C5 convertase, which cleaves C5 to two fragments C5a and C5b. The former is a potent anaphylatoxin. C5b forms a complex with C6, C7 and C8, which facilitates the insertion of a number of C9 molecules in the membrane. The C5b-8 and the multimeric C9 molecules form the membrane attack complex (MAC), creating a lytic pore in the membrane and lysing the target cell. Cells not immediately involved in the process but close to it can also be lysed by seeded MAC, the so-called bystander lysis. Blood cells coated with C3b will bind to cells carrying receptors for C3b (CR1 or CD35). This adherence can lead to extravascular cell destruction, mainly in the liver, but if the bound C3b degrades to its inactive components iC3b and C3dg before the cell is lysed then the cell is protected from lysis. Membrane-bound molecules such as decay accelerating factor (DAF) and membrane inhibitor of reactive lysis (MIRL) protect red cells from lysis in this way. They are of clinical importance as: • DAF (CD55) and MIRL (CD59) are linked to the blood cell membrane via a glycosylphosphatidylinositol (GPI) anchor; • patients with paroxysmal nocturnal haemoglo-

Essential immunology

binuria (PNH) have an acquired mutation in the PIG-A gene in a subset of their HSC which prevents synthesis of the anchor; • progeny from the affected stem cells lack GPIlinked membrane proteins; and • the absence of DAF and MIRL from red cells increases the sensitivity for complementmediated lysis which occurs when the pH is marginally lowered during sleep, resulting in haemoglobinuria. In vitro acidification is used in Ham’s acid test to reveal the presence of a population of erythrocytes with increased sensitivity for complementmediated lysis. Flow cytometric analysis looking for the absence of GPI-linked proteins on a subset of leucocytes derived from mutated stem cells is an alternative test for the diagnosis of PNH.

Cell-mediated immunity Antigen-specific cell-mediated responses are carried out by T cells. They provide the immune system’s main defence against intracellular microorganisms and can lyse cells expressing specific antigens (i.e. cytotoxicity). In addition they release cytokines that can trigger inflammation and are responsible for delayed hypersensitivity and symptoms usually associated with infection such as fever, myalgia and fatigue. Cytotoxic T cells

Cytotoxicity is the job of cytotoxic T cells, which are distinguished by the presence of CD8 on the cell surface. This facilitates their interaction with HLA class I on the surface of cells altered by the presence of antigen, usually as the result of virus infection. Cytotoxic T cells are the main means of protection against virus infection. They are also important mediators of allograft rejection. Like the B-cell response, the cytotoxic T-cell response requires help from T-helper cells and is regulated by these cells. Delayed hypersensitivity

Delayed hypersensitivity is an example of another

crucial role for helper T cells. It is dependent on the secretion of the cytokines interleukin (IL)-1, IL-2, tumour necrosis factor (TNF) and IFN-g, the socalled proinflammatory or Th1-type cytokines. These recruit inflammatory cells, in particular macrophages, to sites of infection and arm them to kill certain bacteria, e.g. Mycobacterium tuberculosis, which normally proliferates inside cells and is resistant to killing after phagocytosis. Helper T cells also function in a Th2 manner, releasing cytokines that promote antibody formation including production of IgE, which is important in protection against parasites. The core Th2 cytokine profile is IL-4, IL-5, IL-10 and IL-13. Cell-mediated immunity in transfusion medicine

Cell-mediated immunity is of much less importance to the transfusion of blood cells than humoral immunity. It is important in the defence against blood-borne virus (discussed below). Th1type cytokines released from leucocytes in stored blood are the main cause of non-allergic febrile transfusion reactions in susceptible individuals. As we have seen, virally infected target cells are marked for recognition by cytotoxic T cells by the presence of oligopeptides derived from viral proteins in the cleft of the HLA class I molecule. This is a process analogous to that which occurs for antigen presentation on class II HLA by APCs. The cytotoxic T-cell response to this antigen is very intense. Recent studies with cytomegalovirus (CMV)-derived peptides captured in HLA class I tetramers have revealed that up to 8% of CD8+ T cells are CMV specific during CMV infection and similar results have been obtained with peptides derived from other viruses. Biologically, the antibody response can afford to lag behind this response and usually does. This is of importance to transfusion practice. Prevention of the transmission of hepatitis B, hepatitis C, human immunodeficiency virus (HIV) types 1 and 2, and CMV by blood transfusion is one of the major challenges of transfusion medicine. Counselling of donors, together with the detection of viral antigens and antibodies and the development of tests for virally derived nucleic acid, is the bedrock for the prevention of viral 21

Chapter 2

transmission. Antibody-based immunoassays are the mainstay but after a first encounter with a virus the formation of viral antibodies will require time; they are not the first line of defence or even the means to recovery from infection for many of the pathogens concerned but the basis of immunity against subsequent infection. There is a critical window, which may be several months, in which the donor carries the virus but is still antibody negative. The sequence of immunodominant oligopeptides which appear in the HLA class I molecule for all four main blood-borne viruses and several others of clinical significance has been defined and applied to treatment. Such short oligopeptides can be used to load APCs for in vitro education and proliferation of virus-specific cytotoxic T cells. These educated T cells can be clinically used for the prevention of viral infection in immunosuppressed transplant patients. • Adoptive immunotherapy for prevention of post-transplant Epstein–Barr virus-associated lymphomas by infusion of virus-specific cytotoxic T lymphocytes has been successfully applied in allogeneic bone marrow transplantation in children. • Preliminary data suggest that CMV infection in allogeneic bone marrow transplant patients can be prevented in a similar way. As blood services are adept at handling cells in vitro, they are increasingly supporting the clinical development of these novel cellular therapies, which are a direct application of the underlying immunobiology.

Antigen presentation and its clinical implications The interaction between APCs and helper T cells is complex (Fig. 2.4). We now understand which of the many interactions are prominent in turning on T-cell activity. This knowledge may provide the means of specific manipulation of immune responses. The requirement is for concurrent signalling over two independent pathways, one antigen specific via TCR and HLA/peptide and the other non-specific, the CD28/CD80 pathway being prominent. This interaction also provides control to prevent inappropriate T-cell proliferation. Initial triggering via CD28 results in T-cell proliferation and IL-2 production, which in turn induces the expression of cytotoxic T-lymphocyte antigen (CTLA)-4 on the expanding T-cell clone. CTLA-4 is a competitive inhibitor of proliferation and competes with CD28 for binding with CD80 and CD86. This molecular competition in the control of T-cell proliferation results in a balanced expansion of antigen-specific T-cell clones. The absence of initial signalling through B7 from the APC when antigen is appropriately presented leads to apoptosis of the helper T cell. This can be exploited in the modulation of immune responses through molecules which block the interaction. The following are examples which are of relevance to transfusion medicine. Clinical studies in HSC transplantation suggest that donor lymphocytes can be tolerated for Cell membrane

Cell membrane

TCR MHC class I or II APC/virus infected target cell

(ICAM-1) CD54 (LFA-3) CD58

CD4 or CD8 (LFA-1 ) CD11a/CD18 CD2

(B7.1) CD80

CTLA-4

(B7.2) CD86

CD28

CD40

CD40L

T cell

Fig. 2.4 Adhesion molecules and

signalling pathways in T-cell activation.

22

Essential immunology

incompatible HLA alloantigens on host cells by ex vivo exposure to them in the presence of a recombinant CTLA-4–IgG fusion protein which blocks B7. In platelet transfusion, alloimmunization to HLA antigens with subsequent failure to increment after further platelet transfusions is a clinical problem. This may be reduced by leucocytedepleting blood components, thus removing APCs along with other white cells. A similar effect is obtained by treatment which modifies cell membranes and particularly the B7 signal, such as ultraviolet irradiation. Conversely, immune responses can be deliberately induced by priming separated dendritic cells with immunogenic peptide derived from the relevant antigen. This approach is under investigation in cancer therapy and again blood services are involved due to their familiarity with the safe handling of such cells under conditions of good manufacturing practice. Similar molecular switches and controls influence the B-cell response to antigen. In the interaction between B cells and T-helper cells the CD40/CD154 (CD40-ligand) pathway exercises control over B-cell isotype switching and subsequent maturation of the antibody response. This pathway is also important in certain aspects of

effector T-cell function. Again there is the potential for clinical exploitation. Anti-CD40 antibodies might be powerful therapeutic reagents in kidney transplantation. HLA-incompatible transplants can be achieved successfully in monkeys if pretreated with anti-CD40 antibodies. The ability of anti-CD40 antibodies to control autoreactive B cells in treatment-refractory autoimmune thrombocytopenia looks promising. Such designer therapeutics look likely to replace less specific therapies, such as polyvalent intravenous immunoglobulin. The amount of antigen required to activate B cells is reduced if C3d is covalently bound to antigen as this leads to the concurrent signalling of CD21 (the complement receptor type 2 on B cells) as well as SIg by antigen. This is also a pathway that is open to manipulation.

Further reading Chapel H, Haeney M, Misbah S, Snowden N. Essentials of Clinical Immunology, 4th edn. Oxford: Blackwell Science, 1999. Janeway CA, Travers P, Walport M, Capra JD. Immuno Biology. The Immune System in Health and Disease, 5th edn. London: Churchill Livingstone, 2004.

23

Chapter 3

Human blood group systems Geoff Daniels

A blood group may be defined as an inherited character of the red cell surface detected by a specific alloantibody. This definition would not receive universal acceptance as cell surface antigens on platelet and leucocytes might also be considered blood groups, as might uninherited characters on red cells defined by autoantibodies or xenoantibodies. However, the definition is suitable for the purposes of this chapter. Most blood groups are organized into blood group systems. Each system represents a single gene or a cluster of two or more closely linked homologous genes. Of the 285 blood group specificities recognized by the International Society for Blood Transfusion, 245 belong to one of 29 systems (Table 3.1). All these systems represent a single gene, apart from Rh, Xg and Chido/ Rodgers, which have two closely linked homologous genes, and MNS with three genes. Most blood group antigens are proteins or glycoproteins, with the blood group specificity determined primarily by the amino acid sequence, and most of the blood group polymorphisms result from single amino acid substitutions, though there are many exceptions. The four types of red cell surface glycoproteins, based on their integration into the red cell membrane, are shown in Fig. 3.1. Some blood group antigens, including those of the ABO, P, Lewis and H systems, are carbohydrate structures on glycoproteins and glycolipids. These antigens are not produced directly by the genes controlling their polymorphisms — the genes encode transferase enzymes that catalyse the final stage in the synthesis of an oligosaccharide chain. The two most important blood group systems from the clinical point of view are ABO and Rh. 24

They also provide good models for contrasting carbohydrate- and protein-based blood group systems.

The ABO system ABO is often referred to as a histo-blood group system because, in addition to being expressed on red cells, ABO antigens are present on most tissues and in soluble form in secretions. At its most basic level, the ABO system consists of two antigens, A and B, indirectly encoded by two alleles, A and B, of the ABO gene. A third allele, O, produces neither A nor B. These three alleles combine to effect four phenotypes: A, B, AB and O (Table 3.2). Clinical significance

Two key factors make ABO the most important blood group system in transfusion medicine. Firstly, almost without exception, the blood of adults contains antibodies to those ABO antigens lacking from their red cells (see Table 3.2). In addition to anti-A and anti-B, group O individuals have an antibody, called anti-A,B, to a determinant common to A and B. Secondly, ABO antibodies are invariably IgM, though they may also have an IgG component, activate complement, and cause immediate intravascular red cell destruction, which can give rise to severe and often fatal haemolytic transfusion reactions (see Chapter 13). Major ABO incompatibility (i.e. donor red cells with an ABO antigen not possessed by the recipient) must be avoided in transfusion and, ideally, ABO matched blood (i.e. of the same ABO group) should be provided.

Human blood group systems Table 3.1 Human blood group systems.

No.

Name

Symbol

No. of antigens

Gene name(s)

Chromosome

001 002 003 004 005 006 007 008 009 010 011 012 013 014 015 016 017 018 019 020 021 022 023 024 025 026 027 028 029

ABO MNS P Rh Lutheran Kell Lewis Duffy Kidd Diego Yt Xg Scianna Dombrock Colton Landsteiner-Wiener Chido/Rodgers H Kx Gerbich Cromer Knops Indian Ok Raph John Milton Hagen I Globoside Gill

ABO MNS P1 RH LU KEL LE FY JK DI YT XG SC DO CO LW CH/RG H XK GE CROM KN IN OK RAPH JMH I GLOB GIL

4 43 1 48 20 25 6 6 3 21 2 2 5 5 3 3 9 1 1 8 12 8 2 1 1 1 1 1 1

ABO GYPA, GYPB, GYPE P1 RHD, RHCE LU KEL FUT3 FY SLC14A1 SLC4AE1 (AE1) ACHE XG, MIC2 SC DO AQP1 LW C4A, C4B FUT1 XK GYPC DAF CR1 CD44 CD147 CD151 SEMA7A GCNT2 B3GALT3 AQP3

9 4 22 1 19 7 19 1 18 17 7 X/Y 1 12 7 19 6 19 X 2 1 1 11 19 11 15 6 3 9

Single-pass Type 1 N

Type 2

Polytopic

GPI-anchored

(multi-pass)

Type 5

Type 3

C

N N C

Fig. 3.1 Diagrammatic representation

of the four types of glycoproteins of the red cell surface membrane, with examples of blood group antigens expressed on those types of glycoproteins. GPI, glycosylphosphatidylinositol.

C Glycophorins A to D, Lutheran, LW, Knops (CD35), Indian (CD44)

N Kell

N

C RhD, RhCcEe, RhAG Kidd, Diego (band 3), Colton (AQP1), Gill (AQP3), Kx

C Duffy

Yt(AChE), Dombrock Cromer (CD55) JMH (CDw108)

25

Chapter 3 Table 3.2 ABO system.

Frequency Phenotype

Genotypes

Europeans*

Africans†

Indians‡

Antibodies present

O A1 A2 B A1B A2B

O/O A1/A1, A1/O, A1/A2 A2/A2, A2/O B/B, B/O A1/B A2/B

43% 35% 10% 9% 3% 1%

51% 18% 5% 21% 2% 1%

31% 26% 3% 30% 9% 1%

Anti-A, -B, -A,B Anti-B Sometimes anti-A1 Anti-A None Sometimes anti-A1

* English people. † Donors from Kinshasa, Congo. ‡ Makar from Mumbai.

ABO antibodies seldom cause haemolytic disease of the newborn and when they do it is usually mild. The prime reasons for this are (i) IgM antibodies do not cross the placenta; (ii) IgG ABO antibodies are often IgG2, which do not activate complement or facilitate phagocytosis; and (iii) ABO antigens are present on many fetal tissues and even in body fluids, so the haemolytic potential of the antibody is greatly reduced.

membrane glycoproteins, mainly the anion transporter band 3 and the glucose transporter GLUT1, but are also on glycosphingolipids embedded in the membrane. The tetrasaccharides that represent the predominant form of A and B antigens on red cells are shown in Fig. 3.2, together with their biosynthetic precursor, the H antigen, which is abundant on group O red cells. The product of the A allele is a glycosyltransferase that catalyses the transfer of N-acetylgalactosamine (GalNAc) from

A and B subgroups

The A (and AB) phenotype can be subdivided into A1 and A2 (and A1B and A2B). In a European population, about 80% of group A individuals are A1 and 20% A2 (see Table 3.2). A1 and A2 differ quantitatively and qualitatively. A1 red cells react more strongly with anti-A than A2 cells. In addition, A2 red cells lack a component of the A antigen present on A1 cells and some individuals with the A2 or A2B phenotype produce an antibody, anti-A1, which agglutinates A1 and A1B cells but not A2 or A2B cells. Anti-A1 is seldom reactive at 37°C and generally considered clinically insignificant. There are numerous other ABO variants, involving weakened expression of A or B antigens (A3, Ax, Am, Ael, B3, Bx, Bm, Bel), but all are rare.

O

Gal

GlcNAc c

R

Gal

GlcNAc

R

Gal

GlcNAc

R

(H) Fuc

A

GalNAc Fuc

B

Gal Fuc

Biosynthesis and molecular genetics

Red cell A and B antigens are expressed predominantly on oligosaccharide structures on integral 26

Fig. 3.2 Tetrasaccharides representing A and B antigens, and their biosynthetic precursor (H), which is abundant in group O. R, remainder of molecule.

Human blood group systems

a nucleotide donor substrate, UDP-GalNAc, to the fucosylated galactose (Gal) residue of the H antigen, the acceptor substrate. The product of the B allele catalyses the transfer of Gal from UDP-Gal to the fucosylated Gal residue of the H antigen. GalNAc and Gal are the immunodominant sugars of A and B antigens, respectively. The O allele produces no transferase, so the H antigen remains unmodified. The ABO gene on chromosome 9 consists of seven exons. The A1 and B alleles differ by seven nucleotides in exons 6 and 7, which encode a total of four amino acid substitutions at positions 176, 235, 266 and 268 of their glycosyltransferase products (Fig. 3.3). It is primarily the amino acids at positions 266 and 268 that determine whether the gene product is a GalNAc-transferase (A) or Gal-transferase (B). The most common O allele (O1) has an identical sequence to A1, apart from a single nucleotide deletion in exon 6, which shifts the reading frame and introduces a translation stop codon before the region of the catalytic site, so that any protein produced would be truncated and have no enzyme activity. Another common O allele, called O1v, differs from O1 by at least nine nucleotides, but has the same single nucleotide deletion as that in O1 and so cannot produce any functional enzyme. O2, which represents about 3% of O alleles in a European population, does not have the nucleotide deletion characteristic of most O alleles and encodes a complete protein

D 261

526

exon 6 176 Arg

D 703 796 803 1059 exon7 235 266 268 Gly Leu Gly

Arg

Gly Leu Gly

Gly

Ser Met Ala

A1 A2 B O

Fig. 3.3 Diagrammatic representation of exons 6 and 7 of

the ABO gene showing the position of the nucleotide deletions (D) responsible for the common form of O (exon 6) and for A2 (exon 7), and the positions of the four nucleotide changes in exon 7 responsible for the amino acid residues that are characteristic of A- and B-transferases. Below are representations of the encoded transferases.

product, but with a charged arginine residue instead of a neutral glycine (A) or alanine (B) at position 268. This amino acid change at a vital position inactivates enzyme activity. The A2 allele has a sequence almost identical to A1, but has a single nucleotide deletion immediately before the translation stop codon. The resultant frameshift abolishes the stop codon, so the protein product has an extra 21 amino acids at its C-terminus, which reduces the efficiency of its GalNActransferase activity and might alter its acceptor substrate specificity. Biochemically related blood group systems: H, Lewis and I

H antigen is the biochemical precursor of A and B (see Fig. 3.2). It is synthesized by an a1,2fucosyltransferase, which catalyses the transfer of fucose from its donor substrate to the terminal Gal residue of its acceptor substrate. Without this fucosylation neither A nor B antigens can be made. Two genes, active in different tissues, produce a1,2-fucosyltransferases: FUT1, active in mesodermally derived tissues and responsible for H on red cells, and FUT2, active in endodermally derived tissues and responsible for H in many other tissues and in secretions. Homozygosity for inactivating mutations in FUT1 leads to an absence of H from red cells and therefore an absence of red cell A or B, regardless of ABO genotype. Such mutations are rare, as are red cell H-deficient phenotypes. In contrast, inactivating mutations in FUT2 are relatively common and about 20% of Caucasians (non-secretors) lack H, A and B from body secretions despite expressing those antigens on their red cells. Very rare individuals who have H-deficient red cells and are also H non-secretors (Bombay phenotype) produce anti-H together with anti-A and -B and create a severe transfusion problem. Antigens of the Lewis system are not produced by erythroid cells, but become incorporated into the red cell membrane from the plasma. Their corresponding antibodies are not usually active at 37°C and are not generally considered clinically significant. Lea and Leb are not the products of alleles. The Lewis gene (FUT3) product is an 27

Chapter 3 RHD Precursor of H and Lea

Gal

GlcNAc

R

Gal

GlcNAc

R

GlcNAc

R

exons

RHCE 3´

5´ 1

10

3´

10

D

1 C/c

H precursor of Leb

5´

E/e

Fuc

Gal Lea

Fuc

Gal

GlcNAc

R

Leb Fuc

N

C N

C

Fig. 3.5 Diagrammatic representation of the Rh genes,

RHD and RHCE, shown in opposite orientation as they appear on the chromosome, and of the two Rh proteins in their probable membrane conformation, with 12 membrane-spanning domains and six extracellular loops expressing D, C/c and E/e antigens.

Fuc

Fig. 3.4 Oligosaccharide structures representing Lea and Leb

expression and their biosynthetic precursors. R, remainder of molecule.

a1,3/4-fucosyltransferase that transfers fucose to the GlcNAc residue of the secreted H precursor in non-secretors to produce Lea and to secreted H in secretors to produce Leb (Fig. 3.4). Consequently, H secretors are Le(a–b+) or Le(a+b+), H nonsecretors are Le(a+b–) and individuals homozygous for FUT3 inactivating mutations (secretors or non-secretors) are Le(a–b–). I antigen represents branched N-acetyllactosamine (Galb1–4GlcNAc) structures in the complex carbohydrates that also express H, A and B antigens. The I gene (GCNT2) encodes a branching enzyme, which only becomes active during the first months of life. Consequently, red cells of neonates are I-negative. Rare individuals are homozygous for inactivating mutations in GCNT2 and never form I on their red cells. This phenotype, called adult i, is associated with production of anti-I, which is usually only active below 37°C, but may occasionally be haemolytic at body temperature.

The Rh system Rh is the most complex of the blood group 28

systems, with 49 specificities. The most important of these is D, and then C, c, E and e. Rh genes and proteins

The antigens of the Rh system are encoded by two genes, RHD and RHCE, which produce D and CcEe antigens respectively. The genes are highly homologous, each consisting of 10 exons. They are closely linked, but in opposite orientation, on chromosome 1 (Fig. 3.5). Each gene encodes a 417 amino acid polypeptide that differ by only 31–35 amino acids, according to Rh genotype. The Rh proteins are palmitoylated, but not glycosylated, and span the red cell membrane 12 times, with both termini inside the cytosol and with six external loops, the potential sites of antigenic activity (see Fig. 3.5). D antigen

The most significant Rh antigen from the clinical point of view is D. About 85% of Caucasians are D+ (Rh-positive) and 15% are D– (Rh-negative). In Africans only about 3–5% are D– and in the Far East D– is rare. The D– phenotype is usually associated with absence of the whole D protein from the red cell membrane. This explains why D is so immuno-

Human blood group systems

genic, as the D antigen comprises numerous epitopes on the external domains of the D protein. In Caucasians, the D– phenotype almost always results from homozygosity for a complete deletion of RHD. D+ individuals are either homozygous or heterozygous for the presence of RHD. In Africans, in addition to the deletion of RHD, D– often results from an inactive RHD (called RHDY) containing translation stop codons within the reading frame. Other genes containing inactivating mutations are also found in D– Africans and in D– Asians. Weak forms of D (previously known as Du) result from amino acid substitutions in the membrane-spanning or cytosolic regions of the D protein. Red cells of some D+ individuals lack some or most of the D epitopes and, if immunized by a complete D antigen, can make antibodies to the epitopes they lack. There are numerous types of these partial D antigens. They result from amino acid changes in the external loops of the D protein. Usually this is due to one or more exons of RHD being exchanged for the equivalent exons of RHCE in a process called gene conversion, but sometimes straightforward missense mutations are responsible. Anti-D

Anti-D is almost never produced in D– individuals without immunization by D+ red cells. However, D is highly immunogenic and about 85% of D– individuals will make anti-D following infusion of 200 mL or more of D+ red cells. Anti-D can cause severe immediate or delayed haemolytic transfusion reactions and D+ blood must never be transfused to a patient with anti-D. Anti-D is the most common cause of severe haemolytic disease of the fetus and newborn (HDN). The effects of HDN caused by anti-D are, at its most severe, fetal death at about week 17 of pregnancy. If the infant is born alive, the disease can result in hydrops and jaundice. If the jaundice leads to kernicterus, this usually results in infant death or permanent cerebral damage. The prevalence of HDN due to anti-D has been substantially reduced by anti-D immunoglobulin prophylaxis.

In 1970, at the beginning of the anti-D prophylaxis programme, there were 1.2 deaths per 1000 births in England and Wales due to HDN caused by antiD; by 1989 this figure had been reduced to 0.02. Prediction of fetal Rh genotype by molecular methods

Knowledge of the molecular bases for D– phenotype has made it possible to devise tests for predicting fetal D type from fetal DNA. This is a valuable tool in assessing whether the fetus of a woman with anti-D is at risk from HDN. Most methods involve polymerase chain reaction (PCR) tests that detect the presence or absence of RHD. It is important to test for more than one region of RHD, so that hybrid genes responsible for partial D antigens do not give a false result, and to test for RHDy, so that this does not give rise to a falsepositive result. Until recently the usual source of fetal DNA has been amniocytes. These are obtained by amniocentesis, which has an inherent risk of fetal loss and of fetomaternal haemorrhage. It is now possible to use sensitive PCR technology to determine fetal D type from the small quantity of free fetal DNA present in maternal plasma, as early as 12 weeks into the pregnancy. This noninvasive form of fetal D typing is now provided as a reference service in a few countries. C and c, E and e

C/c and E/e are two pairs of allelic antigens produced by RHCE. The fundamental difference between C and c is a serine–proline substitution at position 103 in the second external loop of the CcEe protein (see Fig. 3.5), although the situation is more complex than that. E and e represent a proline–alanine substitution at position 226 in the fourth external loop. Taking into account the presence and absence of D, and of the C/c and E/e polymorphisms, eight different haplotypes can be recognized. The frequencies of these haplotypes and the shorthand symbols often used to describe them are shown in Table 3.3. Anti-c is clinically the most important Rh antigen after anti-D and may cause severe HDN. On the other hand, anti-C, anti-E and anti-e rarely 29

Chapter 3 Table 3.3 Rh phenotypes and the genotypes that produce them (presented in DCE and shorthand terminology).

Phenotype

Frequency (%) Asians‡

Genotypes

0.7

56.0

2.3

1.3

3.5

+

2.1

58.9

0.2

+

-

Rare

Rare

Rare

+

-

+

34.9

13.2

8.4

-

+

+

+

11.8

18.3

2.1

+

+

-

+

+

0.2

Rare

1.1

+

+

+

+

-

0.1

Rare

0.3

+

+

+

+

+

13.4

2.1

28.1

-

+ + + + + +

+ + + + + +

+ + + + + +

+ + + + + +

Rare Rare 15.1 Rare 0.1 0.1 Rare Rare Rare

0.1 Rare 4.1 Rare 1.3 Rare Rare Rare Rare

Rare Rare 0.1 Rare 0.1 Rare Rare Rare Rare

DCe/Dce DCe/dCe DcE/DcE DcE/dcE Dce/dce Dce/Dce DCE/DCE DCE/dCE DCe/dce DCe/Dce Dce/dCe DcE/dce DcE/Dce Dce/dcE DCe/DCE DCE/dCe DCe/dCE DcE/DCE DCE/dcE DcE/dCE DCe/DcE DCe/dcE DcE/dCe DCE/dce Dce/DCE Dce/dCE dCe/dCe dcE/dcE dce/dce dCE/dCE dCe/dce dcE/dce dCe/dCE dcE/dCE dcE/dCe dCE/dce

D

C

c

E

e

Europeans*

+

+

-

-

+

18.5

+

-

+

+

-

+

-

+

-

+

+

-

+

+

+

* English donors. † Yoruba of Nigeria. ‡ Cantonese of Hong Kong.

30

Africans†

R1/R1 R1r¢ R2R2 R2r≤ Ror RoRo RzRz Rzry R1r R1Ro Ror¢ R2r R2Ro Ror≤ R1Rz Rzr¢ R1ry R2Rz Rzr≤ R2ry R1R2 R1r≤ R2r¢ Rzr RoRz Rory r¢r¢ r≤r≤ rr ryry r¢r r≤r r¢ry r≤ry r≤r¢ ryr

Human blood group systems

cause HDN and when they do the disease is generally mild, though all have caused severe disease. Other Rh antigens

Of the 49 Rh antigens, 20 are polymorphic (i.e. have a frequency between 1 and 99% in at least one major ethnic group), 21 are rare antigens and eight are very common antigens. Antibodies to many of these antigens have shown themselves to be clinically important and it is prudent to treat all Rh antibodies as being potentially clinically significant.

Other blood group systems Of the remaining blood group systems (see Table 3.1), the most important clinically are Kell, Duffy, Kidd and MNS, and are described below. Kell system

The original Kell antigen, K (KEL1), has a frequency of about 9% in Caucasians, but is rare in other ethnic groups. Its allelic antigen, k (KEL2), is common in all populations. The remainder of the Kell system consists of one triplet and three pairs of allelic antigens: Kpa, Kpb and Kpc; Jsa and Jsb; K11 and K17; K14 and K24; plus 11 highfrequency and four low-frequency antigens. All represent single amino acid substitutions in the Kell glycoprotein. Anti-K can cause severe haemolytic transfusion reactions and HDN. About 10% of K-negative patients who are given one unit of K-positive blood produce anti-K, making K the next most immunogenic antigen after D. About 0.1% of all cases of HDN are caused by anti-K; most of the mothers will have had previous blood transfusions. HDN caused by anti-K differs from Rh HDN in that anti-K appears to cause fetal anaemia by suppression of erythropoiesis, rather than immune destruction of mature fetal erythrocytes. Anti-K is a very rare antibody. It is always immune and has been incriminated in some cases of mild HDN. Most other Kell-system antibodies are rare and best detected by an antiglobulin test.

The Kell antigens are located on a large glycoprotein that crosses the cell membrane once and has a glycosylated C-terminal extracellular domain, maintained in a folded conformation by multiple disulphide bonds. The Kell glycoprotein belongs to a family of endopeptidases, which process biologically important peptides, and are able to cleave the biologically inactive peptide big endothelin-3 to produce endothelin-3, an active vasoconstrictor. Duffy system

Fya and Fyb represent a single amino acid substitution in the extracellular N-terminal domain of the Duffy glycoprotein. Their incidence in Caucasians is 66% Fya and 80% Fyb. About 70% of AfricanAmericans and close to 100% of West Africans are Fy(a–b–) (Table 3.4). They are homozygous for an Fyb allele containing a mutation in a binding site for the erythroid-specific GATA-1 transcription factor, which means that Duffy glycoprotein is not expressed in red cells, although it is present in other tissues (Table 3.5). The Duffy glycoprotein is the receptor exploited by Plasmodium vivax merozoites for penetration of erythroid cells. Consequently, the Fy(a–b–) phenotype confers resistance to P. vivax malaria. The Duffy glycoprotein (also called Duffy antigen chemokine receptor, DARC) is a red cell receptor for a variety of chemokines, including interleukin-8. Anti-Fya is not infrequent and is found in previously transfused patients who have usually already made other antibodies. It can cause haemolytic transfusion reactions, but seldom causes HDN. Anti-Fyb is very rare.

Table 3.4 Duffy system: phenotypes and genotypes.

Frequency (%) Phenotype

Genotype

Europeans

Africans

Fy(a+b-) Fy(a+b+) Fy(a-b+) Fy(a-b-)

Fy a/Fy a or Fy a/Fy Fy a/Fy b Fy b/Fy b or Fy b/Fy Fy/Fy

20 48 32 0

10 3 20 67

31

Chapter 3 Table 3.5 Nucleotide polymorphisms

Allele

GATA box sequence -64 to -69 (promoter)

Codon 42 (exon 2)

Antigen

Fya Fyb Fy

TTATCT TTATCT TTACCT

GGT (Gly) GAT (Asp) GAT (Asp)

Fya Fyb Red cells: none Other tissues: probably Fyb

in the promoter region and in exon 2 of the three common alleles of the Duffy gene.

Kidd system

MNS system

Kidd has two alleles, Jka and Jkb, which represent a single amino acid change in the Kidd glycoprotein. Jka and Jkb antigens both have frequencies of about 75% in Caucasian populations. A Kidd-null phenotype, Jk(a–b–), results from homozygosity for inactivating mutations in the Kidd gene, SLC14A1. It is very rare in most populations, but reaches an incidence of greater than 1% in Polynesians. The Kidd glycoprotein is a urea transporter in red cells and in renal endothelial cells. Anti-Jka is uncommon and anti-Jkb is very rare, but they both cause severe transfusion reactions and, to a lesser extent, HDN. Kidd antibodies have often been implicated in delayed haemolytic transfusion reactions. They are often difficult to detect serologically and tend to disappear rapidly after stimulation.

MNS, with a total of 43 antigens, is second only to Rh in complexity. These antigens are present on one or both of two red cell membrane glycoproteins, glycophorin A (GPA) and glycophorin B (GPB). They are encoded by two homologous genes, GYPA and GYPB, on chromosome 4. The M and N antigens, both with frequencies of about 75%, differ by amino acids at positions 1 and 5 of the external N-terminus of GPA. S and s have frequencies of about 55 and 90%, respectively, in a Caucasian population, and represent an amino acid substitution in GPB. About 2% of black West Africans and 1.5% of AfricanAmericans are S– s–, a phenotype virtually unknown in other ethnic groups, and most of these lack the U antigen, which is present when either S or s is expressed. The numerous MNS variants mostly result from amino acid substitutions in GPA or GPB and from the formation of hybrid GPA–GPB molecules, resulting from intergenic recombination between GYPA and GYPB. GPA and GPB are exploited as receptors by the malaria parasite Plasmodium falciparum. Anti-M and anti-N are not generally clinically significant, though anti-M is occasionally haemolytic. Anti-S, the rarer anti-s, and anti-U can cause HDN and have been implicated in haemolytic transfusion reactions.

Diego system

Diego is a large system of 21 antigens: two pairs of allelic antigens (Dia and Dib, Wra and Wrb) plus 17 antigens of very low frequency. All represent single amino acid substitutions in band 3, the red cell anion exchanger. The original Diego antigen, Dia, is very rare in Caucasians and black people, but relatively common in Mongoloid people, with frequencies varying between 1% in Japanese and 50% in some native South Americans. Anti-Dia and anti-Dib are immune and rare, but can cause HDN. Wra has a frequency of about 0.1%. Its high-frequency allelic antigen, Wrb, is dependent on an interaction of band 3 with glycophorin A for its expression. Naturally occurring anti-Wra is present in approximately 1% of blood donors. Very rarely, anti-Wra causes HDN. 32

Biological significance of blood group antigens The functions of several red cell membrane protein structures bearing blood group antigenic determinants are known, or can be deduced, from their

Human blood group systems

structure. Some are membrane transporters, facilitating the transport of biologically important molecules through the lipid bilayer: band 3 membrane glycoprotein, the Diego antigen, provides an anion exchange channel for HCO3– and Cl– ions; the Kidd glycoprotein is a urea transporter; the Colton glycoprotein is aquaporin 1, a water channel; the GIL antigen is aquaporin 3, a glycerol transporter; and the Rh protein complex might function as an ammonium transporter or a CO2 channel. The Lutheran, LW and Indian (CD44) glycoproteins are adhesion molecules, possibly serving their primary functions during erythropoiesis. The Duffy glycoprotein is a chemokine receptor and could function as a ‘sink’ or scavenger for unwanted chemokines. The Cromer and Knops antigens are markers for decay accelerating factor (CD55) and complement receptor 1 (CD35), respectively, which protect the cells from destruction by autologous complement. Some blood group glycoproteins have enzyme activity: the Yt antigen is acetylcholinesterase and the Kell antigen is an endopeptidase, though their functions on red cells are not known. The Cterminal domains of the Gerbich antigens, GPC and GPD, and the N-terminal domain of the Diego glycoprotein, band 3, are attached to components of the cytoskeleton and function to anchor it to the external membrane. The carbohydrate moieties of the membrane glycoproteins and glycolipids, especially those of the most abundant glycoproteins (band 3 and GPA), constitute the glycocalyx, an extracellular coat that protects the cell from mechanical damage and microbial attack. The structural differences between allelic red cell antigens (e.g. A and B, K and k, Fya and Fyb) are small, often being just one monosaccharide or one amino acid. The biological importance of these differences is unknown and there is little evidence to suggest that the product of one allele confers any significant advantage over the other. Some blood group antigens are exploited by pathological microorganisms as receptors for attaching and entering cells, so in some cases absence or changes in these antigens could be beneficial. It is likely that interaction between cell surface molecules and pathological microorganisms has been a major factor in the evolution of blood group polymorphism.

Summary Blood groups are inherited characters of the red cell surface detected by specific alloantibodies. Most of the authenticated 285 blood group specificities are organized into 29 blood group systems, each representing a single gene or a cluster of closely linked homologous genes. Blood group antibodies are of clinical importance in transfusion medicine as they can cause haemolytic transfusion reactions and HDN. The carbohydrate ABO antigens and protein Rh antigens are the most important from the clinical aspect. Blood group proteins serve a variety of functions, though little is known about the biological significance of blood group polymorphism.

Further reading Avent ND. Fetal genotyping. In: Hadley A, Soothill P, eds. Alloimmune Disorders of Pregnancy. Cambridge: Cambridge University Press, 2002: 121–39. Avent ND, Reid ME. The Rh blood group system: a review. Blood 2000; 95: 375–87. Chester AM, Olsson ML. The ABO blood group gene: a locus of considerable genetic diversity. Transfus Med Rev 2001; 15: 177–200. Daniels G. Functional aspects of red cell antigens. Blood Rev 1999; 13: 14–35. Daniels G. Human Blood Groups, 2nd edn. Oxford: Blackwell Science, 2002. Daniels G, Poole J, de Silva M et al. The clinical significance of blood group antibodies. Transfus Med 2002; 12: 287–95. Daniels GL, Fletcher A, Garratty G et al. Blood group terminology 2004. Vox Sang 2004, in press. Denomme GA. The structure and function of the molecules that carry human red blood cell and platelet antigens. Transfus Med Rev 2004; 18: 203–31. Henry S, Samuelsson B. ABO polymorphisms and their putative biological relationships with disease. In: King M-J, ed. Human Blood Cells: Consequences of Genetic Polymorphism and Variations. London: Imperial College Press, 2000: 1–103. Reid ME, Lomas-Francis C. The Blood Group Antigen Facts Book. San Diego: Academic Press, 1997. Watkins WM (ed.) Commemoration of the centenary of the discovery of the ABO blood group system. Transfus Med 2001; 11: 239–351.

33

Chapter 4

Human leucocyte antigens Cristina V. Navarrete

The human leucocyte antigen (HLA) system consists of a family of cell surface polymorphic molecules involved in the presentation of antigen to T cells and therefore plays a central role in the induction and regulation of immune responses. HLA molecules are also known to be involved in the pathogenesis of certain autoimmune and infectious diseases and they have an important influence on the outcome of solid organ and haemopoietic stem cell transplantation. Furthermore, HLA antigens present in blood cells are responsible for some of the serious clinical complications of blood transfusion. The genes coding for the HLA molecules are located on the short arm of chromosome 6 and span a distance of approximately 4 Mb. This region is divided into three subregions. • Class I subregion contains genes coding for the heavy chain of the classical (HLA-A, -B and -C) and non-classical (HLA-E, -F and -G) HLA class I molecules; • Class II subregion contains genes coding for the HLA class II molecules (DR, DQ and DP) and genes involved in the processing and transport of antigenic peptides. • Class III subregion lies between the other two subregions and contains genes coding for a diverse group of proteins, including complement components (C4Bf), tumour necrosis factor (TNF) and heat-shock proteins. A number of additional genes, including the major histocompatibility complex (MHC) class I chain-related gene A (MICA) and MHC class I chain-related gene B (MICB) also involved in immune responses, have been mapped between the class I and class III subregions (Fig. 4.1). In addition, the non-classical class I-like gene HFE has 34

been mapped to a locus located 4 Mb telomeric to HLA-F. Mutations in this gene have been shown to be responsible for the development of hereditary haemochromatosis. Following the development of recombinant DNA technology, it has been possible to perform a detailed analysis of the HLA region, leading to the unravelling of the genetic complexity and structure of its genes and molecules. These findings, combined with parallel studies on their function genes, have led to a better understanding of the role of the HLA system in clinical medicine.

HLA class I genes The HLA class I genes have been classified according to their structure, expression and function as classical (HLA-A, -B and -C) and non-classical (HLA-E, -F and -G). Both classical and non-classical HLA class I genes code for a heavy (a) chain, of approximately 43 kDa, non-covalently linked to a non-polymorphic light chain, the b2-microglobulin of 12 kDa, which is coded for by a gene on chromosome 15. The extracellular portion the heavy chain has three domains (a1, a2 and a3) approximately 90 amino acids long. These domains are encoded by exons 2, 3 and 4 of the class I gene, respectively. The a1 and a2 domains are the most polymorphic domains of the molecule and they form a peptide-binding groove that can accommodate antigenic peptides approximately eight to nine amino acids long. The exon/intron organization of the non-classical HLA class I genes (E, F and G) is very similar to the classical class I genes but they have a more restricted polymorphism.

Human leucocyte antigens HLA class III subregion

HLA class II subregion DMB DP

B2 A2

DOA DMA

LMP7 TAP1 TAP2 LMP2 DOB

B1 A1

DQ

MICB MICA TNF HSP70 B C

DR

B2 A2 B3 B1 A1

HLA class I subregion

B1 B2 B5 B3 B4

A

A

E

A H G F

B

Class II genes pseudogenes

Class II genes

ABC transporter genes

Proteosomelike genes

TNF

Classical class I genes

Non-classical class I genes

MICA and MICB

HSP70

Fig. 4.1 Map of the human leucocyte antigen complex. HSP, heat-shock protein; TNF, tumour necrosis factor.

α1

Peptide binding region

α2 S S

N Immunoglobulinlike region

β2-m

S

S

S

S

α3

C Transmembrane region P P P

Cytoplasmic region

HLA class A gene Fig. 4.2 HLA class I molecule. b2-m,

b2-microglobulin.

A schematic representation of the classical HLA class I gene and molecule is shown in Fig. 4.2.

HLA class II genes The class II DR, DQ and DPA and DPB genes

C

Regulatory sequences

L

α1

α2

α3

TM

CYT

3'UT

1

2

3

4

5

6 7

8

Exons

code for a heterodimer formed by two noncovalently associated a and b chains of approximately 34 and 28 kDa respectively. The expressed a and b chains consist of two extracellular domains as well as transmembrane and cytoplasmic domains. The a1/b1 and a2/b2 domains are encoded by exon 2 and exon 3 of the class II gene 35

Chapter 4

respectively. The majority of the polymorphism is located in the b1 domain of the DR molecules and in the a1 and b1 domains of the DQ and DP molecules. Similarly to the class I molecules, these domains also form a peptide-binding groove. However, in the case of the class II molecules (DR), the groove is open at both sides and it can accommodate antigenic peptides of varying size, although most of them are approximately 13–25 amino acids long. A schematic representation of the HLA class II gene and molecule is shown in Fig. 4.3.

Expression of HLA class II genes There is one non-polymorphic DRA and nine DRB genes, of which B1, B3, B4 and B5 are highly polymorphic and B2, B6 and B9 are pseudogenes. The main DR specificities (DR1–DR18) are determined by the polymorphic DRB1 gene. Furthermore, the number of DRB genes expressed in each

α1

Peptide binding region

individual varies according to the DRB1 allele expressed, e.g. HLA-DR1, -DR103, -DR10 and -DR8 alleles express the DRB1 gene only. DR15 and DR16 alleles express the DRB1 and DRB5 genes, which code for the DR51 product; HLADR17, -DR18, -DR11, -DR12, -DR13 and -DR14 alleles express the DRB1 and the DRB3 genes, which code for the DR52 specificity; and, finally, the HLA-DR4, -DR7 and -DR9 alleles express the DRB1 and the DRB4 genes coding for the DR53 product (Fig. 4.4). There are a few exceptions to this pattern of gene expression, e.g. a DRB5 gene has been found to be expressed with some DR1 alleles and non-expressed or null DRB5 and DRB4 genes have also been identified. In contrast to the DRB genes, there are two DQA and three DQB genes, of which only the DQA1 and DQB1 are expressed and both are polymorphic. Similarly, there are two DPA and two DPB genes, of which only the DPA1 and DPB1 are expressed and both are polymorphic. There are additional genes located in the class II

β1 NN S S

Immunoglobulinlike region

α2

Transmembrane region

S

S

S

Papain S cleavage sites

C

β2

C

Cytoplasmic region

Exons HLA class A gene HLA class B gene

1 L

2 α1

3 α2

5 3'UT

Regulatory sequences L 1

β1 2

β2 3 Exons

36

4 TM/CYT

TM CYT 3'UT 4 5 6 Fig. 4.3 HLA class II molecule.

Human leucocyte antigens Specifications DRB1

DRB6 Y

Dr1, DR10, DR103, DR15 DRB1 DR15, DR16, DR1

DRB9

DRA

Y

DRB6 Y

DRB5

DRB2 Y

DRB3

DRB9

DRA

Y DR51

DR17, DR18, DR11, DR12 DR13, DR14, DR1403, DR1404

DRB1

DR52 DRB9

DR8 DRB1 Fig. 4.4 Expression of HLA-DRB genes.

The classical HLA class I molecules (A, B, C) are expressed on the majority of tissues and cells, including T and B lymphocytes, granulocytes and platelets. Low levels of expression have been

DRA

Y

DR4, DR7, DR9

Distribution of HLA molecules

DRA

Y

DRB1

region which are involved in the MHC class I antigen presentation pathway. These include the low-molecular-mass polypeptide genes (LMP2 and LMP7) and the transporter associated with antigen-processing genes (TAP1 and TAP2). The LMP2 and LMP7 genes are thought to improve the capacity of the proteosomes to generate peptides of the appropriate size and specificity to associate with the class I molecules. Conversely, the TAP1 and TAP2 genes are primarily involved in the transport of the proteosome-generated peptides to the endoplasmic reticulum, where they associate with the class I molecules. In addition, the DMA and DMB genes, which code for a heterodimer involved in the loading of peptides presented by HLA class II molecules, are also located in this subregion (see Fig. 4.1). The main function of the DM molecules is to facilitate the release of the class II-associated invariant chain peptide from the peptide-binding groove of the HLA class II molecule so that it can be exchanged for the relevant antigenic peptide.

DRB9

DRB7

DRB8

Y

Y

DRB4

DRB9

DRA

Y DR53

detected in endocrine tissue, skeletal muscle and cells of the central nervous system. HLA-E and -F are also expressed on most tissues tested but HLAG shows a more restricted tissue distribution and to date the HLA-G product has only been found to be expressed on extravillous cytotrophoblasts of the placenta and mononuclear phagocytes. The HLA class II molecules are constitutively expressed on B lymphocytes, monocytes and dendritic cells but can also be detected on activated T lymphocytes and activated granulocytes. It is not clear whether they are also present on activated platelets. HLA class II expression can be also be induced on a number of cells and tissues such as fibroblasts and endothelial cells as the result of activation and/or the effect of certain inflammatory cytokines, such as interferon (IFN)-g, TNF and interleukin (IL)-10.

Genetics One of the main features of the HLA genes is their high degree of polymorphism and the strong linkage disequilibrium (LD) in which they segregate. LD is a phenomenon where the observed frequency of alleles of different loci segregating together is greater than the frequency expected by random association. The difference between the 37

Chapter 4

observed and the expected frequencies for a particular combination of alleles is called the delta value and is positive for alleles in LD with each other. Whereas some of the polymorphism and the patterns of LD are expressed with similar frequencies in all populations, others are unique to some population groups. For example, HLA-A2 is expressed at a relatively high frequency in most population groups studied so far, whereas B53 is found predominantly in black people. In addition, all HLA genes are codominantly expressed and are inherited in a mendelian fashion and the genetic region containing all HLA genes on each chromosome is termed the haplotype. Some haplotypes are also found across different ethnic groups, e.g. HLA-B44-DR7, whereas others are unique to a particular population, e.g. HLA-B42DR18 in black Africans. This characteristic is particularly relevant for the selection of HLAcompatible family donors for patients requiring solid organ or bone marrow transplantation (BMT).

Function of HLA molecules HLA molecules are directly involved in the presentation of antigenic peptides to T cells. This is a highly regulated process and requires a fine interaction between the HLA molecules, the antigenic peptide, the T-cell receptor and a number of costimulatory molecules (e.g. CD80, CD86) and adhesion molecules such ICAM-1(CD54) and LFA-3 (CD58). The HLA class I molecules are primarily but not exclusively involved in the presentation of endogenous antigenic peptides to CD8+ cytotoxic T cells. However, it has now been shown that both classical and non-classical HLA class I molecules also interact with a new family of receptors present on natural killer (NK) cells. Some of these receptors, which are polymorphic and differentially expressed, have an inhibitory role whereas others are involved in NK cell activation. The killeractivating and killer-inhibitory receptors belong to two distinct families: the immunoglobulin superfamily called killer immunglobulin receptors

38

(KIRs), and the C-type lectin superfamily CD94NKG2. The interaction between the inhibitory receptors and the relevant HLA ligand results in the prevention of NK lysis of the target cell. Thus NK cells from any given individual will be alloreactive towards cells lacking their corresponding inhibitory KIR ligands, e.g. tumour or allogeneic cells. In contrast NK cells will be tolerant to cells from individuals who express the corresponding KIR ligands. This information is currently being exploited in the clinical setting in order to promote the graft-versus-leukaemia (GVL) effect mediated by NK cells. HLA class II molecules are mostly involved in the presentation of exogenous antigenic peptides to CD4+ helper T cells. Once activated, these CD4+ cells can initiate and regulate a variety of processes leading to the maturation and differentiation of cellular (CD8+ cytotoxic T cells) and humoral effectors by the secretion of proinflammatory cytokines (IL-2, IFN-g, TNF-a) and regulatory cytokines (IL-4, IL-10 and transforming growth factor-b). The nature of the peptide presented by class I and II molecules is largely dependent on the location of the peptide within the processing machinery of the cells.

Identification of HLA gene polymorphism Traditionally, characterization of HLA polymorphisms has been carried out using serological and cellular techniques. There are, however, several limiting factors in the use of serological typing methods: • it is often difficult to obtain antibody of sufficient titre and specificity to distinguish between all described HLA types; • the antibodies available are very often of Caucasian origin, making it difficult to HLA type patients from other population groups; • patients undergoing chemotherapy normally have low white cell counts and cell antigen expression can be affected by the chemotherapy; • the requirement for viable cells, ideally 24 h after venepuncture, means that serological HLA typing results are not always reliable; and

Human leucocyte antigens

• because class II molecules are expressed on B cells and not (resting) T cells, the low number of B cells in a sample makes serological class II typing difficult. However, with the development of gene cloning and DNA sequencing it has been possible to perform a detailed analysis of these genes at the single nucleotide level. This analysis has shown the existence of shared nucleotide sequences between alleles of the same and/or different loci and the existence of certain locus-specific nucleotide sequences in both the coding (exons) and noncoding (introns) regions of the genes. The DNA sequencing of a number of alleles of various loci has also demonstrated that the majority of the polymorphism is located in regions of the a1 and a2 domain of the class I molecules and of the a1 and b1 domain of the class II molecules. These are called hypervariable regions. Based on this information, a number of techniques have been developed to characterize HLA polymorphisms. Most of the described techniques make use of the polymerase chain reaction (PCR) to amplify the specific genes or region to be analysed. These techniques include PCR-SSOP (sequence-specific oligonucleotide probing), PCRSSP (sequence-specific priming), and conformational methods including reference strand conformational analysis (RSCA) and sequencingbased typing (SBT). The number of recognized serologically defined antigens and DNA-identified HLA alleles is shown in Table 4.1. Sequence-specific oligonucleotide probing

In this technique the gene of interest is amplified using generic primers, i.e. primers designed to anneal with DNA sequences common to all alleles of the loci of interest. The amplified PCR product is then immobilized onto support (e.g. nylon) membranes and the specificity of the products analysed by reacting the membranes with labelled oligonucleotides designed to anneal with polymorphic sequences present in each allele. By scoring the probes that bind to specific regions, it is possible to assign the HLA type. A recent modification of this technique, called

Table 4.1 Number of recognized HLA antigens/alleles.

(From Marsh et al. 2002 with permission.) Alleles

Antigens

HLA class I HLA-A HLA-B HLA-C119 HLA-E HLA-F HLA-G

250 490 9 6 1 15

24 50

HLA class II HLA-DRB1 HLA-DRA1 HLA-DRB3 HLA-DRB4 HLA-DRB5 HLA-DQB1 HLA-DQA1 HLA-DPB1 HLA-DPA1

315 3 38 12 150 53 22 99 20

17 — 1 1 1 6 — 6* —

* Cellularly defined.

reverse blot, involves the addition of the PCRamplified product to labelled probes immobilized on membranes (strips) or plates. PCR-SSOP is useful when large numbers of samples need to be HLA typed, e.g. bone marrow or cord blood donors. Sequence-specific priming

This technique involves the use of sequencespecific primers in the PCR step, i.e. primers designed to anneal with DNA sequences unique to each allele and locus. The detection of the PCRamplified product is then carried out by running the product on an agarose gel. This technique allows the rapid identification of the HLA alleles in individual samples since the readout of this method is the presence or absence of the product for which specific primers were used. This technique is therefore ideally suited to HLA typing individual samples, e.g. patients requiring HLAmatched platelets. However, although this is a very rapid procedure, many PCR reactions have to be set up per sample, e.g. at least 24 reactions for low-

39

Chapter 4

resolution DR typing. Furthermore, for PCR-SSP typing the target sequence of the alleles must be known since novel unknown sequences may not always be detected. Conformational analysis methods

Conformational methods depend on the mobility of PCR products in the gels. A number of variations of this technique have been described, including single-strand conformational polymorphism analysis (SSCP), double-strand conformational polymorphism analysis (DSCA), RSCA and heteroduplex formation. In the SSCP technique, PCR-generated DNA products are denatured by heating and rapid cooling to prevent reannealing of the strands. The products are run on a polyacrylamide gel with the mobility depending upon the secondary structure of the single-stranded DNA. The major disadvantage of this approach for HLA typing is the tendency of single-stranded DNA to adopt many conformational forms under the same electrophoretic conditions, resulting in the presence of several bands from a single product. A modification of this technique that compares the mobility in polyacrylamide of duplex molecules generated by mixing PCR products is called DSCA. In this case, the mobility depends on the mismatching of the sequence and the formation of heteroduplex molecules. A further modification of this technique, the RSCA, has recently been developed and successfully applied to HLA class I and II (DP) typing. In this technique the PCR amplification is carried out on the test DNA and on a reference DNA of known sequence, using fluorescently labelled primers for the PCR of the reference DNA. The PCR products are then mixed, allowed to anneal and run in an automated sequencer. Only those duplexes containing a labelled strand are detected, i.e. the homoduplex of reference DNA and the heteroduplexes of reference and test DNA. The mobility of every known allele with the reference DNA is then established and used to compare with that of the DNA under test. Heteroduplexing

Heteroduplexing is another DNA conformational 40

technique based on the fact that mismatched DNA hybrids (heteroduplexes) migrate at a slower rate through gels (due to the presence of singlestranded loops) than fully matched DNA duplexes. Heteroduplexes are formed during the annealing stage of the PCR when the sense strand of an allele binds to the antisense strand of a different allele. The banding pattern of these products following electrophoresis through a gel can be used to identify the alleles present. This technique is particularly useful for performing DNA-based ‘crossmatching’. In this case, HLA genes from the patient and potential donor are amplified, denatured by heating to about 95°C for several minutes and then mixed together under conditions which promote reannealing. Where the donor and patient alleles are similar but not identical, then, as in RSCA, homoduplexes are formed together with chimeric or heteroduplexes where sense strands from the patient anneal to antisense strands from the donor and vice versa. These heteroduplexes would have different migration rates through gels when compared with the homoduplexes. When this mix is electrophoresed through a gel, the banding pattern observed would contain all the bands that would be observed if PCR products from the patient and donor were to be electrophoresed separately. However, in addition to these there will be extra bands that are due to the heteroduplexes. The presence of extra bands in a heteroduplex ‘crossmatch’ would therefore indicate a difference in HLA type between the patient and donor. These extra bands would not be observed if the HLA alleles of the patient and donor were identical. The sensitivity of the technique can be increased by adding DNA from a known HLA allele that is not present in the patient or donor. One advantage of typing by conformational analysis when compared with methods such as PCR-SSP and PCR-SSOP is that new mutations not previously described can be readily picked up. A disadvantage of the technique is that, like SSCP, it cannot detect some HLA class II combinations while detecting too many silent mutations. Sometimes the banding patterns observed are

Human leucocyte antigens

complex and difficult to interpret. It also suffers in that it provides no actual HLA typing information. One of the main disadvantages of all the above described techniques is that they require DNA sequence data in order to design primers and probes. Sequencing-based typing

The principle of DNA sequencing is relatively straightforward and involves the denaturation of the DNA to be analysed to provide a single-strand template. A sequencing primer is then added and the DNA extension is performed by the addition of Taq polymerase in the presence of excess nucleotides. The sequencing mixture is divided into four tubes, each of which contains specific dideoxyribonucleoside triphosphate (ddATP). When this is incorporated into the DNA synthesis, elongation is interrupted with chain-terminating inhibitors. In each reaction there is random incorporation of the chain terminators and therefore products of all sizes are generated. The products of the four reactions are then analysed by electrophoresis in parallel lanes of a polyacrylamide–urea gel and the sequence is read by combining the results of each lane using an automated DNA sequencer. The sequencing products are detected by labelling the nucleotide chain inhibitors with radioisotopes and, more recently, with fluorescent dyes. In HLA SBT, some ambiguous results can be obtained with heterozygous samples and these may need to be retested by PCR-SSP. HLA typing using the RSCA and SBT techniques permit high-resolution HLA typing, which is known to be important in the selection of HLAmatched unrelated donors. A major advantage of all DNA-based techniques is that no viable cells are required to perform detailed HLA class I and II typing. Furthermore, since all the probes and primer are synthesized to order, there is a consistency of reagents used, allowing the comparison of HLA types from different laboratories. However, although serological typing is being rapidly replaced by DNA-based typing techniques, serological reagents may still be required for antigen expression studies.

The advantages and disadvantages of the various techniques described above are given in Table 4.2.

HLA antibodies HLA-specific antibodies may be produced in any situation that exposes the host to these alloantigens, including pregnancy, transplantation, blood transfusions and planned immunizations. However, the affinity, avidity and class of the antibody produced will depend on various factors, including the route of immunization, the persistence and type of immunological challenge and the immune status of the host. Cytotoxic HLA antibodies can be identified in approximately 20% of human pregnancies. The antibodies produced are normally multispecific, high titre, high affinity and of the IgG class. Although these IgG antibodies can cross the placenta, they have not been shown to be harmful to the fetus. Conversely, the antibodies produced following transplantation seem to be largely dependent on the degree of HLA mismatch between donor and recipient and the majority of these antibodies formed are IgG, although a few IgM antibodies have also been identified. In contrast, the majority of HLA antibodies found in multitransfused patients are multispecific IgM and IgG and are mostly directed at public epitopes. The recent introduction of leucocytedepleted blood components (see Chapter 22) may lead to a reduction in alloimmunization in naive recipients but it may not be very effective in preventing alloimmunization in already sensitized recipients, i.e. women who have become immunized as a result of pregnancy. Finally, normal individuals were previously deliberately immunized with HLA-mismatched cells in order to produce potent HLA-specific reagents, but the deliberate immunization of these healthy individuals is nowadays difficult to justify ethically. However, at present, planned HLA immunization is still carried out as a form of treatment for women with a history of recurrent spontaneous abortion. These women are immunized with lymphocytes from their partners or a third party to attempt to induce an immunomodulatory 41

Chapter 4 Table 4.2 Advantages and disadvantages of DNA-based techniques.

Technique

Advantages

Disadvantages

Sequence-specific oligonucleotide probing (SSOP)

Needs only one pair of genetic primers, fewer reactions to set up Larger number of samples can be processed simultaneously Requires small amount of DNA Cheap

Different temperatures required for each probe Probes can cross-react with different alleles Large numbers of probes required to identify specificity Difficult to interpret pattern of reactions

Sequence-specific priming (SSP)

Provides rapid typing with higher resolution than SSOP All PCR amplifications are carried out at same time, temperature and conditions Fast and simple to read and interpret

Too many sets of primers are needed to fully HLA type Requires a two-stage amplification to provide HR typing

Reference strand conformational analysis (RSCA)

Easy to perform Provides higher resolution than SSOP and SSP

Requires expensive equipment Requires established data on viability value of each allele studied

Sequencing-based typing (SBT)

Provides the highest level of resolution Able to identify new alleles Does not require previous sequence data to identify new allele

Not easy to perform Requires expensive reagents and equipment Difficult to interpret Requires DNA sequence data to compare results Slower than SSOP (reverse blot), SSP and RSCA

response that results in the maintenance of the pregnancy.

Detection of HLA antibodies Over the years, a number of techniques to detect HLA antibodies have been described. These include the complement-dependent lymphocytotoxicity (LCT) test, enzyme-linked immunosorbent assay (ELISA) and flow cytometry. More recently, a new method to detect HLA antibodies, Luminex, has been described. Lymphocytotoxicity

The LCT assay, developed by Terasaki and McClelland (1964), is the most commonly used. In this technique, equal volumes of serum and cells are mixed and incubated to allow the binding of the specific antibody to the target cell. This is followed by the addition of rabbit complement and a further incubation step. Complement-fixing antibodies reacting with the HLA antigen present on 42

the cell surface leads to the activation of complement via the classical pathway and results in the disruption of the cell membrane. The lysed cells are then detected by adding ethidium bromide (EB) and acridine orange (AO) at the end of the incubation period. Live cells actively take up AO and under ultraviolet (UV) light they appear green, whereas lysed cells allow the entry of EB, which binds to DNA, and they appear red under UV light. There are a number of alternatives to EB and AO including carboxyfluorescein diacetate and EB, eosin or trypan blue. The reactions are scored by estimation of the percentage of dead cells in each well after establishing baseline values in the negative and positive controls (Table 4.3). One of the main disadvantages of the LCT assay is that it does not discriminate between HLA and non-HLA cytotoxic lymphocyte-reactive antibodies including autoantibodies, which are common in thrombocytopenic patients, particularly in post-BMT patients, in whom there is a dysfunction of the immune system. Fortunately, the majority of these lymphocytotoxic autoantibodies are IgM and can be identified by screening the

Human leucocyte antigens Table 4.3 Lymphocytotoxicity grading.

Cell death (%)

Score

Interpretation

0–10 11–20 21–50 51–80 81–100

1 2 4 6 8 0

Background cell death, negative Doubtful negative Weak positive Positive Strong positive Unreadable/invalid

serum with and without dithiothreitol (DTT). The addition of DTT to the serum results in the breakdown of the intersubunit disulphide bonds in the IgM molecule, leading to the loss of cytotoxicity due to IgM. Prolonged exposure or excess DTT can lead to the breakdown of intramolecular disulphide bonds in the IgG molecules and also inactivate complement, but this can be inhibited by the addition of cystine. In our laboratory, we tested 104 serum samples from immunologically refractory patients receiving HLA-matched platelet transfusions, 50% of the samples being positive by LCT but only 40% being positive when screened with DTT, illustrating the importance of using DTT to remove the reactivity due to IgM, non-HLA lymphocytotoxic antibodies (10%). The presence of lymphocytotoxic autoreactive antibodies in itself is not thought to be of clinical significance in solid organ transplant recipients; however, the clinical relevance of these antibodies in immunological refractoriness to random platelet transfusions is not yet clear. Another disadvantage of the LCT test is that it only detects cytotoxic HLA-specific antibodies and therefore the presence of non-cytotoxic HLAspecific antibodies may be missed. Non-cytotoxic antibodies are best detected by using an ELISA or by flow cytometry, as described below. Enzyme-linked immunosorbent assay

ELISA-based methods have often been the technique of choice for antibody detection for a number of antigen systems, particularly where there has been a requirement for testing large numbers of samples.

The basic principle of the technique is as follows. HLA antigens are purified and immobilized on a microwell plate, directly or via an antibody directed against a non-polymorphic region of the HLA antigen. HLA-specific antibodies bound to the immobilized antigen can be detected with an enzyme-linked secondary antibody which, upon addition of specific substrate, catalyses a colour change reaction that is detected in an ELISA reader (GTI QuickScreen). This ELISA test can only be used to determine the presence or absence of HLA-specific antibody. In order to detect HLA specificity, an alternative technique is used. In this case, HLA antigen is isolated from a selected panel of cell donors or cell lines derived from these donors. Antibodies directed against the non-polymorphic region of the HLA class I molecule, i.e. the a3 domain, are used to immobilize the specific HLA antigen to the microwell, ensuring that the more polymorphic a1 and a2 domains are available for antibody binding. The test is designed to cover all the major HLA specificities at least once; thus it should be possible to determine antibody specificity in those sera which show a restricted panel reactivity. Three commercial kits are now available to perform the specificity screening, SangStat.-PRASTAT, GTI QuickID and Lambda Antigen Tray LAT. One of the main advantages of the ELISA technique is that the non-cytotoxic antibodies detected are HLA specific since it relies on the binding of the antibodies to wells coated with pools of solubilized HLA antigens. It is also more sensitive and in our laboratory the ELISA test achieves a 7% increase in sensitivity over the LCT test (49% vs. 42%). The precise identification of the type, specificity and titre of antibodies is important not only for the diagnosis but also to establish the best course of treatment, since in the case of patients immunologically refractory to platelet transfusions, the majority of these patients can benefit from the provision of HLA matched or crossmatched negative platelets (see below). Luminex

This technique uses fluorochrome-dyed poly43

Chapter 4

styrene beads coated with specific antigens. The precise ratio of these fluorochromes creates 100 distinctly coloured beads, each of them coated with a different antigen. The beads are then incubated with the patient’s serum and the reaction is developed using a PE-conjugated antihuman IgG (Fc specific) antibody. Flow cytometry

The use of flow cytometric techniques was initially investigated as an alternative crossmatch technique and was shown to be more sensitive than LCT. The increased sensitivity may be attributed to a number of factors, one of which is the additional reactivity due to the detection of non-cytotoxic antibodies, some of which may be HLA specific. As with LCT screening, cells and serum are mixed and incubated to allow the binding of the antibody to the target antigen. The bound antibody is detected by using an antibody labelled with a fluorescent marker such as fluorescein isothiocyanate or R-phycoerythrin against human immunoglobulin. The flow cytometer can then be used to identify the different cell populations based on their morphology/granularity and on the fluorescence. Test

sera with median fluorescence values greater than the mean + 3SD of the negative controls are considered positive. A modification of the above procedure can be used to detect antibodies reacting with T or B cells. The main advantages of flow cytometric techniques are the increased sensitivity when compared with LCT and the detection of noncomplement fixing antibodies, allowing the early detection of sensitization. However, one of the disadvantages is that it also detects non-HLA and lymphocyte-reactive antibodies, and the clinical relevance of these antibodies is unclear. The relative advantages and disadvantages of each of these techniques are given in Table 4.4.

Clinical relevance of HLA antigens and antibodies Although the main role of the HLA molecules is to present antigens to T cells, HLA molecules can themselves be recognized as foreign by host T cells by a mechanism known as allorecognition. Two pathways of allorecognition have been identified, direct and indirect.

Table 4.4 HLA antibody screening techniques.

Advantages

Disadvantages

Lymphocytotoxicity test (LCT)

Well established Robust Requires small amount of serum Used for antibody screening and crossmatching Low cost

Viable cells required Needs separated T and B cells for class I and class II antibody screening Large and selected panel of cells required Detects non-HLA cytotoxic antibodies, e.g. autoantibodies

Enzyme-linked immunosorbent assay (ELISA)

Easy to standardize Objective readout Suitable for bulk testing Detects cytotoxic and non-cytotoxic HLA-specific antibodies More sensitive than LCT Medium cost

Not yet well established to define class I and class II specificities Large amounts of serum required Cannot be used for crossmatching

Flow cytometry

Highly sensitive Detects weak and early sensitization Detects cytotoxic and non-cytotoxic antibodies Can define class I and class II antibodies simultaneously

Not well standardized Large panel of cells required to establish antibody specificity Expensive Detects non-HLA antibodies

44

Human leucocyte antigens

In the direct allorecognition pathway, the host’s T cells recognize HLA molecules (primarily class II) expressed on donor tissues, e.g. tissue dendritic cells and endothelial cells. Indirect allorecognition involves the recognition of donor-derived HLA class I and II antigenic peptides presented by the host’s own antigen-presenting cells. Because of this mechanism, HLA antigens are therefore one of the main barriers to the success of solid organ transplantation or BMT and are responsible for the strong alloimmunization seen in patients following transplantation or blood transfusion. Solid organ transplantation

Several studies have confirmed the importance of matching for HLA-A, -B and -DR antigens. Results published by the United Network for Organ Sharing (UNOS) registry have shown that 1-year graft survival in recipients of fully HLA-A, -B and -DR matched kidneys was 94%, as opposed to the 89% and 90% survival rate observed in recipients of one HLA haplotype-matched kidney from a parent or a sibling, respectively. The differences between match grades, which are minimal at 1 year after transplant, become more apparent with increased follow-up half-life figures. Studies on the relative contribution of HLA-A, -B and -DR compatibility on the outcome of graft survival have shown that HLA-DR has the strongest effect. When grafts from cadaver donors are analysed, the 1-year graft survival rate is 88% for HLA-A, -B and -DR matched and 79% for mismatched kidneys. With the application of the PCR-based techniques, it is now possible to identify molecular differences between otherwise serologically identical HLA types of donor and recipient pairs. Correlation of these results with graft survival has shown a higher graft survival rate when recipients and donors are HLA-DR identical by serological and molecular techniques compared with when they are HLA-DR identical by serological but not molecular methods (87% vs. 69%). The use of DNA-based techniques allows the identification of ‘minor’ mismatches that were previously undetected by serological typing, particularly in the DRb1 chain.

The role of HLA antibodies in solid organ transplantation is well established. The presence of circulating HLA-specific antibodies directed against donor antigens in renal and cardiac recipients has been associated with hyperacute rejection of the graft. It is therefore important that these antibodies are detected and identified as soon as the patient is registered on the transplant waiting list to ensure that incompatible donors are not considered for crossmatching. Furthermore, recent data have also suggested that the appearance of donorspecific antibodies after transplantation may be a sign of rejection, indicating the importance of post-transplant monitoring for some groups of patients. HLA and BMT

The main risk factors affecting the survival of patients undergoing BMT are graft-versus-host disease (GVHD), leukaemia relapse, and graft rejection or graft failure. The probability of developing acute GVHD is directly related to the degree of HLA incompatibility. Although BMT between HLA-identical siblings ensures matching for the classical HLA genes, acute GVHD still develops in about 20–30% of patients transplanted with an HLA-identical sibling. This is probably due to the effect of untested HLA antigens, e.g. DP, or minor histocompatibility antigens in the activation of donor T cells. However, patients receiving grafts from HLA-matched unrelated donors have a higher risk of developing GVHD than those transplanted using an HLA-identical sibling. Original studies on the role of HLA in the outcome of BMT indicated that HLA-DR incompatibility was one of the main risk factors associated with the development of GVHD. However, more recent studies have shown that mismatches at the HLA-A alleles, and to a lesser extent HLA-C alleles, are independent risk factors for death in patients following haemopoietic stem cell transplantation from an unrelated donor. Furthermore, mismatches at the HLA-C locus, as detected by DNA sequencing, has been shown to be associated with graft failure in patients receiving an unrelated BMT. 45

Chapter 4

The use of DNA-based methods for the detection of the HLA polymorphisms mentioned above has provided a unique opportunity to improve HLA matching of patients and unrelated donors and to reduce the development of GVHD. However, it has been shown that the increased GVHD seen as a result of HLA mismatch may result in a lower relapse rate, probably due to a GVL response associated with the graft-versushost response. Furthermore, the use of T-celldepleted marrow, which has successfully decreased the incidence of GVHD, also resulted in an increased incidence of leukaemia relapse. Thus it appears that mature T cells in the marrow, which may be responsible for GVHD, may also be involved in the elimination of residual leukaemic cells. More recently, cord blood has been successfully used as a source for haemopoietic stem cells for patients requiring marrow reconstitution. Preliminary clinical data have shown reduced risk and severity of GVHD following HLA-matched and HLA-mismatched cord blood transplantation. It is possible that the immunological immaturity of cord blood mononuclear cells compared with mononuclear cells present in adult bone marrow may result in a reduced GVHD effect. The impact of this on the relapse rates is not yet clear. Conversely, the rate of graft rejection is significantly higher in recipients of an HLA-mismatched transplant than in those receiving a transplant from an HLA-identical sibling (12.3% vs. 2.0%). Graft failure, which is thought to be mediated by residual recipient T and/or NK cells reacting with major or minor histocompatibility antigens present in the donor marrow cells, has also been shown to be due to antibodies reacting with donor’s HLA antigens. Thus, rejection is particularly high in HLA-alloimmunized patients. Furthermore, studies on leukaemic patients with donor-specific antibodies and with a positive T- or B-cell cytotoxic crossmatch have a high incidence of graft failure compared with those with a negative crossmatch. Preformed cytotoxic antibodies can also increase the incidence of graft failure in patients with aplastic anaemia. However, in spite of these reports, HLA antibodies are particularly relevant in the post-BMT setting, where patients 46

receiving multiple transfusions can experience immunological refractoriness to random platelet transfusions due to the presence of HLA antibodies. These patients require transfusions of HLA-matched platelets (see Chapter 9). Blood transfusion

The clinical relevance of HLA antigens and antibodies in blood transfusion has been well documented. White cells present in transfused products express antigens which, if not identical to those present in the recipient, are able to activate T cells and lead to the production of antibodies and/or effector cells responsible for some of the serious complications of blood transfusion. On the other hand, antibodies (and sometimes cells) present in the transfused product may react directly with the relevant antigens in the recipient and in this way provoke a transfusion reaction. The immunological reactions due to the passive transfer of HLA antibodies or HLA-specific effector cells include transfusion-related acute lung injury (TRALI) and transfusion-associated graft-versus-host disease (TA-GVHD). TRALI is a relatively under-reported complication of blood transfusion, characterized by acute respiratory distress, pulmonary oedema and severe hypoxia. The development of TRALI has been associated with the transfusion of blood components containing HLA and HNA antibodies reacting with the recipient white cells, causing complement activation, accumulation of neutrophils in the lungs and oedema. TRALI cases have also been associated with the presence of white cell antibodies in recipients reacting with transfused leucocytes and/or to inter-donor antigen–antibody reactions in pooled platelets. TA-GVHD is a rare but often severe and fatal reaction associated with the transfusion of cellular blood components. It occurs primarily in immunosuppressed individuals, although it can also occur in immunocompetent recipients. In this condition, immunocompetent T lymphocytes present in blood or blood products are able to recognize HLA and/or minor histocompatibility antigens present on the recipient cells and induce a GHVD reaction similar to that seen following haemopoi-

Human leucocyte antigens

etic stem cell transplantation. Diagnosis depends on finding evidence of donor-derived cells, chromosomes or DNA in the blood and/or affected tissues of the recipient. CD4+ and CD8+ cytotoxic as well as CD4+ T-cell clones lacking direct cytotoxicity but with supernatants containing TNF-b have been isolated from the lesion of patients with TA-GVHD. Immunological reactions due to antibodies present in the recipient include non-haemolytic febrile transfusion reaction (NHFTR) and immunological refractoriness to random platelet transfusions. The occurrence of NHFTR has been commonly associated with the presence of HLA antibodies in the recipient reacting with white blood cells present in the transfused product. However more recently, and particularly in countries that have introduced universal leucodepletion, it has been observed that NHFTR can also be triggered by the direct action of cytokines such as IL-1b, TNF-a, IL-6 and/or by chemokines such as IL-8 which are found in transfused products. On the other hand, immunological refractoriness to random platelet transfusions is primarily due to HLA and, to a lesser extent, HPA and hightitre ABO alloantibodies present in the patient and destroying the transfused platelets. This results in the lack of platelet increments following the transfusion of random donor platelets. Although HLA antibodies are found in approximately 20–50% of multitransfused patients, only approximately 10–20% of them become immunologically refractory and require HLA-matched or HLAcompatible platelet transfusions. Following the introduction of universal leucodepletion, the proportion of multitransfused patients with HLA antibodies seems to have decreased to approximately 10–20% and only about 3–5% of these patients are immunologically refractory. These are previously sensitized transplanted or transfused recipients including multiparous women.

HLA and disease HLA genes are known to be associated with a variety of autoimmune, non-autoimmune and, more recently, infectious diseases. A number of

different mechanisms to explain this association have been postulated, including LD with the relevant disease susceptibility gene, the preferential presentation of the pathogenic peptide by certain HLA molecules, and molecular mimicry between certain pathogenic peptides and host-derived peptides. Among those shown to be due to linkage with the relevant HLA-related genes is hereditary haemochromatosis (HH) and among those diseases in which HLA antigens seem to be involved in the preferential presentation of peptides to antigenic T cells is neonatal alloimmune thrombocytopenia. Diseases in which the molecular mimicry mechanism has been postulated include ankylosing spondylitis and Klebsiella infection. However, the precise pathogenic mechanisms involved remain unknown. A number of diseases associated with both HLA class I and II have been described (Table 4.5).

Table 4.5 HLA-associated diseases.

HLA class I genes Birdshot chorioretinopathy: HLA-A29 Behçet’s disease: HLA-B51 Ankylosing spondylitis: HLA-B27 Psoriasis: HLA-Cw6 Malaria: HLA-B53 HLA class II genes Rheumatoid arthritis HLA-DRB1*0401 HLA-DRB1*0404 HLA-DRB1*0405 HLA-DRB1*0408 HLA-DRB1*0101/0102 HLA-DRB1*1402 HLA-DRB1*1001 Narcolepsy: HLA-DQB1*0602/DQA1*0102 Coeliac disease: HLA-DQB1*0201/DQA1*0501 Neonatal alloimmune thrombocytopenia: HLA-DRB3*0101 Malaria: HLA-DRB1*1302/DQB1*0501 Insulin-dependent diabetes mellitus: HLA-DQB1*0302/DQA1*0301 HLA-linked diseases Haemochromatosis: (HLA-A3) HFE gene C282Y, H63D and S65C 21-OH deficiency: (HLA-B47) 21-OH gene

47

Chapter 4 H63

Hereditary haemochromatosis

HH is a common genetic disorder in northern Europe, where between 1 in 200 and 1 in 400 individuals suffer from the disease, with an estimated carrier frequency of between 1 in 8 and 1 in 10. Clinical manifestations include cirrhosis of the liver, diabetes and cardiomyopathy. Detection of asymptomatic iron overload is important since removal of excess iron by phlebotomy can prevent organ damage. Previous screening methods relied on the measurement of iron saturation, confirmed with a fasting sample. Confirmation of HH was by liver biopsy. Hence a non-invasive screening method would be an advantage. Previously, a close association between HLA-A3 and HH had been described and, until recently, HLA-A3 was the only test available to aid diagnosis, although this was not very specific since the majority of HLAA3-positive individuals do not have HH. A number of mutations have been identified in the HFE gene, which is located 3 Mb telomeric of the HLA region. Clinical data indicate that at least three of these mutations (C282Y, H63D and S65C) may predispose and affect the clinical outcome of this condition. Over 90% of HH patients in the UK are homozygous for the mutation that replaces a cysteine (C) with a tyrosine (Y) at codon 282 in the HFE gene. The second and third mutations (H63D and S65C) are thought to be less important, although it may have an additive effect if inherited with the first mutation (Fig. 4.5). Recent studies on blood donors have shown that approximately 1 in 280 donors are homozygous for the mutations. A DNA-based technique to detect these three mutations simultaneously has now been developed in our laboratories and provides a simple, rapid and unambiguous definition of these mutations. Neonatal alloimmune thrombocytopenia

Neonatal alloimmune thrombocytopenia is due to fetomaternal incompatibility for human platelet antigens (see also Chapters 5 and 8). More than 80% of cases occur in women who are homozygous for the HPA-1b allele. Although the majority of cases are associated with the presence of HPA48

a1

D

a2 NH2

NH2

b2m

a3

C282

COOH

Y

COOH Fig. 4.5 HFE molecule. b2-m, b2-microglobulin.

1a antibodies, about 15% of cases are due to anti-HPA-5b. Early studies indicated that the production of HPA-1a antibodies is strongly associated with the HLA-DRB3*0101 allele. However, only approximately 35% of HPA-1a-negative, DRB3*0101positive women develop antibodies upon exposure to the antigen, suggesting that other genes or factors may be involved in the development of alloimmunization against HPA-1a. It has been shown that the amino acid substitution of leucine for proline at position 33 on the GPIIIa chain is responsible for the production of alloantibodies and, more recently, T cells responding to a peptide containing this residue have been identified in an HPA-1b/b patient with an affected child.

Further reading Brown C, Navarrete C. HLA antibody screening by LCT, LIFT and ELISA. In: Bidwell J, Navarrete C, eds. Histocompatibility Testing. London: Imperial College Press, 2000: 65–98. Campbell RD. The human major histocompatibility complex: a 4000-kb segment of the human genome replete with genes. In: Davies KE, Tilghman SM, eds.

Human leucocyte antigens Genome Analysis, Vol. 5. Regional Physical Mapping. New York: Cold Spring Harbor Laboratory Press, 1993: 1–33. Dyer PA, Claas FHJ. A future for HLA matching in clinical transplantation. Eur J Immunogenet 1997; 24: 17–28. Gruen JR, Weissman SM. Evolving views of the major histocompatibility complex. Blood 1997; 90: 4252–65. Harrison J, Navarrete C. Selection of platelet donors and provision of HLA matched platelets. In: Bidwell J, Navarrete C, eds. Histocompatibility Testing. London: Imperial College Press, 2000: 379–90. Howell M, Navarrete C. The HLA system: an update and relevance to patient–donor matching strategies in clinical transplantation. Vox Sang 1996; 71: 6–12. Lardy NM, Van Der Hjorst AR, Ten Berge IJM et al. Influence of HLA-DRB1* incompatibility on the occurrence of rejection episodes and graft survival in serologically HLA-DR-matched renal transplant combinations. Transplantation 1997; 64: 612–16. Madrigal JA, Arguello R, Scott I, Avakian H. Molecular histocompatibility typing in unrelated donor bone marrow transplantation. Blood Rev 1997; 11: 105–17.

Madrigal JA, Scott I, Arguello R, Szydlo R, Little A-M, Goldman JM. Factors influencing the outcome of bone marrow transplants using unrelated donors. Immunol Rev 1997; 157: 153–66. Marsh SG, Albert ED, Bodmer WF et al. Nomenclature for factors of the HLA system. Tissue Antigens 2002; 60: 407–64. Murac, Raguenes O, Ferec C. HFE mutations analysis in 711 haemochromatosis probands: evidence for S65C implication in mild form of hemochromatosis. Blood 1999; 93: 2502–5. Parham P, McQueen KL. Alloreactive killer cells: hindrance and help for haematopoietic transplants. Nat Rev Immunol 2003; 3: 108–21. Suthanthiran M, Strom TB. Renal transplantation. N Engl Med J 1994; 331: 365–76. Terasaki PL, McClelland JD. Microdroplet assay of human serum cytokines. Nature 2000; 204: 998–1000. Thorsby E. HLA associated diseases. Hum Immunol 1997; 53: 1–11.

49

Chapter 5

Platelet and neutrophil antigens David L. Allen, Geoffrey F. Lucas, Willem H. Ouwehand and Michael F. Murphy

Human platelet antigens As on red cells, antigens on human platelets can be categorized according to their biochemical nature. • Carbohydrate antigens on glycolipids and glycoproteins (GPs): A, B and O antigens, P, Le. • Protein antigens: human leucocyte antigen (HLA) class I A, B and C, GPIIb/IIIa, GPIb/IX/V, etc. • Haptens: quinine, quinidine; heparin; some antibiotics, e.g. penicillins and cephalosporins. These antigens can be targeted by some or all of the following types of antibodies: • autoantibodies; • alloantibodies; • isoantibodies; and • drug-dependent antibodies. Many platelet antigens are shared with other cells, e.g. ABO and HLA class I (Table 5.1). This section focuses on protein alloantigens expressed predominantly on platelets, although some of these are also present to a lesser extent on some other blood cells, e.g. human platelet antigen (HPA)-5 on activated T lymphocytes. These antigens are commonly referred to as platelet-specific alloantigens or human platelet alloantigens. Human platelet alloantigens

There are a number of well-characterized biallelic platelet alloantigen systems, and a number of rare, private or low-frequency antigens have also been described (Table 5.2). Most of these antigens were first discovered during the investigation of cases of neonatal alloimmune thrombocytopenia (NAIT). Platelet-specific alloantigens are located on platelet membrane GPs involved in haemostasis 50

through interactions with extracellular matrix proteins in the vascular endothelium and with plasma coagulation proteins. The majority of these antigens are on the GPIIb/IIIa complex, which plays a central role in platelet aggregation as a receptor for fibrinogen, fibronectin, vitronectin and von Willebrand factor. Other important GPs are GPIb/IX/V, the main receptor for von Willebrand factor, which is involved in platelet adhesion to damaged vascular endothelium; GPIa/IIa, which is involved in adhesion to collagen; and CD109, which also appears to be a collagen receptor. Congenital deficiency of these GPs results in bleeding disorders, e.g. lack of GPIIb/IIIa causes Glanzmann’s thrombasthenia and absence of GPIb/IX/V results in Bernard–Soulier syndrome. The expression of platelet alloantigens located on these GPs may be altered in these disorders and HPA typing performed by serological assays (‘phenotyping’) may give discrepant results when compared with results obtained by molecular typing (‘genotyping’). Inheritance and nomenclature

Most of the platelet alloantigen systems reported to date have been shown to be biallelic, with each allele being codominant. Historically, systems were named by the authors first reporting the system, usually using an abbreviation of the name of the patient in whom the antibody was detected. Some systems were published simultaneously by different laboratories and with different names, e.g. Zw and PlA or Zav/Br/Hc, and only later were they found to be the same polymorphism. In 1990, a working party for platelet immunology of the International Society of Blood Transfusion (ISBT)

Platelet and neutrophil antigens Table 5.1 Antigen expression on peripheral blood cells.

Antigens

Erythrocytes

Platelets

Neutrophils

B lymphocytes

T lymphocytes

Monocytes

A, B, H I Rh* K HLA class I HLA class II GPIIb/IIIa GPIa/IIa GPIb/IX/V CD109

+++ +++ +++ +++ -/(+) -

(+)/++ ++ +++ +++ +++ +++ (+)/++§

++ +++ -/+++§ (+)‡ -

+++ +++ -

+++ -/+++§ ++§ -/++§

+++ +++ +++D

* Non-glycosylated. § On activated cells. ‡ GPIIIa (b3) in association with an alternative a chain (av). D Dependent on monoclonal antibody used to assess expression.

agreed a new nomenclature for platelet polymorphisms, the HPA nomenclature. The recently founded international Platelet Nomenclature Committee has published guidelines for acceptance and naming of newly discovered platelet GP alloantigens. In the HPA nomenclature, each system is numbered consecutively (HPA-1, -2, -3 and so on) (see Table 5.2) according to its date of discovery, with the high-frequency allele in each system being designated ‘a’ and the low-frequency allele ‘b’. Newly discovered systems are only officially included when confirmed by a second party and approved by the nomenclature committee. If an antibody against only one allele has been reported, a ‘w’ (for workshop) is added after the antigen name, e.g. HPA-10bw. One possible reason why antibodies against the ‘a’ antigen have not yet been reported for many of the recently discovered systems is that the ‘b’ allele is of such low frequency that ‘bb’ homozygous individuals either do not exist or are extremely rare. In Caucasian populations, the allele frequency for the majority of HPA systems is skewed towards the ‘a’ allele and homozygosity for the ‘b’ allele is below 3%. This places significant pressures on the blood services in the management of alloimmunized patients since compatible red cells and platelets are difficult to obtain for patients with antibodies against the ‘a’ alloantigen. Allele fre-

quencies vary between populations, e.g. GPIIIaproline-33 (HPA-1b) is extremely rare or absent in the Far East, while GPIIIa-glutamine-143 (HPA4b) does not occur in Caucasians. These differences are important when investigating cases of suspected platelet alloimmunity in different ethnic groups. Until the early 1990s, platelet typing was performed by serological assays. These assays required the use of monospecific antisera, which were relatively uncommon as the majority of immunized individuals produced HLA class I antibodies in addition to the platelet-specific antibodies. The typing that could be performed therefore was limited, and many laboratories were only able to phenotype for HPA-1a. The publication of more advanced assays, such as monoclonal antibody-specific immobilization of platelet antigens (MAIPA) (Fig. 5.1) permitted more extensive phenotyping but some antisera were simply not available. With the advent of techniques such as immunoprecipitation of radioactively labelled platelet membrane proteins, MAIPA and the polymerase chain reaction (PCR), the molecular basis of the majority of clinically relevant platelet-specific alloantigen systems was elucidated (Fig. 5.1 and see Table 5.2). The molecular basis for all the HPA alloantigen systems has been determined and in all but one (HPA-14bw) the difference between the 51

Chapter 5 Table 5.2 Platelet-specific alloantigen systems.

System

Antigen

Alternative names

HPA-1

HPA-1a HPA-1b

Zwa, P1A1 Zwb, P1A2

HPA-2

HPA-2a HPA-2b

Kob Koa, Siba

HPA-3

HPA-3a HPA-3b

Baka, Leka Bakb

HPA-4

HPA-4a HPA-4b

Yukb, Pena Yuka, Penb

HPA-5

HPA-5a HPA-5b

Brb, Zavb Bra, Zava, Hca

HPA-6

HPA-6bw

HPA-7

Glycoprotein (GP)

Nucleotide change

Amino acid change

97.9 28.8

GPIIIa

T196 C196

Leucine33 Proline33

>99.9 13.2

GPIba

C524 T524

Threonine145 Methionine145

GPIIb

T2622 G2622

Isoleucine843 Serine843

>99.9 <0.1

GPIIIa

G526 A526

Arginine143 Glutamine143

99.0 19.7

GPIa

G1648 A1648

Glutamic acid505 Lysine505

Caa, Tua

0.7

GPIIIa

G1564 A1564

Arginine489 Glutamine489

HPA-7bw

Mo

0.2

GPIIIa

C1267 G1267

Proline407 Alanine407

HPA-8

HPA-8bw

Sra

<0.01

GPIIIa

C2004 T2004

Arginine636 Cysteine636

HPA-9

HPA-9bw

Maxa

0.6

GPIIb

G2603 A2603

Valine837 Methionine837

HPA-10

HPA-10bw

Laa

<1.6

GPIIIa

G281 A281

Arginine62 Glutamine62

HPA-11

HPA-11bw

Groa

<0.25

GPIIIa

G1996 A1996

Arginine633 Histidine633

HPA-12

HPA-12bw

Iya

0.4

GPIbb

G141 A141

Glycine15 Glutamic acid15

HPA-13

HPA-13bw

Sita

0.25

GPIa

C2531 T2531

Threonine799 Methionine799

HPA-14

HPA-14bw

Oea

<0.17

GPIIIa

D AAG1929–1931

D Lysine611

HPA-15

HPA-15a HPA-15b

Govb Gova

74 81

CD109

C2108 A2108

Serine703 Tyrosine703

HPA-16

HPA-16bw

Duva

<1

GPIIIa

C517 T517

Threonine140 Isoleucine140

Vaa PlT Vis Pea Dya Moua

<0.4 >99.9

* Frequencies based on studies in Caucasians.

52

Phenotype frequency* (%)

80.95 69.8

26

GPIIb/IIIa GPV GPIV GPIba 38 kDa GP Unknown

Platelet and neutrophil antigens

(a)

Patient serum containing anti-HPA-1a

1.

+

Hu

Donor platelets

Fig. 5.1 Monoclonal antibody-specific

immobilization of platelet antigens. (1) A cocktail of target platelets, murine monoclonal antibody (MoMab) directed against the glycoprotein being studied, e.g. GPIIb/IIIa and human serum is prepared: (a) the test serum contains anti-HPA-1a; (b) no anti-platelet antibodies are present. (2) After incubation a trimeric (a) or dimeric (b) complex is formed. Excess serum and MoMab is removed by washing. (3) The platelet membrane is solubilized in a non-ionic detergent, releasing the complexes into the fluid phase and particulate matter is removed by centrifugation. (4) The lysates containing the glycoprotein–antibody complexes are added to the wells of a microtitre plate previously coated with goat antimouse antibody. (5) Unbound lysate is removed by washing and an enzymeconjugated goat anti-human antibody added. (6) Excess conjugate is removed by washing and a substrate solution added. Cleavage of the substrate, i.e. a colour reaction, indicates binding of human antibody to the target platelets

+

MoMab HLA class-1

GPIIb/IIIa

GPIa/IIa

MoMab

GPIIb/IIIa Donor platelets

Microplate

MoMab

Hu

(b)

Inert serum (negative control)

MoMab

2. Donor platelets

Donor platelets

Hu

4.

MoMab

MoMab

3.

Hu

MoMab

MoMab Goat anti-mouse

Goat anti-mouse

Enzyme-conjugated goat anti-human Ig

Hu

5. MoMab

MoMab Goat anti-mouse

Goat anti-mouse

Hu

6.

MoMab

two alleles is based on a single nucleotide difference that results in a single amino acid substitution. Most polymorphisms have been found on GPIIb/IIIa (CD61/CD41) (see Table 5.2 and Fig. 5.2). Based on this knowledge, many molecular typing techniques have been developed over the last decade and these have largely overcome problems in platelet typing. One such assay is PCR using sequence-specific primers (PCR-SSP). This is a fast and reliable molecular typing technique with

MoMab Goat anti-mouse

Goat anti-mouse

minimal post-PCR handling and has become one of the cornerstone techniques in HLA typing and is widely used for HPA genotyping (Fig. 5.3). High-throughput DNA-based typing techniques with automated readout are under development and will be used in platelet immunology reference laboratories in the near future. Knowledge of the genetic basis of plateletspecific antigens makes it possible to carry out molecular genotyping on whatever DNAcontaining material is available, e.g. platelet typing 53

Chapter 5 GPIIb/IIIa aIIbb3 CD61/CD41

Fibrinogen binding site RGD binding site Fig. 5.2 Schematic representation of

GPIIIa

HPA-4 Arg/Gln143

GPIIba Ca++

HPA-7 Pro/Ala407 HPA-16 Thr/Ile140 HPA-10 Arg/Gln62

HPA-9 Val/Met837

Ca++ Ca++

HPA-6 Arg/Gln489

Ca++

HPA-1 Leu/Pro33

S S

HPA-14 Lys611 deletion HPA-8 Arg/Cys636

HPA-11 Arg/His633 COOH

GPIIbb COOH

—

—

Fig. 5.3 PCR-SSP determination of HPA-1–5 genotypes.

The upper band present in all lanes is the 429-bp product of human growth hormone. The lower bands are the products of sequence-specific primers. The results are read from left to right, i.e. lane 1 HPA-1a, lane 2 HPA-1b, etc. The HPA genotype in this case is 1b/1b, 2a/2a, 3a/3b, 4a/4a, 5a/5a. Courtesy of Dr Paul Metcalfe.

using fetal DNA from amniocytes or from chorionic villous biopsy samples. However, when used in the setting of first-trimester fetal HPA typing, extreme caution is required to exclude possible contamination with maternal cells and consequent erroneous typing. Platelet isoantigens, autoantigens and haptens

In some individuals, certain GPs may be absent from the platelet surface, e.g. approximately 4% of black and 10% of Japanese individuals do not express GPIV (CD36) on their platelets. If these individuals are exposed to normal, GPIV-positive 54

HPA-3 Ile/Ser843

platelet GPIIb/IIIa or the aIIbb3 integrin. GPIIIa is recognized by murine monoclonal antibodies of the CD61 cluster and the heterodimer by antibodies of the CD41 cluster. The amino acid substitutions arising from the allelic variation of the GPIIb and GPIIIa genes are depicted by white circles and the name of the HPA system is noted. Amino acids are given in three-letter acronyms. The fibrinogen-binding sites are in light grey and the Arg/Gly/Asp (RGD)binding site is in dark grey. The RGD peptide is the minimal fibrinogenderived peptide which binds GPIIb/IIIa.

platelets through pregnancy or transfusion they may produce anti-GPIV isoantibodies. These isoantibodies are of clinical significance as the survival of transfused normal donor platelets will be reduced or, in pregnancy, NAIT may ensue. Similarly, formation of isoantibodies might complicate the transfusion support of patients with congenital deficiencies of platelet GPs, such as GPIIb/IIIa (Glanzmann’s thrombasthenia), GPIb/IX/V (Bernard–Soulier syndrome) or GPVI. Many platelet membrane glycoproteins are the target of autoantibodies in autoimmune thrombocytopenia. Such autoantibodies will bind to the platelets of all individuals regardless of their HPA type. Platelet autoimmunity is frequently associated with B-cell malignancies and autoantibody formation is not infrequent in the period after haematopoietic stem cell transplantation during immune cell re-engraftment. In both situations, the presence of autoantibodies might contribute to the refractoriness to donor platelets. Some drugs too small to elicit an immune response may bind as a hapten to platelet GPs in vivo. In some patients, the haptenized platelet GP can trigger the formation of antibodies that only bind to the GP in the presence of the hapten. A classic example is quinine and its stereoisomer quinidine. Typically, quinine-dependent antibodies

Platelet and neutrophil antigens

are either anti-GPIIb/IIIa and/or anti-GPIb/IX/V. Similarly, the interaction of heparin with platelet factor 4 can cause antibody formation and lead to a drug-dependent thrombocytopenia. Besides the more classic examples of quinine and heparin, many other categories of drugs including several antibiotics have been associated with hapten-mediated platelet antibody formation. In haemato-oncology patients, who often receive a spectrum of drugs, the unravelling of the causes of persistent thrombocytopenia or poor response to platelet transfusions can be complex and sometimes frustrating. However, resolution of such a problem, for example the identification of drugdependent antibodies such as anti-vancomycin antibodies as the culprit of poor response to platelet transfusions, can be important in improving patient care. Detection of platelet alloantibodies

Over the last four decades, tests for the detection of platelet-specific antibodies have evolved from being non-specific and insensitive, e.g. the platelet agglutination test and chromium release assay, to more sensitive but still non-specific techniques such as the platelet immunofluorescence test (PIFT), which is unable to distinguish between platelet-specific and HLA class I antibodies, and to state-of-the-art GP capture assays such as the MAIPA assay. Despite the limitation described above, the PIFT remains the most widely used assay and, when results are analysed by a flow cytometer, is one of the most sensitive assays available. The principles of PIFT are shown in Plate 5.1 (shown in colour between pp. 304 and 305). However, it is the MAIPA assay that has become the modern cornerstone for the identification of platelet-specific antibodies. This assay captures specific GPs using platelet glycoprotein-specific monoclonal antibodies and can be used to analyse complex mixtures of antibodies in patient sera. The principle of this assay is shown in Fig. 5.1. Detailed knowledge of the molecular basis of platelet-specific alloantigens has enabled thirdgeneration antibody detection assays with purified or recombinant platelet GPs to be developed. However, there remains a disadvantage that the

sensitivity of these assays is not satisfactory for all alloantigen systems and some systems (e.g. HPA15) are not included. Clinical significance of HPA alloantibodies

HPA alloantibodies are responsible for the following clinical conditions: • NAIT (this condition is described in detail below; see also Chapter 8); • refractoriness to platelet transfusions (described in detail in Chapter 9); and • post-transfusion purpura (described in detail in Chapter 17). Neonatal alloimmune thrombocytopenia History The first case of NAIT was described by van Loghem in 1959. The existence of the platelet equivalent of haemolytic disease of the newborn had long been suspected, but its recognition was delayed because the detection of platelet antibodies was more technically demanding than that of red cell antibodies. NAIT is now a well-recognized clinical entity with an estimated incidence of severe thrombocytopenia due to maternal alloantibodies of 1 in 1000 to 1200 live births. Unlike haemolytic disease of the newborn, about 50% of cases of NAIT occur in first pregnancies. Definition NAIT is due to maternal HPA alloimmunization caused by fetomaternal incompatibility for a fetal human platelet alloantigen inherited from the father but which is absent in the mother. Fetal and neonatal thrombocytopenia may result from placental transfer of IgG antibodies against HPAs. Pathophysiology Maternal IgG alloantibodies against a fetal HPA alloantigen cross the placenta and bind to fetal platelets. Dependent on the quantity, affinity and subclass of the IgG antibodies and the density of the target antigen, platelet survival will be reduced. Severe thrombocytopenia and haemorrhage are generally caused by anti-HPA-1a. Anti-HPA-5b 55

Chapter 5

tends to cause mild thrombocytopenia, although intracranial cerebral haemorrhage (ICH) has been reported. NAIT due to alloantibodies against the other HPA alloantigens is infrequent. HLA class I antibodies, which are present in 15–25% of multiparous women, are not thought to be responsible for fetal thrombocytopenia as the majority of antibody is absorbed in the placenta, and if transferred to the fetal compartment will distribute over all HLA class I positive cells. Destruction of IgG-coated fetal platelets takes place in the spleen through interaction with mononuclear cells bearing receptors for the constant domain of IgG-Fc. Platelet antigens are known to be expressed from 16 weeks’ gestation, and placental transfer of IgG antibodies can occur from 14 weeks, so thrombocytopenia can occur very early in pregnancy. There have been reports of ICH before 20 weeks’ gestation. Incidence Prospective screening of pregnant Caucasian women for HPA-1a alloimmunization has shown that about 1 in 1100 neonates have severe thrombocytopenia (< 50 ¥ 109/L) because of anti-HPA1a. Anti-HPA-5b is the platelet alloantibody that most frequently occurs in pregnancy, but it causes clinically significant platelet destruction much less frequently than anti-HPA-1a. Relative immunogenicity HPA-1a and HPA-5b are the most immunogenic platelet alloantigens and are implicated in about 85% and 10%, respectively, of clinically diagnosed cases of NAIT. Alloantibodies in the other systems are less frequently responsible for NAIT. The HPA-15 system was described a decade ago, but its clinical relevance has only recently been demonstrated. HPA-15 has now been shown to be the third most commonly encountered alloantibody specificity and, more importantly, these antibodies were the only possible explanation for the thrombocytopenia in several cases of NAIT. Therefore, testing for HPA-15 alloantibodies should be included in the routine investigation of suspected NAIT cases. 56

The ability of an HPA-1a-negative mother to form anti-HPA-1a is controlled by the HLA DRB3*0101 allele. The chance of antibody formation in an HLA DRB3*0101-negative individual is small when compared with a DRB3*0101-positive individual, with an odds ratio of 140. Such an explicit linkage between an HLA class II type and the formation of alloantibodies has not been observed for any of the other platelet-specific alloantigen systems. Although the negative predictive value of the absence of HLA-DRB3*0101 for HPA-1a alloimmunization in HPA-1a-negative women is greater than 90%, its positive predictive value is only 35%, limiting its potential usefulness in an antenatal screening programme. About 10% of HPA-1anegative pregnant women develop anti-HPA-1a, and about 30% of these have an affected fetus/neonate with a platelet count less than 50 ¥ 109/L. Clinical features A typical case of NAIT presents with skin bleeding (purpura, petechiae and/or ecchymoses) or more serious haemorrhage, such as ICH, in a full-term and otherwise healthy newborn with a normal coagulation screen and isolated thrombocytopenia. There are less common presentations in utero, including ventriculomegaly, cerebral cysts and hydrocephalus. Although rare, hydrops fetalis has been reported in association with NAIT and this diagnosis should be considered if there are no other obvious reasons for the hydrops. Nearly 50% of severe ICHs occur in utero, usually between 30 and 35 weeks’ gestation, but sometimes even before 20 weeks. At the other end of the clinical spectrum, NAIT can be discovered incidentally when a blood count is performed for other reasons. Severe NAIT with a platelet count below 20 ¥ 109/L in a neonate is a serious condition. Rapid correction of the platelet count with HPA-compatible platelets is essential to prevent ICH and the possibility of lifelong disability. This management should be combined with laboratory investigations to confirm the clinical diagnosis, but platelet transfusion of compatible platelets should not be delayed while waiting for the laboratory results.

Platelet and neutrophil antigens

Differential diagnosis Other causes of neonatal thrombocytopenia are infection, prematurity, intrauterine growth retardation, maternal platelet autoimmunity and, rarely, inadequate megakaryocytopoiesis. Precise figures on the incidence of neonatal thrombocytopenia caused by viral infection are unavailable. Maternal platelet autoimmunity is rarely associated with severe thrombocytopenia in the neonate, but should be considered in women with a history of autoimmune thrombocytopenia. Platelet-type von Willebrand’s disease, in which mutations in the GPIba gene are associated with a propensity for in vitro platelet aggregation, can lead to falsely low platelet counts. Laboratory investigations Only antibodies against platelet-specific alloantigens are thought to cause immune thrombocytopenia in the fetus and neonate. For appropriate clinical management, the cause of severe thrombocytopenia in an otherwise healthy neonate should be determined with urgency. Detection of maternal platelet-specific antibodies is usually carried out by two techniques, the indirect PIFT and the MAIPA assay, using a panel of HPA-typed platelets. The most frequently detected antibodies are anti-HPA-1a, anti-HPA-5b and anti-HPA-15b. The serum of the mother is also tested against paternal platelets by both tests so that any alloantibodies against low-frequency alloantigens and private antigens can be detected. The parents are typed for the HPA-1, HPA-2, HPA-3, HPA-5 and HPA-15 alloantigens using PCR-SSP. In Caucasians, typing is usually confined to the aforementioned five systems as the immunogenicity of the other alloantigen systems is comparatively low. Neonatal management A neonatal platelet count below 20 ¥ 109/L should be corrected immediately, preferably with HPA1a- and HPA-5b-negative donor platelets, as these will be compatible with the maternal HPA alloantibody in over 95% of cases. The results of the laboratory investigations should not delay immediate platelet transfusion, as full investigation may be time-consuming and the risk of cerebral bleeds is highest in the first 48 h after delivery. In a typical

case, the platelet count should recover to normal within a week, although a more protracted recovery can occur. Antenatal management In a subsequent pregnancy of a mother with a previously affected pregnancy with NAIT, a decision needs to be taken on the approach to patient management in collaboration with a fetal medicine unit. Treatment during pregnancy is indicated for those cases in which the estimated risk of severe fetal/neonatal thrombocytopenia is considerable and this is based on the history of haemorrhage and thrombocytopenia in previous pregnancies. The possible treatments during pregnancy are: • intrauterine intravascular transfusion of compatible platelets combined with fetal blood sampling (FBS) at weekly intervals or just before delivery; • intravenous immunoglobulin (IVIG) and/or corticosteroids administered to the mother; or • a combination of the above. As few trials have been performed in this field, it is difficult to make definite recommendations. Weekly platelet transfusions by FBS have been shown to prevent ICH even in high-risk cases (Fig. 5.4), but are technically demanding, invasive and associated with a high mortality due to haemorrhage, cardiac dysrhythmias and premature labour. The high incidence of complications associated with FBS and the effectiveness of maternal therapy in about 67% of cases suggest that treatment with IVIG (1 g/kg per week) without an initial FBS should be used as initial management. This should be considered from or before 16 weeks’ gestation in those cases where there is a history of antenatal ICH in previous pregnancies, because the earliest reports of ICH are at 16 weeks. IVIG could be started a few weeks later (20–22 weeks) for those fetuses with a sibling history of severe thrombocytopenia but no antenatal ICH. FBS may be used to monitor the effect of maternal therapy (4–8 weeks after starting IVIG) and to indicate if alternative therapy is required. Failure of maternal therapy may potentially be rescued by the addition of oral prednisolone (0.5 mg/kg daily) and/or an increased dose of IVIG (2 g/kg per 57

Chapter 5

10 000

Platelets x 109/L

1 000 300 100 Platelet transfusion 30 10

1 25

26

27

28 29 30 31 Weeks gestation

32

33

5 12 19 Days postnatal

CS Fig. 5.4 Seventh pregnancy of a patient who has had five

miscarriages. The last of these was shown to have hydrops and hydrocephalus and a platelet count of only 17 ¥ 109/L, and the serological findings supported a diagnosis of neonatal alloimmune thrombocytopenia due to anti-HPA1a. The fetal platelet count was less than 10 ¥ 109/L at 25 weeks’ gestation in the sixth pregnancy, and a cord haematoma developed during fetal blood sampling (FBS) resulting in fetal death. In the seventh pregnancy, prednisolone 20 mg daily and intravenous IVIG 1 g/kg weekly were administered to the mother from 16 weeks

week), serial fetal platelet transfusions or early delivery at an acceptable gestation. Further studies are required to determine the optimal antenatal management for NAIT. The delivery also needs careful planning between obstetric and paediatric teams in close consultation with the consultant haematologist. Counselling Counselling of couples with an index case about the risks of severe fetal/neonatal thrombocytopenia in a subsequent pregnancy needs to be based on the severity of disease in the index case and the outcome of immunological investigations. The following should be taken into account: • thrombocytopenia in subsequent cases is as severe or, generally, more severe; • the best predictors of severe fetal thrombocytopenia in a future pregnancy are the occurrence of 58

until delivery. The figure shows pre-transfusion and posttransfusion platelet counts following serial FBS and platelet transfusions. The fetal platelet count was less than 10 ¥ 109/L at 26 weeks. The aim was to maintain the fetal platelet count above 30 ¥ 109/L by raising the immediate post-transfusion platelet count to above 300 ¥ 109/L after each transfusion. The fetal platelet count fell below 10 ¥ 109/L on one occasion when there were problems in preparing the fetal platelet concentrate and the dose of platelets was inadequate. CS, Caesarean section.

antenatal intracranial hemorrhage and severe thrombocytopenia (platelet count < 20 ¥ 109/L) in a previous pregnancy; • antibody specificity and titre do not reliably correlate with the severity of NAIT, and are currently of no value in the management of individual cases; • the zygosity of the partner. HPA-typed donor panels Establishing donor panels for fetal and neonatal platelet transfusion requires a major commitment from blood services. The frequency of HPA-1anegative donors in Caucasians is 2.5%. Therefore, identification of suitable donors requires simple, affordable and high-throughput typing techniques, as available for red cell phenotyping. Recently such techniques have been developed for HPA-1a phenotyping. Besides donors being

Platelet and neutrophil antigens

negative for the mandatory microbiological tests, they should also test negative for antibodies against red cells, platelets and leucocytes, and be cytomegalovirus seronegative. It has been calculated that in order to recruit one HPA-1a-negative donor who is able to meet all of the above criteria, approximately 1500–2000 donors will have to be phenotyped for HPA-1a. In addition, therapeutic platelets should be RhD matched, as small amounts of red cells present in platelet concentrates may immunize RhD-negative recipients, and negative for high-titre anti-A and anti-B antibodies. In order to recruit a single O RhD-negative HPA-1a-negative donor whose platelets will be suitable for a first fetal or neonatal platelet transfusion, where the fetal/neonatal blood group is unknown, approximately 15 000–20 000 donors need to be phenotyped for HPA-1a.

antigens, e.g. HNA-1a, HNA-1b, HNA-1c polymorphisms on CD16. The current nomenclature for the HNA systems includes polymorphisms that are both cell specific and ‘shared’ (Table 5.3). HNA-1 system

The most immunogenic polymorphisms, the triallelic HNA-1 alloantigen system, are localized on neutrophil FcgRIIIb (CD16), which is one of the two low-affinity receptors (R) for the constant domain (Fc) of human IgG (g). Four amino acid changes with arginine/serine, asparagine/serine, aspartic acid/asparagine and valine/isoleucine substitutions at positions 36, 65, 82 and 106 respectively define the difference between HNA-1a and -1b, and a single amino acid substitution (alanine/asparagine) at position 78 defines the HNA-1c polymorphism (Fig. 5.5). The expression of HNA-1c is frequently associated with the presence of an additional FcgRIIIb gene and increased expression of FcgRIIIb. Two other FcgRIIIb-associated high-frequency alloantigens have been reported: the LAN antigen and SAR antigen. The FcgRIIIb ‘null’ phenotype is rare and is based on a double deletion of the FcgRIIIb gene and is in some cases associated with a deletion of the FcgRIIc gene. The deficiency for the most abundant FcgR on neutrophils can cause immune neutropenia in the newborn due to maternal isoantibodies against FcgRIIIb. PCR-SSP can be used to determine the HNA-1a,

Human neutrophil antigens The antigens on the surface of human neutrophils can, as with platelets, be divided into different categories. There are common antigens that have a wider distribution on other blood cells and tissues, e.g. I and P blood group systems and HLA class I. There are ‘shared’ antigens which have a limited distribution among other cell types, e.g. human neutrophil alloantigen (HNA)-4a and -5a polymorphisms associated with CD11/18. There are also a limited number of truly granulocyte-specific

FcgRIIIbHNA-1b

FcgRIIIbHNA-1a

Ser (36)

Arg (36) S Fig. 5.5 Representation of the amino

acid substitutions resulting in the HNA-1a, HNA-1b and HNA-1c forms of FcgRIIIb. The positions of the amino acid substitutions arising from allelic variation of the FcgRIIIb gene are depicted by black dots. Amino acids are given in three-letter acronyms. The intrachain disulphide bonds create two domains which are closely related to the C-terminal heavy-chain domains of IgG.

Asn (65) Ala (78) Asp (82)

FcgRIIIbHNA-1c Ser (36)

S

S

S Ser (65) Ala (78) Asn (82) Val (106)

S S

Ser (65) S

Asp (78) Ile (106)

S Ile (106)

Asn (82)

S

S

S

S

Phosphatidylinositol Neutrophil surface membrane

59

Chapter 5 Table 5.3 Neutrophil-specific alloantigen systems.

System

Antigen

Original acronym for antigen

Caucasian phenotype frequency (%)

Nucleotide change

Amino acid change

HNA-1

HNA-1a

NA1

46

FcgRIIIb

G108 C114 A197 G247 G319

88

FcgRIIIb

5

FcgRIIIb

C108 T114 G197 A247 A319 A266 C266

Arginine36 None Asparagine65 Aspartic acid82 Valine106 Serine36 None Serine65 Asparagine82 Isoleucine106 Aspartic acid78 Alanine78

HNA-1b

NA2

HNA-1c

SH+

CD177 nk nk nk FcgRIIIb FcgRIIIb nk

nk nk nk nk nk nk nk

nk nk nk nk nk nk nk

70–95 kDa

nk

nk

CD11b CD11b

G302 A302

Arginine61 Histidine61

CD11a CD11a

G2466 C2466

Arginine766 Threonine766

Glycoprotein

SHHNA-2

HNA-3a HNA-4a

HNA-2a ND NE LAN SAR Five

NB1 NB2 ND1 NE1 LANa SARa 5a

Five

5b

Mart

Marta(+)

97 32 98.5 23 >99 >99 — — 99.1

Marta(-) HNA-5a

Ond

Onda(+) Onda(-)

>99

nk, not known.

-1b, -1c, -4a, -4b and FcgRIIIb null genotypes, and transfected cells expressing the FcgRIIIb HNA-1a, -1b and -1c allotypes have been used for alloantibody detection. HNA-2 alloantigen

HNA-2a, formerly known as NB1, is localized on a 58–64 kDa GP (CD177) expressed as a glycosylphosphatidylinositol-anchored membrane GP found both on neutrophil surface membranes and on secondary granules. The percentage of neutrophils expressing HNA-2a varies between indi60

viduals and HNA-2a alloantibodies typically give a bimodal fluorescence profile with granulocytes from HNA-2a-positive donors. The gene encoding the HNA-2a protein has recently been cloned and HNA-2a alloantigen status can be determined by phenotyping with polyclonal or monoclonal antibodies. The HNA-2a-negative phenotype is the result of a point mutation in the mRNA and there is no antithetical antigen to HNA-2a. An anti-NB2 antiserum was originally thought to define the antithetical antigen to HNA-2a (NB1) but this serum may instead recognize human monocyte antigen 1.

Platelet and neutrophil antigens

Alloantigens on CD11a and CD11b

The genes encoding the aM and aL subunits of the b2 integrin or CD11b and CD11a are polymorphic and are associated with HNA-4a and HNA-5a respectively. Alloantibody formation against these two polymorphisms has been observed in transfusion recipients, and recently a case of neonatal neutropenia due to anti-HNA-4a alloantibodies has been described. The low incidence of neonatal neutropenia associated with these antibodies is probably explained by the ubiquitous tissue distribution of these proteins. Detection of neutrophil antibodies

The detection of neutrophil antibodies is difficult. The main problems are the abundant expression of the two low-affinity receptors for human IgG or FcgR, which results in increased binding of circulating immunoglubulins in normal sera and the requirement for fresh and typed donor neutrophils as panel cells. The incidence of antibody-mediated neutropenias is comparatively rare and therefore the best strategy for investigation of clinical cases is in a centralized laboratory so that adequate technical expertise and the required reagents are available. Many techniques have been evaluated over the years for their suitability for neutrophil antibody detection. First-generation assays such as the granulocyte cytotoxicity and agglutination tests had a very low specificity. The neutrophil immunofluorescence and chemiluminescence tests have the advantage of good sensitivity but also detect HLA class I antibodies. For some human neutrophil alloantigen systems, e.g. antigens expressed on CD16, CD177 and C11/18, assays comparable with the MAIPA can be applied but otherwise immunoprecipitation of surface radioactivelabelled neutrophils remains the only reliable technique to determine the nature of the antigen. The principles of the granulocyte immunofluorescence test and the monoclonal antibody immobilization of granulocyte antigens (MAIGA) assay are analogous to the equivalent platelet tests described in Plate 5.1 and Fig. 5.1 respectively. Typing for neutrophil alloantigens is generally

based on the use of monospecific alloantisera derived from immunized patients. However, for some HNA, e.g. HNA-1a, HNA-1b and HNA-2a, murine monoclonal antibodies with allele-specific reactivity are available. Furthermore, for some of the neutrophil alloantigen systems (HNA-1a, -1b, -1c, -4a, -4b) the molecular basis has been determined and PCR-based genotyping can be used. Clinical significance of HNA antibodies

Neutrophil-specific antibodies are implicated in: • neonatal alloimmune neutropenia; • non-haemolytic febrile transfusion reactions (FNHTR) (also see Chapter 14); • transfusion-related acute lung injury (TRALI) (also see Chapter 14); • transfusion-related alloimmune neutropenia; • autoimmune neutropenia; • persistent post-bone marrow transplant neutropenia. Neonatal alloimmune neutropenia

Maternal alloimmunization against neutrophilspecific alloantigens on fetal/neonatal neutrophils is rare as a clinically significant entity. The incidence of the disorder is estimated at 0.1–0.2% of live births but there are no reliable prevalence figures. Clinical presentation is one of mainly bacterial infection with a selective neutropenia on a whole blood count. Severe but reversible neutropenia in the newborn may require treatment with antibiotics and/or granulocyte colony-stimulating factor (GCSF) to control bacterial infection and hasten development of a normal neutrophil count. Left untreated, the neutropenia in some cases has been reported to extend up to 28 weeks, presumably because the target antigen is restricted to a small compartment of antigen-positive cells. FNHTR and TRALI (see Chapter 14)

FNHTR are occasionally associated with the presence of leucocyte (HLA and neutrophilspecific) alloantibodies in the recipient. Serological investigations are of limited clinical value as the diagnostic specificity of tests for leucocyte 61

Chapter 5

antibodies is low. Moreover, the clinical management of FNHTR is to alter product specification if premedication with corticosteroids is not effective. In the UK, where there is universal leucocyte depletion of blood, investigations for other causes of fever associated with transfusion should be carried out, e.g. tests for bacterial contamination. Testing for neutrophil-specific antibodies might be required in the rare cases in which a severe FNHTR cannot be otherwise explained. TRALI is a severe and sometimes life-threatening transfusion reaction. The majority of cases are caused by donor leucocyte alloantibodies against alloantigens present on the patient’s leucocytes although patient alloantibodies might be involved in some cases. Investigations for TRALI are complex and need to include a screen for neutrophilspecific and HLA alloantibodies in samples from donors and the patient. Transfusion-related alloimmune neutropenia

The first case of transfusion-related alloimmune neutropenia was recently described following the infusion of plasma-reduced blood into a 4-week old infant. The blood contained approximately 28 mL of plasma containing granulocyte-specific HNA-1b antibodies and resulted in an absolute neutropenia in the infant, which was resolved after treatment with GCSF after 7 days. The case is of interest since it demonstrates that in some circumstances infused granulocyte-specific antibodies can trigger neutropenia rather than TRALI. Autoimmune neutropenia

Autoimmune neutropenia is a rare condition that can occur as a transient self-limiting autoimmunity in young children or in a chronic form in adults. The autoantibodies tend to target the FcgRIIIb (CD16), CD177 or CD11/18 molecules but can also be HNA specific, especially in children. The most sensitive method for the detection of autoantibodies is to test the patient’s neutrophils using the direct immunofluorescence test. However, the combination of severe neutropenia and the requirement for a fresh sample limits the applicability of this test. Screening of a patient’s serum 62

sample with a panel of typed neutrophils in the indirect immunofluorescence and chemiluminescence tests is a suitable alternative and, in some studies, has been found to be only slightly less sensitive than the direct test. Persistent post-bone marrow transplant neutropenia

Antibody mediated neutropenia may be a serious complication of bone marrow transplantation. In this context, the neutrophil antibodies may be antoimmune and/or alloimmune in nature and laboratory investigation requires serological and typing studies to elucidate the nature of the antibodies involved.

Further reading Birchall JE, Murphy MF, Kaplan C, Kroll H on behalf of the European FMAIT Study Group. European collaborative study for the antenatal management of fetomaternal alloimmune thrombocytopenia. Br J Haematol 2003; 122: 275–88. Bussel JB, Zabusky MR, Berkowitz RL, McFarland JG. Fetal alloimmune thrombocytopenia. N Engl J Med 1997; 337: 22–6. Bux J, Jung KD, Kauth T, Mueller-Eckhardt C. Serological and clinical aspects of granulocyte antibodies leading to alloimmune neonatal neutropenia. Transfus Med 1992; 2: 143–9. Bux J, Behrens G, Jaeger G, Welte K. Diagnosis and clinical course of autoimmune neutropenia in infancy: analysis of 240 cases. Blood 1998; 91: 181–6. Garner SF, Smethurst PA, Merienx, Y et al. A rapid one stage whole blood HPA-1a phenotyping assay using a recombinant monoclonal IgG1 anti-HPA-1a. Br J Haematol 2000; 108: 440–7. Griffin HM, Ouwehand WH. A human monoclonal antibody specific for the leucine-33 (P1A1, HPA-1a) form of platelet glycoprotein IIIa from a V gene phage display library. Blood 1995; 86: 4430–6. Lucas GF, Metcalfe P. Platelet and granulocyte polymorphisms. Transfus Med 2000; 10: 157–74. Maslanka K, Yassai M, Gorski J. Molecular identification of T cells that respond in a primary bulk culture to a peptide derived from a platelet glycoprotein implicated in neonatal alloimmune thrombocytopenia. J Clin Invest 1996; 98: 1802–8.

Platelet and neutrophil antigens Metcalfe P, Watkins NA, Ouwehand WH et al. Nomenclature of human platelet antigens. Vox Sang 2003; 85: 240–5. Murphy MF, Waters AH, Doughty HA et al. Antenatal management of fetal alloimmune thrombocytopenia: report of 15 affected pregnancies. Transfus Med 1994; 4: 281–92. Newman PJ, Valentin N. Human platelet alloantigens: recent findings, new perspectives. Thromb Haemost 1995; 74: 234–9. Shastri KA, Logue GL. Autoimmune neutropenia. Blood 1993; 81: 1984–95. Van Loghem JJ, Hanny Dorfmeijer Jr, van der Hart, Schreuder F. Serological and genetical studies on a platelet antigen (Zw). Vox Sang 1959; 4: 161–9. Von dem Borne AE, Ouwehand WH. Immunology of platelet disorders. Baillieres Clin Haematol 1989; 2: 749–81.

Von dem Borne AE, de Haas M, Roos D, Homburg CH, van der Schoot CE. Neutrophil antigens, from bench to bedside. Immunol Invest 1995; 24: 245–72. Warkentin TE, Smith JW. The alloimmune thrombocytopenic syndromes. Transfus Med Rev 1997; 11: 296–307. Watkins NA, Smethurst PA, Allen D, Smith GA, Ouwehand WH. Platelet alphaIIb/beta3 recombinant autoantibodies from the B-cell repertoire of a post-transfusion purpura patient. Br J Haematol 2002; 116: 677–85. Williamson LM, Hackett G, Rennie J et al. The natural history of fetomaternal alloimmunization to the plateletspecific antigen HPA-1a (PlA1, Zwa) as determined by antenatal screening. Blood 1998; 92: 2280–7.

63

Part 2

Clinical transfusion practice

Chapter 6

The effective and safe use of blood components Brian McClelland and Tim Walsh

Principles This chapter focuses on clinical decisions about the transfusion of blood components and emphasizes the opportunities for minimizing the need to transfuse and safe practice. The reasons for taking this approach will be evident from the chapters on adverse effects of transfusion. There is a high level of concern to avoid the well-publicized risks of transfusion. Although transfusion is very safe in the context of all the risks of hospital care, it is important that patients, and the wider public, understand that like all medical interventions it can never be entirely free of risks. Evidence for effectiveness

Clinical effectiveness, i.e. the measure of how much a treatment can help the patient, is a balance of benefits and risks. Many of the conventional and widely taught indications for transfusing blood components are not supported by reliable evidence of clinical benefit. It is important to take a critical approach to prescribing blood components. Guidelines for practice should, but do not always, reflect the best available evidence for clinical effectiveness. There are few good randomized, controlled, clinical trials of some of the main uses of transfusion. Some important recent trials challenge conventional teaching (see Chapter 36). Systematic review (usually with analysis of the pooled results) of clinical trials or other studies is an important way of examining the basis of current practice, and often reveals that conventional current practice is based on small studies that are not well designed (see Chapter 37).

The value of systematic review evidence as a guide to best practice depends on the quality of the original trials and also on whether they reflect current practice. Systematic reviews may clearly identify therapies that are effective, ineffective or harmful but often show that the available trials are simply inadequate to identify the clinical situations in which a blood component or product is beneficial or harmful. For example, Cochrane reviews of the effects of various intravenous replacement fluids on mortality in severely ill patients are summarized in Table 6.1. These led to considerable controversy, recently resolved by a large clinical trial in Australia that has established that albumin and saline solutions are effectively equivalent in terms of safety and effectiveness when used for resuscitation of critically ill patients. Information and consent

Patients must be informed about the benefits, risks and choices for their treatment whenever possible, although this obviously cannot be done if the patient is unconscious or too sick to communicate. For patients awaiting elective surgery, the preadmission clinic is an ideal opportunity to include information about transfusion as part of the information given to the patient about the whole process of care. Formal consent for transfusion is not required in the UK, but the prescriber has a professional duty to make sure the patient knows if transfusion is intended. The patient should be made aware of the risks of receiving blood (and of not receiving it) and of ways in which transfusion could be avoided, if this is an option. Whatever method is used to give this information, the patient’s notes should record that it has been 67

Chapter 6 Table 6.1 Clinical effectiveness of conventional transfusion

practices: the evidence may challenge conventional wisdom. Human albumin and artificial colloids for infusion in critically ill patients* Meta-analysis of clinical trials shows no evidence to suggest benefit from the use of albumin vs. use of crystalloid, or the use of artificial colloids vs. use of crystalloid. Controversy about the suggestion that the use of albumin may increase mortality has been resolved with the publication of the SAFE study Red cell transfusion† Study of 838 sick patients in intensive care unit. Randomized to transfusion to maintain Hb at 7–9 g/dL or 10–12 g/dL. Mortality at 30 days: 7–9 g/dL, 19% 10–12 g/dL, 23% * Cochrane Injuries Group Albumin Reviewers (1998); Schierhout & Roberts (1998); SAFE Study Investigators (2004). † Hébert et al. (1999).

given and that the patient’s questions have been answered. Suitable information leaflets are available on the websites of the various blood services.

Surgical and medical use of blood Recent studies in the UK show that about half of all red cell transfusions are given to surgical patients and about half for medical indications. The following sections outline some general principles that apply to all patients who may need transfusion and some that relate to specific categories of patients. Good blood management: avoiding the need for transfusion

The guiding principle is that allogeneic (donor) blood should be prescribed only when there are good grounds for believing that the benefits of transfusing outweigh the risks. Put another way, blood should only be prescribed when the clinician is satisfied that the risk of not transfusing is likely to be greater than the risk of transfusing. The decision is straightforward in some situations, for example when a patient has major haemorrhage associated with profound thrombocytopenia, or when a patient being treated for cancer has severe 68

disabling anaemia. However, many transfusions are given in situations where it is very difficult to estimate the probability that the patient will benefit. The challenge in clinical practice is to make this judgement in the many clinical situations where there is real uncertainty. An example is the decision whether or not to transfuse an elderly patient with a moderately low haemoglobin (Hb) level in the postoperative period. Retrospective studies of patients operated on for hip fracture reveal great variability in the use of perioperative red cell transfusion. There is at present no evidence whether such patients are likely to benefit if they are ‘transfused up’ or if they are allowed to remain moderately anaemic. Two pilot randomized controlled trials in hip fracture patients examined the effect of alternative transfusion protocols (transfuse at a ‘trigger’ level of 10 g/dL, transfuse only for symptoms). The use of the 10 g/dL trigger resulted in much more transfusion (Table 6.2). From a public health perspective, good blood management embraces wider issues such as anaemia prevention through programmes on nutrition, malaria, human immunodeficiency virus (HIV) infection, parasite infestation, and so on. Blood management in elective surgery

In elective surgery, transfusion cannot be considered in isolation as attention to the many individual elements of management can greatly influence the need to transfuse. This probably explains the low blood requirements of some surgical teams compared with others. Many studies have revealed large differences in the use of blood components for a given surgical procedure in apparently similar patient populations managed in different hospitals (Fig. 6.1). This can only partly be explained by obvious factors such as the patient’s age, gender and preoperative Hb status or by reported surgical variables such as blood loss or duration of operation. The use of specific blood conservation technologies can, for some patients, further reduce the need for allogeneic transfusion and may offer some other benefits (see Chapter 26). However, when these technologies are introduced they can be a

Effective use of blood components Table 6.2 Influence of changing the

‘transfusion trigger’ on blood use in hip fracture patients: results from two pilot clinical trials. Trial design: patients admitted for repair of hip fracture (mean age 85 years) were randomized to either ‘Symptomatic’ (transfuse if symptoms or if Hb < 8 g/dL and clinician wishes to transfuse) vs. ‘Trigger’ (transfuse to maintain Hb just > 10 g/dL)*.

Scotland †

USA

Number eligible Number consented Postoperative Hb < 10 g/dL > 8 g/dL

73 61 23

192 143 96

Randomized Arm of trial Completed protocol Transfused: n (%)

9 Trigger 9 9 (100)

9 Symptom 9 1 (11)

40 Trigger 37 39 (98)

40 Symptom 35 19 (45)

* Data from Carson et al. (1998) and Palmer et al. (1998). † Two patients randomized in this series are included among the 80 randomized patients in the USA study.

Fig. 6.1 Percentage (mean ± SE) of

operated patients perioperatively transfused with red cell units in each hospital, after adjustment for age, gender, preoperative haematocrit and blood loss, in total hip replacement and right and left hemicolectomy. Each hospital is identified by a country code followed by a letter, e.g. NL (Netherlands), hospital A (NLA). (From Sanguis Study Group 1994 with permission.)

Total hip replacement

Right and left hemicolectomy

100 80 60 40 20 0

DKA DKB GBC GBF GBH GBI NLA NLB Hospital

focus of attention for patients and professionals, with strong encouragement from the companies that develop and market them (Table 6.3). This should not distract attention from the need for very critical evaluation of their ability to improve some important outcomes. A recent systematic review of autologous transfusion techniques raises considerable doubt about their effectiveness. Transfusion protocols or local practice guidelines

A systematic review of the effects of clinical transfusion protocols indicates that the use of a protocol can reduce the use of blood. As an example, a protocol that specifies guideline Hb concentrations for transfusion and a plan for routine measurement of Hb and possibly haemostasis at relevant stages can reduce large and unexplainable varia-

BB

DB EA NLA NLB Hospital

tions in practice. Recorded Hb values and transfusion information provide objective data for clinical quality assurance. Blood management targets and interventions: elective surgery and emergencies

1 Optimize the Hb level before planned surgery. (a) Detect, identify cause and treat anaemia. (b) Erythropoietin with iron. 2 Optimize iron stores before surgery, even if patient not anaemic. 3 Optimize haemostasis before planned surgery. (a) Identify congenital coagulation disorders. (b) Withdraw drugs that impair haemostasis (if safe to do so). 4 Collect patient’s own blood before surgery. 5 Improve haemostasis during surgery. 69

Chapter 6 Table 6.3 In elective surgery, the use of autologous blood and/or drugs reduces bleeding and erythropoietin reduces the use of

allogeneic transfusion. (From Laupacis & Fergusson 1997 with permission.)

Intervention

Type of surgery

No. of trials

No. of patients

Exposure to allogeneic (donor) blood. Odds ratio (95% confidence interval)

Autologous techniques PAD ANH ANH

Miscellaneous Miscellaneous Miscellaneous, methodologically sound

6 16 8

933 615 NR

0.17 (0.08–0.32) 0.31 (0.15–0.62) 0.64 (0.31–1.31)

Cell salvage Washed Unwashed Unwashed

Orthopaedic Orthopaedic Cardiac

7 9 12

429 733 899

0.39 (0.30–0.51) 0.35 (0.26–0.46) 0.85 (0.79–0.92)

Drugs Aprotinin Desmopressin Tranexamic acid Aminocaproic acid Erythropoietin + PAD Erythropoietin + PAD Erythropoietin alone Erythropoietin alone

Cardiac Cardiac Cardiac Cardiac Orthopaedic Cardiac Orthopaedic Cardiac

45 12 12 3 11 5 3 2

5808 793 882 118 825 224 684 245

0.31 (0.25–0.39) 0.98 (0.64–1.50) 0.50 (0.34–0.76) 0.20 (0.004–1.12) 0.42 (0.28–0.62) 0.25 (0.08–0.82) 0.36 (0.24–0.56) 0.25 (0.06–1.04)

ANH, acute normovolaemic haemodilution; NR, not recorded; PAD, preoperative autologous blood donation.

(a) Anaesthetic techniques. (b) Surgical techniques. (c) Positioning. (d) Temperature control. 6 Collect and reinfuse blood lost during surgery: intraoperative blood salvage. 7 Collect and reinfuse blood after surgery: postoperative blood salvage. Autologous transfusion techniques are described in Chapter 26. Erythropoietin (EPO, epoetin, human recombinant erythropoietin)

This is a peptide hormone, normally made in the kidney (see Chapter 35). The therapeutic product is made by genetically engineered expression of the human erythropoietin gene. There are currently three products licensed in the UK, with some differences in the indications. Erythropoietin is a

70

potent stimulator of erythropoiesis, although some recent evidence suggests it may also have other physiological effects. Effective use in the surgical context may require parenteral iron repletion. Erythropoietin, like red cell transfusion, has its risks. If the Hb level is raised too rapidly, or too high, the risk of hypertension and thrombosis increase. Some patients receiving long-term erythropoietin treatment by the subcutaneous route have developed red cell aplasia. In at least some cases this appears to reflect an immunological response to the drug and is a reminder that even infrequent adverse events due to alternatives to conventional transfusion must not be ignored. Newer parenteral iron preparations, often used with erythropoietin to deliver the iron required for rapid erythropoiesis, still have an appreciable risk of moderate and severe acute reactions, especially if the manufacturer’s dose and administration instructions are not closely followed.

Effective use of blood components

Pharmacological agents used to reduce surgical bleeding

Effectiveness of blood management interventions: understanding the systematic reviews

These include aprotinin, which affects platelet function and fibrinolysis, tranexamic acid and e-aminocaproic acid (inhibitors of fibrinolysis), and desmopressin (1-deamino-8-D -arginine vasopressin, DDAVP), which acts centrally to increase plasma factor VIII levels. These agents are described in detail in Chapter 7.

There are two ways of expressing the size of the effect that an intervention has on an outcome (see also Chapter 37). If the intervention has no effect, the ratio between treatment and control groups is 1.0. In Table 6.3, the outcome of interest is the use of allogeneic red cells. If the intervention reduces the use of allogeneic red cells, the odds ratio or relative risk is less than 1.0. If the intervention were to increase transfusion, the ratios would be greater than 1.0. The 95% confidence intervals (CI) are one way of showing the statistical strength of the result. Take the first example in Table 6.3, i.e. preoperative autologous donation: the 95% CI is 0.08–0.32. This indicates that there is a strong effect on the outcome and that there is less than a 5% chance that the intervention is not effective. In contrast, in the third example in Table 6.3 (acute isovolaemic haemodilution) the upper 95% CI is greater than 1.0. It cannot be concluded in this case that the treatment is effective.

Recombinant factor VIIa

This product was developed to assist in managing haemophilia patients with inhibitors, for which it has proved effective. It has also been used to control massive surgical, traumatic or obstetric bleeding, although it is not licensed for these indications. Formal trials of its effectiveness are needed in these indications. Because there may be risks of thrombotic complications and because the drug is currently extremely expensive, hospitals have special procedures for making it available, e.g. consultation with a haematologist with an interest in haemostasis. Do these interventions reduce the need for allogeneic blood transfusion in surgical patients?

Clinical trials to answer this question have been subject to systematic reviews with meta-analysis. A number of recent papers are summarized in Table 6.3. These methods reduce the use of allogeneic transfusion but may have other consequences. For example, predeposit autologous transfusion usually increases the total amount of red cell units transfused when both autologous and allogeneic units are counted. It is also important to note that reductions in red cell transfusion are much more evident in studies that record a high blood use in the control group. In other words, some surgical teams manage their patients using transfusion infrequently and without recourse to any of the above bloodsparing methods.

A ‘total quality management’ approach for minimizing anaemia, bleeding and the need for transfusion

Individual surgical teams can manage their patients so as to minimize the need for transfusion without recourse to specific blood-sparing technologies. This may reflect the use of protocols to guide transfusion decisions and a commitment to attend to the many individual details of management that can reduce the need for blood replacement. A quality assurance system for management of anaemia and bleeding in surgery could be defined as an integrated approach that covers all aspects of management that influence the quality of care provided for the patient and ensures that they are consistently and correctly handled. Figure 6.2 illustrates aspects of care of a surgical patient that individually and collectively have an influence on the use of transfusion during an episode of care.

71

Chapter 6

Intraoperative

Preoperative Preoperative management

Non-emergency surgery

Postoperative

Trauma/ haemorrhage

Postoperative management

Assess patient for conditions that may reduce oxygen delivery, increase blood loss or reduce red cell production: anaemia iron deficiency blood coagulation problems dehydration malaria infections cardiorespiratory disease

Plan surgical and anaesthetic management to: Minimize blood loss anaesthetic technique vasoconstrictors tourniquets control of bleeding points support circulation posture

Assess and resuscitate: clear airway insert oropharyngeal airway or endotracheal tube give oxygen supplement check for chest injury, seal open wound if pneumothorax, insert drain with underwater seal control bleeding; local pressure at bleeding site

Monitor, control and replace fluid and red cell losses; optimize oxygen supply and temperature and analgesia

Diagnose anaemia

Fluid replacement

Give oxygen supplement

Treat anaemia

Salvage and reinfuse blood

Replenish iron stores

Replace red cell losses by transfusion if clinically required

Monitor: pulse, blood pressure, capillary refill conscious level Estimate blood/fluid loss

Stop drugs that impair blood clotting Correct fluid and electrolyte balance Recognize medical conditions: treat or other necessary action hepatitis, other virus infections bacterial infection cardiovascular disease respiratory disease others Decide if transfusion needed: autologous blood donor blood

Monitor blood losses: drains wounds Surgical re-exploration of continued bleeding

Maintain fluid and electrolyte balance Maintain temperature control

Set up intravenous line and infuse normal saline 3x estimated blood loss rapidly

Control pain Treat infection

Decide if transfusion needed: very urgent group O or patient's group less urgent: ensure grouped and compatible blood is available Reassess

Decide if transfusion is needed Treat postoperative anaemia: ferrous sulphate folic acid Treat other medical conditions

Operate to stop bleeding Replace red cell losses by transfusion if clinically required

Order blood, if clinically required Fig. 6.2 Framework for minimizing the need for transfusion in perioperative management.

Indications for the use of blood components Red cells and whole blood (Table 6.4) Acute anaemia and bleeding: clinical assessment and the decision to transfuse

The generally accepted reason for transfusing red cells is to increase the circulating red cell mass as a means of improving oxygen supply to the tissues. 72

On this basis, transfusion is only indicated if reduction in oxygen supply causes a clinically significant problem, and if it is likely that oxygen supply will be improved by a rapid increase in red cell mass. Current practice in surgery and critical care in developed countries, influenced by one large trial on intensive care unit (ICU) patients (see Table 6.1), is to adopt Hb concentrations of 7–8 g/dL as the threshold below which transfusion is indicated.

Effective use of blood components Table 6.4 Use of red cell components. (From Sanguis Study Group 1994 with permission.)

To increase circulating red cell mass with the intention to relieve clinical features caused by insufficient oxygen delivery In special situations such as sickle cell disease, thalassaemia (see Chapter 9) Examples of clinical quality assurance indicators: red cell prescription indicator Observed Patient records contain correct details of red cell transfusion Hb value recorded before transfusion Patient records contain a stated reason for transfusion ‘Avoidable’ red cell transfusions Percentage of red cell units given that result in discharge haematocrit > 33%

95% of records 90% of records 23% of records % of red cell transfusions 23% (lowest clinical unit), 82% (highest clinical unit)

Red cell transfusion during patient admission for specified procedure

Colectomy Transurethral resection of the prostate Coronary artery bypass graft AAA Total hip replacement

Pt/Po

u/Pt

0–79 (41) 0–46 (17) 17–100 (83) 64–100 (83) 29–100 (81)

0.3 (1) 0–1 (0.5) 0–6 (3) 0–6 (3) 0.5 (2)

Po, patients operated; Pt, patients transfused; Pt/Po, percentage of patients who are transfused; u/Pt, number of units received by transfused patients.

Recent studies raise some questions about both the above rationale for red cell transfusion and the adoption of this threshold Hb level. When resting subjects are intentionally haemodiluted to very low Hb concentrations, evidence of myocardial ischaemia is rare and only appears at Hb concentrations of 5–6 g/dL. When stable ICU patients were transfused with 2 units of red cells at an Hb level of 8–9 g/dL, there was no observable improvement in any of the ICU measures of systemic or regional oxygenation. These findings appear to suggest that even lower Hb thresholds could be appropriate for red cell transfusion, at least in some patient groups. On the other side of the debate, it is known that the performance of athletes improves objectively when Hb concentrations are raised to the high normal range by autologous transfusion or erythropoietin. In some clinical situations, the Hb concentration may be a poor indicator of the need to restore red cell mass by transfusion. After some types of major surgery, notably involving cardiopulmonary bypass, there may be a period when Hb concentration is lowered by the effects of haemodilution as well as blood loss and must be interpreted in the light of the patient’s fluid status. In practice, the

decision to transfuse is also influenced by known and estimated blood loss and by the clinical judgement of the risk of further bleeding. Chronic anaemia

Patients with chronic renal failure have improved clinical outcomes if Hb concentrations are maintained near normal. The same may be true of some patients with heart failure. Clinical experience, backed by at least one study, is that many chronically anaemic patients including those with malignant disease have an improved quality of life if Hb concentrations are kept nearer the normal range. Clinical trials have not been done to determine whether such patients benefit equally when the Hb concentration is maintained by erythropoietin treatment or by red cell transfusion. The available evidence does not indicate that the conservative transfusion guidelines currently recommended for surgical and critically ill patients are suitable for patients with chronic anaemia due to a serious underlying disorder. Current guidelines therefore advise that for surgical or critically ill patients with evidence of heart disease, transfusion at a Hb concentration around 10 g/dL is a 73

Chapter 6

reasonable compromise until better evidence is available. Clinical factors relevant to the decision to transfuse (Table 6.5)

One way to aid clinical decision-making is to summarize the factors that are most important in

Table 6.5 Factors in deciding whether a patient needs

transfusion. Blood loss External bleeding Internal bleeding: non-traumatic, e.g. peptic ulcer, varices, ectopic pregnancy, antepartum haemorrhage, ruptured uterus Internal bleeding: traumatic, e.g. chest, spleen, pelvis, femur(s) Red cell destruction, e.g. malaria, sepsis, HIV Haemolyis Malaria Sepsis HIV Cardiorespiratory state and tissue oxygenation Pulse rate Blood pressure Respiratory rate Capillary refill Peripheral pulses Temperature of extremities Dyspnoea Cardiac failure Angina Conscious level Urine output Haemoglobin estimate Clinical: tongue, palms, eyes, nails Laboratory: haemoglobin or haematocrit (PCV) Patient’s tolerance of blood loss and anaemia Age Other conditions, e.g. diabetes, pre-eclampsic toxaemia, renal failure, cardiorespiratory disease, chronic lung disease, acute infection, treatment with b-blockers Anticipated need for blood Is surgery or anaesthesia anticipated? Is bleeding continuing, stopped or likely to recur? Is haemolysis continuing? HIV, human immunodeficiency virus; PCV, packed cell volume.

74

making the decision into a simple checklist or algorithm that can be used (on paper or in the head) to help to focus the decision on whether or not to transfuse. Figure 6.3 is an illustration of this approach. Single-unit transfusion of red cells?

Dogmatic statements have often been made that there is no case for giving a single-unit transfusion. This dogma should be ignored. For example, in the case of a 45-kg patient with hypoxic signs or symptoms attributed to a Hb concentration of 7 g/dL, a single unit of red cells may be quite sufficient to relieve symptoms (and raise the Hb concentration by 1–2 g/dL). Choice of blood components Whole blood versus red cell concentrate (Table 6.6)

The doctrine of blood component therapy (together with the requirement for plasma for fractionation) has encouraged the almost universal use of red cell concentrates in most developed countries. In the UK, there is a case to review the relative merits of whole blood versus red cell concentrates, since UK plasma is not currently used for fractionation as a precautionary measure against the risk of transmitting variant Creutzfeldt–Jakob disease (vCJD) (although current thinking is that minimizing the plasma content of red cell products may offer some reduction of any vCJD risk) (see Chapter 20). Whole blood may be entirely appropriate for a patient with acute bleeding who requires both red cells and expansion of plasma volume. In cases when disseminated intravascular coagulation (DIC) contributes to blood loss, it may be entirely logical to use whole blood (or leucocyte-depleted whole blood) since it contains at least part of the total dose of fibrinogen and stable clotting factors that the patient requires and can reduce the need for plasma units from other donors. The suggestion that whole blood may be appropriate for some patients will be seen by some as highly controversial, although in many parts of the world it is widely used.

Effective use of blood components

Patient Name

Age

Gender

Hospital reference no.

Date of assessment

Time

No

Hb < 11 Yes

No

Signs and/or symptoms of inadequate O2 supply to tissues

Pale Breathless Tachycardia Other

Yes No

Comorbidity Malaria Sepsis Fever Haemolysis Leukaemia Ischaemic heart disease Other

Yes

No

Expected Delivery Bleeding Surgery Haemolysis Bone marrow failure Other Hb g/dL Sample date

Action (based on the information you have recorded above) Doses of red cell concentrate to raise Hb by 1g/dL: Adult: 1 unit (250 mL)/5O kg Infant/child: 3 mL/kg

Fig. 6.3 Example of a transfusion

Decision:

Transfuse

Intended result:

Clinical [Hb] raise to:

Review of result Clinical

Date:

No Yes

Units/mL

g/dL Time: [Hb]

decision chart.

Fresh or stored red cells for transfusion?

A widely quoted clinical study has suggested that transfusion of stored red cells could actually impair regional oxygenation. The clinical trial summarized in Table 6.1 suggested that some ICU patients maintained at a lower Hb concentration,

and so receiving less transfusion, may have improved outcomes. One interpretation of this observation is that this could be associated with some adverse effect of transfusing stored red cells. This could not be confirmed in a recent blinded, randomized, controlled study comparing the effect of fresh versus stored leucodepleted red cells 75

Chapter 6 Table 6.6 Whole blood or red cell concentrates?

Whole blood

Red cell concentrates

Contains plasma Replaces fibrinogen and other stable coagulation factors Volume-expanding effect of plasma may be an advantage in a hypovolaemic patient Red cells and plasma from the same donor

Minimal plasma content Does not replace coagulation factors If colloid volume expansion is shown to be harmful, the absence of plasma could be a benefit If both red cells and plasma are required, the patient is exposed to more donors Less volume load per dose of red cells, so safer if the patient is normovolaemic Provides plasma (FFP is likely to be used inappropriately if available!)

Colloid volume expansion effect could cause volume overload Does not produce plasma Does not contain platelets or factor VIII In UK and some other countries supplied as leucocyte-depleted Little or no good clinical trial evidence to compare the effectiveness of whole blood vs. red cell concentrates Whole blood is intrinsically simpler and cheaper to prepare.This may be extremely important in countries with restricted health budgets FFP, fresh frozen plasma.

on systemic and regional oxygenation in ICU patients. At present there is no clear evidence to support the selection of fresh red cells for critically ill patients, although the ‘fresh versus stored’ question cannot yet be considered to be fully answered. Fresh frozen plasma (Table 6.7)

Worldwide, the largest avoidable risk to patients from transfusion is probably due to the transfusion of fresh frozen plasma (FFP) for unproven clinical indications. In any area where blood safety testing may be unreliable, transfusion of FFP can be an important source of transmission of these infections. A recent systematic review suggests that there are few well-supported indications for transfusing FFP and this is reflected in the recent UK clinical guideline. Plasma is just as likely as whole blood to transmit HIV or hepatitis. FFP should be used only to replace rare clotting factor deficiencies for which no virus-safe fractionated plasma product is available or when there is a multifactor deficiency due to severe bleeding and DIC. Other indications for FFP (see Chapter 11)

76

Table 6.7 Use of fresh frozen plasma (FFP) and cryoprecipitate.

Replacement of plasma coagulation factors if a suitable licensed virusinactivated product is not available Some special indications, e.g. thrombotic thrombocytopenia purpura: infusion of plasma or plasma exchange with FFP (see Chapter 11) Contraindications To replace circulatory fluid volume To raise plasma albumin level As an alternative to total parenteral nutrition Example of clinical quality assurance indicator: FFP prescription Observed*

Indicator

Pre-transfusion (%)

Post-transfusion (%)

Prothrombin ratio recorded Prothrombin ratio > 2

94 68

94 11

* 19 FFP transfusion episodes in 12 consecutive patients (unpublished local audit report).

are the management of thrombotic thrombocytopenic purpura (TTP) and haemolytic–uraemic syndrome, in which plasma infusion or plasma exchange with FFP is effective. A recent systematic

Effective use of blood components

review concludes that there is little sound evidence for other uses of FFP. Does FFP have to be used immediately after thawing? After thawing, the level of factor VIII falls rapidly. Factor V also falls, but levels of fibrinogen and the other haemostatic proteins are maintained. New UK guidelines permit the use of plasma that has been stored in the bloodbank for up to 24 h after thawing. This has the advantage that plasma can be released quickly when required for urgent management of massive bleeding. Minimizing vCJD risk in the UK: importation of plasma and pathogen-reduced plasma In the UK, because of concerns about the risks of transmitting vCJD (see Chapter 20), the Department of Health has recommended that plasma imported from countries not affected by bovine spongiform encephalopathy (BSE) be used for all patients born after 1 January 1996 for whom FFP is indicated. To safeguard against viral infections undetected by testing, this plasma is treated to further reduce the risk of infectivity and is termed ‘pathogen-reduced plasma’. This additional processing causes reduced levels of plasma procoagulants such as fibrinogen, possibly necessitating higher volumes of plasma to achieve a given effect. In one pathogen-reduced plasma product, the levels of natural anticoagulants such as protein S were reduced. This was associated with an increased risk of thrombosis in patients undergoing liver transplantation. A similar problem may have occurred in patients undergoing plasma exchange for TTP. Cryoprecipitate (see Table 6.7)

If virus-inactivated plasma fractions are available, cryoprecipitate is only indicated as a source of fibrinogen in the management of DIC, e.g. in obstetric haemorrhage. If no virus-inactivated plasma fraction is available, cryoprecipitate is used to replace factor VIII in haemophilia A and von Willebrand’s disease (see Chapter 11). Pathogenreduced cryoprecipitate from non-UK plasma

should be used for patients born after 1996 according to Department of Health guidance. Platelets (Table 6.8)

Use of platelet transfusions continues to grow annually. Audit against current guidelines may reveal possible ways of reducing prescribing within current guidelines. A brief summary of indications for platelet transfusion follows (see also Chapter 9). Prophylaxis of bleeding due to bone marrow failure with thrombocytopenia Recent studies indicate that the clinically stable patient is unlikely to benefit from prophylactic platelet transfusion if the platelet count is greater than 10 ¥ 109/L. A higher threshold for transfusion is appropriate with sepsis and other complications. Trials in progress are assessing the benefits of prophylactic platelets versus platelets given only for bleeding. Surgery in the thrombocytopenic patient UK guidelines provide recommendations to minimize the risks due to bleeding in critical surgical sites such as head, neck and spinal canal. Table 6.8 Use of platelets (platelet concentrate).

Treat bleeding due to thrombocytopenia, for example platelet count < 10 ¥ 109/L due to bone marrow failure platelet count < 50 ¥ 109/L prior to surgery in critical area (head and neck) or invasive procedure (see Chapter 9) in management of haemorrhage (‘massive transfusion’) during/after surgery on cardiopulmonary bypass, where ‘pump’ damages platelets (see Chapters 9 and 11) Example of clinical quality assurance indicator: platelet prescription Indicator (platelet count recorded) Observed (% of transfusion episodes)* Within guideline Pre-transfusion Post-transfusion Increased by at least 20 ¥ 109/L

70 89 85 70

* 1701 episodes of platelet transfusion in 138 patients (unpublished local audit report).

77

Chapter 6

Urgent and emergency transfusion Examples of clinical scenarios in which transfusion may be needed are summarized in Table 6.9, Table 6.9 Some clinical situations where component

transfusion may be needed.

Situation Emergencies Obstetric haemorrhage

Factors that may influence the decision to transfuse

Clinical assessment during resuscitation

Trauma Ruptured aortic aneurysm Upper gastrointestinal bleeding Elective surgery Cardiac Orthopaedic Solid tumours

Bone marrow failure Malignant Myelodysplastic Drugs or chemicals Infective ITU patient Postoperative recovery Sepsis syndrome Inherited haemoglobin disorders Infections Malaria HIV Immunological disorders Immune haemolysis Immune cytopenias Neonatal problems Haemolytic disease of the newborn Blood sampling HIV, human immunodeficiency virus.

78

Previously fit patient Elderly, cardiovascular disease Liver disease Other comorbidity Pre-existing anaemia Bleeding tendency Jehovah’s Witness Stable Fever Splenomegaly Platelet count Stage of treatment Oxygenation measurements Haemoglobin Platelet count

Age Pregnancy, delivery Surgery

and discussed in detail in other chapters. This section concludes by reviewing some of the issues that arise in clinical situations where there is acute major bleeding, or the expectation of it, so that blood is required very rapidly. There is little or no systematic evidence for the effectiveness of the measures suggested and trials to evaluate them are required, although they will be extremely difficult to do. Clinical and blood bank experience is that delays in providing blood in a life-threatening emergency can put patients at risk, although there are few published reports that adequately document this. Consistent use of a simple, agreed terminology and for urgent contacts between, for example, the labour ward and the blood bank may save lives. Hospital Major Haemorrhage Protocol

The use of a Hospital Major Haemorrhage Protocol communicated effectively to all relevant staff is a sensible measure to improve the management of such cases. Occasional ‘fire drills’ to familarize staff and test the protocol have been found useful by staff in at least one large teaching hospital. Analysis by national haemovigilance schemes of ‘near miss’ incidents in supplying blood in emergencies would be a valuable way of sharing experience and building knowledge of effective strategies for providing blood safely in emergencies. Dealing with emergencies

A single patient with catastrophic bleeding is a major challenge for the clinical team. Road traffic accidents and other less common disasters may bring several desperately ill patients in quick succession, adding the risks that can result from uncertainties in patient identification. These are situations when it is vital for all the team to know and use the established local emergency procedures for: • requesting blood; • completing request forms and labelling samples; • communicating precisely with the transfusion laboratory; and

Effective use of blood components

• checking blood components before collection and transfusion. It is extremely important to have clear communication between clinicians and the blood bank, and essential to have a simple standard procedure, such as that described below. All staff (medical, nursing, laboratory and transport) play a very important part and should be trained in the use of this procedure, including an understanding of the importance of their role. Emergency transfusion procedures

Resuscitation and management of massive blood loss are dealt with in Chapters 7 and 11 and more information about hospital transfusion procedures can be found in Chapter 24. An example of a protocol for massive blood transfusion is shown in Table 6.10, and some key principles are described below. 1 In an emergency, insert an intravenous cannula, use it to take the blood sample for crossmatching, set up the intravenous infusion, and get the blood sample and blood request form to the blood bank as quickly as possible. 2 Make sure yourself that the blood bank staff know when the blood is required and why. 3 For each patient, the crossmatch sample tube and the blood request form must be clearly labelled. If the patient is unidentified, some form of emergency admission number should be used. Use the patient’s name only if you are sure you have correct information. 4 If another request for blood is needed for the same patient within a short period, use the same identifiers as on the first request form and blood sample so that blood bank staff will know it is the same patient. 5 If there are several staff working with emergency cases, one person should take charge of ordering blood and communicating with the blood bank about the incident. This is especially important if several injured patients are involved at the same time. 6 If there is a special stock of ‘emergency O negative’ blood, e.g. in the labour ward, use this first in an emergency. Do not wait for crossmatched blood if the patient is exsanguinating.

7 Tell the blood bank how quickly the blood is needed for each patient. Communicate using words that have been previously agreed with the blood bank to explain how urgently blood is needed. 8 Do not ask for crossmatched blood in an emergency. Ask the blood bank to supply what can be provided most quickly with reasonable safety according to the local policy. 9 Make sure that both you and the blood bank staff know: (a) who is going to bring the blood to the patient; and (b) where the patient will be, e.g. if your patient is about to be transferred to another part of the hospital for an X-ray, make sure the blood will be delivered to the X-ray room. 10 The blood bank may send group O RhD negative blood, especially if there is any risk of errors in patient identification. In an emergency, this may be the safest way to avoid a serious mismatched transfusion.

Avoiding the biggest risks of transfusion A full review of transfusion risks is given in Chapters 13–20. This section deals with risks that are common and which clinical staff can prevent by adhering to simple procedures. The first is the risk of patients receiving a blood component that was intended for someone else, and the second is that a special requirement, e.g. gamma-irradiated or cytomegalovirus (CMV)-seronegative blood, is not specified when blood is ordered. Following the simple rules shown in Fig. 6.4 and below will go a long way to preventing avoidable harm to patients. • When completing the blood request form, make sure the need for the transfusion and any special product, such as CMV-seronegative or gammairradiated blood components, is clearly stated. • Recording your decision to prescribe and administer a transfusion. There are two important reasons for writing in the patient’s notes the reason for giving a transfusion. First, it concentrates the mind to have to write and sign a permanent record of the basis of your clinical decision. Second, should the patient have a problem related to the 79

Table 6.10 Example of a hospital policy for massive blood loss. (Adapted from Stainsby et al. 2000 with permission.)

Immediate actions

Key points

Other considerations

Arrest bleeding

Early surgical or obstetric intervention Upper gastrointestinal tract procedures Interventional radiology

Contact key personnel

Most appropriate surgical team Duty anaesthetist Blood transfusion laboratory

Restore circulating volume NB In patients with major vessel or cardiac injury, it may be appropriate to restrict volume replacement after discussion with surgical team

Insert wide-bore peripheral cannulae Give adequate volumes of crystalloid/blood Aim to maintain normal blood pressure and urine output > 30 mL/h in adults (0.5 mL/kg per h)

Blood loss is often underestimated Refer to local guidelines for the resuscitation of trauma patients and for red cell transfusion Monitor CVP if haemodynamically unstable

Request laboratory investigations

FBC, PT,APTT, fibrinogen; blood bank sample, biochemical profile, blood gases Ensure correct sample identity and use of red label for transfusion samples Repeat FBC, PT, APTT, fibrinogen every 4 h, or after one-third blood volume replacement, or after infusion of FFP

Take samples at earliest opportunity as results may be affected by colloid infusion Misidentification is commonest transfusion risk May need to give FFP and platelets before the FBC and coagulation results available

Request suitable red cells NB All red cells are now leucocyte depleted. Volume is provided on each pack, and is in the range 190–420 mL

Blood needed immediately Use ‘emergency stock’ group O RhD negative Blood needed in 15–60 min Uncrossmatched ABO group specific will be provided when blood group known (15–60 min from receipt of sample in laboratory) Blood needed in 60 min or longer Fully crossmatched blood will be provided

Contact blood transfusion laboratory or on-call laboratory scientist and provide relevant details Collect sample for group and crossmatch before using emergency stock Emergency use of RhD positive blood is acceptable if patient is male or postmenopausal female Blood warmer indicated if large volumes are transfused rapidly Consider use of cell salvage

Consider use of platelets

Anticipate platelet count < 50 ¥ 109/L after 1.5–2¥ blood volume replacement Dose: 10 ml/kg body weight for a neonate or small child, otherwise one ‘adult therapeutic dose’ (one pack)

Target platelet count > 100 ¥ 109/L for multiple/ CNS trauma, > 50 ¥ 109/L for other situations May need to use platelets before laboratory results available: take FBC sample before platelets transfused

Consider use of FFP

Anticipate coagulation factor deficiency after blood loss of 1.5¥ blood volume Aim for PT/APTT < 1.5¥ mean control Allow for 30 min thawing time Dose: 12–15 mL/kg body weight (1 L or 4 units for an adult)

PT/APTT > 1.5¥ mean control correlates with increased surgical bleeding May need to use FFP before laboratory results available: take sample for PT, APTT, fibrinogen before FFP transfused

Consider use of cryoprecipitate

To replace fibrinogen and factor VIII Aim for fibrinogen > 1.0 g/L Allow for 30 min thawing time Dose: 1 pack/10 kg body weight

Fibrinogen < 0.5 g/L strongly associated with microvascular bleeding

Suspect DIC

Treat underlying cause if possible Mortality of DIC is high

Shock, hypothermia, acidosis: increased risk of DIC

APTT, activated partial thromboplastin time; CNS, central nervous system; CVP, central venous pressure; DIC, disseminated intravascular coagulation; FBC, full blood count; FFP, fresh frozen plasma; PT, prothrombin time.

80

Effective use of blood components

Step 1: Ask the patient to tell you their full name and date of birth 8E

9 01 G ALD 10 RA ON No. 56 MO ACD AL 7/19 M SPIT 1/0 HO B 1 DO

Step 2: Check these details against the patient’s wristband Step 3: Check the hospital ID number on the patient’s wristband against documentation e.g. patient case notes or request form

(a) Step 1: Complete the blood collection form (or follow the local collection procedure) with the following information: Any NHS Trust Blood and Blood Products Collection Form To be completd for each unit collected

• Forename • Surname • Date of birth • Hospital number

Patient Name: MORAG McDONALD….. DOB: ………….1/07/56……………………. Hospital Number:…100198E………… Addressograph label may be used Please tick Red cells: … ………. Fresh frozen Plasma: …. Platetes: …….….. Other ( specify: ………..

For each blood component collected

Step 2: Check patient ID details against compatibility label attached to the blood component

SUR NAM FOR ENA

ADD RES S

D.O .B.

G10

ABO

1 602

5 97

E

MA CD

E W IT H

22 9

ON AL

MORA

N

G101 602 597 229 N

D

G

1001

98E 25 HI RED CELLS IN AD D ITIVE SO LU TIO N , TOWNLL ST RE CE NT ET LEU CO CYTE D EPLETED RE STORE AT 4OC ± 2O C (SAGM) 11/07 /1956 WA RD

PO S

0 43 33

INS TRUC TION

ITU Alw ays check patient / component NUM BE R OF UNIT IS SU S ED

1

compatibility / identity. Inspect pack for signs of deterioration or damage. Risk of adverse reaction / infection.

Rh D PO SITIVE

CT 664/2

DAT E

WEST STO RE AT o GLA 4 C. UNIVER SGOW HOS PIT SIT Y ALS NHS TRU ST

O

Volume 263 ml

RHE SUS

O DER ESE RVE

PA TIBL

ME

HOS No. PIT AL

Do Not Use After: 03 DEC 2002 23:59

OF

2

CM V Ne ga ti ve

LOT

B1 08 0 21 0 62 0 + 6 B

LOT REF

REF

C00105107B

S NBTS

E TEGB 00 5

Step 3: Document removal of unit on blood fridge register or electronic release system

B

Maa cc o

P h a rm a

CO M

DON PAC O R No. K

Da te B led 29 Oct 2 002

(b) Fig. 6.4 Hospital procedures for safe transfusion. (a) Collection of blood sample for pretransfusion testing. Be extra vigilant

when checking the identity of the unconscious/compromised patient. (b) Procedure for collection of blood from the blood refrigerator. NB Follow procedure for each blood component collected.

81

Chapter 6 Step1: Check the blood component has been prescribed Step2: Undertake baseline observations Step3: Before approaching the patient check the component for:

M Ma

B

co

P h a rm aa

COM

DO NO P ACK R No .

SU RN

G 10

AM E

10 01 ESS

D.O .B. AB O

229

IT H G101 602 597 229 N

AG

98E

25 HI RED CELLS IN AD DITIVE SOLU TION , LL ST TO W N CE RE ET LEUCOCYTE D EPLETED NT RE STORE AT 4 OC ± 2O C (SAGM)

11 /07

/1956

RH ES

POS

WA RD

0 4 33 3

IN STR UCTION

IT U Always check patient / component NU MB O F UNIER I SS UE TS D

W EST ST OR E AT o GL 4 C. UN IVE AS GO W HOSP RSITY ITA NH S TR US LS T

O

Volume 263 ml

US

1

compatibility / identity. Inspect pack for signs of deterioration or damage. Risk of adverse reaction / infection.

Rh D POSITIVE

CT 664/ 2

O DE RE SERV E DA TE

Expiry date

Do Not Use After: 03 D EC 2002 23:59

OF

2

CMV Negative

B1 0 80 2 10 6 2 0+ 6B

LOT REF

REF

C00105107B

SNBTS

ETEG B005

LOT

(c)

Leaks Discoloration Clumping

N

DONA LD

M OR

HO SP No . ITA L

LE W

1 6 02 59 7

M AC

FO RE NA ME

AD DR

P A TIB

Date Bled 2 9 O ct 200 2

If there is any discrepancy, do not transfuse

transfusion, such as development of hepatitis in the future, the record of the clinical decision may prove to be important medicolegal evidence. A record that the patient has been given information about the transfusion (see above) may equally be important medicolegal evidence, as doctors have been criticized for alleged past failures to inform patients about the risks of blood products. Conclusions

It may be useful to think of a blood component transfusion as a tissue transplant to emphasize that prescribing is not a trivial decision. When you are uncertain if transfusion is likely to give clinical benefit to a patient, it may be helpful to ask yourself some critical questions. Before ordering blood in preparation for planned surgery

Can I reduce this patient’s need for transfusion by correcting anaemia, stopping warfarin or aspirin, checking for a coagulation disorder or arranging in advance for intraoperative cell salvage to be available in theatre?

82

Fig. 6.4 (c) Procedure for the administration

of blood.

Before prescribing blood or blood products for a patient

• What improvement in the patient’s clinical condition am I aiming to achieve? • Are there any other treatments I should give before making the decision to transfuse, such as intravenous replacement fluids or oxygen? • What are the specific clinical or laboratory indications of transfusion for this patient? • Do the benefits of transfusion outweigh the risks for this particular patient? • What other options are there if no blood is available in time? • Will a trained person monitor this patient and respond immediately if any acute transfusion reactions occur? • Have I recorded my decision and reasons for transfusion on the patient’s chart and the blood request form? • Finally, ask yourself ‘If this was myself or my child, would I agree to the transfusion?’

Effective use of blood components Step1: Ask the patient to tell you their full name and date of birth (where possible) Step2: Check ID details against the patient’s wristband and compatibility label Compatibility label COMPATIBLE WITH DONOR PACK No.

Surname

G101 602 597 229 N

SURNAME

MACDONALD

Forename

FORENAME

MORAG HOSPITAL No.

Hospital number

100198E

0 10 LD o. N 56 N A R D O A L /1 9 O 7 T M A C P I 1 /0 M OS 1 H OB D G

19

8E

A

ADDRESS

25 HILL STREET TOWN CENTRE D.O.B.

Date of birth

11/07/1956 ABO

Rh

WARD

O

POS

ITU NUMBER OF UNITS ISSUED

DERESERVE DATE

1 STORE AT 4oC. ANY NHS TRUST

OF

2

B

Ma c o

Ph a rm a

COMP A

SU RN AM E

G101

M OR

HOSP No. ITAL ADDR ES

229

S

ABO

MACDONALD

AG

FORENAME

MORAG STORE AT 4 OC ± 2OC

11 /07/ 1956 WARD

POS

0 4 3 33

INSTRUCTION ITU Always check patient / component

NUMB ER OF UN ISSUEITS D

1

compatibility / identity. Inspect pack for signs of deterioration or damage. Risk of adverse reaction / infection.

HOSPITAL No.

Blood group

Rh D POSITIVE

CT 664/2

WE ST ST ORE AT GL 4 oC. UNIVE AS GOW HO RSITY SP NHS TR ITALS UST

O

(SAGM)

Volume 263 ml

RHES US

E DATE

G101 602 597 229 N

G101 602 597 229 N

25 HI RED CELLS IN ADDITIVESOLUTION, LL ST TOW N CE REET LEUCOCYTE DEPLETED NT RE

O

COMPATIBLE WITH DONOR PACK No.

SURNAME

N

LD

10 01 98 E

D.O.B.

DE RE SE RV

602 59 7

M AC DONA

FORE NAME

Donor component number

TIB LE WITH

Do Not Use After: 03 DEC 2002 23:59

25 HILL STREET TOWN CENTRE D.O.B.

11/07/1956

OF

2

CMV Negative

B1080210620+ 6B

LO T REF

REF

C0 0105 107B

SNBTS

ETEGB005

LOT

RhD group

100198E

ADDRESS

ABO

O

DONO PACK R No.

Rh

POS

Date Bled 29 Oct 2002

DERESERVE DATE

WARD

ITU NUMBER OF UNITS ISSUED

1 STORE AT 4o C. ANY NHS TRUST

If there is any discrepancy, do not transfuse

OF

2

(d)

Fig. 6.4 (d) Identification checks at the bedside before transfusion. Be extra vigilant when checking the identity of the

unconscious/compromised patient.

83

Chapter 6 Example compatibility report form SCOTTISH NATIONAL BLOOD TRANSFUSION SERVICE CLINICAL SERVICES

EXAMPLE OF A TRANSFUSION COMPATIBILITY REPORT

COMPATIBLE WITH DONOR PACK No.

G101 602 597 229 N

SURNAME

MACDONALD

FORENAME

MORAG HOSPITAL No.

100198E

N SB TS /T xR ep Fm 98( 1 )

SURNAME

Ensure that you sign the compatibility report form to say you have checked the blood component against the patient’s wristband

ADDRESS

DOB

O

WARD

POS

ITU NUMBER OF UNITS ISSUED

1 STORE AT 4oC. ANY NHS TRUST

COMMENTS:

O Positive

ANTIBODY SCREEN

DONATION No . /

21/05/03 11:00

Issue No.

REQUESTED FOR GRO UP

COMPO NENT

PO OL No.

EXP IRY DATE /

G101 602 597 229 N

O POS

G101 603 703 984N

O POS

1803905

Lab Ref . COMPONENTS

Forrest

OAS

OAS

TIME

19/06/03

I.D. CHECK (S ig n at ur e)

R GNurs e

000082898

I .D. CHECK

DATE

TIME

2 1/05 /0 3

1 1. 10 am

(Sig natur e)

J Doctor

21/06/03

3

Rh

DERESERVE DATE

BLOOD GROUP.

11/07/1956

CONSULTANT

2

11/07/1956 ABO

MORAG

HOSPITAL No. 100198E

4C HOSP. Any Hospital WARD

1

25 HILL STREET TOWN CENTRE D.O.B.

FIRST NAME(S)

MACDONALD

OF

2

4

Complete documentation, e.g. ensure donor component number is recorded on compatibility report form

5

6

7

8

TRANSFUSION REPORT FORM SAMPLE COLLECTED : 20/05 /03

PLEASE FILE IN CASE NOTES AFTER COMPLETION OF TRANSFUSION RECEIVED : 20/05/03

(e)

REPORTED : 20/05/031

©

Fig. 6.4 (e) Documenting the transfusion on the compatibility report form.

Further reading Website www.transfusionguidelines.org.uk. Public site giving access to UK Handbook of Transfusion Medicine, UK standards for blood components, links to BCSH Guidelines, Cochrane library, etc.

Systematic reviews Hill SR, Carless PA, Henry DA et al. Transfusion thresholds and other strategies for guiding allogeneic red blood cell transfusion. Cochrane Database of Systematic Reviews 2003; 3. Henry DA, Carless PA, Moxey AJ et al. Pre-operative autologous donation for minimising perioperative allogeneic blood transfusion. Cochrane Database of Systematic Reviews 2003; 3. Henry DA, Moxey AJ, Carless PA et al. Desmopressin for minimising perioperative allogeneic blood transfusion. Cochrane Database of Systematic Reviews 2003; 3. Henry DA, Moxey AJ, Carless PA et al. Anti-fibrinolytic use for minimising perioperative allogeneic blood transfusion. Cochrane Database of Systematic Reviews 2003; 3.

84

Laupacis A, Fergusson D, Henry DA et al. Cell salvage for minimising perioperative allogeneic blood transfusion. Cochrane Database of Systematic Reviews 2003; 3. EBM reviews. Perioperative recombinant human erythropoietin in hip replacement. ACP Journal Club 1993; 119: 46. NHS Centre for Reviews and Dissemination. A metaanalysis of the effectiveness of cell salvage to minimize perioperative allogeneic blood transfusion in cardiac and orthopedic surgery (provisional record). Database of Abstracts of Reviews of Effectiveness 2003, issue 3. NHS Centre for Reviews and Dissemination. The efficacy of technologies to minimise peri-operative allogeneic transfusion (structured abstract). Database of Abstracts of Reviews of Effectiveness 2003, issue 3.

Original papers and reviews Beris P. The use of iron to increase red cell mass. Can J Anaesth 2003; 50 (Suppl.): S3–S9. Carson JL, Duff A, Berlin JA et al. Perioperative blood transfusion and postoperative mortality. J Am Med Assoc 1998; 279: 199–205. Carson JL, Terrin ML, Barton FB et al. A pilot randomized trial comparing symptomatic vs. haemoglobin-level-

Effective use of blood components driven red blood cell transfusions following hip fracture. Transfusion 1998; 38: 522–9. Carson JL, Terrin ML, Magaziner J. Anaemia and post operative rehabilitation. Can J Anaesth 2003; 50 (Suppl.): S3–S9. Cochrane Injuries Group Albumin Reviewers. Human albumin administration in critically ill patients: systematic review of randomised controlled trials. Br Med J 1998; 317: 235–40. Hébert PC, Schweitzer I, Calder L, Blajchman M, Giulivi A. Review of the clinical practice literature on allogeneic red blood cell transfusion. Can Med Assoc J 1996; 156 (11 Suppl. 1): S9–S26. Hébert PC, Wells G, Blajchman MA et al. A multicentre randomized controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999; 340: 409–17. Laupacis, Fergusson. Anesth Analg 1997; 85: 1258–67. McClelland DBL. Albumin: don’t confuse use with the facts (editorial). Br Med J 1998; 317: 829–30. McLellan SA, Walsh TS, McClelland DBL. Should we demand fresh red blood cells for perioperative and critically ill patients (editorial)? Br J Anaesth 2002; 89: 537–40. Palmer JB, Maciver CR, Scott R et al. Hip fracture and

transfusion trial (HATT) (abstract). Transfus Med 1998; 8 (Suppl. 1): 36. SAFE Study Investigators. A comparison of albumin and saline for fluid resuscitation in the intensive care unit. N Engl J Med 2004; 317: 235–40. Sanguis Study Group. Use of blood products for elective surgery in 43 European hospitals. Transfus Med 1994; 4: 251–68. Schierhout G, Roberts I. Fluid resuscitation with colloid or crystalloid solutions in critically ill patients: a systematic review of randomised trials. Br Med J 1998; 31: 961–4. Serious Hazards of Transfusion Scheme (SHOT). SHOT Annual Report. SHOT Office, 1998. Stainsby D, MacLennan S, Hamilton PJ. Management of massive blood loss: a template guideline. Br J Anaesth 2000; 85: 487–91. Walsh TS, McClelland DBL. When should we transfuse critically ill and perioperative patients with known coronary artery disease (editorial)? Br J Anaesth 2003; 90: 719–22. Walsh TS, McArdle F, Maciver C, Maginnis M, McClelland DBL. Age of transfused red blood cells does not affect indices of oxygenation after transfusion to critically ill patients: randomised controlled trial. Crit Care Med 2004; 32: 364–71.

85

Chapter 7

Bleeding associated with trauma and surgery Beverley J. Hunt

In the past, attempts at perioperative blood conservation have been driven by fears of transfusiontransmitted infection, concerns about the clinical efficacy and cost. The current concerns about variant Creutzfeldt–Jakob disease (vCJD) have stimulated interest in transfusion triggers, blood conservation techniques and the use of pharmacological agents to obviate the need for transfusion (Table 7.1). The use of blood should be minimized, but not at the risk of increasing morbidity and mortality in the patient from non-availability of blood or under-transfusion. The use of nearpatient haemostatic assessment in the operating theatre is of increasing interest and can result in more appropriate use of blood components. Nearpatient haemostatic testing has developed to such an extent that it is a prerequisite in orthotopic liver transplantation anaesthesia. Its use is common in cardiac surgery, and is being explored in other surgical fields.

Causes of perioperative bleeding Excessive bleeding can be due to surgical causes (i.e. suture deficiency) and/or derangement of haemostasis. In practice, blood loss is highly variable between surgeons, and there is increasing interest in training surgeons in adequate surgical haemostasis. A few minutes of the surgeon’s attention to careful haemostatic control may well save a patient from returning to theatre for surgical reexploration to find a bleeding point. It may also prevent the development of haemostatic problems associated with continued bleeding, the use of blood components and increased risk of morbidity 86

and mortality associated with excessive perioperative bleeding. However, there is a subset of patients in whom generalized oozing in the surgical field cannot be attributed to demonstrable bleeding vessels. There are no adequate definitions of an ‘excessive bleeder’ and yet many surgeons complain when it occurs. Some relevant factors in this respect may include the following. • Perioperative haemostatic changes have not been extensively studied, although haemostatic activation is thought to occur as a result of the hyperadrenergic state induced by the stress of surgical stimulation. • Increased fibrinolytic activity occurs during and shortly after an operation. Increased levels of tissue plasminogen activator (t-PA) and plasminogen activator inhibitor (PAI)-1, with a net increase in free t-PA during and after an operation, have been noted. • Meninges, cerebral and prostatic tissue contain relatively high concentrations of tissue activators of plasminogen. Elimination of t-PA from the blood is mainly regulated by the liver, with a normal half-life of 3.5 min.

Identifying those at risk of excessive bleeding • A history of previous bleeding problems in patients and/or their family, together with drug therapy (see below for advice about withdrawing aspirin and warfarin) and a full blood count, are the usual prerequisites for a standard surgical procedure.

Bleeding in trauma and surgery Table 7.1 Methods used to reduce allogeneic blood use

perioperatively. Accepted methods Acceptance of lower levels of haemoglobin postoperatively Improved surgical haemostasis Use of pharmacological agents Autologous transfusion (see Chapter 26)

N

Electromagnetic transducer

S

Pin

Other methods* Mechanical methods of intraoperative red cell salvage (see Chapter 26) Acute normovolaemic haemodilution Controlled hypotension Preoperative erythropoietin

Cup 0.36 mL whole blood (clotted) 4°45'

* Little evidence as yet for efficacy and safety.

Fig. 7.1 Operation of the thromboelastograph. See text for

explanation.

a°

A30

30 min MA

20 mm

• The most important determinants of the haemorrhagic risk are the patient’s diagnosis, the procedure planned and whether previous surgery has been performed. • Prior to cardiac surgery or other major surgery in an unwell individual, especially one who is actively bleeding, prothrombin time (PT), activated partial thromboplastin time (APTT) and fibrinogen level are necessary. • No other assays have been shown to be of value (including preoperative bleeding time), unless an underlying congenital bleeding disorder is suspected. Non-surgical perioperative bleeding has been poorly understood and poorly managed. The reasons for this include the following. • There is failure to recognize the limitations of laboratory tests. • There are no quick and reliable laboratory assays to investigate some components of haemostasis that may be abnormal, e.g. fibrinolysis and platelet dysfunction. This has stimulated the use of near-patient assessment such as the thromboelastograph (TEG) and other near-patient haemostatic assessment devices. The principle of the TEG and its interpretation is as follows. A blood sample is placed in a rotating pot and a piston is suspended in this (Fig. 7.1). As small clots form between pot and piston, the movement of the pot is transmitted to the piston. The piston is connected by a wire to a trace. Thus, as the piston is oscillated due to the clot linking it

r k

Thrombosis

Fibrinolysis

Fig. 7.2 Thromboelastograph trace. See text for

explanation.

to the pot, the movement is translated into lateral movements on the trace. Individual parameters (Fig. 7.2) include the r time, the time taken for blood to clot (some equivalence to laboratory clotting times), the maximal amplitude (K), and the a angle, which indicates platelet–fibrin interactions. Fibrinolysis can be measured by the reduction in the amplitude of the trace as the clot is lysed, 60 min after maximal amplitude (MA). The use of the TEG as a guide to intraoperative transfusion has been shown to reduce postoperative transfusion requirements in high-risk surgical patients. 87

Chapter 7

Patients on antithrombotic medication Surgeons prefer their patients to have normal haemostasis at the time of operation. If patients are receiving aspirin, ideally this should be stopped at least 10 days prior to surgery to allow for new platelets to develop that are not affected by aspirin’s inhibition of cyclooxygenase, otherwise blood loss is greater. Ideally, other antiplatelet drugs such as clopidogrel and non-steroidal antiinflamamtory drugs should be stopped 2–3 days prior to surgery. Patients receiving aspirin at the time of surgery do have a greater blood loss. If patients are receiving warfarin prior to surgery, then the following can apply. 1 If the patient can stop warfarin without the need for any perioperative thromboprophylaxis, then warfarin should be stopped at least 3 days preoperatively. 2 If the patient requires perioperative thromboprophylaxis, then the patient can be switched to heparin for the perioperative period. Depending on the underlying prothrombotic state the patient can switch to either of the following. (a) Full-dose intravenous heparin to maintain an APTT ratio of 2–2.5. This is still the required management for patients with artificial heart valves. No low-molecular-weight heparin is licensed for this indication and thrombotic events have occurred in patients with mechanical valves receiving low-molecular-weight heparin. As the half-life of heparin is approximately 2 h, heparin can be stopped 2 h preoperatively and restarted as the patient’s wound is being sewn up. (b) The modern alternative for prothrombotic states other than artificial heart valves would be to switch to a low-molecular-weight heparin the day after stopping warfarin. The dose used will depend on the patient’s underlying prothrombotic state. A full treatment dose can be given in those with severe prothrombotic states, and thromboprophylaxis doses to those with lesser prothrombotic states. Alternatively, the latter patient group could be given subcutaneous unfractionated heparin, but this will require more frequent injections. If the patient is taught how to administer subcutaneous injections, this 88

regimen will allow the patient to prepare at home and be admitted to hospital the day before, reducing the pressure on surgical beds. 3 If a patient has antithrombin deficiency or other thrombophilic disorder, then advice from a specialist in thrombosis should be sought. As heparin acts through potentiating the effect of antithrombin, then patients with antithrombin deficiency may need antithrombin concentrates as well as heparin thromboprophylaxis.

When red cell transfusion should be given • Previously, anaesthetists have tended to transfuse patients when their haemoglobin concentration has fallen below 10 g/dL or when the haematocrit is less than 30%. However, there is evidence that much lower levels of haemoglobin can be tolerated without adverse effects. • Experimental evidence shows that in a healthy human the cardiac output does not increase until haemoglobin falls to less than 7 g/dL. In critically ill patients, a recent randomized trial indicated that a threshold as low as 7 g/dL is as safe as, and possibly superior to, a haemoglobin threshold of 10 g/dL, with the possible exception of those with severe cardiac disease. • Moderate haemodilution is tolerated well by young healthy individuals undergoing elective surgery, in whom there would be greater concerns about the potential long-term effects of transfusion. • Concern remains about moderate haemodilution in those with compromised respiratory or myocardial function. • The ‘transfusion trigger’ should be a composite of variables specific for an individual patient. • The large variability in use of blood transfusion between different units and surgeons in the UK needs to be addressed. This is one of the recommendations in the Department of Health’s Better Blood Transfusion initiative in the UK (see Further reading). Each hospital should have a hospital transfusion committee to develop protocols that include agreed triggers for transfusion of red cells and the management of massive blood loss (see below and Chapter 25).

Bleeding in trauma and surgery

Special situations Transfusion management of acute blood loss and massive blood loss

This covers situations where blood loss is profuse for a limited period, or situations where massive blood loss occurs. The latter is arbitrarily defined as the replacement of the patient’s total blood volume in less than 24 h. The usual presentation is as an emergency in accident and emergency, labour ward or operating theatre, but in some situations, such as liver transplantation, it can be predicted and thus sophisticated monitoring and protocols are possible. Management is aimed at prompt resuscitation (Table 7.2). It should be remembered that the most frequent cause of death in massive transfusion is inadequate replacement of circulating volume and red cells. Preventing tissue hypoxia by maintaining an adequate circulating volume of red cells is the most important part of resuscitation. It is imperative that the prime goal is the treatment of shock; achieving surgical control of bleeding and managing a consumptive coagulopathy can wait. Whole blood is unavailable in many countries including the UK and following the implementation of universal leucodepletion of blood components it would not contain platelets in any case. There is no advantage in using fresh whole blood, and the use

Table 7.2 Principles of management of massive blood loss.

Replace and maintain oxygen-carrying capacity by: Maintaining blood volume Optimizing PCV > 20% Maintain haemostasis by: Platelet count > 50 ¥ 109/L INR and APTT ratios < 1.5 Fibrinogen > 1.0 g/L Avoid metabolic disturbances Hypocalcaemia Hyperkalaemia Acid–base disturbances Hypothermia Treat the cause of the blood loss APTT, activated partial thromboplastin time; INR, international normalized ratio; PCV, packed cell volume.

of whole blood from ‘walk-in’ donors is now considered an unacceptable and dangerous practice due to the risks of transfusion-transmitted disease (see also Chapter 6). Practical management

• Obtain maximum venous access. One or two large-bore intravenous cannulae should be inserted and if possible a central line. • Should crystalloid or colloid be used as initial replacement therapy? This thorny argument has continued for many years and has been fuelled by the recent concerns about the use of albumin. If crystalloids are used, then larger volumes are required. • Prevent tissue hypoxia by giving adequate volume to achieve an acceptable systolic blood pressure. Initially this can be done using crystalloid or colloid until red cell transfusion is available. • Blood is required. Send a blood sample to the blood transfusion laboratory for ABO group and RhD group (can usually be performed within 5 min) and telephone the transfusion laboratory to indicate the need for blood. If possible, wait for ABO- and RhD-compatible blood. In emergency cases, use group O RhD-negative red cells until the patient’s ABO and RhD groups are known. Switch to blood of the same ABO and RhD groups as the patient as soon as possible to avoid inappropriate use of group O RhD-negative red cells as their supply may be limited. • Also send a baseline ethylenediamine tetraacetic acid (EDTA) sample for full blood count (FBC), and a citrate sample for coagulation screen and sample to biochemistry for urea and electrolytes. • Administration of blood: when a very fast rate of transfusion is required (>50 mL/kg per h in adults or >15 mL/kg per h in children), a blood warmer should be used. • Once the patient’s blood pressure has been restored, consider surgical control of the bleeding. • Haemostasis (see also Chapter 11). An early coagulation screen and platelet count or TEG will provide a guide to the use of blood components. It is important to understand that at least 1.5 blood 89

Chapter 7

volumes (i.e. 7–8 L in adults) must be transfused before the platelet count falls below 50 ¥ 109/L in an average healthy individual. There is often time to assess coagulation fully; transfusion of blood components should be given as necessary according to the results of the screening coagulation tests. • Aim to (i) keep the platelet count greater than 50 ¥ 109/L by administering platelet concentrates; and (ii) maintain PT and APTT ratios less than 1.5 times the control value by giving fresh frozen plasma (FFP). Fibrinogen in the form of cryoprecipitate can be given if fibrinogen levels are disproportionately low, in order to maintain fibrinogen concentrations greater than 1.0 g/L. Coagulation problems (see also Chapter 11)

These occur in patients with extensive bleeding because of: • loss of haemostatic factors; • consumption in clot formation; • dilution with blood components and volume expanders; • depletion of coagulation factors due to inadequate synthesis, although factor VIII deficiency is partially compensated for by increased synthesis and release as part of the stress response; • acidosis and hypothermia precipitate disseminated intravascular coagulation (DIC). DIC may occur but cannot be predicted. Massively transfused patients do not form a homogeneous group; delayed or inadequate treatment of shock is probably the common predisposing factor, while extensive tissue damage, particularly head injuries, and pre-existing hepatic and renal failure may contribute to a deterioration in haemostasis. Volume expanders may produce other haemostatic hazards, apart from dilution. Dextrans, and to a lesser extent hydroxylethyl starch, have a fibrinoplastic effect; they accelerate the action of thrombin in converting fibrinogen to fibrin, which makes clots more amenable to fibrinolysis. Both are adsorbed on platelet surfaces and on von Willebrand factor (vWF), causing decreased platelet function and an acquired von Willebrand syndrome. Gelatins produce few problems, al90

though they decrease plasma fibronectin activity, but this has little clinical significance. Consider the use of adjunct haemostatic agents. Persistent bleeding will stimulate fibrinolytic activity. Ideally D-dimers or a TEG trace could be used in this situation as a guide to fibrinolytic activity. The use of an antifibrinolytic agent such as aprotinin 500 000 kallikrein inhibitory units (KIU) intravenously or tranexamic acid 1 g intravenously may be beneficial. Other possible complications of blood transfusion

• Hypocalcaemia. Calcium gluconate (2 mL of 10% solution per unit of blood) should be given when ionized calcium or calcium concentration can be measured and shown to be low or there are clinical signs or electrocardiographic changes. • Hyperkalaemia may occur due to the high concentration (~ 40 mmol/L) in stored blood. This is usually only a problem in those with hepatic or renal disease. • Acid–base disturbances. Despite the presence of lactic acid in transfused blood, fluid resuscitation usually improves acidosis in shocked patients. Furthermore, transfused citrate produces an alkalosis once it is metabolized. • Avoid hypothermia. Warm the patient, and the blood if a fast rate of transfusion is required. Cardiopulmonary bypass

Cardiopulmonary bypass (CPB) has evolved into a reliable method for total body perfusion, maintaining an oxygenated blood supply during the time heart surgery is performed. This is normally achieved by draining venous blood under gravity from the right atrium into a reservoir, and then pumping blood through an oxygenation device back into the patient’s arterial system (Fig. 7.3). This procedure bypasses the heart and lungs and creates a bloodless surgical field. Haemorrhage is one of the most important complications of cardiac surgery since re-exploration for bleeding, which occurs after 2–6% of coronary artery bypass grafting procedures, has been associated with a case fatality rate as high as 22%. This is especially relevant now there are an increasing

Bleeding in trauma and surgery

Direction of blood flow Venous return from patient under gravity

Cannula to right atrium

Cardiotomy suction lines for return of blood from the open chest Venous reservoir and cardiotomy filters

Arterial line pressure gauge Oxygenator Cannula to aorta Heat exchanger

Fig. 7.3 Cardiopulmonary bypass

Centrifugal pump

Blood returned to patient's systemic circulation bypassing the lungs

circuit.

number of patients requiring reoperation and of patients who have received anticoagulant, antiplatelet or thrombolytic therapy prior to surgery or who undergo complex surgery such as combined heart and lung transplantation. Factors associated with the bleeding diathesis of CPB

• The extensive contact between blood and the synthetic surfaces of the circuit causes coagulation activation, which necessitates the use of heparin. Intravenous heparin is administered to the patient prior to CPB at a dose of 3 mg/kg, with repeated doses of heparin being given during CPB (approximately equivalent to heparin levels of 3 U/mL). During CPB the activated clotting time (ACT) is used to monitor anticoagulant therapy and is maintained above 350–400 s. • Thrombocytopenia and defects in platelet function are proportional to the duration of CPB, and probably related to the level of hypothermia. In addition, abnormalities of platelet function include a reduced response to aggregation stimuli

owing to discharge of a granules and loss of platelet membrane receptors such as glycoprotein (GP)Ib and GPIIb/IIIa. • Fibrinolytic activation measured by D-dimer levels has been shown to peak during CPB and decrease at the end of CPB. There is wide variation in fibrinolytic response to CPB. The increased fibrinolytic activation is mainly due to an increase in t-PA. • Haemodilution is also a consequence of extracorporeal circulation but the fall in haematocrit, platelet count and plasma proteins, including coagulation factors, is about 30% and usually not sufficient to cause bleeding. There is increasing use of minimally invasive or ‘beating’ heart surgery, which avoids the use of CPB. For example, using a piece of equipment known as an ‘Octopus’, which once placed on the heart will reduce movement in the area in which the surgeon wishes to operate, allows for surgery to proceed without stopping the heart beating. The haemostatic changes of these procedures are less severe than those associated with CPB. 91

Chapter 7

Practical management of bleeding after cardiac surgery

• Check blood loss from all the chest drains. If one is filling at a greater rate than others, then this suggests a surgical cause, so discuss with the surgeon. • Obtain a history of preoperative drugs. The use of aspirin and non-steroidal anti-inflammatory drugs is associated with increased blood loss. • If there is a steady blood loss in all the chest drains of more than 300 mL/h, arrange for further blood to be made available by calling the blood transfusion laboratory. An FBC and clotting screen could be requested. While awaiting the results of these laboratory assays, an ACT may be performed to check that heparin has been adequately reversed with protamine. If heparin has not been adequately reversed, then give the appropriate dose of protamine. • Usually after cardiac surgery there is a thrombocytopenia and platelet function defect, so the bleeding time will be prolonged and therefore of little value in differentiating the cause of bleeding. A TEG trace, if available, may help. • Discuss rationale for the use of platelet concentrates and if necessary order an adult therapeutic dose of platelets. While waiting for this to arrive, if the patient has not received an antifibrinolytic agent perioperatively consider using aprotinin 500 000 KIU. • If bleeding continues at a rate of more than 300 mL/h despite correction of any pre-existing haemostatic defects and adequate haemostatic therapy, i.e. platelet transfusion given and platelet count greater than 100 ¥ 109/L and PT and APTT ratios less than 1.5, with a fibrinogen level in the normal range (normally 2–4 g/L), then discuss reexploration of the wound with the surgeon. Hyperfibrinolysis

Bleeding may occur if there is excessive generation of plasmin secondary to the release of tissue and urokinase plasminogen activators. Plasmin is a non-specific proteolytic enzyme and will split peptides with arginyl-lysyl amino acid sequences. These include fibrinogen, factors V and VIII, and the first component of complement. 92

Some regions, especially prostatic and pelvic tissues, are rich in plasminogen activator, excessive liberation of which may occur during the following. • Pelvic and prostatic surgery, especially for carcinoma of the prostate. • Extensive surgery. • CPB. • Liver transplantation. Extremely high levels of tPA occur during the anhepatic phase of orthotopic liver transplantation, probably as a result of increased endothelial release and decreased hepatic clearance. • Iatrogenic fibrinolytic bleeding can occur through the use of exogenous fibrinolytic activators such as streptokinase or urokinase in the management of thrombosis. Useful investigations include the following. • A global test of fibrinolytic activity such as the TEG should ideally be available. • Levels of fibrin degradation products (D-dimers) are greatly increased. • PT, APTT and thrombin time are mildly prolonged due to fibrinogenolysis. • Factors V and VIII may be normal or moderately reduced, in contrast to findings in DIC. • Often it is difficult to exclude DIC, especially as the most useful fibrinolytic assays are complex, time-consuming and not widely available. Practical management

Treatment with an antifibrinolytic agent should be considered if primary fibrinolytic bleeding is suspected. Suggested management options are: • tranexamic acid up to 1 g slowly intravenously; and • aprotinin 500 000 KIU as an intravenous bolus.

Pharmacological agents to reduce bleeding These have been used in two ways, either to prevent excessive bleeding or to treat established bleeding. The majority have been used in patients having cardiac surgery with CPB. The agents used can be broadly classified into four groups: anti-

Bleeding in trauma and surgery

fibrinolytics, desmopressin, haemostatic sealants and recombinant factor VIIa.

Antifibrolytics Aprotinin

• High-dose aprotinin in cardiac surgery (2 000 000 KIU to the patient, 2 000 000 KIU into the CPB circuit and 50 000 KIU/h during CPB) reduces postoperative drainage loss by 81%, and total haemoglobin loss by 89%. It also has benefit in vascular surgery and liver transplantation. Shorter operating times were also seen. This may result from the striking reduction of oozing that is normally seen; the operative fields remain ‘bone dry’. • Aprotinin is a basic polypeptide extracted from bovine lung. It is able to inhibit certain serine proteases by binding to their active site. In low concentrations, aprotinin is a powerful inhibitor of plasmin: its molar potency in vitro is 100 and 1000 times that of tranexamic acid and e-aminocaproic acid. Its main mechanism of action is through an antiplasmin effect. In high doses, it also inhibits kallikrein. The currently licensed high-dose regimen was designed to achieve blood levels that inhibit kallikrein (about 200 KIU/mL). Kallikrein is formed during the activation of coagulation by CPB and has a central role in the activation of the inflammatory response. • Aprotinin has no effect on the fall in platelet count, but may have a minor effect on preserving platelet function by preserving platelet membrane receptors, possibly by inhibiting plasmin-mediated degradation. • Aprotinin, by inhibiting kallikrein, will prolong in vitro tests of the intrinsic system and the ACT, because kallikrein normally operates a positive feedback on the generation of factor XII. In order to allow for adequate levels of heparin, the ACT timer should be run greater than 750 s. The activator in the ACT has traditionally been celite. Recently, kaolin has been used instead in some ACT tubes, for it is less affected by aprotinin and thus ACTs can be monitored in the normal way. • Since aprotinin is a bovine protein and thus can provoke an immunological reaction, a test dose

should be given. Consideration should be given to the future need of this drug, e.g. if a patient requires repeat cardiac surgery but in the long term requires a cardiac transplant, the surgeon may wish to reserve the use of aprotinin for the transplant operation. • The risks of using a prothrombotic agent perioperatively have not been defined, especially those of increased postoperative thromboembolic disease and particularly in relation to graft patency after coronary artery bypass grafting. Until these risks are defined the use of aprotinin to prevent blood loss should be limited to its licensed indication, i.e. the prevention of blood loss in high-risk cardiac surgery. • Aprotinin can also be used in established fibrinolytic bleeding. An intravenous dose of 500 000 KIU is a good antiplasmin dose. Lysine analogues

The lysine analogues, e-aminocaproic acid and tranexamic acid, are competitive inhibitors of plasmin binding to fibrin. A continuous infusion of tranexamic acid perioperatively in open cardiac surgery reduces bleeding significantly by about one-third, although it is not as effective as aprotinin. The dose given is 10 mg/kg over 20 min preoperatively followed by 1 mg/kg for 10 h. Both can be given to treat established fibrinolysis. The recommended dose for tranexamic acid is up to 1 g by slow intravenous infusion. Desmopressin

Desmopressin acetate (DDAVP) is a synthetic vasopressin analogue that is relatively devoid of vasoconstrictor activity. It increases the plasma concentrations and activity of vWF, probably by inducing the release of vWF from Weibel–Palade bodies in the endothelium. Plasma levels of vWF increase from two to five times from the baseline within an hour. It also improves platelet function. It thus leads to shortening of the bleeding time in patients with von Willebrand’s disease, platelet function defects and uraemia. Trials of DDAVP 0.3 mg/kg given after CPB to reduce bleeding had variable results; overall it was 93

Chapter 7

not beneficial. It may well be useful in patients with platelet function defects preoperatively, but this is not proven as yet. However, DDAVP may yet have a place in reducing bleeding, for it may be useful in reducing blood loss in patients with functional platelet disorders, notably those patients who have received aspirin preoperatively. Adverse effects include flushing and an antidiuretic effect. Haemostatic sealants

Haemostatic sealants mimic the final part of the coagulation cascade. • Fibrin sealants provide a source of thrombin and fibrinogen that when mixed together in the presence of calcium form a clot. They can be administered by a ‘gun’ which produces mixing of the reagents. Fibrin glue is best suited to secure haemostasis in patches and suture lines, for in situations where there are high blood flow rates then it is in danger of being washed off before it has ‘set’. A systematic review has shown that they do reduce allogeneic blood use and reduce bleeding but generally the trials were small and of poor methodological quality. • The initial source of thrombin was of bovine origin, which led to the development of a bleeding diathesis postoperatively. This is due to the formation of antibodies to bovine thrombin, which cross-react with human factor V, leading to acquired factor V deficiency in the recipients. Currently human thrombin is used in most fibrin sealants. • Methods of preparing autologous fibrin glue have been developed. They are currently made by units with large transfusion centres and commercial companies also exist to manufacture components. • Other haemostatic sealants, such as Floseal (Baxter), use bovine thrombin and gelatin. These are reconstituted and mixed together. The mixture is applied to the tissue surface, the gelatin expands to physically restrict the flow of blood and then thrombin converts endogenous fibrinogen to fibrin. The structural integrity of the gelatin–fibrin matrix enables it to remain in place at the tissue surface. The clot is resorbed within 6–8 weeks, 94

consistent with the time frame of normal wound healing. Recombinant activated factor VII

Recombinant activated factor VIIa (rVIIa) was initially used to treat haemophiliacs with inhibitors. The mechanism of action of factor VIIa in this setting is not fully understood but it does bypass the need for factor VIII and IX by generating thrombin and thus fibrin via direct activation of factor X. It is unclear whether tissue factor is necessary for this. The use of rVIIa has been explored ‘off licence’ for uncontrolled bleeding in a number of clinical scenarios. A group in Israel first described it as being efficacious in trauma patients who continued to bleed despite conventional component therapy. Subsequently, it has been used to reverse over-anticoagulation with warfarin, uncontrolled bleeding in hepatic dysfunction, orthotopic liver transplantation, cardiac surgery and in patients with thrombocytopenia and platelet dysfunction. There is a paucity of double-blind randomized trials to assess efficacy and safety in these multiple potential applications, and a number of such trials are currently in progress. The major concern about safety is the theoretical risk of thrombosis: rVIIa will bind to exposed tissue factor and initiate local thrombosis. Thus in patients with atheroma, tissue factor in plaques may be exposed to blood at the time of plaque rupture, while in DIC tissue factor may be exposed on monocytes. Accumulating safety data suggest that the risk of thrombosis with rVIIa is low, but this may relate to its low use in those with atheromatous disease. A randomized double-blind trial explored the use of rVIIa in preventing perioperative bleeding in 36 patients undergoing retropubic prostatectomy randomized to receive a perioperative dose of 20 or 40 mg/kg rVIIa or placebo. Median blood loss was significantly reduced in those receiving 40 mg/kg and none of this group required transfusion. There were no adverse events in the study group. This study suggests that rVIIa requires further study to assess its efficacy and safety in preventing perioperative bleeding and reducing the use of allogeneic blood.

Bleeding in trauma and surgery

Organization of transfusion for patients with trauma and for major accidents (see also Chapter 6) In a major accident, large numbers of people may be injured within a short space of time. The rescue and management of patients requires a coordinated approach from the rescue services and the hospital designated for admission of the casualties. A ‘major accident procedure’ is a necessity within every hospital. It should be tested periodically by holding a ‘major accident exercise’. The following must be incorporated into the procedure. • The telephone numbers of those who ‘need to know’ should be held by the hospital switchboard and called. • Suspend the issue of blood for non-emergency cases. • Increase the stocks of blood components to a predefined level by arranging deliveries from the nearest transfusion centre and maintain the necessary stocks of blood and blood products throughout the emergency. The blood transfusion laboratory must have a telephone line independent of the hospital switchboard because the main hospital switchboard may be inundated with telephone calls. • The risk of errors in the identification of patients and blood samples can be high in an emergency. Special care must be taken in the identification of casualties, and in labelling blood samples. In accident and emergency every attempt to maintain good clinical practice should be made. The minimum patient identification details for the request form and blood sample are the hospital number of the patient (or accident and emergency or major incident number) and their gender. • A telephone call from accident and emergency to the blood transfusion laboratory to inform them of estimates of potential future blood requirements is essential. • To avoid errors, the practice of issuing blood in a major disaster should not be changed from the routine practice of providing blood for urgent requests. • When the recipient’s blood group is not known, group O RhD-negative blood should be given to

girls and women of reproductive age, unless there is life-threatening bleeding and group O RhDnegative blood is not available. Group O RhDpositive blood can be given to males with unknown blood groups. • Blood components such as FFP and platelet concentrates need to be available for those who are receiving massive transfusion. • Dealing with requests to donate blood: following a major accident, there may be a number of telephone calls from the public, offering to donate blood. These potential donors should be given the telephone number of the Blood Service National Call Centre so that they can attend one of the routine blood donor clinics.

Conclusions There have been some exciting advances in monitoring and managing bleeding in surgical and trauma patients in the last few years. Despite this, many more adequately powered, prospective studies are required to investigate the utility and safety of these approaches in different surgical settings.

Further reading Carless PA, Anthony DM, Henry DA. Systematic review of the use of fibrin sealant to minimize perioperative allogeneic blood transfusion. Br J Surg 2000; 89: 695–705. Carson JL, Poses RM, Spence RK, Bonavita G. Severity of anaemia and operative mortality and morbidity. Lancet 1996; 348: 1055–60. Clark RAF. Fibrin sealant in wound repair: a systematic survey of the literature. Exp Opin Invest Drugs 2000; 9: 2371–91. Cochrane Injuries Group Albumin Reviewers. Human albumin administration in critically ill patients: systematic review of randomised controlled trials. Br Med J 1998; 317: 235–40. [This article received an enormous critical response, and the reader is referred to one of the reviewers of the original article whose response is typical of the criticism: Soni N. Human albumin administration in critically ill patients. Validity of review methods must be assessed. Br Med J 1998; 317: 883–4.]

95

Chapter 7 Friederich PW, Henny CP Messelink EJ et al. The effect of recombinant activated factor VII on perioperative blood loss in patients undergoing retropubic prostatectomy: a double-blind placebo-controlled randomised trial. Lancet 2003; 361: 201–5. Ghorashian S, Hunt BJ. Off license use of recombinant activated factor VII. Blood Rev 2004; 18: 245–59. Hébert PC, Wells G, Blajchman MA et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999; 340: 409–17. Horrow JC, Hlavecek J, Strong MD et al. Prophylactic tranexamic acid decreases bleeding after cardiac operations. J Thorac Cardiovasc Surg 1990; 99: 70–4. Martinowicz U, Kenet G, Segal E et al. Recombinant activated factor VII for adjunctive haemorrhage control in adults. J Trauma 2001; 51: 431–9.

96

Oz MC, Delos M, Cosgrove DM et al. Controlled clinical trial of a novel hemostatic agent in cardiac surgery. Ann Thorac Surg 2000; 69: 137–82. Porte RJ, Leebeek. Pharmacological strategies to decrease transfusion requirement in patients undergoing surgery. Drugs 2002; 62: 2193–211. Sanguis Study Group. Use of blood products for elective surgery in 43 European hospitals. Transfus Med 1994; 4: 251–68. Segal H, Hunt B. Aprotinin: pharmacological reduction of perioperative bleeding. Lancet 2000; 355: 1289–90. Shore-Lesserson L, Manspeizer HE, DePerio M, Francis S, Vela-Cantos F, Ergin MA. Thromboelastography-guided transfusion algorithm reduces transfusion in complex cardiac surgery. Anesth Analg 1999; 88: 312–19.

Chapter 8

Prenatal and childhood transfusions Irene Roberts

Obstetrics One of the most important aspects of obstetric transfusion medicine is the prevention, recognition and treatment of haemolytic disease of the newborn (HDN), which causes at least 50 neonatal deaths per year in the UK. This is considered in detail in this chapter. Other topics covered include: • aspects of maternal platelet and white cell disorders relevant to transfusion; • maternal haemorrhagic disorders, including major obstetric haemorrhage; and • transfusion requirements during pregnancy of patients with major haemoglobinopathies. Antenatal red cell antibody testing

The three factors essential in the pathogenesis of HDN are: • maternal red cell alloantibodies that cross the placenta; • fetal red blood cells that express antigens against which the antibodies are directed; and • antibodies which are able to mediate red cell destruction. Clinically relevant alloantibodies are almost always IgG and are reactive at 37°C. Women develop these antibodies as a result of previous transfusions, previous pregnancies or both. Identification of such antibodies is the main goal of antenatal screening.

• Identify RhD-negative women who require antiD prophylaxis (around 16% women are RhD negative). • Ensure swift provision of compatible blood for obstetric emergencies. • Identify the fetus requiring treatment in utero or in the neonatal period. • Identify additional red cell alloantibodies. • Identify new red cell antibodies induced by intrauterine transfusion. Red cell serology at the booking visit

At the booking visit, which should take place before week 16 of pregnancy, all women should have their ABO and RhD group determined and should be screened for red cell alloantibodies. If red cell antibodies are detected at the booking visit and/or if there is a history of HDN, the antibodies should be identified, quantified and monitored as outlined below. It is particularly important to monitor women with anti-D, anti-c and anti-K, since these antibodies may be associated with severe HDN. If no red cell alloantibodies are detected at booking, all pregnant women should be retested at 28–36 weeks of gestation. Further testing of women without detectable antibodies is unnecessary since immunization later in pregnancy is unlikely to result in antibody levels sufficient to cause HDN requiring treatment. Partial/weak D

Objectives of red cell antibody testing in pregnancy

• Identify the pregnancy at risk of fetal or neonatal HDN.

Du (weak D) individuals are regarded as RhD positive and do not form immune anti-D. Individuals with known partial D status, e.g. DVI, are more likely to make anti-D; therefore it is important that 97

Chapter 8

reagents for RhD grouping do not detect DIV (so that these individuals group as RhD negative). ABO antibodies

There is no need to test for ABO immune antibodies in antenatal samples as their presence is not predictive of HDN and such antibodies very rarely cause significant haemolysis in utero. Samples at delivery

At the time of delivery a maternal and a cord blood sample should be collected from all pregnancies in RhD-negative women. Women with no previously detected anti-D should have prophylactic anti-D administered if the infant is RhD positive. A Kleihauer test should also be carried out on all such women to assess the requirement for additional anti-D. A direct antiglobulin test (DAT) should be performed on the cord blood: a positive DAT is a good predictor of HDN. It is important to note that in women who have had prophylactic anti-D during pregnancy the anti-D remains detectable in serum for up to 12 weeks and may cause confusing serological results in the mother and a positive DAT in the baby in the absence of HDN. In the case of women with other clinically significant red cell alloantibodies (see below), a DAT should be carried out on cord blood; if the DAT is positive, a red cell eluate may help identify the red cell antibody. Infants born to mothers with clinically significant antibodies should be monitored for 48–72 h for the presence of haemolysis. Clinically relevant red cell alloantibodies

The main antibodies implicated in HDN include the following. • Rh group: D, c, C, e, E, Ce and Cw. • Kell group: K1, K2 and Kpa. • Duffy group: Fya. • Kidd group: Jka. The antibodies most commonly implicated in severe to moderate HDN are anti-D, anti-c and anti-Kell. Anti-D is the commonest cause of HDN (approximately 40% of cases in the UK). This is because anti-D is highly immunogenic and the high 98

proportion of women who are RhD negative (16%). Most anti-D antibodies are IgG1 or IgG1 plus IgG3. The presence of IgG3 alone, which has 100 times the destructive ability of IgG1, is uncommon and rarely associated with HDN in utero, but can cause severe postnatal manifestations of HDN. Anti-c is found most commonly in women with the R1R1 genotype (CDe/CDe), which occurs in 20% of pregnant women. Such women also have the propensity to make anti-E. HDN due to anti-E is both less common and less severe. However, anti-E and anti-c in combination cause more severe HDN than either antibody alone. Note that in such cases only the anti-E is detectable in eluates from cord blood red cells. Anti-K1 is the most common red cell alloantibody outside the ABO and Rh system. K1 is the principal antigen of the Kell blood group system and is highly immunogenic; 5% of K1-negative individuals will produce anti-K1 if transfused with K1-positive blood. K1 has around twice the potency of c and E and 20 times the potency of Fya. Anti-K1 often causes severe HDN; the haemolytic anaemia is compounded by suppression of erythropoiesis due to anti-K1 inhibiting the growth of erythroid progenitor cells. Anti-K titres can be an unreliable predictor of the severity of HDN. Therefore it is important to identify the fetuses at risk of HDN by determining the fetal Kell genotype in all mothers with anti-K1 whose partners are heterozygous for K1 (since only 50% of such fetuses will be K1 positive). Moderate to severe HDN may also be caused by anti-K2 (anti-cellano) and anti-Kpa. A number of other red cell alloantibodies have also been reported to cause HDN of variable severity, e.g. anti-U. These initially present with a positive indirect antiglobulin test (IAT) in maternal serum; therefore all women with a positive IAT should have further investigation to try to identify any clinically relevant red cell alloantibodies. Red cell alloantibodies not implicated in HDN

These include the following: • anti-Lea and anti-Leb; • anti-Lua; • anti-P; • anti-N;

Prenatal and childhood transfusions

• anti-Xga; • anti-Gerbish. Management of pregnant women with red cell alloantibodies Anti-D

• Women with anti-D should have their anti-D titres monitored monthly until 28 weeks of gestation and then every 2 weeks. • All samples should be checked in parallel with the previous sample. • An increase in anti-D by 50% or more compared with the previous sample is a significant increase irrespective of gestation. It is important to note that these are guidelines. Titres of anti-D do not always correlate closely with the development of HDN. Therefore all women with affected pregnancies should be referred early to specialist fetal medicine units for fetal assessment by ultrasound, amniocentesis or fetal blood sampling as indicated (for management of the fetus and neonate, see pp. 107–9). Anti-c

• Women with anti-c should have their anti-c titres monitored monthly until 28 weeks of gestation and then every 2 weeks. • All samples should be checked in parallel with the previous sample. • An increase in anti-c by 50% or more compared with the previous sample is a significant increase irrespective of gestation. • Anti-c titres of greater than 10 IU/mL are associated with a moderate risk of HDN and may require intrauterine transfusion (IUT). • All women with anti-c should be referred to a specialist fetal medicine unit early in pregnancy. Anti-Kell

• Women with anti-K1 should have their anti-K titre monitored monthly until 28 weeks of gestation and then every 2 weeks. • Anti-K titres may not accurately reflect the degree of fetal anaemia.

• Fetal K typing by chorionic villous sampling (CVS), amniocentesis or fetal blood sampling should be performed where the father is heterozygous for K1. • Fetal growth, fetal anaemia and the presence of hydrops should be monitored by serial ultrasound and Doppler and anaemia confirmed by fetal blood sampling as indicated. • Amniocentesis is not a good indicator of the severity of fetal anaemia since anaemia due to antiK results from a combination of haemolysis and red cell hypoplasia. • All women with anti-K should be referred to a specialist fetal medicine unit early in pregnancy. Other red cell alloantibodies

• If the antibody is likely to cause problems with provision of blood to cover an obstetric emergency, it is important to inform the obstetrician in charge of the case and the transfusion laboratory in the hospital and efforts should be made to ensure that appropriate blood products can be supplied. • Any babies born to mothers with an IATreacting antibody must be assessed at birth for evidence of HDN. Blood transfusion support for mother and baby Mother

• Red cell components of the same ABO and RhD group must be selected. • Group O blood may be used, provided it is plasma depleted and does not contain high-titre agglutinins. • Note that in pregnancy, immunization following a transfusion is most likely to occur in the third trimester. • Samples used for pretransfusion testing should ideally be taken immediately before transfusion and must never be more than 7–10 days old (Table 8.1). Fetus and neonate: crossmatching and general considerations

Management of the fetus and neonate at risk of 99

Chapter 8 Table 8.1 Pretransfusion testing of maternal samples.

Timing of last transfusion

Timing of pretransfusion sample

3–14 days before 14–28 days before 28 days to 3 months

24 h before transfusion 72 h before transfusion 1 week before transfusion

HDN is discussed in detail below. The general principles of the blood products used are summarized here. 1 Prior to the first transfusion, samples should be obtained from the mother for ABO, RhD grouping and antibody screening and from the fetus/neonate for ABO, RhD and DAT (plus an antibody screen if no maternal sample is available). 2 In the fetus/neonate the ABO group is determined on the cells only (as reverse grouping can detect passive maternal antibodies). 3 Red cells which are ABO compatible with maternal and neonatal plasma, RhD negative (or RhD identical with neonate) should be used. (NB If exchange or ‘top-up’ transfusion is required for HDN due to ABO incompatibility, group O red cells with low titre anti-A and B or group O red cells suspended in AB plasma should be used.) 4 Group O blood is acceptable; units with hightitre anti-A/anti-B must be excluded. 5 If the mother’s blood group is unknown, blood for the fetus/neonate should be crossmatched against the baby’s serum. 6 If no atypical antibodies are present in the maternal (or infant) sample, and if the DAT of the infant is negative, crossmatching is not necessary for the first 4 months of postnatal life. 7 If the antibody screen or DAT is positive, full serological investigation and compatibility testing are necessary. 8 An electronic crossmatch is not advisable unless an appropriate algorithm has been created, as ABO-identical adult blood transfused to an infant with maternal anti-A or anti-B may haemolyse even if the pretransfusion DAT is negative, due to stronger ABO antigen expression on adult cells. 9 Red cells (and platelets if given) should be cytomegalovirus (CMV) negative and leucocyte depleted. 100

10 Note that alloantibody formation is rare in the fetus and neonate and is usually associated with massive transfusion or with the use of fresh or whole blood. 11 Gamma-irradiation of cellular blood components to reduce the risk of transfusion-associated graft-versus-host disease (TA-GVHD) is recommended for: (a) all IUT; (b) all transfusions to neonates previously transfused in utero; (c) exchange transfusions as long as gammairradiation would not result in a delay in transfusion; (d) all transfusions from a family member; (e) all neonates with known inherited immune deficiencies (e.g. severe combined immunodeficiency). These precautions are due to the immaturity of the fetal and neonatal immune system which may lead to a reduced ability to reject transfused allogeneic lymphocytes, immune tolerance and the persistence of donor lymphocytes for up to 6–8 weeks after exchange transfusion. HDN: guidelines for prevention

The introduction of anti-D prophylaxis for recently delivered RhD-negative women in the UK in 1969 led to a reduction in new immunizations against anti-D from 17% of pregnancies to 1.5%. Every year in the UK 80 000 RhD-negative women have a RhD-positive infant and despite national guidelines sensitization still occurs, largely due to non-compliance with the guidelines. A dose of anti-D of 125 IU (25 mg) suppresses immunization by 1 mL of RhD-positive red cells (i.e. 2 mL of whole blood). (Note that in the UK the dose of anti-D is given in IU, whereas in other countries it is expressed in milligrams.) While anti-D is extremely effective as prophylaxis, it cannot reverse immunization once it has occurred and has no effect on the development of non-D antibodies. Indications for anti-D immunoglobulin (Table 8.2)

Anti-D should be given to all RhD-negative

Prenatal and childhood transfusions

women without anti-D antibodies after the following sensitizing events: • abortion (see below); • CVS; • ectopic pregnancy; • amniocentesis; • external cephalic version; • abdominal trauma; • antepartum haemorrhage; • premature labour; • pre-eclampsia; and • intrauterine death (associated with chronic fetomaternal haemorrhage). Anti-D should be administered following all abortions after 12 weeks, both spontaneous and induced, and following abortion at any gestation following surgical or medical treatment, including the use of abortifacients. Anti-D should also be administered in cases of threatened abortion if there is any bleeding after 12 weeks of gestation. Current UK guidelines also recommend the administration of anti-D immunoglobulin as antenatal prophylaxis since fetomaternal bleeding can happen at any gestation. Dose and schedule of administration of anti-D during the antenatal period

Therapeutic anti-D immunoglobulin to prevent the development of immune anti-D after sensitizing events should be given within 72 h of the sensitizing event; however, anti-D may still be

Table 8.2 Antenatal and postnatal prophylaxis with anti-D.

Indications for anti-D

Dose and schedule of administration

Sensitizing event <26 weeks’ gestation Threatened abortion continuing to term RhD-negative women without anti-D or a sensitizing event Standard postnatal prophylaxis

250 IU (50 mg) i.m. 500 IU every 6 weeks until term or 1250 IU every 10 weeks until term 500 IU at 28 weeks and 34 weeks

500 IU (1500 IU in USA and some European countries)

worthwhile up to 10 days after the event. In threatened abortion continuing to term, it is important to repeat the dose every 6 weeks (or use higher doses; see Table 8.2) since low antibody levels can augment the antibody response. The dose of anti-D is as follows. • Sensitizing event before 26 weeks of gestation: 250 IU (50 mg) i.m. • Threatened abortion continuing to term: 500 IU (100 mg) every 6 weeks or 1250 IU (250 mg) every 10 weeks until term. • It is now recommended in the UK that routine antenatal prophylaxis is provided in RhD-negative women without anti-D in the absence of a sensitizing event: 500 IU (100 mg) at 28 weeks and 34 weeks (no antibody screen is necessary before the dose at 34 weeks or at delivery). Dose and schedule of administration of anti-D in the postnatal period

Therapeutic anti-D immunoglobulin to prevent the development of immune anti-D should be given within 72 h of delivery; however, anti-D may still be worthwhile up to 10 days after the event. The dose of anti-D is as follows. • 500 IU (100 mg) is the standard dose to cover a fetomaternal bleed of 4 mL or less. • A Kleihauer test should always be performed to detect larger bleeds (bleeds >4 mL occur in 0.8% and of >15 mL in 0.3% of deliveries) so that an additional dose of anti-D (125 IU/mL blood loss) can be given. • The standard dose of anti-D in the USA and some European countries is higher (1500 IU). It takes 48 h following an intramuscular dose of anti-D to reach a good level and 72 h for clearance of sensitized red cells. If clearance of the RhDpositive cells is not complete, further anti-D must be given until RhD-positive cells can no longer be detected in maternal blood. Kleihauer test

This simple and inexpensive test is used to detect whether there has been a fetomaternal haemorrhage and the size of that haemorrhage. The principle and method for the test is as follows. 101

Chapter 8

• Fetal haemoglobin (HbF)-containing fetal red cells resist acid elution and therefore stain dark pink in comparison with HbA-containing cells, which appear as unstained ‘ghost’ cells (Plate 8.1, shown in colour between pp. 304 and 305). • To quantitate fetomaternal haemorrhage, the numbers of pink-staining HbF-containing cells in each single low-power field are counted; using this method a count of 200 HbF-containing cells or less in 50 low-power fields is equivalent to a fetomaternal haemorrhage of 4 mL or less. • Samples of maternal blood for the Kleihauer test must be taken within 2 h of administration of antiD to avoid a falsely low estimate of the size of the fetomaternal bleed. • Maternal hereditary persistence of fetal haemoglobin may cause a false-positive Kleihauer test due to maternal HbF-containing red cells.

Anti-D immunoglobulin not indicated

Anti-D immunoglobulin is not indicated in the following circumstances: • patients who are already sensitized; • those classified as weak D (e.g. Du); • if the infant is RhD negative; • for women not capable of child-bearing (following transfusion of RhD-positive blood); • for complete abortions before 12 weeks of gestation if there has been no surgical treatment. Preparation of anti-D immunoglobulin

Anti-D is a polyclonal antibody prepared by plasmapheresis of hyperimmunized donors, 95% of whom are women who have been sensitized during pregnancy. Anti-D is now prepared using US donor plasma because of concerns about transmission of variant Creutzfeldt–Jakob disease.

Large fetomaternal bleeds

Larger bleeds (>4 mL) may be measured using flow cytometry as well as the Kleihauer test. Larger fetomaternal bleeds are associated with: • amniocentesis; • abdominal trauma; • antepartum haemorrhage; • stillbirth; • twin pregnancy; • manual removal of the placenta. For any fetomaternal haemorrhage >4 mL an appropriate supplementary dose of anti-D must be given immediately AND a repeat test for fetal cells and free anti-D should be carried out on the mother 48 hours after the initial anti-D injection. A further appropriate dose of IgG anti-D should be given to the mother: • if fetal cells are no longer present but there is no residual free anti-D detectable (to make sure there is sufficient anti-D to eliminate small numbers of fetal cells below the limits of detection); • if fetal cells are still present but there is no detectable anti-D. Note that if fetal cells are still present after 48 hours but there is still detectable anti-D, a repeat test for fetal cells and free anti-D should be carried out after a further 48 hours to determine whether more anti-D IgG should be given to the mother. 102

Platelet and white cell disorders in pregnant women Differential diagnosis of thrombocytopenia in pregnancy

The most common causes of maternal thrombocytopenia are the following. • Gestational. • Pregnancy induced: pre-eclampsia, eclampsia, HELLP (haemolysis with elevation of liver enzymes and low platelets) syndrome. • Immune: immune thrombocytopenia (ITP) and systemic lupus erythematosus (SLE). • Virus associated, e.g. human immunodeficiency virus (HIV). Less common causes of maternal thrombocytopenia include: • antiphospholipid syndrome; • thrombotic thrombocytopenic purpura; • disseminated intravascular coagulation (DIC); • type IIb von Willebrand’s disease; • congenital bone marrow failure (e.g. Fanconi’s anaemia); • heparin-induced thrombocytopenia; • folate/B12 deficiency; • myelodysplasia/acute leukaemia. It may be difficult to distinguish between gestational, pregnancy-induced and immune thrombo-

Prenatal and childhood transfusions

cytopenia in pregnancy. ITP is more likely if the platelet count was subnormal prior to or during the first trimester of pregnancy. Further investigation depends on careful evaluation of the blood film and marrow smear, which may reveal characteristic changes (e.g. acute leukaemia). The disorders of particular relevance to transfusion medicine are ITP, HELLP and type IIb von Willebrand’s disease.

moderate not severe. Fulminant pre-eclampsia precipitating early delivery may be associated with DIC and require treatment with platelet transfusion and fresh frozen plasma (FFP) (with or without cryoprecipitate). Thrombocytopenia in HELLP syndrome is more often severe and platelet transfusion may be indicated, particularly at delivery, which is usually by urgent Caesarean section, and postpartum.

Management of maternal ITP

ITP usually presents in an otherwise well mother with or without a previous history of ITP or, less commonly, SLE. In those without a previous history, the diagnosis may be difficult and platelet antibody studies are of limited value since they may be positive even in the absence of ITP. The management of the mother with active ITP and who is thrombocytopenic should be as conservative as possible. The most common approach to therapy is with intravenous immunoglobulin (0.3–0.5 mg/kg daily) for 3–5 days. However, prednisolone (1 mg/kg) can also be used. The indications for treatment of maternal ITP are: • platelets less than 20 ¥ 109/L in the first, second or early third trimester; • aim to have platelets above 80 ¥ 109/L in the late third trimester; • avoid epidural or spinal anaesthesia if the platelet count is less than 80 ¥ 109/L; • platelet transfusion is very rarely indicated; • splenectomy should be postponed until after delivery if possible; • fetal blood sampling and elective Caesarean section for maternal ITP are unnecessary since significant fetal thrombocytopenia is uncommon (12%) and intracranial haemorrhage is even less common (1%). The fetal platelet count cannot be predicted from maternal platelet counts nor from platelet serology. The most important factor predicting the presence and severity of fetal thrombocytopenia is a history of maternal ITP prior to pregnancy: in this higher-risk group, 10–30% of babies will have significant thrombocytopenia (<50 ¥ 109/L).

Type IIb von Willebrand’s disease

Type IIb von Willebrand’s disease (vWD) is characterized by: • low factor VIII and von Willebrand factor (vWF); • thrombocytopenia due to platelet activation/ consumption; • reduced high-molecular-weight vWF multimers; • the need for prophylactic treatment at delivery with vWF or factor VIII. Leukaemia in pregnancy

When chemotherapy is administered during an ongoing pregnancy, both the mother and fetus may develop pancytopenia. Platelet transfusion for the mother should be given according to the usual guidelines, aiming to have a platelet count greater than 80 ¥ 109/L at delivery and for the first few days postpartum. The fetus should be monitored and fetal blood sampling with or without blood/platelet transfusion of irradiated blood products may be indicated on rare occasions. Maternal haemorrhagic problems

Guidelines for the investigation and management of haemorrhagic disorders in pregnancy have been published by the British Committee for Standards in Haematology (BCSH) Haemostasis and Thrombosis Task Force (1994) and are summarized here. Inherited disorders

Pre-eclampsia and HELLP syndrome

In pre-eclampsia, thrombocytopenia is usually

The most common inherited coagulation disorder is vWD. The prevalence of vWD is around 1% but 103

Chapter 8

many are only mildly affected. There are three main types of vWD: type I is the commonest (75%); type IIa occurs in 10%; type IIb occurs in 7% and may be accompanied by severe thrombocytopenia; and type III is the most severe. Most women with vWD have increased factor VIII and vWF during pregnancy and do not bleed excessively. Bleeding in pregnant women with vWD, when it does occur, usually causes problems: • during invasive procedures in the first trimester (e.g. CVS); • after birth, particularly after surgical delivery and/or perineal damage. Management of vWD in pregnancy can be summarized as follows. • For first-trimester procedures factor VIIIC activity should be raised to 50 U/dL. • At delivery, operative procedures should be avoided except for obstetric indications and trauma should be minimized. • Type I vWD: for vaginal delivery no blood product support is necessary if factor VIIIC is greater than 40 U/dL; if less than 40 U/dL or if Caesarean section is planned, factor VIIIC should be given to raise the level to at least 50 U/dL. • Type II and type III vWD: prophylactic factor VIIIC should be started at the onset of labour in all type III patients, most type IIa patients and some type IIb patients; the aim should be to raise the level of factor VIIIC to more than 40 U/dL; products with high levels of large vWF multimers should be used (e.g. 8Y, intermediate purity factor VIII concentrate containing vWF); for Caesarean section, factor VIIIC should be raised to more than 50 U/dL. Acquired disorders

The most common acquired disorder is DIC. This occurs secondary to: • severe eclampsia/pre-eclampsia; • intrauterine death; • placental abruption; • amniotic fluid embolism; • hydatidiform mole (chronic DIC). Haemorrhage may be severe. Treatment is as for DIC in non-pregnant patients except that the use of heparin is not recommended. Women who 104

develop HELLP syndrome with associated hepatic failure may also develop severe DIC with low fibrinogen and extremely low antithrombin levels. Maternal mortality approaches 30%; fetal mortality is around 50%. Major obstetric haemorrhage protocol

Severe bleeding is an obstetric emergency. It is essential to establish a major obstetric emergency haemorrhage protocol in any hospital where pregnant women are likely to deliver (see also Chapters 6 and 7). This protocol must be agreed by haematologists, including the haematology and transfusion laboratory staff, obstetricians and midwives, and those responsible for transporting urgent specimens and blood products. The most important points include the following. • If the blood group is known and there are no atypical antibodies: group-compatible red cells (6 units) should be issued. • If there are atypical antibodies: phenotyped red cells should be issued whenever possible. • If the blood group is unknown: group O RhDnegative blood (6 units) should be issued. • Samples should be obtained from the patient as soon as possible for blood group, full blood count and coagulation screen. • Once results are available further units of red cells, FFP, cryoprecipitate and/or platelets should be issued as indicated. • An antibody screen and retrospective crossmatch should be performed on the units issued as soon as time allows. Management of the major haemoglobinopathies during pregnancy Sickle cell disease

The frequency of maternal complications related to sickle cell disease, particularly vaso-occlusive crises, is not particularly increased during pregnancy. However, pregnant sickle cell disease patients do have an increased risk of: • placental insufficiency causing intrauterine growth restriction; • eclampsia/pre-eclampsia;

Prenatal and childhood transfusions

• preterm delivery; • stillbirth and neonatal death; • urinary tract infection (SC disease); • pulmonary thromboembolism (particularly SC disease).

• Desferrioxamine is discontinued during pregnancy because of the potential for nephrotoxicity, hepatotoxicity and ear/eye damage.

Fetal and neonatal transfusion Indications for red cell transfusion in pregnant women with sickle cell disease • Exchange transfusion should be carried out prior to planned operative delivery. • Exchange transfusion should be continued in any patient already on a regular exchange programme (e.g. for prevention of recurrent stroke). • Exchange transfusion should be considered for severe prolonged vaso-occlusive crises during pregnancy. • ‘Top-up’ transfusion should be considered for severe anaemia if symptomatic. • Regular prophylactic transfusion therapy during pregnancy does not reduce pregnancyrelated complications or improve outcome. • Red cells should be phenotyped (for Rh, Kell, Duffy, Kidd and MNS) to reduce the risk of alloimmunization, and red cells from sickle trait donors should not be used. Thalassaemia major and intermedia

Both maternal and fetal morbidity are higher in pregnant women with thalassaemia major and intermedia. In the mother, the degree of anaemia, the cardiovascular changes associated with pregnancy and pre-existing cardiac damage may aggravate the multiorgan damage secondary to iron overload. For the fetus, most reports suggest an increase in intrauterine growth restriction and preterm delivery. The rate of Caesarean section is increased, possibly because of cephalopelvic disproportion. Management is complex, requiring close liaison between haematologists and obstetricians. Management guidelines for pregnant women with thalassaemia major or intermedia • Haemoglobin should be maintained above 10–12 g/dL to optimize fetal outcome. • Transfusion requirements usually increase during pregnancy.

Causes of fetal and neonatal anaemia

The principal causes of fetal and neonatal anaemia are shown in Table 8.3. The commonest causes of fetal anaemia in the UK are parvovirus infection, HDN and twin-to-twin transfusion and the commonest causes of neonatal anaemia are iatrogenic blood loss and anaemia of prematurity. Management of neonatal anaemia (HDN is discussed in the next section)

The only available treatment for neonatal anaemia is red cell transfusion. Prevention or amelioration Table 8.3 Principal causes of fetal and neonatal anaemia.

Impaired red cell production Diamond–Blackfan anaemia Congenital infection (e.g. parvovirus, CMV)* Congenital dyserythropoietic anaemia Pearson’s syndrome Haemolytic anaemias Alloimmune: haemolytic disease of the newborn (Rh,ABO, Kell, other)* Autoimmune (e.g. maternal autoimmune haemolysis) Red cell membrane disorders (e.g. hereditary spherocytosis) Red cell enzyme deficiencies (e.g. pyruvate kinase deficiency) Some haemoglobinopathies (e.g. a-thalassaemia major*, HbH disease) Infection (e.g. bacterial, syphilis, malaria, CMV, Toxoplasma, herpes simplex Anaemia due to haemorrhage Occult haemorrhage before or around birth (e.g. twin-to-twin*, fetomaternal) Internal haemorrhage (e.g. intracranial, cephalhaematoma) Iatrogenic: due to frequent blood sampling Anaemia of prematurity Due to impaired red cell production, impaired erythropoietin production and reduced red cell lifespan Haemoglobin nadir usually 6.5–9 g/dL * Commonly present in the fetus; other causes may present during fetal life but neonatal presentation is more common. CMV, cytomegalovirus.

105

Chapter 8

of anaemia sufficient to reduce red cell transfusion requirements may be achieved by using recombinant erythropoietin, although this is not useful in the acute setting due to the 2-week delay in increasing haemoglobin using erythropoietin. Indications for ‘top-up’ transfusion

Guidelines for ‘top-up’ transfusion have been devised by committees in a number of countries, including the UK, Canada and the USA. Since there is almost no objective evidence, these guidelines are based on clinical experience and represent consensus views. Table 8.4 summarizes the indications for ‘top-up’ transfusions used in many UK neonatal intensive care units. Hazards of transfusion in neonates

Transfusion of neonates is becoming safer, especially with the more widespread use of satellite packs. The most important hazards of transfusion in neonates are: • infection (bacterial or viral); • hypocalcaemia (more common in neonates than in infants or children); • volume overload; • citrate toxicity; • rebound hypoglycaemia (high glucose levels from blood additives); • thrombocytopenia (after exchange transfusion); • TA-GVHD (if non-irradiated products given to those at risk; see above). Strategies to minimize transfusion risk in neonates

The following strategies have been shown to Table 8.4 Indications for neonatal ‘top-up’ transfusions.

Clinical situation

Transfuse at

Anaemia in the first 24 h Neonate receiving mechanical ventilation Acute blood loss Oxygen dependency (not ventilated) Late anaemia, stable patient (off oxygen)

Hb < 12 g/dL Hb < 12 g/dL 10% blood volume lost Hb < 8–10 g/dL Hb 7 g/dL

106

reduce the need for red cell transfusion and/or donor exposure in neonates: • development and implementation of local transfusion guidelines for each neonatal unit; • minimizing iatrogenic blood letting; • prevention and treatment of haematinic deficiencies (iron and folic acid); • use of dedicated satellite packs (Paedipacks); • improved antimicrobial screening; • judicious use of erythropoietin (see below); • autologous cord blood transfusion: delayed clamping of the cord at delivery. Neonatal ‘top-up’ transfusions: product specification

• Small volume ‘top-up’ transfusions can be given without further testing provided that there are no atypical maternal antibodies in maternal/infant serum and the infant’s DAT is negative. • Red cells should be 35 days old or less (if in SAG-M or similar additive system) or 28 days old or less (if in CPD). • Paedipacks (aliquotted donations from a single unit) should be used wherever possible for repeated transfusions in order to minimize donor exposure. • Recommended haematocrit for red cells for neonatal ‘top-up’ transfusion is 0.5–0.7. • Volume of a neonatal ‘top-up’ transfusion is usually 10–20 mL/kg. T-antigen activation

Severe haemolytic transfusion reactions are occasionally seen in neonates or young children transfused with adult blood or FFP containing anti-T antibodies. This may be due to exposure or ‘activation’ of the T antigen on neonatal red cells, usually as a result of infection with clostridia, streptococci or pneumococci and/or in association with necrotizing enterocolitis. Up to 25% of infants with necrotizing enterocolitis have T-antigen activation but so do many healthy neonates and haemolysis is extremely rare. Therefore, although there remains some controversy, the majority of centres worldwide feel that no special provision for neonates with necrotizing enterocolitis is necessary and

Prenatal and childhood transfusions

neither screen neonates for T activation nor donors for high-titre anti-T.

There have been numerous clinical trials of erythropoietin for prevention or amelioration of neonatal anaemia, particularly anaemia of prematurity, since endogenous erythropoietin production is low in preterm babies for the first 6–8 weeks of life. These trials show that erythropoietin (250 units/kg daily three times per week for the first 6 weeks of life) can reduce red cell transfusions in well preterm babies but has a negligible effect on transfusion requirements of sick preterm babies, particularly those of less than 26 weeks’ gestation at birth. In practice this means that erythropoietin has a limited role in neonates as it works best in those that need it least. Therefore most neonatal units no longer use erythropoietin routinely. The situations in which erythropoietin can be useful are: • in neonates whose parents refuse permission to use blood products; • to prevent ‘late anaemia’ in babies with HDN.

Management of HDN in the fetus Fetal monitoring of at-risk pregnancies

The aims are to prevent hydrops developing in utero and to time delivery so that the baby has the best chance of survival. Fetal monitoring includes the following. • Regular ultrasound scans for fetal growth, hepatosplenomegaly and/or hydrops. • Amniocentesis to measure amniotic fluid bilirubin as an indirect measure of fetal haemolysis (not reliable before 25 weeks and after 34 weeks of gestation); the bilirubin is plotted on a graph of Liley zones modified by Whitfield (Fig. 8.1) in order to predict the severity of HDN and plan management. • Fetal blood sampling if severe HDN before 24 weeks’ gestation is suspected, if there is a rapid rise in maternal antibody, or if there has been a previous intrauterine death due to HDN (note that fetal

0.6 0.5 0.4

O.D. 450 nm of liquor bilirubin

Role of erythropoietin in reducing neonatal red cell transfusion

1.0 0.8

0.3 0.25 0.02 0.16 0.14 0.12 0.08

Zone 1 Zone 2

0.06 0.05 0.04

Zone 3

0.03 0.025 0.002 0.016 0.012 0.010

20

25

30 Gestation (weeks)

35

40

Fig. 8.1 Whitfield’s (1970) curved action line is

superimposed on Liley zones of severity of fetal haemolysis based on estimations of amniotic fluid bilirubin concentrations. This gives guidelines for appropriate management related to gestation, i.e. intrauterine transfusion or, if the fetus is viable, delivery and exchange transfusion. (Adapted from Hann et al. 1991.)

blood sampling carries a 1–2% fetal loss rate and may cause fetomaternal haemorrhage with further sensitization). • IUT if anaemia is severe and delivery is not possible due to extreme prematurity. Intrauterine transfusion

The aims of IUT are: • to prevent or treat fetal hydrops before the fetus can be delivered; • to enable the pregnancy to advance to a gestational age that will ensure survival of the neonate (in practice, up to 36–37 weeks) with as few invasive procedures as possible (because of the risk of fetal loss). These are achieved by starting the transfusion programme as late as safely possible but before hydrops develops and by maximizing the intervals between transfusions by transfusing as large a 107

Chapter 8

volume of red cells as is considered safe. Transfusions may be intravascular, intraperitoneal or intracardiac. All transfusions are carried out using ultrasound guidance (Fig. 8.2). During transfusion the point of the needle and fetal heart should be watched closely for signs of needle displacement, cardiac tamponade and bradycardia. The fetal loss rate associated with IUT is around 5%. IUT is generally indicated when the haematocrit falls to below 0.25 between 18 and 26 weeks of gestation or to less than 0.3 after 26 weeks of gestation. The aim of the transfusion is to raise the haematocrit to 0.45 and repeat transfusion is often necessary after 2–3 weeks. Intrauterine transfusion: product specification

• Plasma-reduced red cells with a haematocrit of 0.75 or greater should be used. • The red cells should be 5 days old or less, in CPD anticoagulant and sickle screen negative. • The red cells should be group O (low-titre haemolysin) or ABO identical with the fetus (if known) and RhD negative (unless HDN is due to

Donor blood

3-way tap

anti-c, in which case RhD-positive, c-negative blood should be used). K blood is recommended to reduce additional maternal alloimmunization risks. • An IAT crossmatch compatible with maternal serum and negative for the relevant antigen(s) determined by maternal antibody status should be carried out. • Red cells for IUT should always be irradiated because of the risk of TA-GVHD. • Red cells for IUT should be warmed to 37°C immediately prior to transfusion and transfused at a rate of 5–10 mL/min. Management of HDN in the neonate

The severity of HDN varies considerably, from a hydropic infant with gross hepatosplenomegaly who needs immediate exchange transfusion to mild jaundice with or without anaemia. The following samples should be collected at delivery from all suspected cases: • ABO and Rh group; • DAT; • serum bilirubin; • full blood count, reticulocyte count and blood film. Affected babies should be monitored by checking their bilirubin and haemoglobin every 3–4 h. A rising bilirubin level may require treatment with exchange transfusion and/or phototherapy depending upon gestational age, postnatal age and birth weight (action charts are available for guidance). Phototherapy should be given from birth to all Rh-alloimmunised infants with haemolysis as the bilirubin can rise steeply after birth and this expectant approach will prevent the need for exchange transfusion in some infants. ‘Late’ anaemia presents at a few weeks of age in some babies with milder haemolytic disease who do not require exchange transfusion and in babies who have had earlier exchange transfusion; ‘top-up’ transfusion may be required. The blood film shows evidence of ongoing haemolysis and the anaemia is aggravated by the normal postnatal suppression of erythropoiesis. Exchange transfusion in neonates with HDN

Fig. 8.2 Intrauterine transfusion. (From Alter 1989.)

108

Exchange transfusion is used to treat severe

Prenatal and childhood transfusions

anaemia at birth, particularly in the presence of heart failure, and severe hyperbilirubinaemia. The aim is to remove both antibody-coated red cells and excess bilirubin. Exchange transfusion is a specialist procedure and should be undertaken only by experienced staff. Double-volume exchange (160–200 mL/kg) gives the best reduction in bilirubin (50%) and removes 90% of the infant’s circulating RhD-positive cells. The pH of whole blood or plasma-reduced red cells used in exchange transfusion is around 7.0, which does not cause acidosis in the infant.

• Jaundice may be severe but anaemia is usually mild. • The blood film shows very large numbers of spherocytes with little or no increase in nucleated red cells. • The DAT is usually, but not always, positive. • Severe HDN requiring exchange transfusion occurs in only 1 in 3000 births. • If an exchange transfusion is required, this should be with group O red cells, with low-titre anti-A and B or with group O red cells suspended in AB plasma.

Indications for exchange transfusion

Neonatal thrombocytopenia

• Cord haemoglobin less than 8 g/dL. • Cord bilirubin greater than 100 mmol/L. • Rapidly rising bilirubin.

Thrombocytopenia occurs in 1–4% of neonates. It is much more common in sick preterm infants, 40–50% of whom will develop thrombocytopenia in the first 4 weeks of life. Causes of neonatal thrombocytopenia are shown in Table 8.5. The most common cause presenting in the first few days of life is that associated with intrauterine growth restriction or maternal hypertension; however the most important cause of severe thrombocytopenia (platelets < 50 ¥ 109/L) at birth is neonatal alloimmune thrombocytopenia (NAITP).

Exchange transfusion in the neonate: product specification

• Plasma-reduced red cells with a haematocrit of 0.5–0.6 should be used, as packed cells may have a haematocrit up to 0.75 and cause a very high postexchange haematocrit. • The red cells should be less than 5 days old, collected into CPD anticoagulant and sickle screen negative. • The most recent BCSH guidelines (2003) state that red cells for neonatal exchange transfusion should be gamma-irradiated (and transfused within 24 h of irradiation); gamma-irradiation is essential in the case of neonates who have previously received IUT and in all other cases is advisable unless to do so would lead to clinically relevant delay. • Red cells for exchange should be warmed to 37°C immediately prior to transfusion. Special features of HDN due to ABO antibodies

• ABO haemolytic disease occurs only in offspring of women of blood group O and is confined to the 1% of such women that have high-titre IgG antibodies. • Haemolysis due to anti-A is more common (1 in 150 births) than anti-B.

Investigation of neonatal thrombocytopenia

In most cases the following tests will identify the diagnosis. • Full blood count and film: the combination of mild neutropenia and large numbers of nucleated red cells together with moderate thrombocytopenia suggest that the cause is intrauterine growth restriction or maternal hypertension. The presence of neutrophil left shift and toxic granulation with more severe thrombocytopenia suggests that the cause is bacterial infection with or without DIC. • Congenital infection screen: the most common congenital infection associated with neonatal thrombocytopenia is CMV. • Screening for NAITP (see below) should be carried out on any case of severe thrombocytopenia (platelets < 50 ¥ 109/L) presenting in the first week of life unless there is very clear evidence of acute infection. 109

Chapter 8 Table 8.5 Causes of neonatal thrombocytopenia.

Early onset (<72 h after birth) Placental insufficiency (pre-eclampsia, IUGR, diabetes)* Neonatal alloimmune thrombocytopenia* Birth asphyxia Perinatal infection (group B Streptococcus, Escherichia coli, Listeria) Congenital infection (CMV, Toxoplasma, rubella) Maternal autoimmune (ITP, SLE) Severe rhesus HDN Thrombosis (renal vein, aortic) Aneuploidy (trisomy-21, -18, -13) Congenital/inherited (TAR,Wiskott–Aldrich) Late onset (>72 h after birth) Bacterial and fungal sepsis* Necrotizing enterocolitis* Congenital infection (CMV, Toxoplasma, rubella) Maternal autoimmune (ITP, SLE) Congenital/inherited (TAR,Wiskott–Aldrich) * The most common causes. CMV, cytomegalovirus; HDN, haemolytic disease of the newborn; ITP, idiopathic thrombocytopenic purpura; IUGR, intrauterine growth restriction; SLE, systemic lupus erythematosus; TAR, thrombocytopenia with absent radii.

Neonatal alloimmune thrombocytopenia

• NAITP is analogous to HDN: maternal alloantibodies to antigens present on fetal platelets cause immune destruction of platelets in utero. • The five principal human platelet antigens (HPA1–5) show biallelic autosomal inheritance. • Alloantibodies to HPA-1a, HPA-5b and HPA-3a account for almost all cases of NAITP, the commonest being anti-HPA-1a (80–90% of cases of NAITP). • NAITP affects around 1 in 1000 pregnancies and occurs in the first pregnancy in almost 50% of cases. • Thrombocytopenia is frequently severe (platelets < 30 ¥ 109/L) and may present prenatally (as early as 20 weeks’ gestation) or at birth. • The ability of an HPA-1a-negative woman to form anti-HPA-1a is controlled by the HLA DRB3*0101 allele: HLA DRB3*0101-positive women are 140 times more likely to make antiHPA-1a than HLA DRB3*0101-negative women. • The main clinical problem in NAITP is intracra110

nial haemorrhage; this occurs in 10% of cases with long-term neurodevelopmental sequelae in 20% of survivors. • The diagnosis of NAITP is made by demonstrating platelet antigen incompatibility between mother and baby serologically or by polymerase chain reaction (PCR) and is carried out in reference transfusion laboratories (see Chapter 5 for methods). Management of NAITP • In all suspected cases the platelet count must be monitored for at least 72 h after birth as it may continue to fall during this time. • Severely thrombocytopenic babies (platelets <30 ¥ 109/L) should be transfused with HPAcompatible platelets (available ‘off the shelf’ from transfusion centres). • Babies with an intracranial haemorrhage in association with NAITP should have their platelet count maintained above 50 ¥ 109/L with HPAcompatible platelets. • If there is ongoing severe thrombocytopenia and/or haemorrhage despite HPA-compatible platelets, intravenous IgG (total dose 2 g/kg over 2–5 days) is often useful until spontaneous recovery occurs 1–6 weeks after birth. • All babies with severe thrombocytopenia due to NAITP should have a cranial ultrasound to look for evidence of intracranial haemorrhage (Fig. 8.3). Management of pregnancies at risk for NAITP (see also Chapter 5) • Prenatal management of NAITP remains controversial and all pregnancies should be monitored in a specialist fetal medicine centre with experience of NAITP. • The principal options are an invasive approach using fetal blood sampling plus fetal transfusion with HPA-compatible platelets if thrombocytopenia is detected or a non-invasive approach relying on non-invasive monitoring by fetal ultrasound and maternal intravenous IgG therapy; each approach has evidence to support it. • Paternal HPA genotyping is helpful since all fetuses fathered by men homozygous for HPA-1a (HPA-1a/1a) will be HPA-1a positive and there-

Prenatal and childhood transfusions

Fig. 8.3 MRI studies: inversion

recovery sequence (IR 1800/600/33) showing subacute haematoma (black arrow) and chronic haematoma (open arrow). (From de Vries et al. 1988.)

fore at high risk of developing severe thrombocytopenia, whereas only 50% of fetuses will be at risk if the father is heterozygous for HPA-1a (HPA-1a/1b). • There is no clear correlation between the titre of maternal anti-HPA antibodies and the severity of fetal thrombocytopenia and/or the development of intracranial haemorrhage. Neonatal thrombocytopenia due to maternal ITP

• Around 10% of infants of mothers with ITP or SLE develop neonatal thrombocytopenia secondary to transplacental passage of maternal platelet autoantibodies. • The thrombocytopenia is usually mild and intracranial haemorrhage occurs in less than 1% of at-risk babies. • The platelet count of babies born to mothers with ITP or SLE should be checked at birth and

monitored daily for 2–3 days if below 200 ¥ 109/L at birth. • If the baby is well, treatment is unnecessary unless the platelet count falls below 20 ¥ 109/L. • If the baby has severe thrombocytopenia (platelets <20 ¥ 109/L), treatment with intravenous IgG (0.4–1 g/kg per day, total dose 2–4 g/kg) is usually effective. • Cranial ultrasound to look for intracranial haemorrhage should be performed in all babies with severe thrombocytopenia (platelets <20 ¥ 109/L). • Platelet transfusion is reserved for life-threatening haemorrhage and should be given in conjunction with intravenous IgG. Indications for platelet transfusion in neonates

Published guidelines for neonatal platelet transfusion acknowledge the lack of evidence on which 111

Chapter 8

to base recommendations and aim for a safe approach. Suggested guidelines based on clinical experience are shown in Table 8.6. There is some evidence to suggest that prophylactic platelet transfusions are not required for healthy neonates until the platelet count falls to 20–30 ¥ 109/L. However, a higher trigger level (<50 ¥ 109/L) should be used for babies with the greatest risk of haemorrhage, especially extremely low birth weight neonates (<1000 g) in the first week of life. Neonatal platelet transfusion: product specification

• ABO identical or compatible. • RhD identical or compatible. • HPA compatible in infants with NAITP. • Produced by standard techniques without further concentration. • Irradiated if appropriate. • Volume transfused usually 10–20 mL/kg. Neonatal neutropenia

Normal neutrophil levels vary with postnatal age, falling in healthy babies from around 5–10 ¥ 109/L at birth to 2–6 ¥ 109/L by the end of the first week of life. Neutropenia is therefore variably defined Table 8.6 Guidelines for platelet transfusion in neonatal

thrombocytopenia. Platelet count < 30 ¥ 109/L in otherwise well infants, including NAITP, if no evidence of bleeding and no family history of intracranial haemorrhage Platelet count < 50 ¥ 109/L in infants with: Clinical instability Concurrent coagulopathy Birth weight < 1000 g and age < 1 week Previous major bleeding (e.g. GMH–IVH) Current minor bleeding (e.g. petechiae) Planned surgery or exchange transfusion Platelet count falling and likely to fall below 30 ¥ 109/L NAITP if previous affected sib with intracranial haemorrhage Platelet count < 100 ¥ 109/L in infants with major bleeding GMH–IVH, Germinal matrix–intraventricular haemorrhage; NAITP, neonatal alloimmune thrombocytopenia.

112

depending on postnatal age: less than 2.0 ¥ 109/L at birth and less than 1.0 ¥ 109/L from 1 week of age. The most common causes are neutropenia secondary to intrauterine growth restriction or maternal hypertension and neutropenia secondary to severe sepsis. The presence of neutrophil left shift and toxic granulation in a neutropenic neonate suggests acute bacterial infection. Alloimmune neonatal neutropenia (see also Chapter 5)

• This is analogous to HDN: there is maternal sensitization to fetal neutrophil antigens during pregnancy. • The most common implicated antibodies are anti-NA1 and anti-NA2. • The estimated incidence of neonatal alloimmune neutropenia is 3% of live births but most cases are mild and asymptomatic and the diagnosis may be missed. • Infants with severe neonatal alloimmune neutropenia develop severe cutaneous, respiratory or urinary tract infection. • Treatment is with antibiotics and, if necessary, granulocyte colony-stimulating factor. Granulocyte transfusions in neonates

There is no good evidence for benefit of granulocyte transfusions for treatment of neonatal infection. Both granulocyte colony-stimulating factor and granulocyte–macrophage colony-stimulating factor can be used to increase the neutrophil count in neutropenic neonates but there is no clear evidence that this improves outcome.

Coagulation problems in the newborn Causes of haemorrhage in the newborn

In well infants the most common causes of bleeding are: • vitamin K deficiency (haemorrhagic disease of the newborn); • inherited disorders, particularly haemophilias; • NAITP. In sick infants the most common causes are:

Prenatal and childhood transfusions

• DIC, secondary to perinatal asphyxia, necrotizing enterocolitis or, less commonly, sepsis; • liver disease. Vitamin K deficiency

Vitamin K deficiency remains a clinical problem largely because of the controversy in recent years surrounding the possible tumorigenic effects of intramuscular vitamin K administered to newborn infants to prevent haemorrhagic disease of the newborn. Vitamin K is necessary for the posttranslational carboxylation of factors II, VII, IX and X and of the natural anticoagulants protein C and protein S. Levels of vitamin K and of all these factors are physiologically low at birth. This physiological deficiency can be exacerbated by breastfeeding, prematurity and liver disease, resulting in haemorrhagic disease of the newborn, often now referred to as vitamin K-dependent bleeding (VKDB). There are three patterns of VKDB: early, classical and late. • Early VKDB presents in the first 24 h of life usually with severe haemorrhage, including gastrointestinal bleeding and intracranial haemorrhage. It is caused by severe vitamin K deficiency in utero, usually as a result of maternal medication that interferes with vitamin K, e.g. anticonvulsants (phenobarbital, phenytoin), antituberculous therapy and oral anticoagulants. • Classical VKDB presents at 2–7 days old in babies who have not received prophylactic vitamin K at birth. The risk is increased in breast-fed babies and in those with poor oral intake. The incidence in babies not receiving vitamin K supplementation is 0.25–1.7%. Classical VKDB can be prevented by a single intramuscular dose of vitamin K at birth. • Late VKDB occurs 2–8 weeks after birth. It usually presents with sudden intracranial haemorrhage in an otherwise well, breast-fed term baby or in babies with liver disease. Late VKDB in healthy breast-fed babies can be prevented either by a single intramuscular dose of vitamin K or by repeated oral doses of vitamin K over the first 6 weeks of life; babies with chronic liver disease or malabsorption require prolonged vitamin K supplementation. • Diagnosis of VKDB is based on clotting studies, which show a prolonged PT with normal platelets

and fibrinogen; in severe deficiency the APTT may also be prolonged. • Treatment of VKDB depends on the severity of the bleeding. Mild cases should be given vitamin K (1 mg) intravenously or subcutaneously as this increases the levels of active vitamin K-dependent coagulation factors within a few hours; where there is significant bleeding, FFP may be given in addition to vitamin K. Vitamin K prophylaxis

Guidelines for the prevention of early VKDB recommend a single intramuscular injection of vitamin K at birth together with antenatal administration of oral vitamin K to the mother during the last 4 weeks of pregnancy. For classical and late VKDB there are several options because although intramuscular vitamin K at birth prevents classical and late VKDB, some studies have suggested a link between intramuscular vitamin K at birth and later childhood malignancies. Although other studies have not confirmed the link with malignancy, the controversy is unlikely to be resolved unequivocally in the short term. There is no link between oral vitamin K and malignancy. Both the American Academy of Pediatrics and the Royal College of Paediatrics and Child Health recommend vitamin K supplementation at birth. In healthy babies the choice of which route of administration is left to parents, who have to balance a possible risk of leukaemia [odds ratio between 1.6 (CI 0.89–1.25) and 1.16 (CI 0.97–1.39)] with intramuscular vitamin K against the slightly higher risk of VKDB (2.7 per 100 000) in infants given three doses of 1 mg vitamin K orally at birth, 1 week and 1 month of age. Use of FFP, cryoprecipitate and human albumin solution in neonates

Guidelines for the use of FFP, cryopreciptate and albumin in neonates have been published by national committees in a number of countries. The guidelines aim to minimize their risks in the newborn by using pathogen-inactivated products and by recommending their use for a small number of clinical indications. The only indications for 113

Chapter 8

FFP in neonates recommended in the recent BCSH guidelines and supported by evidence are DIC, VKDB and inherited deficiencies of coagulation factors. Prophylactic FFP administered to preterm neonates at birth does not prevent intraventricular haemorrhage or improve outcome at 2 years of life. Similarly, FFP is not superior to other colloid or crystalloid solutions as a volume replacement solution in standard neonatal practice and there is no evidence to support its use to ‘correct’ the results of abnormal coagulation screens. Product specifications

• The current BCSH guidelines state that FFP for transfusion to neonates should be group AB (since this contains neither anti-A nor anti-B) or the same ABO blood group as the neonate. • The volume transfused is usually 10–20 mL/kg, with the larger dose given if possible in order to limit donor exposure where repeated dosing is likely. • FFP may be standard or pathogen inactivated. In England at the present time the Department of Health has indicated that single-unit methylene blue-treated FFP should be used for neonates and for children born after 1 January 1996. FFP for this group will be sourced from plasma from the USA from early 2004. Standard FFP for other patients is prepared from leucocyte-depleted whole blood donations from UK donors. It carries a small residual risk of transmitting transfusiontransmissible viruses (with the exception of CMV). Coagulation factor levels are 20–25% lower in pathogen-inactivated FFP than untreated FFP. Human albumin solution (HAS) is associated with excess mortality in adults receiving intensive care but data about the risks of HAS in neonates are not available. Current studies suggest that there is no good indication for the use of HAS in standard neonatal practice.

Transfusion in children General points

Although most children never require blood trans114

fusion, there are several groups who are frequently transfused, including those with inherited transfusion-dependent disorders, such as thalassaemia major, and those undergoing intensive chemotherapy for haematological malignancies. For many of these patients, including those with thalassaemia major and sickle cell disease, bone marrow transplantation (BMT) is a possible future treatment. Therefore all such children for whom BMT is a possible option should receive CMV-negative blood components. All children on regular transfusions should be vaccinated against hepatitis B as early as possible. Those on chronic transfusion therapy, particularly those with haemoglobinopathies but also those with congenital dyserythropoietic anaemia, aplastic anaemia and other bone marrow failure syndromes, should have an extended red cell phenotype (see below) performed prior to, or as soon as possible after, commencing regular transfusions. The formula for calculating red cell transfusion volume in children is as follows: Desired Hb (g/dL) – actual Hb (g/dL) ¥ weight (kg) ¥ 3 where ‘3’ represents 3 mL of red cells, which has been calculated to raise the haemoglobin by 1 g/dL. The normal rate of red cell transfusion is around 5 mL/kg per h.

Transfusion support for children with haemoglobinopathies Thalassaemia major

By definition all patients with thalassaemia major are transfusion dependent. Transfusion therapy is determined by the degree of anaemia and evidence of failure to thrive. Most children start transfusion when their haemoglobin drops below 6 g/dL. Current BCSH and Thalassaemia International Federation guidelines recommend the following: • maintain an average haemoglobin of 12 g/dL; • maintain a pretransfusion haemoglobin of 9–10 g/dL; • transfusion should prevent marrow hyperplasia, skeletal changes and organomegaly;

Prenatal and childhood transfusions

• extended red cell phenotyping should be carried out before starting transfusions (for Rh and Kell antigens); • red cell requirements should be adjusted to accommodate growth; • splenectomy may be considered if hypersplenism develops and causes a sustained increase in red cell requirements; • iron chelation therapy should be considered after 10 transfusions and started once ferritin is greater than 1000 ng/mL (if possible starting after the age of 2 years because of desferrioxamine toxicity); • since BMT is the only cure, families should be offered HLA typing of siblings as possible bone marrow donors. Sickle cell disease

Red cell transfusion in children with sickle cell disease should not be routine but reserved for specific indications (Table 8.7). Extended red cell phenotyping before the first transfusion is very important because up to 50% of patients otherwise develop red cell alloimmunization and may be very difficult to crossmatch. The majority of antibodies are in the Rhesus or Kell system and may be transient and very difficult to detect, leading to a risk of delayed transfusion reactions. Indications for ‘top-up’ transfusion in sickle cell disease Indications include splenic or hepatic sequestration, and aplastic crisis. The aim is to raise the haemoglobin to the child’s normal steady state (the haemoglobin should never be raised acutely to >10 g/dL, since this is likely to cause an increase in blood viscosity). Indications for exchange transfusion in sickle cell disease • Acute chest syndrome. • Mesenteric (abdominal) syndrome. • Stroke. • Selected patients preoperatively. • Priapism (occasionally). The aim is to reduce sickling and increase oxygen carriage without an increase in viscosity.

Table 8.7 Indications for transfusion in sickle cell disease.

‘Top-up’ Splenic sequestration* Hepatic sequestration* Aplastic crises* Exchange transfusion Chest syndrome* Stroke* Priapism Mesenteric syndrome Hypertransfusion Stroke (to prevent recurrence)* Renal failure (to prevent/delay deterioration) Chronic sickle lung disease Surgery Selected patients preoperatively (e.g. joint replacement) * Proven value.

Indications for hypertransfusion in sickle cell disease • To prevent recurrence of stroke (i.e. secondary prevention of stroke). • To prevent the development of stroke in children with sickle cell disease with Doppler and/or magnetic resonance evidence of cerebrovascular infarction/haemorrhage in the absence of clinical evidence of stroke (i.e. primary prevention of stroke). • To delay or prevent deterioration in end-organ failure (e.g. chronic sickle lung). The aims are to maintain the percentage of HbS below 25% and the haemoglobin between 10 and 14.5 g/dL. After 3 years a less intensive regimen maintaining HbS at 50% or less may be sufficient for stroke prevention. Indications for preoperative transfusion in sickle cell disease The BCSH guidelines are based on observational studies and one large randomized controlled study as there are no other available data. These guidelines state that: • ‘top-up’ transfusion (haemoglobin 8–10 g/dL) is as effective as exchange transfusion and may be safer; • minor and straightforward procedures (e.g. ton115

Chapter 8

sillectomy, possibly cholecystectomy) can be safely undertaken without transfusion in most patients; • exchange transfusion should be performed preoperatively for major procedures, e.g. hip/knee replacement, organ transplantation, eye surgery, and considered for major abdominal surgery. Practical aspects of transfusion in sickle cell disease • Extended red cell phenotyping (for Rh, K, Fy, Jk and MNS) should be carried out; this should be done before the first transfusion and may be usefully arranged at outpatient clinic follow-up during the first year of life. • In particular patients should be typed for U. • The R0 blood group (cDe/cDe) is common in patients of African or Caribbean origin: all R0 patients should receive C-negative, E-negative blood (i.e. rr or R0). • The use of sickle trait-positive blood should be avoided by testing for HbS in blood centres or hospitals. • During exchange transfusion in the acute situation a total exchange of 1.5–2 times their blood volume is required to achieve an HbS level of 20% or less; this may take two to three procedures if carried out manually. Automated exchange using a cell separator allows the exchange to be completed as a single procedure. The volume of packed cells (in mL) for each exchange is determined by: weight (kg) ¥ 30. • Normal saline (not FFP or albumin) should be used as volume replacement at the beginning of the exchange prior to starting venesection to avoid lowering the circulating blood volume. Leukaemia, chemotherapy and BMT

All children being treated with high-dose chemotherapy/radiotherapy or those with aplastic anaemia may at some time be candidates for future BMT. While components leucocyte depleted to less than 5 ¥ 106/unit are widely considered to be CMV-safe, not all BMT centres agree. Gamma-irradiation of blood components is not necessary for most children receiving chemotherapy for leukaemia or solid tumours, although there are several important situations where irradiation of blood products (25 Gy) is necessary. 116

Gamma-irradiated blood components should be given to the following children

• For 2 weeks before allogeneic haemopoietic stem cell transplant (SCT) and during conditioning for all types of SCT until at least 6 months postSCT or until all immunosuppressive agents have been discontinued, whichever is later. • For 2 weeks before autologous SCT irradiation and during conditioning until 3 months post-SCT (6 months if total body irradiation given). • For SCT in children with severe combined immunodeficiency, irradiation should continue for at least a year following SCT or until normal immune function has been achieved. • For 7 days prior to harvesting of autologous bone marrow or peripheral blood stem cells. • For children with Hodgkin’s disease during treatment and thereafter (susceptibility to transfusion-associated GVHD is now considered to be lifelong). • During treatment with fludarabine and for at least 2 years or until full recovery of cellular immune function. • Where blood products from relatives are being used. Transfusions for children undergoing blood groupmismatched BMT

• Major incompatibility arises when the recipient has antibodies to the donor cells (e.g. patient group O, donor group A). • Minor incompatibility arises when the donor has antibodies to recipient cells (e.g. patient group A, donor group A). • Current BCSH and European Blood and Bone Marrow Transplantation Group guidelines recommend that in ABO-incompatible SCT group O red cells should be given (irrespective of the ABO group) until ABO antibodies to the donor ABO type are undetectable and the DAT is negative; thereafter red cells of the donor group are given (high-titre anti-A/anti-B donor units must be excluded). • RhD-negative red cells are given if the patient is RhD negative and/or the donor is RhD negative. • After an ABO-incompatible SCT, platelets of the

Prenatal and childhood transfusions

recipient’s ABO group should be given until there is conversion to the donor ABO group and ABO antibodies to the donor ABO group are undetectable; thereafter platelets of the donor group should be given. • After an ABO-incompatible SCT, FFP of the recipient’s ABO group should be given. If there is both a major and minor mismatch, group AB should be given. Platelet transfusion in children undergoing chemotherapy or BMT

• Indications for platelet transfusion in children are consensus based; those developed by the BCSH are shown in Table 8.8; in general, in non-infected well children a platelet count of 5–10 ¥ 109/L can be used as a transfusion trigger but higher thresholds are used for children who are sick and/or bleeding. • Platelets should be ABO compatible where possible because of the risk of haemolysis (see above for ABO-incompatible SCT patients). • Platelets should be RhD compatible and RhDnegative girls must receive RhD-negative platelets because of the risk of sensitization by contaminating red cells. • A transfusion of 10–20 mL/kg is given to children under 15 kg and an apheresis unit for children over 15 kg.

Table 8.8 Indications for platelet transfusion in children

with thrombocytopenia.

Granulocyte transfusion in children undergoing chemotherapy or BMT

• There is no evidence to support the use of prophylactic granulocyte transfusions. • Empirical data from some studies support their use where there is severe bacterial or fungal infection in neutropenic children, including SCT, but they increase the risk of platelet refractoriness. • Granulocytes for transfusion should be ABO and RhD compatible. • Granulocytes for all recipients should always be irradiated. Haemopoietic stem cell donors

• Children who act as bone marrow donors for their sibling(s) usually require blood transfusion to cover blood lost during the procedure; allogeneic blood transfused to the donor during the bone marrow harvest should always be irradiated and CMV negative unless both the patient and donor are known to be CMV positive. • In older children (>25 kg and more than 8 years old), autologous blood donation should be considered around 2 weeks prior to marrow/peripheral blood stem cell donation. • For autologous donation children should have no unstable cardiovascular or pulmonary problems and a haemoglobin of more than 11 g/dL. The maximum collected at each donation should be 12% of the estimated blood volume and the amount of citrate anticoagulant in the pack should be adjusted to maintain the appropriate ratio of blood to anticoagulant.

Platelet count < 10 ¥ 109/L Platelet count < 20 ¥ 109/L and one or more of the following: Severe mucositis DIC Anticoagulant therapy Platelets likely to fall < 10 ¥ 109/L before next evaluation Risk of bleeding due to local tumour infiltration Platelet count 20–40 ¥ 109/L and one or more of the following: DIC in association with induction therapy for leukaemia Extreme hyperleucocytosis Prior to lumbar puncture or central venous line insertion DIC, disseminated intravascular coagulation.

Further reading Alagappan A, Shattuck KE, Malloy MH. Impact of transfusion guidelines on neonatal transfusions. J Perinatol 1998; 18: 92–7. Alcock GS, Liley H. Immunoglobulin infusion for isoimmune haemolytic jaundice in neonates. Cochrane Database Syst Rev 2002; 3: CD003313. Alter BP. Methods in Haematology: Prenatal Haematology. Edinburgh, Churchill Livingstone, 1989. BCSH Blood Transfusion and Haematology Task Forces. The estimation of fetomaternal haemorrhage. Transfus Med 1999; 9: 87–92.

117

Chapter 8 Boralessa H, Modi N, Cockburn H et al. RBC T activation and hemolysis in a neonatal intensive care population: implications for transfusion practice. Transfusion 2002; 42: 1428–34. British Committee for Standards in Haematology Haemostasis and Thrombosis Task Force. The investigation and management of neonatal haemostasis and thrombosis. Br J Haematol 2002; 119: 295–309. British Committee for Standards in Haematology Transfusion Task Force. Transfusion guidelines for neonates and older children. www.bcshguidelines.com/ British Committee for Standards in Haematology. Blood Transfusion Task force. Guidelines for pre-transfusion compatibility procedures in blood transfusion laboratories. Transfusion Medicine 1996; 6: 273–283. Bruce M, Chapman JF, duguid J, Kelsey P, Knowles S, Murphy M, Williamson L Addendum for guidelines for blood grouping and red cell antibody testing during pregnancy. BCSH Transfusion Task Force. Transfus Med 1999; 9: 99. Bussel JB. Alloimmune thrombocytopenia in the fetus and newborn. Semin Thromb Hemost 2001; 27: 245– 52. de Vries LS, Connell J, Bydder GM et al. Recurrent intracranial haemorrhages in utero in an infant with alloimmune thrombocytopenia. Case report. Br J Obstet Gynaecol 1988; 95: 299–302. Gottstein R, Cooke RWI. Systematic review of intravenous immunoglobulin in haemolytic disease of the newborn. Arch Dis Child 2003; 88: F6–F10. Grant SR, Kilby MD, Meer L, Weaver JB, Gabra GS, Whittle MJ. The outcome of pregnancy in Kell alloimmunisation. Br J Obstet Gynaecol 2000; 107: 481–5. Hann IM, Gibson BES, Letsky EA. Haemolytic disease of the newborn. In: Fetal and Neonatal Haematology. London: Baillière Tindall, 1991; 106. Heegaard ED, Brown KE. Human parvovirus B19. Clin Microbiol Rev 2002; 15: 485–505. Kadir RA. Women and inherited bleeding disorders: pregnancy. Semin Hematol 1999; 36: 28–35. Kelton JG. Idiopathic thrombocytopenic purpura complicating pregnancy. Blood Rev 2002; 16: 43–6. Kulkarni R, Lusher J. Perinatal management of newborns with haemophilia. Br J Haematol 2001; 112: 264–74. Lee D, Contreras M, Robson SC, Rodeck CH, Whittle MJ. Recommendations for the use of anti-D immunoglobulin for Rh prophylaxis. British Blood Transfusion Society and the Royal College of Obstetricians and Gynaecologists. Transfus Med 1999; 9: 93–7. Meyer MP, Sharma E, Carsons M. Recombinant erythropoietin and blood transfusion in selected preterm infants. Arch Dis Child 2003; 88: F41–F45.

118

Murray NA, Howarth LJ, McMcloy M, Letsky EA, Roberts IAG. Platelet transfusion in the management of severe thrombocytopenia in neonatal intensive care unit (NICU) patients. Transfus Med 2002; 12: 35–41. National Institute for Clinical Excellence (NICE). Guidelines on the use of routine antenatal anti-D prophylaxis for RhD-negative women. Technology Appraisal Guidance No 41. www.nice.org.uk/pdf/ prophylaxisFinalguidance.pdf. 2002. Paul DA, Leef KH, Locke RG, Stefano JL. Transfusion volume in infants with very low birthweight: a randomized trial of 10 versus 20 ml/kg. J Pediatr Hematol Oncol 2002; 24: 43–6. Puckett RM, Offringa M. Prophylactic vitamin K for vitamin K deficiency bleeding in neonates. Cochrane Database Syst Rev 2000; 4: CD002776. Ramasethu J, Luban NLC. Red blood cell transfusion in the newborn. Semin Neonatol 1999; 4: 5–16. Roberts IAG, Murray NA. Thrombocytopenia in the newborn. Curr Opin Pediatr 2003; 15: 17–23. Stockman JA, de Alarcon PA. Overview of the state of the art of Rh disease: history, current clinical management, and recent progress. J Pediatr Hematol Oncol 2001; 23: 385–93. Sutor AH, von Kries R, Cornelissen EA, McNinch AW, Andrew M. Vitamin K deficiency bleeding (VKDB) in infancy. ISTH Pediatric/Perinatal Subcommittee. International Society on Thrombosis and Haemostasis. Thromb Haemost 1999; 81: 456–61. Thompson J. Haemolytic disease of the newborn: the new NICE guidelines. J Fam Health Care 2002; 12: 133–6. Vamvakas EC, Strauss RG. Meta-analysis of controlled clinical trials studying the efficacy of rHuEPO in reducing blood transfusions in the anemia of prematurity. Transfusion 2001; 41: 406–15. Waldron P, de Alarcon P. ABO hemolytic disease of the newborn: a unique constellation of findings in siblings and review of protective mechanisms in the fetal–maternal system. Am J Perinatol 1999; 16: 391–8. Wee LY, Fisk NM. The twin–twin transfusion syndrome. Semin Neonatol 2002; 7: 187–202. Wong W, Fok TF, Lee CH et al. Randomised controlled trial: comparison of colloid or crystalloid for partial exchange transfusion for treatment of neonatal polycythaemia. Arch Dis Child 1997; 77: F115–F118. Zipursky A. Prevention of vitamin K deficiency bleeding in newborns. Br J Haematol 1999; 104: 430–7. Zuppa AA, Maragliano G, Scapillati ME et al. Recombinant erythropoietin in the prevention of late anaemia in intrauterine transfused neonates with Rh-haemolytic disease. Fetal Diagn Ther 1999; 14: 270–465.

Chapter 9

Haematological disease Michael F. Murphy and Simon J. Stanworth

Patients with haematological diseases are major users of blood products. Over 15% of all red cell units are transfused to patients with haematological disease, mostly to patients with malignant disorders. The requirement for blood transfusions in this group are related to both the underlying condition itself and the myelosuppressive/myeloablative effects of specific treatments used. This chapter considers the following topics: • the indications for red cell, platelet and granulocyte transfusions in haematology patients; • the approaches to the management and prevention of complications associated with transfusions in haematology patients, including the use of special types of blood components.

Indications for transfusions in patients with haematological diseases Haematological diseases requiring transfusion support cover a whole spectrum of clinical disorders: fetal, neonatal and paediatric practice (Chapter 8), haemophilia (Chapter 11), haemoglobinopathies, immune disorders, and bone marrow failure syndromes, in addition to haematological malignancies. The haemopoietic system has a dramatic capacity for increasing the production of mature blood cells, but this capability varies between different diseases. The scenario of anaemia related to marrow ablation following chemotherapy is very different to anaemia in an individual with a wellcompensated chronic haemolytic process. It is also acknowledged that while much of the current impetus in transfusion practice is aimed at reducing inappropriate transfusions, there is now evi-

dence indicating a risk of ‘under-transfusing’ certain groups of patients, for example those with coexisting cardiac disease. Red cell transfusions

The ready availability of red cell concentrates means that anaemia in haematology patients can be easily relieved. There are some specific considerations in the management of anaemia. • Its cause should be established, and treatment other than blood transfusion should be used where appropriate, for example patients with iron deficiency or megaloblastic or autoimmune haemolytic anaemias. Anaemia of malignancy may be due to a number of causes including the effects of marrow infiltration or therapy and ‘inhibitory’ cytokine-mediated influences leading to the secondary anaemias (of chronic disorders), or low erythropoietin. • There is no universal ‘trigger’ for red cell transfusions in haematology patients, i.e. a given level of haemoglobin at which red cell transfusion is appropriate for all patients. Clinical judgement balancing factors such as quality-of-life indices plays an important role in the decision to transfuse red cells or not. • The clinical use of recombinant erythropoietin might be considered in some situations, e.g. delayed erythroid engraftment after allogeneic bone marrow/peripheral blood progenitor cell transplantation, the treatment of anaemia in patients with myeloma or myelodysplasia, and in the management of Jehovah’s Witnesses with haematological disorders. Evidence supports an association between increases in haemoglobin concentration, reduced red cell transfusion require119

Chapter 9

ments and improvement in quality-of-life indices with erythropoietin therapy, although the findings concerning quality-of-life measures are more difficult to compare between studies. Uncertainties also remain about the factors predicting responsiveness, since a number of individuals fail to show adequate responses to erythropoietin. Overall cost-effectiveness studies have not documented a major cost–benefit for erythropoietin, but this balance may alter in the light of changes to the supply and costs of donor blood and a better understanding of the potentially beneficial effects of erythropoietin on tumour response and overall survival. Patients receiving intensive myelosuppressive/myeloablative treatment

There are specific considerations relating to the use of red cell transfusions in patients receiving intensive myelosuppressive/myeloablative treatment, including the need to provide a ‘reserve’ in case of severe infection or haemorrhage, and the convenience of having a standard policy for red cell transfusion in the setting of an acute haematology service, even if this may result in some patients being over-transfused. The level of haemoglobin used as the ‘trigger’ for transfusion varies from centre to centre, but is usually in the range 8–10 g/dL. There are no definite data to support the use of a higher level, although studies in animal models of thrombocytopenia and in uraemic patients suggest that correction of anaemia also results in correction of prolonged bleeding times. Red cell transfusions and chronic anaemias

In patients with chronic anaemia requiring regular transfusions, red cell transfusions should be used to maintain the haemoglobin level just above the lowest level not associated with symptoms of anaemia. There is considerable variation in this level depending on the patient’s age, level of activity and coexisting medical problems, such as cardiovascular and respiratory disease; for example, some young patients are asymptomatic with a haemoglobin level below 8 g/dL, while some 120

elderly patients are symptomatic even at haemoglobin levels above 10 g/dL. Haemoglobinopathies (in adults)

Similar principles guide red cell transfusion support for adult patients with transfusiondependent thalassaemia as for children (see Chapter 8). Pretransfusion haemoglobin concentrations should be 9–10 g/dL, and iron-chelation therapy will be required in all patients (see below). All patients on long-term transfusion programmes should be vaccinated against hepatitis B. Transfusions of red cells in sickle cell disease are used to correct anaemia or as part of the treatment of vaso-occlusive or other complications of severe sickling. Factors other than the number or percentage of sickle cells may affect the risk of vasoocclusion, e.g. baseline severity of anaemia, infection. Transfusion support for sickle disease may be classified by the clinical indication (e.g. aplastic crisis, acute chest syndrome, stroke, priapism, before surgery), the predominant pathophysiological reason (anaemia or sickling or iron overload), or by the chosen transfusion protocol (simple or additive transfusion, exchange transfusion or hypertransfusion). Exchange transfusions may be carried out manually at the bedside or by apheresis using a cell separator (with the possible attendant problems relating to obtaining good venous access). Careful attention to fluid balance is required, and saline should be given prior to venesection at the start of an exchange procedure to prevent falls in blood volume. The rationale for differing transfusion requirements as part of the management of these clinical complications of sickle cell disease does not relate solely to the desired post-transfusion reduction in HbScontaining cells, and the exact mechanism by which exchange transfusion, for example, is beneficial in acute chest syndrome is not fully defined. Transfusion support may be indicated in relation to surgery, because anaesthesia and surgery may increase the risk of sickling complications. Transfusion support may not be required prior to minor surgical procedures, e.g. tonsillectomy. Transfusions when indicated are given either as a

Haematological disease

‘top-up’ or as an exchange. A large randomized trial noted that an aggressive transfusion regimen did not reduce the number of postoperative vaso-occlusive complications compared with topup transfusions to a haemoglobin concentration of 10 g/dL, and had the disadvantage that red cell alloimmunization was twice as common. One of the major concerns about transfusions in sickle cell disease is red cell alloimmunization, which may be as high as 20–35% in both adults and children. Providing phenotyped donor red cells matched specifically for individual patients (for Rh and Kell antigens as a minimum) reduces this risk, but it is a major logistic exercise to ensure availability of appropriate units of donor red cells. The R0 group is more common in Afro-Caribbean patients and such individuals should receive C-negative Enegative blood (which should also be sickle cell trait negative). Providing matched blood may be unnecessary for ‘non-responder’ individuals, who rarely make red cell antibodies, but it is difficult to identify these individuals. Hypertransfusion programmes are used to prevent recurrence of stroke, and may also have a role in preventing deterioration in end-organ failure, e.g. renal, lung. The results of the Stroke Prevention Trial in Sickle Cell Anaemia (STOP) have indicated a role for regular transfusions in the primary prevention of stroke, but uncertainty remains about the duration of transfusion support. Immune blood disorders

In immune haemolytic anemia, antibodies bind to red blood cell surface antigens and initiate destruction via the complement system and/or the macrophage system. Immune haemolytic anemia may be alloimmune, autoimmune or drug induced. Alloimmune haemolytic anaemia occurs in haemolytic disease of the newborn (see Chapter 8), haemolytic transfusion reactions and after allogeneic bone marrow, renal, liver or cardiac transplantation when donor lymphocytes transferred in the allograft (‘passenger lymphocytes’) may produce red cell antibodies against the recipient and cause haemolytic anaemia (see Chapter 13). Autoimmune haemolytic anaemias (AIHAs) are uncommon, with estimates of the incidence at 1–3

per 100 000 per year. They are characterized by the production of antibodies directed against highfrequency red cell antigens, and often exhibit reactivity against donor red cells. The degree of haemolysis depends on a number of factors, including the characteristics of the bound antibody (e.g. class, quantity, specificity, thermal amplitude), the target antigen (e.g. density, expression), and other host-related genetic factors (e.g. markers of macrophage activity). The antibody class in turn will affect the degree of classical complement activation (IgM) or binding to splenic and other tissue macrophages via Fc receptors (IgG1 and IgG3 antibodies). AIHA is divided into ‘warm’ and ‘cold’ types, depending on whether the antibody attaches better to red cells at body temperature (37°C) or at lower temperatures. In warm antibody AIHA, IgG antibodies predominate and the direct antiglobulin test is positive with IgG alone (20%), IgG and complement (67%), or complement only (13%); the red cell autoantibodies usually have Rh specificity. In cold AIHA, the antibodies are usually IgM. They easily elute off red cells, leaving complement, which is detected as C3d. The cause of warm antibody AIHA remains unknown in more than 30% of cases, but may be associated with lymphoid malignancies or diseases such as rheumatoid arthritis and systemic lupus erythematosus or drugs. Therapy of warm antibody AIHA depends on the severity of the haemolysis. Treatment is usually required once symptomatic anaemia develops. Steroids are the first-line treatment (e.g. prednisolone in doses of 1 mg/kg daily) and are effective in inducing a remission in about 80% of patients. Steroids reduce both production of the red cell autoantibody and destruction of antibody-coated cells. Splenectomy may be necessary if there is no response to steroids or if remission is not maintained when the dose of prednisolone is reduced. Other immunosuppressive drugs, such as azathioprine and cyclophosphamide, may be effective in patients who fail to respond to steroids and splenectomy. Blood transfusion may be required if there is fulminant haemolytic anaemia or severe anaemia not responding to steroids or other therapy. The presence of red cell autoantibodies on the patient’s red 121

Chapter 9

cells and in the plasma can cause problems in the identification of compatible blood. It is important to exclude the presence of red cell alloantibodies, and autoabsorption of autoantibodies in the plasma using enzyme treatment of the patient’s red cells may be necessary to permit the investigation of the plasma for alloantibodies (see Chapter 25). Cold antibody AIHA is usually due to IgM antibodies. Normally, low titres of IgM cold agglutinins reacting at 4°C are present in plasma and are harmless. At low temperatures these antibodies can attach to red cells and cause their agglutination in the cold peripheries of the body. In addition, activation of complement may cause intravascular haemolysis when the cells return to the higher temperatures in the core of the body. After certain infections, e.g. Mycoplasma, cytomegalovirus (CMV), Epstein–Barr virus (EBV), there is increased synthesis of polyclonal cold agglutinins producing a mild to moderate transient haemolysis. Chronic cold haemagglutinin disease usually occurs in the elderly, with a gradual onset of haemolytic anaemia owing to the production of monoclonal IgM cold agglutinins, usually with anti-I specificity. After exposure to cold the patient develops an acrocyanosis similar to Raynaud’s as a result of red cell autoagglutination. The underlying cause should be treated, if possible, and patients should avoid exposure to cold. Treatment with steroids, alkylating agents and splenectomy is usually ineffective. Regular blood transfusion is occasionally required to prevent symptoms of anaemia. Paroxysmal cold haemoglobinuria is a rare condition more commonly associated with childhood infections, such as measles, mumps and chickenpox, but was originally described in association with syphilis. Intravascular haemolysis is associated with polyclonal IgG complement-fixing antibodies. These antibodies are biphasic, reacting with red cells in the cold in the peripheral circulation, with lysis occurring due to complement activation when the cells return to the central circulation. The antibodies have specificity for the P red cell antigen. The lytic reaction is demonstrated in vitro by incubating the patient’s red cells 122

and serum at 4°C and then warming the mixture to 37°C (Donath–Landsteiner test). Haemolysis is self-limiting, but supportive transfusions may be necessary. P-negative blood should be considered if there is no sustained response to transfusion of P-positive crossmatch compatible blood. The issue of whether it is necessary to use an inline blood warmer when transfusing patients with cold antibody AIHA is controversial. It is logical to keep the patient warm, and common practice to use a blood warmer if the patient has florid haemolytic anaemia. There is also debate about the need to use washed or leucocyte-depleted blood when transfusing patients with AIHA; it seems reasonable to use leucocyte-depleted blood to avoid febrile reactions that may result in delay in completing the transfusion. Platelet transfusions

In general, platelet transfusions are indicated for the prevention and treatment of haemorrhage in patients with thrombocytopenia or platelet function defects. The cause of the thrombocytopenia should be established before platelet transfusions are used because they are not always appropriate treatment for thrombocytopenic patients, and in some instances are contraindicated, for example in thrombotic thrombocytopenic purpura, haemolytic–uraemic syndrome, and heparininduced thrombocytopenia. Bone marrow failure

Therapeutic platelet transfusions are established as effective treatment for patients who are bleeding. The issue of the benefit of prophylactic platelet transfusions for the prevention of haemorrhage in chronically thrombocytopenic patients with bone marrow failure remains controversial. There have been no recent randomized trials comparing the frequencies of bleeding events and patient survival in patients receiving either prophylactic or therapeutic platelet transfusions. A strategy of transfusing platelets only for therapeutic indications in the context of clinical bleeding is appropriate for some patients with chronic persisting thrombocytopenia due to bone marrow failure syndromes.

Haematological disease

Prophylaxis for invasive procedures depends on the type of procedure. • No increase in platelet count required: bone marrow aspiration and biopsy. • Platelet count should be raised to 50 ¥ 109/L: lumbar puncture, epidural anaesthesia, insertion of intravascular lines, transbronchial and liver biopsy, and laparotomy. • Platelet count should be raised to more than 100 ¥ 109/L: surgery in critical sites such as the brain or the eyes. Immune thrombocytopenias

• Autoimmune thrombocytopenias: platelet transfusions should only be used in patients with major haemorrhage. • Post-transfusion purpura: platelet transfusions are usually ineffective in raising the platelet count, but may be needed in large doses to control severe bleeding in the acute phase (see Chapter 17). • Neonatal alloimmune thrombocytopenia: human platelet antigen (HPA)-matched platelet concentrates are the most appropriate treatment for this condition (see Chapter 8). Massive blood transfusion

• Clinically significant dilutional thrombocytopenia only occurs with the transfusion of more than 1.5 times the blood volume of the recipient. • The platelet count should be maintained above 50 ¥ 109/L in patients receiving transfusions for massive acute blood loss (see Chapter 7). Disseminated intravascular coagulation

• In acute disseminated intravascular coagulation (DIC), where there is bleeding associated with severe thrombocytopenia, platelet transfusions should be given in addition to coagulation factor replacement (see Chapter 11). • In chronic DIC, or in the absence of bleeding, platelet transfusions are not indicated. Cardiopulmonary bypass surgery

• Platelet function defects and some degree of

thrombocytopenia frequently occur after cardiac bypass surgery, but prophylactic platelet transfusions are not indicated. • Platelet transfusions should be reserved for patients with bleeding not due to surgically correctable causes. Granulocyte transfusions

Severe persisting neutropenia is the principal limiting factor in the use of intensive treatment of patients with haematological malignancies. It may last for 2 weeks or more after chemotherapy or bone marrow/peripheral blood progenitor cell transplantation, and during this period the patient is at risk of life-threatening bacterial and fungal infections. The use of haemopoietic growth factors, such as granulocyte colony-stimulating factor (G-CSF), may reduce the duration and severity of severe neutropenia, but they are only effective if the patient has sufficient numbers of haemopoietic precursors. Moreover, the time to response may be several days. Supportive treatment with granulocyte transfusions is a logical approach, although a number of factors have limited its application: • difficulties in the collection of neutrophils, which are present in low numbers in normal individuals and which are difficult to separate from red cells because of their similar densities (commercially available long-chain starch solutions now facilitate this separation); • the short half-life of neutrophils after transfusion, coupled with short storage times and negative effects on function of prolonged storage; • the frequent occurrence of adverse effects such as febrile reactions, including occasional severe pulmonary reactions and human leucocyte antigen (HLA) alloimmunization causing platelet refractoriness. Various methods have been used in the past to increase the number of neutrophils collected, including obtaining granulocytes from patients with chronic myeloid leukaemia, treating donors with steroids, and using hydroxyethyl starch to promote sedimentation of red cells. However, a number of clinical trials of granulocyte transfusions in the 1970s and 1980s suggested they had 123

Chapter 9

limited efficacy in adults, and interest in their usage declined. Some centres continued to use granulocyte transfusions for small children and neonates because concentrates collected from adult donors produced a relatively much greater dose per recipient weight, and sometimes appeared to be clinically effective. There has recently been a resurgence of interest in granulocyte transfusions because of the accumulating evidence that G-CSFs can be safely administered to normal individuals. Much larger doses of granulocytes can then be collected from donors using regimens including G-CSF administered 12–16 h prior to apheresis, together with oral steroids such as dexamethasone to further improve the yields. Further evidence of the safety of this approach for donors, and the efficacy of granulocyte transfusions collected in this way, are required before granulocyte transfusion therapy becomes accepted in the care of patients with severe neutropenia and fungal infection, in conjunction with other potential approaches such as improved diagnostic strategies and organismtargeted antimicrobials. Trials to evaluate evidence of survival benefit following granulocyte transfusions are clearly needed, but their design is complicated by several issues including the numbers of patients required to power a trial and the methodological difficulties related to incorporating blinding in such studies. High-dose granulocyte transfusions collected using donors treated with G-CSFs might therefore be considered as indicated in patients of any age with severe neutropenia due to bone marrow failure under the following circumstances: • proven bacterial or fungal infection unresponsive to antimicrobial therapy, or probable bacterial or fungal infection unresponsive to appropriate blind antimicrobial therapy; • neutrophil recovery not expected for 5– 7 days; • children and lighter adults might be expected to show better incremental responses to granulocyte transfusions. Granulocyte transfusions might be considered inappropriate for: • patients with haematological disease resistant to treatment; 124

• ventilated patients; • patients with known HLA alloimmunization.

Approach to complications associated with blood transfusion in haematology patients Transfusion-transmitted CMV infection Clinical features and risk factors

CMV infection may cause significant morbidity and mortality in immunocompromised patients, mainly due to pneumonia. Patients who have never been exposed to CMV are at risk for primary infection transmitted by blood components prepared from blood donors who have previously had CMV infection and still carry the virus. Patients who have been previously exposed to CMV and are CMV seropositive are at risk of reactivation of CMV during a period of immunosuppression. The extent to which CMV-seropositive patients are at risk from reinfection with different strains of CMV remains unknown, but this risk is generally considered to be low. The patients at risk of transfusion-transmitted CMV infection are shown in Table 9.1, and the generally accepted indications for the use of CMV-seronegative blood components are shown in Table 9.2. Prevention

The use of CMV-seronegative blood components has been shown to reduce the incidence of CMV infection in groups at risk for transfusiontransmitted CMV infection to 1–3%. This incomplete prevention may be due to: • occasional failure to detect low-level CMV antibodies; • loss of antibodies in previously infected blood donors; • transfusion of blood components prepared from recently infected donors. CMV is transmitted by leucocytes, and a number of studies have found that leucocyte depletion of blood components is as effective as the use of CMV-seronegative blood components in the prevention of transfusion-transmitted CMV infection in neonates, patients undergoing remission

Haematological disease Table 9.1 Patients at risk for transfusion-transmitted

cytomegalovirus (CMV) infection. Risk well established CMV-seronegative recipients of allogeneic bone marrow/peripheral blood progenitor cell transplants from CMV-seronegative donors CMV-seronegative pregnant women Premature infants (<1.2 kg) born to CMV-seronegative women CMV-seronegative patients with HIV infection Risk less well established CMV-seronegative patients receiving autologous bone marrow/peripheral blood progenitor cell transplants CMV-seronegative patients who are potential recipients of allogeneic or autologous bone marrow/peripheral blood progenitor cell transplants CMV-seronegative patients receiving solid organ (kidney, heart, lung liver) transplants from CMV-seronegative donors Risk not established CMV-seronegative recipients of allogeneic bone marrow/peripheral blood progenitor cell transplants from CMV-seropositive donors CMV-seropositive recipients of bone marrow/peripheral blood progenitor cell transplants CMV-seropositive recipients of solid organ transplants

Table 9.2 Indications for the use of cytomegalovirus

(CMV)-seronegative blood components. Transfusions in pregnancy Intrauterine transfusions Transfusions to neonates and to infants in the first year of life Transfusions to the following groups of CMV-seronegative patients After allogeneic bone marrow/peripheral blood progenitor cell transplants where the donor is also CMV seronegative After autologous bone marrow/peripheral blood progenitor cell transplants Potential recipients of allogeneic bone marrow/peripheral blood progenitor cell transplants Patients with HIV infection

induction therapy for acute leukaemia and after bone marrow transplantation (the only prospective randomized trial was conducted in transplant recipients using bedside filtration, which cannot be adequately quality controlled). These data suggest that leucocyte-depleted blood components can be accepted as a substitute for CMV-seronegative

blood components for patients at risk of transfusion-transmitted CMV infection when CMV-seronegative blood components are not available. Further information about the effectiveness of leucocyte depletion of blood components in the prevention of transfusion-transmitted CMV infection in different patient groups is required before CMV-seronegative blood components can be discontinued. A recent consensus conference in Canada recommended that where universal leucocyte depletion had been implemented, both leucocyte-depleted and CMV-seronegative blood should be used for CMV-seronegative pregnant women, intrauterine transfusions, and CMV-seronegative allogeneic haemopoietic cell transplant recipients.

Transfusion-associated graft-versus-host disease Pathogenesis and clincal features

Transfusion-associated graft-versus-host disease (TA-GVHD) is a rare but serious complication of blood transfusion. As discussed in Chapter 18, there is engraftment and proliferation of donor T lymphocytes, and interaction with recipient cells expressing HLA antigens causing cellular damage particularly to the skin, gastrointestinal tract, liver and spleen, and the bone marrow. Clinical manifestations usually occur 1–2 weeks after blood transfusion, and early features include fever, maculopapular skin rash, diarrhoea and hepatitis. At-risk transfused haematology patients are those who are undergoing transplantation, have Hodgkin’s disease, or have received therapy with certain drugs, e.g. purine analogues. Prevention

The dose of donor lymphocytes sufficient to cause TA-GVHD is unknown, but may be lower than is achievable by current techniques for leucocyte depletion of blood components, since there are case reports of TA-GVHD following transfusions of leucocyte-depleted blood. Gamma-irradiation to remove the proliferative capability of donor lymphocytes remains the method of choice to prevent TA-GVHD (see Chapter 18). The currently recommended indications for the use of 125

Chapter 9 Table 9.3 Indications for gamma-irradiation of blood

components in haematology patients. Indications Acute leukaemia: only for HLA-matched platelets or donations from first- or second-degree relatives Allogeneic bone marrow/peripheral blood progenitor cell transplantation: from the time of initiation of conditioning therapy and continuing while the patient remains on GVHD prophylaxis (usually 6 months) or until lymphocytes are greater than 1 ¥ 109/L. It may be necessary to irradiate blood components for SCID patients for up to 2 years, and for patients with chronic GVHD if there is evidence of immunosuppression Donors of allogeneic bone marrow: to prevent TA-GVHD mediated by lymphocytes in donor blood transfused before or during the harvest Autologous bone marrow/peripheral blood progenitor cell transplantation: during and 7 days before the harvest of haemopoietic cells, and then from the initiation of conditioning therapy until 3 months post-transplant (6 months if total body irradiation is used) Hodgkin’s disease Patients treated with purine analogues Non-indications Aplastic anaemia (even if treated with antilymphocyte globulin) Non-Hodgkin’s lymphoma (although this may be reviewed following some recent reports of TA-GVHD in patients with B-cell nonHodgkin’s lymphoma) HIV infection SCID, severe combined immunodeficiency; TA-GVHD, transfusion-associated graft-versus-host disease.

gamma-irrradiated blood for patients are shown in Table 9.3.

haematology

How to ensure that patients receive the correct ‘special’ blood?

An important issue for haematology departments and hospital blood banks is how to ensure that patients receive special blood components (e.g. CMV-seronegative, gamma-irradiated) when they are indicated and that standard blood components are not transfused as this may have devastating consequences. Each hospital needs to establish its own procedures so that patients receive the correct special blood components, where they are indicated. These should include the following: • education of ward medical and nursing staff 126

about the indications for special blood components, and the importance of receiving the correct type of blood component; • requests for blood components to include the patient’s diagnosis and any requirement for special blood components; • storing of individual patient’s requirements for special blood components in the blood bank computer; • the prescription for blood components should include any requirement for special blood components, enabling the ward staff to check that the blood component to be transfused complies with these requirements; • providing patients with cards indicating their special blood requirements, particularly for those patients receiving shared care between two hospitals and those with a long-term requirement for gamma-irradiated blood, e.g. patients with Hodgkin’s disease. HLA alloimmunization and refractoriness to platelet transfusions

Platelet refractoriness is the repeated failure to obtain satisfactory responses to platelet transfusions, and occurs in more than 50% of patients receiving multiple transfusions. Various methods are used to assess response to platelet transfusions. If the patient is bleeding, the clinical response is an important indication of the effectiveness of the transfusion. The response to a prophylactic platelet transfusion is assessed by measuring the increase in platelet count after the transfusion. Various formulas have been used to correct for the variation in response dependent on the patient’s size and the number of platelets transfused; these include platelet recovery and corrected count increment. However, in practice, a (nonsustained) increase in the patient’s platelet count of less than 5 ¥ 109/L at 20–24 h after the transfusion can be used as a simple measure of a poor response. Causes

Many causes of platelet refractoriness have been described, and they can be subdivided into

Haematological disease

immune mechanisms, most importantly HLA alloimmunization, and non-immune mechanisms involving platelet consumption (Table 9.4). Platelet consumption is the most frequent mechanism of platelet refractoriness, usually associated with sepsis. However, immune-mediated platelet destruction remains an important cause of platelet refractoriness; HLA antibodies are the commonest immune cause, and the other immune causes are rare. The precise mechanism of HLA alloimmunization remains uncertain, but primary HLA alloimmunization appears to be initiated by intact cells expressing both HLA class I and class II antigens such as lymphocytes and antigen-presenting cells. Platelets only express HLA class I antigens, and leucocyte-depleted blood components will not cause primary HLA alloimunization. However, secondary HLA alloimmunization does not require the presence of HLA class II antigens, and may occur in patients who have been pregnant or previously transfused with non-leucocyte-depleted blood components. Investigation and management

If platelet refractoriness occurs, the following algorithm can be used for investigation and management (Fig. 9.1). 1 A clinical assessment should be made for clinical factors likely to be associated with non-immune platelet consumption. 2 If non-immune platelet consumption appears

likely, an attempt should be made to correct the clinical factors responsible, where possible, and platelet transfusions from random donors should be continued. If poor response to random donor platelet transfusions persists, the patient should be tested for HLA antibodies. 3 If non-immune platelet consumption appears to be unlikely, an immune mechanism should be suspected, and the patient’s serum should be tested for HLA antibodies. If HLA antibodies are present, the specificity of the antibodies should be determined as this may help in the selection of HLAcompatible donors. However, HLA antibodies stimulated by repeated transfusions are often

Table 9.4 Causes of platelet refractoriness.

Immune Platelet alloantibodies HLA HPA ABO Other antibodies Platelet autoantibodies Drug-dependent platelet antibodies Immune complexes Non-immune Infection and its treatment, especially amphotericin B Splenomegaly Disseminated intravascular coagulation Fever Bleeding

Poor response to random donor platelets (e.g. 24-h platelet increment <5 x 109/L with two or more consecutive transfusions) Clincal evaluation

For presence of infection, DIC, splenomegaly

Positive clinical factors

Fig. 9.1 Algorithm for the

investigation and management of patients with platelet refractoriness. DIC, disseminated intravascular coagulation.

Continue with random platelets Responses to platelet transfusions should improve when clinical factors resolve

Negativeclinical factors

HLA antibodies present Use HLA-compatible platelet transfusions

No HLA antibodies

1 Consider ABO incompatibility 2 Test for non-cytotoxic HLA antibodies and HPA antibodies 3 Ifall negative, consider causes such as drug-dependent and platelet autoantibodies

127

Chapter 9

‘multispecific’, and it is not possible to determine their specificity. 4 Platelet transfusions from HLA-matched donors (matched for the HLA-A, -B antigens of the patient) should be used for patients with apparent immune refractoriness, and the response to further transfusions should be observed carefully. Figure 9.2 shows improved responses to HLA-matched platelet transfusions in a patient with platelet refractoriness due to HLA alloimmunization. If responses to HLA-matched transfusions are not improved, the reason should be sought, and platelet crossmatching of the patient’s serum against the lymphocytes and platelets of one of the HLA-matched donors may be helpful in determining the cause, and the selection of compatible donors for future transfusions. 5 If there are no factors for non-immune platelet consumption and HLA antibodies are not detected, consideration should be given to less frequent causes of immune platelet refractoriness. (a) High-titre ABO antibodies in the recipient. This is an unusual cause of platelet refractoriness, and can be excluded by switching to ABO-compatible platelet transfusions, if ABO-

% platelet recovery (20 h after platelet transfusions)

100

Random donors

HLA-matched donors

80 60 40 20 0 0

20

10

30

Days Fig. 9.2 Responses to platelet transfusions in a female

patient with acute myeloblastic leukaemia undergoing remission induction therapy. There were poor responses to the initial platelet transfusions, and the patient was found to have HLA antibodies. There were improved responses to platelet transfusions from HLA-matched donors.

128

incompatible transfusions have been used for previous transfusions. (b) HPA antibodies, which usually occur in combination with HLA antibodies, but sometimes occur in isolation. (c) Drug-dependent platelet antibodies, which may be underestimated as a cause for platelet refractoriness. Alloimmunization to red cell antigens Incidence

Alloimmunization to red cell antigens is another important consequence of repeated transfusions in haematology patients. The incidence of red cell alloimmunization in adult haematology patients is in the range of 10–15% and is similar to other groups of multitransfused patients, e.g. patients with renal failure. However, a higher proportion of children requiring long-term transfusion support develop red cell alloimmunization. In sickle cell disease, the incidence is in the range of 20–30%. The implications of these observations include the following. • Patients with sickle cell disease should be phenotyped for Rh, Kell, Fy, Jk and MNS antigens before the first transfusion, and patients with thalassaemia and other children requiring chronic transfusion support should be phenotyped for Rh and Kell antigens. • Blood for transfusion to children requiring longterm transfusion support, including patients with haemoglobinopathies, should be matched for Rh and Kell antigens to prevent alloimmunization. • Phenotyping and antigen matching to prevent red cell alloimmunization is not required for other groups of patients requiring repeated transfusions. Timing of sample collection for compatibility testing

In patients with haematological disorders receiving repeated transfusions, an important issue is the timing of blood sample collection in relation to the previous transfusion. • Where the patient is receiving very frequent transfusion, e.g. daily, it is only necessary to request a new sample every 72 h.

Haematological disease

• Where the previous transfusion was 3–14 days earlier, the sample should ideally be taken within 24 h of the start of the transfusion, although some laboratories stretch this to 48 h for patients who have been repeatedly transfused without developing antibodies. • Where the previous transfusion was 14–28 days earlier, the sample should be taken within 72 h of the start of the transfusion. • Where the previous transfusion was more than 28 days ago, the sample should be taken within 1 week of the planned transfusion. ABO-incompatible bone marrow/peripheral blood progenitor cell transplants

ABO-incompatible bone marrow/peripheral blood progenitor cell transplants present particular problems (Table 9.5). The transplant may provide a new A and/or B antigen from the donor (major mismatch) or a new A and/or B antibody (minor mismatch). Recommendations for the use of donor

blood components are given in Fig. 9.3, and can be briefly summarized as follows. • Major ABO mismatch: use red cells of patient’s ABO type until recipient ABO antibodies are undetectable and the direct antiglobulin test is negative, and platelets and plasma from donors of recipient’s ABO type. • Minor ABO mismatch: use red cells of donor ABO type throughout, and plasma and platelets of recipient type until recipient-type red cells are no longer detectable. • Major and minor ABO mismatch: use group O red cells until recipient ABO antibodies are undetectable, and then switch to donor-type red cells. Use group AB plasma and platelets until recipienttype red cells are undetectable. RhD-incompatible transplants can also cause difficulties. It is recommended that RhD-negative blood components should be used for RhDpositive recipients with RhD-negative donors. However, no cases of immunization have been reported when RhD-negative recipients have received RhD-positive transplants, and RhDpositive blood components may be used.

Table 9.5 Problems associated with ABO-incompatible

bone marrow/peripheral blood progenitor transplants. Major ABO incompatibility (e.g. recipient O, donor A) Failure of engraftment: risk not increased in ABO-incompatible transplants Acute haemolysis at time of reinfusion: avoided by processing donor bone marrow/peripheral blood progenitor cells Haemolysis of donor-type red cells: avoid by using red cells of recipient type in the early post-transplant period Delayed erythropoiesis: may be due to persistence of anti-A in the recipient, but minimize transfusion of anti-A by using platelets and plasma from group A donors Delayed haemolysis due to persistence of recipient anti-A: only switch to donor red cells when recipient anti-A undetectable and direct antiglobulin test undetectable Minor ABO incompatibility (e.g. recipient A, donor O) Graft-versus-host disease: risk not increased in ABO-incompatible transplants Acute haemolysis at time of reinfusion: avoid by removing donor plasma if the donor anti-A titre is high Delayed haemolysis of recipient cells due to anti-A produced by donor lymphocytes (passenger lymphocyte syndrome): maximum haemolysis usually occurs between days 9 and 16 post-transplant, and occasionally there is severe intravascular haemolysis

Iron overload

A major adverse consequence of repeated red cell transfusions over a long period is iron overload. The body has no mechanism for excreting excess iron, and regular transfusion inevitably leads to accummulation of iron. Each unit of blood contains about 200–250 mg of iron, and clinical symptoms and signs of iron overload do not generally occur until iron stores are in the range of 20–30 g, i.e. about 100 units of blood transfused. The clinical picture of iron overload resembles that of idiopathic haemochromatosis, and the organs mainly affected include: • skin (causing pigmentation); • endocrine glands (causing diabetes mellitus, hypogonadism, poor growth, hypothyroidism); • liver (causing cirrhosis); • heart (causing heart failure). The haematology patient group at greatest risk for iron overload is thalassaemia major, but patients with other inherited haematological diseases such as sickle cell disease or with acquired 129

Chapter 9

Recipient group

Group O

Group AB

Donor group

Major ABO incompatibility Red cells Plasma/platelets Minor ABO incompatibility Red cells Plasma/platelets Major and minor ABO incompatibility Red cells Plasma/platelets 1

2

3

4

1 Begin pretransplant chemotherapy 2 Bone marrow transplant

Fig. 9.3 Recommendations for ABO

3 ABO antibodies to donor RBC not detected. Direct antiglobulin test negative 4 RBC of recipient group no longer detected

haematological disorders such as myelodysplasia may also require long-term red cell transfusions and are at risk. Iron chelation therapy is the only way to remove excess iron in haematology patients and prevent heart failure, which is the major cause of death in transfusional iron overload. Venesection, as used in inherited haemochromatosis, is inappropriate for most patients, but has been occasionally used for some sickle cell patients with iron overload. Only a small proportion of the body iron is available for chelation. Parenteral desferrioxamine is established for promoting negative iron balance, reversing cardiac toxicity and prolonging life expectancy. Subcutaneous desferrioxamine (30–50 mg/kg) is typically administered as an infusion using a syringe driver pump over 8–12 h (or longer if tolerated to sustain the chelation process) at least five times a week, usually overnight. However, the efficiency of currently used chelation regimens is generally very low, and most of the administered subcutaneous desferrioxamine is excreted without binding to iron atoms. 130

type of blood components in ABOincompatible bone marrow/peripheral blood progenitor cell transplants. (From Warkentin 1983 with permission.)

Vitamin C (200 mg adults, 100 mg/day children or orange juice) is given to increase the availability and hence excretion of chelatable iron, starting 2 weeks after iron chelation has begun. New systems of delivering desferrioxamine are being developed because of the recognized problems of compliance, including the lightweight silent balloon infusor which provides continuous pressure without the need for a battery or other mechanical device. Administration of intravenous desferrioxamine only at the time of blood transfusion is an ineffective approach to iron chelation, but may be indicated as a continuous infusion (e.g. through Port-a-Cath) for patients with consistently high levels of iron overload. It should be recalled that sickle cell (and to a degree) thalassaemia patients requiring Port-a-Cath have a prothrombotic predisposition. The decision when to initiate iron chelation therapy is not always easy, particularly in elderly patients with acquired disorders such as myelodysplasia where the patient’s life expectancy from their underlying condition may be less than the

Haematological disease

time it would take to develop clinically significant iron overload. The serum ferritin level is the simplest way of monitoring iron overload, but assay of iron stores using liver biopsies is a more reliable measure. Newer imaging techniques may provide this information non-invasively in the future. Iron chelation therapy should aim to maintain the liver iron concentration at about 3–7 mg/g liver dry weight, or in practice a serum ferritin level of 1000–1500 mg/L. Desferrioxamine may cause adverse effects such as local swelling and irritation at the sites of infusion. Excessive doses of desferrioxamine may cause: • disturbances of vision and hearing, and therefore regular (yearly) audiograms and retinal examination are recommended; • bone deformities and growth retardation in children. Long-term effects of desferrioxamine can largely be avoided by adjusting the dose to take account of the degree of iron overload, ensuring that the mean daily dose does not exceed 40–50 mg/kg (until growth is completed), or by calculating a therapeutic index of the mean daily dose of desferrioxamine (mg/kg) divided by the serum ferritin (mg/L) every 6 months and keeping this ratio below 0.025. Difficulties of compliance with infusion of desferrioxamine for practical and psychological reasons remain major issues for many patients and have stimulated an ongoing search for alternatives; the ideal would be an effective and safe iron chelator that can be taken orally.

Further reading British Committee for Standards in Haematology. Guidelines on gamma irradiation of blood components for the prevention of graft-versus-host disease. Transfus Med 1996; 6: 261–71. British Committee for Standards in Haematology. Guidelines on the clinical use of red cell transfusions. Br J Haematol 2001; 113: 24–31. British Committee for Standards in Haematology. Guidelines for platelet transfusions. Br J Haematol 2003; 122: 10–23. British Committee for Standards in Haematology. Transfusion guidelines for neonates and small children. Br J Haematol 2004; 124: 433–53. Dale DC, Conrad Liles W, Price TH. Renewed interest in granulocyte transfusion therapy. Br J Haematol 1997; 98: 497–501. Laupacis A, Brown J, Costello B et al. Prevention of posttransfusion CMV in the era of universal WBC reduction: a consensus statement. Transfusion 2001; 41: 560–9. Novotny VMJ. Prevention and management of platelet transfusion refractoriness. Vox Sang 1999; 76: 1–13. Oliveri NF, Brittenham GM. Iron-chelating therapy and the treatment of thalassaemia. Blood 1997; 89: 739–61. Pamphilon DH, Rider JR, Barbara JAJ, Williamson LM. Prevention of transfusion-transmitted cytomegalovirus infection. Transfus Med 1999; 9: 115–23. Petz LD. Hemolysis associated with transplantation. Transfusion 1998; 38: 224–8. Porter JB. Practical management of iron overload. Br J Haematol 2001; 115: 239–52. Samol J, Littlewood TJ. The efficacy of rHuEPO in cancerrelated anaemia. Br J Haematol 2003; 121: 3–11. Warkentin PI. Transfusion of patients undergoing bone marrow transplantation. Hum Pathol 1983; 14: 261–6.

131

Chapter 10

Transfusion strategies in organ transplant patients Derwood H. Pamphilon

Transplant patients frequently require blood component transfusions. These may be given preoperatively or perioperatively. Blood transfusions are an important consideration for transplant recipients because: • they may sensitize potential transplant recipients to histocompatibility antigens (HLA), resulting in an increased risk of graft rejection; • induce modulation of immune responses and promote graft acceptance; • cause the acquisition of viruses such as hepatitis C (HCV) and cytomegalovirus (CMV). Blood component order schedules for orthotopic liver transplantation (OLT) place large demands on transfusion services, while in kidney transplants perioperative blood transfusions are often not required. In some types of transplant surgery, autologous predeposit, as well as intraoperative blood salvage and the use of pharmacological agents to minimize the use of allogeneic products, may be considered.

Renal transplantation Pretransplant period

Patients with renal disease are frequently anaemic and receive transfusions preoperatively. In the case of transplant from cadaveric donors, third-party blood transfusions are given. In addition, patients who receive live donor transplants from siblings or other relatives may be given transfusions from the potential donor (see below). These are known as donor-specific transfusions (DST). In 1973 it was reported that graft survival was better in transfused patients irrespective of matching at HLA-A, -B or -DR, although the benefit of transfu132

sion was greatest when there was more HLA mismatch. More recently it has been appreciated that this ‘transfusion effect’ is less apparent when cyclosporin is used after grafting for immunosuppression. It has been argued that pretransplant transfusions may sensitize a proportion of patients and that the benefit ascribed to transfusions could then be due to identification of non-responders and high responders. The latter would either be excluded or transplanted with a kidney from a crossmatch-negative donor. In 150 patients treated with post-transplant cyclosporin, graft survival was not related to a history of blood transfusion but the presence of lymphocytotoxic antibodies was a major risk factor for graft failure. In a large study of 4015 children receiving either live donor or cadaveric transplants reported by the North American Paediatric Renal Transplant Cooperative Study, it was shown that patients who received between one and five transfusions had a reduced risk of graft failure. Blood transfusion may reduce the rate of graft failure by downregulating the immune response. This could be due to: • generation of suppressor cells; • reduction in the number of cytotoxic T lymphocytes reactive with donor-type HLA; or • formation of anti-idiotype antibodies. There is debate as to whether the impact of pretransplant transfusion is greater if there is HLA haplotype sharing between transfused blood and the patient. It has been shown that the frequency of anti-donor cytotoxic T lymphocytes was markedly reduced in patients who received transfusions where one HLA haplotype or at least one HLA-B and -DR was shared.

Organ transplantation

DST is thought to induce a state of immunological tolerance between donor and recipient, perhaps as a result of persistence of donor cells after transfusion (see below). In DST, between 8 and 13% of patients are sensitized after transfusion, fewer if concurrent treatment with immunosuppressive drugs is given. Graft and patient survival of 90% at 5 years following DST has been reported where combinations of immunosuppressive drugs were given after transplantation. In patients who became sensitized to third-party DST, the antibody can be removed and the patient ‘desensitized’ by plasma exchange procedures or immunoadsorption. Although these patients may have a lower overall rate of durable renal engraftment, desensitization at least allows transplants to be done in patients who would otherwise be excluded as lymphocytotoxic crossmatch positive. Recently, it has been shown that transfusion of donor bone marrow cells to patients who receive kidney and other organ transplants is associated with a lower incidence of acute cellular graft rejection and lowered reactivity against donor cells in the mixed lymphocyte reaction. This also suggests that transfusion of donor cells before transplantation may induce a state of specific tolerance. This could be partly due to the persistence of donor cells, which may be detected by sensitive molecular analysis of chimeric status. Recombinant human erythropoietin (EPO) is now used extensively in the preoperative period in renal transplant recipients. It reduces the degree of anaemia and therefore the chance of sensitization induced by random blood transfusions. It is generally recommended that all renal transplant patients who require transfusion, despite the use of EPO, should receive leucocyte-depleted blood components before transplantation to avoid HLA alloimmunization. In the UK, this presents no difficulties as all blood components have been routinely leucocyte depleted since 1999 (see Chapter 23). However, one or more nonleucocyte-depleted transfusions may be given deliberately to induce tolerance before transplantation.

Viral infections Cytomegalovirus

CMV is of importance after renal transplantation and can be acquired from: • the donor kidney if seropositive; • CMV-positive blood transfusions; • reactivation of the patient’s CMV. A review of 1145 patients showed active infection in 85% of those with, and 53% of those without, CMV antibodies at the time of transplant, suggesting that latent virus is reactivated during the transplant process. Reactivation itself does not appear to be detrimental to graft survival. If CMVseronegative blood is given to CMV-seronegative donor–recipient pairs, then no CMV infection results. Primary infection transferred with the donor kidney is of greater severity than infection associated with reactivation. In one report, 26 of 74 patients who received CMV-seronegative blood and a kidney from a donor of unknown CMV serostatus developed primary infection, 20 of 26 were symptomatic, and three died. CMV infection can be prevented in CMVseronegative donor–recipients by: • provision of CMV-seronegative blood components; or • leucocyte depletion to less than 5 ¥ 106 leucocytes/transfused component. Either approach is satisfactory, although the use of leucocyte-depleted blood components will reduce the rate of HLA alloimmunization as well. Moreover, in several European countries universal leucocyte depletion of blood components is now standard practice. In ‘at-risk’ transplants, where either the patient or donor is CMV seropositive, administration of ganciclovir or valaciclovir for 3 months is effective in preventing CMV disease. Intravenous immunoglobulin from CMV hyperimmune globulin (CMVIg) may reduce the chance of CMV infection but its role is controversial. Hepatitis viruses

HCV, hepatitis B (HBV) and hepatitis G (HGV, also called GBV-C) may be acquired from blood transfusion. An increased incidence of post133

Chapter 10

transplant liver disease has been reported in HCVseropositive patients but this does not appear to influence either the graft or patient survival. Perioperative period

The use of EPO has had a significant impact on preoperative anaemia and at our institution blood is not routinely crossmatched for the transplant procedure unless the haemoglobin is below 9 g/dL. As transfusions are infrequently required during surgery, the use of autologous predeposit, intraoperative blood salvage or pharmacological agents to reduce blood loss is not appropriate.

Orthotopic liver transplantation Pretransplant period

Precise HLA matching between donor and recipient does not influence the outcome in liver transplantation. However, if the lymphocytotoxic crossmatch against donor cells is positive (possibly as a result of previous transfusion), it is associated with: • poorer graft survival; • higher retransplant rate; • reduced 1-year overall survival. The presence of lymphocytotoxic antibodies is not necessarily associated with an increased chance of a positive lymphocytotoxic crossmatch and patients who receive transfusions more than 90 days preoperatively have a reduced incidence of severe or recurrent rejection episodes. This is not reflected in an improvement in allograft or overall patient survival. However, there is considerable interest in induction of donor tolerance pretransplant and it is now appreciated that the persistence of transfused leucocytes (donor microchimerism) may have a clinical role in suppressing host immune responses. Efforts have been made to augment this with preoperative or perioperative donor bone marrow infusion. This can be done without inducing clinical graft-versus-host disease in many types of whole organ transplant recipients.

134

Viral infections Cytomegalovirus

Liver transplant recipients who are CMV seronegative may acquire infection via a seropositive graft or via blood transfusions. The former route seems to be of greater importance. Recipients of CMVseropositive grafts are more likely to have: • CMV pneumonia; • CMV hepatitis; • invasive fungal infections; • reduced survival. The incidence of CMV pneumonia has been shown to correlate with both the total number of units of blood and the number of CMV-positive units transfused perioperatively. CMV infection is also associated with an increased chance of graft rejection. CMV infection may also occur because of reactivation in CMV-positive recipients. It appears that seroconversion to human herpes virus (HHV)-6 is a marker for CMV disease. CMV infection may be prevented in seronegative donor–recipient pairs by transfusion of leucocyte-depleted or CMV-seronegative blood components and this seems to be a sensible precaution. In patients at risk of CMV infection, use of CMVIg has been shown to reduce severe CMVassociated disease and improve long-term survival. In addition, the use of prophylactic ganciclovir in at-risk patients significantly reduces the incidence of CMV disease. Hepatitis viruses

• HCV: a number of patients with HCV infection develop severe liver disease and undergo OLT. The overall outcome in this group of patients is excellent, with survival as high as 75% at 5 years after transplantation. Surprisingly, grafts from HCVseropositive donors do not result in decreased survival when compared with grafts from HCVseronegative donors and there is no difference in the rate of graft rejection. • HGV is also acquired by blood transfusion. It is present in a large number of patients after transplantation compared with before transplantation. It appears not to be associated with significant

Organ transplantation

liver disease or to impact the outcome of transplantation. Perioperative period Blood compatibility

• It is important in OLT to ensure that there is ABO compatibility between donor and recipient. Mismatching for ABO results in an increase in hyperacute rejection, vascular thrombosis and biliary injury. It may be appropriate in some instances to use ABO-mismatched OLT in patients with fulminant liver failure, where a matched liver is unavailable. • Where there is ABO mismatch, viable lymphocytes in the graft may induce clinically significant immune-mediated haemolysis in recipient red cells. Usually this occurs where the donor is group O and the patient is group A. • Because of preceding blood transfusions OLT patients have a significant incidence of red cell alloantibody formation (6.3% in one series) and this adds to the difficulty of providing large inventories of compatible red cells. Blood use

OLT patients usually have disordered coagulation before transplantation due to their underlying liver disease and associated portal hypertension. Frequent findings are prolonged prothrombin and activated partial thromboplastin times and thrombocytopenia. There is activation of fibrinolysis during surgery, which causes increased blood loss. In the first 100 patients transplanted at the Mayo Clinic the average number of units of red cells, fresh frozen plasma (FFP), cryoprecipitate and platelets transfused was over 10. In addition, a mean of 5.6 units of blood were salvaged intraoperatively. Some centres prepare modified whole blood by separating cryoprecipitate and platelets and returning the plasma to the red cells. The aim of this is to reduce donor exposure. In one centre the unit standing order for components in low-risk adult and paediatric OLT was red cells 10 units,

FFP 10 units, cryoprecipitate 10 units and platelets 12 adult therapeutic doses. Usage of blood components is summarized in Table 10.1. FFP treated with solvent–detergent to inactivate viruses is as effective as standard FFP in OLT. Intraoperative blood salvaging may significantly reduce the demand for allogeneic blood components. Massive blood loss during OLT is significantly associated with previous upper abdominal surgery. It results in poorer graft function in the immediate postoperative period, an increased incidence of infection, gastrointestinal and intraabdominal complications, rejection and lower overall survival. Pharmacological agents

These are discussed in more detail in Chapter 7. • Aprotinin: both standard and high doses have been shown to inhibit fibrinolysis, thereby reducing the requirement for red cells, FFP, cryoprecipitate and platelet concentrates. • Tranexamic acid (an antifibrinolytic drug) in high but not low dose has been shown to reduce intraoperative blood loss. Low-dose tranexamic acid does inhibit fibrinolysis but is not sufficient to influence transfusion requirements. • e-Aminocaproic acid: in one report where this was given to patients there was no significant difference in blood loss and blood requirement compared with those patients who did not receive it.

Table 10.1 Transfusion support in liver transplantation.

Report

Red cells

FFP

Cryoprecipitate

Platelets

1 2 3

12.9 24.5 25.0

13.8 38.7 24.0

17.4 12.2 9.0

16.9* 26.2† 20.0‡

Figures given are mean number of units transfused. * Apheresis platelet units. † Random donor platelet units. ‡ Type of platelets not stated.

135

Chapter 10

Cardiac transplantation Pretransplant period

There is some evidence that pretransplant transfusions may induce tolerance and improve the outcome of transplantation. This is poorer if there is a positive lymphocytotoxic crossmatch or disparity at HLA-DR, but not HLA-A or -B. • Administration of pretransplant transfusions where HLA-DR is shared results in better overall outcome and reduced vascular graft rejection. • DST has, in certain strain combinations in animals, been shown to downregulate immune responses and improve graft acceptance. • A detailed analysis in human cardiac allografting showed that the mixed lymphocyte culture response to donor-type cells was significantly reduced at 2 years after transplantation compared with pretransplant testing. Responses to thirdparty lymphocytes were still intact. As in renal and liver transplantation (see above), preoperative or perioperative infusion of donor bone marrow statistically reduces the chance of acute cellular rejection when compared with control patients, and responses in mixed lymphocyte culture are reduced. Finally, where donor HLA antibodies could be effectively removed by plasma exchange before transplantation, survival at 1 year was 87% compared with 25% of patients in whom plasma exchange was ineffective in reducing donor antibodies.

Viral infection Cytomegalovirus

CMV infection adversely affects the outcome of cardiac allografting. In 91 of 301 patients who developed CMV infection at one centre over an 8year period there was an increase in: • graft rejection; • development of atherosclerosis resulting in decreased survival. CMV infection may, as indicated above, be prevented by provision of: • CMV-seronegative blood components; • leucocyte-depleted blood components containing less than 5 ¥ 106 leucocytes/transfusion. 136

There is also evidence that reperfusion injury may be reduced by the use of leucocyte-depleted blood components. Perioperative period Blood losses and blood order schedules

The requirement for blood components is much lower than in OLT. Typical requirements are shown in Table 10.2. Heart–lung patients require more transfusions as shown. Pharmacological agents

Aprotinin has been shown to reduce bleeding and blood component requirement after both primary and reoperative cardiac surgery (see also Chapter 7). Autologous blood

It has been shown that the haemodynamic status in patients with cardiac or pulmonary disease who were candidates for cardiac transplants was comparable with controls following autologous blood donation. This may therefore be an appropriate strategy to reduce the requirement for allogeneic blood products during surgery. Blood collection could be augmented by the use of EPO, although this is expensive. For a more detailed discussion of autologous transfusion strategies see Chapter 26. Platelet concentrates may be collected immediately preoperatively in the anaesthetic room, using an apheresis machine, and infused during surgery. There is evidence that this will reduce the requirement for allogeneic platelets and may also impact Table 10.2 Transfusion support in cardiac transplantation.

Report

Red cells

FFP

Platelets*

1 2 3†

3.9 5.0 8.0

3.6 4.0 4.0

4.6 None 5.0

Figures given are mean number of units transfused. * Type of platelet component not stated. † Heart–lung transplants.

Organ transplantation

on transfusion requirements since the platelets themselves are fresh.

Further reading Blajchman MA, Singal DP. The role of red blood cell antigens, histocompatibility antigens, and blood transfusions on renal allograft survival. Transfus Med Rev 1998; 3: 171–9. Chavers B, Sullivan EK, Tejani A, Harmon WE. Pretransplant blood transfusion and renal allograft outcome: a report of the North American Paediatric Renal Transplant Cooperative Study. Pediatr Transplant 1997; 1: 22–8. Deeg HJ, Sayers MH. Transfusion support in transplant patients. In: Pamphilon DH, ed. Modern Transfusion Medicine. Boca Raton, FL: CRC Press, 1995: 177–92. Falagas ME, Snydman DR, Ruthazer R et al. Cytomegalovirus immune globulin (CMVIG) prophylaxis is associated with increased survival after orthotopic liver transplantation. The Boston Center for Liver Transplantation CMVIG Study Group. Clin Transpl 1997; 11: 432–7. Gratton MR, Moreno-Cabral CE, Starnes VA, Oyer PF, Stinson EB, Shumway ME. Cytomegalovirus and its association with cardiac allograft rejection and atherosclerosis. J Am Med Assoc 1998; 261: 3561–6. Kanj SA, Sharara AI, Clavien P-A, Hamilton, JD. Cytomegalovirus infection following liver

transplantation: review of the literature. Clin Infect Dis 1996; 22: 537–49. Koneru B, Harrison D, Rizwan M et al. Blood transfusions in liver recipients: a conundrum or a clear benefit in the cyclosporin/tacrolimus era? Transplantation 1997; 15: 1587–90. McCarthy JF, Cook DJ, Massad MG et al. Vascular rejection post heart transplantation is associated with positive flow cytometric cross-matching Eur J Cardiothorac Surg 1998; 14: 197–200. Opelz G, Senger DPS, Mickey MR, Terasaki PI. The effect of a blood transfusion on subsequent kidney transplantation. Transplant Proc 1973; 5: 253–9. Palomo Sanchez JC, Jimenez C, Moreno Gonzalez E et al. Effects of intraoperative blood transfusion on postoperative complications and survival after orthotopic liver transplantation. Hepatogastroenterology 1998; 45: 1026–33. Salgar SK, Shapiro R, Dodson F et al. Infusion of donor leucocytes to induce tolerance in organ allograft recipients. J Leuk Biol 1999; 66: 310–14. Scudamore CH, Randall TE, Jewesson PJ et al. Aprotinin reduces the need for blood products during liver transplantation. Am J Surg 1995; 169: 546–9. van Twuyver E, Mooijaart RJD, ten Berge IJM et al. Pretransplantation blood transfusion revisited. N Engl J Med 1991; 325: 1210–13. Wallington TB. Cytomegalovirus and transfusion. In: Cash JD, ed. Progress in Transfusion Medicine. London: Churchill Livingstone, 1987: 26–45.

137

Chapter 11

Inherited and acquired coagulation disorders Joanne E. Joseph and Samuel J. Machin

Blood components have long been used to treat both inherited and acquired disorders of coagulation. Attempts are continually being made to produce as ‘safe’ a product as possible, while providing specific concentrates of the highest purity. Advances in recombinant DNA technology have expanded the types of products available for clinical use. With so many types of product now available, it can be difficult deciding which is the most appropriate to use. This chapter deals with the most common disorders of coagulation (both inherited and acquired) as well as detailing the most appropriate therapies for each condition. The use of platelet concentrates will not be specifically addressed in this chapter (see Chapter 9).

Normal haemostasis Haemostasis is a complex process involving the interaction of many components: blood vessels, platelets, coagulation factors, coagulation factor inhibitors and fibrinolytic enzymes. The procoagulant cascade (Fig. 11.1) is activated when tissue factor (TF) expressed on damaged or stimulated cells (vascular cells or monocytes) comes in contact with circulating factor VII and VIIa (which accounts for approximately 1–2% of circulating plasma factor VII). This TF–factor VIIa complex activates limited quantities of factors IX and X. Newly generated factor IXa forms a complex with factor VIIIa (activated by traces of thrombin generated slowly by factor Xa) in the presence of calcium and membrane phospholipid. This complex subsequently also activates factor X to Xa, and is known as 138

tenase. Factor Xa binds to factor Va (again activated by thrombin) which, with calcium and phospholipid, rapidly converts prothrombin to thrombin. The initial TF–VIIa complex is quickly inhibited by the TF pathway inhibitor; however, by this time, the thrombin that has already been produced activates factor XI as well as factors V and VIII, therefore augmenting the formation of factor Xa and ultimately the production of more thrombin. Factor XI can also be activated by factor XIIa, formed from the high-molecular-weight kininogen–prekallikrein complex on endothelial cells; however, this contribution to physiological haemostasis is minimal. The ultimate function of thrombin is to cleave fibrinogen to fibrin and activate factor XIII that results in the cross-linked stable clot. Fibrinolysis is also part of the normal haemostatic response. Circulating plasminogen is activated to form the serine protease plasmin, which digests cross-linked fibrin to form D-dimers and other fibrinogen fragments.

Investigation of abnormal haemostasis A careful clinical history and physical examination should be undertaken in order to differentiate between bleeding caused by a local factor and that due to an underlying haemostatic defect. Continued oozing from venepuncture and injection sites or from wound drains suggests the possibility of generalized haemostatic failure. Initially, some simple ‘screening’ laboratory tests that are easy to perform and which give quick reliable results should be undertaken (Table 11.1).

Coagulation disorders IX XIa TF

Ca2+

VII/VIIa

IXa PL, Ca2+, VIIIa

TF/VIIa Ca2+ X

X

Xa PL, Ca2+

Va

Prothrombin

Thrombin

Fig. 11.1 The procoagulant pathway.

PL, phospholipid; TF, tissue factor.

Table 11.1 Simple laboratory haemostasis screening tests.

Coagulation Prothrombin time (PT) Activated partial thromboplastin time (APTT) International normalized ratio (INR): only in patients receiving oral anticoagulation Thrombin time (TT) Fibrinogen assay Platelets Platelet count Blood film inspection Platelet function (using PFA-100, which measures in vitro ‘high shear’ bleeding time) Fibrinolysis D-dimers Euglobulin clot lysis time Global haemostasis Thromboelastogram

If one or more of these tests suggests an abnormality, then further specialized investigations (such as specific coagulation factor assays) should be performed in order to define precisely the defect and its severity.

Fibrinogen

Fibrin

In high-dependency units, the availability of near-patient testing devices to rapidly assess coagulation, e.g. prothrombin time (PT) and activated partial thromboplastin time (APTT), and overall global haemostasis screening (thromboelastogram) allows rapid treatment decisions to be made without sending a citrated sample to the laboratory. It is important to note that in some disorders, such as mild forms of haemophilia or von Willebrand disease (vWD), ‘screening’ tests such as APTT may not be overly prolonged, and hence if a bleeding disorder is strongly suspected from the patient’s history and clinical picture, specific factor assays and/or immunological tests should be performed regardless of the ‘screening’ test result.

Transfusion support for patients with acquired haemostatic defects Disseminated intravascular coagulation

Disseminated intravascular coagulation (DIC) is a disorder resulting from inappropriate and excessive activation of the haemostatic system that can be manifested by both thrombotic and haemor139

Chapter 11

rhagic pathology. DIC may be acute (uncompensated), with decreased levels of haemostatic components, or chronic (compensated), with normal or sometimes elevated levels of coagulation factors. The main triggering mechanism for DIC is the exposure of blood to a source of TF that initiates coagulation. This can occur as a result of the following: • synthesis of TF on the surface of endothelial cells or monocytes stimulated by endotoxins and cytokines as a result of sepsis; • release or exposure of TF as a result of direct tissue injury (as in placental abruption, cerebral trauma) or from malignant cells; • snake venoms may cause DIC as a result of direct activation of coagulation factors such as factor X or prothrombin. The final consequence of coagulation activation is thrombin generation and fibrin formation, which may result in microthrombus formation (e.g. gangrene of fingers or toes, renal failure). Following intravascular thrombosis, secondary activation of the fibrinolytic pathway occurs with subsequent lysis of fibrin and the formation of cross-linked complexes such as D-dimers, which can be detected by a number of assays. Raised levels of these fibrin degradation products further add to the bleeding diathesis as they inhibit the action of thrombin and also inhibit platelet function by binding to the platelet membrane. Due to ongoing activation of the coagulation cascade, hepatic synthesis of coagulation factors is unable to fully compensate for their consumption, and so there is a reduction in levels of all coagulation factors, but particularly factors V, VIII and XIII and fibrinogen. The bone marrow is unable to maintain a normal platelet count, and thrombocytopenia eventuates. This combination of coagulation factor deficiency, thrombocytopenia and the inhibitory actions of raised fibrin degradation products causes the generalized and continued bleeding tendency characteristic of DIC. These events are summarized in Fig. 11.2. The main causes of DIC are listed in Table 11.2. Laboratory abnormalities seen in DIC include: • prolonged thrombin time (TT), variably prolonged PT and APTT; 140

Table 11.2 Main causes of disseminated intravascular

coagulation. Infection Septicaemia (~60% of all cases) Viraemia Malignancy Leukaemia (especially acute promyelocytic) Metastatic carcinomas Obstetric disorders Septic abortion Abruptio placentae Eclampsia Amniotic fluid embolism Placenta praevia Shock Extensive surgical trauma Burns Heat stroke Liver disease Cirrhosis Acute hepatic necrosis Transplantation Tissue rejection Extracorporeal circulation Cardiac bypass surgery Extensive intravascular haemolysis ABO-incompatible transfusion Certain snake bites

• reduction of fibrinogen levels, increased levels of D-dimers; • thrombocytopenia; • anaemia, fragmented red cells, raised reticulocyte count. The most important aspect of management is removal or alleviation of the triggering event or underlying cause, as well as treatment of any associated infection, hypovolaemia, etc. Obstetric emergencies should be attended to immediately. In the presence of widespread bleeding, specific replacement therapy should be given, which includes the following. • Fresh frozen plasma (FFP): almost all procoagulant factors and inhibitors are contained within

Coagulation disorders

Trigger factors

Activation of coagulation cascade

Vessel wall damage

Fibrin–platelet thrombosis

Platelet activation

End-organ damage

Lysis and repair

Coagulation factor deficiency

Thrombocytopenia

Fibrinolysis activation

Generalized bleeding tendency

FDPs generated

Fig. 11.2 Pathogenesis of acute

disseminated intravascular coagulation. FDPs, fibrin degradation products.

FFP, and approximately 4–5 units should be rapidly infused. • Cryoprecipitate: contains fibrinogen in a ‘concentrated’ form and 5–10 units should be infused with the initial FFP. • Platelet concentrates: approximately one to two complete adult doses to be given. • Following initial replacement therapy, laboratory tests should be repeated and any further treatment guided by both the clinical and laboratory response. Heparin anticoagulation may also be useful in situations where initial replacement therapy has failed to control excessive bleeding or when DIC is complicated by microvascular thrombosis or large-vessel thrombosis. Low-dose continuous intravenous therapy (500–1000 IU/h) is one suggested regimen.

Specific clotting factor inhibitor concentrates (including antithrombin and activated protein C) may have a role in the management of certain groups of patients (e.g. those who do not respond to simple replacement therapy or who have overwhelming sepsis or meningococcaemia). Liver disease

All coagulation factors, except factor VIII and von Willebrand factor (vWF), and protease inhibitors are synthesized by hepatocytes. The liver also serves to remove activated intermediates of coagulation from the bloodstream. In liver disease, coagulopathy may result from a number of mechanisms: reduced synthesis of coagulation factors; cholestasis and subsequent malabsorption resulting in vitamin K deficiency; and acquired ‘dysfib141

Chapter 11

rinogenaemia’. The platelet count is often reduced due to hypersplenism. Laboratory abnormalities seen in liver disease include the following. • Prolonged PT and APTT. • Prolonged TT: this may result from low fibrinogen concentration or dysfibrinogenaemia. A prolonged reptilase time despite a normal fibrinogen concentration implies dysfibrinogenaemia. • Elevated D-dimers. It is important to note that abnormal coagulation tests are not always associated with bleeding and, in such cases, patients do not require replacement therapy. However if there is active bleeding, then replacement of clotting factors with FFP and platelet transfusions to maintain a platelet count above 50 ¥ 109/L should be instituted. Vitamin K in doses of 10–20 mg may produce some improvement in the coagulation abnormalities. Prothrombin complex concentrates (which contain factors II, IX and X) may sometimes be used in severe cases, but if so, with great caution, as they may precipitate DIC.

Complications of anticoagulant and thrombolytic drugs Oral anticoagulants

Coumarin and phenindione derivatives act by blocking the g-carboxylation of glutamic acid residues of vitamin K-dependent coagulation factors, resulting in decreased biological activity of factors II, VII, IX and X, as well as proteins C and S. The international normalized ratio (INR) monitors their effect on the haemostatic system. Some clinical situations may be associated with an increased risk of bleeding during anticoagulation and these are listed in Table 11.3. Management of excessive anticoagulation depends on the INR and whether there is minor or major bleeding. In the absence of haemorrhage, warfarin should be stopped for a few days and recommenced when the INR falls into the desired range. Small doses of vitamin K (1–2.5 mg) may be given intravenously/orally if the INR is greater than 5.0, as there is a significantly greater risk of serious haemorrhage at this level. 142

Table 11.3 Conditions associated with increased risk of

bleeding during anticoagulation. Age (possible) Uncontrolled hypertension Alcoholism Liver disease Poor drug or clinic visit compliance Active major bleeding Previous intracranial bleeding Potential bleeding lesion (e.g. aneurysm, internal ulcer) Thrombocytopenia Platelet dysfunction (e.g. use of aspirin)

If the patient is bleeding, then the anticoagulant effect should be reversed. Since the action of vitamin K is not maximal for at least 24 h, additional measures are required. • FFP (10–15 mL/kg body weight) will immediately supply the necessary coagulation factors. However, there are some potential problems with this type of therapy. Very large amounts of plasma (1–2 L) may need to be infused in order to correct the coagulopathy; and even though the INR may correct into the normal range, this is misleading since it is not sensitive to factor IX, the concentration of which is only minimally increased by treatment with FFP. • As a result, alternative therapies with clotting factor concentrates have been used in some patients. These include a combination of prothrombin complex concentrate (PCC), which contains variable amounts of factors II, IX and X, together with specific factor VII concentrates; or Prothromplex T (Immuno, Vienna), which contains factors II, VII, IX and X. The dose of concentrate used is calculated at approximately 50 IU factor IX per kilogram body weight. These concentrates are able to correct the defect caused by oral anticoagulants; however, they carry the potential risk of inducing thromboembolism as they often contain activated coagulation components. Therefore if using these products, caution should be exercised especially in high-risk groups. • Haemorrhage occurring in a warfarinized patient with an INR in the therapeutic range should be managed as above; however, additional

Coagulation disorders

investigations to exclude any underlying local lesions should also be performed. Thrombolytic agents

These agents generally cause a state of systemic lysis. However, the degree to which this is affected varies according to the particular drug used. Streptokinase has a greater effect on the laboratory markers of systemic lysis than does tissue plasminogen activator; however, this does not appear to correlate with the incidence of bleeding. Laboratory tests such as the TT and fibrinogen levels will detect the presence of a systemic lytic state, although they do not predict the likelihood of haemorrhage, and nowadays most protocols use fixed-dose schedules. Haemorrhage complicating these agents is most commonly local (e.g. at the site of catheterization in the groin), although intracranial or gastrointestinal bleeding may occur. Measures such as pressure packs will often control local bleeding; more serious bleeding usually necessitates discontinuing thrombolysis. Most agents have a short half-life (minutes) and so the fibrinolytic state will reverse within a few hours of drug cessation. The exception to this is acylated plasminogen– streptokinase activator complex (APSAC), which has a half-life of 90 min. In the case of life-threatening haemorrhage, infusions of cryoprecipitate or FFP can be given to reverse the hypocoagulable state. Antifibrinolytic drugs such as e-aminocaproic acid may or may not provide some additional benefit. Uraemia

Bleeding is a relatively common complication of renal failure: the major cause is that of platelet dysfunction as well as a defect in platelet–vessel wall interaction. Deficiencies of coagulation factors are not a common feature, unless there is complicating liver disease or DIC. Many qualitative platelet defects can be demonstrated in vitro, including impaired aggregation in response to agonists as well as storage pool defects. However, these abnormalities do not appear to correlate well with clinical bleeding. It is

also thought that plasma from uraemic patients contains an inhibitor that interferes with normal vWF–platelet interaction. Dialysis is useful in reversing the haemostatic defects in uraemia, although this may not correct them entirely. Anaemia, particularly when packed cell volume (PCV) is below 20%, should be corrected by either blood transfusion or erythropoietin as this improves platelet function and shortens bleeding time. Infusions of 1-deamino-8-Darginine vasopressin (DDAVP) (0.4 mg/kg) have been used successfully to provide short-term correction of the bleeding time and decreased symptoms of bleeding. Massive transfusion

This is generally defined as blood loss requiring the replacement of the patient’s total blood volume in less than 24 h, and may be associated with coagulation abnormalities. Thrombocytopenia can occur reasonably quickly and usually results from dilution, but increased consumption of platelets may also occur. The use of plasma-reduced red cell concentrates can result in significant dilution of coagulation factors, although generally the haemostatic concentration of the coagulation factors is well maintained and products such as FFP should not be given prophylactically. In order to prevent the indiscriminate use of component therapy, patients receiving massive transfusions should have routine tests of haemostasis performed early in order to define precise abnormalities. Microvascular bleeding and general oozing from wounds or venepuncture sites is particularly likely when the platelet count falls below 50 ¥ 109/L and platelet concentrates should be infused in order to control any microvascular bleeding. If laboratory tests show that a coagulopathy is present, or if clinical judgement is such that treatment cannot wait for results of tests, then FFP at a dose of 10–15 mL/kg will replace both fibrinogen and other critical clotting factors. If hypofibrinogenaemia is present or suspected, then cryoprecipitate should also be given. The need for ongoing haemostatic treatment should be guided by the 143

Chapter 11

patient’s clinical response and results of repeated laboratory tests. In recent years, recombinant factor VIIa (NovoSeven) has been used for patients with uncontrollable life-threatening haemorrhage. This product was originally developed for use in haemophilia patients with inhibitors to factor VIII or IX; however, a number of clinical trials for indications other than haemophilia are currently in progress. Due to its expense, its use is generally limited to ‘rescue’ therapy for massively transfused patients with persistent bleeding despite appropriate blood component transfusion, haemostatic and pharmacological measures and surgical intervention. Experience based solely on case reports indicates a haemostatic effect of recombinant factor VIIa in doses of 60–120 mg/kg; its relatively short half-life means that a repeat dose may need to be given every 2–3 h to decrease bleeding significantly. The decision to use this product should generally be made in consultation with a haematologist.

Transfusion support for patients with inherited haemostatic defects Haemophilia A

This sex-linked recessive disorder results in absent/low levels of factor VIII, although it may also arise from a spontaneous mutation in up to one-third cases. Clinical features vary according to the factor VIII level and patients can be classified into mild, moderate or severe according to their factor VIII coagulant activity (Table 11.4). The severity and type of bleeding is related to the absolute level of factor VIII. The minimal effective level for haemostasis is generally about 25–30%. Investigations

Laboratory abnormalities seen in haemophilia A include: • prolonged APTT; • reduction of factor VIII coagulant activity; and • normal vWF activity (it is important to measure vWF activity in order to exclude vWD, which will also give low factor VIII levels). 144

Table 11.4 Clinical manifestations in haemophilia A.

Factor VIII level (% normal)

Clinical manifestation

<1% (severe disease)

Usual age of onset <1 year Spontaneous bleeding common (haemarthrosis, muscle haematoma, haematuria) Bleeding post surgery and dental extraction Post-traumatic bleeding Crippling joint deformity if inadequate treatment

1–5% (moderate disease)

Usual age of onset <2 years Occasional spontaneous bleeding Bleeding post surgery and dental extraction Post-traumatic bleeding

6–40 % (mild disease)

Usual age of onset >2 years Bleeding post surgery and dental extraction Post-traumatic bleeding

Management

The mainstay of treatment is to raise the concentration of factor VIII sufficiently to arrest spontaneous and traumatic bleeds or to cover surgery. There are a number of products currently available which can be used to treat this condition, including: • recombinant factor VIII preparations; • plasma-derived factor VIII concentrates (which may vary in degree of purity); • DDAVP (for mild disease only, baseline factor VIII above 15%). The choice of which product to use is generally determined by cost and availability. Recombinant products are the treatment of choice if available and affordable, because they eliminate the risk of transmission of human and animal infectious agents. They are generally recommended as the product of choice for prophylaxis to prevent spontaneous joint bleeding for children with severe haemophilia, as well as replacement therapy for previously untreated patients. Factor VIII inhibitors may develop in up to one-third of patients receiving recombinant products.

Coagulation disorders

Plasma-derived clotting factor concentrates are currently considered to be ‘safe’ in terms of human immunodeficiency virus (HIV) and hepatitis viruses due to effective donor screening and specific viral ‘killing’ procedures, and are used when recombinant products are unavailable. DDAVP (0.3 mg/kg) given intravenously, subcutaneously or intranasally can be used to control bleeding in mild haemophiliacs. Hyponatraemia and water intoxication are adverse effects of this drug, and hence it is not recommended for children under 2 years of age. It is also thought to have thrombogenic potential and may need to be avoided in the elderly or those with known vascular disease. Patients with moderate/severe haemophilia will require treatment with recombinant or plasmaderived factor VIII concentrates for bleeding, prior to invasive procedures, surgery, etc. It is known that a dose of factor VIII of 1 unit/kg will result in an increase in plasma factor VIII level by 2%. The level of factor VIII concentrate required to achieve adequate haemostasis will depend on the type of bleeding, but can be calculated according to the formula: Units of factor VIII required = weight (kg) ¥ desired level (%) ¥ 0.5 The plasma half-life of factor VIII is 8–12 h and thus repeated doses at 12-h intervals are usually needed. Alternatively, a continuous infusion of factor VIII can be given. For major soft tissue bleeds, levels above 50% are generally sufficient; however, for major surgery, a preoperative level of 100% is necessary and thereafter levels of 50–100% are sufficient for adequate wound healing. Haemophilia B

This sex-linked recessive disorder results in a deficiency of factor IX. The clinical features are identical to those of haemophilia A. Investigations

Laboratory abnormalities seen in haemophilia B include:

• prolonged APTT; • reduction of factor IX coagulant activity. Management

The main types of products currently used for treatment include: • intermediate-purity factor IX complex concentrates (PCC); • high-purity factor IX concentrates; • recombinant factor IX products. The choice of factor IX replacement product has been influenced by the fact that PCC contains significant amounts of activated coagulation factors II, VII and X, and can cause thromboembolism and DIC in some patients. Highly purified factor IX concentrates (which contain very little or no other coagulation factors) and recombinant factor IX (unique in containing no animal or human proteins) are also available. Although more costly, they are less thrombogenic than PCC and are the products of choice for prophylactic regimens, surgery and those patients at high risk of venous thrombosis/DIC. The dosage of factor IX required can be calculated according to the formula: Units of factor IX required = weight (kg) ¥ desired level (%) ¥ 1.0 The plasma half-life of factor IX is 18–30 h, and therefore if repeated doses are needed they should be given every 12–24 h or by continuous infusion. Treatment of patients with inhibitors

Patients with haemophilia can develop inhibitor antibodies to factor VIII or IX, sometimes making their treatment quite difficult. Inhibitors are most common in patients with severe haemophilia A. Haemophilia A

• If the inhibitor is of low titre (i.e. <10 Bethesda units), then bleeding episodes can be treated with higher than normal doses of human factor VIII or porcine factor VIII (based on the species specificity of factor VIII inhibitors). • If the inhibitor is of high titre (i.e. >10 Bethesda 145

Chapter 11

units), human factor VIII is ineffective to control bleeding, and management strategies include use of porcine factor VIII, recombinant factor VIIa, FEIBA or PCC. For major haemorrhage, recombinant factor VIIa (70–90 mg/kg) is generally recommended as first-line therapy (if available). Eradication of inhibitors with ‘immune tolerance induction’ using factor VIII concentrates alone or together with immunosuppressives is considered the best long-term treatment option for these patients.

Table 11.5 Variants of von Willebrand disease.

Haemophilia B

Type 3 Autosomal recessive inheritance Severe quantitative deficiency of vWF Severe haemophilia-like bleeding disorder

• Immune tolerance using factor IX concentrates can be attempted, otherwise PCC or recombinant factor VIIa can be used for bleeding. von Willebrand disease

This is the most common of the inherited bleeding disorders and is due to a quantitative and/or qualitative defect in the vWF protein. vWF has two main functions: firstly, it promotes the adhesion of platelets to the subendothelium by binding to the platelet receptor glycoprotein Ib; secondly it protects factor VIII:c from proteolytic degradation by forming a non-covalent association. vWD is a heterogeneous group of disorders and is classified into three different types (Table 11.5). Depending on the type of vWD, some patients may be asymptomatic whereas others will have haemophilia-like bleeding. Laboratory abnormalities seen in vWD include: • prolonged PFA-100 closure time (an in vitro ‘high shear’ bleeding time device); • reduction of vWF antigen; • reduction of vWF ristocetin cofactor activity; • reduction of factor VIII coagulant activity (which can cause prolonged APTT); • abnormal vWF multimers in some subtypes. The goal of therapy in patients with vWD is to correct the dual defect of haemostasis, i.e. the abnormal platelet adhesion and the abnormal coagulation due to low fctor VIII levels. It is important to distinguish between the various types of vWD as treatment will differ. • For patients with type 1 disease, DDAVP is the 146

Type 1 Autosomal dominant inheritance Partial quantitative deficiency of von Willebrand factor (vWF) Normal vWF multimers Mild bleeding disorder that decreases during pregnancy, elderly Type 2 Autosomal dominant inheritance Qualitative deficiency of vWF Numerous subtypes Abnormal vWF multimers Generally mild bleeding disorder

treatment of choice and a dose of 0.3 mg/kg is usually given. It is important to first test an individual patient’s response to DDAVP prior to using it to ‘cover’ procedures. • For patients with types 2 and 3 disease, vWF ‘replacement therapy’ is generally required. Either a factor VIII concentrate rich in vWF or a purified vWF concentrate is the treatment of choice. One problem with factor VIII concentrates is that they rarely contain the highest molecular weight multimers of vWF, which are very important for adequate haemostasis. Similarly, the very low content of factor VIII in purified vWF concentrates may make it necessary (in the event of acute bleeding or emergency surgery) to infuse vWD patients with a single first dose of purified factor VIII concentrate to ensure immediate correction of the low factor VIII levels. In the past, cryoprecipitate was used to treat patients with vWD; however, it is now unacceptable to use such untreated plasma derivatives when there are ‘safer’ alternatives available. Other inherited disorders

Hereditary deficiencies of other coagulation factors are generally rare. Factor XI deficiency is particularly common among Ashkenazi Jews and is transmitted as an autosomal recessive trait.

Coagulation disorders

There is a poor correlation between factor XI levels and bleeding tendency, which usually presents following surgery or dental procedures. If available, plasma-derived factor XI concentrates should be given to treat bleeding, otherwise FFP will suffice. Deficiencies of factors II, V, VII, X and XIII and fibrinogen can all be treated with FFP; however, if there are more specific therapies available then they should be used. Currently, there are specific factor concentrates for factors VII, X and XIII and fibrinogen, although these may not always be available. Cryoprecipitate can be used for fibrinogen deficiency/dysfibrinogenaemias; PCC can be given to patients with factor II or X deficiency (although thromboembolic risks should be considered). Patients with deficiencies of ‘contact factors’ (factor XII, prekallikrein, and high-molecularweight kininogen) do not bleed excessively and do not require any treatment.

Appropriate and inappropriate use of FFP, cryoprecipitate and coagulation factor concentrates Fresh frozen plama

FFP is obtained from either single donations or plasmapheresis collections and is prepared by freezing plasma to a temperature of –30°C or less within 6 h of collection in order to preserve the activity of coagulation factors. Each unit contains all of the coagulation factors and has a volume of approximately 200 mL. When ready for use it must be thawed at 37°C and — ideally it should be infused within 2 h of thawing. It can be stored at 4°C for up to 24 h as long as factor VIII replacement is not required. The dose of FFP required will depend on the clinical indication; however, an initial dose of 10–15 mL/kg is usually given. As this is an empirical dose, laboratory tests should be used to monitor its efficacy, and results of these tests as well as the patient’s clinical response should guide any further dosing requirements. ABO-compatible FFP should be used, although compatibility testing is not required. In case of an

emergency when the patient’s blood group is not known, group AB plasma may be safely given. RhD-compatible plasma should be given to women of child-bearing age since FFP contains small amounts of red cell stroma, which can boost the response to RhD in previously immunized patients. Single-unit FFP is not a virally inactivated product. Other alternatives include quarantined FFP, methylene blue-treated single unit, or solvent–detergent (SD)-treated pooled plasma. All these products have been shown to be clinically effective, although SD plasma lacks the highmolecular-weight forms of vWF. In the UK, virally inactivated FFP from bovine spongiform encephalopathy (BSE)-free countries is recommended by the Department of Health for children born after 1 January 1996; the viral inactivation will be carried out by methylene blue treatment method. Virus-inactivated FFP sourced from non-UK untransfused male donors is the safest product for the UK to avoid the risks of transfusion-related acute lung injury (see Chapter 14) and variant Creutzfeldt–Jakob disease (see Chapter 20). There are difficulties in establishing specific criteria to determine which patient groups should receive the safest available FFP. Although extending the use of virus-inactivated FFP sourced from non-UK donors to all recipients remains under consideration, the main constraint is cost. While the use of FFP has become widespread, it may be transfused inappropriately. There are a limited number of ‘definite’ indications for the use of FFP and these are described below. ‘Definite’ indications Immediate reversal of overdosing with oral anticoagulant In the event of serious or life-threatening bleeding requiring immediate reversal of the anticoagulant effect, PCC is regarded as the treatment of choice. PCC does not usually contain factor VII, and thus a separate factor VII concentrate should be given (if available). The recommended dose is 50 IU factor IX activity per kilogram body weight. PCC is potentially thrombogenic and this should be 147

Chapter 11

borne in mind if used. FFP is an alternative if PCC is not available, using a dose of 10–15 mL/kg. Further doses of PCC or FFP should be given according to laboratory results and patient’s clinical state. Vitamin K deficiency Conditions that impair vitamin K absorption (e.g. biliary tract obstruction) as well as haemorrhagic disease of the newborn can result in a coagulopathy similar to that seen with warfarin overdosage. Any serious or life-threatening bleeding should be treated in the same manner as described above. Acute DIC In acute DIC, activation of the coagulation and fibrinolytic systems results in depletion of platelets and coagulation factors (especially factors V and VIII and fibrinogen). Treatment is always aimed at removing the underlying cause. In cases of haemorrhage associated with laboratory abnormalities, replacement therapy with FFP, cryoprecipitate and platelet concentrates is indicated. If there is no bleeding, then replacement therapy should not be given in an attempt to correct the coagulopathy. Thrombotic thrombocytopenic purpura FFP in conjunction with plasma exchange has been the treatment of choice for patients with thrombotic thrombocytopenic purpura (TTP). Large daily doses are needed, usually in the order of 3 L. In some series, the supernatant portion formed during the preparation of cryoprecipitate (cryosupernatant) has been shown to be more effective than standard FFP when used as the replacement fluid. Cryosupernatant plasma is depleted in factor VIII and fibrinogen; but whereas the factor VIII concentration may only be about 0.11 IU/mL, proportionately less fibrinogen may be removed, leaving up to 70% remaining. All forms of FFP contain the metalloproteinase enzyme, which is deficient or inhibited in TTP and is responsible for degrading the ultra-large multimers of vWF that cause the excessive platelet activation and consumption in this condition. Cryosupernatant is deficient in vWF multimers but contains metalloproteinase. The reduced activity of protein S in SD-treated FFP has been associated with the devel148

opment of venous thromboembolism in patients with TTP. Replacement of single factor deficiencies As more specific factor concentrates become increasingly available, FFP should only be used as replacement when specific/combined factor concentrates are unavailable. Currently there are no specific factor concentrates available for deficiencies of factors II (prothrombin) and V. Inherited deficiencies of inhibitors of coagulation Previously, FFP has been used as a source of antithrombin, protein C and protein S for patients with inherited deficiencies of these inhibitors who may be receiving heparin therapy for spontaneous thrombosis or who are undergoing surgery. Now that specific concentrates are being manufactured (antithrombin and protein C), FFP should only be used when these are not available. Other indications

There are several other circumstances where FFP may be used appropriately to treat bleeding in the presence of abnormal coagulation. Liver disease Coagulation abnormalities occur quite frequently in patients with severe liver disease; however, bleeding is often precipitated by an event such as surgery or liver biopsy, and is rarely attributable to the haemostatic defect alone. If there is bleeding (or a very strong possibility that bleeding will occur), then FFP is indicated. Large volumes are often required to control the bleeding/correct the defect, and this can be problematic in patients who may already have an expanded plasma volume. Complete normalization of a prolonged PT is often not possible, and the use of PCC may be considered. However, one must be aware of the potential risks of inducing thrombosis or DIC in these patients, particularly since they already suffer from impaired clearance of activated clotting factors and reduced levels of antithrombin. Since thrombocytopenia and platelet function defects are also a feature of hepatic disease, platelet concentrates may also need to be given.

Coagulation disorders

Massive transfusion This is defined as the replacement of a patient’s total blood volume within 24 h. Coagulation abnormalities seen in this setting are more closely related to the clinical condition necessitating the transfusion rather than the volume of blood transfused. Prophylactic ‘replacement’ regimens with FFP are not indicated. Instead, treatment with both FFP and cryoprecipitate should be guided by the patient’s clinical status as well as the results of laboratory tests (if available). Thrombocytopenia is a frequent occurrence and should be treated with platelet concentrates if necessary (usually when the platelet count falls below 50 ¥ 109/L). More recently, recombinant factor VIIa (NovoSeven) has been used for massively transfused patients with persistent bleeding despite appropriate blood component transfusion, haemostatic and pharmacological measures, and surgical intervention. Cardiopulmonary bypass Haemostatic disturbances during cardiopulmonary bypass are usually due to platelet dysfunction. If there is persistent bleeding (despite adequate platelet transfusion) and a coagulopathy other than that caused by heparin has been demonstrated, then FFP should be used. Special paediatric conditions FFP can be used in the treatment of neonatal DIC, and is also sometimes used in newborn/premature infants with severe sepsis, although FFP should be avoided in neonates with T-antigen activation (see Chapter 8). Special paediatric packs of FFP that contain smaller volumes than standard FFP should be used. Inappropriate use

The use of FFP is inappropriate in the following conditions: • hypovolaemia; • routine plasma exchange; • part of a ‘predetermined’ replacement protocol; • treatment of immunodeficiency or proteinlosing states.

Cryoprecipitate

Cryoprecipitate is prepared from plasma that is frozen quickly and then thawed slowly at 1–6°C, leaving behind a small amount of precipitated protein. The cryoprecipitate protein is then resuspended in a small volume of residual supernatant plasma (usually 9–16 mL). It contains factor VIII, fibrinogen, vWF, factor XIII and fibronectin in higher concentrations than are found in plasma. The main use for cryoprecipitate is as a source of fibrinogen and a single unit usually contains 200–250 mg fibrinogen. It can be used appropriately in cases of hypofibrinogenaemia such as DIC, reversal of fibrinolytic agents and advanced liver disease. Plasma levels of fibrinogen above 1.0 g/L are generally considered as adequate for haemostasis. When using cryoprecipitate, it is usual to thaw and pool approximately 10–30 donor units for infusion; however, the dose will vary according to the clinical condition and the patient’s fibrinogen level. An empirical rule is to give one bag of cryoprecipitate for every 5 kg of body weight. In the past, cryoprecipitate was used to treat vWD and haemophilia A; however, with the advent of drugs such as DDAVP and the availability of specific factor concentrates, it should no longer be used for the treatment of these disorders. Cryoprecipitate has more recently been used for the production of fibrin surgical adhesive (‘glue’), which can be used in various surgical procedures and can be prepared from autologous donors. Coagulation factor concentrates

Both specific and non-specific coagulation factor concentrates are prepared from plasma using a number of different techniques in an attempt to produce a ‘purified’ product that has undergone viral inactivation. Specific factor concentrates for many of the coagulation factors have now been developed and are in clinical use as replacement therapy for inherited deficiencies. Their use has been discussed previously in the treatment of inherited haemostatic defects. 149

Chapter 11

Non-specific factor concentrates such as PCC generally contain factors II, IX and X together with variable amounts of factor VII. They are generally used in conditions associated with deficiencies of one or more of these factors (e.g. treatment of overdosage with warfarin); however, since they often contain ‘activated’ forms of coagulation factors, they have thrombogenic potential, which can limit their use.

Summary Coagulation disturbances are relatively common and it is important to know when to treat as well as the most appropriate therapy. Simple laboratory tests will often give an indication of the problem and more specific tests will indicate the exact diagnosis. With the widespread availability of plasma and factor concentrates, it is important for individual hospital centres to set up guidelines in order to minimize wastage and inappropriate use of these products. Cost is another important issue that needs to be addressed when drafting guidelines. Wherever possible, specific factor concentrates should be used to treat coagulation factor deficiencies in preference to plasma. Although advances in viral inactivation procedures have resulted in the development of factor concentrates that are ‘safe’, it is important to remember that there are potential risks associated with the use of these products and that not all inactivation procedures inactivate all known viruses. Recombinant technology allows the provision of safe and pure treatment, but does not effect cure of the underlying condition.

150

Further reading Bolton-Maggs PHB, Pasi KJ. Haemophilias A and B. Lancet 2003; 361: 1801–9. British Committee for Standards in Haematology. Guidelines for the use of fresh frozen plasma. Transfus Med 1992; 2: 57–63 (revision in preparation, see www.bcshguidelines.org). Davis S. Use of factor VIIa in uncontrollable haemorrhage. A position statement of the NSW Therapeutic Assessment Group Inc., October 2002 (see www.nswtag.org.au). Machin SJ. Acquired coagulation, non-immune platelet disorders and vascular purpuras. In: Hoffbrand AV, Lewis SM, Tuddenham EGD, eds. Postgraduate Haematology. Oxford: Butterworth-Heinemann, 1999: 636–52. Makris M, Watson HG. The management of coumarininduced over-anticoagulation Br J Haematol 2001; 114: 271–80. Makris M, Greaves M, Phillips WS et al. Emergency oral anticoagulant reversal: the relative efficacy of infusions of fresh frozen plasma and clotting factor concentrates on correction of the coagulopathy. Thromb Haemost 1997; 77: 477–80. Practice guidelines for blood component therapy. A report by the American Society of Anesthesiologists Task Force on Blood Component Therapy. Anesthesiology 1996; 84: 732–47. Smith OP, Hann IM, Machin SJ. The use of factor concentrates in the management of hemophilia A and other inherited coagulation disorders. In: Pamphilon D, ed. Modern Transfusion Medicine. Boca Raton, FL: CRC Press, 1995: 165–76. Wallington TB. Transfusion of plasma and its products. In: Pamphilon D, ed. Modern Transfusion Medicine. Boca Raton, FL: CRC Press, 1995: 139–64.

Chapter 12

Uses of intravenous immunoglobulin David J. Unsworth and Tim B. Wallington

The potential list of clinical indications for intravenous immunoglobulin (IVIG) is expanding rapidly. Evidence justifying the use of IVIG in many clinical situations (e.g. systemic lupus erythematosus, IgA nephropathy, chronic fatigue syndrome, asthma, multiple sclerosis and others) is contentious and based on poorly designed, small or uncontrolled non-blinded clinical studies. There are also concerns that demand will outstrip supply, since production depends on the availability of safe human plasma, from which normal IgG is fractionated. Hepatitis C outbreaks associated with IVIG in the recent past remind us that (as with cellular blood products) transmission of infection is a major potential adverse effect. This is not surprising given that IVIG production depends on pooling of plasma immunoglobulin from several thousand blood donors. One gram of IVIG is estimated to contain at least 7 million different antibody specificities. One simple and expedient way of reducing the risk of viral transmission and other adverse effects is to restrict use of IVIG to those clinical situations where use is of proven value. This chapter is limited to those diseases for which the published evidence supporting efficacy is scientifically compelling (Table 12.1).

Clinical indications for IVIG Antibody-deficient patients

The only cases that merit consideration for treatment are those with documented IgG deficiency and a clear associated increased vulnerability to infection. Cases of isolated IgA or IgM deficiency

tend not to be associated with recurrent bacterial infection, and do not merit IVIG treatment. Deficiency may be due to a primary/inherited defect where IgG deficiency is an irreversible feature. The commonest primary immunodeficiency which features IgG deficiency and susceptibility to bacterial infection is common variable immunodeficiency (CVID). Sometimes there is an obvious associated T-cell defect. CVID is an umbrella term, probably encompassing several different lymphoid defects. Alternatively, IgG deficiency may arise secondarily to a reversible condition. For example, marrow-toxic drugs, and other drugs including gold, sulphasalazine and certain anticonvulsants, can affect immunoglobulin synthesis. Immunoglobulin deficiency may also be a feature of lymphoproliferative disorders such as myeloma and chronic lymphocytic leukaemia (CLL). Recovery may occur, for example once a particular drug is withdrawn, but it may take several years. Licensed uses include primary immunodeficiencies, recent bone marrow transplantation, myeloma or CLL, where documentated low IgG is associated with increased frequency and severity of bacterial infection. Most cases of IgG deficiency present with global or ‘pan-hypogammaglobulinaemia’. It is the persistently low IgG levels rather than deficiency in IgA or IgM which, over a period of years, leaves patients particularly vulnerable to recurrent bacterial (and some viral) infections. Haemophilus influenzae and Streptococcus pneumoniae are frequently grown from sputum of IgG-deficient cases. Untreated, the majority will develop bronchiectasis. In children, frequent bilateral ear infections

151

Chapter 12 Table 12.1 Clinical indications for IVIG.

Prophylaxis against bacterial infection Primary immunodeficiency states with IgG deficiency, e.g. Bruton’s hypogammaglobulinaemia and SCID Some secondary immunodeficiency states (see text, includes paediatric HIV, CLL, myeloma) Post bone marrow transplantation Immunomodulation of autoimmune conditions Idiopathic thrombocytopenia purpura and certain other autoantibodymediated cytopenias, including post-transfusion purpura* and neonatal cytopenias Kawasaki vasculitis* Guillain–Barré syndrome* and some cases of chronic inflammatory demyelinating polyneuropathy Dermatomyositis Myasthenia gravis * Recommended first-line treatment without need to first trial prednisolone/other treatments. CLL, chronic lymphocytic leukaemia; HIV, human immunodeficiency virus; SCID, severe combined immunodeficiency.

with need for grommets is typical. Permanent hearing impairment is a risk. Failure to thrive and weight loss (cachexia in adults) are seen after a long history of recurrent infection. Parasitic infections are also seen. Giardia lamblia and certain other gastrointestinal infections may lead to diarrhoea and even malabsorption. Patients also frequently complain of a tendency to a productive cough and perennial nasal blockage secondary to bacterial rhinitis. In some rarer cases of adult-onset antibody deficiency, there may be an associated thymoma, which can precede or postdate the diagnosis of IgG deficiency. In some IgG-deficient patients, lymphadenopathy and/or hepatosplenomegaly is seen as a complication of IgG deficiency. Histology may show widespread granulomatous inflammation. The prognosis is much improved if IgG is replaced passively by parenteral infusion. Replacement can be either intravenously or by subcutaneous infusion. Both are effective in reducing infection rates. Experience is greatest with IVIG infusions repeated every 2–4 weeks, rather than the subcutaneous alternative, which is required weekly to maintain adequate plasma IgG levels. IVIG as microbiological prophylaxis in cases of 152

IgG deficiency is used according to body weight, at a dose of 0.2–0.4 g/kg per month. Infused IgG is consumed or catabolized, with a half-life of arround 20 days, so that repeat infusions every 2–4 weeks are required. Treatment will be lifelong or until the associated underlying condition responsible for IgG deficiency is successfully treated. Commercially available IVIG preparations are largely (>95%) composed of IgG. Repeat intravenous infusions will keep trough levels of plasma IgG above the lower limit of normal (>6 g/L) and at a level sufficient to reduce vulnerability to infection, usually around 8 g/L in our practice. The exact dose and intervals between infusions can be determined by serial measurement of trough levels and appropriate adjustments to the treatment protocol. It is important to remember that individual cases differ in their requirements, and perhaps the most important criterion is the minimum dose which achieves clear clinical benefit. This may equate to a trough level considerably higher (or lower) than 8 g/L. In cases of deficiency secondary to reversible causes (e.g. drug-induced marrow dysfunction), the requirement may persist for many months or years but should be regularly reassessed and infusions stopped if endogenous B-lymphocyte function recovers. IgG deficiency related to protein leakage syndromes, such as nephrotic syndrome, should only exceptionally be treated with IVIG, as the infused IVIG is likely to be wasted. Immunomodulation of autoimmune disease

Serendipitous observations in a case of hypogammaglobulinaemia that happened to be associated with idiopathic thrombocytopenic pupura (ITP) led to the realization that IVIG can have immunomodulatory effects. Within 24 h of IVIG infusion in ITP, platelet numbers typically rise. The degree of response varies from patient to patient but, in the majority, a beneficial elevation of the platelet number is seen. This can be short-lived (days to weeks) but is often sufficient to prevent or reverse serious bleeding. The popular explanation for this effect is that infused IgG directly blocks the antiplatelet autoantibody. The possible modes of

Intravenous immunoglobulin

action of IVIG as an immunomodulator are discussed below. The commonly used dose schedule is 0.4 g/kg for five successive days, mainly because this is effective in the prototype condition — ITP. This represents high-dose treatment when compared with the lower monthly doses used in cases of immunodeficiency. Many haematologists now use a single dose of 1 g/kg over 24 h, and this is as effective as the 5day treatment in most cases. The main problem is that beneficial effects in autoimmune disease are short-lived and repeat courses are often required. Also, many of the adverse effects occur more frequently with higher doses (Table 12.2). Little research has been directed at determining minimum effective dose schedules. There are a number of autoimmune diseases where high-dose IVIG therapy is widely accepted, although it is regarded as first-line therapy only in a minority (see Table 12.1). Guillain–Barré syndrome (GBS) (see Chapter 29), Kawasaki vasculitis and post-transfusion purpura (see Chapter 17) are the major examples of where first-line use of IVIG is generally agreed. Although the evidence is compelling, not all preparations are licensed for all these indications, in all countries. Licensed immunomodulatory uses include ITP, Kawasaki Table 12.2 Adverse effects of IVIG.

Treatment Common Hypersensitivity reactions (urticarial rashes, wheeze, other) Fevers/chills in septic patients Uncommon Transmission of viral agents Renal tubular damage* Aseptic meningitis* Cerebrovascular accident Transmission of autoantibody in IVIG (haemolytic anaemia, other) * Seen with high-dose IVIG.

Slow down infusion or stop antihistamine ± steroid as treatment or prophylaxis Ideally treat with antibiotics and defer IVIG May be ‘brand’ specific Consider switching to different brand/product

Stop infusion, report to manufacturer. Check other patients receiving same batch

disease and GBS. Reputable clinical trials in GBS, comparing plasma exchange with IVIG, clearly show that both treatment modalities provide impressive and comparable therapeutic effects (see Chapter 29). The financial costs of these two treatments are similar but IVIG is easier to organize, especially out of hours and remote from specialist regional centres. IVIG has now replaced plasma exchange as the favoured treatment in GBS. As a generalization, conditions that respond to plasma exchange tend also to respond to IVIG; this includes myasthenia gravis. Chronic inflammatory demyelinating polyneuropathy is essentially a relapsing/remitting GBS-type illness. Patients may respond to IVIG or plasma exchange and sometimes one treatment modality works best in a particular patient. A simplistic general overview would suggest that plasma exchange removes autoantibodies or other pathogenetic plasma factors, while IVIG blocks the same factors. IVIG in transplantation

IVIG is predictably useful as bacterial prophylaxis, and is used for that reason after bone marrow transplantation. Studies in this context have also shown reduced risk of cytomegalovirus infection and beneficial effect on graft-versus-host disease. IVIG has also been beneficial before renal allograft transplantation, employing IVIG to block high titres of anti-HLA antibody capable of causing hyperacute antibody-mediated rejection.

Risks and adverse effects Viral transmission

Safety concerns now dictate that all plasma donations processed for IVIG production are tested for human immunodeficiency virus (HIV)-1 and HIV2 and hepatitis B and C antibodies. Although this helps by identifying potentially infected donors, it is disadvantageous because removal of antiviral (neutralizing) antibody at the same time makes viral transmission (from a donor before seroconversion for example) more likely. The hepatitis C outbreak in 1994 in the USA seemed to arise for exactly that reason, with 100 IVIG recipients 153

Chapter 12

wordwide developing clinically significant hepatitis C transmission. Polymerase chain reaction (PCR) testing of plasma pools for hepatitis C and other agents has therefore become standard practice. Different virucidal strategies are additionally used by the manufacturers to provide further confidence (Table 12.3). The Food and Drug Administration in the USA requires manufacturers to employ model viruses to demonstrate effectiveness of their viricidal strategies in vitro. Nonetheless, vigilance is still required as aplastic anaemia due to transmitted parvovirus B19 infection has been reported. This virus is unaffected by solvent–detergent treatment, and can partially resist heat treatment. Despite use of HIV-contaminated plasma in the early 1980s, HIV has never been transmitted by IVIG. Cold ethanol fractionation, the basis of

extracting IgG from plasma, inactivates HIV and markedly reduces hepatitis B load. Prion transmission

Outbreaks of variant Creutzfeldt–Jakob disease (vCJD) in the UK in the 1990s led to concerns that the prion responsible might be transmissible by blood or blood products (see Chapter 20). Concerns were such that use of UK plasma in manufacture was banned, and UK producers purchased North American plasma instead. Recently, however, occasional cases of bovine spongiform encephalopathy have been identified in the USA. Follow-up of patients in the UK who received blood from donors who subsequently died of vCJD has provided evidence that transmission may occur by red cell transfusion. Fractionation Table 12.3 Virus precautions in

Commercial preparations:

1

2

3

4

5

6

Antibody tests on each donation HAV HBV HCV HIV ALT/AST Other

N Y Y Y Y —

N Y Y Y Y —

— Y Y Y Y —

Y Y Y Y Y Syphilis

— Y Y Y Y HIV p24

Y Y Y Y Y —

PCR tests on donor minipools HCV HIV Other

Y Y HBV

Y N HAV

Y N —

Y N —

— — —

— — —

Virus-inactivating steps Cold/ethanol Solvent–detergent Ion-exchange chromatography Incubation at low pH pH4/pepsin Pasteurization PEG precipitation DEAE-Sephadex chromatography

Y Y Y Y — — — Y

Y — — — Y — — —

Y — — — — Y Y Y

Y Y — Y — — — —

Y Y — — — — — —

Y — — — Y — — —

Plasma source USA EC (non-UK)

Y —

Y —

Y —

Y Y

Y Y

Y Y

ALT/AST, alanine aminotransferase/aspartate aminotransferase; HAV, hepatitis A virus; HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus.

154

different commercial preparations.

Intravenous immunoglobulin

models, using prion-spiked plasma, suggest that prion if present will be removed by most routine IVIG manufacturing processes.

documented pre-existing renal dysfunction and is therefore a potential pitfall of considering IVIG treatment in myeloma patients.

Hypersensitivity reactions

Aseptic meningitis

Cutaneous itching or wheezing occurring during or soon after infusion occurs in up to 5% of patients, although these reactions tend to be mild. Slow infusion rates (0.01 mL/kg per min or less) tend to minimize such reactions. Severe reactions might be expected in IgA-deficient individuals, because they are at risk of generating anti-IgA antibodies against the low levels of contaminating IgA found within certain IVIG preparations. Fortunately, however, the problem is very rarely seen in practice and not all experts agree that routine pretreatment screening for IgA deficiency is worthwhile. Where severe IgA deficiency does exist, it seems prudent to select an IVIG product with a very low (<0.005 mg/mL) content. Where an anaphylactoid reaction (see Chapter 15) occurs in association with IVIG administration the infusions should be stopped and, after appropriate resuscitation, investigations undertaken to check for antiIgA antibodies. Septic patients who have been given IVIG may well develop feelings of severe malaise, with shivers and chills. This problem is well recognized in antibody-deficient patients. Treatment should, if possible, be deferred until antibiotics have reduced the bacterial load in these circumstances. Similar symptoms can arise if complement-activating IgG aggregates in IVIG preparations are infused, even in the absence of sepsis. This arises, for example, when lyophilized IVIG preparations are reconstituted inefficiently. Both these types of reaction are thought to involve immune complex formation with complement activation.

This is also seen generally with high-dose IVIG and often in cases with a history of migraine. Patients report very severe meningitic symptoms, typically 48 h after infusion. Cerebrospinal fluid analysis shows high protein and white cells (neutrophils and lymphocytes). Switching to a different IVIG preparation often fails to resolve the problem. Very slow infusion rates sometimes help.

Renal dysfunction

High-dose treatments may cause acute oliguric renal failure because sugars used to stabilize the IgG present a high solute load, which damages the proximal renal tubule. The defect is not always reversible. It is most likely to develop in cases of

Hyperviscosity/cerebrovascular accident

This problem is fortunately rare. It is reduced by slow infusion and the risk increases with higher doses. Caution is sensible in arteriopathic patients. Deep vein thrombosis has also been reported. IVIG-containing ‘autoantibodies’ or haemolysins

Uveitis due to antineutrophil cytoplasmic antibody and haemolysis due to high-titre ABO antibodies have been the subject of case reports. ABO crossmatching of IVIG preparations against the patient’s red cells is not however required in routine practice.

Possible immunomodulatory mechanisms of action The popular theories relating to mode of action of IVIG in suppressing certain autoimmune diseases are summarized in Table 12.4. Certain hypotheses are more suited to explaining the beneficial effect in one disease, while being less plausible in another. For example, while IgG Fc receptor blockade is a good explanation in ITP, it does not adequately explain the mode of action in myasthenia gravis, where the autoantibody targets the acetylcholine receptors at nerve endings. An anti-idiotype block is the more likely explanation in myasthenia gravis. Of interest, in ITP for example,

155

Chapter 12 Table 12.4 Immunomodulation and IVIG: possible modes

of action. Fc receptor blockade IgG anti-idiotype effect Interference in immune complex clearance Direct effect on lymphoid cells (T, B or NK) Cytokine block Enhanced catabolism of autologous IgG Block of complement activation

infusion of anti-D can be effective, supporting the hypothesis of Fc blockade in ITP. Although IVIG is predominantly composed of IgG, it also contains contaminating factors that might play a crucial immunomodulatary role. These incude soluble HLA molecules, soluble cytokine receptors capable of specific cytokine block, soluble CD4 molecules and others. One key observation is that in diseases with an immunomodulatory action beyond dispute, the benefit is seen typically within hours or a few days. Direct and/or indirect effects of IVIG on lymphocye function and particularly T-cell function are suspected but ill understood. T lymphocytes do carry IgG Fc receptors. One possibility is that cytokine production is blocked or switched to favour an anti-inflammatory effect.

Effects of different manufacturing methods Viral inactivation

All production methods are based on Cohn fractionation (see Table 12.3). This inactivates HIV and markedly reduces the titre of viable hepatitis B. Some products have good safety and efficacy records extending back many years, including those using a pH4 pepsin step, which seems to be fortuitously virus inactivating. Storage in low pH solution is also probably virus inactivating. Newer products, however, include more premeditated specific antiviral steps such as pasteurization or a solvent–detergent step. The latter method inactivates enveloped viruses (including hepatitis C) only. 156

Therapeutic benefit

Different products are assumed to show comparable clinical efficacy, though trials have not been carried out. There is no evidence to the contrary. Levels of contaminating IgA

No product is entirely IgA free, but some products have very low levels indeed (<0.005 mg/mL) and are therefore preferred for truly IgA-deficient patients.

General basic measures prior to IVIG use • Baseline renal and liver biochemistry and a full blood count. • Storing serum and whole ethylenediamine tetraacetic acid (EDTA) samples before treatment for future baseline comparison may be useful in patients receiving lifelong repeated IVIG (i.e. primary immunodeficiency states) to provide an audit trail in case of suspected viral/prion transmission. • In long-term treatment, monitor liver function to check for the possibility of viral hepatitis transmission. • Monitor the underlying disease itself (platelet counts in ITP, and blood counts to detect recovery of marrow function in patients after treatment with cytotoxics), and regularly reassess whether continued IVIG treatment is indicated.

Cost considerations Three basic considerations will ensure that patients are treated in the most cost-effective manner. • Restrict treatment to those illnesses where the treatment is of proven benefit (established by controlled clinical trial). • Alternative, more conventional (and effective) treatments (prednisolone in ITP) should be tried prior to IVIG whenever possible. • The dose (mg/kg), route, formulation, frequency and duration of treatment should be the minimum

Intravenous immunoglobulin

that is clinically effective (based partly on published trial data). Other factors such as adverse effects and compliance also need to be considered. Whenever possible, in the context of long-term treatment, home therapy or hospital day-case treatment is the most cost-effective.

Further reading Duhem C, Dicto MA, Ries F. Side-effects of intravenous immune globulins. Clin Exp Immunol 1994; 97 (Suppl. 1): 79–83. Dwyer JM. Manipulating the immune system with immune globulin (review). N Engl J Med 1992; 326(2): 107–16. Haskin JA, Warner DJ, Blank DU. Acute renal failure after large doses of intravenous immune globulin. Ann Pharmacother 1999; 33: 800–3.

Knezevic-Maramica I, Kruskall MS. Intravenous immune globulins: an update for clinicians. Transfusion 2003; 43: 1460–80. Mouthon L, Kaveri SV, Spalter SH et al. Mechanisms of action of intravenous immune globulin in immune mediated diseases. Clin Exp Immunol 1996; 104 (Suppl. 1): 3–10. Sharief MK, Ingram DA, Swash M, Thompson EJ. Intravenous immunoglobulin reduces circulating proinflammatory cytokines in Guillain–Barré syndrome. Neurology 1999; 52: 1833–8. Spellberg B. Mechanism of intravenous immune globulin therapy. N Engl J Med 1999; 341: 57–8. Yap PL. The viral safety of intravenous immune globulin. Clin Exp Immunol 1996; 104 (Suppl. 1): 35–43.

157

Part 3

Complications of transfusion

Chapter 13

Haemolytic transfusion reactions Sue Knowles and Geoff Poole

A haemolytic transfusion reaction (HTR) is the occurrence of lysis or accelerated clearance of red cells in a transfusion recipient. With few exceptions, these reactions are caused by immunological incompatibility between the blood donor and the recipient. HTRs can be classified with respect to either the time of their occurrence following the transfusion or the predominant site of red cell destruction: • acute HTRs (AHTRs) occur during or within 24 h of the transfusion; • delayed HTRs (DHTRs) typically occur 5–7 days following the transfusion; • haemolysis can be predominantly intravascular, when it is characterized by gross haemoglobinaemia and haemoglobinuria, or predominantly extravascular, when the only feature may be a fall in haemoglobin; • in general, intravascular haemolysis is seen in AHTRs and extravascular haemolysis in DHTRs.

Pathophysiology of HTRs There are three phases involved (Fig. 13.1): • antibody binding to red cell antigens, which may involve complement activation; • these opsonized red cells interacting with and activating phagocytes; and • production of inflammatory mediators. Antigen–antibody interactions

Where an immunological incompatibility is responsible, the course of the reaction depends upon:

• class and subclass (in the case of IgG) of the antibody; • blood group specificity of the antibody; • thermal range of the antibody; • number, density and spatial arrangement of the red cell antigen sites; • ability of the antibody to activate complement; • concentration of antibody in the plasma; • amount of red cells transfused. Characteristics of the antibody and antigen

The characteristics of the antibody (such as immunoglobulin class, specificity and thermal range) and of the antigen sites against which antibody activity is directed (such as site density and special arrangement) are interrelated. Antibodies of a certain specificity, from different individuals, are often found only within a particular immunoglobulin class and have similar thermal characteristics. Red cells of a certain blood group phenotype, from different individuals, tend to be relatively homogeneous regarding the attributes of the relevant antigen. It is for this reason that a knowledge of the specificity of an antibody can be highly informative in predicting its clinical significance. Three examples illustrate this. • Anti-A, anti-B and anti-A,B antibodies are regularly present in moderate to high titre in the plasma of group O persons. These antibodies are often both IgM and IgG, having a broad thermal range up to 37°C, and are often strongly complement binding. The A and B antigens are often present in large site numbers (e.g. up to 1.2 ¥ 106 A1 antigen sites per cell) and are strongly immunogenic (provoking an immune response in an individual

161

Chapter 13

Phase 1

Red cells coated with antibody

No complement bound

Complement bound

FcgR binding to monocytes, macrophages and NK cells

FcgR and CR1/CR3 binding to monocytes, macrophages and NK cells

Extravascular destruction phagocytosis fragmentation lysis (ADCC)

Extravascular destruction phagocytosis fragmentation lysis (ADCC)

Activation to MAC

Intravascular destruction—lysis

C3a

Phase 3

Phase 2

Activation to C3 only

C3a, C5a

Cytokine release

Fig. 13.1 Pathophysiology of haemolytic transfusion reaction. ADCC, antibody-dependent cell-mediated cytotoxicity; MAC,

membrane attack complex; NK, natural killer.

lacking the antigen). Anti-A, anti-B and/or anti-AB are frequently implicated in AHTRs. • Anti-Jka antibodies may be produced following immunization of a Jk(a–) person. They are usually IgG (but may also have an IgM component), are active at 37°C and may be complement binding. In Jk(a+b–) persons, there are about 1.4 ¥ 104 Jka antigen sites per cell. Jka antigens are not particularly immunogenic. However, the antibody is sometimes difficult to detect in pretransfusion testing (because of the low titre of antibody); consequently Jk(a+) blood may be inadvertently transfused to patients with pre-existing anti-Jka. These antibodies are frequently implicated in DHTRs. 162

• Anti-Lua antibodies may be produced following the immunization of a Lu(a–) person, or may be ‘naturally occurring’. They are usually IgM (but often have IgA and IgG components), are only sometimes reactive at 37°C and are not usually complement binding. The Lua antigens show variable distribution on the red cells of an individual, and are poorly immunogenic. The antibody may not be detected in pretransfusion testing, because of the fact that screening cells usually do not possess the Lua antigen and because antibody levels fall after immunization. Anti-Lua antibodies have not been implicated in AHTRs and only rarely in (mild) DHTRs.

Haemolytic transfusion reactions

Complement activation

Cytokines

Antibody-mediated intravascular haemolysis is caused by sequential binding of complement components (C1–C9). IgM alloantibodies are more efficient activators of C1 than IgG, since the latter must be sufficiently close together on the red cell surface to be bridged by C1q in order to activate complement. Activation to the C5 stage leads to release of C5a into the plasma and assembly of the remaining components of the membrane attack complex on the red cell surface, leading to lysis. Extravascular haemolysis is caused by noncomplement-binding IgG antibodies or those which bind sublytic amounts of complement. IgG subclasses differ in their ability to bind complement, with the following order of reactivity: IgG3 > IgG1 > IgG2 > IgG4. Activation of the C3 stage leads to C3b and iC3b deposition on red cells, promoting binding to two complement receptors, CR1 and CR3, which are both expressed on macrophages and monocytes, and to the release of C3a into the plasma. Hence, C3b and iC3b augment macrophagemediated clearance of IgG-coated cells, and antibodies binding sublytic amounts of complement (e.g. Duffy and Kidd antibodies) often cause more rapid red cell clearance and more marked symptoms than non-complement-binding antibodies (e.g. Rh antibodies). C3a and C5a are anaphylatoxins with potent proinflammatory effects, including oxygen radical production, granule enzyme release from mast cells and granulocytes, nitric oxide production and cytokine production.

Cytokines are generated during an HTR as a consequence of both anaphylatoxin generation (C3a, C5a) and monocyte FcgRI interaction with red cell-bound IgG. Some biological actions of cytokines implicated in HTRs are given in Table 13.1. ABO incompatibility stimulates the release of high levels of tumour necrosis factor (TNF)-a into the plasma, within 2 h, followed by interleukin (IL)-8 and monocyte chemotactic protein (MCP)1. In IgG-mediated haemolysis, TNF-a is produced at a lower level together with IL-1b and IL-6. IL-8 production follows a similar time course to that in ABO incompatibility. IgG-mediated haemolysis, as opposed to ABO incompatibility, also results in the production of Table 13.1 Cytokines implicated in haemolytic transfusion

reactions. Terminology

Proinflammatory cytokines TNF, IL-1 Fever Hypotension, shock, death Mobilization of leucocytes from marrow Activation of T and B cells Induction of cytokines (IL-1, IL-6, IL-8,TNF-a, MCP) Induction of adhesion molecules IL-6 Fever Acute-phase protein response B-cell antibody production T-cell activation Chemokines IL-8

Fc receptor interactions

IgG alloantibodies bound to red cell antigens interact with phagocytes through Fc receptors. The affinity of Fc receptors for IgG subclasses varies, with most efficient binding to IgG1 and IgG3. After attachment to phagocytes, the red cells are either engulfed or lysed external to the monocyte membrane by lysosomal enzymes excreted by the monocyte, i.e. antibody-dependent cell-mediated cytotoxicity.

Biological activity

MCP-1

Chemotaxis of neutrophils Chemotaxis of lymphocytes Neutrophil activation Basophil histamine release Chemotaxis of monocytes Induction of respiratory burst Induction of adhesion molecules Induction of IL-1

Anti-inflammatory cytokines IL-1ra Competitive inhibition of IL-1 type I and II receptors IL, interleukin; IL-1ra, IL-1 receptor antagonist; MCP, monocyte chemotactic protein;TNF, tumour necrosis factor.

163

Chapter 13

the IL-1 receptor antagonist (IL-1ra). The relative balance of IL-1 and IL-1ra may also, at least in part, account for some of the clinical differences between intravascular and extravascular haemolysis.

Antibody specificities associated with HTRs These are given, together with the site of red cell destruction, in Table 13.2.

Acute haemolytic transfusion reactions Aetiology and incidence

These reactions arise as a result of existing antibodies, in either the recipient or donor plasma, which are directed against red cell antigens of the other party. The majority of AHTRs are due to the transfusion of ABO-incompatible transfusions, predominantly red cells, but can also be due to the administration of plasma containing high titres of ABO haemolysins. ABO-incompatible transfusions are the result of the ‘wrong’ blood being given to the ‘wrong’ patient because of clerical or administrative errors, occurring at any stage during the transfusion process. The Serious Hazards of Transfusion Table 13.2 Antibody specificities associated with haemolytic

transfusion reactions. Blood group system

Intravascular haemolysis

ABO, H Rh Kell Kidd Duffy MNS Lutheran Lewis Cartwright Vel Colton Dombrock

A, B, H

164

K Jka

Extravascular haemolysis

All K, k, Kpa, Kpb, Jsa, Jsb Jka, Jkb, Jk3 Fya, Fyb M, S, s, U Lub

Lea Vel

Yta Vel Coa, Cob Doa, Dob

(SHOT) confidential reporting scheme has shown that in ABO-incompatible transfusions, 20% of errors occur in prescription, sampling and request, 29% in the transfusion laboratory and 48% in collection and administration. The reports have also highlighted that multiple errors contribute to ‘wrong blood’ incidents. Examples of reported errors from several series are given in Table 13.3. Estimates of ABO-incompatible transfusions vary and may be underestimates, since some may be unrecognized or not reported, but two recent surveys have found a frequency of approximately 1 in 30 000 transfusions. Not all ABO-incompatible transfusions cause morbidity and mortality; mortality is dependent upon the amount of incompatible red cells transfused, and is reported to be 25% in recipients receiving 1–2 units of blood and reaches 44% with more than 2 units. However, as little as 30 mL group A cells given to a group O recipient can be fatal. Less frequently, Kell, Kidd and Duffy antibodies can be responsible and the acute reaction is due to a failure to detect, or take account of, the red cell

Table 13.3 Errors resulting in ‘wrong blood’ incidents.

Prescription, sampling and request Failure to identify correct recipient at sampling Correct patient identity at sampling but incorrectly labelled sample Selection of incompatible products in an emergency Transfusion laboratory Took a correctly identified sample and aliquoted it into an improperly labelled test tube for testing Took a wrongly identified sample through testing Tested the correct sample but misinterpreted the results Tested the correct sample but recorded the results on the wrong record Correctly tested the sample but labelled the wrong unit of blood as compatible for the patient Incorrect serological reasoning, e.g. O-positive FFP to non-O-positive recipient Collection of unit Failure to check recipient identity with unit identity Bedside administration error Recipient identity checked through case notes or prescription chart, and not wristband Wristband absent or incorrect

Haemolytic transfusion reactions

alloantibody in either the antibody screen or crossmatch. Details of the incompatibilities resulting in deaths reported to the Food and Drug Administration (FDA) in the USA between 1976 and 1985 are provided in Table 13.4. In the UK, with voluntary reporting to the SHOT scheme, there have been five deaths definitely attributable to an ABOincompatible transfusion, with two further probable deaths and eight possible deaths attributable to the transfusion, between 1996 and 2002. Over the same period there have been 41 cases of major morbidity due to an ABO-incompatible transfusion and five others attributable to Kidd and Duffy antibodies. Symptoms and signs

These may become apparent after receiving as little as 20 mL of ABO-incompatible red cells. Initial clinical presentations include the following: • fever, chills or both; • pain at the infusion site, or localized to the loins, abdomen, chest or head (the aetiology is unclear, but may be related to rapid complement activation at the site of infusion, and the generation of bradykinin following complement activation); Table 13.4 Fatal acute haemolytic transfusion reactions

reported to the FDA between 1976 and 1985. Incompatibility O recipient and A red cells O recipient and B/AB red cells B recipient and A/AB red cells A recipient and B red cells O plasma to A/AB recipient B plasma to AB recipient Total ABO incompatibilities

Number of deaths 80 26 12 6 6 1 131

anti-K anti-E + K + P1 anti-Jkb anti-Jka + Jkb + Jk3 anti-Fya

5 1 1 1 1

Total non-ABO incompatibilities

9

• hypotension, tachycardia or both; • agitation, distress and confusion, particularly in the elderly; • nausea or vomiting; • dyspnoea; • flushing; and • haemoglobinuria. In anaesthetized patients, the only signs may be uncontrollable hypotension or excessive bleeding from the operative site, as a result of disseminated intravascular coagulation (DIC). These symptoms and signs can also be features of a reaction to bacterial contamination of the unit. Complications

Renal failure develops in up to 36% of patients as a result of acute tubular necrosis induced by both hypotension and DIC. Thrombus formation in renal arterioles may also cause cortical infarcts. DIC develops in up to 10% of patients. TNF-a can induce tissue factor expression by endothelial cells and together with IL-1 can reduce the endothelial expression of thrombomodulin. Thromboplastic material is also liberated from leucocytes during the course of complement activation.

Immediate management of suspected AHTR Actions for nursing staff

In the presence of a fever greater than 38°C, and/or any symptoms or signs mentioned above, the nursing staff should: 1 stop the transfusion, leaving the giving set attached; 2 use a new giving set and keep the intravenous infusion running with normal saline; 3 call a member of the medical staff; 4 check that the patient identity as provided on the wristband corresponds with that given on the label on the blood pack and on the compatibility form; 5 save any urine the patient passes; and 6 monitor the pulse, blood pressure and temperature at 15-min intervals.

165

Chapter 13

than the infusion site for the investigations listed in Table 13.6.

Actions for medical staff

The immediate actions depend upon the presenting symptoms and signs, and are summarized in Table 13.5. Investigation of suspected AHTR

Blood samples should be taken from a site other

Other reactions characterized by haemolysis

In patients with autoimmune haemolytic anaemia, transfusion may exacerbate the haemolysis and be associated with haemoglobinuria. Donor units of red cells may also be haemolysed as a result of:

Table 13.5 Immediate medical management of an acute transfusion reaction.

Symptoms/signs

Likely diagnosis

Actions

Isolated fever or fever and shivering, stable observations, correct unit given Fever with pruritus, urticaria

Febrile non-haemolytic transfusion reaction (FNHTR)

Any other symptoms/signs, hypotension, or incorrect unit

Assume to be an acute haemolytic transfusion reaction in first instance

Paracetamol 1 g orally, continue transfusion slowly, observation of pulse, blood presure and temperature every 15 min for 1 h, then hourly. If no improvement call haematology medical staff Chlorpheniramine 10 mg i.v. and other actions as for suspected FNHTR Discontinue transfusion, nornal saline to maintain urine output >1 mL/kg per h. Full and continuous monitoring of vital signs. Call haematology medical and transfusion laboratory staff immediately for further advice/action. Send discontinued unit of blood with attached giving set and other empty packs, after clamping securely, to transfusion laboratory

Allergic transfusion reaction

Table 13.6 Laboratory investigation of suspected acute haemolytic transfusion reaction.

Blood test

Rationale/findings

Full blood count Plasma/urinary haemoglobin, haptoglobin, bilirubin Blood group

Baseline parameters, red cell agglutinates on film Evidence of intravascular haemolysis

Direct antiglobin test Compatibility testing

Urea/creatinine and electrolytes Coagulation screen Blood cultures

IAT, indirect antiglobulin test.

166

Comparison of post-transfusion and retested pretransfusion samples, to detect ABO error not apparent at bedside. Unexpected ABO antibodies after transfusion may result from transfused incompatible plasma. The donor ABO group should be confirmed Positive in majority, pretransfusion sample should be tested for comparison. May be negative if all incompatible cells destroyed An IAT antibody screen and IAT crossmatch using the pretransfusion and post-transfusion sample provide evidence for the presence of alloantibody. Elution of antibody from post-transfusion red cells may aid identification of antibody, or confirm specificities identified in serum in cases of non-ABO incompatibility. Red cell phenotype should also be performed on recipient pretransfusion sample and unit in cases of non-ABO incompatibility in order to confirm absence in patient and presence in unit of corresponding antigen Baseline renal function Detection of incipient disseminated intravascular coagulation In the event of septic reaction caused by bacterial contamination of unit, which may be suspected from inspection of pack for lysis, altered colour or clots

Haemolytic transfusion reactions

• • • • • •

bacterial contamination; excessive warming; erroneous freezing; addition of drugs or intravenous fluids; trauma from extracorporeal devices; or red cell enzyme deficiency.

Management of a confirmed AHTR

1 Maintain adequate renal perfusion by: (a) fluid challenges; (b) frusemide infusion (250 mg over 4 h); and (c) if necessary, inotropic support. 2 Transfer to a high-dependency area where continuous monitoring can take place. 3 Repeat coagulation and biochemistry screens every 2–4 h. 4 If urinary output cannot be maintained at 1 mL/kg per h, seek expert renal advice. 5 Haemofiltration or dialysis may be required for the acute tubular necrosis. 6 In the event of the development of DIC, blood component therapy may be required.

Prevention of AHTRs Prevention of ‘wrong blood’ incidents

Prevention of the multiplicity of errors which can contribute to the transfusion of ABO-incompatible red cells must depend upon the creation of an effective quality system for the entire process, which will involve: • adherence to national guidelines and standards; • local procedures which are agreed, documented and validated; • training and retraining of key staff; • regular error analysis and review; • reporting to local risk management committee; • reporting to national haemovigilance schemes to contribute to the understanding of the extent and underlying causes. These aspects are specifically covered in Chapter 25. Since the majority of errors leading to an ABOincompatible transfusion are due to misidentification of the patient or patient’s sample, due attention must be paid to the comprehensive use of

unique patient identifiers throughout the hospital and automation within the laboratory. Access to previous transfusion records containing historical ABO groups should be available at all times. In the future, it is desirable that computerized systems are used to verify at the bedside the matches between the patient and the sample taken for compatibility testing, and at the time of transfusion between the patient and the unit of blood. Prevention of non-ABO AHTRs

In the case of recurrently transfused patients, due attention should be paid to the interval between sampling and transfusion in order to optimize the detection of newly developing antibodies. For patients transfused within the previous 72 h, the following pretransfusion sample should not be taken more than 24 h before the next transfusion. Pretransfusion samples should also only be kept for 3 days, if the patient has been transfused within the previous 14 days. In the presence of multiple red cell alloantibodies, and when it is not feasible to obtain compatible red cells in an emergency, intravenous immunoglobulin (1 g/kg daily for 3 days) with or without steroids (hydrocortisone 100 mg 6hourly) has been used with anecdotal reports of preventing a haemolytic episode.

Delayed haemolytic transfusion reactions Aetiology and incidence

With few exceptions, DHTRs are due to secondary immune responses following re-exposure to a given red cell antigen. The recipient has been primarily sensitized to the antigen in pregnancy or as a result of a previous blood transfusion and a few days after a subsequent transfusion there is a rapid increase in the antibody concentration, resulting in the destruction of red cells. • The antibodies most commonly implicated and reported to SHOT between 1996 and 2002 were those from the Kidd blood group system, followed by those from the Rh, Duffy and Kell systems. One analysis showed that in approximately 10% of 167

Chapter 13

reported cases, more than one alloantibody was found in the serum. • Frequently, there are no clinical signs of red cell destruction, but subsequent patient investigations reveal a positive direct antiglobulin test (DAT) and the emergence of a red cell antibody. This situation has been termed a delayed serological transfusion reaction (DSTR). • Kidd and Duffy antibodies are more likely to cause symptoms and be associated with a DHTR rather than a DSTR. • Estimates of the frequency of DHTR and DSTR vary, but in a series reported from the Mayo Clinic the frequency of DHTR was 1 in 5405 units and of DSTR was 1 in 2990 units, giving a combined frequency of 1 in 1900 units transfused. • DHTRs are in themselves rarely fatal, although in association with the underlying disease can lead to mortality. • Of transfusion fatalities reported to the FDA between 1976 and 1985, 10% were due to DHTR; in 75% of cases, more than one alloantibody was present in the serum, and the same proportion involved non-Rh antibodies. • Six deaths reported to SHOT between 1996 and 2002 have been due to DHTRs. Tragically, in some instances, there were delays in diagnosis, investigation and provision of compatible units which led to marked anaemia and contributed to mortality. Signs and symptoms

These usually appear within 5–10 days following the transfusion, but intervals as short as 24 h and as late as 21 days have been recorded. The exact onset may be difficult to define since haemolysis can be initially insidious and may only be appreciated from results of post-transfusion samples. The commonest features are: • fever; • fall in haemoglobin concentration; and • jaundice and haemoglobinuria. Hypotension and renal failure are uncommon (6% of cases). In the postoperative period in particular, the diagnosis may be overlooked and the symptoms and signs incorrectly attributed to continuing haemorrhage or sepsis. 168

Management

The majority of DHTRs require no treatment because red cell destruction occurs gradually as antibody synthesis increases. However, particularly in a bleeding patient, haemolysis will contribute to the development of life-threatening anaemia and urgent investigations are required to ensure the timely provision of antigen-negative units. Expert medical advice may be required for treatment of the hypotension and renal failure. When accompanied by circulatory instability and renal insufficiency, a red cell exchange transfusion with antigen-negative units can curtail the haemolytic process. Future transfusions of red cells should also be negative for the antigen in question. Investigation of suspected DHTR

• The peripheral blood film is likely to show spherocytosis. • Other evidence of haemolysis, namely hyperbilirubinaemia, reduced serum haptoglobin, haemoglobinaemia, haemoglobinuria and haemosiderinuria, is useful to confirm the nature of the reaction and to monitor progress. • The DAT usually becomes positive within a few days of the transfusion until the incompatible cells have been eliminated. • Further serological testing on pretransfusion and post-transfusion samples should be undertaken in accordance with the schedule provided for AHTR. • The antibody may not be initially apparent in the post-transfusion serum but can be eluted from the red cells. If the red cell eluate is inconclusive, then a repeat sample should be taken after 7–10 days, to allow for an increase in antibody titre. However, additional, more sensitive techniques may have to be employed to detect the antibody and it is advisable to seek the help of a reference laboratory. • Since a significant proportion of cases have more than one alloantibody in the serum, it is important that the panels used for antibody identification have sufficient cells of appropriate phenotypes to exclude additional specificities.

Haemolytic transfusion reactions

Prevention

Access to previous transfusion records may disclose the presence of antibodies undetectable at the time of crossmatching, and all patients should be questioned regarding previous transfusions and pregnancies. Patients found to have developed a clinically significant red cell alloantibody should be provided with an antibody card. When the care of patients requiring transfusion support is shared between hospitals, there must be adequate communication between laboratories and clinical teams. Laboratories should ensure that their antibody screen is effective in detecting weak red cell alloantibodies and that screening cells are taken from homozygotes where the corresponding antibodies show dosage.

Delayed haemolysis following organ transplantation (passenger lymphocyte syndrome) Donor-derived B lymphocytes within the transplanted organ may mount an anamnestic response against the recipient’s red cell antigens. Donorderived antibodies are usually directed against antigens within the ABO and Rh systems. Haemolysis occurs 7–10 days after transplantation, with an unpredictable and abrupt onset. In minor ABO-incompatible transplants (O donors and recipients of other groups), pretransplant isohaemagglutinin titres do not appear to predict the incidence or severity of haemolysis. In both haemopoietic stem cell and solid organ transplants, the haemolytic syndrome is almost exclusively associated with the use of cyclosporin and tacrolimus. The ex vivo removal of T cells has a similar enhancing effect on the function of transplanted donor memory B lymphocytes. Haemopoietic stem cell transplants (see also Chapter 9)

Most patients transplanted with minor ABOincompatible marrow develop a positive DAT but

only 10–15% develop clinically significant haemolysis. Haemolysis in minor ABO incompatibility is short-lived and exchange transfusion is rarely required. Plasma-containing components should be of the recipient type, and red cells group O. It has been suggested that the use of peripheral blood stem cells may increase the risk of significant haemolysis since the number of lymphocytes infused with the graft is increased, and three deaths due to AHTR have been reported between 1997 and 1999 in minor ABO-incompatible transplants. Several cases due to anti-D have been described, and antibody production has persisted for up to 1 year. Solid organ transplants

In ABO-unmatched organs, the frequency of occurrence of donor-derived antibodies and haemolysis increases with the lymphoid content of the graft, from kidney to liver to heart–lung transplants. The figures for haemolysis are 9%, 29% and 70%, respectively. The frequency of haemolysis increases with an O donor and A recipient. The ABO antibodies, which appear 7–10 days after transplant, last for approximately 1 month. Haemolysis is usually mild, although several cases of renal failure and one death have been reported. It can be prevented by switching to group O cells, either at the end of surgery or postoperatively if the DAT becomes positive. Rh antibodies have been described following kidney, liver and heart–lung transplants. They can cause haemolysis for up to 6 months, which can be sufficiently severe to merit therapy.

Haemolytic transfusion reactions in sickle cell disease The frequency of alloimmunization in sickle cell anaemia is dependent upon the nature and success of the extended red cell antigen matching policy employed. Approximately 40% of patients who are alloimmunized have experienced or will experience a DHTR. Although DHTRs are characteristically mild in 169

Chapter 13

other groups of recipients, they can be responsible for major morbidity in sickle cell disease. The term ‘sickle cell haemolytic transfusion reaction syndrome’ has been suggested to capture some of the distinctive features which can be seen to accompany a reaction. These features are as follows. • Symptoms suggestive of a sickle cell pain crisis develop or are intensified during the HTR. • Marked reticulocytopenia (for the patient). • Development of a more severe anaemia after transfusion than was present before: this may be due to the suppression of erythropoiesis as a result of the transfusion, although hyperhaemolysis of autologous red cells (bystander immune haemolysis) has been suggested. • Subsequent transfusions may further exacerbate the anaemia and it may become fatal. • Patients often have multiple red blood cell alloantibodies and may also have autoantibodies, which makes it difficult or impossible to find compatible units of red blood cells. However, in other patients no alloantibodies are identified. • Serological studies may not provide an explanation for the HTR: even red cells which are phenotypically matched with multiple patient antigens may be haemolysed. • Withholding further transfusion and corticosteroids (hydrocortisone 100 mg 6-hourly) and intravenous immunoglobulin (1 g/kg daily) has been beneficial in some cases. • It is recommended that patients with sickle cell disease are phenotyped prior to transfusion and that blood is selected for Rh and K.

Summary • HTRs are the commonest cause of immediate morbidity and mortality following a transfusion.

170

• The clinical presentations are diverse and they can be unrecognized or misdiagnosed. • Most fatal AHTRs are due to the transfusion of ABO-incompatible red cells. • The transfusion of ABO-incompatible red cells is the result of an error occurring at any stage in the transfusion process. • Devising and successfully implementing measures to overcome these preventable and fatal errors are a priority and challenge for those involved.

Further reading Beauregard P, Blajchman MA. Haemolytic and pseudohaemolytic transfusion reactions: an overview of the haemolytic transfusion reactions and the clinical conditions that mimic them. Transfus Med Rev 1994; 8: 184–99. Davenport RD. Haemolytic transfusion reactions. In: Popovsky MA, ed. Transfusion Reactions. Bethesda, MA: AABB Press, 1996: 1–44. Linden JV, Kaplan HS. Transfusion errors: causes and effects. Transfus Med Rev 1994; 8: 169–83. Mollison PL, Engelfriet CP, Contreras M. Haemolytic transfusion reactions. In: Blood Transfusion in Clinical Medicine. Oxford: Blackwell Science, 1997: 358–89. Petz LD, Calhoun L, Shulman IA, Johnson C, Herron RM. The sickle cell haemolytic transfusion reaction syndrome. Transfusion 1997; 37: 382–92. Ramsey G. Red cell antibodies arising from solid organ transplants. Transfusion 1991; 31: 76–86. Sazama K. Reports of 355 transfusion-associated deaths: 1976 through 1985. Transfusion 1990; 30: 583–90. Serious Hazards of Transfusion. Annual Report 2002–2003. Manchester: SHOT Office, 2004 (www.shotuk.org). Vamvakas EC, Pineda AA, Reisner R, Santrach PJ, Moore SB. The differentiation of delayed haemolytic and delayed serologic transfusion reactions: incidence and predictors of haemolysis. Transfusion 1995; 35: 26–32.

Chapter 14

Febrile reactions and transfusion-related acute lung injury Michael F. Murphy and Sheila MacLennan

A common adverse reaction to a blood transfusion is the development of fever, which can occur after transfusion of any type of blood component. The development of fever without haemolysis, a febrile non-haemolytic transfusion reaction (FNHTR), is relatively innocuous in itself, but it is important to differentiate it from more serious complications of transfusion, particularly acute haemolytic reactions and transfusion-transmitted bacterial infection. FNHTRs are associated with the presence of leucocytes in the blood components, and it is likely that the incidence will be significantly reduced following the implementation of universal leucocyte depletion of blood components in the UK and other countries. Transfusion-related acute lung injury (TRALI) is a severe pulmonary reaction associated with the transfusion of blood components containing donor plasma. Although TRALI occurs infrequently, it is one of the commonest causes of death associated with blood transfusion.

Febrile non-haemolytic transfusion reactions Definition

FNHTRs are febrile episodes where there is a temperature rise of 1°C or more during or soon after a transfusion, and where there is no obvious cause such as a haemolytic transfusion reaction. Incidence

of patients receiving standard non-leucocytedepleted blood components. Clinical features

FNHTRs associated with red cell transfusions occur in patients who have had previous pregnancies and/or transfusions; FNHTRs associated with platelet transfusions may occur in patients who have not had previous pregnancies and/or transfusions. Typical clinical manifestations are: • flushing; • fever; • tachycardia; and • sometimes rigors. Symptoms usually occur about 30 min to 2 h after starting a red cell transfusion, and even earlier after a platelet transfusion. In the mildest reactions, patients are febrile but asymptomatic. The temperature usually settles 2–12 h after discontinuation of the transfusion. FNHTRs are considered to be relatively innocuous as they are transient and do not lead to more serious clinical effects, although they can be uncomfortable and distressing for patients. Many variables influence the severity of clinical symptoms, including: • the number of leucocytes (red cell transfusions) or quantity of cytokines (platelet transfusions) transfused; • the speed of the transfusion; and • recipient factors such as the titre of antileucocyte antibodies.

FNHTRs have been reported to occur with an incidence as high as 6.8% after red cell and 37.5% after platelet transfusions in prospective studies 171

Chapter 14

of clinical staff of a transfusion-transmitted bacterial infection. Fever due to coincidental infection and other causes unrelated to blood transfusion should also be considered in the differential diagnosis of fever occurring during a transfusion.

Differential diagnosis

Fever is the commonest early manifestation of haemolytic transfusion reactions. Later effects include back pain, headache, shortness of breath, haemoglobinuria, shock and bleeding due to disseminated intravascular coagulation (DIC). Concern that fever may be the first indication of a haemolytic transfusion reaction is the reason why it is usual practice to discontinue transfusions associated with fever. There has been considerable debate about how to manage patients developing fever during a transfusion. • Should the blood component be returned to the blood bank? • What investigations should be carried out to exclude a haemolytic reaction? • Should the transfusion of the implicated blood component be recommenced if the investigations prove negative? Transfusion-transmitted bacterial infection is an important cause of fever during transfusion (see Chapter 16). Not all bacterially contaminated blood components cause symptoms but, if they occur, fever and chills are common, starting during or shortly after a transfusion. Subsequent symptoms may include nausea, vomiting, diarrhoea, shock, respiratory symptoms and bleeding due to DIC; such symptoms should arouse the suspicion

Pathogenesis

The pathogenesis of FNHTRs following red cell and platelet transfusions is different. Red cell transfusions

The importance of leucocytes in causing FNHTR associated with red cell transfusion was demonstrated many years ago (Fig. 14.1). The rise in temperature is directly related to the number of leucocytes transfused, and leucocyte antibodies are detectable in most patients. In one early study, the reactions were prevented in all patients with a history of recurrent FNHTRs by reducing the number of leucocytes to 0.25 ¥ 109/unit, which is approximately 10% of the number of leucocytes in a non-leucocyte-depleted red cell concentrate. FNHTRs had been assumed to be due to the reaction between recipient HLA or granulocytespecific antibodies with donor leucocytes, and the subsequent release of ‘pyrogens’, principally interleukin (IL)-1. Evidence that IL-1 is not stored in

103.0

Oral temperature (°F)

102.0 Buffy rich (fraction II) 101.0

Infusion begun

100.0 Fig. 14.1 Febrile (non-haemolytic)

99.0 Buffy poor (fraction I) 98.0 0

172

1

2 3 4 5 Hours after beginning of infusion

6

7

8

transfusion reaction after transfusion of blood containing the buffy coat (fraction II), and no reaction after transfusion of buffy coat-poor blood (fraction I). (From Brittingham & Chaplin 1957 with permission.)

Febrile reactions and TRALI

leucocyte granules but is synthesized and released following cell activation suggests that FNHTRs are not mediated by pyrogens released from donor leucocytes. An alternative explanation is that recipient antibodies bind to donor leucocyte antigens and fix complement, and the resulting antigen–antibody–complement complexes activate recipient macrophages to release pyrogens. The release of recipient pyrogens in response to immune complexes formed by recipient antibody and donor antigen may serve as a model for fever following the transfusion of incompatible red cells and platelets as well as leucocytes. Platelet transfusions

The major cause of FNHTRs after platelet transfusion is the presence of pyrogenic cytokines, such as IL-1, IL-6 and tumour necrosis factor, released from leucocytes during the 5 days of platelet storage. A role for cytokines or other mediators in the pathogenesis of FNHTRs following platelet transfusion is supported by the following observations: • cytokines reach very high levels during storage of platelet concentrates, suggesting that they are synthesized and not just leaked from leucocytes; • when stored platelet concentrates are separated into plasma and cellular portions and then transfused, the plasma portion is associated with the majority of reactions; • there is a strong correlation between the concentration of cytokines in the plasma of stored platelet concentrates and the occurrence of FNHTRs; • FNHTRs after transfusions of stored platelet concentrates are not reliably prevented by bedside filtration; • increases in cytokine levels have not been found in platelet concentrates leucocyte depleted before storage. Management

The incidence of FNHTRs after red cell and platelet transfusions is now much reduced following the implementation of universal leucocyte depletion of blood components. This means that it is even more important not to ignore the occur-

rence of fever during transfusion as it may be the first manifestation of a serious complication of transfusion, e.g. a haemolytic transfusion reaction associated with a red cell transfusion, or bacterial contamination of a platelet concentrate. Mild febrile reactions (rise in temperature of 1–1.5°C but without severe symptoms such as rigors, lumbar pain) occurring during a red cell transfusion should be managed by temporarily stopping the transfusion. A clerical check should be carried out to ensure that the intended patient has received the correct unit of blood. A sample should be sent to the blood bank to: • check the blood group of the patient; • examine the supernatant plasma for haemoglobinaemia, which is facilitated by comparing it with the plasma of the pretransfusion sample; and • carry out a direct antiglobulin test on the posttransfusion sample. These investigations should take no longer than 1 h. Symptomatic measures, such as keeping the patient warm and administering aspirin or paracetamol to reduce the fever, may be used while the investigations are being carried out. Paracetamol rather than aspirin should be used in patients with thrombocytopenia. If the results of the investigations have been completed within an hour and are negative, and the patient’s symptoms are settling, the transfusion can be recommenced in the knowledge that a haemolytic reaction has been excluded. Mild febrile reactions associated with platelet transfusions can be managed by slowing the transfusions and administering paracetamol. The transfusion should be completed if there is no progression of symptoms. Where there has been a severe febrile reaction (rise in temperature >1.5°C with or without other symptoms) to a red cell or platelet transfusion, the same unit should not be restarted. Investigations should be carried out for a haemolytic transfusion reaction. Bacterial contamination of the unit should also be suspected, and the unit sent for microscopy and culture along with a sample of blood from the patient. Prevention

For the prevention of FNHTRs in patients having 173

Chapter 14

recurrent FNHTRs after the transfusion of standard red cell concentrates before the implementation of universal leucocyte depletion of blood components, a reduction in the number of leucocytes in future transfusions was recommended. Leucocyte depletion to less than 5 ¥ 106 leucocytes/unit was not usually necessary to prevent FNHTRs; a reduction in the number of leucocytes to 5 ¥ 108 leucocytes/unit was sufficient in most cases. This was achieved most cost-effectively using buffy coat-depleted red cell concentrates, if they were available, or by filtration at the bedside. Leucocyte-depleted red cell concentrates are effective in the prevention of FNHTRs in patients dependent on long-term transfusion support, e.g. patients with b-thalassaemia major. For the prevention of FNHTRs after platelet transfusions, the introduction of routine use of pooled platelets derived from buffy coats in the UK reduced the rate of FNHTRs to 3–5%, and the use of platelet concentrates leucocyte depleted prior to storage is now associated with an even lower incidence (possibly <1%). This allows platelet transfusions to be given with little concern about the occurrence of FNHTRs even in multiply transfused patients. Pre-storage leucocyte depletion has been routine for some time in Leiden, Netherlands; in addition, most of the plasma is removed by centrifugation before platelet transfusions, and the platelets are resuspended in 20 mL of residual plasma. This practice is associated with a very low incidence of FNHTRs and urticarial reactions (<0.5%), enabling the transfusion to be given within 30 min including the collection of a sample for a post-transfusion platelet count 10 min after the transfusion.

Transfusion-related acute lung injury Definition

TRALI is a severe acute reaction characterized by respiratory distress, hypoxia and pulmonary infiltrates soon after transfusion with no other apparent cause. In many cases, preformed leucocyte antibodies have been found in the plasma of the donors, and these are thought to cause pul-

174

monary leucostasis and leucocyte activation resulting in capillary leak and pulmonary damage. Incidence

TRALI is considered to be a rare complication of blood transfusion, but this may be due to underrecognition and under-reporting. In the 1997–98 Serious Hazards of Transfusion (SHOT) report (see Chapter 25), nine cases were fully reported, giving an approximate incidence of 1 in 250 000 units transfused; in the subsequent four SHOT reports a further 30 cases were described. The SHOT report of 2001–2 described a significantly increased number of cases (30 in the one year). It is likely that this increase represents an improved awareness of the condition rather than a true increase in incidence, a finding supported by evidence reported from a hospital in the USA with a particular interest in TRALI of an incidence of 0.02% (1 in 5000 units of blood transfused). Clinical features

TRALI is characterized by: • acute respiratory distress occurring within 6 h of starting a transfusion; • severe bilateral pulmonary oedema; • severe hypoxia; • fever; and • chest X-ray typically shows perihilar and nodular shadowing in the mid and lower zones (Fig. 14.2). Male and female patients are equally affected, and TRALI has been reported in all age groups. Most patients have no history of previous transfusion reactions. It had been thought that there were no common underlying clinical conditions in patients affected by TRALI. However, a recent study found that all patients had a predisposing clinical condition such as active infection, cytokine administration, recent surgery or massive transfusion; it has been proposed that a predisposing clinical condition may be required for the development of TRALI. Data from SHOT suggests that the most common underlying diagnoses which may predispose patients to develop TRALI are

Febrile reactions and TRALI

common after transfusion of plasma-rich components such as: • whole blood; • fresh frozen plasma (FFP); • cryoprecipitate; and • platelet concentrates. Differential diagnosis

(a)

(b) Fig. 14.2 Chest X-rays of a patient with transfusion-related

acute lung injury (a) 1 day before a platelet transfusion and (b) shortly after transfusion showing diffuse bilateral shadowing of the lungs and a normal-sized heart. (From Virchis et al. 1997 with permission.)

haematological malignancy and elective surgery, including cardiac surgery. TRALI has been reported to be associated with all blood components (including one report after intravenous immunoglobulin), but is more

TRALI is clinically indistinguishable from acute respiratory distress syndrome (ARDS), and the occurrence of ARDS with a possible association with transfusion provides grounds to consider TRALI as the cause. Unlike ARDS, TRALI is self-limiting and tends to result in no long-term sequelae, and there is usually clinical improvement within 48–96 h provided prompt respiratory support has been provided. In one series 70% of patients required mechanical ventilation. Cardiogenic causes of pulmonary oedema should be ruled out. In contrast to patients with cardiogenic pulmonary oedema, patients with TRALI have normal central venous pressure and normal or low pulmonary wedge pressure. The identification of HLA and/or granulocytespecific antibodies in an implicated donor or plasma-containing blood component, and which are reactive with the patient’s leucocytes, provides support for the diagnosis. Female donors with a history of previous pregnancies and male donors with a history of transfusion should be screened for HLA and granulocyte-specific antibodies. If antibodies are detected, the donor plasma should be tested for reactivity against the patient’s lymphocytes and neutrophils. Donor leucocyte antibodies are not detected in about 10% of cases of TRALI, suggesting that other mechanisms may be involved in its pathogenesis. Pathogenesis

Leucocyte (HLA and/or granulocyte-specific) antibodies in donor plasma are found in nearly 90% of cases of TRALI; donor leucocyte antibodies were detected in seven of the nine cases of TRALI in the 1997–98 SHOT report. Multiparous women are usually the source of the antibody-containing

175

Chapter 14

plasma. In a small proportion of cases the antibody is identified in the recipient’s plasma, and the corresponding antigen has been transfused in the component. The antibodies are thought to react with the patient’s leucocytes, causing pulmonary leucostasis and complement-mediated leucocyte activation, resulting in endothelial damage in pulmonary capillaries through release of proteolytic enzymes and toxic oxygen metabolites from neutrophils (Fig. 14.3). However, the transfusion of HLA and/or granulocyte-specific antibodies does not account for all cases of TRALI.

It has been shown that during storage of cellular blood components, lipids are generated which rapidly prime neutrophil oxidases. Animal models of ARDS have suggested that neutrophil activation in pulmonary vessels is involved in its pathogenesis. Given the clinical similarity between ARDS and TRALI, it is now proposed that TRALI is a two-step process, involving: • a predisposing clinical condition that causes release of cytokines (and possibly other factors) which prime patients’ neutrophils and cause them to become attached to vascular endothelium, particularly in pulmonary vessels; and

(a)

Fig. 14.3 Thin sections of fixed lung

from a patient with transfusionrelated acute lung injury. There is acute diffuse alveolar damage with intra-alveolar oedema and haemorrhage. There was no histological evidence of infection, and all postmortem cultures (bacterial, viral and fungal) were negative. Magnification: a, ¥40; b, ¥440. (From Silliman et al. 1997 with permission.)

(b)

176

Febrile reactions and TRALI

• lipids, and possibly cytokines, and/or leucocyte antibodies from the plasma of stored blood cause further neutrophil priming, neutrophil activation and pulmonary damage. Management

Prompt and vigorous respiratory support is essential. Intubation, oxygenation and/or mechanical ventilation is frequently necessary. Steroids are sometimes used, but are probably of marginal value, and diuretics are not indicated. TRALI usually improves clinically within 48–96 h, and pulmonary infiltrates recede within 1–4 days. However, in approximately 20% of patients, hypoxia and pulmonary infiltrates persist for at least 7 days. For those recovering rapidly, there are no long-term sequelae. Mortality is approximately 5%, although death occurred in two of the nine cases reported by SHOT in 1997–98. Prevention

Blood donors with HLA and/or granulocytespecific antibodies implicated in cases of TRALI should not be used again as donors. Additional measures have been proposed: • exclusion of multiparous women as donors of whole blood, FFP and platelet concentrates prepared by apheresis (these blood components all contain 200–300 mL of plasma); • testing multiparous women for HLA and granulocyte-specific antibodies before accepting them as donors of plasma-rich components; • pooling of plasma, which would theoretically reduce the risk of TRALI because of the dilution of leucocyte antibodies; and • removal of plasma from cellular blood components. Each of these measures may be more suited to some components than others. In the UK, a review is being undertaken of which risk reduction measure is most applicable to each type of component, and it is anticipated that some steps to decrease the incidence of TRALI will be taken in the near future, such as selection of male donor plasma for FFP production. Wholesale withdrawal

of female donors who have had three or more pregnancies is not feasible as it is estimated that this would decrease blood collection by approximately 5%, but ensuring that these donors are not used for plasma-rich components is likely to decrease the risk of TRALI. Pooling of plasma theoretically increases the risk of transfusion-transmitted infection, but could be used in combination with pathogen-reduction techniques. Removal of all plasma from blood components, such as red cell and platelet concentrates, would be logistically complex and costly but would have the potential to reduce other complications of transfusion such as FNHTRs and the risk of transmission of variant Creutzfeldt–Jakob disease (see Chapter 20). Leucocyte depletion of blood components may reduce the risk from the small number of cases of TRALI thought to be caused by the presence of leucocyte antibodies in the recipient, but will not affect the majority of cases caused by the transfusion of donor leucocyte antibodies.

Further reading Boshkov LK. Transfusion-associated acute lung injury (TRALI): an evolving understanding of the role of antileucocyte antibodies. Vox Sang 2002; 83 (Suppl. 1): 299–303. Brand A. Passenger leucocytes, cytokines, and transfusion reactions. N Engl J Med 1994; 331: 670–1. British Committee for Standards in Haematology. Guidelines on the clinical use of leucocyte-depleted blood components. Transfus Med 1998; 8: 59–71. British Committee for Standards in Haematology. Guidelines for the use of platelet transfusions. Br J Haematol 2003; 122: 10–23. Brittingham TE, Chaplin H. Febrile transfusion reactions caused by sensitivity to donor leukocytes and platelets. J Am Med Assoc 1957; 165: 819–25. Dzik WH. Is the febrile response to transfusion due to donor or recipient cytokine? Transfusion 1992; 32: 594. Dzik WH, Anderson JK, O’Neill EM et al. A prospective, randomized clinical trial of universal WBC reduction. Transfusion 2002; 42: 1114–34. Heddle NM, Klama LN, Griffith L, Roberts R, Shukla G, Kelton JG. A prospective study to identify the risk factors associated with acute reactions to platelet and red cell transfusions. Transfusion 1993; 33: 794–7. Heddle NM, Klama L, Singer J et al. The role of plasma

177

Chapter 14 from platelet concentrates in transfusion reactions. N Engl J Med 1994; 331: 625–8. Heddle NM, Blajchmann MA, Meyer RM et al. A randomized controlled trial comparing the frequency of acute reactions to plasma-removed platelets and prestorage WBC-reduced platelets. Transfusion 2002; 42: 556–66. Kopko PM, Holland PV. Transfusion-related acute lung injury. Br J Haematol 1999; 105: 322–9. Muylle L, Joos M, Wouters E, de Bock R, Peetermans ME. Increased tumour necrosis factor alpha (TNF-alpha), interleukin 1, and interleukin 6 (IL-6) levels in the plasma of stored platelet concentrates: relationship between

178

TNF-alpha and IL-6 levels and febrile transfusion reactions. Transfusion 1993; 33: 195–9. Popovsky MA, Chaplin HC Jr, Moore SB. Transfusionrelated acute lung injury: a neglected, serious complication of haemotherapy. Transfusion 1992; 32: 589–92. Silliman CC, Paterson AJ, Dickey W et al. The association of biologically active lipids with the development of transfusion-related acute lung injury. Transfusion 1997; 37: 719–26. Virchis AE, Patel RK, Contreras M et al. Acute noncardiogenic lung oedema after platelet transfusion. Br Med J 1997; 314: 880–2.

Chapter 15

Urticarial and anaphylactic reactions David J. Unsworth

Allergic reactions are not infrequent complications of transfusion. For example, data from the Serious Hazards of Transfusion (SHOT) scheme indicate that of 478 incidents reported between 1996 and 2002, 48 were cases of acute transfusion reaction; of these, fresh frozen plasma (FFP), platelets or both were implicated in 31 cases and of these, 26 were allergic or anaphylactic reactions. The majority, however, were associated with little or no morbidity. In other studies, platelet components have been the commonest cause of allergic/anaphylactic reactions, and there is speculation that plateletderived membrane microparticles (also detectable in FFP) are involved. These negatively charged particles may activate complement. Solvent–detergent treatment removes the particles, and has been reported to be associated with fewer acute reactions. Allergic reactions may also complicate autologous transfusion. Over a 5-year period in a single large institute in the USA, reactions were reported in 15 of 9353 (0.16%) patients receiving preoperative donations and in 5 of 18 506 (0.027%) receiving intraoperative salvaged blood. These reactions accounted for 2% of all reported transfusionassociated reactions in that institute over the 5year period. Of the 20 reactions, eight seemed to be unrelated to the transfusion per se and four were acute/allergic.

Anaphylaxis This is caused by mast cell degranulation due to cross-linking of mast cell surface IgE by specific antigens. Hence, anaphylaxis can only occur once

prior exposure has allowed generation of specific IgE antibody. This term tends to be reserved for life-threatening reactions (Table 15.1). Passive transfer of specific IgE or allergen by transfusion is another cause of anaphylaxis. Food antigens, for example, can enter the bloodstream and persist for several hours after food, and even tiny quantities could provoke anaphylaxis. Reactions associated with transferred specific IgE would be more likely following FFP and other immunoglobulin/allergen-rich blood products. However, there is very little evidence for anaphylaxis by this mechanism from detailed clinical reports. Following an apparent allergic reaction, it is worth asking the recipient (and the donor, if traceable) about a history of severe food allergy to determine whether this could be the cause. ‘Allergy’ (type 1 IgE-mediated immediate hypersensitivity) transfer to previously non-allergic individuals has also been noted following bone marrow or liver transplantation. In one case, 3 months following combined liver and kidney transplantation, the previously non-allergic recipient developed urticaria and laryngeal oedema after eating peanuts. The donor had actually died of peanut anaphylaxis. Although the recipient became skin-prick sensitive to peanut (implies possession of IgE antibody), the time course is too long for passively transferred specific IgE antibody. DNA from recipient skin biopsy samples confirmed microchimerism, and transfer of allergen-specific donor T lymphocytes seems more likely. In some cases following bone marrow transplantation, the donor is non-allergic, pointing to altered immunoregulation in the recipient. 179

Chapter 15 Table 15.1 Life-threatening allergic

Anaphylaxis

Anaphylactoid

IgE-mediated mast cell/basophil degranulation Requires previous exposure/sensitization Life-threatening: due to release of pharmacologically active mast cell constituents Sudden onset Clinical features vary, but mortality associated with laryngeal oedema, bronchospasm, cardiac arrhythmias, and hypovolaemic shock due to increased vascular permeability

IgE independent Can occur on first exposure Clinically indistinguishable from true anaphylaxis

Anaphylactoid This is clinically indistinguishable from an anaphylactic episode and is also life-threatening. The pharmacology is the same, being due to global, major release of mast cell constituents. However, the pathogenesis differs, and IgE, by definition, is not involved. Any other mechanism, be it immunological or non-immunological, leading to sudden massive global mast cell degranulation can lead to an ‘anaphylactoid’ reaction. In contrast to true anaphylaxis, anaphylactoid reactions may occur on first exposure. For example, IgG anti-IgA antibodies produced by IgA-deficient individuals exposed to blood products containing normal donor IgA may lead to generation of IgG/IgA immune complexes, with complement activation. The complement fragments C3a and C5a are ‘anaphylotoxins’ that act directly at dedicated mast cell surface receptors, triggering degranulation, an example of an immunologically based anaphylactoid reaction. In contrast, anaphylactoid reactions seen after administration of iodine-containing intravenous radiocontrast media arise by a direct chemical (non-immunological) effect on mast cells, and thus can occur on first exposure.

Mild allergic reactions These are common in the context of blood transfusion, and include local cutaneous reactions (urticaria or angio-oedema) or wheeze, or combinations thereof. Again, the cause may be 180

reactions.

immunologically based (IgE or other) or nonimmunological. It is misleading to call these reactions ‘mild anaphylaxis’.

Patient management Even with a mild reaction: • stop the transfusion; and • review and treat the patient according to the guidelines in Table 15.2. Treatment is according to the severity of the reaction, and a matter of clinical judgement.

Investigation of the cause of the transfusion reaction Table 15.3 lists some general considerations in a patient receiving blood and suspected of an allergic transfusion reaction. Appropriate samples should be obtained for future investigation (Fig. 15.1). • Time 0 sample (i.e. sample immediately prior to Table 15.3 Allergic reactions associated with transfusion.

Transfusion cause Antibody to transfused blood cells (anti-HLA, platelet, other) Recipient deficient in plasma protein (e.g. IgA, complement) Recipient/donor allotypic differences (IgG allotypes, complement C4d, Chido, haptoglobin) Other cause Latex allergy Drug associated (asprin,ACE inhibitor, other)

Urticarial and anaphylactic reactions Table 15.2 Management of allergic reactions.

Grade

Reaction

Symptoms/signs

Treatments

1 2

Non-systemic Mild systemic

3

Moderate systemic

4

Severe systemic (anaphylaxis)

Focal urticaria/angio-oedema Chest tightness (wheeze), generalized urticaria/ angio-oedema Wheeze/breathlessness/obstructive laryngeal oedema Severe breathing difficulty; shock; arrhythmias; loss of consciousness

Antihistamine (p.o.) (chlorpheniramine 4 mg) Antihistamine (p.o.), salbutamol and/or inhaled steroid, e.g. prednisolone 20 mg (p.o.) All above including prednisolone (p.o.) or hydrocortisone (i.v.); consider adrenaline (i.m.) Adrenaline (i.m.) and all others listed above

Note that in a life-threatening reaction, be it anaphylactic or anaphylactoid, adrenaline (epinephrine) i.m. without delay is the key treatment. Use 1 in 1000 strength (concentration 1 mg/mL) at a dose of 0.01 mg/kg body weight. Intravenous adrenaline should only be considered if intramuscular treatment has failed, and will require slow cautious infusion of a more dilute solution (1 in 10 000), while monitoring for possible cardiac arrhythmias.

Possible allergic transfusion reaction

1 Stop infusion 2 Treat (see Table 15.2)

1 Full clinical history 2 Obtain blood samples (see above)

Complement tests (C3, C4, C3a, C5a)

Check mast cell tryptase

Check recipient IgA

Fig. 15.1 Investigation of the cause of

IgE anti-latex

If IgA < 0.05 mg/dL check for anti-IgA

an allergic transfusion reaction.

transfusion, if available) of recipient’s blood (serum and plasma samples), and further timed samples (1 h, 3–6 h and 12–15 h). • Sample from the recipient used for pretransfusion testing. • Sample of the infused (donor) blood product. Iatrogenically haemodiluted recipient samples (either after transfusion or as a consequence of attempted resuscitation) should be avoided and measurement of an uninvolved serum protein (e.g. IgG total) should be carried out. A low C4 is more likely to be genuine, pointing to possible immune complex-type reactions, if the IgG is normal. Similarly, a very low IgG and normal mast cell tryptase

(MCT) may be dilutional and mask a true elevation in the level of tryptase. The choice of tests critically depends on the type of reaction. Thus by definition the detection of IgE antibodies is useful, for example in anaphylaxis caused by exposure to latex, but is useless if what was observed was an anaphylactoid process (e.g. anaphylatoxin driven). False-negative tests are avoided if samples taken before the transfusion or 1 month after the event are evaluated. Investigation is further complicated because adverse reactions temporally associated with transfusion may not be triggered by the blood products per se, but rather by some other 181

Chapter 15

concurrent aspects of patient treatment or parallel drug (antibiotics, etc.) administration. Consider, for example, a patient with thrombotic thrombocytopenic purpura receiving red cell transfusion while undergoing plasma exchange. After 1 h, marked angio-oedema and a tendency to hypotension develop. Unless a full history is taken, the fact that the patient has been taking an angiotensin-converting enzyme (ACE) inhibitor for long-standing hypertension may be overlooked. It is now appreciated that ACE inhibitors block kininase II, allowing accumulation of plasma kinins and the risk of an anaphylactoid episode. Certain synthetic membranes, used for example in renal dialysis, blood filtration and plasma exchange, trigger kinin generation, particularly in patients taking ACE inhibitors. This all occurs independently of the immune system and does not involve mast cells/basophils. Latex allergy is an increasing problem. Healthcare professionals could plausibly provoke an IgE-mediated response, mild or anaphylaxis, in presensitized patients. Wheezing (powdered gloves especially are associated with aerosols of latex) or local contact urticaria while delivering a blood product are the most likely. MCT assays

In severe life-threatening allergic reactions, with a mast cell or basophil pathogenesis (either anaphylactoid or true anaphylaxis), MCT is typically markedly but transiently elevated in the bloodstream (Table 15.4). False-positive test results are a rarity, and all positive results merit detailed further investigation to identify the cause and prevent

recurrence. The only slight drawback is that MCT has a short half-life (~90 min) and thus diagnostic samples will only be of use if taken very early (ideally within the first 3 h) in the time course. Immunoassays are commercially available. IgA antibodies

Though relative (partial) IgA deficiency is common (1 in 400 in the UK), most of these individuals are not at risk of transfusion reactions. In contrast, those with profound IgA deficiency (< 0.05 mg/dL) are (Table 15.5). The same principle extends to other plasma protein deficiencies (a1-antitrypsin, IgG subclasses, haptoglobin and others). Indeed, at least in Japan, haptoglobin deficiency is far more common than IgA deficiency as a cause of immune-mediated acute reaction. AntiIgA may be of IgE or IgG isotype, but most assays in use only detect IgG. Tests for anti-IgA antibodies of IgG isotype

A range of assays are available including enzymelinked immunosorbent assay (ELISA) and passive haemagglutination. They tend to use purified myeloma IgA paraprotein as antigen and, unless specifically designed to detect IgE antibody, they favour the detection of IgG antibody. Note that this approach detects IgG anti-IgA in approximately 1 in 200 individuals with a total IgA below 0.05 mg/dL. This contrasts with the observed rarity of severe transfusion reactions and suggests that these assays are oversensitive. False-negative results are less of a problem. Approximately 75% of IgA-deficient cases with a documented transfusion reaction test positive.

Table 15.4 Mast cell tryptase. Table 15.5 IgA deficiency, anti-IgA antibody and

Mast cell/basophil specific Very low plasma levels normally (< 5 mg/L) Short half-life (< 2 h) Immunoassays commercially available Requires fresh or frozen (–20°C) serum High levels strongly suggest that anaphylaxis/anaphylactoid reaction has occurred (> 20 mg/L within first hour) Suggest four serial samples (pretransfusion, 1 h, 3–6 h, 12–15 h)

182

anaphylaxis. Rare: between 1 in 20 000 and 1 in 47 000 routine transfusions Highly unlikely unless IgA is < 0.05 mg/dL Majority of routine immunoassays detect IgG (rather than IgE) anti-IgA, and overestimate the problem 1 in 200 IgA-deficient individuals have IgG anti-IgA, though few of these react when given IgA-containing blood products

Urticarial and anaphylactic reactions

Blood products for at-risk IgA-deficient individuals

Those with a previously documented transfusion reaction associated with IgA deficiency and antibodies to IgA need specially prepared blood products as below. • Predonated autologous blood can be provided for an elective procedure. • Red cells can be supplied after thorough washing to remove contaminating plasma protein, including IgA. • Platelets are more difficult to wash with retention of therapeutic efficacy. Better to obtain platelets from a compatible IgA-deficient donor source.

Further reading Fisher M. Treatment of acute anaphylaxis. Br Med J 1995; 311: 731–3. Gilstad WG. Anaphylactic transfusion reactions. Curr Opin Hematol 2003; 10: 419–23. Laroche D, Vergnaud MC, Sillard B, Soufarapis H, Bricard H. Biochemical markers of anaphylactoid reactions to drugs. Anesthesiology 1991; 75: 945–9. Owen HG, Brecher ME. Atypical reactions associated with angiotensin-converting enzyme inhibitors and apheresis. Transfusion 1994; 34: 891–4. Sandler GS, Mallory D, Malamut D, Eckrich R. IgA anaphylactic transfusion reactions. Transfus Med Rev 1995; 9: 1–8. Serious Hazards of Transfusion. 2003. Annual Report 2001–2002. Serious Hazards of Transfusion, Manchester, UK (www.shotuk.org). Wibaut B, Mannessier L, Horbez C et al. Anaphylactic reactions with anti-Chido antibody following platelet transfusions. Vox Sang 1995; 69: 150–1.

183

Chapter 16

Bacterial contamination Patricia E. Hewitt

Bacterial contamination is a long-established and well-recognized complication of blood transfusion. Many factors contributed to this problem in the early days of blood transfusion, including open preparation techniques, reuse of equipment and lack of refrigerated storage facilities. Despite the disappearance of all these contributory factors, bacterial transmission has not been eliminated and remains a significant problem in terms of mortality and morbidity. Indeed, bacterial contamination is the most common microbiological complication of transfusion medicine, and the current estimated risks of bacterial contamination of blood components in Europe exceed the risks of human immunodeficiency virus (HIV), hepatitis C virus (HCV) and hepatitis B virus (HBV) contamination combined. Bacterial contamination of blood components is responsible for more immediate morbidity and mortality than is transfusiontransmitted viral infection. Since 1995, bacterial contamination has been responsible for 29 of 50 (58%) transfusion-transmitted infections reported to the Serious Hazards of Transfusion (SHOT) scheme in the UK. Bacterial infection caused seven of nine deaths due to transfusion-transmitted infections in that period. Platelets were implicated in 25 of 29 cases of bacterial infection and the platelets were 3 days old or more in 23 of 25 cases. Staphylococcus epidermidis was isolated in 9 of 25 platelet cases. The French Blood Agency haemovigilance surveillance system attributed 16 deaths to bacterially contaminated blood components between 1994 and 1998, and 15.9% of all transfusion-related fatalities in the USA in the years 1986–91 were accounted for by this problem. There is certainly no evidence that the problem of bacterial contamination is disappear184

ing. Furthermore, this is not a risk that can be avoided by the use of autologous transfusion.

Incidence of bacterial contamination Although serious clinical sequelae of bacterial contamination after red cell transfusion are relatively infrequent, there is a much higher incidence after platelet transfusion. This is mainly due to the different storage requirements of platelets, at a temperature (22°C) that supports bacterial survival or growth. In contrast, red cells are stored at 4°C, a temperature which does not support the growth of the majority of bacteria. Nevertheless, contamination of red cells does occur and can (rarely) lead to serious consequences. The prevalence of serious episodes is probably in the order of 1 in 5000 for platelet units and 1 in 500 000 (or less) for red cell units. Rarely, bacterial transmissions from cryoprecipitate and fresh plasma units have been reported, due to contamination of the water baths used to thaw frozen units. Recipients of platelet transfusions, particularly those who require a prolonged period of support after chemotherapy or bone marrow transplantation, are exposed to significant risk during this period, especially if receiving pooled platelet preparations. • The incidence of contamination assessed by routine bacteriological surveillance of red cell and platelet preparations has been reported as 0.4% in studies from Canada and Germany. • Pooled platelet preparations have higher contamination rates than single-donor (apheresis) platelets, as would be predicted. In a retrospective study at the National Blood Service, North London Centre, the contamination rate of time-

Bacterial contamination

expired (i.e. >5-day-old) pooled platelets was 0.7% compared with 0.4% for apheresis platelets. The incidence of reactions in recipients of pooled platelets was 1 in 4200 platelet transfusions in a study from the USA and 1 in 2100 platelet transfusions in a prospective study in Hong Kong. In the latter study, each recipient was regularly evaluated for possible signs of a transfusion reaction, and observed reactions were investigated with bacterial cultures. • The incidence of reported reactions in transfusion recipients is much less than predicted from the figures obtained from routine surveillance monitoring of contamination rates of blood components or from the prospective clinical follow-up study from Hong Kong. Reasons for this discrepancy are discussed below. Possible explanations for apparent infrequency of clinical events

• Non-pathogenic bacteria. • Insufficient numbers of bacteria to cause clinical sequelae. • Patients premedicated with steroids/antipyretics, which mask the typical signs of a reaction. • Patients already taking antibiotics effective against the contaminating bacteria. • Patients are immunosuppressed and expected to have infections, and episodes are therefore underinvestigated/reported. Factors contributing to frequency of bacterial contamination of platelets

• Storage conditions of platelets: at 22°C for 5 days. • Many platelet preparations are pooled; pooling increases risk by increasing donor exposure.

Other species (14.3%)

Pseudomonas putida (4.1%) Yersinia enterocolitica (51%)

Treponema pallidum (4.1%)

Pseudomonas fluorescens (26.5%)

Fig. 16.1 Bacterial species associated with sepsis from

erythrocyte transfusions.

• Other Gram-negative rods. A number of factors contribute to Y. enterocolitica contamination of red cells: • the bacterium can grow well at 4°C; • donor leucocytes phagocytose living microorganisms in vivo and release them into the blood when the leucocytes disintegrate; and • potent endotoxin is formed under the conditions of red cell storage. Platelets (Fig. 16.2)

• Coagulase-negative staphylococci. • Serratia marcescens. • Streptococci. • Bacillus cereus. • Gram-negative rods. The contaminants reported as causing clinical reactions in platelet preparations differ from those reported following routine bacteriological monitoring of platelets (Fig. 16.3). This is largely accounted for by non-pathogenic bacteria detected during routine monitoring.

Types of bacterial contaminants Red cells (Fig. 16.1)

• Yersinia enterocolitica (accounts for approximately 50% of reports). • Pseudomonas fluorescens. • Serratia liquefaciens.

Sources of bacterial contamination It is not always possible to identify the source of bacterial contamination, but cases should be carefully investigated, to cover the following areas: 185

Chapter 16

Salmonella cholerae-suis (13.5%) Serratia marcescens (9.6%) Staphylococcus epidermidis (25.0%)

Other species (36.5%)

Staphylococcus aureus (5.8%) Bacillus cereus (5.8%) Streptococcus viridans group (3.8%)

Fig. 16.2 Bacterial species associated with sepsis from

platelet infusions.

Other (13.3%) Coliforms (3.9%) Diphtheroids (25%)

Staphylococcus epidermidis (25%)

• This is an uncommon source of bacterial contamination of blood components, since most bacteraemic individuals would not be fit enough to donate blood. • Exceptions are usually due to ‘episodic bacteraemia’ in well individuals, for example transient bacteraemia following dental treatment or associated with a chronic low-grade infection such as osteomyelitis due to Salmonella. • The other main source of bacteraemia in ‘well’ individuals is that associated with enteric infections, particularly Y. enterocolitica and, less commonly, Campylobacter jejuni and some Salmonella. These infections may give rise to mild non-specific symptoms of abdominal pain or diarrhoea, which would not necessarily be considered relevant in a healthy individual presenting as a blood donor. To exclude all donors with such symptoms in the period immediately preceding blood donation would result in the loss of significant numbers of blood donations that would present no risk at all.

Pseudomonas sp. (3.9%) Enterobacter cloacae (5.9%) Staphylococcus aureus (7.2%)

Fig. 16.3 Bacteria isolated following monitoring of platelet

products.

• donor bacteraemia: • skin contaminants on the donor’s arm (the most common source); • contamination of the pack or its contents at source (faulty manufacturing): and • environmental contamination during collection, processing or storage (faulty heat seals, pin-hole defects, etc.).

186

Donor bacteraemia

Skin contamination of donor arm

• Inadequate skin cleaning, due to heavy bacterial contamination at the venepuncture site, the presence of bacterial spores not sensitive to the disinfectant used, or faulty technique. • Scarring of the venepuncture site, more common in regular apheresis donors, preventing effective surface skin cleaning. • Entry of a small core of skin from the venepuncture site into the phlebotomy needle and into the collection bag. The core may harbour bacteria that would not be affected by surface cleaning of the skin. Contamination of pack and/or contents

• Case reports have previously cited contaminated anticoagulant fluid within the collection bag as the source of bacterial contamination. • Other suggested routes of contamination have included grossly contaminated outer surface of

Bacterial contamination

packs/tubing, leading to contamination of the operator’s hands, and contaminated vacuum tubes that were then used to collect samples from the primary pack after the collection was complete, leading to ‘retrograde’ contamination of the pack contents. Environmental contamination

• Since sterile closed systems are used, blood contamination during storage and handling is thought to be very rare. • Contamination of water baths used to thaw units of cryoprecipitate or fresh frozen plasma may lead to entry of organisms into the pack, possibly through pin-hole defects or cracks in the plastic.

Clinical features of bacterial contamination The symptoms of a reaction due to bacterial contamination usually appear immediately, during transfusion of the implicated unit (Table 16.1). Less often, symptoms are delayed until after the end of the transfusion. The most commonly reported symptoms and signs are: • fever, temperature elevation usually greater than 2°C; • chills, rigors; • hypotension, collapse, shock; • nausea, vomiting; and • disseminated intravascular coagulation, intravascular haemolysis, renal failure. In anaesthetized patients, the clinical signs may be masked. Severe reactions are usually the result of endotoxins produced by contaminating bacteria during the storage period of the component.

Immediate management • Stop the transfusion. NB retain all packs for investigation. • Give general supportive treatment, as required, which may include intravenous fluids, inotropic agents and diuretics to maintain urine output.

Table 16.1 Transfusion reaction due to bacterial

contamination: clinical example. Clinical condition 48-year-old male; acute myeloid leukaemia Admitted for isolation during neutropenic phase following fourth course of induction chemotherapy Infected Hickman line site causing intermittent pyrexia Pancytopenic: Hb 9.7 g/dL,WBC 0.3 ¥ 109/L; platelet count 23 ¥ 109/L Management Blood cultures and skin swabs taken; broad-spectrum antibiotics started Uneventful transfusion of 2 units of red cells with an additional platelet transfusion Continued intermittent pyrexia over following 24 h, Hickman site infection not controlled Blood cultures repeated; started on second-line antibiotics Continuing pancytopenia, but clinically stable Further red cell and platelet transfusion prescribed Collapsed during platelet transfusion with: copious diarrhoea tachycardia hypotension (blood pressure 70 mmHg systolic) epistaxis rigors; temperature 37°C Treatment: transfusion discontinued, resuscitation with plasma expanders, blood and faeces obtained for culture 90 min later was alert, clinically recovered, blood pressure 120/80 mmHg, pulse rate 88/min Provisional diagnosis: severe platelet reaction or sepsis or haemolytic reaction. Platelet packs referred for culture 6 h after reaction developed haematuria; Hb 3.7 g/dL, platelet count 66 ¥ 109L. Sample taken for investigation of possible haemolytic transfusion reaction 18 h after reaction developed dyspnoea, oliguria, hypoxia 21 h after reaction developed petechial haemorrhage, peripheral oedema, thrombocytopenia, worsening hypoxia. Clinical diagnosis is septicaemia Transferred to intensive care for respiratory and renal support; commenced on benzylpenicillin/metronidazole/ciprofloxacin Gram stain of remnants of platelet pack shows Gram-positive rods and Gram-negative organisms Cultures of recipient’s blood and contents of platelet pack yielded Clostriduim perfringens. Further investigation revealed presence of toxin Progressive multiorgan failure. Died 9 days after reaction Cause of death: shock due to overwhelming septicaemia caused by infected platelet transfusion (toxin mediated)

187

Chapter 16

• Give broad-spectrum antibiotics until the results of blood cultures are known. • Assess need for intensive care. • Record carefully and accurately all actions taken.

Investigations Patient

• Blood count, coagulation studies, urine output and blood cultures. Blood pack

• Perform Gram stain for immediate evidence of bacterial contamination. This investigation is relatively insensitive, but any contamination leading to clinical symptoms will be significant and therefore readily detected. • Culture of pack contents, taking care not to contaminate the pack during sampling. • Inform the blood centre that supplied the pack as soon as bacterial contamination is suspected, as other components of the same blood donation may not yet have been transfused. • Blood centres will generally require the pack to be returned for further investigation, including examination for pack defects, etc. Source of contamination

• The blood centre will wish to investigate the possible origin of any contamination by a systematic procedure addressing all the possible sources. • The identity of the contaminating organism may indicate the likely source of the contamination and the priorities for investigation.

Measures to reduce the risk of bacterial contamination

Yersinia infection in donors are unlikely to be effective, since the symptoms are non-specific. Blood collection

• Avoidance of obviously scarred venepuncture sites. • Attention to skin cleaning techniques, including evaluation of both the disinfectant solution and the technique of cleaning by monitoring of effectiveness through pre- and post-cleaning swabbing of the skin, training and regular retraining of staff. • The use of collection devices that allow rejection of the first 20–30 mL of the donation or which diverts this volume into a sampling device. There is evidence that this approach decreases the prevalence of contamination. Component storage times

• Bacterial proliferation in platelet preparations exhibits a lag phase of 24–48 h. The incidence of significant levels of contamination increases steeply after 3 days of storage. • In the mid-1980s, availability of improved plastics for the manufacture of platelet packs led to an extension of the shelf-life of platelets in the USA from 5 to 7 days. A dramatic increase in the number of severe reactions due to bacterial contamination led to a rapid return to the 5-day limit, which has not subsequently been changed. • For red cells, where Yersinia is the major offending organism, the current additive solutions and plastics used in packs allow a shelf-life of 35 days. Yersinia exhibits a lag phase of 21 days, after which bacterial numbers and endotoxin production increase dramatically. Reduction of the shelflife of red cells to 21 days would be expected to significantly reduce the problem, but would seriously jeopardize the availability of red cells for patients.

Donor selection procedures

• Exclusion of donors at risk of bacteraemia, particularly those who have had recent dental treatment (within 24 h of donation). • Questions to identify possible symptoms of 188

Component storage temperatures

The storage temperature of platelets is a significant factor in the risk of bacterial contamination sufficient to cause clinical symptoms. If the storage

Bacterial contamination

temperature could be reduced, then the risk of contamination would be reduced. Unfortunately, storage of platelets at lower temperatures significantly reduces their haemostatic function and viability. Pretransfusion screening of components

There is evidence that the prevalence of significant contamination rises dramatically after 3 days of storage (for platelets) and 21 days (for red cells). It has been suggested that components stored for longer than these periods should be cultured before issue/transfusion. There are various methods of detecting contamination, some of which (visual inspection, Gram staining) are relatively insensitive. There are automated techniques that might be appropriate for this situation, such as the BacT/Alert system (Organon Technika, Durham, NC). Other suggested strategies include ribosomal and polymerase chain reaction (PCR) assays. There are many unanswered questions, such as when to sample, how to sample (without increasing the chance of contaminating the pack contents) and timing in relation to issue from the blood centre or prior to transfusion of the recipient. Conversely, sterility verification of platelet units could allow an extension of the current shelflife of platelets to 7 days, which would relieve the supply problems often encountered, and there would be a potential gain additional to patient safety. This approach has been introduced into routine use in the Netherlands and is being evaluated in the UK with a view to its introduction in the future. Leucocyte filtration

Phagocytosis appears to play an important role in the elimination of viable bacteria which might be present in a unit of blood. There is experimental evidence that leucocyte filtration of cellular blood components reduces the level of bacterial contamination, and that such filtration should be carried out at least 8 h after collection of the blood to allow sufficient time for phagocytosis to take place. There have been no prospective studies to determine the clinical efficacy of this strategy.

Photochemical decontamination of cellular blood components

Much development work is being carried out on the use of photodynamic methods for inactivation of bacteria. Such methods will also be effective against viruses and protozoa, and therefore present a real attraction in terms of blood component safety (see Chapter 23). Photodynamic methods include irradiation with ultraviolet B light, methylene blue, phthalocyanines, merocyanine 540 and a combination of psoralen and irradiation with ultraviolet A light. Data on the use of psoralen-treated platelets in animal models indicate that in vitro platelet function is maintained, as is in vivo platelet recovery and survival. Although other potential agents have been studied with similar results, clinical trials are clearly needed.

Summary Unlike many other infectious risks of blood transfusion, bacterial contamination has continued to be a significant problem in recent years. Current initiatives, including leucocyte filtration, diversion of the first 20–30 ml of the donation and the investigation of potential methods for decontamination, are expected to have an impact in reducing serious reactions due to bacterial contamination. However, care must be taken not to ignore the primary problem, which is the source of contamination. Since the most common source is probably skin contamination of the donor arm, concentration on the basic procedure of skin cleaning prior to venepuncture and on evaluation of the effectiveness of such cleaning is a vital quality standard, which must not be ignored.

Further reading Blajchman MA. Transfusion-associated bacterial sepsis: the phoenix rises yet again (editorial.) Transfusion 1994; 34: 940–2. Blajchman MA, Ali AM. Bacteria in the blood supply: an overlooked issue in transfusion medicine. In: Nance SJ, ed. Blood Safety: Current Challenges. Bethesda: American Association of Blood Banks, 1992.

189

Chapter 16 Goldman M, Blajchman MA. Blood productassociated bacterial sepsis. Transfus Med Rev 1991; 1: 73–83. Hogman CF. Serious bacterial complications from blood

190

components: how do they occur? (Editorial) Transfus Med 1998; 8: 1–3. Sazama K. Bacteria in blood for transfusion: a review. Arch Pathol Lab Med 1994; 118: 350–65.

Chapter 17

Post-transfusion purpura Michael F. Murphy

In 1959 van Loghem and colleagues described a 51-year-old woman who developed severe thrombocytopenia 7 days after elective surgery. The thrombocytopenia did not respond to transfusion of fresh blood but there was spontaneous recovery after 3 weeks. The patient’s serum contained a strong platelet alloantibody, which enabled the description of the first human platelet antigen (HPA) (Zw, see Chapter 5). However, the relationship of platelet alloimmunization to post-transfusion thrombocytopenia was not recognized until 2 years later when Shulman and colleagues studied a similar case, naming the antibody anti-PlA1 (later shown to be the same as anti-Zw), and coined the term ‘post-transfusion purpura’ (PTP).

of the Serious Hazards of Transfusion (SHOT) scheme during which approximately 20 million blood components were transfused, giving an approximate incidence of 1 in 465 000 transfusions. Since the introduction of universal leucocyte depletion of blood components in the UK in 1999, there has been a reduction in the annual number of reported cases from about 11 to 3. The low incidence of PTP relative to the 2.5% of the population who are HPA-1a negative and at risk of the condition raises the question of individual susceptibility. As in neonatal alloimmune thrombocytopenia (NAIT), the antibody response to HPA-1a is strongly associated with a certain human leucocyte antigen (HLA) class II type (HLA-DR3*0101) (see Chapter 5).

Definition Clinical features PTP is an acute episode of severe thrombocytopenia occurring about a week after a blood transfusion. It usually affects HPA-1a-negative women who have previously been alloimmunized by pregnancy. The transfusion precipitating PTP causes a secondary immune response, boosting the HPA-1a antibodies, although the mechanism of destruction of the patient’s own HPA-1a-negative platelets remains uncertain.

Incidence PTP is considered to be a rare complication of transfusion. Over 200 cases had been reported in the literature up to 1991. However, this may not reflect the true incidence of PTP, which is not known; 43 cases were reported in the first 6 years

PTP typically occurs in middle-aged or elderly women (mean 57, range 21–80 years), although it has also been reported in a small number of males. All patients, apart from rare exceptions, have had previous exposure to platelet antigens through pregnancy and/or transfusion. The interval between pregnancy and/or transfusion and PTP is variable, the shortest being 3 years and the longest 52 years. The initial maternal sensitization to platelet antigens during pregnancy in females subsequently developing PTP is rarely of sufficient degree to cause NAIT. Blood components implicated in causing PTP are: • whole blood; • packed red cells; and • red cell concentrates. 191

Chapter 17

There are two reports of PTP following the transfusion of plasma. Severe thrombocytopenia and bleeding usually occur about 5–12 days after transfusion; shorter or longer intervals are rare. The onset is usually rapid, with the platelet count falling from normal to less than 10 ¥ 109/L within 12–24 h. Haemorrhage is very common and sometimes severe. There is typically widespread purpura and bleeding from mucous membranes and the gastrointestinal and urinary tracts. In many cases the precipitating transfusion has been associated with a febrile non-haemolytic transfusion reaction, probably due to the presence of HLA antibodies stimulated by previous pregnancy and/or transfusion. Megakaryocytes are present in normal or increased numbers in the bone marrow and coagulation screening tests are normal in uncomplicated PTP. In untreated cases the thrombocytopenia usually lasts between 7 and 28 days, although it occasionally persists for longer.

Laboratory investigations A preliminary diagnosis of PTP on clinical grounds needs to be confirmed by the detection of plateletspecific alloantibodies. The majority (80–90%) of cases of PTP are associated with the development of HPA-1a antibodies in HPA-1a-negative patients. Antibodies against HPA-1b, HPA-3a, HPA-3b, HPA-4a, HPA-5a, HPA-5b, Gova and Naka have been associated with PTP, and occasionally multiple antibodies are present, e.g. antiHPA-1a, anti-HPA-2b and anti-HPA-3a were found in one case. HLA antibodies are often present in patients with PTP. There is no evidence that they are involved in causing PTP but their presence complicates the detection of platelet-specific antibodies. Modern platelet serological techniques, such as the monoclonal antibody immobilization of platelet antigens assay, are useful for resolving mixtures of antibodies in patients with PTP (see Chapter 5).

Pathophysiology Differential diagnosis The rapid onset of severe thrombocytopenia in a middle-aged or elderly woman should arouse suspicion of PTP and a history of recent blood transfusion should be sought. The differential diagnosis includes other causes of acute immune thrombocytopenia such as the following: • autoimmune thrombocytopenia; • drug-induced thrombocytopenia, e.g. heparininduced thrombocytopenia; • non-immune platelet consumption, e.g. disseminated intravascular coagulation and thrombotic thrombocytopenic purpura; • a less likely possibility is passively transfused platelet-specific alloantibodies from an immunized blood donor when thrombocytopenia occurs within the first 48 h after the transfusion; • pseudothrombocytopenia due to ethylenediamine tetra-acetic acid (EDTA)-dependent antibodies should be excluded in any patient with unexplained thrombocytopenia by examination of the blood film. 192

The time course of events in PTP is shown in Fig. 17.1. A blood transfusion triggers a rapid secondary antibody response against HPA-1a, and there is acute thrombocytopenia about a week after the transfusion. It is difficult to understand why the patient’s own HPA-1a-negative platelets are destroyed. There remains no generally accepted mechanism to explain this, although a number of suggestions have been made. • Transfused HPA-1a-positive platelets release HPA-1a antigen, which is adsorbed on to the patient’s HPA-1a-negative platelets, making them a target for anti-HPA-1a. Support for this hypothesis comes from observations such as the elution of anti-HPA-1a from HPA-1a-negative platelets in some cases of PTP, and the demonstration of the adsorption of HPA-1a antigen on to HPA-1anegative platelets after incubation with plasma from HPA-1a-positive stored blood. • The released HPA-1a antigen forms immune complexes with anti-HPA-1a in the plasma and the immune complexes become bound to the patient’s platelets, causing their destruction.

Post-transfusion purpura

Blood transfusion

Anti-HPA-1a

Platelet count

transfusion purpura. Purpura and severe thrombocytopenia occurred 5–10 days after a blood transfusion. The diagram indicates the secondary antibody response of anti-HPA-1a, and the postulated transient appearance of free HPA-1a antigen in the plasma, which binds to HPA-1anegative platelets, HPA-1a/anti-HPA1a immune complexes, platelet autoantibodies or cross-reacting HPA1a antibodies.

Transient appearance of: 1 Free HPA-1a 2 HPA-1a/anti-HPA-1a immune complexes 3 Autoantibodies, or 4 Cross-reacting anti-HPA-1a

0

• The transfusion stimulates the production of platelet autoantibodies as well as anti-HPA-1a. Evidence in favour of this mechanism is the detection of positive reactions of some PTP patients’ sera from the acute thrombocytopenic phase with autologous platelets. • In the early phase of the secondary antibody response, anti-HPA-1a may be produced which has the ability to cross-react with autologous as well as allogeneic platelets.

Management Immediate treatment is essential as the risk of fatal haemorrhage is greatest early in the course of PTP. In a review of 71 cases of PTP, five died within the first 10 days because of intracranial haemorrhage. The main aim of treatment is to prevent severe haemorrhage by shortening the duration of severe thrombocytopenia. No randomized controlled trials of treatment for PTP have been carried out. Comparison of various therapeutic measures is complicated because it may be difficult to differentiate a response to treatment from a spontaneous remission in individual cases. High-dose intravenous IgG (IVIgG) (2 g/kg given over 2 or 5 days) is the current treatment of choice, with responses in about 85% of cases; there is often a rapid and prompt increase in the platelet count (Fig. 17.2). Steroids and plasma exchange were the preferred treatments before the

5

Antibody titre

Platelet count

Fig. 17.1 A typical time course of post-

10 Days

availability of IVIgG, and plasma exchange in particular appeared to be effective in some but not all cases. Platelet transfusions are usually ineffective in raising the platelet count but may be needed in large doses to control severe bleeding in the acute phase, particularly in patients who have recently undergone surgery before there has been a response to high-dose IVIgG. There is no evidence that platelet concentrates from HPA-1a-negative donors are more effective than those from random donors in the acute thrombocytopenic phase; the dose of platelets may be more important than the platelet type of the donor platelets. There is no evidence to suggest that further transfusions in the acute phase prolong the duration or severity of thrombocytopenia. Platelet transfusions have been reported to cause severe febrile and occasionally pulmonary reactions in patients with PTP; these were probably due to HLA antibodies reacting against leucocytes in non-leucocyte-depleted platelet concentrates.

Prevention of recurrence of PTP Recurrence of PTP has been reported. However, it is unpredictable and has usually occurred following a delay of 3 years or more after the first episode. The patient should be issued with a card to indicate that he/she has previously had PTP and ‘special’ blood is required for future transfusions. 193

Chapter 17

15

500

12

400

9

300

6

200

3

100

0

0 –5

0

5

10 15 20 25 30 35 40 45

160

Days

Future transfusion policy should be to use red cell and platelet concentrates from HPA-compatible donors or autologous transfusion. If these are not available, leucocyte-depleted blood components are considered to be safe. There have been occasional reports of recurrence of PTP with leucocyte-poor red cell concentrates but the implicated components would not have complied with current standards for leucocyte depletion.

Further reading Becker T, Panzer S, Maas D et al. High-dose intravenous immunoglobulin for post-transfusion purpura. Br J Haematol 1985; 61: 149–55. Berney SI, Metcalfe P, Wathen NC, Waters AH. Posttransfusion purpura responding to high-dose intravenous IgG: further observations on pathogenesis. Br J Haematol 1985; 61: 627–32. Borne AEG Kr von dem, Plas-van Dalen CM van der.

194

286

Fig. 17.2 Haematological course of a

Platelets x 109/L

Hb g/dL

Prednisolone (mg) 60 0 Transfusion Plasmapheresis Sandoglobulin (30 g)

patient with post-transfusion purpura showing the onset of profound thrombocytopenia 6 days after a blood transfusion. Initial treatment with random platelet concentrates caused rigors and bronchospasm, and there was no platelet increment. There was no response to prednisolone (60 mg/day) or plasma exchange (2.5 L/day for 3 days), but there was a prompt remission following high-dose intravenous IgG (30 g/day for 3 days). (Redrawn from Berney et al. 1985 with permission.)

Further observations on post-transfusion purpura. Br J Haematol 1986; 62: 374–5. Kickler TS, Ness PM, Herman JH, Bell WR. Studies on the pathophysiology of post-transfusion purpura. Blood 1986; 68: 347–50. Loghem JJ van, Dorfmeijer H, Hart M van der, Schreuder F. Serological and genetical studies on a platelet antigen (Zw). Vox Sang 1959; 4: 161–9. Mueller-Eckhardt C. Post-transfusion purpura. Br J Haematol 1986; 64: 419–24. Shulman NR. Post-transfusion purpura: clinical features and the mechanism of platelet destruction. In: Nance SJ, ed. Clinical and Basic Science Aspects of Immunohaematology. Arlington, VA: American Association of Blood Banks, 1991: 137–54. Shulman NR, Aster RH, Leitner A, Hiller MC. Immunoreactions involving platelets. V. Post-transfusion purpura due to a complement-fixing antibody against a genetically controlled platelet antigen. A proposed mechanism for thrombocytopenia and its relevance in ‘autoimmunity’. J Clin Invest 1961; 40: 1597–620. Waters AH. Post-transfusion purpura. Blood Rev 1989; 3: 83–7.

Chapter 18

Immunomodulation and graft-versus-host disease Lorna M. Williamson and Cristina V. Navarrete

Immunomodulatory effects of transfusion There is an increasing body of evidence that transfusion results in both temporary and long-standing changes in immune function in the recipient. These effects may be stimulatory or suppressive depending on a number of factors related to the blood donor, the component transfused and the immune status of the recipient. Stimulation of the patient’s immune system is exemplified by alloimmunization to red cell, platelet or HLA antigens. Suppressive effects mainly affect cell-mediated immunity, including: • transient reduction in the CD4 : CD8 ratio of circulating T cells; • reduced natural killer cell function; • impaired lymphocyte mitogenic responses; and • suppression of delayed hypersensitivity. The precise conditions that determine the immunological outcome are not known but suppressive effects are thought to be mediated primarily by donor leucocytes, although plasma factors may also play a role. Different models provide evidence for: • clonal deletion of alloreactive T cells; • induction of suppressor T cells or other immunoregulatory cells in the recipient leading to anergy; • production of idiotypic antibodies; • generation of veto cells in the donor, causing T-cell inactivation or death upon presentation of a major histocompatibility complex (MHC)– antigen; • inhibitory effects of supernatants of stored red cells on the function of neutrophils; • an immunosuppressive effect of soluble HLA class I molecules.

Soluble HLA (sHLA) class I and class II molecules are present in the serum of healthy individuals. Liver is the main source of sHLA found in circulation. High levels of these molecules have been found in the serum or plasma of patients with a variety of conditions, including inflammatory, autoimmune and infectious diseases and transplanted patients. sHLA are also found in blood and blood components in direct proportion to the number of cells present and to the length of storage. The biological significance of these molecules has not been fully established, although it has been reported that they may be involved in the downregulation of the immune response and/or induction of tolerance. sHLA molecules have been shown to induce apoptosis of both alloreactive and Epstein–Barr virus (EBV)-specific CD8+ cytotoxic T lymphocytes. At the clinical level, both beneficial and adverse effects of transfusion-induced immunosuppression have been described (Table 18.1). Renal transplantation

The beneficial effect of prior transfusion on renal allograft survival was first described more than 20 years ago, and was confirmed by animal data and clinical experience worldwide. It became standard policy in many renal units to deliberately expose patients on transplant waiting lists to one or more transfusions. The beneficial effect of pretransplant transfusion was thought to be less important in the era of cyclosporin and other potent immunosuppressive drugs and many renal units discontinued its use. However, recent publications have confirmed improved 1-year and 5-year renal allograft survival in previously transfused patients. The 195

Chapter 18 Table 18.1 Clinical diseases in which an

immunomodulatory effect of transfusion has been proposed. Beneficial effect Before renal transplantation to reduce graft rejection Crohn’s disease: fewer relapses Adverse effect Increased tumour recurrence, particularly colorectal Increased postoperative infection and prolonged hospital stay

benefit is greatest when there is one HLA haplotype or one HLA-DR shared between blood donor and recipient, but there is also an advantage for HLA-matched sibling allografts. The role of transfusions deliberately mismatched for one HLA haplotype or one HLA-DR antigen is a current topic of investigation. The exact mechanism of the beneficial effects of transfusion in renal allografting is unknown. Possible explanations include the following. • Transfusion selects immunological nonresponders (patients remaining HLA-antibody negative after transfusion), who are inherently likely to show less graft rejection. However, this cannot explain the benefit seen in HLA-compatible grafts. • Deletion of alloreactive T-cell clones is no longer thought to be the primary explanation, while the post-transfusion production of suppressor T cells is only a temporary effect. • One longer-term phenomenon is the presence of anti-idiotypic antibodies, but their role in the promotion of renal allograft survival is unclear. • Observations from transfusions where there is HLA-DR or haplotype sharing suggest recipient downregulation. This could be mediated via cytokines, as animal data have shown that transfusion leads to induction of T-helper type 2 (Th2) responses, with production of interleukin (IL)-4 and IL-10, which favour humoral immunity. Simultaneously, Th1 responses important for cellmediated immunity are downregulated. • More recent studies have also proposed that longer-term immunosuppression might be mediated by the establishment of a state of permanent microchimerism involving stem cells present in the transfusion. 196

• Finally, plasma present in transfused components may play a part. A temperature-sensitive plasma factor has been shown to suppress mitogen-induced T-cell responses in vivo. However it is mediated, the transfusion effect in renal allografting appears to require viable leucocytes, in that patients awaiting renal transplants derive less immunological benefit from transfusions which are leucocyte depleted, washed or frozen/thawed. However, the need to include leucocytes in the transfusion exposes patients to the risk of HLA sensitization. Before erythropoietin was available to treat the anaemia of chronic renal failure, many patients had received multiple transfusions by the time the transplant was performed. It was observed that the beneficial effects of transfusion were not seen in patients who had received more than 10 transfusions, possibly because of the presence of HLA cytotoxic antibodies. Such antibodies in the patient may result in a positive cytotoxic crossmatch with potential kidney donors and greatly limit the possibility of finding a compatible organ. Alternatively, HLA-sensitized patients might have an increased risk of rejection episodes following transplantation. The impact of the policy of universal leucocyte depletion of blood components being introduced in a number of countries on potential renal transplant recipients remains to be seen.

Autoimmunity Crohn’s disease

Because Crohn’s disease is immunologically mediated, and because many patients require both surgery and transfusion, a number of groups have examined whether postsurgical relapse might be reduced in patients who have been transfused at the time of primary surgery. Five retrospective studies performed during the 1980s gave conflicting data, with two studies showing a significant benefit of transfusion, and three showing none. A large prospective randomized trial would be needed to answer this question definitively. At present, systematic transfusion is not practised as part of the therapy of Crohn’s disease.

Immunomodulation and GVHD

Recurrent spontaneous abortion

Recurrent spontaneous abortion (RSA) occurs in approximately 1% of couples, and in approximately half of the cases the cause is unknown. During pregnancy, there is maternal recognition of the fetus, as demonstrated by the presence in the mother’s serum of antibodies against paternal antigens. However, generally this does not lead to immunological rejection of the fetus, suggesting that normal mechanisms of allograft rejection are suppressed. Initial studies suggested an increased frequency of parental HLA haplotype sharing in RSA families, but this was not confirmed in later studies. However, women with RSA have reduced cellular reactivity against paternal HLA antigens in the mixed lymphocyte reaction (MLR). This may be due to blocking antibodies, which can inhibit the MLR response against paternal and third-party cells, and which are commonly seen in such patients. Based on these observations, women with RSA have been treated by infusions of their partner’s or third-party leucocytes in order to induce immunological tolerance of the fetus. Although the majority of leucocyte infusion programmes have been uncontrolled open studies, successful pregnancy outcome has been reported in 50–90% of cases. The precise mechanisms responsible for the immunomodulatory effect are poorly understood but could shed light on the immunomodulatory effect of transfusion in general. The use of intravenous IgG for the treatment of unexplained RSA has been assessed in a number of studies. However, the results have shown no conclusive evidence of the efficacy of this treatment and a well-controlled multicentre study is required to determine whether intravenous IgG may have a role in the treatment of patients with RSA. Tumour recurrence

A number of well-controlled experimental studies in animals have demonstrated the deleterious effect of allogeneic transfusion on tumour growth. This effect appears to be immunologically mediated, in that the tumour-promoting effect can be adoptively transferred using spleen cells. In addi-

tion, the transfusion effect is reduced by the prestorage (but not poststorage) removal of leucocytes from the transfusion. Over 70 clinical studies have examined the possible effect of transfusion on recurrence of a number of cancers. Colorectal cancer is the single tumour most commonly examined, with over 30 published studies. More than 20 of them have been retrospective analyses of the effect of perioperative transfusion on either tumour relapse or overall patient survival. Approximately two-thirds of studies demonstrated an adverse effect of transfusion, with the remainder showing neither detriment nor benefit. Two separate meta-analyses of relevant studies have both suggested a real, independent effect of transfusion on colorectal cancer recurrence, with a cumulative odds ratio for cancer recurrence of 1.8 in one review and a 37% increased risk in the other. Prospective randomized studies have examined the possible benefit of either autologous or leucocyte-depleted blood on an assumed transfusion effect in colorectal cancer (Table 18.2). Both studies confirmed the adverse effect of transfusion but failed to show any benefit from either autologous or leucocyte-depleted blood. There are inherent difficulties in designing such studies, which must correct for other variables, such as anaemia and blood loss, that might lead to greater use of transfusions but which might also influence tumour outcome. The effect of transfusion in other cancers has been examined in over 40 studies, with approximately half showing an adverse effect and the remainder showing none. This must be separated from the possible beneficial role of transfusion in correcting anaemia and thus improving oxygen delivery to tumours prior to radiotherapy. This separate effect appears to be particularly important in head and neck cancers, so transfusion should not be withheld from anaemic patients with such tumours because of fears of promoting tumour recurrence. Postoperative infection

The evidence for transfusion as an independent risk factor for the types of infection commonly 197

Chapter 18 Table 18.2 Examination of the impact of either autologous or leucocyte-depleted transfusions on recurrence of colorectal

cancer. 1

2

Transfusion randomization

Allogeneic

Autologous

Buffy coat depleted

Leucocyte depleted

Number of patients

236

239

360

337

Disease-free survival

56% at 4 years 62% at 4 years Untransfused = 73%*

64% at 3 years

Cancer-specific survival

64% at 4 years 68% at 4 years Untransfused = 88%‡

Overall 3-year survival, 73% Overall 3-year survival, 70% Untransfused = 84%†

60% at 3 years Untransfused = 81%†

* P = 0.01 compared with patients receiving either type of transfusion. † P = 0.001 compared with patients receiving either type of transfusion. ‡ P = 0.04 compared with patients receiving either type of transfusion. In both studies, there was no effect of type of transfusion on outcome.

Table 18.3 Retrospective analyses of

Study number:

1

2

3

4

Type of surgery Number of centres Number of patients in study Infection rates Not transfused Allogeneic transfusion Autologous transfusion

Spinal, hip, knee 2 376

Spinal 1 102

All types 1 100*

Hip replacement 1 84

4.9% 6.9%† 5.0%

4.0% 20.8%‡ 3.3%

— 16.0%§ 4.0%

— 32.0%¶ 3.0%

postoperative infection in relation to transfusion.

* Paired study design. † No difference between the three groups. ‡ P = 0.01 compared with autologous or no transfusion. § P < 0.05. ¶ P = 0.0029.

seen after elective surgery is increasing. The studies shown in Table 18.3, in both clean orthopaedic and potentially contaminated abdominal surgery, have all demonstrated significantly fewer infections in non-transfused patients in the postoperative period. This effect is potentially important from a health economic viewpoint since infection is an important predictor of duration of hospital stay, which is in turn a major determinant of operative costs. Since the effect is assumed to be mediated via allogeneic leucocytes, the effect of transfusions which are either autologous or leucocyte depleted has been compared with standard blood in a number of randomized controlled trials. 198

In addition, both Canada and the UK, in which universal leucocyte depletion of blood components has been implemented, have performed very large cohort ‘before and after’ studies to see whether the results of smaller randomized trials hold true in routine practice. Such studies have challenging design issues to ensure that all confounding risk factors for infection are considered. These include risk factors for local infection such as urinary catheters, endotracheal tubes and wound drains. Definitions of infection can also be difficult since clinical signs alone may have other causes, while not all apparent infections have positive bacterial cultures. The evidence for a benefit on postopera-

Immunomodulation and GVHD

tive infection has been conflicting. Although the Canadian study found that leucocyte depletion appeared to reduce mortality, this was not confirmed across a number of studies by a metaanalysis. The current position appears to be that if any benefit of leucocyte depletion exists, it seems to be small in comparison to other patient- and procedure-related factors which predispose to postoperative infection. At present, no country that still practises leucocyte depletion only for selected patients recommends elective surgery as a specific indication for leucocyte-depleted blood.

Transfusion-associated graft-versus-host disease Perhaps surprisingly, transfusion-associated graftversus-host disease (TA-GVHD) emerged as the commonest transfusion-related cause of death reported to the UK Serious Hazards of Transfusion scheme in its first 3 years, with 12 reports, all fatal. This serves as a reminder that although the incidence of TA-GVHD is low compared with other

transfusion complications, there is currently no effective treatment. Although this condition has become even rarer since universal leucocyte depletion was implemented in 1999 (Table 18.4), it remains an important complication of transfusion. The classical features of skin rash, diarrhoea and deranged liver function are almost universally followed by bone marrow hypoplasia, pancytopenia and death from infection within 3–4 weeks of transfusion. Although some novel approaches are being taken towards therapy, mortality in established cases is likely to remain close to 100% for the foreseeable future. Identification of risk factors, such as immune impairment in the patient, and HLA haplotype sharing between donor and patient, allows preventive measures to be taken. TA-GVHD can be prevented by gammairradiation of cellular blood components to a dose of 25 Gy in high-risk situations as defined below. Pathogenesis: alloreactive lymphocytes and the cytokine response

The graft-versus-host reaction was first described in mice in the 1950s, following the observation

Table 18.4 Cases of TA-GVHD

reported to the UK haemovigilance scheme between 1996 and 2002. Universal leucocyte depletion was introduced in the UK during 1999.

Year

No cases

Diagnoses

Shared haplotype

Outcome

1996–97

4

Congenital immunodeficiency No risk factors B-cell non-Hodgkin’s lymphoma B-cell non-Hodgkin’s lymphoma

Yes NK NK NK

Died Died Died Died

1997–98

4

Waldenström’s macroglobulinaemia B-cell non-Hodgkin’s lymphoma Cardiac surgery Immune thrombocytopenic purpura

NK* Yes Yes NK

Died Died Died Died

1998–99

4

Myeloma Uncharacterized immunodeficiency Cardiac surgery Cardiac surgery

NK NK NK Yes

Died Died Died Died

1999–00

0

2000–01

1

B-cell acute lymphoblastic leukaemia

NK

Died

2001–02

0

* Donor homozygous. NK, not known.

199

Chapter 18

that infusion of marrow from parent to F1 hybrid caused a syndrome of erythema, rash and jaundice, with death from wasting and infection, but marrow transfer from F1 hybrid to parent did not. The requirements for a GVHD reaction were defined by Billingham as: • the graft must contain immunologically competent cells; • the host must possess important transplantation alloantigens lacking in the donor graft, with the host able to be an antigenic stimulus to the graft; and • the host must be at least temporarily incapable of mounting an effective immunological reaction against the graft, such that donor cells have an opportunity to engraft. To a large extent, clinical descriptions of TAGVHD have fulfilled Billingham’s criteria. TAGVHD was first described in children with severe combined immunodeficiency and, in the following years, virtually every case described had some degree of predisposing immunodeficiency or immunosuppression. The recognition that the condition could also occur in immunocompetent patients came from Japanese reports of TA-GVHD in patients transfused ‘fresh’ blood for cardiac surgery. The predisposing factor in these cardiac patients was the sharing of an HLA haplotype between donor and recipient. This is much more likely to occur by chance in countries such as Japan where there are relatively few HLA haplotypes in the population. Transfusion of HLA homozygous blood to an individual possessing the same haplotype appeared to present a particular risk. The third requirement, the proliferative capacity of alloreactive lymphocytes in the transfusion, is demonstrated by the success of gamma-irradiation in preventing the TA-GVHD reaction in otherwise high-risk transfusions. Definition of the cell types involved has come from Japanese work in which donor-derived lymphocyte clones were grown from the peripheral blood of patients with advanced TA-GVHD. At least three different types of T-cell clones were defined in these studies: 1 CD8+ clones with direct cytotoxicity against patient-specific HLA class I epitopes (type I);

200

2 CD4+ clones cytotoxic for patient-specific HLA class II epitopes (type II); and 3 CD4+ clones lacking direct cytotoxicity but with lytic supernatants, the actions of which were blocked by monoclonal antibodies to tumour necrosis factor (TNF)-b (type III). Studies of T-cell receptor (TCR) gene rearrangements in lymphocytes taken from TA-GVHD patients have shown marked oligoclonality of response. The precise mechanisms of lymphocyte cytotoxicity remain to be elucidated. In other contexts, two pathways of lymphocyte killing have been recognized: cytolytic mediators (perforin and granzymes), and a molecule, designated Fas ligand, which triggers apoptotic cell death via an interaction with its receptor Fas on the surface of the target cell. This latter mechanism may be involved in TA-GVHD, since the cytotoxicity of the type I and II clones described above is blocked by monoclonal antibodies to Fas. From one patient, a B-cell line with cytotoxic activity in its supernatant was also established. As well as direct lymphocytotoxicity, cytokines almost certainly play a key role in the tissue damage which is such a striking feature of TAGVHD. Levels of inflammatory cytokines such as IL-1, IL-2, TNF and interferon-g are greatly increased in TA-GVHD, a process which may be amplified by tissue injury due to infection, chemotherapy/radiotherapy or tumour invasion. Such high cytokine levels upregulate HLA expression, and recruit other donor-derived T cells and macrophages. A positive feedback loop is thus established, leading to the full clinical picture (Fig. 18.1). Lymphocyte dose and HLA haplotype sharing in human TA-GVHD

The precise dose of lymphocytes required to initiate TA-GVHD in humans is not known precisely. In the mouse, more than 107 lymphocytes are required and although most reported clinical cases received 1010 cells, a child with immunodeficiency apparently developed the condition after only 104 cells/kg. In stored blood, lymphocytes can retain their mitotic capacity for 17–22 days at 4°C,

Immunomodulation and GVHD

Autologous recognition

Allogeneic recognition

Tissue injury (from chemotherapy, infection tumour, etc.)

Upregulation of histocompatibility antigens

T-cell activation

Production of cytokines (IL-1, IL-2, TNF, g-IFN)

be accounted for totally by under-recognition and under-reporting. Protective mechanisms in the host presumably prevent the vast majority of potential TA-GVHD reactions associated with HLA haplotype sharing. Indirect evidence for this comes from observations using the polymerase chain reaction (PCR) to differentiate donor and recipient cells. Donor leucocytes can be found in the peripheral blood of immunocompetent recipients for 2–3 days after transfusion. At this point a transient 1-log increase can be demonstrated, normally followed by rapid clearance thereafter. Specific risk factors for TA-GVHD in apparently immunocompetent recipients have not yet been identified.

GVHD

Clinical features and diagnosis Fig. 18.1 Interaction of cells and cytokines and creation of

positive feedback loops in the pathogenesis of graft-versushost disease (GVHD). IFN interferon; IL, interleukin; TNF, tumour necrosis factor. (Data from author’s laboratory.)

although many reports involve either blood transfused in the first few days of its shelf-life or platelets, which have a shelf-life of only 5 days at 20–22°C. Cases most similar to the parent-to-F1 hybrid model of TA-GVHD are represented by transfusion from HLA homozygous random donors coincidentally haploidentical with their recipients, although they may differ at DP or in subtypes of DR and DQ. HLA-DP allelic sequences are important in the MLR, and DP mismatching along with a single-allele HLA class I mismatch has been associated with fatal TA-GVHD in an immunocompetent individual. Estimates of HLA homozygous transfusions being given to unrelated individuals heterozygous for the same haplotype have been calculated as 1 in 17 700 to 1 in 39 000 for the white population in the USA, 1 in 6900 to 1 in 48 500 for Germans, and 1 in 1600 to 1 in 7900 for Japanese. Numbers of reported cases of TA-GVHD in immunocompetent recipients in the USA are several thousandfold less than predicted, which could not possibly

As in GVHD following transplantation of bone marrow (BMT) or of peripheral blood-derived progenitor cells, the classical features are fever, erythematous skin rash, diarrhoea and biochemical hepatitis with or without jaundice. These symptoms do not arise until 1–2 weeks after transfusion and, since these may have other causes in sick and/or immunosuppressed transfusion recipients, the correct diagnosis may not be considered for a number of days. A critical difference between TA-GVHD and GVHD following BMT is the severe bone marrow hypoplasia and pancytopenia in TA-GVHD, since in this setting the bone marrow is the ‘host’ rather than the ‘graft’. Once this is established, the disease usually follows a downhill course, with death from infection and/or multiorgan failure the usual outcome. In neonates, there may be early hepatosplenomegaly and lymphadenopathy followed by lymphoid regression. This may be difficult to differentiate from congenital immunodeficiency. Similarly fever, failure to thrive and diarrhoea may be mistaken for infection in premature neonates. Histological diagnosis is most easily made by skin biopsy, although the features are not specific and may be reported as ‘consistent with’ rather than diagnostic of TA-GVHD. There is epidermal cell dyskeratosis and eosinophilic necrosis, satellite

201

Chapter 18

cell necrosis and a dermal infiltrate of mononuclear cells. It would therefore be useful to obtain direct evidence of donor-derived cells in the recipient’s tissues, or at least in the patient’s blood. However, blood analysis can be difficult because the patient is often profoundly leucopenic before the diagnosis is considered. In addition, HLA typing alone may be non-informative if the donor is homozygous for a haplotype shared with the recipient. For this reason, minisatellite probes to non-HLA polymorphisms have been developed to demonstrate differences between donor-derived circulating cells and recipient tissue. Techniques for testing recipient tissue non-invasively using hair or nail clippings have been described. An alternative technique is comparison of variable number tandem repeat (VNTR) profiles between donor, recipient and recipient posttransfusion samples, which can highlight the presence of donor DNA following a transfusion. Following PCR of highly polymorphic VNTRs using fluorescently labelled primers, samples can be electrophoresed in an automated genetic analyser. An example of this technique is shown in Fig. 18.2. Therapeutic options in established TA-GVHD

No therapy has yet emerged as curative in established cases of TA-GVHD. The rarity of the condition makes assessment of new therapies difficult. Recently reported cases have virtually all received high-dose steroids, while azathioprine, methotrexate, cyclosporin, antithymocyte globulin, CAMPATH (anti-CD52) antibodies and granulocyte colony-stimulating factor (G-CSF) have all been used unsuccessfully in individual cases. BMT has been tried in two cases, without success. A new way of testing possible new therapies has emerged from experiments performed in Japan using T-cell lines derived from TA-GVHD cases. Two therapeutic strategies have been examined in this model, the protease inhibitor nafomostat mesilate, licensed in Japan as an anticoagulant, and the antimalarial drug chloroquine. Now widely used in autoimmune rheumatoid disorders, chloroquine has been reported to inhibit presentation of antigen to the TCR, T-cell granule exocytosis and TNF production by T cells. Both 202

nafomostat mesilate and chloroquine inhibited CD4+ and CD8+ cytotoxic T cells against HLAexpressing target B-cell lines at concentrations achievable in humans. The two drugs act synergistically at low concentration to inhibit the cytotoxic T-lymphocyte response and, when used separately, reduced TNF-b synthesis in a separate T-cell clone. Two patients have been treated with nafomostat mesilate and, although both died, their survival was prolonged to 35 and 47 days respectively, with recovery of mononuclear cell counts and liver function, and resolution of fever and rash. Prevention of TA-GVHD by gamma-irradiation of cellular blood components

The lack of effective therapy emphasizes the importance of prevention of TA-GVHD in highrisk situations. To date, the only reliable means of prevention is gamma-irradiation of cellular components to prevent donor lymphocyte proliferation. A dose of 25 Gy to all parts of the pack is required, since there have been ‘breakthrough’ cases associated with lower doses. No part of the pack should be exposed to more than 50 Gy, since this dosage is associated with cellular damage. The process is best carried out in a dedicated blood irradiator. Dosimetry studies are required to define the best dose field. Although the number of TA-GVHD cases in the UK appears to have fallen with the implementation of universal leucocyte depletion, this is almost entirely accounted for by the disappearance of cases in individuals with normal or only mildly suppressed immunity, in whom gamma-irradiated blood components are not recommended. For patients with greater degrees of immunosuppression, leucocyte-depleted blood components are not an adequate substitute for gamma-irradiation, as the residual lymphocyte numbers may still be above the threshold dose for TA-GVHD. There has been at least one reported case of TA-GVHD associated with leucocyte-depleted red cells. Ultraviolet (UV)B irradiation has been successfully used to reduce HLA alloimmunization by inactivation of antigen-presenting cells. UVB also reduces T-cell responses in the MLR and to phytohaemagglutinin.

Immunomodulation and GVHD

Locus = FGA

260

280 Donor

1200 800 400 0

Patient

2500 2250 1500 750 0

Patient: post-transfusion

270 180 90 0

Locus = vWA31 150 900

Donor

600 300 0 2250

Patient

1500 750 0 630

Patient: post-transfusion

420 210 0

Fig. 18.2 Variable number tandem repeat profiles for two different loci, FGA and vWA31, in samples taken from the donor,

recipient and recipient post bone marrow transplantation showing the presence of donor DNA after transfusion. Detection of 0.1–1% donor cells can be obtained routinely.

Blood components fall into three categories. 1 Those which carry a high intrinsic risk and so should be irradiated for all recipients. Granulocytes, from either apheresis or buffy coats, are

always transfused within hours of collection, and are contaminated with large numbers of viable lymphocytes. This, combined with some inevitable degree of immune suppression in the recipient, 203

Chapter 18

means that they should always be irradiated. Because of the importance of HLA haplotype sharing as a risk factor for TA-GVHD, all transfusions of any cellular components between family members, and all HLA-selected platelets, must be irradiated. 2 Those from which the risk is so low that irradiation is never required. Fresh frozen plasma (FFP), cryoprecipitate and fractionated plasma products fall into this category. There has been one report of fresh plasma causing TA-GVHD in a child with congenital immunodeficiency but, in standard FFP, lymphocyte responses to phytohaemagglutinin are absent so gamma-irradiation of FFP is not currently recommended for any patient group. 3 Those which should be irradiated for susceptible patients, as defined below.

exchange transfusion (ET), the red cells should be transfused within 24 h of irradiation. Gamma-irradiation has no undesirable effects on platelet quality within normal storage conditions, so irradiation may be performed on platelets at any stage in their shelf-life, which can be maintained for a total of 5 days. At dose rates currently used (3–4 Gy/min), there is no evidence of radiation-induced malignant change, activation of latent viruses or leakage of plasticizer in irradiated blood units. These may therefore be safely returned to stock if not used for the intended recipient. The use of radiation-sensitive labels is recommended to provide assurance that a given pack has been subjected to the irradiation process. Labels can be purchased for different irradiation doses, but their use should supplement formal dosimetry rather than replace it.

Effect of gamma-irradiation on component quality

Gamma-irradiation causes accelerated loss of intracellular potassium from red cells, at approximately twice the rate of that in non-irradiated cells. This means that supernatant potassium levels at any point during the 35-day storage period are twice as high as in standard red cells. Infusion of a large potassium load, particularly if given rapidly and/or into a central vein, may cause dangerous cardiac dysrhythmias. An extreme case is intrauterine transfusion, where central administration of red cells with a haematocrit of 80–90% may be associated with a supernatant potassium concentration as high as 9 mmol/L. However, removal of supernatant plasma with washing of red cells is not recommended as such manipulations increase the risk of error and bacterial contamination. Irradiated red cells also show reduced posttransfusion recovery, but this can still be above the required 75% level, depending on the age of the blood at the time of irradiation and the duration of postirradiation storage. The current UK recommendation is that red cells should be irradiated at any time up to 14 days after collection, and thereafter can be stored for up to a further 14 days. Where the patient is at particular risk from hyperkalaemia, e.g. intrauterine transfusion (IUT) or 204

Indications for gamma-irradiated blood components (Table 18.5) Paediatric transfusion

Clinically recognized TA-GVHD following IUT from an unrelated donor is extremely rare, but it Table 18.5 Indications for gamma-irradiated cellular

components for prevention of transfusion-associated graftversus-host disease. General recommendations Congenital immunodeficiencies affecting T cells Intrauterine and exchange transfusions Transfusions from family members HLA-matched platelets Allogeneic bone marrow (or peripheral blood stem cell) transplantation Autologous bone marrow (or peripheral blood stem cell) transplantation Hodgkin’s disease Recommended in some countries HIV positivity and AIDS Aplastic anaemia Acute leukaemia Chronic leukaemias Non-Hodgkin’s lymphoma All neonatal transfusions Patients receiving fludarabine and related drugs

Immunomodulation and GVHD

has been suggested that the fetus is less able to reject allogeneic lymphocytes. In addition, very fresh blood must be used if hyperkalaemia is to be avoided. These two factors together have led most authorities to recommend irradiation for all IUT of red cells and platelets. An additional point is that, for reasons which are unclear, infants who have had IUT may be more prone to TA-GVHD from any subsequent exchange or ‘top-up’ transfusions. Cases have also been described associated with ET without prior IUT. Many centres now irradiate all ET blood, provided this does not cause undue delay in making the component available. The postirradiation shelf-life for IUT and ET blood in the UK is 24 h, and the blood must be transfused within 5 days of collection. For small-volume ‘top-up’ transfusions, the risk seems to be considerably less. Blood for smallvolume transfusion of infants can now be used up to the normal 35-day shelf-life, so the risks associated with very fresh blood do not apply, and the volumes infused are much less. Provided the infant has not had previous IUT and provided the donor is not a family member, it seems unnecessary to irradiate these units routinely. More importantly, prestorage irradiation, by shortening the shelf-life to 14 days, severely limits the possibility of allocating a prealiquotted adult unit (in ‘paedipacks’) to a single infant requiring repeated transfusions over a 3–4 week period. Non-irradiated red cells can be used in this way right up to their 35-day expiry, and this strategy greatly reduces donor exposure. The risk of TA-GVHD from neonatal platelet transfusion seems low, and irradiation is not required unless there has been prior IUT. Most reports of TA-GVHD in early childhood have been associated with an underlying primary immune defect of T-cell function. A variety of other similar T-cell deficiencies exist in which TAGVHD has not been reported. The immune features of all these disorders are very similar, so it is recommended that all infants with suspected or diagnosed T-cell immune deficiencies receive irradiated blood. Infants with pure B-cell defects appear not to be at risk. Transfusions in normal infants to cover either cardiac surgery or extracorporeal membrane oxygenation do not pose a significant risk of TA-

GVHD. However, there should be a high index of suspicion concerning coexisting cardiac defects and immunodeficiency (DiGeorge’s syndrome, where there may be immunodeficiency even when the lymphocyte count is normal). If there is any suspicion of immunodeficiency, irradiated components should be provided. Haemopoietic progenitor cell allografting and autografting

Bone marrow allograft recipients have received irradiated blood as a standard of care for over 20 years. However, there is no consensus as to how long this needs to be continued, and surveys in the USA and UK have revealed wide variations in practice. It would be reasonable to continue provision of irradiated blood while the patient requires prophylaxis against transplant-induced GVHD (usually 6 months), or until the lymphocyte count is greater than 1 ¥ 109/L. However, irradiation of blood components may need to be continued for longer if the transplant is performed for an immune defect, or if there is chronic GVHD from the allograft. Gamma-irradiation of blood components for autograft recipients is also now standard practice, and should be commenced 1 week prior to bone marrow or peripheral blood stem cell harvesting. This will prevent harvesting of viable blood donorderived T cells with the autograft. As with allograft recipients, there is no precise evidence as to how long irradiation of blood need continue. Many patients require long periods of transfusion support, particularly with platelets. It would be wise to continue irradiation of those until there is unequivocal evidence of haemopoietic engraftment, which may take at least 3 months, or even up to 6 months if the patient has had total body irradiation. Leukaemia and lymphoma

TA-GVHD associated with acute or chronic myeloid leukaemias in the absence of a transplant procedure is extremely rare, with only 11 reports in the world literature. No cases of TA-GVHD in such patients could be recalled in a UK survey in 205

Chapter 18

1995, and no such cases have been reported since. Current UK recommendations therefore do not require provision of irradiated components for patients with myeloid leukaemias, but some countries include all acute leukaemias in their TAGVHD guidelines. In contrast, cases of TA-GVHD have been reported in association with all lymphoproliferative disorders, but some types carry a higher risk than others. Hodgkin’s disease, at whatever stage, is an independent risk factor for TA-GVHD, and all patients with the condition should receive irradiated components throughout treatment. Fewer reports have appeared of TA-GVHD in association with non-Hodgkin’s lymphoma, despite this being a more common disease. Non-Hodgkin’s lymphoma is therefore not included in the current UK indications for irradiated components. However, four cases of TA-GVHD in association with both low-grade and high-grade B-cell lymphoproliferative disorders have been reported in the UK in the last 6 years, so this recommendation remains under review. New chemotherapies and immunotherapies

It became apparent during UK trials of purine antagonists (such as fludarabine, cladribine and 2¢deoxycoformycin) for low-grade B-cell malignancies that an unexpected number of cases of TA-GVHD were occurring. These drugs cause profound depression of peripheral blood T-cell numbers, which may last for several years, but which may be masked by the underlying B-cell lymphocytosis. These drugs are now used in other chemotherapy regimens, and are now included in the UK list of indications for irradiated components. However, TA-GVHD in patients receiving purine antagonists has not been observed in the USA, for reasons which are not apparent. In general, patients receiving chemotherapy or immunotherapy for solid tumours or autoimmune disorders have a very low risk of TA-GVHD. However, dose escalation of chemotherapy regimens, especially in paediatric oncology, may expose patients to an increased risk. The same might apply to the use of therapeutic antibodies

206

against T cells in non-transplant settings such as autoimmune disorders and aplastic anaemia. Each regimen must be considered in terms of extent and duration of suppression of T-cell numbers and function. Immunosuppression associated with organ transplantation using drugs such as cyclosporin might be expected to increase the risk of TAGVHD, because of inhibitory effects on T cells. However, no such cases have been described. Rare cases of GVHD following transplantation of liver or heart/lung have been shown to be due to passenger lymphocytes in the transplanted organ. Human immunodeficiency virus infection and AIDS

A survey of blood banks in the USA in 1990 showed that 25% of institutions were already providing irradiated components for patients with human immunodeficiency virus (HIV) or AIDS. However, the risk of TA-GVHD in HIV-positive patients receiving non-irradiated components appears not to be increased. The reasons for this are not entirely clear, but may relate to virus uptake by donor lymphocytes following transfusion, limiting their capacity to proliferate. Monitoring of TA-GVHD cases

Many countries have developed guidelines for prescription of irradiated components. An essential step in ensuring that these remain relevant is the development of a mechanism for collation of cases of TA-GVHD at national or international level, and rapid feedback for update of guidelines should new risk factors emerge.

Further reading Blumberg N. Transfusion-induced immunomodulation: an overview. Transfus Med Rev 2001; 15: 109–35. Brennan DC, Mohanakumar T, Wayne M. Donor-specific transfusion and donor bone marrow infusion in renal transplantation tolerance: a review of efficacy and mechanisms. Am J Kidney Dis 1995; 26: 701–15. British Committee for Standards in Haematology.

Immunomodulation and GVHD Guidelines for gamma irradiation of blood components for the prevention of transfusion-associated graft-vs.-host disease. Transfus Med 1996; 6: 261–71. Contini P, Ghio M, Merlo A et al. Soluble HLA class I/CD8 ligation triggers apoptosis in EBV-specific CD8+ cytotoxic T lymphocytes by Fas/Fas-ligand interaction. Hum Immunol 2000; 61: 1347–51. Dzik S, Blajchman MA, Blumberg N, Kirkley SA, Heal JM, Wood K. Current research on the immunomodulatory effect of allogeneic blood transfusion. Vox Sang 1996; 70: 187–94. Ghio M, Contini P, Mazzei C et al. Soluble HLA class I, HLA class II and Fas ligand in blood components: a possible key to explain the immunomodulatory effects of allogeneic blood transfusions. Blood 1999; 93: 1770–7. Hill GR, Krenger W, Ferrara JLM. The role of cytokines in acute graft-vs.-host disease. Cytokines Cell Mol Ther 1997; 3: 257–66.

Vamvakas EC. Meta-analysis of randomised controlled trials investigating the risk of postoperative infection in association with white blood cell-containing allogeneic blood transfusion: the effects of the type of transfused red blood cell product and surgical setting. Transfus Med Rev 2002; 16: 302–14. Vamvakas EC, Blajchman MA. Deleterious effects of transfusion-associated immunomodulation: fact or fiction? Blood 2001; 97: 1180–95. Vogelsang GB, Hess AD. Graft-versus-host disease: new directions for a persistent problem. Blood 1994; 84: 2061–7. Williamson LM, Warwick RM. Transfusion-associated graft-vs.-host disease and its prevention. Blood Rev 1995; 9: 251–61. Zavazava N, Krönke M. Soluble HLA class I molecules induce apoptosis in alloreactive cytotoxic T lymphocytes. Nat Med 1996; 2: 1005–10.

207

Chapter 19

Transfusion-transmitted infections Alan D. Kitchen and John A.J. Barbara

This chapter aims to describe the range of infectious agents currently known to be transmissible by blood transfusion, transfusion-transmissible infectious agents (TIAs), considering briefly their biology and epidemiology and the options available for their detection in donated blood and blood products. It is only over the last 20 years that the true significance of the potential of blood transfusion as a vehicle for the transmission of infectious agents has been recognized widely. While transmission of syphilis and hepatitis B virus (HBV) have been recognized for many years, it was the identification of human immunodeficiency virus (HIV) and hepatitis C virus (HCV) that opened many eyes not only to the potential of transfusion as a route of infection, but also to its startling efficiency in the transmission of a whole range of infectious agents. Viruses, bacteria and protozoa have been clearly demonstrated to be transmitted by transfusion of blood or blood products. More recently, the potential for the transmission by blood transfusion of protein-based agents, prions, has been highlighted and it is in the area of these and novel viruses that much current interest is focused. The detection of TIAs, although described in Chapter 22, merits a brief mention here in the context of understanding the roles of donor selection and laboratory screening, and the strengths and limitations of each in relation to specific TIAs. • The process of identifying donors potentially carrying TIAs begins with donor selection. • The identification and deferral of donors considered to be ‘high risk’, by activity or association, is central to donor selection, which aims to minimize any risk of TIA from donations which may have been collected from donors in the ‘window period’ 208

of an infection (infectious but not at that time detectable by routine screening tests). • The brief assessment of donors prior to donation, even when performed thoroughly, can only be an initial filtering process to identify any donors who have an identifiable exposure risk to a TIA. • Laboratory screening remains the key step in identifying donations from infected individuals. • The implementation of a formal quality system with effective quarantine and disposal of unsuitable donations and products is critical for minimizing the risk of releasing donations containing TIAs. • The increasing development and use of effective and safe microbial inactivation procedures, and their application to a wider range of components, is a new and pivotal strategy in reducing any remaining risk of TIA. • In many countries, the screening of donated blood for a minimum set of markers of infectious agents is mandatory. • The identification of those infectious agents for which blood needs to be screened is based upon prevalence and incidence data for the particular donor population, surveillance data for other potentially transmissible agents and the pathogenic potential of the agent, together with any political, social or ethical considerations that may be relevant. Thus in most countries blood is not screened for all potential TIAs identified. The choice in screening represents a compromise between risk and available resources.

Transmissibility of infectious agents The reasons why specific agents are transmissible

Transfusion-transmitted infections

and their specific characteristics need to be considered for a full understanding of TIAs. There are four main properties that generally need to be met for an agent to be transmitted by transfusion: • it gives rise to asymptomatic infection; • it is present in the bloodstream; • it is transmitted parenterally; and • it is able to survive during storage of the blood. Asymptomatic infection

The agent must be capable of giving rise to asymptomatic infection in the infected individual, such that an actively infected and thus potentially infectious individual may present as a blood donor. Any potential donor who has any recognizable clinical symptoms which could be due to an infection should be deferred. However, this does assume that the donor selection and questioning procedures are adequate, and that the donor would declare any symptoms appearing in the few weeks before donation. Thus, any infectious agent that always gives rise to clinical symptoms, i.e. a symptomatic infection, is very unlikely to be transmitted by transfusion because an infected donor should always be identified and deferred prior to donation.

Carried in leucocytes

Free virions or latent forms integrated into cellular nucleic acid can be carried in leucocytes. In some cases the same agent may be found in both forms, but at different stages of infection. Latent agents persist even in the presence of specific neutralizing antibody. Although some types of leucocytes will actively engulf individual infectious agents in their role as scavengers, agents that can be found in leucocytes generally specifically infect these cells as part of their life cycle. Because leucocytes are nucleated and have a normal cellular cytoplasmic organization, once latently infected the infection may persist for the life of the cell, and ultimately the life of the individual. However, reactivation can subsequently occur and may result in acutephase infection. Carried in erythrocytes

Some protozoan infections include a phase in which the agent is present and actively dividing in red cells. During this phase the agent usually matures into the next stage of its life cycle before being released from red cells, either while they are still circulating or in the liver or spleen.

Presence in the bloodstream

Parenteral transmission

An infectious agent must be present in the blood of the donor at the time of donation, in an infectious or potentially infectious form, to be transmitted by blood transfusion. TIAs are carried in the blood in a number of different ways, depending on the particular agent and the stage of infection: free in the plasma, within the leucocytes (either as infectious virions or in a latent form) or within red cells.

In general, only those infectious agents transmitted parenterally are considered to be TIAs and are of most concern in relation to the safety of donor blood. However, there are exceptions to this; for example, hepatitis A virus (HAV) has been reported to be transmitted by transfusion. Although HAV is an enterovirus transmitted by the faeco-oral route and is not normally considered to be transmissible by transfusion, cases of transmission through large-pool plasma products as well as single-donor products have been reported.

Free in the plasma

Several viruses, bacteria and protozoa may be carried free in the plasma, as a means of directly infecting other tissues or as part of the life cycle of the organism when it is released from infected tissues into the blood.

Survival during storage

Blood and blood products may be stored in a number of different physical states (e.g. whole blood, plasma, high-concentration protein solution, lyophilized material) and at a number of 209

Chapter 19

different temperatures (–40°C to 25°C). Any agent present in donated blood must be able to survive at least some of these storage conditions and the conditions during processing in order to infect recipients of the products. Viruses are best suited to this and certain viruses, if present in the original donation, may be found in virtually all products prepared. This is especially true of non-enveloped viruses such as HAV and parvovirus B19. Bacteria also persist but often they may also multiply during storage and as they die leave endotoxins in the products, which can then rapidly cause severe illness in recipients. Treponema pallidum, the causative agent of syphilis, is one of the more unstable organisms and will usually survive for no longer than 72 h at 4°C.

Table 19.1 Infectious agents reported to have been

transmitted by blood transfusion. Viruses Hepatitis viruses Hepatitis A virus Hepatitis B virus Hepatitis C virus Hepatitis D virus (requires coinfection with hepatitis B virus) Retroviruses Human immunodeficiency virus 1 and 2 (plus other subtypes) Human T-cell leukaemia virus I and II Herpes viruses Human cytomegalovirus Epstein–Barr virus Human herpesvirus 8 Parvoviruses Parvovirus B19

Types of infectious agents transmitted Viruses, bacteria and protozoa have been demonstrated to be transmitted by transfusion. Fungi have not been reported to have been transmitted and there is still no firm evidence that prions are actually transmitted. However, it is difficult to produce a definitive list of transmissible agents as increasing numbers of cases of transmission of unusual infectious agents are being reported in addition to the more common ‘established’ TIAs. An important factor in the increasing range of transmitted agents is international travel. As travel increases, individuals are being exposed to an everincreasing range of infectious agents including many potential TIAs. Migration of individuals from endemic areas with a high prevalence of infection is also a potential risk to blood safety in areas of lower endemicity. Thus blood transfusion services, especially in countries whose donor populations travel extensively, face an expanding range of infectious agents that need to be considered to ensure a safe blood supply. Unfortunately, in vitro screening for many of these is either not practical or not appropriate, and donor selection processes are key in the identification and deferral of ‘at-risk’ donors. Table 19.1 provides an up-todate list of viral, bacterial and protozoan infectious agents reported to have been transmitted by

210

Miscellaneous viruses GBV-C: previously referred to as hepatitis G virus TTV West Nile virus Bacteria* Endogenous Treponema pallidum (syphilis) Borrelia burgdorferi (Lyme disease) Brucella melitensis (brucellosis) Yersinia enterocolitica/Salmonella spp. Exogenous Environmental species: staphyloccocal spp./pseudomonads/Serratia spp. Rickettsiae: Rickettsia rickettsii (Rocky Mountain spotted fever), Coxiella burnetii (Q fever) Protozoa Plasmodium spp. (malaria) Trypanosoma cruzi (Chagas’ disease) Toxoplasma gondii (toxoplasmosis) Babesia microti/divergens (babesiosis) Leishmania spp. (leishmaniasis) * For a detailed review of bacterial species and frequency in relation to blood transfusion see Chapter 16.

Transfusion-transmitted infections

Viruses Hepatitis viruses

The hepatitis viruses are a diverse group of viruses, including hepadnaviruses, flaviviruses and picornaviruses, all of which have been transmitted by transfusion. Although HBV and HCV are transmitted parenterally and are typical candidate transfusion-transmissible viruses, HAV may also (rarely) be transmitted parenterally when sufficiently high titres of virus are present. Hepatitis B virus

HBV is a DNA virus and a member of the hepadnavirus family. The infectious particle, the Dane particle, comprises the DNA genome encapsulated in core protein, which is then covered by an envelope of surface proteins. The hepadnaviruses are characterized by the production of a vast excess of the surface proteins, hepatitis B surface antigen (HBsAg) in the case of HBV, and these are released

Window period

into the blood along with the infectious Dane particles. The virus is transmitted parenterally and infection may follow one of two courses: acute infection with the subsequent clearance of the virus and development of immunity (Fig. 19.1) or chronic infection with persistence of virus replication for extended periods even during the lifetime of the individual (Fig. 19.2). Chronic infection may resolve spontaneously and the individual may then develop immunity. Alternatively, a stable chronic infection may reactivate, with a resulting further acute episode. Although HBV infection can lead to severe disease, i.e. cirrhosis, hepatocellular carcinoma and liver failure, asymptomatic infections are very common, with most individuals resolving infection and developing immunity without, or with only mild, symptoms. Detection of HBV infection in donated blood is by screening for HBsAg, the first marker to appear in the blood, which persists throughout the period of infectivity and eventually marks chronic infection. The other markers of HBV infection are of use in confirming infection and determining the type and stage of infection, but apart from anti-

transfusion or which are considered to be potentially transmissible.

Incubation Transaminases Concentration of virus markers

Jaundice Prodrome

HBeAg Anti-HBc HBsAg Anti-HBe

Fig. 19.1 Markers of HBV infection

during acute infection. Typical course of an acute infection with HBV. HBeAg, hepatitis Be antigen; HBsAg, hepatitis B surface antigen; anti-HBc, antibody to hepatitis B core antigen; anti-HBe, antibody to HBe; anti-HBs, antibody to HBsAg.

Anti-HBs Infection

0

3

4

5

6

7

8

9

Time (months)

211

Concentration of virus markers (usually asymptomatic)

Chapter 19

Anti-HBc

HBsAg

HBeAg, for variable period, usually converting to anti-HBe Fig. 19.2 Markers of HBV infection

Infection 0

3

4

5

6

7

Time (months)

HBc (antibody to hepatitis B core antigen) have no value in routine screening of blood donors. Anti-HBc screening For many years the subject of anti-HBc screening of donations, in addition to HBsAg, has been considered. Before specific anti-HCV screening became possible, anti-HBc screening was considered as a surrogate marker for reducing the number of cases of post-transfusion non-A, non-B hepatitis then seen in large numbers, but it was demonstrated subsequently that in most populations this strategy had little specific value. Anti-HBc screening may have value in identifying the small number of donors who are either resolving an acute infection or clearing a chronic infection (see Figs 19.1 and 19.2). These donors are apparently HBsAg negative on screening, but may still have a low-level viraemia and be infectious (‘tail-end carriers’). Anti-HBc may be the only detectable circulating marker of infection in such individuals, and they may only be identifiable by anti-HBc screening. In addition, anti-HBc screening may be of value in the detection of HBVinfected donors but who have mutant HBsAg (see section below). In cases of HBsAg mutations it is only the HBsAg protein that may be sufficiently 212

Several years

during carrier state. Typical course of a chronic infection with HBV leading to a carrier state (see Fig. 19.1 for explanation of abbreviations).

altered in structure to render it undetectable by some HBsAg assays; the expression of the other markers of HBV infection expected in such individuals is unaltered. The above situations could explain cases of post-transfusion HBV reported as resulting from the transfusion of donations screened as HBsAg negative where it is often not easy to demonstrate that the patient had no other risks of infection. However, a major issue with anti-HBc screening is the identification of those donors who are truly ‘anti-HBc only’ or who are naturally immune following infection earlier in life. It is generally agreed that individuals with low-level anti-HBs (usually <100 miu/mL), although there is little scientific evidence for the setting of this level, cannot be considered to be sufficiently immune to be used as blood donors. Currently a number of countries do screen all donations for anti-HBc but its true value is still debatable. HBV DNA screening The screening of blood donations for HBV DNA has been considered to reduce even further the risk of transmission of HBV by reducing further the window period in early infection. However early in infection the levels of HBV DNA are signifi-

Transfusion-transmitted infections

cantly lower than the other major TIAs and the sensitivity of the assays is only very marginally, if at all, greater than the current HBsAg assays. In situations where pools of samples are screened for viral nucleic acids, there would be no benefit from HBV DNA screening. In situations where individual donations are screened the benefit would be marginal and certainly not cost-effective in terms of significant disease prevented. As mentioned above there is no value in HBV DNA screening with the current commercially available screening assays in the case of resolving acute infections and anti-HBc would be a more reliable and costeffective marker. HBV mutants There are currently two main groups of HBV mutants identified: the core group of mutants and the surface antigen mutants. The core mutants have normal HBsAg expression and are not currently considered to present any threat to the blood supply. However, the HBsAg group of mutants are of concern as HBsAg expression is altered such that some assays may fail to detect some examples. The extent of the problem is difficult to assess and although mutants are being identified and variable reactivity with assays reported, their frequency is very low. A proportion of those that have been identified was a result of specific searches for such mutants and not as a result of transmissions from donations screened as HBsAg negative. There is no evidence currently available of a significant problem of post-transfusion HBV infection associated with HBsAg mutants. Hepatitis delta virus

Hepatitis delta virus (HDV) is a small RNA virus, currently not firmly classified, that requires coinfection with HBV and replicates only in the hepatocytes. Although transmissible by transfusion, viral replication has an absolute requirement for coinfection with HBV and screening for HBsAg will also prevent transmission of HDV. Hepatitis C virus

HCV is an enveloped RNA virus that has been

classified as a separate genus within the flavivirus family. Although infections with HCV have been recognized for many years, the virus has only been identified and characterized within the last 15 years. It is transmitted parenterally and although there was initially some debate about the risk of transmission through sexual contact, the routes of infection are essentially the same as for HBV. Following infection there is an incubation period of about 3 weeks to 3 months prior to the appearance of HCV RNA and HCV antigen, and a further 0.5– 4 weeks before the development of anti-HCV. Infection with HCV can follow one of two courses: acute infection followed by resolution of infection or chronic persistent infection. In about 50% of cases the infection is acute and resolves, usually within a year. Whether true immunity follows the resolution of acute infection is not yet clear. Although HCV RNA is no longer detectable, the anti-HCV remaining may not protect the individual from subsequent reinfection. Although much is now known about the virus, the agent itself has still not been isolated and studied as an intact virion; HCV RNA and HCV antigen can be detected in the serum of infected individuals, but so far complete HCV virions have not been recovered. Current knowledge about HCV and diagnostic assays that are now available derive from isolation of the complete viral genome from human plasma and its in vitro expression and subsequent characterization of the proteins expressed. The viral genome has been found to have certain areas with marked variation in the nucleic acid sequence, which has been shown to give rise to a number of distinct virus genotypes. Currently six have been well characterized, and at least two more are being investigated. These genotypes tend to be geographically associated, and importantly do show some differences in the course and severity of infection and in the response to interferon therapy, for example subtype 2 appears to be more resistant to current interferon therapy. Whether these variations are further reflected in the immune response to the virus, possibly a contributory factor to the classification of some HCV ‘indeterminates’, is not yet known. The humoral immune response to HCV appears 213

Chapter 19

to be relatively weak compared with HBV and HIV; antibodies are not present at high titres and their reactivity against the currently identified individual epitopes is very variable. Our knowledge and understanding of the serology of HCV is based primarily upon the reactivity of infected individuals against the specific and defined antigens used in current screening tests. This has limited the development of our understanding, because only a limited number of HCV-specific antigens are currently available for use in diagnostic tests. Although the introduction of anti-HCV screening reduced dramatically the number of cases of post-transfusion HCV, occasional cases still occur. A feature of HCV infections is the window period during which HCV RNA is present prior to an immune response. Although donations collected from recently infected individuals may transmit HCV to recipients, the incidence of such HCV RNA-positive, antibody-negative donations is very low. Nonetheless, because of this the use of HCV RNA screening, currently using either individual donation testing or pools of samples (usually from 16 to 96), is increasing since in countries with well-defined and resourced healthcare systems, blood transfusion services are under political and commercial pressure to improve further the safety of the blood supply. In many populations the cost of this appears to outweigh the benefit. HCV antigen testing Concurrent with the appearance of HCV RNA is that of HCV core antigen (HCVcAg). The antigen is a normal constituent of the virion and conventional serological tests now exist to detect HCVcAg, both when complexed with specific antibody and when circulating in the plasma in the uncomplexed form. In addition, in a similar way to HIV, at least one HCV antigen–antibody combination assay is close to commercial release. Although not always correlating totally with viral RNA, HCVcAg can be detected, as expected, just as early during infection. In a number of studies using seroconversion panels where the early members were HCV RNA and antibody negative, HCVcAg was detected either at the same time as or within just a few days of viral RNA. This, of course, is not sur214

prising as the antigen is an integral part of the virion. Any residual sensitivity difference between HCV RNA and HCVcAg detection is most likely to be due to the inherent sensitivity of the assay system rather than absence of antigen, i.e. failure to detect the low level of antigen present. Thus the detection of HCVcAg is a real alternative to the detection of HCV RNA in closing the antibody window period in HCV infection and increasing blood safety. Unfortunately for many developed countries the appearance of the HCVcAg assays came too late, as the infrastructure for HCV nucleic acid amplification technology (NAT) had already been developed and in most cases testing had started. Importantly, from a scientific and technical perspective, the detection of HCVcAg is not only a totally acceptable alternative to nucleic acid detection but is also technically easier to implement and sustain, and significantly cheaper. However, the litigious and regulatory nature of blood safety now outweighs any scientific values or judgement, and in countries where HCV RNA screening has already been introduced it is very unlikely that it would be replaced by HCVcAg screening. Nonetheless it is hoped that countries considering introducing HCV RNA screening would evaluate the potential of HCVcAg screening as an equivalent, but more cost-effective and technically simpler and therefore easier to control, methodology. Hepatitis A virus

HAV is a picornavirus that has been classified as a separate genus, hepatovirus. The virus is a nonenveloped RNA virus that is very stable and may persist in virally inactivated fractionated plasma products, especially those subjected only to solvent–detergent (SD) treatment. The clinical symptoms of HAV resemble those of HBV, although the onset is usually more abrupt and the preicteric stage is less prolonged. Most infections occur in children and the severity of disease increases with age. About 0.1% of infections lead to death due to fulminant hepatitis. However, chronic infection does not develop, and most infected individuals recover completely with no long-term sequelae. Most infections last for

Transfusion-transmitted infections

Retroviruses Human immunodeficiency virus

HIV is a retrovirus that primarily infects lymphocytes. The virus is transmitted parenterally mainly through sexual contact, mother-to-infant transmission and less commonly by blood or intravenous drug use. Unfortunately HIV has been transmitted through the transfusion of blood or products to thousands of people across the world. The virus infects lymphocytes and integrates into the host cell DNA using the cellular machinery to replicate. Most HIV-infected individuals recover from this initial infection within 2–3 weeks, seroconvert and then remain asymptomatic often for extended but variable periods. However, during this period, individuals actually appear to undergo a persistent chronic infection, which causes a gradual decline in CD4+ T-cell numbers. When these numbers fall below a certain level, the individual becomes susceptible to a number of infections and the symptoms marking the start of AIDS appear. Following HIV infection, seroconversion occurs usually 1–3 months later (Fig. 19.3). Prior to seroconversion, viral RNA can be detected in the

Concentration of markers

Window period

about 1 month, although rarely some infections may relapse and last for as long as 6 months. After 2–3 weeks, IgM antibodies against HAV start to appear, followed by IgG anti-HAV. By about 6 months, the IgM antibodies have disappeared but the IgG titre has stabilized, providing lifelong immunity to HAV. The virus is normally only spread by the faecooral route, but some cases of transmission by blood products have been reported, e.g. by SD factor VIII preparations and rarely by blood components. Natural transmission of HAV is very dependent on poor sanitation and standards of hygiene; it has a very high prevalence (>90%) in most developing countries, but a falling prevalence in developed countries. Although the virus is normally excreted in highest titres in the faeces, in some individuals a high-titre viraemia may be present. While asymptomatic, such individuals may be the source of transmissions of HAV, and with the falling numbers of previously exposed (i.e. immune) individuals, neutralization of any free virus in large-pool products may be incomplete.

Anti-HIV envelope

Primary infection anti-HIV core

HIV antigen

HIV antigen

Fig. 19.3 Serological markers of HIV

infection. Note that the concentrations of HIV antigen are not on the same scale as for anti-HIV. Concentrations of the latter are much greater than are concentrations of antigen.

0 1 Infection

2

3 Time (months)

Years Persistent antigenaemia

215

Chapter 19

bloodstream and proviral DNA can be detected in lymphocytes. In addition, 1–2 weeks before seroconversion, HIV p24 antigen can be detected. This is viral core antigen and is one of the most abundant proteins produced by infected cells. As the antibody levels rise, the level of p24 antigen declines as it is bound by increasing levels of circulating antibody. Although not detectable by direct methods, HIV antigen production usually continues until AIDS develops. A number of types and subtypes of HIV have been identified, and these may exhibit significant serological differences. The major division is into HIV-1 and HIV-2 and while there is significant serological cross-reactivity, there are also major differences and most anti-HIV tests used for blood screening incorporate antigens from both virus types. In addition there are subtypes that also show some serological differences, although these are not so clear-cut. The identification of HIV-1 subtype O demonstrated the need for surveillance of these emerging types and subtypes because of its variable reactivity with the then current anti-HIV-1 and -2 assays. Today all anti-HIV assays from the major international diagnostic manufacturers specifically detect subtype O, and no recent cases of failure to detect subtype O samples have been reported. A significant proportion of blood donations in developed countries are now screened for HIV infection using combined HIV antigen–antibody combination assays, specific for HIV-1 and HIV2. These assays are commonly referred to as enhanced antibody assays as generally they retain the sensitivity of the antibody-only assay but the antigen component is slightly less sensitive than an antigen-only assay. They represent an effective compromise between having separate antibody and antigen assays, and running just one assay. There is significantly higher overall sensitivity at detecting HIV-infected donors than using only an antibody assay. However, in some countries, those with a high incidence of HIV in the population, the use of separate anti-HIV-1 and -2 and HIV p24 antigen assays is more appropriate as the probability of early infections is greater, the presence of HIV p24 antigen prior to the development of the humoral response, and thus a higher sensitivity and broader screen is essential. 216

Additionally, in some countries HIV NAT is now routine, both on pools and individual donations. The value of HIV NAT is not yet clear in most countries where the incidence of HIV in blood donors is low and where blood is collected from low-risk donors. Although cases of HIV RNA-positive, antibody-negative donations are identified, the cost-effectiveness of HIV NAT is far from clear. Human T-cell leukaemia viruses (types I and II)

Human T-cell leukaemia virus (HTLV)-I was the first human retrovirus identified. It is an oncogenic virus causing adult T-cell leukaemia and lymphoma (ATLL) and tropical spastic paraparesis, also known as HTLV-I associated myelopathy. A second virus, HTLV-II, has also been identified in specific groups of individuals, for example intravenous drug users, although no significant disease process has yet been associated with this virus. Most infections with HTLV-I are asymptomatic. However, there is a small risk that disease may develop any time up to 40 years after infection. ATLL can present as an acute leukaemia of CD4+ lymphocytes and death usually occurs within a year of the onset of symptoms. Tropical spastic paraparesis is a progressive disease involving the degeneration of neurones in the spinal cord, leading to gradual paralysis of the lower limbs. More recently, HTLV infection has also been associated with certain inflammatory diseases. It is thought that HTLV is cell-associated, infecting CD4+ lymphocytes, and is transmitted in these cells parenterally via blood or semen, or from mother to infant via breast milk. Transmission by breast milk is a major route of infection in some areas where HTLV-I is endemic. Studies on the transmission of HTLV-I indicate that the virus is not normally transmitted in utero, but is transmitted in early life through breast milk equally to both male and female children. However, later in life sexual transmission is almost exclusively from males to females. Blood transfusion is another potentially significant route of infection. Early studies demonstrated the efficiency of transmission by blood transfusion, and that cell-free products, such as plasma, do not transmit infection. Fresh

Transfusion-transmitted infections

components from infected individuals are those most likely to transmit the virus. Following infection with HTLV, there is an incubation period of 30–90 days before seroconversion. Prior to seroconversion viral RNA can be detected in lymphocytes. At the time of seroconversion, antibodies to HTLV appear and detection of antibodies are used as the main diagnostic test for HTLV infection. After seroconversion the antibodies generally persist for life, even if clinical disease subsequently develops only much later. The serological responses to HTLV-I and HTLV-II are very similar, but like HIV-1 and HIV-2 there are sufficient differences to enable tests to be developed to specifically detect anti-HTLV-I and anti-HTLV-II, and thus discriminate between infections. The potential significance of blood transfusion as a route of transmission has meant that in a number of endemic countries screening of dona-

tions for anti-HTLV-I and anti-HTLV-II has been carried out for some time. In some non-endemic developed countries with mixed populations screening has also been introduced, in some instances restricted to previously untested donors. In other countries debate continues over the need for, and value of, screening donations. A costeffective approach to screening has been implemented in the UK, where a sensitive anti-HTLV enzyme immunoassay (EIA) is used to test the pooled samples that are prepared for NAT. Herpes viruses

The human herpesviruses are a family of large DNA viruses that almost always give rise to latent infections following acute infection (Table 19.2). Their pathogenic and clinical significance vary from mild and insignificant to severe disease, often depending on the immune status of the individual.

Table 19.2 Human herpesviruses.

Designation

Subfamily

Common name

Transmission

Major disease

HHV-1

a

Herpes simplex type I

Respiratory Person–person contact with active lesions

Oral and ocular lesions Encephalitis

HHV-2

a

Herpes simplex type II

Respiratory Person–person contact with active lesions

Genital lesions

HHV-3

a

Varicella-zoster

Respiratory route

Chickenpox Shingles

HHV-4

g

Epstein–Barr

Respiratory Person–person salivary contact

Glandular fever Implicated in Burkitt’s lymphoma and nasopharyngeal carcinoma

HHV-5

b

Cytomegalovirus

Respiratory Person–person salivary contact Parenteral

Congenital infection: neural tube defects Infectious mononucleosis-type disease

HHV-6

b

Human herpesvirus 6

Currently unclear, probably respiratory Person–person salivary contact

Roseola Exanthum subitum

HHV-7

b

Human herpesvirus 7

Currently unclear, probably respiratory Person–person salivary contact

Roseola (reactivation of HHV-6)

HHV-8

g

Human herpesvirus 8 Kaposi’s sarcomaassociated herpesvirus

Parenteral Sexual

Kaposi’s sarcoma Body cavity B-cell lymphoma

217

Chapter 19

Transmission by blood transfusion of herpesviruses has been demonstrated clearly for human cytomegalovirus (HCMV) and Epstein–Barr virus (EBV). Transmission of the most recently identified herpesvirus, human herpesvirus 8 (HHV-8), is uncertain and currently under investigation. While transmission of the other viruses in the family is less likely, it cannot be ruled out. Human cytomegalovirus

HCMV is the herpesvirus of most significance in blood transfusion, and was previously transmitted widely, with serious consequences for some patients. The virus is widely distributed in populations, with increasing prevalence and lower age of infection in poorer socioeconomic conditions. Prevalence figures range from 20% (though falling) in industrialized countries to 90% or more in rural areas of economically restricted countries. Although a serious infection in immunocompromised individuals, HCMV causes a largely asymptomatic infection in immunocompetent individuals, rarely with any significant long-term sequelae. The incubation period generally lasts from 1 week to 1 month, after which the infection normally lasts up to a month, but most often with no or only very mild and limited symptoms. Antibodies appear following resolution of infection and the development of the latent state of infection. Not only will a significant proportion of donors have been exposed to the virus, but also a significant proportion of patients. However, patients who have not been previously exposed will normally only be at risk if they are immunosuppressed in any way. The main reason for the significance of HCMV in transfusion medicine is because leucocytes (including lymphocytes, monocytes and neutrophils) are one site of latency of the virus. Transfusion of blood containing leucocytes has been shown to lead to HCMV infection. Although antibody to the virus can be demonstrated in previously infected individuals, these antibodies do not necessarily prevent recrudescence or reinfection, and reactivation of the latent virus may occur. As with most of the herpesviruses, the key trigger to

218

reactivation/reinfection appears to be immunosuppression, whether due to specific immunosuppressive treatment or ‘natural’ immunosuppression. Although previously screening for both specific IgM and IgG antibodies to HCMV was considered essential, it is now accepted that screening blood for IgG antibodies is effective in identifying previously exposed and thus potentially infectious donors. However, screening is not generally applied to all donations because the percentage of patients requiring screened blood is relatively low. Epstein–Barr virus

EBV is the cause of infectious mononucleosis and is associated with other diseases, in particular Burkitt’s lymphoma and nasopharyngeal carcinoma. Like HCMV, its sites of latency include leucocytes, in this case B cells, but recrudescence appears to be a lot less common. The virus is globally widespread, with prevalence levels from 40% in industrialized countries to more than 90% in economically restricted countries. Only occasional cases of post-transfusion EBV infection have been reported. Donor screening is not performed, and is considered to be of limited value because of the high prevalence of the virus and because donors with active infection, i.e. those with infectious mononucleosis, are generally symptomatic. Human herpesvirus 8

HHV-8 (or Kaposi’s sarcoma-associated herpesvirus), first isolated in 1994, is the most recently identified herpesvirus. Our knowledge and understanding of this virus are incomplete, and its significance to transfusion is unclear. It causes Kaposi’s sarcoma, body cavity-based lymphoma and some severe forms of lymph node enlargement. Although clearly a different virus to HIV, there is a high frequency of HHV-8-related diseases among homosexually infected HIV patients. Infection rates vary significantly, ranging from less than 5% in northern Europe, 10% in the USA and increasing through southern Europe into Africa where over 50% of some populations are infected.

Transfusion-transmitted infections

The virus appears to be transmitted by both sexual and non-sexual routes, being more commonly transmitted through homosexual practices than heterosexual, but also through oral contact and even from mother to child at birth in higher prevalence countries. Transmission by organ transplantation has been reported, although it is rare, but this does suggest caution when considering transmission by blood transfusion.

Parvoviruses Parvovirus B19

The parvoviruses are one of the smallest DNA viruses that infect humans. They are very stable non-enveloped viruses that are resistant to many chemical and physical inactivation techniques. Parvovirus B19 is the only definite member of the genus erythrovirus (the virus replicates in erythroid progenitor cells). Clinically, parvovirus B19 infection gives rise to the following. • A range of generally mild symptoms including rash, vomiting, aching joints and limbs, fatigue, general malaise and leucopenia, and in many individuals the infection passes largely unnoticed. • Aplastic crisis in sickle cell and thalassaemia patients, cases of chronic haemolytic anaemia and other conditions with red cell membrane defects. It may also cause aplastic anaemia in immunocompromised individuals. • Severe fetal anaemia, death or malformation in infants infected in utero (mainly in the second trimester). B19 is fairly widespread among the general population, with regular community outbreaks across most countries. The prevalence of antibody to B19 in blood donors ranges from 50 to 98% from developed to developing countries. It is transmitted mainly via the release of virus particles from the upper respiratory tract, but may also be transmitted parenterally, by blood transfusion, at times of high viraemia. Viraemia usually appears within the first week of infection and persists generally for 1–2 weeks, although longer-term viraemia is not uncommon. The humoral immune

response normally begins after 1–2 weeks with the appearance of IgM, followed closely by IgG antibodies. Chronic infections do not occur. The detection of antibody to B19 indicates immunity to the virus. IgM testing may identify recently infected, and thus potentially still infectious, individuals. IgG antibodies persist for many years, possibly for life, and identify previously infected rather than infectious individuals. B19 has been demonstrated to be transmitted by transfusion, although the significance of any resulting disease is related to the immune status of the patient transfused. There is little clinical significance in immunocompetent individuals, with a greater but still relatively low clinical significance in immunocompromised individuals. Because of its low clinical significance, B19 is another virus, like GBV-C (see below), that is clearly transmissible by transfusion but with limited clinical relevance. The only area of major concern is that of large-pool fractionated products, where because of its natural resistance to inactivation, the virus can be found in some products in high titres during outbreaks. There is a concern that this may be of clinical relevance to immunocompromised patients receiving significant amounts of such products. Some fractionators have introduced lowered sensitivity NAT procedures to reduce the viral load prior to inactivation without excluding excessive numbers of donors. Parvovirus B19 over recent years has been used as a model for viral inactivation procedures, because of its high resistance to many inactivation methods. Miscellaneous viruses

There are a number of viruses that can be considered in this group: transmissible but which cause little or no clinical disease, or very restricted in their distribution or period of highest potential exposure. These viruses to a degree pose a dilemma as their significance varies significantly in different countries or even regions within a country, and to effectively remove the risk of their transmission from the blood supply may require significant intervention and resources, in some

219

Chapter 19

cases far in excess of any potential clinical consequence. In addition some, like West Nile virus (WNV), are not persistent but can be more frequent in epidemic periods in certain countries. GBV-C and hepatitis G virus

GBV-C and hepatitis G virus (HGV) are now known to have been simultaneous independent isolates of the same virus. Additionally, although originally called hepatitis viruses, there is doubt as to whether they do have any role in liver disease and their classification as a hepatitis virus may have been premature. In this text, the term GBV-C is used to refer to both the GBV-C and HGV original isolates. GBV-C is a recently identified flavivirus that has been shown to be present at a relatively high prevalence and incidence worldwide. The virus is transmitted parenterally and viral RNA has been found at a high prevalence in the expected high-risk groups, such as transfused individuals, transplant recipients, intravenous drug users, haemodialysis patients, infants born to infected mothers and sexual partners of infected individuals. The course of infection varies from an acute selflimited infection, with development of antibody to the viral envelope, to long-term chronic infection, with viral RNA production persisting for a number of years in the absence of an antibody response. Following infection, there is a short incubation period followed by the appearance of GBVC RNA together with a rise in serum alanine aminotransferase, although the levels do not reach the high levels seen in HBV and HCV infections. In acute infection, the RNA levels start to fall 4–6 months after infection, and levels of anti-E2 start to rise to a maximal level which may be maintained for 5 years or more. In a proportion of individuals, currently 20–40%, there is no clearance of the viraemia, anti-E2 is not detected and a chronic infection is established. A high proportion of infected individuals in the high-risk groups have also been found to be coinfected with HCV, but this is a consequence of the particular shared routes of infection. The virus has many general similarities with HCV in areas such as structure, genome organization, epidemiology 220

and general routes of infection, but there are also some significant differences, notably the lack of a core gene. GBV-C has been clearly demonstrated to be transmitted by transfusion, although any clinical significance associated with these transmission events has yet to be established. The high prevalence of the virus in ‘at-risk’ groups has not been shown to be, on its own, a significant contributory factor to any morbidity in infected individuals, nor does it exacerbate coinfection with HAV, HBV or HCV. The laboratory detection of active or recent infection is based on detection of viral RNA. Unlike HCV infection, specific circulating antibody has not been detected during the period of viraemia; the appearance of antibody appears to mark the clearance of viral RNA. The humoral immune response to GBV-C has not yet been characterized fully, but most of the seropositive individuals appear to produce antibodies only to the structural E2 antigen; humoral immunity to the non-structural proteins has not been found. In individuals on antiviral therapy viral RNA has been found to decline without the subsequent appearance of antibody. Although a lot of information about the virus has been collected in a relatively short time, there is still a great deal that is unknown, for example whether significant clinical disease is associated with it and, especially, the apparent current lack of disease association following transmission by blood transfusion. TT virus

TT virus (TTV) is a DNA virus that was first isolated in 1997 from a patient (T.T.) with posttransfusion hepatitis of unknown aetiology. The virus is non-enveloped and although it has some similarities with both parvoviruses and circoviruses, it is considered to be a member of a new virus family. The prevalence of infection varies from 3–4% in blood donors in the UK and the USA to as high as 80% in rural areas in some developing countries. Although transmitted parenterally, there is evidence of non-parenteral transmission and the contribution of each of these to transmission through

Transfusion-transmitted infections

populations is not fully understood. Transmission by blood transfusion has been demonstrated where it has been cited as the cause of non-A–G post-transfusion hepatitis, but the pathogenicity and role of TTV in liver disease is unclear. Detection of infection is currently by nucleic acid detection; serological tests have not yet been developed. Because transmission is not exclusively parenteral, the significance of TTV to blood transfusion is unclear. Sen V

This virus appears to be related to TTV and there are some indications that it might be of significance in cases of non A–E post-transfusion hepatitis, although there have been no cases reported in the UK. West Nile Virus

WNV is a flavivirus primarily transmitted by mosquitoes. Birds act as intermediate hosts, in which viral titres reach high levels. Mosquitoes feeding on the birds can then become infected, completing the cycle. Although they become infected by mosquitoes, humans and those other large mammals that are infected appear to effectively stop the cycle of the virus because viral titres never rise above the threshold needed for feeding mosquitoes to become infected. Historically, WNV has caused epidemics across much of the world and a fatality rate of 5–15% has been seen. Currently, attention has been focused on the USA, particularly New York and around the east coast, where WNV has been increasing over the last few years since its first appearance in 1999. There is a defined viraemia following infection and prior to the development of the humoral response. The viraemia generally lasts no more than 28 days, usually a lot less, and is followed by the rapid appearance of IgM followed by IgG. Infection is often asymptomatic and there is no chronic stage. Although identification of infectious donors/donations is by NAT, the relatively short period of infectivity followed by immunity enables alternative strategies to be considered, i.e. the deferral of potentially exposed donors to beyond

the viraemic period. However, the problem is identifying specific risk exposure; at-risk individuals need to be able to be identified clearly by the donor selection criteria adopted. Although there are specific geographical and seasonal factors, such a system is not exact and because of the high numbers of infected individuals who are asymptomatic, they may present with no identifiable risk factors and thus may not be deferred from donating. In the USA, where the major problem with WNV currently resides, cases of transfusion transmission from donors in the early window period and tested by pooled NAT have been reported. Because viral titres are generally low in humans, individual donor NAT may be required in this instance. Severe acute respiratory syndrome

Severe acute respiratory syndrome (SARS) is respiratory infection caused by the newly emergent SARS coronavirus (SARS-CoV). The disease has severe morbidity and mortality, but presents with non-specific signs and symptoms and there is no clear-cut diagnostic approach to prospectively identifying cases prior to the appearance of symptoms. The virus has a viraemic phase, which is found prior to symptoms and which persists into the symptomatic phase. Viraemic individuals may transmit if blood is collected during the early phase of infection prior to symptoms appearing. Screening of individuals for SARS-CoV is possible using NAT, but is not practical bearing in mind the current geographical restrictions/exposure risk of the virus. Deferral of donors who may have been potentially exposed to the virus is currently the most effective way of minimizing risk of transmission.

Bacteria The presence of bacteria in donated blood and tissues is assuming a greater importance in transfusion medicine, because although it is uncommon, post-transfusion bacterial sepsis has a high mortality rate (see also Chapter 16). Furthermore, the potential for contamination is increasing because 221

Chapter 19

of the increased manipulation of blood during the preparation of blood components. Numerous instances of infected red cell products have been reported, and the storage of platelet concentrates at 20–24°C for up to 5 days has provided an ideal environment for the growth of any contaminating bacteria present. There are two broad routes of bacterial contamination: • endogenous contamination due to bacteria present in the donor’s blood at the time of donation (bacteraemic donor), and which may include such organisms as Treponema pallidum and Yersinia enterocolitica; and • exogenous contamination due to bacteria that have entered the blood pack from the environment during collection, processing or other handling, storage or transport, and which may include such organisms as Pseudomonas and Staphylococcus. In either case, bacterial growth within the blood pack gives rise to endotoxins that are generally the major cause of the post-transfusion sepsis seen after transfusion of a contaminated unit. In these cases growth is restricted and the bacteria themselves do not survive past a few multiplication cycles, but the toxins they produce can have clinical significance. Endogenous bacteria

If a donor is bacteraemic at the time of donation, except for certain diseases such as syphilis, it is most likely to be due to a low-grade asymptomatic infection, often with only a short period of bacteraemia. Treponema pallidum (syphilis)

The spirochaete T. pallidum is the causative agent of syphilis. Treponemes: • have thin flexible helical walls and are extremely motile; • cannot be cultivated on artificial media although they can be cultivated in cell culture or in animals; and • can be seen easily under dark-field microscopy. Syphilis can follow several stages, leading to 222

primary, secondary and tertiary syphilis following the initial infection. The primary site of infection is usually marked by a lesion known as a chancre, which is full of treponemes. Although this may heal and disappear completely, the regional lymph nodes may still be infected and continuing treponemal division may give rise to secondary syphilis and, if still untreated, to tertiary syphilis. Because treponemes are released into the bloodstream as part of their life cycle, there is the potential for transmission by transfusion. They are particularly fragile, but can be transmitted by transfusion if they are present in the donation, and many transmissions were reported in the early days of blood transfusion. However, storage at 4°C soon destroys the organisms because they are very heat sensitive. It is generally considered that any spirochaetes present in the pack would be destroyed within 72 h of storage at 4°C. As spirochaetes can only be seen for short periods during infection, identification of infected individuals relies on serology. Blood is therefore screened using either non-specific tests for indirect evidence or specific tests for direct evidence (antibody to T. pallidum) of current or previous infection with T. pallidum. In countries with a low incidence of syphilis the vast majority of cases identified in blood donors are due to ‘old’ infections that have been treated successfully and present no risk of transfusion transmission, although some cases of recent primary acute syphilis are occasionally identified. In such countries, the value of continuing syphilis screening is often questioned as cases of post-transfusion syphilis are rare and any that may occur can be successfully treated with no lasting sequelae. However, syphilis screening of donated blood, no matter what the incidence in the donor population, has been considered to have value as a ‘lifestyle’ indicator, because individuals exposed to syphilis may also have been exposed to other sexually transmitted diseases and therefore should not donate. Non-specific screening tests include simple and rapid tests such as the Venereal Disease Reference Laboratory (VDRL) test and rapid plasma reagin test, which essentially detect current or recent infection, measuring the amount of anticardiolipin

Transfusion-transmitted infections

free in the blood and produced in the early course of infection in response to the organism. Although non-specific cardiolipin tests may give rise to a high number of false-positive reactions, they are useful because they do give an indication of current status; in treated acute infections anticardiolipin titres fall and the tests become negative. Specific screening tests detect antibody to T. pallidum and include both haemagglutination and EIA formats. Other tests in use, such as the T. pallidum immobilization test or the fluorescent treponemal antibody absorption test, are primarily confirmatory tests. Borrelia burgdorferi (Lyme disease)

Like syphilis, Lyme disease is caused by a spirochaete, Borrelia burgdorferi. The organism is carried by a number of insect vectors, mainly by ticks of the Ixodes genus, but it has been increasingly found in other blood-feeding insects such as horseflies and mosquitoes. It is likely that human transmission is possible via these routes. The disease was first identified around Lyme, Connecticut, in the USA, but is now the most common ticktransmitted infection in the USA, and is also known to be endemic in many other parts of the world. The disease is generally seasonal and marked by unique skin lesions, rash, fever and lymphadenopathy. This may progress to meningoencephalitis or myocarditis, and then arthritis. A high percentage of infected individuals develop chronic joint disorders. Although no case of post-transfusion Lyme disease has yet been reported, the potential for transmission remains as B. burgdorferi can retain viability in blood stored for up to 6 weeks. EIAs are available which detect specific antibody against the organism, but as cases of spirochaetaemia are generally symptomatic, careful donor selection should ensure that any potential risk of transmission is minimized. Brucella melitensis (brucellosis)

Brucellosis (undulant fever) is caused by the bacterium B. melitensis and has the following characteristics.

• It is usually acquired from an infected animal source. • It is not usually transmitted from person to person but because there is a period of bacteraemia, transmission by blood transfusion may occur and has been reported in endemic regions. • It normally enters through the mucous membranes of the throat, migrates to the regional lymph nodes, where it multiplies before being released into the blood, from where it enters and resides in the reticuloendothelial systems of different tissues. • Infection is characterized by general malaise and an undulating fever. • Chronic infection normally follows, which may last for many years with bouts of sometimes quite serious illness. • While the organism is prevalent in many parts of the world, brucellosis has only rarely been reported after transfusion. This may reflect poor reporting of post-transfusion infections and the true incidence of transmission may be at a higher level. The active deferral of donors at risk of exposure to, or previously infected with, Brucella should minimize any risk of transmission. Yersinia enterocolitica

Yersinia enterocolitica is a Gram-negative bacterium that may be present as an asymptomatic bacteraemia in infected donors at the time of donation. The extent and significance of such infections is difficult to assess as effective mass screening procedures are currently not feasible, and only sporadic monitoring is performed. Yersinia enterocolitica has been recognized as a potentially serious microbial agent that may be present in donated blood. It was thought that the organism persisted free in the blood and because it was psychrophilic (able to grow at low temperatures) it was able to multiply in the blood pack during storage, eventually reaching high enough numbers to cause posttransfusion sepsis in the recipient. However, it is possible that in an infected individual the bacteria are phagocytosed but survive intracellularly in the circulating leucocytes. During storage of the blood, the natural breakdown of the leucocytes 223

Chapter 19

releases the bacteria, which are then able to grow at 4°C. This results in the build-up of large numbers of bacteria and their toxins in the stored unit, with the potential for causing both posttransfusion sepsis and septic shock in the recipient. The screening of donations for Y. enterocolitica is not currently feasible; the deferral of potentially infected donors relies upon the routine donor selection procedures. Other endogenous bacteria

Cases of post-transfusion sepsis caused by a number of other endogenous bacterial species are occasionally reported. Donors may have a transient bacteraemia, possibly as a result of: • a low-grade gastrointestinal infection; • following dental procedures; and • a minor wound. Donor bacteraemia may subsequently result in post-transfusion sepsis due to organisms such as Salmonella, Campylobacter, Streptococcus and Staphylococcus. Exogenous bacteria

A number of bacterial species, for example Pseudomonas, Serratia and Staphylococcus, may be introduced into the blood pack from the environment during or after donation, either at venepuncture or during subsequent processing, storage or transportation of the blood. These environmental organisms are generally more likely to cause problems in blood components stored at higher temperatures, such as platelet concentrates, because they are potentially able to multiply to large numbers, producing high levels of bacterial toxins. There have been many cases of post-transfusion septicaemia involving single or multiple recipients infected by the same source, and involving a large range of bacterial species. However, while a large number of bacterial species have been implicated, in most cases the precise source of the contamination could not be determined. Platelet concentrates are a common source of post-transfusion septicaemia, although the ratio of cases of sepsis to

224

death is higher than with cases of post-transfusion septicaemia following transfusion of red cell concentrates. Some years ago an outbreak of post-transfusion septicaemia due to Serratia marcescens was reported. The outbreak affected transfusion centres in both Denmark and Sweden and was traced to blood bags contaminated on their outer surfaces during manufacturing or packaging. Six patients were affected, one of whom died, and 4000 units of blood and an unknown number of platelet concentrates had to be discarded. Studies on the growth of S. marcescens in artificially infected blood packs have since demonstrated that not only does the organism grow well at both 4 and 22°C, but that natural protection mechanisms such as phagocytosis and complement-mediated killing are not effective in destroying it. Rickettsiae

The rickettsiae are smaller than most other bacteria. They are most closely related to Gramnegative bacteria, but are unique in that they grow only inside animal cells. All rickettsial diseases are transmitted between animals via a blood-feeding insect vector, with the exception of Q fever. Transmission of rickettsial infections by blood transfusion has occurred, but cases are extremely rare and only transmission of Q fever and Rocky Mountain spotted fever has been reported. It occurs because infectious organisms are shed into the blood, and donations taken during this phase of rickettsaemia may transmit the infection. However, most infections at this stage are symptomatic and donor selection procedures should identify any potentially infectious donors. Some laboratory screening tests are available but are very specialized and not ideally suited for the screening of blood donations.

Protozoa Protozoan infections have always been a major problem in developing countries. However, with increased global travel and population migration,

Transfusion-transmitted infections

protozoan infections are now becoming a concern to all countries, and this concern extends to the safety of the blood supply. Plasmodium spp. (malaria)

Protozoa of the Plasmodium species cause malaria. There are four known Plasmodium species that are agents of human malaria: P. falciparum, P. malariae, P. vivax and P. ovale. Although there are some basic similarities in the life cycles of the organisms and the clinical features of infection, there are also significant differences. The incubation periods range from 12 days for P. falciparum, 15 days for P. vivax and P. ovale, and as long as 30 days for P. malariae. In general, complications from P. falciparum infection are more serious than from the other three species and are often fatal. The agent is transmitted to humans through the bite of the Anopheles mosquito, the life cycle of Plasmodium being split between the two hosts, with the sexual replication phase in the female mosquito and the asexual replication phase in the human. The merozoite form of the parasite infects the red cells, where it replicates, subsequently causing the red cell to burst, releasing more organisms into the blood. Only P. malariae persists for extended periods in humans (up to 30 years), while after 1–2 years plasmodia of the other three types normally die, and the individual, if not reinfected, becomes free from malaria. The transmissibility of malaria by blood transfusion has long been recognized because of the erythrocytic phase of the life cycle, and cases of post-transfusion P. falciparum are often fatal due to the complication of cerebral malaria. Immunity to plasmodia builds up in adults living in endemic areas, but is quickly lost upon moving to a non-endemic area. However, the recent development of immunity can be used to identify infected individuals by screening for specific antiplasmodial antibody. This is only of significant value when applied to individuals from lowprevalence areas who have travelled to endemic areas and when sufficient time between returning from an endemic area and testing being performed

has been allowed. Individuals from some endemic areas where the exposure rate is high may be found to be semi-immune, a situation in which parasites may be found circulating at the same time as lowlevel antibody. Such individuals are among the highest risk for transfusion transmission as the antibody levels may fluctuate to below detectable. However, such individuals may be identified by geography alone and deferred permanently. Direct detection of parasites in blood is possible, but in most cases is far too insensitive to detect any but the most parasitaemic individual, generally in a period of crisis and hence symptomatic, and is of no significant value in the screening of blood donations in most countries. Trypanosoma cruzi (Chagas’ disease)

The protozoan T. cruzi causes Chagas’ disease. The disease is confined mainly to the American subcontinent, where it is endemic in Latin America and increasing its presence in the southern states of the USA as migrant workers from Latin America move north. The agent is transmitted to humans by reduviid bugs (triatomides), who carry the parasites in their gut and excrete them in their faeces as they feed; the open feeding site acts as the site of entry for the parasite. Similarly to Plasmodium, the organism has a life cycle split between two hosts, the gut of the reduviid and the tissues and organs of humans. The blood acts as the transport system disseminating the organism around the body and provides a new source of organism for any feeding reduviids. Although the liver and spleen are usually infected, the most characteristic infected organ is the heart, and congestive heart failure plays a significant part in the morbidity and mortality of the disease. Although transmission by reduviids is generally the major route of infection, blood transfusion is considered to be the second most important route of transmission in endemic areas. Donor selection can be problematical due to the high proportion of asymptomatic infections (>20%), and in vitro screening of all blood donations is carried out in some parts of Latin America. This spread of the disease by travelling migrant

225

Chapter 19

workers is starting to affect other countries. Recent data from the USA include the report of a case of endemic Chagas’ disease in an infant in one of its southern states who had no identifiable risk of infection except insect bites. The risk of infection in travellers from non-endemic regions also has to be considered, and donor selection procedures amended accordingly. Once infected, an individual generally remains infected for life, although the morbidity and mortality associated with Chagas’ disease vary significantly. Like malaria, immunity to T. cruzi does occur, and can be used to identify individuals previously exposed to the organism and therefore not suitable as blood donors. Toxoplasma gondii (toxoplasmosis)

Toxoplasma gondii, the causative agent of toxoplasmosis, is globally one of the most widespread vertebrate protozoan parasites; in some countries up to 95% of adults may have been infected with the parasite. Members of the cat family are the hosts of T. gondii, and mice are thought to act as intermediate hosts helping to maintain its life cycle. The organism has a sexual replication phase in intestinal cells of cats and an asexual phase in another mammal. During this phase the sporozoites infect and multiply in a wide variety of other cell types, including those of the reticuloendothelial system, leucocytes and eventually the central nervous system (CNS). The acute infection in healthy individuals is generally asymptomatic and not associated with any morbidity. However, in immunocompromised individuals infection is far more severe, with the possibility of CNS involvement, myocarditis and pneumonia. Congenital infection can give rise to serious complications involving the liver and the CNS, and even abortion or stillbirth. Transmission by blood transfusion has occasionally been documented in immunosuppressed individuals, including some fatalities, due to the presence of the organism in leucocytes. Following resolution of acute infection, circulating antibodies appear but the organism persists latently in the circulating leucocytes, and reactivation has been reported. Although in vitro screening 226

for antibody to T. gondii is available, selective donor screening and leucocyte depletion of blood components may be more appropriate for the small group of recipients with a significant risk of disease. Babesia microti/divergens (babesiosis)

The tick-borne protozoan parasite Babesia (B. microti in North America and B. divergens in Europe) is the causative agent of babesiosis. The organism is transmitted by tick bite, and currently it is thought that the same reservoirs of Lyme disease are also the source of Babesia, and that the disease is restricted to the USA and northern Europe. Babesiosis is generally symptomatic, although symptoms range from a mild to severe malaria-like illness with the red cells acting as sites of replication of the organism. Studies with B. microti have shown that the organism can survive in red cells for at least 1 month under normal blood bank storage conditions and, like Plasmodium, can be transmitted by transfusion of blood from an infected asymptomatic individual. Identified transmissions are uncommon but not rare, as the disease is not widespread and symptoms are generally present, but no deaths have been reported. Although no cases of transfusion-transmitted babesiosis have been reported in Europe, it is quite likely that a number of cases have occurred and not been identified following asymptomatic infection in the recipient. Laboratory screening is currently not possible and donor selection procedures have to be relied on to minimize any risks of transmission. Leishmania spp. (leishmaniasis)

Infection with protozoa of the Leishmania spp. gives rise to leishmaniasis, which causes infection of the reticuloendothelial system and which exists in three main forms: cutaneous, mucocutaneous and visceral (kala-azar). It is thought that the basic differences between the three types of infection result from the differing ability of the Leishmania species to invade the body. Although a number of species exist, morphologically they are almost identical, and differences are only apparent when

Transfusion-transmitted infections

molecular techniques are used to examine their DNA. The organisms are transmitted through the bite of infected sandflies of the genus Phlebotomus, but each Leishmania species is restricted to a particular Phlebotomus species. The reservoirs for the organism vary among different regions but include rodents and other small wild mammals, although in urban areas dogs and even humans can serve as reservoirs. The life cycle is split between the two hosts, with the flagellated forms in the sandfly and non-flagellated forms in the vertebrate host. The organism invades the reticuloendothelial system, where it replicates and is released back into the blood. Although potentially a threat to the blood supply in endemic areas, parasitaemia is generally transient and at a low level, and consequently there is a low risk of transmission. This is supported by the lack of reports of transmission by blood transfusion even in endemic areas. Laboratory screening is currently not possible, and donor selection procedures have to be relied upon to minimize any risks of transmission.

Prions Prions have been included here as they have, in the shape of variant Creutzfeldt–Jakob disease (vCJD), had a significant impact on transfusion practice, at least in the UK, despite the absence of any firm evidence of transfusion transmission in humans. Variant Creutzfeldt–Jakob disease

vCJD was identified in 1996 and is the most recently identified human prion disease (see also Chapter 20). Prion diseases occur in a number of animal species and occur when the naturally occurring benign form of the prion protein (PrP) changes to an insoluble protease-resistant form (PrPSc). This leads to the formation of plaques in the brain. vCJD differs from classical CJD in that the age of onset is earlier and disease progression is slower, and higher levels of PrPSc are found in the brain. Higher levels of PrPSc than in CJD have been

found in tonsillar tissue, although these glycoforms differ from those found in the brain. The demonstration of lymphoid association in scrapie, together with the finding of PrPSc in tonsillar tissue, has led to a relationship being postulated between PrPSc and B lymphocytes. The transmissibility of vCJD by blood transfusion has not been demonstrated. Its presence in the blood is suspected, especially if the B-lymphocyte association is correct, but has not been demonstrated conclusively in anything but artificially produced situations using mice as models. Whether transmissions have occurred, or will do so, is as yet unknown. Certainly it would appear that the association of the disease with younger individuals raises some issues, as these individuals are more likely to be blood donors than older people. Surveillance data on classical CJD are available and no cases of transmission have been found, even in multiply transfused individuals. However, there are differences between the two diseases and insufficient time has passed for similar analyses to be performed for vCJD. Any potential risk of transmission of vCJD by transfusion will be exacerbated if large numbers of the population prove to be infected, and monitoring of the epidemic in the human population is therefore essential. At this time there is no in vitro test that can be applied to donated blood to detect infected individuals, and currently the approach taken by the UK is to try to remove any agent present by leucocyte depletion of all donated blood, removing the cells possibly harbouring the agent. However, experimental data supporting the benefit of this approach have not yet been obtained. Continued surveillance of vCJD is essential to determine if this approach to blood transfusion has been successful and has prevented transmission by transfusion.

Summary Our knowledge of the range of infectious agents that can infect donated blood is relatively broad but is still growing. However, the clinical significance of infection by a number of these agents is 227

Chapter 19

still unclear and thought has to be given to the responses of blood transfusion services to this everincreasing burden. If an infectious agent is not shown to be associated with pathogenicity and transmission events do not lead to clinical disease, should blood be screened for the agent? This question is currently being posed in response to the identification of GBV-C and TTV. Quite how to respond is unclear and there is the ever-present danger of public pressure in response to perceived rather than actual risks. Blood transfusion services do have to be alert and actively monitor changes, but reacting to such pressures without due thought must be avoided.

Further reading Alter HJ. G-pers creepers, where did you get those papers? A reassessment of the literature on the hepatitis G virus. Transfusion 1997; 37: 569–72. Arguin PM, Singleton J, Rotz LD et al. and the TransfusionAssociated Tick-Borne Illness Task Force. An investigation into the possibility of transmission of tickborne pathogens via blood transfusion. Transfusion 1999; 39: 828–33. Barbara J. Transfusion transmitted diseases: prions. In: The Compendium. Philadelphia: AABB, 1998: 308–12.

228

Hollinger FB, Kleinman S. Transfusion transmission of West Nile virus: a merging of historical and contemporary perspectives. Transfusion 2003; 43: 992–7. Infectious agents transmitted by transfusion. In: Mollison PL, Engelfreit CP, Contreras M, eds. Blood Transfusion in Clinical Medicine, 10th edn. Oxford: Blackwell Science, 1997: 509–57. Kitchen AD, Barbara JA. Transfusion transmitted nonviral infections. Curr Opin Infect Dis 1994; 7: 493–8. Murphy MF. New variant Creutzfeldt–Jakob disease: the risk of transmission by blood transfusion and the potential benefit of leucocyte-reduction of blood components. Transfus Med Rev 1999; 13: 75–83. Nishizawa T, Okamoto H, Konishi K et al. A novel DNA virus (TT) associated with elevated transaminase levels in post-transfusion hepatitis of unknown aetiology. Biochem Biophys Res Commun 1997; 241: 92–7. Pamphilon DH, Rider JR, Barbara JA, Williamson LM. Prevention of transfusion-transmitted cytomegalovirus infection. Transfus Med 1999; 9: 115–23. Reesink HEW, Engelfreit CP, Vreeslink H et al. Consequences of nucleic acid amplification testing for blood transfusion centres. Vox Sang 1998; 74: 263–70. Soldan K, Barbara JA, Heponstall J. Incidence of seroconversion to seropositivity for hepatitis C antibody in repeat blood donors in England, 1993–5. Br Med J 1998; 316: 1413–17. Williamson LM, Lowe S, Love EM et al. Serious Hazards of Transfusion (SHOT) initiative: analysis of the first two annual reports. Br Med J 1999; 319: 16–19.

Chapter 20

Variant Creutzfeldt–Jakob disease Marc L. Turner

Variant Creutzfeldt–Jakob disease (CJD) is one of a variety of transmissible spongiform encephalopathies (TSEs) described in animals and humans (Table 20.1). Bovine spongiform encephalopathy (BSE) was first described in cattle in the UK in 1986, though in retrospect the first cases probably appeared as early as 1982. It remains unclear whether it arose from scrapie in sheep or from a sporadic case of BSE in cattle, but it is thought the disease was transmitted through the food chain via rendered meat and bonemeal. In the UK over 180 000 clinical cases of BSE have been described with around 300 cases in other European countries, and occasional cases elsewhere in the world, probably related to exported UK cattle or meat and bonemeal. The UK epidemic peaked in 1992 and is now subsiding as a result of a ban on the use of ruminant protein in cattle feed. However, mathematical projections suggest that up to a million infected cattle could have entered the human food chain prior to the development of clinical disease. Unlike scrapie, BSE has proved itself capable of crossing species barriers by infecting up to 20 other species including exotic and domestic cats (feline spongiform encephalopathy) and exotic ruminants in zoos (exotic ungulate encephalopathy). In humans several forms of TSE have been described. Sporadic or classical CJD was first described in the early 1920s. It presents at a median age of 68 years as a rapidly progressive dementia with a duration of illness of around 6 months. The incidence of CJD is around 1 in 1 million per annum throughout the world, with no clear link to the incidence of TSEs in domestic animals. In the 1950s a form of TSE called Kuru was

described in the Foré of Papua New Guinea. This disease presented at a much younger age, with cerebellar ataxia as a prominent feature and a more prolonged clinical course and at one time was the leading cause of death in these tribal people. Kuru was transmitted from person to person probably through the cannibalistic funereal rites practised by the tribe at that time. It is informative to note that children died from Kuru and that although cannibalistic feasts discontinued around 1959–60, there are still occasional patients presenting with clinical disease. This points to a very wide range of incubation periods, with an upper limit of 40–50 years or perhaps even beyond normal human lifespan. In the 1980s a number of iatrogenic transmissions of CJD were described. These fell broadly into two groups. Direct central nervous system (CNS) transmission due to contaminated neurosurgical instruments, EEG electrodes and dura mater grafts led to a rapidly progressive dementia reminiscent of sporadic CJD after a short incubation period of around 2 years and death within about 6 months of presentation. Peripheral transmission from cadaveric pituitary-derived growth and follicle stimulating hormone gave rise to a clinical picture more reminiscent of Kuru, with a prolonged incubation period of some 13–15 years. Finally, a number of familial forms of CJD have been described including familial CJD, Gerstmann–Sträussler–Scheinker (GSS) disease and fatal familial insomnia (FFI), which arise due to polymorphisms in the gene for prion protein (PrP). TSEs are therefore an interesting type of disease from an aetiological perspective in that they can arise spontaneously, are transmissible and can also arise due to genetic polymorphism. 229

Chapter 20 Table 20.1 Transmissible spongiform

Animals

Human

Scrapie Chronic wasting disease Transmissible mink encephalopathy Bovine spongiform encephalopathy Feline spongiform encephalopathy Exotic ungulate encephalopathy

Sporadic Creutzfeldt–Jakob disease Kuru Iatrogenic Creutzfeldt–Jakob disease Variant Creutzfeldt–Jakob disease Familial Creutzfeldt–Jakob disease Gerstmann–Sträussler–Scheinker disease Fatal familial insomnia

Variant CJD The UK government instituted routine surveillance for CJD in 1989 in response to the BSE epidemic, with the aim of monitoring any change in the incidence or pattern of disease in the UK population. In 1995 the first cases of variant CJD were described. The clinical features differ from those of human sporadic CJD. Patients are younger, with a median age at presentation of 28 years (range 14–74 years). They often present with behavioural change, such as depression and anxiety, or with dysaesthesia. The disease progresses to cerebellar ataxia, involuntary movements, dementia and death over a period of 7–38 months. Over 135 cases of variant CJD have been described in the UK thus far, though the incidence of new cases appears to be falling. Elsewhere there have been eight cases described in France, one in Italy, two in the Republic of Ireland, one in the USA and one in Canada. The American, Canadian, and one of the Irish patients spent a considerable time in the UK, whereas the other Irish, French, and Italian patients did not, and probably contracted the disease in their own countries. Though original estimates of the number of people who may eventually develop the disease gave an upper limit of around 130 000, the recent downturn in the number of new cases in the UK has led to a revised prediction of just over 500 cases. These individuals are presumed to be currently incubating the disease and therefore at risk of passing it on to others via contaminated surgical instruments or blood transfusion. A considerable amount of epidemiological, clinical, neuropathological and experimental data now supports the view that variant CJD is the 230

encephalopathies.

same strain of disease as BSE, and that these are different from the TSE strains which give rise to other forms of CJD in humans or scrapie and chronic wasting disease in animals.

Aetiology and pathophysiology TSEs are associated with a change in the secondary structure of prion protein (PrP). PrP is a widely expressed 30–35 kDa glycoprotein with two Nlinked oligosaccharides. It is normally linked to the cell membrane by a glycosylphosphatidylinositol (GPI) anchor, though transmembrane anchorage has also been described. The normal secondary structure of PrP contains around 40% a-helices and 3% b-pleated sheets, with the membranedistal part of the molecule largely unstructured. The development of TSE is associated with a change in the secondary structure of the PrP glycoprotein, with an increase in the proportion of bpleated sheets to some 40–50% of the molecule largely at the expense of the unstructured region (PrPSc) (Fig. 20.1). This changes the physicochemical characteristics of the molecule, giving it increased resistance to both physical and biological degradation. In vitro treatment with proteinase K removes the membrane-distal part of the molecule, but is unable to digest the 30–32 kDa core. PrPSc accumulates in vivo giving rise to a form of amyloidosis. The pathophysiology of the disease remains debated. Some authorities propose the presence of a small DNA molecule associated with PrPSc (termed a virion), but this has not yet been identified and the infectious agent does appear to be resistant to physical conditions that would nor-

Variant Creutzfeldt–Jakob disease

PrP

C

Conformational transformation PrPSc Proteolytic degradation PrPSc

Fig. 20.1 The prion hypothesis. PrPc (top) is a 30–35 kDa

glycoprotein with two N-linked glycosylation sites, anchored by glycosylphosphatidylinositol to the cell membrane, with 40% a-helix and 3% b-pleated sheet. Transmissible spongiform encephalopathies are associated with conformational change in the secondary structure, with an increase in the amount of b-pleated sheet to some 40–50% of the molecule (middle). This changes the physicochemical and biological properties of the molecule rendering it resistant to degradation by enzymes such as proteinase K (bottom).

mally degrade viruses. The prion hypothesis proposes that the abnormal isoform of the protein is itself the infectious agent, changing the structure of the normal form either through heterodimer formation or through a physicochemical process of nuclear polymerization. Accumulation of amyloid plaques consisting of PrPSc leads to the classical neuropathological features of neuronal death, astrogliosis and spongiform degeneration of the CNS (Plates 20.1 and 20.2, shown in colour between pp. 304 and 305). In sporadic, iatrogenic and familial forms of CJD, abnormal PrP accumulation appears to be confined to the CNS. In variant CJD, abnormal PrPSc accumulation has been demonstrated in follicular dendritic cells (FDCs) in the tonsil, spleen, cervical, mediastinal, para-aortic and mesenteric lymph nodes and gut-associated lymphoid tissue of the appendix up to 2 years prior to the onset of clinical disease (Plate 20.3, shown in colour between pp. 304 and 305). This observation is consistent with what we know about the pathophysiology of transmission of TSE by peripheral routes in experimental systems. Experimental peripheral transmission of scrapie strains in murine models leads to the presence of

infectivity and/or PrPSc in the spleen and lymph node from a very early stage of infection, well before detection of infectivity or PrPSc in the CNS. Interestingly, immunosuppression and splenectomy have long been known to decrease the efficiency of peripheral transmission, whereas irradiation and thymectomy do not. In recent years a series of experiments have demonstrated that mice with severe combined immunodeficiency are resistant to peripheral but not central TSE challenge and that sensitivity is regained after allogeneic bone marrow transplantation. Similarly, PrPnegative mice with a PrP-positive CNS implant can only be affected by peripheral transmission following PrP-positive allogeneic bone marrow transplant, whereas PrP-positive mice are resistant to peripherally transmitted disease following PrPnegative bone marrow transplant. Detailed knockout experiments have demonstrated that Rag 1, Rag 2 and mMT knockout mice are resistant to peripheral challenge whereas CD4, CD8, bmicroglobulin and perforin knockout models display normal sensitivity. These data led to the suggestion that B lymphocytes were essential to peripheral transmission whereas T lymphocytes were not. However, B lymphocytes are also essential for FDC survival and more recent studies have demonstrated that PrP-positive FDCs are essential to peripheral transmission whereas PrP-positive B lymphocytes are not. Indeed peripheral transmission can be inhibited even by temporary FDC inactivation by lymphotoxin b receptor blockade and also by depletion of complement and complement receptors. These data convincingly support the seminal role of FDCs in the early stages of peripheral transmission.

Assessing the risk that variant CJD may be transmissible by blood transfusion It has been demonstrated that PrP is present in the peripheral blood of normal individuals at a concentration of 100–300 ng/mL, with the majority found in the platelets and plasma and 4–5% in the mononuclear leucocytes (Fig. 20.2). It has not, as yet, proved possible to demonstrate accumulation of PrPSc in the peripheral blood of humans, though 231

Chapter 20 10–300 ng/mL Plasma 68%

Leucocytes 3% Platelets 27% Red cells 2%

Fig. 20.2 PrP quantification in human peripheral blood components using DELFIA. (Reproduced with permission of Dr Ian Macgregor.)

one study has demonstrated such accumulation in the peripheral blood of scrapie-infected sheep. Most of the information on peripheral blood infectivity comes from animal experiments where it has proved possible to demonstrate infectivity in the peripheral blood of sheep and rodents with experimental scrapie and BSE, and in rodents with experimental GSS, during both the clinical and incubation phases of disease. However, no infectivity has been demonstrated in natural scrapie in sheep and goats, natural transmissible mink encephalopathy or natural or experimental BSE in cattle. The reason for these differences is not clear. Levels of peripheral blood infectivity have also been investigated in the Fukuoka-1 strain of GSS in experimental mice and have been found to be in the order of 100 infectious units/mL during the clinical phase of disease and 5–10 infectious units/mL during the incubation period. A fourfold to fivefold higher level of infectivity has been demonstrated in the buffy coat (containing the leucocytes and platelets) compared with plasma. Plasma itself shows a 10-fold higher concentration of infectivity compared with any of the Cohn fractions in an experimental fractionation system. The distribution of infectivity has been shown to be similar, albeit at lower concentration, during the incubation phase of disease. Similar findings have been demonstrated in a 293K infected hamster model, which also showed that five to seven times more buffy coat is required to transmit disease via the intravenous compared with the intracranial route. 232

More recently, sheep experimentally infected with BSE or scrapie by oral ingestion have been bled during the incubation and clinical phases of disease and whole-blood donations administered intravenously to secondary recipients; 10–20% of the secondary recipients in both cohorts have subsequently developed the relevant TSEs, amounting to proof in principle that certain forms of TSE can be transmitted by blood transfusion. In humans, of 37 reported attempts to transmit sporadic CJD from peripheral blood of patients with clinical disease by intracerebral inoculation into rodents there have been five positive reports. Interestingly, transmission of CJD from human peripheral blood to primates by intracerebral inoculation has not proved possible and this has thrown some doubt on the validity of the aforementioned rodent experiments. Thus far there have been no successful transmissions of variant CJD from human peripheral blood to rodents or primates, though experiments are ongoing.

Clinical transmission of CJD from the peripheral blood of patients There have been three anecdotal case reports of patients who have developed sporadic CJD some time after receiving blood components or plasma products. In none of these, however, has it been shown that the donors themselves developed CJD. In comparison a large number of epidemiological case–control, lookback and surveillance studies over the past 20 years have shown little evidence of increased risk of sporadic CJD in blood or plasma product recipients, even where a donor is known to have subsequently developed sporadic CJD. There are now a number of patients identified with variant CJD who, in the past, were blood donors. Recipients of blood components and plasma products from these donors have been traced and thus far one of their recipients has developed clinical variant CJD and another has shown evidence of abnormal prior accumulation in the spleen and a cervical lymph node.

Variant Creutzfeldt–Jakob disease

Assessing the level of risk If one assumes that the total number of individuals in the UK currently incubating variant CJD is up to 500, that 5% of the population are current blood donors and that 20% have given blood in the past, between 25 and 100 individuals currently incubating variant CJD have or are continuing to donate blood. This would give an incidence of potentially infected donations in the order of 1 in 100 000. In 1998 a risk assessment carried out for the UK Department of Health estimated that overall 2.6 recipients would be exposed per infected blood donation (prior to discontinuation of UK plasma for manufacture), but that the risk of exposure varied according to subgroups of patients with some having a risk as low as 1 in 1 million and others, such as haemophiliacs, having a risk of exposure approaching unity. However, if the incubation period is in the order of 13–15 years, many recipients of blood components in particular are unlikely to live sufficiently long to develop clinical disease due to their age or underlying disease. Thus it was felt that only 0.8 recipients per donation would be at risk of developing clinical variant CJD, of which 50% would be related to blood components and 50% to plasma products. Since the use of UK plasma for plasma product manufacture was discontinued in 1999, we can estimate that in the UK perhaps around 20 individuals per annum are being exposed to blood from a donor who will go on to develop variant CJD, of whom 10 are likely to live long enough to develop clinical variant CJD if the disease is transmissible by blood transfusion.

Strategies for risk containment While the risk of transmission of variant CJD by blood components remains uncertain and the number of patients exposed to such risk appears relatively low, the blood transfusion service has felt it prudent to consider precautionary policies to contain that risk. However, the introduction of such policies requires careful evaluation, both in terms of likely efficacy in reducing the risk of secondary transmission by blood transfusion and in

terms of the potential increase in other risks including that of blood shortages. Consideration also needs to be given to the cross-impact of different policies and the opportunity costs incurred. Donor selection

The UK blood services use a number of criteria for excluding blood and tissue donors who are at risk of sporadic, iatrogenic, or familial CJD (Table 20.2). There are no epidemiological risk factors described thus far that would discriminate a highrisk group for development of variant CJD within the UK. For example, there is no evidence that veterinary surgeons, cattle farmers, abattoir workers or others with high risk of exposure to infected bovine materials are at higher risk of developing variant CJD than the general population. In comparison, some individuals who have been vegetarians for prolonged periods of time have developed variant CJD. A number of other countries have taken the precautionary step of excluding blood donors who have spent more than a defined period in the UK between the beginning of 1980 and the end of 1996. The defined period varies depending Table 20.2 UK criteria for excluding blood and tissue donors who have, or may have had, contact with sporadic, iatrogenic, or familial CJD. Obligatory Permanently exclude donors with CJD or other prion-associated disorder Permanently exclude anyone identified at high risk of developing a prion-associated disorder: Recipients of dura mater, corneal or scleral grafts Recipients of human pituitary-derived extracts such as growth hormone and gonadotrophins Individuals at familial risk of prion-associated diseases.This includes individuals who have had two or more blood relatives develop a prion-associated disease and individuals who have been informed that they are at risk following genetic counselling Exceptions Individuals who have had two or more blood relatives develop a prionassociated disease but whom, following genetic counselling, have been informed that they are not at risk.This requires confirmation by the consultant with responsibility for donors

233

Chapter 20

on the frequency and pattern with which individuals in that country visit the UK and the likely risk of disease among endogenous donors. These are factors which impact upon the efficacy of UK donor exclusion in terms of risk reduction and on the likely negative impact on the blood donor base. Subsequent to the two probable transmissions of variant CJD priors by blood transfusion, the UK Blood Services have moved to defer blood donors who have themselves received blood transfusions. This has led to the loss of approximately 5–10% of the donor base. Importation of blood components from countries free of BSE/variant CJD

An alternative approach would be to source some or all blood components from countries free of BSE/variant CJD. It is considered impractical to source all UK red blood cell concentrates (some 2.5–3 million donations per annum) from non-remunerated donors from a country free of BSE/variant CJD. Consideration needs to be given to the risk of other infectious agents in the proposed alternative donor population and the longterm security of supply. The short shelf-life of platelet concentrates mitigates against sourcing these products from outwith the country. It is practicable to source plasma from a country free of variant BSE/variant CJD since the product can be virus inactivated and stored and transported frozen with a relatively long shelf-life. Another possibility would be to source blood components from outwith the country only for selected recipients considered most ‘vulnerable’. In the UK it has been decided to import methylene blue-inactivated plasma for neonates and children born after 1 January 1996. The rationale for this policy is that neonates in particular receive a relatively high number of blood components due to prematurity and surgery for congenital disorders, that this group of patients are likely to have a low exposure to BSE through the food chain and that they have the longest prospective lifespan during which to develop clinical variant CJD should they become infected.

234

Development of peripheral blood assays for donor screening

There is no conventional immune response to TSE infection and no DNA has been detected associated with the disease. Hence conventional serological and molecular approaches to the development of peripheral blood assays, utilized to such good effect in screening for microbiological disease, are not applicable to TSEs. A number of non-specific markers of CNS damage are known to be elevated in the peripheral blood of patients with CJD, including 14-3-3 and S100, but given that patients with CNS damage are excluded by existing donor selection criteria it seems unlikely that these would have much to offer in the context of screening normal healthy blood donors. Surrogate markers could allow exclusion of individuals at risk of development of variant CJD. Reduced transcription of erythroid differentiationassociated factor (EDAF) has recently been described in the bone marrow and peripheral blood of scrapie-infected sheep and rodents and of BSE-infected cattle, suggesting that EDAF may itself prove a useful marker of early disease and prompting a search for other subtle abnormalities in the peripheral blood. The gold standard is of course detection of PrPSc in the peripheral blood, although there are a number of fundamental problems. First is the analytical specificity in discrimination of PrPSc. It has proved problematic to develop monoclonal antibodies with a high degree of specificity for the abnormal conformer, probably because PrP itself is a widely expressed normal protein. Other agents such as plasminogen have been shown to bind differentially to PrPSc but the specificity appears to be low. Most groups have chosen to adopt physicochemical methods for discriminating normal and abnormal conformers, such as insolubility in nonionic surfactants, resistance to proteinase K digestion and changes in monoclonal binding affinity consequent on alteration of conformation by chaotropic agents. The analytical sensitivity is also problematic. If one assumes infectivity in human blood of 1–

Variant Creutzfeldt–Jakob disease

10 IU/mL during the incubation period of disease and that the ratio of infectivity to PrPSc is similar to that seen in animal models, then one can calculate that concentration of PrPSc in infected peripheral blood is in the order of 0.01–0.1 pg/mL (in the context of 100–300 ng/mL of PrPc). Nevertheless, a number of assays are under development, including Western blot, capillary immunoelectrophoresis and conformation-dependent dissociation-enhanced lanthanide fluorescence immunoassay (DELFIA), which are beginning to achieve the levels of sensitivity required. Many of these approaches use PrPSc concentration steps to increase the analytical sensitivity of the assay. A second problem is how such assays would be validated given that normally one would do so using samples from patients with the disease in question. Variant CJD assays are likely to have to be validated using animal model systems and human blood spiked with homogenized variant CJD-infected tissues given the rarity of the condition and the limited volumes of blood available from patients with clinical disease. There are a number of further issues that will need to be addressed as or when a putative blood donor screening assay becomes available. One is practicality of implementation: some assay systems are incompatible with current technology, some require large volumes of blood or other tissues from which to concentrate PrPSc, for others the length of time taken to carry out the assay would preclude their use as a release criterion for some blood components. Moreover, the overall sensitivity and specificity of an assay is dependent not only on its analytical features but also on the population under study. For example, an assay that has a high level of specificity in the clinical context of a patient with suspected disease may have a very low level of specificity (i.e. have a high false-positive rate) in the context of healthy blood donors. This point is made in Fig. 20.3, which illustrates the consequences of screening 3 million blood donors in the UK each year with an assay with 90% sensitivity and specificity. Assuming an incidence of preclinical variant CJD of 1 in 1 million, around three true variant CJD-positive donors per annum would be detected, while 0.3

3,000,000 donors vCJD1/million (0.0001%) Assay 90% effective True Positives 90% prevalence 0.00009% 2.7 donors

False Negatives 10% prevalence 0.00001% 0.3 donor

True Negatives False Positives 90%(1 – prevalence) 10%(1 – prevalence) 89.99991% 9.99999% 2,699,997 donors 300,003 donors

• Negative predictive value: 99.99998% • Positive predictive value: 0.00089% Fig. 20.3 Impact of a putative variant CJD assay with 90%

sensitivity and specificity on the blood donor population.

donors per annum (or one donor every 3 years) would be missed (false negative). The majority of donors would of course be true negatives, but a sizeable minority (just over 300 000 per annum) would be falsely positive. This brings into perspective the requirement for confirmatory assays, preferably based on different analytical principles. Even with confirmatory assays it may still be impossible to predict whether a repeat reactive individual will every go on to develop clinical variant CJD (and/or whether they are infective to others). The implications of positivity in terms of the psychological and social impact on the donor and the overall impact on donor recruitment and retention also need to be taken into consideration. Component processing

Universal leucodepletion was introduced in the UK in July 1998 and is predicated on the thesis that if variant CJD infectivity is present in the peripheral blood it is likely to be mainly associated with the mononuclear leucocyte population. Notwithstanding the evidence that PrP is widely expressed in peripheral blood, there is evidence to suggest both that the peripheral immune system is intimately involved in the early pathogenesis of peripheral transmitted TSEs and that in experimental animal models infectivity is present in the buffy coat at five to seven times the concentration seen in other components (as described previously). Modern leucocyte-depletion filters remove 3–4log10 of leucocytes with no evidence of

235

Chapter 20

selective subset removal but some evidence of cellular fragmentation. However, there is no proof that leucodepletion significantly reduces the risk of transmission of variant CJD and indeed the only experimental data available has shown no benefit of leucodepletion in reducing infectivity in plasma, suggesting its only likely impact would be on cellassociated infectivity. There are few other options in terms of component processing. Washing of red blood cells is impracticable and of debatable value. It would be feasible to supply platelet concentrates in optimal additive solution rather than plasma, though how effective this would be as a risk reduction measure given the high concentration of PrP in platelets themselves is unclear. Plasma products

Plasma product recall is not indicated if a blood donor develops sporadic CJD, based on the accumulated clinical evidence of a low risk of transmission in lookback and surveillance studies. In December 1997 the UK Committee for the Safety of Medicines recommended product recall if a donor became infected with variant CJD in view of the uncertainties surrounding transmissibility of the disease. In autumn 1999 the use of UK plasma for fractionation was discontinued altogether because of the recognition that a significant number of cases of variant CJD among donors would lead to multiple recalls and critical product shortages irrespective of the transmissibility of the disease. Most studies in fact suggest a reduction in infectivity by Cohn fractionation processes. Studies with 263K and 301V spikes using Western blot, DELFIA and infectivity bioassays as readouts suggest that cold ethanol precipitation, ionexchange chromatography, depth filtration and nanofiltration all give a 3–4log10 reduction, though whether these steps are additive is unclear. Criticisms of these studies surround the use of homogenized brain as the spike because infectivity may not be in the same physical form as that seen in naturally infected blood. The studies of Brown et al. referred to earlier have shown an overall reduction of up to 3–4log10 using plasma from 236

mice infected with the Fukuoka 1 strain of GSS. Although endogenous infection in this model may be predicted to be in a more relevant form to that seen in patients with variant CJD, the starting levels of infectivity are low and so estimates of the reduction in infectivity by plasma processing steps are likely to be conservative. A number of patients with variant CJD have donated plasma for product manufacture. The implicated batches have been identified and, where possible, the recipients traced and notified. Optimal use of blood components

There remains an urgent need to reduce blood usage both in order to manage the risk of unnecessary exposure to variant CJD and to reduce pressure on the blood supply at a time when significant reduction in the number of blood donors due to the introduction of new donor selection or screening criteria is a real possibility. Key issues to be addressed include better evidence of the efficacy of current clinical transfusion practice, reduction in blood outdate and discard rates, increased use of autologous blood and adoption of bloodconserving approaches.

Further reading Brown P. The pathogenesis of transmissible spongiform encephalopathy: routes to the brain and the erection of therapeutic barriers. Cell Mol Life Sci 2001; 58: 259–65. Brown P, Cervenakova L, McShane LM et al. Further studies of blood infectivity in an experimental model of transmissible spongiform encephalopathy with an explanation of why blood components do not transmit Creutzfeldt Jakob disease in humans. Transfusion 1999; 39: 1169–78. Brown P, Will RG, Bradley R et al. Bovine spongiform encephalopathy and variant Creutzfeldt Jakob disease: background, evolution and current concerns. Emerging Infect Dis 2001; 7: 6–16. Brown P, Cervenakova L, Diringer M. Blood infectivity and the prospects for a diagnostic screening test in Creutzfeldt Jakob disease. J Lab Clin Med 2001; 137: 5–13. Collee JG, Bradley R. BSE: a decade on. Part 1. Lancet 1997; 349: 636–41. Collee JG, Bradley R. BSE: a decade on. Part 2. Lancet 1997; 349: 715–21.

Variant Creutzfeldt–Jakob disease Collinge J. Variant Creutzfeldt Jakob disease. Lancet 1999; 354: 317–23. Foster PR. Prions and blood products. Ann Med 2000; 32: 501–13. Houston F, Foster JD, Chong A et al. Transmission of BSE by blood transfusion in sheep. Lancet 2000; 356: 999–1000.

Macgregor I. Prion protein and developments in its detection. Transfusion 2001; 11: 3–14. Turner ML. Variant Creutzfeldt Jakob disease and blood transfusion. Curr Opin Hematol 2001; 8: 372–9.

237

Part 4

Practice in blood centres and hospitals

Chapter 21

Donors and blood collection Liz Caffrey and Moji Gesinde

Categories of blood donor A repeat, voluntary, non-remunerated donor source for all blood and plasma donations has long been widely advocated because it offers the lowest risk for transfusion-transmitted infection. It is the policy of the International Society of Blood Transfusion, the World Health Organization and the International Federation of Red Cross and Red Crescent Societies and has been achieved for all cellular and fresh frozen components in some countries including the UK and France. In the USA all cellular products are collected from unpaid donors but the majority of plasma for fractionation is still obtained from paid donors, although with stringent quality specifications. Since the 1998 ban on the use of UK plasma for fractionation, because of the theoretical risk from variant Creutzfeldt–Jakob disease (vCJD), all plasma fractionated in the UK has been sourced from paid donors from the USA. Paid blood donors are still widely used in many other parts of the world, usually in developing countries with hospital-based transfusion services that depend on replacement or directed donations. Table 21.1 defines different types of both whole blood and apheresis donors. The majority of donors donate 450–500 mL whole blood at intervals of 3–6 months. A small number of donors volunteer instead to donate by apheresis. Blood components are obtained from them using a variety of cell separator machines which collect specific components, usually platelets and/or plasma, and return the rest of the donation to the donor. These donors may donate as often as once every fortnight. In England, there were 1 700 000 active whole blood donors and

10 000 apheresis donors in 2003; 40% of the national demand for platelet concentrates was prepared by apheresis.

Blood donor motivation Altruism, an unselfish concern for other people, is repeatedly quoted as the prime motivation for volunteer unpaid blood donors. In qualitative research, donors show a greater awareness of the clinical need for blood and are more willing to take personal action to meet that need than individuals who do not donate blood. They may act from ‘enlightened self-interest’, in the knowledge that the need may one day be personal. Peer pressure is also a strong motivator and is confirmed by the fact that first-time donors are frequently recruited by friends or colleagues who already donate. Blood donor sessions based at, or publicized through, work or college maximize the role of peers and provide convenient opportunities to give blood. For committed blood donors, external recognition does not appear to serve as a powerful motive whereas intrinsic factors such as self-image are more important. Specific conditions that are likely to encourage donation include: • family tradition; • close personal and professional experience of transfusion; and • appreciation of any characteristic that makes their donation special, e.g. rare blood group or selected for neonatal use. Some donors may be motivated by time off work, others by the unsubstantiated belief that blood donation is beneficial to their health. Individuals unwilling to donate commonly give health 241

Chapter 21 Table 21.1 Types of blood donor.

Table 21.2 Donor recruitment strategies.

Volunteer, unpaid Donate for the benefit of others Receive no financial gain

Increase public awareness TV/radio advertising Posters Leaflets

Paid Receive payment for donation Replacement Donors recruited from patients’ friends and relatives to replace units used Directed Family or friends of patient donate specifically for that patient’s use

Direct marketing Write or telephone preselected potential donors Personal contacts Friends and family Workplace/college Street campaigns Local organizers

Autologous Patient’s blood is collected for their own use

reasons, fear of needles, inconvenient session times and simple laziness as reasons not to donate. Historically, volunteer unpaid donors have had fewer positive test markers of transfusiontransmitted diseases and their donations have caused less post-transfusion hepatitis than those from paid donors. However, with the relentless need to recruit donors there is ongoing debate about the use of financial and other incentives and their impact, if any, on recipient safety. There is evidence that first-time donors may be attracted by such inducements. In contrast, regular donors may be demotivated by rewards which contradict their self-image as altruistic people. They are rewarded by a sense of well-being following donation and by the feeling of having contributed to society. Nevertheless, recognition in the form of small nonmonetary gifts, such as lapel badges, often presented at award ceremonies to recognize landmark donations, may encourage regular donors to volunteer more frequently.

Donor recruitment Effective donor recruitment depends on public awareness of the need for blood, which can be: • instilled in children through the school curriculum (young children as parental motivators and teenagers as future donors themselves);

242

• maintained in the public mind by a readily recognized national identity; and • reinforced by locality-based publicity campaigns. Advertising and publicity material should be consistent with local cultural and motivational factors. There are many common misconceptions surrounding eligibility to donate. Widespread community education to dispel these myths is an essential part of successful donor recruitment strategies. Information systems that provide demographic data, such as age and lifestyle profiles for populations in specified geographical areas, can be used to obtain a detailed knowledge of the potential for donor recruitment. Modern marketing tools help blood services plan recruitment strategies targeted at individuals most likely to respond to direct approaches by mail or telephone. Traditionally, a personal approach by friends or colleagues has been the mainstay of donor recruitment and remains an important tool. Volunteers working with the local community also have a very important role, particularly in rural areas. The various types of donor recruitment strategies are listed in Table 21.2. However, donor recruitment is only successful if blood services respond promptly to enrolment requests and enquiries about donor eligibility. Therefore, they must be supported by a robust communication infrastructure able to enrol and advise potential donors and initiate an early invitation to a con-

Donors and blood collection

venient blood donor session. In England, a 24-h Donor Helpline established in 1999 takes 600 000 calls per annum. It provides immediate answers to most eligibility queries and up-to-date information about session dates and times. In addition, it has facilitated the development of an effective appointment system.

Donor retention Donor retention is a more cost-effective way of maintaining the donor base than donor recruitment. In the UK 15% of donors lapse each year. The principal reasons are summarized in Table 21.3 and discussed below. Legitimate medical conditions exclude some donors either temporarily or permanently. There is evidence that temporary deferral, e.g. for low haemoglobin, deters donors from returning. It is important that they are not deferred unnecessarily and that temporarily deferred donors are encouraged to return. A warm, welcome, caring but professional attitude from blood collection staff and a genuine ‘thank you’ all contribute significantly to donors’ well-being and encourage them to return. Small tokens of appreciation and donor awards marking milestone donations also encourage donor loyalty. Conversely, negative experiences of donation lead to loss of donors. This can be caused by administrative problems or adverse reactions to donation. Administrative problems can be improved by:

Table 21.3 Why donors lapse.

Health Medical condition/pregnancy Negative experience of donation Moved Left work/changed job Moved house Inconvenience Busy lifestyle Venue/times unsuitable

• holding donor sessions in pleasant accessible venues which are clearly sign-posted; • preventing wasted journeys to give blood by providing effective predonation literature and answering donors’ queries in advance, e.g. through the use of a donor helpline; • reducing waiting times by use of an appointment system and good queue management. Adverse reactions to donation (see below) can be minimized by: • well-trained and sympathetic staff; • prompt and effective first aid and advice; • careful follow-up by blood service clinical staff of donors who have experienced a significant incident such as arterial puncture or severe faint; and • appropriate advice about future donations. The most important factor in donor retention is availability of convenient opportunities to donate. Moving home or job may make the donor’s usual donation venue unsuitable in terms of opening hours and/or location. Creative session programming that provides donors with choice and information about alternative venues should be widely available.

Organization of a blood collection programme Blood collection targets are based on information from hospitals about expected demand for blood components. Collection programmes are planned to meet anticipated seasonal and local variations in demand. Apheresis activity is driven by the demand for components, usually platelets and/or hyperimmune plasma. The donor database is confidential and a valuable resource that must be carefully maintained. Donors receive personalized invitations to attend and must be eligible in terms of age and donation interval and must not have been deferred for any reason. The number of invitations issued takes into account the predicted rates of response and donor deferral based on historical data. In large organizations this can only be achieved efficiently with sophisticated IT support. Response is variable,

243

Chapter 21

ranging from 20 to 50% in the UK and can be enhanced by reminder cards or a telephone call or an SMS text message a few days before the session. Donor deferral rates depend on national donor selection criteria and the effectiveness with which they are applied. In England, 3% of 2.6 million donors attending sessions in 2002–03 were deferred because of low haemoglobin and a further 9% for medical and other reasons. In the UK, more than 60% of blood is collected by mobile units visiting local businesses, colleges and community centres. They travel to venues with a team of staff and all the equipment and disposables needed to run the day’s session. Advanced planning is needed to ensure that venues are booked and provide suitable facilities. The staff unload and set up the beds and equipment before the first blood donor is scheduled to arrive. At the end of the day the staff reload the vehicles and travel back to base. Some staff operate from specially designed vehicles called bloodmobiles. These are used if suitable venues are not available, such as industrial estates where appointment-based sessions can be planned for staff from a number of small companies. Fixed-site clinics in areas of high population density offer shoppers and commuters increased opportunities to donate. However, they carry high overheads and facilities are therefore often shared with donor apheresis programmes, which historically have only been undertaken at fixed-site clinics.

Donor selection guidelines All blood services apply donor selection criteria with the aim of maintaining a safe and adequate blood supply and avoiding harm to donors. The selection criteria should be evidence-based, relevant to the local population and subject to regular review. New criteria should be carefully evaluated to balance any reduction in risk to patients against donations lost, cost and impact on donors. For instance, in the UK, if recipients of blood transfusions were permanently excluded as donors because of concern about the unknown risk of transfusion-transmitted vCJD, the donor base would be reduced by at least 6%, compromising 244

the blood supply and creating a real risk to patients. Criteria relevant in one part of the world may be inappropriate or impractical elsewhere. In Europe, visitors to malaria endemic areas are deferred as donors for 6 months to avoid the risk of transfusion transmission. A similar rule in a malaria endemic country would reduce the blood supply to zero. Donors’ health must not be harmed by giving blood and selection criteria reflect this. Donors should be fit and well, meet local requirements for weight, age, haemoglobin and minimum donation interval. Where medical history or medication indicate that a potential donor might be made worse or react adversely to donation, he or she must be deferred. In the 1960s it was recognized that transfusiontransmitted viral hepatitis could in part be eliminated by excluding donors with a history of jaundice. In the 1970s routine testing for hepatitis B showed a high prevalence among prisoners so that by the mid-1970s blood was no longer collected in prisons in much of Europe (not France) and the USA. The emerging AIDS epidemic in the early 1980s led to recognition that certain patterns of behaviour were associated with a high risk of human immunodeficiency virus (HIV) and hepatitis. This still forms the basis of today’s donor selection criteria, the main purpose of which is to eliminate transfusion-transmitted infection. Table 21.4 lists information provided in a leaflet for potential donors in the UK explaining the high-risk behaviours that are cause for permanent or temporary exclusion. In countries where blood donors are relatively plentiful and the prevalence of transfusion-transmitted infection is low, selection criteria now also take account of other small or uncertain risks such as vCJD and malignancy. The donor selection rules for all four countries in the UK have been agreed, reviewed and published annually since 1993. Some of the most important reasons for permanent and temporary exclusion are listed in Table 21.5. Apheresis donors are usually recruited from established whole blood donors. In addition to meeting normal selection criteria, they are subject to more detailed physical assessment and more

Donors and blood collection Table 21.4 High-risk behaviours which exclude a donor

Table 21.5 Outline of donor exclusion criteria (UK).

permanently or temporarily in the UK. You must never give blood if: You are HIV positive You are a hepatitis B carrier You are a hepatitis C carrier You are a man who has ever had oral or anal sex with another man, even if you used a condom or other protective You have ever received money or drugs for sex You have ever injected, or been injected with, drugs; even a long time ago or only once. This includes body-building drugs You must not give blood for at least 12 months after sex (even if you used a condom or other protective) with: A partner who is, or you think may be: HIV positive A hepatitiis B carrier A hepatitis C carrier (If you are a woman) a man who has ever had oral or anal sex with another man, even if they used a condom or other protective A partner who has ever received money or drugs for sex A partner who has ever injected, or been injected with, drugs; even a long time ago or only once.This includes body-building drugs A partner who has, or you think may have been, sexually active in parts of the world where HIV/AIDS is very common.This includes most countries in Africa

Donor safety Weight less than 50 kg Under age 17 or over 69 (59 for new donors) Minimum donation interval (usually 16 weeks) Pregnancy and until child is 9 months old Cardiovascular disease Surgery Medical investigations/waiting list Recipient safety Permanent exclusions High-risk behaviours (see Table 21.4) Chronic infections, e.g. HIV, HBV, brucellosis, Chagas’ disease Risk of CJD, e.g. recipients of human pituitary hormones or dura mater grafts Diseases of unknown or viral aetiology, e.g. malignancy, ulcerative colitis Temporary exclusions Sexual contacts of high-risk partners (see Table 21.4) Ear or body piercing Tattoo, acupuncture Blood transfusion Travel to malaria-endemic country Infectious diseases and contacts of infectious diseases Vaccination CJD, Creutzfeldt–Jakob disease; HBV, hepatitis B virus.

stringent criteria against which they are regularly monitored (e.g. full blood count, plasma proteins). Donors are further selected on the basis of blood group, cytomegalovirus status, human leucocyte antigen (HLA) type, human platelet antigen (HPA) type and/or the potential to produce multiple doses of platelets from a single procedure. Donors homozygous for common HLA types, e.g. HLAA1-B8, are particularly useful in supporting alloimmunized patients who require HLAmatched platelet transfusions and are encouraged to attend regularly.

Donor selection methods Effective donor selection procedures depend on good communication between the blood service and potential donors. Methods used to achieve appropriate deferral are given below. The effectiveness of all these methods depends on the use of

staff trained in interview skills and able to assess donors’ comprehension and suitability immediately prior to donation, but is reliant upon the honesty and understanding of the donor. Maintaining total confidentiality throughout the whole procedure is vital to achieving honest answers. Predonation information

Potential donors should be given sufficient information to allow self-exclusion. Selection criteria must be a prominent feature of recruitment campaigns and all communications with donors. Table 21.6 is an extract from information printed on the reverse of donor invitation letters in England in 2003. Donor assessment

At the collection session potential donors are 245

Chapter 21 Table 21.6 Extract of information provided to potential donors (England 2003).

You should not give blood if • You have already given blood in the last 12 weeks (normally you must wait 16 weeks) • You have a chesty cough, sore throat or active cold sore (the end of a cold is OK) • You are taking antibiotics now or you have finished taking them within the last 7 days • In the last 12 months you have had hepatitis or jaundice, ear piercing, body piercing, tattooing, or you have received a blood transfusion yourself • You have had acupuncture in the last 12 months outside the NHS (unless you can produce the approved certificate from your acupuncturist) • If two family members (parent, brother, sister, child) have suffered with CJD You may not be able to give blood if • You have had a serious illness or major surgery in the past or are on medication at present. Please discuss this with our clinical staff.The reason you are taking medicines may prevent you from donating • You have had complicated dental work (simple fillings and scale and polish are OK after 24 h, simple extractions are OK after 7 days) • You have been in contact with an infectious disease or been given certain immunizations in the last 4 weeks • You are presently on a hospital waiting list or undergoing medical tests Travel abroad If you answer YES to any of these questions, please ring the Donor Helpline to check when you can donate 1 Have you ever had malaria or an unexplained fever associated with travel? 2a Have you ever lived in or visited Africa or Papua New Guinea for a period of 6 months or more? 2b If yes, have you visited any malarious area in the last 5 years? 3 Have you visited any malarious area in the last 12 months? 4 Have you ever visited Central/South America for a period of 4 weeks or more? CJD, Creutzfeldt–Jakob disease.

questioned to ensure conformance with local selection criteria. Questions are administered either during a face-to-face structured interview and/or by questionnaire. Research in the USA suggests that direct questions (e.g. Did you? Have you?) are more effective than indirect questions. Written questions are of no use if a donor is unable to read or comprehend due to lack of education, blindness or language barriers. Staff must be sensitive and alert to these possibilities and identify donors who need assistance. Interactive computer interviews are claimed to be more successful than these standard methods. It is not only ‘at-risk’ lifestyles that are uncovered by these methods but also other risks such as recent travel to malarious areas, contact with infectious diseases, vaccination and relevant medication. A donor health check questionnaire (DHCQ) was introduced in England in 1998. Since July 2002 this has been mailed to donors’ homes with the invitation to attend and they complete the DHCQ themselves at home. In addition, new donors and those who have not donated within the 246

last 2 years also undergo a detailed confidential interview at the session. This predonation interview for regular donors is similar but shorter, recognizing their familiarity with the process and the selection criteria. Haemoglobin testing

The selection procedure must include a validated method to ensure that the donor’s haemoglobin is above the minimum level acceptable. A quick semi-quantitative gravimetric method using copper sulphate solutions (with specific gravities equivalent to 12.5 and 13.5 g/dL for women and men, respectively) is commonly used. If a drop of capillary blood sinks, the donor can be accepted for donation. If not, further testing of venous blood in a portable haemoglobinometer provides a haemoglobin value that enables the donor to be managed appropriately. Acceptable values for venous blood samples in UK are 12.0 and 13.0 g/dL for women and men respectively.

Donors and blood collection

Physical examination

In some countries, the donor interview is supplemented by a medical examination. It may be limited to measurement of the pulse and blood pressure or can be more thorough. In the UK, reliance is placed on simple visual assessment of the donor. Confidential unit exclusion

Confidential unit exclusion (CUE) procedures were developed in the mid-1980s, prior to the availability of anti-HIV tests, to enable individuals at risk of HIV infection, but who had nevertheless given blood, to indicate that their donation should not be used for transfusion. This can be achieved either by a confidential questionnaire completed at the time of donation or a telephone callback system. Only a small percentage of donations excluded by CUE questionnaire are subsequently found to be positive for markers of transfusiontransmitted infection. Conversely, only 25% of marker-positive donations are identified by CUE questionnaires. The apparent low sensitivity and low specificity of CUE in identifying donations at risk of transmitting HIV infection have fuelled controversy about its value. However, CUE is a method which can identify some window period donations. Local evaluation of the added value of CUE is necessary to prove its usefulness in the context of current local selection criteria, interview methods and testing strategies. The CUE questionnaire is not used in the UK but donors are encouraged to report any information relevant to their donation and a 24-h telephone helpline facilitates confidential contact with blood centre clinical staff.

The donation process The principles of quality assurance and good manufacturing practice are as important in the management of blood collection as they are in the laboratories which test and process the blood. Failure to follow approved procedures at session can lead to severe and potentially fatal conse-

quences for patients. Whole blood collection and apheresis programmes are integral parts of the blood service and subject to the same inspections. This demands suitably qualified staff, trained for the purpose, using validated written instructions called standard operating procedures (SOPs). These are controlled documents; this means that only the current authorized version is in use. A process of regular review and update of SOPs is required as well as a system to issue new versions and recall previous issues in a controlled way. SOPs are important training tools that enable staff to undertake procedures in a standardized way. It is a challenge for staff to work continually to maintain these essential standards while at the same time creating a relaxed and friendly environment for donors, especially in hired community venues with variable facilities. Important quality aspects of the collection process are: • checks of donor identity; • donor selection; • check for integrity of blood pack/apheresis set; • labelling of packs and samples; • data capture (paperwork or computer); • arm cleansing; • venepuncture technique; • mixing of packs to prevent clot formation; • storage and transport conditions; and • health and safety considerations.

Complications of donation Blood donation is painless and uneventful for the majority of donors. Donors should be given information about possible adverse reactions and encouraged to report all such incidents so that first aid and follow-up is given by blood service clinical staff to ensure full resolution and appropriate advice about future donations. • Minor bruising at the venepuncture site is not unusual, but occasionally can be extensive and painful. This can be minimized by ensuring that phlebotomists are highly skilled and that staff and donors are made aware of the importance of immediate sustained pressure to the venupuncture site after donation. 247

Chapter 21

• Faints and feeling faint are the commonest adverse reactions and are much more frequent in first-time donors, teenagers and nervous people. Low body weight and delays during the donation procedure also contribute. Faints complicated by, for instance, convulsions or head injury occur much less frequently. The majority are dealt with by blood service staff but approximately 1 in 150 000 donors may require hospitalization. Delayed faints do occur and may place the donor, and others, at risk of injury. To minimize this, donors whose work or hobbies could make a delayed faint more hazardous (e.g. scaffolder, train driver) should be advised not to donate unless work is finished for the day. • Neurological needle damage occurs in 1 in 6300 donations. Symptoms include numbness and tingling, pain and, less often, weakness of arm or hand. Recovery may take several weeks or months but is usually complete. • Arterial puncture is a very rare but potentially serious complication that can lead to arteriovenous fistula formation requiring surgical correction. In addition to the above, apheresis donors may suffer reversible symptoms due to citrate toxicity and, very rarely, haemolysis or clotting. However, the incidence of severe complications for both whole blood and apheresis donors remains extremely low.

Future developments Shrinking donor numbers and increasing demand from hospitals mandate that blood services develop ever more inventive programmes to encourage recruitment and retention of donors. Smaller mobile collection teams visiting a number of different venues in one day, e.g. doctors’ surgeries or small companies, could provide increased, more convenient opportunities for donors to donate. Flexible opening hours, appointment systems and efficient streamlined collection procedures will make it more practical for people to fit blood donation into already busy lives. Strategies to improve the donors’ experience of donation, e.g. by employing non-invasive technology for the 248

predonation haemoglobin assessment and minimizing adverse events, would also encourage donors to return. The cautious introduction of limited incentives to donate may be necessary to ensure an adequate blood supply. This might take the form of health promotion at collection sessions, credits towards some aspect of healthcare or health screening programmes. An ageing population and innovative treatment regimens for disease suggests that the demand for blood components will increase. Apheresis allows the flexible collection of tailored components from individual donors. For example, a single apheresis donor, carefully selected for platelet count and haematocrit, can donate the equivalent of 2–3 adult therapeutic doses (ATD) of platelets (equivalent to platelets recovered from 8–12 whole blood units). In addition, combinations of components such as 1 unit of red cells and 1–2 ATD of platelets, or 2 units of red cells can be collected at a single visit. Additional, more stringent selection criteria are required to ensure that donors tolerate these donations and avoid any adverse effects. Most manufacturers of current apheresis systems are developing newer machines that are much more portable than the previous generation. Although they use the same principle of separation of blood by centrifugation, their portability means they can be used on bloodmobiles and at mobile session venues. The artificial divide between whole blood collection and apheresis, traditionally at mobile sessions and fixed-site clinics respectively, is no longer relevant. Many years of experience with apheresis have confirmed its excellent safety record. The future is likely to see the introduction of sessions at which portable automated blood component collection (apheresis) systems are used side by side with the traditional way of collecting whole blood into plastic bags. One benefit of automated collections is that they can produce leucocyte-depleted components at the bedside that require no further processing, reducing the need for transport to large-scale processing laboratories. Another benefit is the ability to obtain red cell products of preselected haematocrit, which could standardize the product issued to hospitals, particularly important in the management of transfusion-dependent patients.

Donors and blood collection

Another strategy to maximize the donation potential of donors will be the introduction of more personalized selection criteria. For example, selected heavier donors with high haematocrits could safely donate 2 units of red cells at each session or donate 1 unit at more frequent intervals. Conversely, smaller donors, usually women, with borderline iron stores should be encouraged to donate less frequently, preventing deferral for low haemoglobin which often demotivates them from volunteering again. These initiatives cannot be undertaken on a large scale without the parallel development of more sophisticated IT systems and the collection of more precise donor details, such as weight, dietary history, a full blood count and, perhaps, a suitable measure of iron stores.

Further reading Chiavetta JA, Deeks S, Goldman M et al. Proceedings of a consensus conference: blood-borne HIV and hepatitis. Optimising the donor selection process. Transfus Med Rev 2003; 17: 1–30. Donor Selection Guidelines. Guidelines for the Blood Transfusion Service in the United Kingdom (8th version) (DSG 008). UKBTS/NIBSC, July 2001. Eastland T. Monetary blood donation incentives and the risk of transfusion transmitted infection. Transfusion 1998; 38: 874–82. Gimble JG, Friedman LI. Effects of oral donor questioning about high-risk behaviours for human immunodeficiency virus infection. Transfusion 1992; 32: 446–9. Glynn SA, Williams AE, Nass CC et al. For the Retrovirus Epidemiology Donor Study. Attitudes toward blood

donation incentives in the United States. Transfusion 2003; 43: 7–16. James V, Hewitt PE, Barbara J. How understanding donor behaviour should shape donor selection. Transfus Med Rev 1999; 13: 49–64. McCullough J. National blood programs in developed countries. Transfusion 1996; 35: 1019–32. McLeod BC, Price TH, Weinstein R (eds). Apheresis: Principles and Practice, 2nd edn. Bethesda: AABB Press, 2003. Newman BH. Donor reactions and injuries from whole blood donation. Transfus Med Rev 1997; 11: 64–75. Nilsson Sojka B, Soika P. The blood-donation experience: perceived physical, psychological and social impact of blood donation on the donor. Vox Sang 2003; 84: 120–8. Petersen LR, Lackritz E, Lewis WF et al. The effectiveness of the confidential unit exclusion option. Transfusion 1994; 34: 865–9. Pindyck J, Waldman A, Zang E, Oleszko W, Lowy M, Bianco C. Measures to decrease the risk of acquired immunodeficiency syndrome transmission by blood transfusion. Transfusion 1985; 25: 3–17. Silvergleid AJ, Leparc GF, Schmidt PJ. Impact of explicit questions about high-risk activities on donor attitudes and donor deferral patterns. Transfusion 1989; 29: 362–4. Soldan K, Sinka K. Evaluation of the de-selection of men who have had sex with men from blood donation in England. Vox Sang 2003; 84: 265–73. Voak D, Caffrey EA, Barbara JAJ, Pollock A, Scott M, Contreras MC. Affordable safety for the blood supply in developed and developing countries. Transfus Med 1998; 8: 73–6. Zuck TF, Cumming PD, Wallace EL. Computer-assisted audiovisual health history self-interviewing. Results of the pilot study of the Hoxworth Quality Donor System. Transfusion 2001; 41: 1469–74.

249

Chapter 22

Blood donation testing and the safety of the blood supply David Wenham and Simon J. Stanworth

This chapter describes the aims and methods of laboratory testing of blood donations. It focuses not only on the range of tests currently employed but also on operational aspects crucial for the safe and efficient application of this process to the thousands of samples received in a blood centre laboratory each day. Testing is dealt with under three headings: • red cell serological testing; • microbiological testing; • operational and quality control issues.

Red cell serological testing In the UK and most countries it is mandatory to test every blood donation for: • ABO blood group; • RhD blood group; • presence of irregular red cell antibodies. In practice, most UK blood services also perform full Rh and Kell typing on all donations. The results from these grouping tests are necessary as baseline information for safe transfusion practice, in order to reduce the risk of premature destruction of the transfused donor red cells in a recipient’s circulation due to immunological incompatibility towards the major red cell antigens. Samples

Tests are carried out on anticoagulated venous blood samples collected at the time of donation. The samples are identified by a unique barcoded identification system, which in most countries is an International Society for Blood Transfusion (ISBT) 250

128 number consistent with the aims of international conformity in blood group labelling and which ensures that each donation has a unique number. The UK Blood Transfusion Service Guidelines (‘Red Book’) provide specifications and guidance on the testing reagents required for blood grouping, as described in the following sections. ABO grouping

Donor red cells are tested with monoclonal anti-A and anti-B antibodies, which are capable of detecting all subgroups of these red cell glycoproteins. A reverse grouping is also performed by testing the donor plasma with A1, A2, B, and A1B reagent red cells. Most blood services make use of automated systems for serology testing where batched samples are divided into separate microtitre plate wells. The test results are read photometrically and the pattern of results obtained from testing donor red cells and donor plasma analysed by microprocessors to establish the ABO blood group result for a particular donation. In the case of repeat donors, such a system also allows the results for ABO groupings to be compared with those generated previously. In the case of first-time donors, the ABO and RhD groups are tested twice and validated only when the two sets of results are in agreement. RhD grouping

RhD is performed by testing donor red cells with two different monoclonal anti-D reagents. These two reagents are selected with the requirement for

Blood donation testing

high sensitivity in order to optimize the detection of weak or partial D-bearing red cells. This would include all the weaker Rh variants, including category DVI. It is felt to be essential that blood services correctly identify all such red cells as RhD positive in view of the highly immunogenic capability of the Rh system. Detection of irregular blood group antibodies

Donor samples are tested to exclude the presence of red cell antibodies that could cause reduced red cell survival or haemolysis when transfused into recipients whose red cells are positive for the relevant antigen(s). This is a screening and not an antibody identification step, and involves the testing of donor plasma with a group O R1R2 K-positive red cells which are also positive for the majority of other red cell antigens thought to be clinically significant. It is essential that all Rh and Kell antibodies above a threshold level of detection should be identified. The control system for UK blood services is set at a level of 0.5 IU anti-D, which is a higher threshold than that defined for hospital blood bank practice (0.1 IU). However, blood services are largely concerned with the detection of high levels of antibodies; weak antibodies will be considerably diluted during processing or transfusion. In contrast, hospital blood bank practice initially requires stringent detection of any antibodies in a potential recipient, irrespective of the level. Blood for neonatal transfusion is tested for irregular antibodies to a higher level of sensitivity over standard testing, in order to further minimize the very small risk of transfusion reactions due to passive transfer of antibodies in this specific group of patients. Most automated blood grouping processes are based on the detection of antibodies on enzymetreated red cells at 30°C. These techniques are known to be less sensitive for antibodies such as anti-Fy, anti-N, anti-M and anti-S. Clearly antibodies directed against blood group antigens not present on the screening cells will not be detected, but these may not be clinically significant and are a lower priority in donation testing for the reasons

mentioned above. However, the screening cells chosen do ensure that Rh and Kell antibodies are detected, as these antibodies have very occasionally been found to cause passive transfusion reactions. High-titre anti-A and anti-B

Standard practice in hospitals is to transfuse group-specific red cells to all recipients. Group O red cells may also be transfused for all transfusions to certain groups of patients, such as neonates and patients requiring urgent transfusion, before their blood group is known. However, it is recognized that some group O donors may have high titres of anti-A and anti-B in their plasma that could cause lysis of A and/or B cells, particularly where large volumes of plasma are transfused, e.g. fresh frozen plasma (FFP) and platelet transfusions, or after exchange transfusion of red cells. In practice, because most red cell packs are stored in optimal additive solutions for preservation, the amount of plasma after dilution ultimately transfused is very small. Plasma containing high-titre haemolysins can be screened in the blood service laboratory by observing the reactions between donor plasma and a diluted sample of reagent A1B red cells, and the products labelled accordingly. Recent refinements to testing for high-titre haemolysins include methods to assess only the more clinically relevant IgG (rather than a combination of IgG and IgM) fraction. There is no standard method of testing for high-titre haemolysins and the acceptable cut-off titre varies greatly with the technique used, thereby requiring local assessment of the procedures used. In practice, an automated system lends itself to universal testing of all donations for hightitre anti-A and anti-B, not just from group O donations. Very occasionally, high-titre anti-A may be found in group B donations (and vice versa). Supplementary testing

Not infrequently, anomalies appear in some of the above test results and will preclude accurate grouping of a donation. For example, it has been 251

Chapter 22

estimated that 1 in 10 000 blood donors have a positive direct antiglobulin test (DAT) at the time of donation, which could interfere with the above assays. Weakly positive DAT may not be detected in the routine grouping test, which is based on a control channel for donor red cells mixed with inert serum. These donations may cause problems in hospital blood banks, since they would appear incompatible after crossmatching by indirect antiglobulin test (IAT). Subsequent donations from these donors will be ‘flagged’, and monitored as the positive DAT may be transient. In many cases, the blood service laboratory has to resort to manual techniques to correctly identify the blood group or antibodies. In general, only antibodies reacting in the IAT are considered to be clinically significant. It is standard practice to establish whether the corresponding blood group antigen is absent from the donor’s red cells. In the case of identified anti-D or anti-c, quantitation of the antibody is performed because in those cases where levels are found to be not significantly raised, the red cells may still be released for transfusion, since during component preparation the antibody-containing plasma may be replaced with optimal additive medium. Phenotyped red blood cells

Many blood centres also undertake a more comprehensive red cell antigen phenotyping service in order to identify donors whose red cells could be used for transfusion to recipients whose plasma is known to contain clinically significant blood group antibodies or to patients at high risk of forming multiple alloantibodies, e.g. sickle cell disease. This may involve screening up to 10–15% of samples from all donations received at the laboratory in a day, and should ensure that most requests for antigen-negative blood from hospitals can be met. In particular, testing for S, s, Fya/b, Jka/b, Kpa and Lua in these red cells is often performed, as well as an additional sickle test to identify the presence of HbS (see later). In selected donations further specific red cell phenotyping may be arranged. Individuals tested for HbS on the basis of their ethnic origin could be screened for the U antigen, which is far more likely 252

to be absent in Afro-Caribbeans than in Caucasians. This facilitates the provision of U-negative blood required for transfusion to those individuals who have developed anti-U, which is a clinically relevant antibody. As mentioned, testing is performed to identify donations positive for HbS, in order to ensure that this blood is not transfused either to adults with sickle cell disease or to neonates during exchange transfusions. The need for a sickle cell screening test in a blood service will depend on the prevalence within the donor population. An additional consideration is the need to provide counselling support to inform donors found to be carriers of HbS. Of recent interest, it has been found that sickle-trait (HbAS) blood significantly interferes with the function of the filters currently used for leucocyte depletion. Such ‘failed’ donations would be discarded, but the pattern of red cell antigens in these individuals could be unique and very useful as a transfusion resource.

Microbiological testing of blood donations A wide range of infectious agents have been documented as transmissible by blood transfusion and these are described in Chapter 19. Donor selection and the use of established guidelines to defer individuals at risk of infection by these agents are the important first steps aimed at reducing the risk of collecting blood donations with the potential to transmit infection. This is particularly important with respect to the collection of blood from donors in the ‘window period’ of an infection who may be asymptomatic and infectious, but without the viral load, with regard to hepatitis B surface antigen (HBsAg), or level of antibodies, with regard to hepatitis C virus (HCV) or human immunodeficiency virus (HIV), to be detected by serological screening tests. The laboratory screening tests form the core of the process to identify infected blood components prior to transfusion. From the perspective of the transfusion recipient, sensitivity is the most important criteria for a laboratory screening test, i.e. it will accurately identify most, if not all, infected donations. Maximal sensitivity in a test has to be

Blood donation testing

balanced against specificity. However, many of the newer techniques and kits currently used in blood centres show remarkably high levels of both specificity and sensitivity. Indeed, risks of viral transmission by blood products remain extremely low (see Chapter 19). All positive screening tests require further confirmatory or reference testing to establish whether the result represents a genuine positive case. It must be appreciated that in contrast to blood grouping, in which every sample produces a grouping result, most reports generated in microbiological screening are negative, and this has implications for quality control. On the other hand, blood centres screen large numbers of samples, which means that even though the screen tests exhibit high sensitivity and specificity, there will be significant numbers of samples from donors being identified as reactive on initial screening but where subsequent reference tests are found to be negative (i.e. false positives). Donor counselling

Screening results found to be initially and then repeatedly reactive on further testing, carried out because some initial reactive screen results may reflect assay or instrument-related problems, will be followed up by reference laboratory testing. The number of these repeat-reactive samples far exceeds the numbers of confirmed positives (see later section about results from screen tests). However, donors need to be aware that their blood has been found to be repeatedly reactive in one of the microbiological screen tests, although the likelihood is that the reactive result represents false positivity. This can be a complicated issue to explain and should only be undertaken when it has been confirmed that the reactive result represents false positivity. Blood centres must also develop strategies for dealing with donors found to be confirmed positive (sometimes unexpectedly) and for initiating the counselling and involvement of specialist treatment centres. Principles of the screen test methodology

Most viral serology tests are based on immunoas-

say principles, using enzyme or chemiluminescent techniques of detection. Donor plasma is mixed first with captured antigen of the virus and then with antigen/antihuman globulin specific for the presence of antibody to the infectious agent. Detection may involve a conjugate linked to an enzyme, usually peroxidase, which can be detected photometrically after addition of substrate which produces colour, or by chemiluminescence, in which the optical measuring device detects photons emitted by the chemiluminescent reaction. Other changes to the methodology include improved ways of presenting the captured antigen and the use of antibodies to detect both an IgG and IgM immune seroconversion response to the virus. These changes aim to enhance the detection of the earliest possible immune responses to infection. Additional strategies aimed at reducing the window period of detection entail the development of even more sensitive techniques for detection of specific antibodies and the development of screen assays to look for the viral antigen itself or the nucleic acid. In the USA, HIV antigen (p24) testing is now a standard part of blood donor testing and in many countries a ‘combination’ assay that can detect both HIV antibody and antigen has been introduced. Many countries including the UK have now developed programmes to use nucleic acid amplification technology (NAT testing) on pooled samples of donor plasma to detect viral-associated DNA or RNA (see below). Screening tests

Table 22.1 lists the screening tests used in transfusion microbiology. These tests have been divided into those mandatory in the UK, those mandatory in other countries and those used on selected groups of donors, as discussed later. Screening tests for specific antibodies, such as anti-hepatitis B and anti-varicella-zoster, are performed by some blood services in order to identify donors whose plasma may be collected for issue as high-titre specific immunoglobulin. The final decision to implement a particular screening test in a country will depend on a number of factors, including the prevalence of the infectious disease in the donor population. For 253

Chapter 22 Table 22.1 Screening tests in transfusion microbiology.

Test

Microorganisms

HBsAg* Anti-HIV-1, -2 and -0* (in some countries, only HIV-1 poses a threat) Anti-syphilis* Anti-HTLV*

HBV HIV-1, HIV-2 and HIV-0 Treponema pallidum HTLV-I (and cross-reactivity for HTLV-II)

Mandatory in some countries

Anti-HCV* Anti-HBc Alanine aminotransferase HIV antigen (p24) NAT testing

HCV Surrogate tests for non-A, non-B hepatitis.Anti-HBc will detect some HBsAg-negative, HBV-infected donors HIV HCV; other agents to follow

Discretionary

Anti-CMV Anti-malaria Anti-Chagas’ disease Specific antibodies (hepatitis B, varicella-zoster)

CMV Plasmodium species Trypanosoma cruzi High-titre immunoglobulins

Generally mandatory (where affordable)

* Mandatory screening tests for all blood donations in the UK. CMV, cytomegalovirus; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus; HTLV, human T-cell leukaemia/lymphoma virus; NAT, nucleic acid amplification technology.

example, following extensive discussion, testing for anti-human T-cell leukaemia/lymphoma virus (HTLV) was recently introduced as an additional centralized test in the UK, although the prevalence of this infection is considered low in comparison with other countries. Testing for hepatitis B does not involve detection of antibody but direct detection of HBsAg. Some of the issues concerning the limitations of testing for HBsAg alone are discussed in the Chapter 19. Laboratory screening for infection by Treponema pallidum (syphilis) remains mandatory in many countries including the UK. Actual risks of transmission are low for many reasons, and treponemes survive for only short periods at the low temperatures used for red cell storage. Nevertheless, the incidence of syphilis is rising in different areas of the world and it can be argued that syphilis screening of donated blood is of value as a lifestyle indicator, since individuals exposed to syphilis may also have been exposed to other sexually transmitted diseases. Immunological responses to cytomegalovirus (CMV) infection are detected by the presence of

254

CMV antibodies in donor blood. As CMV is a latent infection, the presence of antibody indicates not only a previous but also a current and potentially infectious state. Screening is undertaken both on random donations and selected previously negative donations. Being a cell-associated virus, it is to be expected that leucocyte depletion should significantly reduce the risk of CMV transmission by blood components. However, as for all procedures and tests, there is a ‘failure’ rate (albeit very low) and it remains a controversial issue whether leucocyte-depleted blood components provide equal or lesser risk of CMV transmission compared with antibody testing. As a consequence of the measures taken to minimize the potential risk of transfusion-transmitted variant Creutzfeldt–Jakob disease (see Chapter 20), plasma for younger-aged recipients will be imported from the USA for use as products such as FFP and cryoprecipitate. Such plasma will be subjected to the same microbiological screening tests as for other blood products derived from UKsourced plasma and will be virally inactivated using a methylene blue technique

Blood donation testing Table 22.2 Donation testing data (England), November 2002 to April 2003. (From NBA/PHLS CDSC Infection Surveillance

Report.)

Assay HBsAg Anti-HCV Anti-HIV Treponemal antibodies (syphilis) Any marker

Number of donations tested

Initially reactive donations Number

%

1 171 607 1 171 412 1 174 877 1 168 216

1359 853 1477 3236

0.12 0.07 0.13 0.28

4 686 112

6925

0.5915

Repeatedly reactive donations

Confirmed positive donations

Number

Number

380 488 684 22 1574

% 0.03 0.04 0.06 0.00 0.1342

Rate per 100 000

%

32 47 22 26

0.0027 0.004 0.0019 0.0022

127

0.0108

2.731 4.012 1.873 2.226 11

HBsAg, hepatitis B surface antigen; HCV, hepatitis C virus. Table 22.3 Confirmed positive rates

Scotland

(from 1/10/95, except Scotland from 1/4/96 and HTLV from 1/8/02 to 30/4/03). (From NBA/PHLS CDSC Infection Surveillance Report.) HBsAg HCV HIV Syphilis HTLV

Rest of UK plus Ireland

New donors

Repeat donors

New donors

Repeat donors

1 : 5230 1 : 1236 1 : 30 541 1 : 9295 XX*

1 : 64 553 1 : 20 224 1 : 120 125 1 : 50 095 XX*

1 : 3451 1 : 1946 1 : 23 651 1 : 6698 1 : 68 361

1 : 126 888 1 : 65 625 1 : 231 825 1 : 86 159 1 : 44 814

* 0 positives in 256 555 tests. HBsAg, hepatitis B surface antigen; HCV, hepatitis C virus; HTLV, human T-cell leukaemia/lymphoma virus.

Results from screen tests in the UK

Tables 22.2 and 22.3 show the results obtained by screening and confirmatory testing on donations in the UK. In the case of screening tests (Table 22.2), the results were taken over a period of 1 month in 1999. By comparison, the data on confirmed cases were derived over 1 year (Table 22.3). The tables highlight two points. 1 Confirmed cases are a small fraction of the numbers of initial reactive results. 2 The frequency of confirmed cases shows wide variation between new and repeat donors. This forms the basis for preferentially processing donations from repeat and not new donors for use in selected at-risk groups of recipients such as infants and neonates in order to further minimize the risk

of transmitted infection. However, the data from these tables also indicate the need to have systems in place to continuously monitor the results for the rates of positive cases. For example, there is concern that the numbers of confirmed cases of HIV in donors have risen over the last year. Nucleic acid amplification technology

The application of polymerase chain reaction (PCR) methodology has the theoretical potential of identifying viral agents in the blood at the earliest possible stages of infection, since the sensitivity for detection is significantly higher than that obtained even by antigen assays. The window period of infection is longest for HCV (75 days compared with 21 days for HIV), and this has been

255

Chapter 22

one factor in driving the recent introduction of NAT testing for HCV in the UK. All plasma products, including FFP, must now be derived from pools of plasma found negative for HCV RNA before issue for clinical use. NAT testing for HIV remains under consideration. A number of operational, technical and economic factors have raised concerns about the actual effectiveness of NAT testing, despite the potential benefits. Some of these issues include the following. • Staff training, in view of the general complexity of this form of testing. • The need for scrupulous methods to avoid crosscontamination. • Appreciation of the time constraints inherent in obtaining NAT results and which crucially may be after the issue of blood components (currently often only 1 day). • Cost analyses, which take into consideration the implications of patenting of the PCR methodology and the operational requirements for traceability of reactive results to the single donation level. These issues are linked to the approach taken by most blood centres undertaking NAT testing, which is now based on testing pooled samples of approximately 50 donations. Such a system must allow traceability, but could also affect the potential sensitivity of the assay for a number of reasons including dilution and the presence of inhibitors within some donor serum samples. Risk estimates had previously suggested that around 1 in 250 000 donations will be derived from HCV-seroconverting donors in the window period. However, recent results obtained after the introduction of NAT testing in the UK indicate that the proportion of samples that are NAT positive but antibody negative may be considerably lower, of the order of 1 in 1 million. This reduction in apparent risk following the introduction of a new test is a not uncommon feature of microbiological testing, since the parameters used to develop a level of risk in the first instance tend to be based on less accurate initial input data. In many respects, if we aim towards a system of zero risk, screen testing for microbiological agents should be based around both NAT testing and antibody testing. However, 256

the details of the benefits and the cost-effectiveness of such an approach is beyond the scope of this chapter (but see Further reading). A final consideration affecting the whole role of NAT testing for transfusion microbiology concerns the increasing application of validated specific viral inactivation steps within component processing. Effective inactivation would have the advantage of destroying transmissible viral agents for which there is currently no suitable test or those which are presently uncharacterized. In contrast, piecemeal addition of NAT testing for new agents would clearly require prior identification and characterization of the virus.

Quality framework and operational issues Figure 22.1 shows a framework for maintaining quality in donation testing. Ultimately the microbiological and blood group safety of the blood supply depends on the input and interaction of the various quality and operational factors shown. The quality system needs to meet the requirements of the Medicines and Health Care Regulatory Authority (MHRA) in the UK. Testing must only be performed by staff trained in approved standard operating procedures (SOPs). Document control systems must be in place to ensure only current procedures are used and any changes documented and approved. Any errors that occur in laboratory procedures must be logged using a quality incident report (QIR) system which requires corrective and preventative action to be taken. Fully automated sample and test processing is now standard practice in the UK blood centres for both microbiology and blood grouping, and the similarity of testing systems and requirements for quality and operational control has led to the integration of these previously separate departments. Systems in use may vary from modular to fully integrated, but all have specific sample identification and a system for tracking the sample throughout the testing procedure in order to provide a complete audit trail and documented evidence of testing to statutory requirements. Operational advantages of automation include:

Blood donation testing

Quality requirements Quality system

Testing

Training, staff competency assessment and review

Mandatory tests Sample ID

Service/maintenance log

Operational requirements automated sample processing In process controls Sample addition monitoring

Plate ID Calibration log

Reagent addition monitoring Plate result Sample result

SOPs and staff training records

Electronic transfer

Quality incident reporting (QIR) Document control system for changes

Integrated computer system controlling Quarantine

Batch validation

Donor/donation history Statistical process control (SPC) and performance review

Discard of hazardous material

Electronic events-logging of critical process steps Data management Electronic interpretation and decision making Go–no-go national standards Validating repeat testing procedures Document archive Sample archive

External audit, NEQAS Issue Integrated release

Available for issue Fig. 22.1 A quality framework for donation testing.

• better reproducibility; • lower staff costs; and • the ability to cope with high volume. Test result interpretation and decision-making must be performed electronically by a data management system that can transfer individual results to a main computer. This integrated computer system will provide facilities that cover the whole blood donation process, providing essential controls and operational information to ensure safe working practices. Test results can be compared with previous laboratory findings and medical history to ensure that any hazardous material cannot be issued to hospitals. Automated equipment must be validated before use and serviced and maintained to manufacturers’ recommendations. Logs of servicing and regular

calibration checks must be kept to provide evidence of satisfactory operational performance. Several critical steps are involved in the performance of microbiological assays and blood grouping tests. These need to monitored, verified and logged by the testing processors. Microbiology test kit suppliers are now required to provide assays that have sample addition monitors and coloured reagents; these can be measured photometrically on the test processor and provide evidence in an event log that the critical steps have been performed and that the tests are valid. Appropriate quality control samples are essential. There are nationally agreed standards for all mandatory microbiology assays, and these must be performed with each batch of tests as ‘go–no-go’ controls, providing documented evidence of satisfactory 257

Chapter 22

test-run sensitivity. These standards are supplied by the National Institute of Biological Standards and Controls (NIBSC), are low-level positive controls and are used to monitor and validate day-today performance. They are also used to assess new batches of test kits prior to use within the laboratory, although acceptance testing of kits has now also been streamlined and centralized in the UK. Records of test results must be maintained in a readily accessible archive, usually in both paper and electronic format. In the UK, a fully traceable archive of all samples tested must be maintained for a minimum of 3 years. The levels of process control now employed by the blood centre donation testing laboratories give a very high level of confidence that the test result for a donation is both valid and correct and that any potentially hazardous material will be discarded. Ultimately, the critical measure of the quality of microbiological screening must be the rate of post-transfusion infection. It is therefore very important that good links are maintained with hospitals so that possible post-transfusion infections are reported and fully investigated.

258

Further reading Brennan M, DeSilva M, Barbara JAJ. Safety of blood components. In: Pamphilon DH, ed. Modern Transfusion Medicine. Boca Raton, FL: CRC Press, 1995. Chamberland M, Khabbaz RF. Emerging issues in blood safety. Infect Dis Clin North Am 1998; 12: 217–29. Courouce AM, Pillonel J, Lemaire JM, Saura C. HTLV testing in blood transfusion. Vox Sang 1998; 74 (Suppl. 2): 165–9. Dow B. Microbiology confirmatory tests for blood donors. Blood Rev 1999; 13: 91–104. Flanagan P. Genomic screening of blood donations: the new dawn arrives. Vox Sang 1999; 76: 135–7. Guidelines for the UK Blood Transfusion Services, 6th edn. London: HMSO, 2002. NBA/PHLS CDSC Infection Surveillance Reports. London: Colindale PHLS. Pamphilon DH, Rider JR, Barbara JAJ, Williamson LM. Prevention of transfusion-transmitted cytomegalovirus infection. Transfus Med 1999; 9: 115–23. Van den Burg PJ, Vrielink H, Reesink HW. Donor selection: the exclusion of high risk donors? Vox Sang 1998; 74 (Suppl. 2): 499–502.

Chapter 23

Production and storage of blood components Lorna M. Williamson and Rebecca Cardigan

Whole blood and its processing to components Guidelines from the UK, Council of Europe and American Association of Blood Banks define a blood donation as 450–500 mL ± 10% of blood collected into citrate anticoagulant also containing phosphate and dextrose. There are no absolute indications for transfusion of whole blood, and the vast majority of blood units collected in the UK are processed to components: red cell and platelet concentrates, and plasma. Such plasma is suitable for either fractionation to plasma derivatives, or for freezing as whole fresh frozen plasma (FFP). Component production from whole blood consists of centrifugation to separate plasma and cells of different density, followed by manual or automated transfer of components from the primary collection pack to transfer packs. Collection and transfer packs are manufactured as a single closed unit to maintain sterility. Whole-blood donations from which platelets are to be harvested must be held and processed at 20–24°C but, for other donations, preprocessing storage and centrifugation can be at either 20 or 4°C. Some countries hold all blood overnight at 20°C prior to component production. This yields components of high quality, but is not yet permitted in either the UK or the USA because of concerns regarding bacterial proliferation. There remains a small but finite risk of transmission of viruses via single-unit or small-pool blood components. Unlike fractionated plasma products, it has not been possible to subject components to a virus inactivation step. Techniques for doing so are now available for FFP and platelets, and under

development for red cells. These are discussed in the appropriate sections below. Collection of components by apheresis

Apheresis involves separation of the blood into components during collection on specially designed equipment, the usual process being to collect plasma and platelets and return the red cells to the donor. Because there is no loss of iron, apheresis donors can donate monthly, and plateletpheresis permits collection of 1–3 adult doses per procedure, depending on the platelet count of the donor. Apheresis has been regarded as a more risky procedure than whole-blood donation, and tended to be undertaken only in donor clinics with trained nursing and medical staff available. However, apheresis equipment has developed into small portable machines drawing only a low extracorporeal volume so that they can be used safely on mobile sessions. Such equipment can be programmed flexibly to collect red cells and platelets and/or plasma, with double-dose red cells being another option. Thus the distinction between whole-blood donation and apheresis is becoming less, and it is likely that ‘near donor processing’ will expand in future years. Quality monitoring

Specifications for the key parameters of each component type are generally set out in a national document such as the Guidelines for the Blood Transfusion Services in the United Kingdom. For other variables, there is freedom to set local specifications, which should be defined in a compendium by the manufacturing blood service. 259

Chapter 23

Usually 1% of components produced are subjected to quality monitoring, although for those which are produced only occasionally a set number each month must be tested. For key parameters, 75% of units tested must fall within the specification limits, although a different approach is applied to monitoring of leucocyte removal.

Leucocyte depletion of blood components Many developed countries (although notably not the USA) have implemented universal leucocyte depletion of components. In the UK and Ireland, the risk that variant Creutzfeldt–Jakob disease (vCJD) might be transmissible by blood, and in particular by leucocytes, was the major factor in this decision. In other countries, additional benefits, such as a reduction in immune-related complications and removal of cell-associated viruses, were considered equally important. Because leucocyte depletion fundamentally alters all aspects of component production and quality monitoring, it is considered in some detail here. The clinical benefits of leucocyte-depleted components are considered below and in Chapter 18. Specifications for leucocyte-depleted blood components

A specification for leucocyte-depleted components is based on the premise that it is currently not feasible to perform a leucocyte count on every unit if these are manufactured on a large scale. Monitoring compliance with the specification depends on random sampling and statistical process monitoring, which predicts the percentage of components which will have leucocyte counts below a stated level. This means that the specification must have three elements: • the maximum permitted number of leucocytes in each component; • the percentage of components produced that must achieve this level of leucocyte depletion; and • the degree of statistical confidence attached to this prediction. Specifications set by the UK and Council of Europe appear different, but are in fact broadly 260

similar (Table 23.1). Some of the practical aspects of preparing leucocyte-depleted blood components include the following. 1 Components manufactured from whole-blood donations are leucocyte depleted by passage through purpose-designed filters. Blood passes through a ‘prefilter’ designed to remove small fibrin clots and cellular aggregates (Fig. 23.1), followed by passage through an irregular meshwork of fibres, usually of polyester or polyurethane (Fig. 23.2). Mechanisms of filtration are thought to include platelet activation causing secondary adhesion of granulocytes and monocytes, direct adhesion, and physical trapping of the more rigid lymphoid cells in the fibre mesh. 2 Transportation of collected blood to the processing centre allows a period of contact between leucocytes and any bacteria that have entered the pack from the donor’s skin. Phagocytosis of bacteria, followed by subsequent removal of leucocytes and ingested bacteria by filtration, may lead to a reduced risk of bacterial transmission. However, systems for collection of platelets by apheresis which perform leucocyte depletion of platelets ‘on the machine’ also appear to provide a bacterially safe component. 3 Leucocyte removal should be performed while the cells are still intact. Whole blood or red cells should ideally be filtered either on the day of collection (day 0) or day 1, and always by the end of day 2. Platelets produced by the buffy coat method are filtered during processing on day 1. Storage conditions prior to filtration are determined by the requirements of component processing and/or the conditions under which the filter can operate.

Table 23.1 The three elements of user specifications for

leucocyte-depleted components.

White blood cell level Percentage components in which achieved Statistical confidence that it has been achieved

UK

Council of Europe

5 ¥ 106 99%

1 ¥ 106 90%

95%

Not stated

Production and storage of components

Inlet port

Outlet port

L L L L L L L L L L L L L L L L L L L

Leucocyte removal Aggregate removal Gel removal

Leucocyte removal

Prefilter

L L L L L L L L L L L L L L L L L L

A A A A A A A A A A A A A A A A A A

Inlet port

G G G G G G G G G G G G G G G G G G

Outlet port Fig. 23.1 Cross-section of a filter designed for leucocyte

depletion of whole blood, red cells or platelets showing areas of multiple laminates of filter media within the filter housing. (Courtesy of Pall Europe.)

4 Most filters designed to leucocyte deplete whole blood or red cells also remove more than 95% of platelets, so whole-blood filtration cannot be applied to units from which platelet manufacture is required. Depending on whether platelets are to be produced, blood is processed by one of two production strategies (Fig. 23.3). (a) In process A (see Fig. 23.3a), whole blood is passed through a leucocyte-depleting filter, which also removes platelets, and subjected to a single ‘hard’ centrifugation step, followed by expression

of the plasma to yield red cells and plasma which are both leucocyte depleted. Additive solution may be added to the red cells at this stage. (b) In process B (see Fig. 23.3b), termed ‘bottomand-top’ (BAT) processing, a ‘hard’ centrifugation step is followed by expression of the plasma from the top of the pack, and removal of the red cells through the bottom of the primary pack into a transfer pack containing an additive solution. The buffy coat remains in the primary pack and, if separation is completed within 8 h of blood collection, this may be used as start material for platelet production (see below). The red cells and plasma can then be passed separately through leucocytedepleting filters. Filters that permit platelets to pass through are now becoming available, with the potential to leucocyte deplete all donations through a single filter prior to processing. However, these are not yet in general use. 5 The Food and Drug Administration (FDA) in the USA has issued a guideline on implementation of leucocyte depletion. 6 The same processing options apply for nonleucocyte-depleted components, except that the filters are omitted. Measurement and quality monitoring of leucocytedepleted components

• Residual leucocytes are too few to be counted on standard haematology analysers. One technique suitable for leucocyte counting on a large scale is flow cytometry, which depends on detection of nucleic acid-bound fluorescent dye in leucocytes. Microscopic counting using large-volume counting chambers (Nageotte) is not suitable for largescale use, but can be used for research purposes. Quality assurance schemes for leucocyte counting should be developed as an important part of the leucocyte depletion process. • It would be desirable to measure the subtypes of leucocytes in leucocyte-depleted components, as these may determine the risk of viral and immune complications. Studies using either flow cytometry or quantitation of mRNA encoding markers of leucocyte subsets suggest that all leucocyte depletion methods effectively remove 3–4log10 of all 261

Chapter 23

Fig. 23.2 Scanning electron micrograph of a filter designed for leucocyte depletion of red cell concentrates showing adhesion of platelets and leucocytes to the fibres. (Reproduced with permission of Pall Europe.)

granulocytes, monocytes and B and T lymphocytes. In filtered red cells, most residual leucocytes appear to be granulocytes. • Systems from different manufacturers should firstly be subjected to formal assessment under the operating conditions of the blood centre. The objective is to create an ‘approved list’ of systems showing high capability and consistency of performance of filter and pack. Approved systems also require some local validation at each manufacturing site. • Once a system is implemented, data on its ongoing performance should be collated and monitored. Systems scoring well on capability and consistency are suitable for statistical process monitoring. This involves testing 1–5% of each component type produced, and plotting the results on a statistical process monitoring chart (Fig. 23.4). Systems with either poor overall capability 262

or outliers giving a bimodal distribution require 100% of components to be counted and are therefore best avoided. • A system for implementing quality monitoring of leucocyte depletion is provided in guidelines published by the Biomedical Excellence for Safer Transfusion (BEST) group of the International Society for Blood Transfusion. Problems with leucocyte depletion

• One consistent feature of leucocyte depletion is retention of up to 15% of the component volume in the filter, with cellular loss in proportion. This can be a particular problem in red cells that are filtered after prior removal of the buffy coat. This product is best avoided for transfusion-dependent patients, in whom transfusion freqency might increase slightly; these patients are best provided

Production and storage of components

(a) Whole blood filtration Whole blood

L/D RCC (also minus platelets)

L/D FFP

(b) Component filtration Whole blood

Buffy coat

Red cells

Plasma

L/D RCC (also minus platelets)

L/D FFP

Pool 4

L/D platelet concentrate

Fig. 23.3 (a) Production of red cell concentrates and plasma

from blood donations. (b) Production of red cell concentrates (RCC), platelet concentrates and plasma from blood donations. L/D, leucocyte depleted.

with red cells separated after whole-blood filtration. • Whole blood or red cells from donors with haemoglobin sickle trait may be difficult to filter, with blockage seen in approximately 50% of units, and possibly a higher than usual rate of failure to meet the leucocyte depletion specification. • There have been remarkably few reports in the world literature of patient complications of leucocyte-depleted components. Reports have appeared of hypotensive reactions in patients receiving platelets or red cells through bedside filters, possibly due to generation of bradykinin on the filter surface, particularly those carrying a negative surface charge. Some reported patients were on medication with antihypertensive drugs of the angiotensin-converting enzyme (ACE) inhibitor type. Such patients may have been more vulnerable since ACE is involved in the bradykinin degradation pathway. The FDA advises that this complication is unlikely to be seen with prestorage leucocyte depletion, because bradykinin is rapidly degraded in normal plasma. • Reports of an acute syndrome of eye pain and visual impairment affecting over 100 recipients of filtered red cells were investigated by the FDA. The incidents related only to one manufacturer’s filters, and the precise cause was not established.

Fig. 23.4 Statistical process

log10 WBC

X-bar

log10 WBC

monitoring charts for leucocytedepleted red cells in SAG-M using log10 white blood cell (WBC)/unit. The upper panel is the X-bar chart showing control line (cl), upper control limit (ucl), and upper specification limit (USL). Here the USL is 6.7, the log10 of 5 ¥ 106. Each point is a mean of five values from units selected according to a systematic sampling protocol. Every point above the ucl is an indication that the process is not in control, and requires increased quality monitoring. The lower panel is the range chart, which demonstrates the difference between the highest and lowest values within each group of five. Again, any points above the ucl are indications for increased quality monitoring.

USL 6.5 ucl 6.0 cl 5.5 5.0 2/11/98 2/11/98 3/11/98 3/11/98 4/11/98 4/11/98 5/11/98 5/11/98 6/11/98 Date Range 2 ucl 1.5 1.0 cl 0.5 0 2/11/98 2/11/98 3/11/98 3/11/98 4/11/98 4/11/98 5/11/98 5/11/98 6/11/98 Date

cl: 5.82056 ucl: 6.27716 LCL: None

263

Chapter 23

However, the episode illustrates the possibility of unusual types of reaction to leucocyte-depleted components. National haemovigilance schemes are now being established in many countries, and have an important role in collating and disseminating reports of rare categories of transfusion events. Removal of cell-associated viruses

Viruses associated with different leucocyte subtypes include cytomegalovirus (CMV), mainly in monocytes, and other DNA herpesviruses such as Epstein–Barr virus and human herpesvirus 8 (in B cells), and T-cell viruses such as human T-cell leukaemia/lymphoma virus (HTLV) I and II. Most studies of prestorage leucocyte depletion have demonstrated its efficacy in preventing transfusion-transmitted CMV, although a recent study has suggested that if enough patients are studied, a small increase in risk might emerge (Table 23.2). Bedside filtration appears to be unreliable in this regard. The Council of Europe, the American Association of Blood Banks and the British Committee for Standards in Haematology all consider that components leucocyte depleted at source are equivalent in safety to those tested as CMV seronegative.

However, this view has not been endorsed by either the FDA or the UK guidelines for the transfusion services. A recent consensus conference in Canada concluded that it would be premature to discontinue CMV testing, which continues to be performed by most transfusion services. Information on removal of other viruses by leucocyte depletion is limited, although one study of HTLV-I removal showed incomplete clearance of virus from some asymptomatic carriers by leucocyte depletion. Variant Creutzfeldt–Jakob disease

This disease is believed to be caused by the same prion agent responsible for bovine spongiform encephalopathy (BSE) in cattle (see Chapter 20). A lookback study has revealed two possible cases attributable to blood transfusion. Unlike classical CJD, vCJD shows affinity for lymphoid tissue, with abnormal prion (PrPSc) demonstrated in the tonsils of affected individuals. Normal prion (PrPc) has been demonstrated on CD34+ progenitor cells, red cells, lymphocytes and monocytes. It is also found on platelets, from which it is released on activation with physiological agonists, but its

Table 23.2 Transmission of cytomegalovirus (CMV) by standard and leucocyte-depleted components. (Modified from

Pamphilon et al. 1999 with permission.) Incidence of CMV infection (%) Study

Patient group (number of patients)

Standard components

Leucocyte-depleted components

1 2 3 4 5 6 7 8 9 10 11

BMT (autologous and allogeneic) (29) Leukaemia (54) Leukaemia/lymphoma (59) BMT (autologous) (24) BMT (allogeneic) (23) BMT (autologous) (37) BMT (autologous and allogeneic) (60) BMT (allogeneic) (62) BMT (autologous and allogeneic) (502) BMT (children) (93) BMT (allogeneic and autologous) (807)

ND 12–22 ND 23 ND ND ND ND 0.8 (CMV seronegative) ND 1.7†

0 0 0 0 0 0 0 0 1–2* 0 4

* Study 9 employed bedside filtration, so it is uncertain whether components fulfilled the criteria for leucocyte depleted. † Study 11 compared filtered and apheresis platelets, which were transfused only if seronegative components were not available. BMT, bone marrow transplantation; ND, not determined.

264

Production and storage of components

functional role remains unclear. PrPc is also found in plasma; its source may be either platelets or endothelial cells. Experimental transmission studies in rodents and sheep suggest that in these species infectivity is present in blood, with both plasma and buffy coat able to transmit the infection. Thus the possibility exists for transfusion-transmitted vCJD via leucocytes within blood components. A large study of leucocyte depletion processes has demonstrated consistent removal of normal prion protein, but for technical reasons the degree of removal of abnormal prion has not yet been established. Immunological effects of transfusion which may be modified by leucocyte depletion

Adverse immunological effects attributed to leucocytes include human leucocyte antigen (HLA) alloimmunization, transfusion-associated graftversus-host disease and immunosuppression, which in turn may lead to increased postoperative sepsis and tumour recurrence. However, universal leucocyte depletion may remove some of the beneficial effects of transfusion-induced immunomodulation, such as improved survival of transplanted kidneys and suppression of Crohn’s disease. These are discussed in detail in Chapter 18.

Red cell components (Table 23.3 for specifications) Leucocyte-depleted red cell concentrates, with or without additive solution

The vast majority of red cell units are leucocyte

depleted during processing, most of the plasma is removed, and approximately 100 mL of an additive solution containing saline, adenine, glucose and mannitol (SAG-M) is added to maintain red cell viability. The most important changes that occur during storage are loss of intracellular potassium, and a reduction in red cell recovery following transfusion. Red cell concentrates in additive solution have a 35–42-day shelf-life (depending on the storage solution), at a controlled temperature of 2–6°C. Red cells stored only in plasma have a 28-day shelf-life. To minimize the possibility of bacterial proliferation and to maintain viability, red cells should be removed from refrigeration as little as possible. Washed red cells

Red cells washed in saline are prescribed only for patients with uncontrollable febrile or anaphylactic reactions to leucocyte-depleted red cells or for patients with IgA deficiency. The objective of washing is to remove as much plasma protein as possible, as such reactions can be due to antibodies to plasma proteins. The current specification is that less than 0.5 g protein/unit should remain. If the washing is performed in an ‘open’ system, the shelf-life is shortened to 24 h. However, at least one closed system for cell washing is now available, which allows red cells to be stored after washing, albeit with a shortened shelf-life. Washed red cells are no longer considered necessary for patients with paroxysmal nocturnal haemoglobinuria, for whom leucocyte-depleted red cells in SAG-M are an acceptable alternative.

Table 23.3 UK specifications for red

cell and platelet concentrates. Red cells Red cells in additive solution

Apheresis platelets Platelets derived from pooled buffy coats

Volume (mL)

Hb (g/unit)

WBC (per unit)

220–320 220–340

40 40

< 5 ¥ 106 < 5 ¥ 106

Platelet count (¥ 109)

pH at end of shelf-life

WBC

> 240/unit > 240/unit

6.4–7.4 6.4–7.4

< 5 ¥ 106/unit < 5 ¥ 106/unit

265

Chapter 23

Frozen and washed red cells

This process is used only for red cells from donors with rare phenotypes or from occasional patients with multiple red cell alloantibodies for whom provision of compatible donor blood is extremely difficult. Red cells may be stored for years in the vapour phase of liquid nitrogen at –80°C provided a cryoprotectant (usually glycerol) is added prior to freezing. Such red cells must be washed several times in saline after thawing, which renders the final product leucocyte depleted. The same considerations regarding shelf-life apply to this product as to red cells washed without prior freezing. Pathogen inactivation in red cells

Red cells present a particular challenge for pathogen inactivation. Photochemical methods suitable for platelets and plasma cannot be applied to red cells because of the high degree of light absorption by haemoglobin. A number of compounds that do not depend on light activation are in development or in clinical trial, but none is yet available for routine use.

Platelet concentrates Sources of platelets for transfusion

Platelets may be produced either from wholeblood donations or by apheresis, in which platelets with or without plasma are collected and the red cells returned to the donor. Specifications for platelet yield and residual leucocyte count are similar for the two methods (see Table 23.3). Apart from exposing the patient to fewer donors and the possibility of HLA/human platelet antigen (HPA) matching with the patient, apheresis platelets are not intrinsically of higher quality. Platelet production from whole blood

Platelet production from whole blood may be carried out either from pooled buffy coats generated by BAT processing (see Fig. 23.3) or from platelet-rich plasma (PRP) as an intermediate step. 266

Buffy coat-derived platelets have long been favoured in Europe, and are now standard in the UK, while the PRP method is standard in North America. Leucocyte depletion by filtration may be routinely incorporated into either process. An adult therapeutic dose of platelets (2.5–3 ¥ 1011) can be manufactured from buffy coats, or by the PRP method, from four to six whole-blood donations, whereas the same dose can be harvested from an apheresis donor in an hour. Platelet production by apheresis

Systems are now available to collect one, two or even three adult doses of platelets during one collection procedure. This may require some preselection of donors according to platelet count and haematocrit. In the USA, donors were for a time administered thrombopoietin to boost platelet yields. This is no longer permitted, and has never been practised in the UK. Platelets are collected into citrate supplemented with adenosine and dextrose. Leucocyte depletion is performed either on the apheresis machine by physical separation, or by filtration soon after collection. Platelet storage

• After production and resting, platelet concentrates are stored in plasma in incubators set at 20–24°C for up to 5 days. Platelet concentrates should never be placed in the refrigerator. • Platelets must be agitated during storage, either in the horizontal plane or by tumbling rotation. Storage packs for platelet concentrates are manufactured from plastic designed to maximize gaseous exchange. • During storage, platelets undergo a fall in pH due to accumulation of lactate, express increased surface expression of activation markers such as Pselectin (CD62p), and change shape from discoid to round. Many different laboratory assays have been advocated to monitor development of this socalled ‘platelet storage lesion’ but few have been demonstrated to correlate with in vivo survival (Table 23.4). pH remains the only quantitative

Production and storage of components Table 23.4 International Society of Blood Transfusion BEST

(Biomedical Excellence for Safer Transfusion) proposals for platelet quality assurance. Routine quality assurance pH at day 5 Platelet count White cell count Volume of concentrate Swirling Evaluation of new component As routine plus: Surrogate for viability Morphology score Hypotonic shock Shape change ATP Platelet activation Surface P-selectin b-Thromboglobulin release Platelet lysis Supernate lactate dehydrogenase In vitro function Aggregation to pairs of agonists Metabolic activity PO2 and PCO2 Lactate production

change which must be monitored routinely, and must be between 6.4 and 7.4 at outdate. Visual inspection to look for the ‘swirling’ effect of discoid platelets has been recommended, but this is highly subjective and changes only when the platelets have been grossly damaged. • Bacterial contamination of platelets has been demonstrated with a frequency as high as 1 in 1000 platelet pools. Now that viral risks are so low, this is the commonest transfusion-transmitted infection. The clinical effects range from mild fever and rigors to acute septicaemia with hypotension and death, and several cases each year are reported to the Serious Hazards of Transfusion (SHOT) scheme. Platelets showing turbidity or unusual colour should be returned to the blood bank for culture. Because the major source of bacteria is the skin of the donors, improved skin cleansing and diversion of the first 20–30 mL of the donation into a side pouch which can be used for testing are

estimated to reduce the bacterial risk by as much as 70%. Bacterial screening of platelets is under consideration in a number of countries, including the UK, since semi-automated methods suitable for mass screening are now available. Because bacterial culture delays issue of the platelets by 1–2 days, there is increased interest in extending shelf-life to 7 days, which was permitted before the bacterial risks became apparent. With prestorage leucocyte depletion and the improved gas exchange offered by modern storage packs, platelets stored for 7 days in plasma maintain their pH well, and several countries are performing additional studies to assess their functionality. If this proves satisfactory, it might be possible to return to a 7-day shelflife for bacterially screened platelets. ‘Washed’ platelets and platelet additive solution

For patients with severe anaphylactic-type reactions, which are usually due to plasma proteins, it is possible to prepare platelets to be virtually plasma-free using one of a number of new platelet additive solutions. This product is sometimes referred to as ‘washed’ platelets, although washing is unnecessary and may lead to platelet activation. These solutions differ from red cell additive solutions in that they contain some or all of potassium, acetate, citrate, phosphate, gluconate and chloride. Platelets in 100% additive solution have only a 24-h shelf-life, but a number of countries have begun to produce platelets in a mixture of 30% plasma and 70% platelet additive solution. This strategy makes more plasma available for fractionation and allows a normal 5-day shelf-life. Data to day 7 and beyond are limited, although one new solution promises storage to day 9 and beyond. These solutions have great potential, but require careful validation, which may need to include volunteer and patient studies. Pathogen reduction in platelets

A group of compounds called psoralens have been developed for their virus and bacterial killing properties, and a system for pathogen inactivation of platelet concentrates using a second-generation psoralen called amotosalen (S-59) has been 267

Chapter 23

licensed in Europe. Psoralens form adducts with DNA and RNA; when activated by exposure to ultraviolet light (UV)A, binding becomes irreversible and nucleic acid replication is blocked. Amotosalen/UVA treatment results in a high degree of killing of the major transfusiontransmitted pathogens human immunodeficiency virus (HIV), hepatitis C virus (HCV) and hepatitis B virus (HBV), including intracellular pathogens such as CMV and HTLV. However, there is no effect on prions, which lack nucleic acid. Additional properties of S-59 photoinactivation include killing of Gram-positive and Gram-negative bacteria, inactivation of the antigen-presenting cells important in HLA alloimmunization, and inhibition of the donor T-cell proliferation that characterizes transfusion-associated graft-versus-host disease (TA-GVHD). Thus there could potentially be multiple benefits from this approach, and in clinical trials of S59-photoinactivated platelets it was considered unnecessary to perform gammairradiation for prevention of TA-GVHD. Randomized clinical trials, albeit on small numbers of patients, have shown that S59-treated platelets, whether prepared by apheresis or from pooled buffy coats, are effective in preventing haemorrhage in thrombocytopenic patients with haematological malignancies. However, platelet increments and intertransfusion intervals were less favourable than in control patients, raising the possibility that increased numbers of platelet units might be required to support such patients. This question can only be answered by further large-scale clinical studies. Other systems for pathogen reduction in platelets are also in development.

Fresh frozen plasma Definition and specification

FFP is the plasma from a single donation, usually 250–300 mL, which has been frozen soon after collection (usually within 8 h) without pooling to a core temperature of less than –30°C. To minimize virus risk, FFP is not manufactured from first-time donors in the UK. FFP can also be derived from apheresis collections, in 300-mL or 600-mL volumes. It is used primarily as a source of multiple 268

coagulation factors in situations such as massive transfusion, disseminated intravascular coagulation and liver disease (see Chapter 11). • The permitted shelf-life depends on storage temperature, e.g. less than –30°C for 12 months; 24month storage is now permitted in Europe. • At least 75% of units processed must contain more than 0.7 IU/mL of factor VIII. Although most FFP is prescribed for patients with normal or elevated factor VIII levels, it is selected for quality monitoring purposes as it is labile and hence sensitive to exposure to adverse conditions. • A standard unit of FFP may contain over 107 leucocytes/mL. There are no specific indications for leucocyte-depleted FFP but if universal leucocyte depletion is implemented, a filtration step is required. This can be done as whole blood prior to separation or by use of a specific plasma filter. • FFP is thawed (in a protective overwrap to prevent bacterial contamination) in a water bath, either with or without agitation; purpose-designed microwave ovens are also available. It has been traditional to recommend that, once thawed, FFP should be maintained at room temperature and transfused as soon as possible. However, more recent data suggest that post-thaw storage at 4°C for up to 24 h results in an acceptable product. Virus inactivation

Two virus-inactivated FFP preparations are now available in the UK, methylene blue (MB) treated and solvent–detergent (SD) treated, with other systems in development. Both methods offer good virus protection but are associated with loss of clotting factors. The key features of MB FFP and SD FFP compared with untreated FFP are shown in Table 23.5. MB is a phenothiazine dye that, when exposed to white light, generates reactive oxygen species which damage nucleic acids, preventing viral replication. Treatment is applied to single unpooled units of plasma, and requires prior removal of white blood cells by filtration or freeze–thawing. MB is contained in or added to the integral pack system, mixed with the plasma, then placed on a light box for activation. MB is generally removed using an adsorption filter prior to transfusion to

Table 23.5 Comparison of standard fresh frozen plasma (FFP) with methylene blue (MB)-treated FFP and solvent–detergent

(SD)-treated FFP. Standard FFP

MB FFP

SD FFP

UK donors, all previously virus tested. Single-unit format

US volunteer donors, all male. Single-unit format

Non-UK donors; pools of up to 380 L (600–1500 ABO identical donations)

Donation testing Serology Genomic

HIV, HBV, HCV, HTLV HCV

HIV, HBV, HCV, HTLV HCV, HIV

HIV, HBV, HCV, HTLV HAV, HCV, B19, HIV, HBV

Virus risk HIV-1 and HIV-2 HCV HBV HAV Parvovirus B19

1 in 10 million 1 in 50 million 1 in 1.2 million Rare event Rare event

No proven cases reported to date for HIV, HBV, HCV (1 possible HCV transmission)

No reported transmissions to date of HIV, HBV, HCV in SD FFP or SD-treated plasma products None reported Batch withdrawals due to possible B19 content. Seroconversion in patients no greater than with untreated FFP

Product characteristics Volume Coagulation factor content

180–300 mL + 50 mL paediatric size Variable between units; 75% units, factor VIII > 0.7 IU/mL

235–305 mL + 50 mL paediatric size Variable between units; 75% units, factor VIII > 0.5 IU/mL; all other factors > 0.5 IU/mL; no reduction ATIII, protein C, protein S. No coagulation factor/complement activation May become available

Not available

Source

No greater than for standard FFP. None reported to date

200 mL; no paediatric size Constant within batch. All factors > 0.5 IU/mL

Cryoprecipitate/ cryosupernatant

Available

Residual additives

None

< 0.3 mmol/L MB. No toxicity seen or predicted at this level, even in premature neonates

< 2 mg/mL TNBP; < 5 mg/mL Triton-X 100. Residual levels not toxic

Allergic reactions

May be reduced by leucocyte depletion 1% 0.1%

Reactions attributable to cells would be expected to be reduced No data No data

Probably less frequent than FFP

As for standard FFP

Pooling reduces all of these risks High-titre anti-A,B not a problem since donations pooled Only 1 possible TRALI case reported

Mild Severe Adverse reactions due to antibody Red cell

Tested for high-titre anti-A,B

Not tested for high-titre anti-A,B

TRALI Thrombocytopenia

> 20 cases/year (SHOT) Very rare

None reported to date

Cellular content

Leucocyte depleted

Leucocyte depleted

No intact cells or fragments; no need to RhD match

Product licence

Not required

Medical device; CE marked

Licensed, batched product

As for FFP

As for FFP

> 1 million units in Europe

3 million units in Europe

Indications Usage to date

300 000 units/year in UK

ATIII, antithrombin III; B19, parvovirus B19; HAV, hepatitis A virus; HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus; HTLV, human T-cell leukaemia/lymphoma virus; SHOT, Serious Hazards of Transfusion; TNBP, tri(N-butyl)phosphate; TRALI, transfusion-related acute lung injury.

269

Chapter 23

the patient, leaving residual MB concentrations of less than 0.3 mmol/L. At these concentrations, no toxicity has been demonstrated or is predicted. Glucose 6-phosphate dehydrogenase deficiency is not a contraindication to use of this product. There is approximately 30% loss of activity of factors VIII and XI; of particular note is the effect of MB on fibrinogen, with up to 30% loss of activity. It is, however, unclear as yet whether these changes require increased volumes of the product to be prescribed. SD treatment can be applied only to pools of several hundred ABO-identical units; as the treatment destroys the lipid envelope of red cells, no RhD matching is required. Exposure to SD destroys the lipid envelope of HIV, HBV and HCV, and no such transmissions have been reported. Non-lipid-coated viruses such as parvovirus B19 and hepatitis A are not specifically inactivated but their titre may be reduced in downstream processing. In addition, plasma pools with high genomic titres of these viruses are rejected, and pools contain specified levels of viral antibodies, which may be at least partially protective. No increase in clinical cases of hepatitis A virus or B19 in SD FFP recipients is evident. Concerns have been expressed in the USA that SD FFP might be associated with increased thrombotic risk in certain clinical situations, attributed to loss of proteins C and S during SD treatment. However, these complications have not been prominent in recipients of SD FFP manufactured by the European method, which results in greater preservation of these proteins. A variant of SD FFP in which ABO groups are mixed is in development. By neutralization of anti-A and anti-B by soluble A and B substance, it is intended to produce a ‘universal FFP’ which could be given to patients of any ABO group. vCJD and FFP

In animal studies, plasma was found to contain infective prion. Therefore, a precautionary measure for the UK was announced in 2002 whereby FFP for children born on or after 1 January 1996 would receive FFP imported from

270

the USA, where no cases of BSE or vCJD have been reported. This is now being implemented, with plasma coming from volunteer donors extensively tested for transfusion-transmitted viruses, including West Nile virus. Plasma is pathogen inactivated by the MB process on arrival in the UK.

Cryoprecipitate and cryosupernatant Cryoprecipitate

Cryoprecipitate is manufactured from single units of FFP by rapid freezing to less than –30°C then slow thawing overnight at 4°C. This precipitates out the so-called cryoproteins, namely factor VIII, fibrinogen, fibronectin and factor XIII. By removing most of the supernatant plasma (‘cryosupernatant’), a component providing a high concentration of these clotting factors is obtained. Current UK guidelines specify that 75% of cryoprecipitate units must contain over 70 IU of factor VIII and 140 mg of fibrinogen, with a 24-month storage period. Although originally developed for factor VIII deficiency (haemophilia A), most cryoprecipitate is now prescribed to treat congenital or acquired hypofibrinogenaemia, usually in the context of liver disease, disseminated intravascular coagulation or massive transfusion. An adult dose of 10–12 packs is generally indicated once the fibrinogen level falls to less than 0.5–1.0 g/L. No licensed preparations of fibrinogen concentrate are yet available in the UK. Cryosupernatant

This is the portion of FFP remaining after separation of cryoprecipitate. It is generally discarded, although contains sufficient fibrinogen to be used as start material for production of fibrinogen concentrate. Cryosupernatant has been used succesfully as replacement fluid in plasma exchange procedures for thrombotic thrombocytopenic purpura. It may have advantages over FFP, possibly because it lacks the highest molecular weight multimers of von Willebrand factor.

Production and storage of components

Virus inactivation of cryoprecipitate and cryosupernatant

Production of cryoprecipitate from SD FFP has been performed experimentally. Such cryoprecipitate contains insufficient von Willebrand factor to treat patients with von Willebrand’s disease. Fibrinogen levels are reduced but acceptable. Manufacture of cryoprecipitate and cryosupernatant from MB plasma is under assessment, but is challenging because of the fibrinogen loss in the start plasma. However, potentially altered recovery in the cryoprecipitation process and the inhibitory effect of MB on fibrin polymerization mean that it may be difficult to achieve adequate concentrations of functional fibrinogen in the MB-treated cryoprecipitate.

Granulocytes for transfusion The use of transfused granulocytes is now uncommon. They are sometimes used for severely neutropenic patients (granulocyte count < 0.5 ¥ 109/L) with focal bacterial or fungal infection refractory to antimicrobial therapy, but there are difficulties in obtaining sufficient functional cells from donors and administering them frequently enough to the patient. Production options have been to use 10 or 12 buffy coats from random donors, or to harvest apheresis granulocytes from family members using a red cell sedimenting agent such as hydroxyethyl starch. Animal studies suggest that greater than 1 ¥ 1010 granulocytes once or twice daily are required to treat an adult, but apheresis can produce no more than 0.5 ¥ 1010 per dose. Therefore unstimulated apheresis granulocytes are directed towards children, in whom an adequate dose can be achieved, with adult patients receiving buffy coat-derived granulocytes. There has been renewed interest in the use of granulocytes in studies of granulocyte colonystimulating factor (G-CSF)-mobilized granulocytes collected by apheresis. Administration to the donor of a single subcutaneous injection of 10 mg/kg of G-CSF plus oral dexamethasone 8 mg 12–24 h prior to apheresis raises the peripheral

leucocyte count to more than 25 ¥ 109/L. This, coupled with Pentaspan sedimentation, allows collection of a therapeutic dose of granulocytes of 5–20 ¥ 108/kg body weight of the recipient. This can result in a measurable rise in the peripheral granulocyte count in the patient, and recovery of migrated cells from saliva. Clinical trials of such granulocytes are ongoing. At present, use of GCSF for granulocyte collection is not permittted in volunteer donors unrelated to the patient. All granulocyte preparations should be released for issue as soon as possible after collection, which may mean that certain time-consuming screening assays such as HCV genome testing cannot be done prior to release. They must be gammairradiated to prevent TA-GVHD and should be administered to the patient without delay. A short period of storage is unavoidable, which should be at 22°C without agitation. Because of red cell contamination, a red cell crossmatch should be performed.

Components for intrauterine transfusion, and for neonates and infants General requirements (Table 23.6)

• In the UK, such components are not manufactured from ‘first-time’ donors. FFP has not been shown to transmit CMV so provision of CMV antibody-negative FFP is not critical. • For components other than those in additive solution, they must be free of clinically significant red cell antibodies, including high titre-anti-A and anti-B. • Blood from haemoglobin sickle heterozygous donors should not be used. • Although components are leucocyte depleted at source, they should still be administered through a 170–200 mm filter to remove any microaggregates formed during storage. • Gamma-irradiation is required for intrauterine and exchange transfusions. ‘Top-up’ red cell transfusions need be irradiated only if there has been a previous intrauterine transfusion (IUT) or if the component is prepared from a family member. Family donations are not encouraged except in

271

Chapter 23 Table 23.6 UK specifications for red cells for intrauterine transfusion (IUT), exchange/large-volume transfusions and ‘top-up’

transfusions for neonates.

Virology-negative donation in previous 2 years Free of high-titre anti-A, anti-B Use of additive solution permitted Leucocyte depleted Gamma-irradiation Haemoglobin S negative Cytomegalovirus seronegative Shelf-life

IUT

Exchange transfusion

Top-up transfusion

Yes Yes No Yes Yes Yes Yes 24 h after irradiation and < 5 days total

Yes Yes No Yes Yes Yes Yes 24 h after irradiation and < 5 days total

Yes No Yes Yes No Yes Yes 35 days

rare cases of fetomaternal alloimmunization where the infant’s requirements cannot be met from donor blood.

should apply to other large-volume transfusions in neonates, such as for cardiac surgery or extracorporeal membrane oxygenation.

Intrauterine transfusions

Top-up transfusions for neonates

Red cells are given in utero to treat severe fetal anaemia due to haemolytic disease of the newborn or parvovirus B19 infection. Red cells for IUT are prepared from blood less than 5 days old to a haematocrit of more than 0.7–0.9 and are gammairradiated. They should be administered within 24 h of irradiation, and always by the end of day 5. In cases of fetomaternal alloimmunization to platelets, weekly transfusions of selected platelets (usually HPA-1a negative) are given in utero via the umbilical vessels. Production of hyperconcentrated platelets by apheresis of a genotyped panel of platelet donors is now possible, yielding a platelet ‘hyperconcentrate’ of more than 120 ¥ 109 platelets in 60 mL of plasma.

Premature neonates are among the most heavily transfused patients in any hospital. Most red cell transfusions are given to replace repeated samples taken for laboratory testing. As each infant may require multiple small transfusions, adult packs are split into four to eight ‘paedipacks’ of 30– 60 mL, which are allocated to one infant for the duration of transfusion dependence. Such a strategy reduces donor exposure considerably. For these small-volume transfusions, red cells in additive solution may be used, up to the normal 35-day shelf-life. Gamma-irradiation of these is not required in the UK, unless there has been a previous IUT or the blood comes from a family member. Studies of erythropoietin in premature infants have not convincingly shown a reduction in transfusion requirements.

Exchange/large-volume transfusion of neonates

This is undertaken to treat hyperbilirubinaemia due to either haemolytic disease or prematurity. Either whole blood or partially packed red cells with a haematocrit of 0.6 may be used. Red cells in additive solution are not recommended by some paediatricians, because of concerns regarding the adverse effects of mannitol, but some countries use this component for exchange transfusion without apparent problems. The same considerations 272

Platelet concentrates and FFP

These are most simply prepared from apheresis donations. Multiple aliquots can be allocated to the same infant if required. An alternative strategy for platelets is to prepare a platelet concentrate from a single buffy coat or from a unit of whole blood using the PRP method. These components are generally used for sick babies with multiple

Production and storage of components

coagulation defects. Platelets from a panel of HPA1a- and HPA-5b-negative donors are available ‘off the shelf’ for rapid availability for suspected cases of neonatal alloimmune thrombocytopenia (see Chapter 5).

Further reading British Committee for Standards in Haematology. Guidelines on the clinical use of leucocyte-depleted blood components. Transfus Med 1998; 8: 59–71. British Committee for Standards in Haematology. Guidelines for the use of platelet transfusions Br J Haematol 2003; 122: 10–23. British Committee for Standards in Haematology. Transfusion guidelines for neonates and older children. Br J Haematol 2004; 124: 433–53. British Committee for Standards in haematology. Guidelines for the use of fresh-frozen plasma. Br J Haematol 2004; 126: 11–28. Council of Europe. Guide to the Preparation, Use and Quality Assurance of Blood Components, 9th edn. Strasbourg: Council of Europe Publishing, 2003.

Dumont LJ, Dzik WH, Rebulla P, Brandwein H, and members of the BEST Working Party of the ISBT. Practical guidelines for process validation and process control of white cell-reduced blood components: report of the Biomedical Excellence for Safer Transfusion (BEST) Working Party of the International Society of Blood Transfusion (ISBT). Transfusion 1996; 36: 11–20. McClelland DBL, ed. Handbook of Transfusion Medicine, 3rd edn. London: The Stationery Office, 2001. Pamphilon DH. Viral inactivation of fresh frozen plasma. Br J Haematol 2000; 109: 680–93. Pamphilon DH, Rider J, Barbara JAJ, Williamson LM. Prevention of transfusion-transmitted cytomegalovirus infection. Transfus Med 1999; 9: 115–23. United Kingdom Blood Transfusion Services/National Institute for Biological Standards and Control. Guidelines for the Blood Transfusion Services in the United Kingdom, 6th edn. The Stationery Office: London, 2002. Williamson LM. How should the safety and efficacy of platelet transfusions be assured? Blood Rev 1998; 12: 203–14.

273

Chapter 24

Medicolegal aspects Patricia E. Hewitt

The provision of blood and blood products from donation to transfusion must be based on sound ethical principles and quality guidelines, and ideally should be controlled within a robust regulatory framework based on statute.

Ethical principles The International Society of Blood Transfusion (ISBT) some years ago instituted a code of ethics setting out the guiding principles for blood donation and transfusion. Following revision, this code of ethics has been adopted by the World Health Organization (WHO). It has also been used to support ethical standards in the drafting of the European Blood Directive. It is recommended that all blood transfusion provision is in accordance with the principles included in this code. The main provisions are listed below. • There should be no coercion to donate blood. • Both donors and recipients must be adequately informed. • Confidentiality must be maintained. • Adequate standards should be enforced. • Clinical need must be the determinant of transfusion therapy.

Quality guidelines Uniformity and process control can be achieved by compliance with detailed quality guidelines. The Guidelines for the Blood Transfusion Services in the United Kingdom and the American Association of Blood Banks Technical Manual are two examples of such documents. Although such 274

guidelines do not possess legal status, they can be interpreted as the definitive statement of minimum professional standards, and set out the requirements to be met for good manufacturing practice. Deliberate non-compliance would be regarded very seriously. Unavoidable non-compliance should be carefully documented and should include a clear explanation of the reasons for noncompliance.

Regulatory framework in the UK The regulatory framework in the UK is briefly described. Similar arrangements will be in place in all developed countries. Medicines Act 1968

The Medicines Act provides the framework for the regulation and control of all dealings with medicinal products. It is clear that cellular and fractionated blood products are included within the terms of the Act. Fractionated products (e.g. albumin, coagulation factor concentrates, intravenous immunoglobulin preparations) are individually licensed. The provision of labile blood components (red cells, platelets, fresh frozen plasma) is enabled by means of an organizational licence awarded to individual blood centres by the Medicines and Healthcare Products Regulatory Agency (MHRA) following appropriate inspection and demonstration of compliance with the standards of good manufacturing practice. This Manufacturers Special Licence is valid for 5 years and individual facilities are reinspected and revalidated on a 2-year cycle. Failure to comply with the licensing

Medicolegal aspects

regulations can lead to the imposition of a fine or closure of an organization, and could proceed to a claim of negligence.

although the means of detecting the defect (the availability of a hepatitis C test) was not necessarily available during the whole period.

Consumer Protection Act 1987

NHS Act 1999

The Consumer Protection Act creates a strict liability action against manufacturers and suppliers when physical injury or property damage is caused by a defective product. The Consumer Protection Act 1987 was enacted as a result of a European Community Directive in 1985 and clearly includes within its terms the provision of all blood and blood products. Its premise is the principle of product liability, that is that there is no need to prove that a negligent action has taken place, merely that the end product is defective and has caused harm. In section 3 (1) of the Act a ‘defect’ is defined as follows ‘There is a defect in the product . . . if the safety of the product is not such as persons generally are entitled to expect . . . ’. Blood providers can be held liable under the terms of the Act as producers, suppliers or keepers. The liability therefore extends from the blood centre producing the product to the hospital blood bank which stores and issues products. There are possible defences within the terms of the Act, such as the ‘state of the art defence’. In essence this means that if a product is found to be defective based on current knowledge, that information cannot be used to prove that the same product was defective sometime previously when the current knowledge was not available (Table 24.1). This defence was not held to apply in the case of the hepatitis C litigation in England, since the defect (the transmission of hepatitis C) was apparent at the time of the claimants’ transfusions (1988–91),

This Act modernizes the NHS in England, Wales and Scotland. Raising standards in the quality of NHS care is at its heart. A statutory duty of quality is now placed on all NHS providers, monitored by means of the Commission for Healthcare Audit and Inspection (CHAI). Additionally, the National Institute of Clinical Excellence (NICE) ensures minimum standards of healthcare are developed throughout the UK. It is accepted that the spur to set up these statutory provisions had been inequality in care. All aspects of healthcare including blood transfusion will no doubt in time be scrutinized; the emphasis is likely to be on clinical aspects of transfusion therapy.

Table 24.1 Relevant UK statutes.

Act

Terms

Medicines Act 1968

Regulates the donation, testing, processing and issuing of blood Encompasses all aspects of provision of blood and blood products from donation to hospital blood bank Concentrates on clinical use of blood

Consumer Protection Act 1987

NHS Act 1999

Duty of care Putting aside the ethical principles, quality guidelines and regulatory framework described above, there remains the clear duty of care which must be at the heart of the provision of blood transfusion. This duty must be according to an accepted standard. At present the standard is determined according to the Bolam principle: ‘The test is the standard of the ordinary skilled man exercising and professing to have that special skill.’ This defines the standard as that of a responsible body of doctors skilled in the same specialty. The standard of care can be supported by the application of professional guidelines, although currently the latter have no legal standing. It must be remembered that inexperience is not a defence and a junior doctor in a specialty is deemed to be equally liable with a senior doctor in the same specialty. The duty of care of blood services, according to the defined standard, is both to the blood donor and to the recipient patient. Duty to the donor

The two general principles that underpin blood 275

Chapter 24

donation are that there should be no harm to the health of the donor, and no risk to the health of the recipient patient. The duty to the donor would therefore include compliance with strict medical selection procedures. The donor needs to be informed about the screening tests performed on the donation and should provide a written consent to human immunodeficiency virus (HIV) testing. Information should be provided on any situation that could potentially pose risk to the donor, for example the administration of growth factors prior to stem cell donation or the use of general anaesthetic during bone marrow donation. In these instances a donor would need to consent formally to the procedures. The blood services in the UK make leaflets available at all routine blood donation sessions, to inform the prospective blood donor of relevant issues. Additionally, the blood service has a duty to maintain the confidentiality of a donor, particularly in the event of a recipient patient being harmed by blood obtained from a single donor. The duty of care to the donor also extends to the clinician prescribing the blood, to ensure appropriate use, particularly as that donation is provided on a voluntary basis with no expectation of monetary gain. Duty to the recipient patient

In the UK the standard of care for patients receiving blood transfusion is likely to be addressed under the legislation referred to above. It is suggested that, as a minimum, this standard of care should include the provision of adequate information to the recipient patient and ensuring appropriate clinical use of individual blood components.

Consent to transfusion Any patient being asked to consent to a medical treatment or investigation has the right to be informed of the aims, benefits and risks of the treatment, and to be given details of any alternatives. Without such information, consent is not valid. The standard NHS ‘Patient agreement to investigation or treatment’ form includes a section completed by the health professional who has the 276

discussion with the patient. This section documents that an explanation has been given to the patient about the proposed investigation or treatment, including the possibility of extra procedures which may be found necessary, and blood transfusion is specifically mentioned at this point. The patient signs a general consent to the procedure/investigation described, embracing the possibility of additional procedures, but has the opportunity to list any procedures for which he or she witholds consent without further discussion. The patient must have the capacity to consent. It is presumed that an adult will have that capacity. No doctor should force a competent adult to accept any treatment even if that adult’s decision appears to be irrational. An adult could be incapacitated and therefore unable to give consent because of loss of consciousness or mental retardation. In general no other person can give consent on behalf of an incapacitated adult. (In some countries, e.g. Scotland, the power of parens patriae applies, where another adult can take responsibility as a parent for an incapacitated individual.) Prior wishes may be taken into consideration where the adult has previously been competent and the treatment is regarded as noncontroversial. Treatment may be given to an adult incapable of consenting if the treatment is urgent and in the patient’s best interests. In an elective situation, however, it would be best to seek a ruling from a court of law. The General Medical Council has published a comprehensive guide to consent and this is recommended for more detailed reading (Table 24.2). In the case of children, the Family Law Reform Act 1969 makes it lawful for a minor to consent to, or refuse, treatment when he or she reaches the age of 16 years. In the case of a child below 16 years of age, the parents usually give consent, although such children are able to give valid consent in their own right if they are capable of understanding Table 24.2 Informed consent must include these elements.

Capacity to understand Should be based on adequate information Should be obtained without coercion

Medicolegal aspects

clearly the nature of the proposed treatment. Here the difficulty could be where parents could have specific religious beliefs that prevent them from consenting to blood transfusion for their child. In this instance, a doctor can decide to provide a treatment, including blood transfusion, in the child’s best interests. To fulfil these criteria, the treatment must be carried out in order to save life or to ensure improvement of or to prevent deterioration in the physical or mental health of the child. This would form the basis of a doctor’s individual decision during an emergency situation, but in the case of a planned blood transfusion it would be appropriate to seek a ruling from a court of law. In these circumstances it is recommended that the doctor’s medical defence body is consulted for advice on how to proceed. For consent to be informed and valid, it must be based on adequate information. Various attempts have been made to define what should be included as adequate information and it is legally acceptable that the explanation need not include all the potential adverse consequences if the risk of them occurring is small or immaterial. Minor insignificant reactions to transfusion occur relatively commonly, whereas the risk of complications with serious or fatal long-term consequences, e.g. transmission of HIV, is extremely low. However, there is heightened public awareness of such low risk, and it would therefore be appropriate for these events to be included in a preliminary explanation. Again, the standard that applies in the UK regarding the provision of information is the standard of a responsible body of skilled doctors, the Bolam principle. In the USA, however, a different rule applies, the standard being judged not according to the information that a reasonable doctor would think relevant to impart, but rather according to that which a prudent patient would think relevant to receive, a situation which is likely to develop within the UK within the next few years. There must be no coercion in obtaining consent. A competent adult is able to accept or refuse treatment even if that decision could lead to harm or indeed death. Should an individual doctor decide to treat an adult without consent, then that doctor should be prepared to explain and justify the decision.

Jehovah’s Witnesses Jehovah’s Witnesses, because of their religious beliefs, will never accept normal blood transfusion therapy, although in appropriate circumstances could find cell salvage in continuous circulation acceptable. Many Jehovah’s Witnesses carry an Advance Medical Directive which states the individual’s views and requirements to be followed in the event that the individual is unconscious or otherwise unable to express his or her views. What should be done in this instance? Where the situation is one relating to a competent adult, as long as it is clear there is no coercion, the decision to refuse treatment must be respected, even if it would lead to harm or indeed death of the patient. In an emergency situation, if the patient is unconscious, and therefore incapacitated, then prior previously held beliefs must be taken into account and blood transfusion should not be prescribed if those beliefs made it clear that it was unacceptable. In the situation of a child, where the parents’ religious beliefs could prevent the child from being given a necessary blood transfusion, it would be advisable to seek a proper legal ruling, which would usually mean the child becoming a ward of court and therefore decisions on the treatment being taken by the court.

Patient recourse Despite compliance with standards and appropriate care, things do and will go wrong. In some countries, for example New Zealand and the Scandinavian countries (Sweden, Norway, Finland and Denmark), compensation for medical accidents is provided under a ‘no fault’ system. However, in most countries there is a need to prove liability. In these circumstances, liability will rest either with the individual doctor or with the health employer if vicarious liability applies (this is the current position in the UK for all NHS work). There have been examples of ‘no fault’ compensation awarded in the UK in specific circumstances, for example the Vaccine Damage Payment Scheme, which provides payment of a fixed lump sum where serious mental or physical damage has been caused by the 277

Chapter 24

administration of specified vaccines. In relation to UK blood transfusion, payments were provided to recipients infected by HIV through transfusion both before and after the introduction of mandatory screening of the blood supply in the UK. These were ex gratia payments to those affected, with the government emphasizing that they should not be regarded as an admission of liability or as compensation, but as a response to a particular and tragic situation. Requests for similar treatment for individuals infected with hepatitis C (and hepatitis B) were, until recently, refused. However, an announcement in August 2003 indicated that patients who had been infected with hepatitis C through treatment with blood products would receive payments in a similar manner. At the time of writing, it has been stressed that similar treatment will not apply to recipients infected with, for example, hepatitis B (or any other agent) through blood transfusion, although such cases are small in number and it becomes difficult to explain the different treatment of patients who have suffered apparently similar unfortunate and unexpected adverse effects through their treatment with blood transfusion. Furthermore, the announcement of ex gratia payments for hepatitis C infection followed the successful claims under the Consumer Protection Act (see below) and has led to a number of questions which remain unresolved. For example, can recipients receive payments twice over and how will the level of payments be decided in the hepatitis C scheme? A patient who has suffered harm can bring an action either in medical negligence or under product liability. If brought in negligence, there would be a need to prove a breach in the duty of care, and that the breach directly caused harm to the patient. If brought under product liability, negligence need not be present; a defective product must have directly caused the harm (Table 24.3). An example of an action that could be brought under medical negligence would be that of the transfusion of a unit of red cells to the wrong recipient patient because of failure to check patient identification. Here there would be a clear breach of the duty of care, that is in checking the patient identification against the red cell unit, and it would also be simple to demonstrate that harm, in the 278

Table 24.3 Comparison of medical negligence versus

product liability. Medical negligence

Product liability

Duty of care Breach of the duty Harm caused directly by the breach

Defective product Harm caused directly by the defect

form of a haemolytic transfusion reaction, had occurred as a direct result of the breach. The recipient patient would be able to seek damages for the injury and compensation for any consequent financial loss. Cases of product liability in relation to blood transfusion in Europe are few. The most notable case was that of a number of recipients (114) in England and Wales who brought a claim under the Consumer Protection Act which was heard in 2000–01. In his judgement, Burton found that the Blood Service was liable for the damage because the product (i.e. the blood) did not provide the safety that the consumer (patient) was ‘generally entitled to expect’. The claimants were awarded damages on a provisional basis according to the damage (extent of hepatitis C disease) present at the time of the action. Provisional damages allow for the claimants to return with a future claim should their medical condition deteriorate. This case has attracted much attention within Europe and will surely be used as a precedent in future claims. Other case law within Europe is scarce, although some European countries (e.g. France and the Scandinavian countries) provide for ‘no fault’ compensation in relation to infection acquired through medical treatment. The judgement in the hepatitis C litigation was not appealed. Future similar claims are likely to be made and will be settled, unless the judgement is contested.

Note added in proof The European Blood Safety Directives (2002/93 and 2004/33) will be transposed into UK law as the Blood Safety and Quality Regulations 2005, due to come into force on 8th February 2005.

Medicolegal aspects

The new regulations impose safety and quality requirements on human blood collection, testing, processing, and storage. The requirements apply to blood transfusion service in England, Scotland, Wales, and Northern Ireland. Many of the provisions of the regulations also apply to hospital blood banks. A further Direcive on haemovigilance and traceability, and on quality systems is being discussed by the European Commission’s Expert Group on Blood. It is likely that this further Directive will not be finalized until 2005 and will be then transposed into UK law by separate amending legislation. The regulations will replace some of those currently covered under the Medicines Act (inspection, licensing and accreditation). It is intended that these activities will cease to be regulated under the Medicines Act, which will therefore need to be amended. At the time of writing, the new regulations are subject to public consultation. They will have wide-reaching implications for both Blood Services and hospital blood transfusion laboratoris.

Further reading

American Association of Blood Banks. Technical Manual, 14th edn. Bethesda, MA: AABB, 2002. Bolam v Friern Barnet Hospital Management Committee [1957] 2 All ER 118. Braithwaite M, Beresford N. Law for Doctors: Principles and Practicalities. London: Royal Society of Medicine Press, 2002. The Consumer Protection Act 1987. In: Halsbury’s Statute of England. HMSO: London. F v West Berkshire Health Authority [1989] 2 All ER 545. General Medical Council. Seeking Patients’ Consent: the Ethical Considerations. London: GMC, 1999. Gillick v West Norfolk and Wisbech Area Health Authority [1984] 3 All ER 402. Goldberg R. Paying for bad blood. Strict product liability after the hepatitis C litigation. Med Law Rev 2002; 10: 165–200. Grubb A, Pearl DA. Blood Testing, Aids and DNA Profiling. Bristol: Family Law (Jordan and Sons Ltd), 1990. Guidelines for the Blood Transfusion Services in the United Kingdom, 6th edn. The Stationery Office: Norwich, 2002. The Medicines Act 1968. In: Halsbury’s Statute of England. HMSO: London. NHS Conferation. The coming year in Parliament for health. Briefing 1998; issue 24. Warden J. HIV infected haemophiliacs: 90 million more. Br Med J 1989; 299: 1358. Williams FG. Consent for transfusion. Br Med J 1997; 315: 380–1.

A and others v National Blood Authority [2002] 3 All ER 289.

279

Chapter 25

Blood transfusion in hospitals Sue Knowles and Geoff Poole

Transfusion medicine in a hospital setting is focused on ensuring that a patient receives the correct blood component support which is clinically indicated, in a safe, timely and cost-efficient manner. Specialists in all branches of medicine and surgery are involved in prescribing blood components and the transfusion process itself involves multiple steps and the cooperative action of several groups of staff. Haemovigilance reports from around the world have confirmed that mistransfusion, i.e. giving the patient the incorrect unit of blood, resulting in an ABO-incompatible transfusion, is the most frequent cause of mortality and morbidity resulting from blood transfusion. Mistransfusion is the result of human error that can occur at all steps in the transfusion process, due to failures to comply with clerical or technical procedures and often compounded by systems that are either poorly constructed or not understood. Multiple errors can be made in the transfusion process, some of which can be detected during an effective bedside check at the time of administering blood. It has been observed that as many as 1 in 19 000 red cells are given erroneously and 1 in 33 000 will involve ABO-incompatible units. Estimates of mortality from mistransfusion range from 1 in 600 000 units to 1 in 1.8 million. While there has been enormous progress made in reducing the risks of transfusion-transmitted infections, there is no evidence from serial annual haemovigilance reports that the safety of blood transfusion in a hospital setting is improving. Although there are numerous guidelines and directives available, and an understanding of measures which can be taken to improve transfusion safety, in practice an effective transfusion quality assurance programme is required to minimize the risk of 280

mistransfusion and to avoid the other patient risks and wastages associated with the transfusion process (Table 25.1). This in turn requires committed leadership and adequate resources.

Key features of a quality assurance system for transfusion practice A quality assurance system can be described as the sum of the activities planned and performed to provide confidence that all processes and their elements that influence the quality of transfusion practice are working as expected. In the UK, the Department of Health emphasized this requirement in a health service circular in 2002 entitled ‘Better Blood Transfusion’ (HSC 2002/009). The key features for assuring safe and effective transfusion therapy can be summarized as follows. • Hospital or Trust executive/board: to provide commitment and appropriate resources for the quality system. • Hospital transfusion committee (HTC): to provide multidisciplinary ownership, leadership, and review of the quality system. • Hospital transfusion team (HTT), consisting of the specialist practitioner(s) of transfusion (hospital transfusion safety officer), the lead consultant for blood transfusion and the blood bank manager: to deliver the HTC’s objectives of improving transfusion medicine practice and to implement and monitor the processes of delivering patient care. • Staff involved in the transfusion process: to deliver and/or receive continuous education in transfusion medicine, training in specific procedures and assessment of their competencies.

Blood transfusion in hospitals Table 25.1 Errors in the transfusion process and some potential outcomes.

Problem

Outcome

Unnecessary prescriptions

Patient subjected to unnecessary risk Wastage of blood components

Failure to prescribe specialist components

Risk of, for example, transfusion-associated graft versus host disease

Failure to keep blood in a controlled environment

Wastage of blood components

Incorrect interval between sampling for pretransfusion testing and transfusion

Potential for acute and delayed haemolytic transfusion reactions

Sample for pretransfusion testing taken from incorrect patient. Transposition of samples or other errors in the laboratory. Incorrect unit of blood collected for and/or administered to the patient

Mistransfusion and potential for an ABO- or RhD-incompatible transfusion

Insensitive techniques in pre-transfusion testing

Potential for acute and delayed haemolytic transfusion reactions

Poor laboratory stock control

Wastage of units Inappropriate use of group O and national shortages of that group

Delays in provision of blood components in an emergency

Patient morbidity from hypoxia or coagulopathy

• Policies and guidelines: documented principles and recommendations which guide all activities. • Standard operating procedures/integrated care pathways: documented work instructions and steps which should be followed. • Equipment/material: planned maintenance and calibration of equipment (e.g. blood bank refrigerators, blood warmers and centrifuges), validation of reagents and techniques used in pretransfusion testing. • System review: (a) performance monitoring, e.g. utilization statistics, wastage, results of laboratory external quality assessment; (b) audits, i.e. compliance with guidelines or procedures; and (c) incident and error investigation, analysis and reporting.

Healthcare Organizations. In the UK, the Department of Health initially issued a health services circular ‘Better Blood Transfusion’ (HSC 1998/224) in December 1998, requiring all hospitals/Trusts to have an HTC in place by March 1999. Since then, HSC 2002/009 has expanded the requirements to encompass the key features of a quality infrastructure. To be effective, the HTC requires a dedicated HTT and adequate resources, including IT and clerical support to facilitate data retrieval and audit. Given the importance of the HTC with respect to clinical governance, it should report to the hospital/Trust management board, through options which include the clinical governance committee, the risk management committee or the clinical executive. Terms of reference

Hospital Transfusion Committee The HTC is the focal point for overseeing transfusion practice. In the USA, an HTC has been a requirement since 1972 for hospital accreditation by the Joint Commission on Accreditation of

The HTC is charged with the review of: • clinical transfusion practice; • performance of the hospital transfusion service; • performance of the local blood centre as a provider; and • legal implications of transfusion practice. 281

Chapter 25

Functions Clinical transfusion practice

• To ensure that local policies and procedures are in place, based upon national guidelines and regulations. These should include the following: (a) the administration of blood and blood components and the management of the transfused patient, including the collection of samples for compatibility testing and the management of adverse events related to blood transfusion; (b) a maximum surgical blood ordering schedule (MSBOS); (c) the appropriate use of blood components and blood products; (d) indications for specialist blood components (gamma-irradiated, CMV seronegative and phenotyped); (e) management of massive transfusion; (f) use of autologous blood; (g) use of pharmacological agents to reduce blood usage during surgery. • To ensure that all relevant groups of staff are trained in accordance with local policies, and their competencies are assessed. • To commission audits of compliance with these policies and procedures. • To organize continuing education in transfusion medicine for all members of the hospital staff involved in prescribing or administering blood. • To ensure that investigations are conducted of near misses and adverse events, which in turn will focus the need for further education or amendments to existing procedures. • To ensure that transfusion incidents are reported to the national haemovigilance scheme, i.e. Serious Hazards of Transfusion (SHOT) in the UK. • To review performance regularly, e.g. wastage rates, inappropriate blood component usage, utilization of blood components by directorate, user or surgical procedure. Monitoring the performance of the hospital transfusion service

• To review the operational effectiveness of the service, e.g. response times for emergency requests, elective work undertaken on call, crossmatch to 282

transfusion ratios for individual surgical procedures and users. • To review quality assurance measures including performance in external quality assessment schemes and the outcome of accreditation inspections and other external audits. Monitoring the performance of the local blood centre as provider

• To review the adequacy and timely provision of blood and blood components. • To review the adequacy of diagnostic reference services and consultant advisory services. Legal implications of transfusion practice

• To ensure that patients undergoing transfusion are provided with access to information in relation to the risks and benefits of transfusion. • To ensure that an audit trail of documentation exists to trace the ultimate fate of all blood components received. • To ensure that all relevant aspects of product liability and health and safety are adequately addressed. Composition

The Chair should be appointed by the Trust chief executive and should have a good understanding and experience of transfusion medicine practice. Ideally, the Chair should not be the consultant responsible for the hospital transfusion laboratory, who could be perceived to have a vested interest. The following membership is suggested: • representatives of major speciality users of blood in all directorates; • lead consultant haematologist for blood transfusion; • hospital blood bank manager; • specialist practitioner(s) of transfusion; • senior nursing officer; • representative from clinical risk management; • representative of junior medical staff; • representative of Trust management; • local blood centre consultant (ex officio);

Blood transfusion in hospitals

• other representatives may be co-opted as required, e.g. from medical records, portering staff, clinical audit, training or pharmacy.

Administration of blood and blood components and the management of the transfused patient This process involves several steps: • counselling the patient of the need for a blood transfusion, given that alternative approaches (autologous transfusions and/or erythropoietin) are insufficient or inappropriate for their circumstances; • the prescription of blood and blood components; • requests for blood and blood components; • sampling for pretransfusion compatibility testing; • collection and delivery of blood and blood components from transfusion issue refrigerator to clinical care area; • administration of blood and blood components; and • monitoring of transfused patients. Errors can occur at each of these steps and data provided in the UK SHOT report for 2002–03 shows that 20% of errors are made at the time of

prescription, sampling and request, 29% arise in the hospital blood bank and 48% are made at the time of collecting and administering the component. The Trust should have written procedures to cover all these steps, to which the relevant staff are trained and assessed. The responsibilities, actions, documents and potential errors are provided in Tables 25.2–25.7. The blood transfusion compatibility form, the prescription chart and the nursing observations related to the transfusion should be kept in the patient’s medical case notes as a permanent record. Since correct identification of both patient and blood unit are critical control processes, computerized systems have been developed for the blood administration process, which includes bedside verification of the match between patient identification and blood unit identification.

Technologies to reduce errors in administering blood Additional manual systems of patient identification

Sets of red labels with the same unique number can be allocated to a transfusion episode. A label can be incorporated into an additional patient wristband at the time of phlebotomy, affixed to the

Table 25.2 Prescription of blood and blood components.

Responsibility

Action

Document

Potential errors

Medical officer

Ensure patient is aware of the need for a blood transfusion and has read and understood information related to the risks and benefits of transfusion Prescribe component, any specialist requirements, quantity, and duration of transfusion

Patient information leaflet Hospital consent form

Failure to take account of patient’s religious beliefs or other views

Prescription chart

Unnecessary prescription, in the failure to follow hospital guidelines or as a result of an error in baseline blood count or coagulation screen. Lack of awareness of or failure to prescribe specialist components

Document rationale for transfusion

Patient case notes

Related hospital procedures and documents Guidelines for the use of blood and blood components, including specialist components. Practice guidelines/procedures for individual diseases/treatments.

283

Chapter 25 Table 25.3 Requests for blood and blood components.

Responsibility

Action

Documentation

Potential errors

Doctor or registered nurse

State full patient identification, location, diagnosis, details of type and quantity of component and time required. State previous obstetric and transfusion history when requesting red cells

Written/electronic request form or laboratory telephone log in an emergency

Incomplete or incorrect patient information leading to failure to recognize historical laboratory record, recording requirement for specialist components or phenotyped units Failure to request specialist components Failure to comply with hospital MSBOS

Hospital blood transfusion laboratory staff

Review historical record and whether a further sample for pretransfusion testing required

Previous laboratory record

Patient identification error in transcribing telephone request Failure to locate/heed information contained in historical record Failure to request a new sample in a recently transfused patient and potential for overlooking newly developed red cell antibodies

Related hospital procedures and documents Timing of pretransfusion sampling with respect to previous transfusion. Maximum surgical blood ordering schedules (MSBOS).

Table 25.4 Sampling for pretransfusion compatibility testing.

Responsibility

Action

Medical, nursing or phlebotomy staff, authorized and trained

Direct questioning of patient to provide surname, first name and date of birth when judged capable. Check that details given match those on patient wristband and on request form Take blood sample and immediately label at bedside with the required patient information

Hospital blood transfusion laboratory staff

To determine that sample labelling meets requirements for pretransfusion testing. If unacceptable, to inform requester of the need for another sample

Documentation

Potential errors Patient misidentification as a result of failing to positively identify patient or as a result of wristband missing or with incomplete information

Sample labelled and signed

Patient misidentification as a result of: • Prelabelled sample tube with another patient’s ID • Labelled away from the bedside with another patient’s ID • Addressograph label affixed from incorrect patient

If unacceptable, to document reasons and log

Potential to issue inappropriate unit if inadequately labelled samples accepted Failure to provide blood in a timely manner if clinicians unaware of the need for another sample

Related hospital procedures and documents Trust sample labelling policy. Trust policy for allocation and maintenance of unique patient identifiers and for resiting wristbands in theatre or intensive care.

284

Blood transfusion in hospitals Table 25.5 Collection and delivery of blood and blood components from transfusion issue refrigerator to clinical care area.

Responsibility

Action

Documentation

Potential errors

Staff authorized and trained

Take documentation bearing patient identification to the issue refrigerator Check that unit removed and accompanying blood transfusion compatibility form bear the identical patient identification details Record time and sign that correct unit has been collected

Prescription chart or a completed collection slip

Incorrect unit collected if no documentation bearing patient ID Incorrect unit removed

Lack of audit trail from failure to sign out unit from issue refrigerator

Related hospital procedures and documents Trust blood collection policy.

Table 25.6 Administration of blood and blood components.

Responsibility

Action

Documentation

Potential errors

Doctor or registered nurse

At the bedside, direct questioning of patient to provide surname, first name and date of birth when judged capable. Check that this identity is identical with documents

Prescription chart Case notes Compatibility form Compatibility label Patient’s wristband

Unit transfused to wrong patient if unit checked away from bedside or no verification of patient identity

Check that blood group is compatible

Case notes Compatibility form Compatibility label Base label on blood pack

Incorrect ABO/D group transfused if failure to detect laboratory grouping or labelling error

Check that special requirements are fulfilled

Prescription chart Blood pack

Inappropriate component transfused if failure to note laboratory issuing error

Check that unit of blood has of not passed its expiry date, and it is intact with no evidence visual discoloration Document date and time of commencement of unit and sign

Failure to note transfusion of timeexpired component Failure to note unit potentially contaminated with bacteria Compatibility form and/or prescription chart

Related hospital procedures and documents Trust blood administration policy.

request form, sample tube and into the current medical notes, and the unique number can also be printed onto the compatibility labels and compatibility report form. At the time of administration, the additional unique number provides a supplementary means of cross-checking.

Barcoded patient administration systems

The use of hand-held computers and portable barcode label and wristband printers provide the means for improving patient identity and patient safety. At the stage of sample collection, the 285

Chapter 25 Table 25.7 Monitoring of transfused patients.

Responsibility

Action

Documentation

Potential errors

Staff authorized and trained

Measure temperature, pulse and blood pressure before the start of each unit Explain to the patient possible adverse effects to be reported and keep patient under close visual observation in first 15–20 min of each unit Measure temperature and pulse 15 min after start of each unit Measure temperature, pulse and blood pressure at the end of each unit

Observation chart, recording date and time

In absence baseline observations, cannot detect any change giving a warning of transfusion reaction Patient not aware of symptoms to be reported that can provide first warning of a transfusion reaction In absence early observation, potential to miss a serious transfusion reaction In absence of timed final observation, cannot know whether any subsequent changes in patient’s condition are temporarily related to an ongoing transfusion

Observation chart, recording time Observation chart, recording time

Related hospital procedures and documents Trust policy on monitoring transfused patients. Management of transfusion reactions.

phlebotomist’s and patient’s identity can be scanned and a barcoded label generated at the bedside to attach to the tube. In the laboratory, the allocated unit is labelled to incorporate the patient’s unique identification barcode and the unit number. At the time of administration, the member of staff is prompted by the hand-held computer to scan their own identification barcode, the barcoded patient wristband, the compatibility label and the unit number. The computer confirms whether the unit is correct for the patient and also provides prompts to check for special requirements, pretransfusion observations and the expiry date of the pack.

Influencing clinical practice There are several potential factors that influence transfusion medicine practice and decision-making: • physician knowledge; • physician perception based upon clinical experience; • peer pressure and feedback; • effectiveness of the hospital governance framework; 286

• educational prompts at the time of the decision; • financial pressures or incentives; and • public and political perceptions and the fear of litigation. Improving transfusion medicine practice within a hospital community requires a planned consistent approach which is endorsed and implemented through clinical governance. Guidelines, algorithms and protocols

Guidelines are defined as systematically developed statements to assist practitioner and patient decisions about appropriate healthcare for specific clinical circumstances. Controlled data are unavailable to assess the impact of professional guidelines, but most would agree that nationally derived documents rarely lead to change unless there is a local implementation and dissemination strategy, which requires time and resources. Developing a local strategy to implement the guidelines is a useful opportunity to gain ownership, in that it can provide educational opportunities in examining the evidence basis, and identify dissension and other local barriers to its implementation, e.g. staff resources, laboratory turnaround times.

Blood transfusion in hospitals

Local groups should adopt the recommendations of pre-existing guidelines but customize them for local use. This may involve separating a guideline into several sections or incorporating some of its recommendations into other local protocols for specific conditions, e.g. a fresh frozen plasma (FFP) guideline incorporated into protocol for the management of disseminated intravascular coagulation (DIC), massive haemorrhage and obstetric haemorrhage. Experience in other medical fields has also demonstrated that embedding the recommendations of a guideline into documents in use at the time of the clinical consultation/decision can significantly improve compliance. Examples of this approach could include: • listing the indications for specialist blood components on the blood transfusion request forms or electronic request screens; • listing the nursing actions to be taken in the event of a transfusion reaction on a specific transfusion observation chart; and • detailing the checks to be made prior to administering blood on the compatibility form. Intraoperative algorithms for the use of platelet concentrates and FFP to correct microvascular bleeding during and after cardiac bypass surgery have also proved to be successful in reducing inappropriate use of these components, when combined with near-patient testing and the rapid availability of results to feed into the decision tree. The local documents should be disseminated alongside training events for all involved staff. A list of relevant guidelines is included at the end of this chapter. Audit

The audit cycle consists of defining the area to be studied and comparing observed practice with a standard. Analysis of the findings should lead to recommendations for improved practice, which may include a revision of the content and clarity of the standard (Fig. 25.1). The audit process has been criticized since it has been said to consume considerable resources and result, at best, in only a transient change in clinical behaviour. However, audits frequently fail because

What are we aiming to do? Standard and subsequent review

Have we made things better? Further comparison

Doing something to make things better Corrective action

Are we doing it? Audit project; comparison of activity with standard

Why are we not doing it? Analysis of causes of non-compliance

Fig. 25.1 The audit cycle.

the cycle is not completed, i.e. when the results and educational messages are not disseminated and discussed by those whose practice could be improved, when no analysis is undertaken of the corrective actions which should be introduced to improve practice, and no resources are provided to implement the actions identified. Audits can also be made more effective when they are conducted repeatedly or on an ongoing basis. Audits can be conducted retrospectively or prospectively. If adopting a prospective approach in monitoring the appropriateness of the requests for blood components, this can be considered to be intrusive and potentially delay the delivery of patient care, but does prevent unnecessary transfusion. Immediate retrospective audit of component utilization does not prevent unnecessary transfusion but, if conducted in a timely and individual fashion, can provide effective educational feedback and a sustained change to a physician’s prescribing habit. Ongoing computerized prospective or retrospective auditing, using agreed algorithms, is the only realistic way of monitoring and providing individual clinician feedback on the inappropriate use of blood components. Regular audit cycles have been shown to 287

Chapter 25

improve compliance with the bedside procedures for administering blood. Regular audits of an MSBOS are essential if the schedule is to be kept in line with changing practice. Audits that involve several healthcare organizations are particularly effective since peer pressure is applied in comparing practice. Participating hospitals’ results can be made anonymous for all other participants. Suitable audits would include the percentage of inappropriate use of blood components, of patients with correct wristband identification and of identification checks made at the time of administering blood. Surveys

Many activities which fall under an ‘audit’ banner are not comparing practice with a standard but are monitoring or surveying practice. These activities, many of which can be quantified, often lead to the development of quality indicators or performance indicators. Trend analysis, or comparison of one organization with another, or one blood user with another, are powerful means of exerting peer pressure and influencing practice (benchmarking). Performance indicators or benchmarking can be applied, for example, to: • percentage of patient wristband errors; • percentage of mislabelled samples; • hospital blood wastage; • percentage of group O usage; • number of units crossmatched to number of units transfused (C : T) ratio; • red cells used per surgical procedure (for each surgical team); • percentage of primary arthroplasties requiring allogeneic transfusion; • percentage of patients receiving platelets after coronary artery bypass grafting; • percentage of patients with refractory anaemia having received a trial of treatment with erythropoietin. National schemes

A number of national schemes in the UK set out to monitor or assure transfusion practice. Such

288

schemes can be used to influence policy and educate users within hospitals. They are generally made anonymous to encourage full participation. National schemes include the following. • The SHOT scheme was launched in November 1996. It is a voluntary system for collecting data on serious adverse events in the transfusion of blood and blood components. It produces an annual report of its findings and recommendations. • External Quality Assurance (EQA) schemes, for example the National External Quality Assurance (NEQAS) scheme in Blood Transfusion Laboratory Practice (BTLP), provide ‘clinical’ material to laboratories on a regular basis. Laboratory results are returned to the scheme organizers for analysis and collated reports are disseminated to users. • The Blood Stocks Management Scheme is a joint venture between hospital laboratories and the National Blood Service. It collates and publishes through its website details of blood stock inventory and wastage and allows participants to compare their practice with that of comparably sized hospitals. • The Royal College of Physicians and the National Blood Service in England have established a national comparative audit initiative, which has concentrated so far on benchmarking aspects of safe transfusion practice, and will soon encompass usage of blood in certain common medical and surgical conditions. Public and political perceptions and fear of litigation

The knowledge that human immunodeficiency virus (HIV) could be transmitted by blood transfusion in 1982 led to a decline in the use of allogeneic red cells in the USA, from 12.2 million units in 1986 to 11.4 million units in 1997. This decline is even more significant if the growth and ageing of the population in the USA during this period are taken into account. Over the same period, autologous donations increased by a factor of more than 30. Individual physicians were sued in the USA if their patient contracted HIV through the blood supply and their transfusion was not clinically indicated. Concern about variant Creutzfeldt–Jakob

Blood transfusion in hospitals

disease (vCJD) being transmissible through blood led in 1998 to the Department of Health in the UK requiring that all hospitals/trusts should have HTCs, implement good transfusion practice and that they should have explored the feasibility of autologous blood transfusion. Universal leucocyte depletion of blood, when introduced in the UK in 1999 as a preventive measure for vCJD, led to a significant increase in the price of red cells, and encouraged a more judicious approach to transfusion and the use of transfusion alternatives. As a consequence, total red cell usage in the UK reached a plateau between 2000 and 2001 and has subsequently fallen by about 1% per year, despite the increase in surgical procedures performed over this period. Local investigation and feedback following critical incidents and ‘near misses’

The UK SHOT scheme defines a ‘near miss’ as any error which, if undetected, could result in the determination of a wrong blood group, or the issue, collection or administration of an incorrect, inappropriate or unsuitable component but which was recognized before transfusion took place. ‘Near-miss’ errors are common, and if systematically analysed and collated provide the opportunity to understand the potential weaknesses in the process of blood transfusion. Corrective action can then be taken to minimize the occurrence of a critical incident. Identified weaknesses include staff misconceptions or ignorance, defective or risky protocols or processes. Sample errors, most importantly those where the tube is labelled with the intended patient’s details but is subsequently found to contain blood from another patient, are the commonest detectable errors. These inevitably arise as a result of a failure to systematically and positively identify the patient at the bedside. Corrective action should involve counselling and educating the individual concerned who failed to comply with the correct procedure. However any investigation will also uncover compounding latent factors contributing to the event, which need to be collectively understood and addressed, for example:

• the practice of not positively identifying patients, since healthcare workers perceive this as denoting an inadequate knowledge of the patients under their care; • reduced junior doctors’ hours and shift patterns of all those involved in direct patient management leading to unfamiliarity with patients; and • patient ‘hot bedding’ in the UK, which frequently leads to preoperative patients having to be sampled for pretransfusion testing before case notes are made available on the wards or wristbands are applied. Sadly, exposure to avoidable patient morbidity or fatality is often the trigger for affecting the local medical and nursing community’s perception of the risks of blood transfusion and for instigating a change in procedures. Education and continuing professional development

Education of all individuals in the transfusion process is difficult to achieve unless it is made an integral part of a hospital/Trust documented mandatory training programme and the process is subject to external inspection. Even then, it requires a dedicated resource and a flexible and pragmatic approach to accommodate shift patterns, staff shortages and agency staff. Education is nevertheless an essential component of every strategy to gain clinician compliance with clinical procedures and guidelines and to modify practice as a result of audit, surveys or investigations into errors. Educational interventions have been found to be more successful when they are interactive, focused on a specific objective and directed at groups of individuals with reflections on their own practice. Continuing professional development schemes exist for the various professional groups involved in healthcare. The schemes vary but all are intended to encourage knowledge acquisition. In a typical scheme, such as the one introduced by the British Blood Transfusion Society (BBTS) in April 2001, members keep a portfolio in which accredited activity in educational, professional and vocational areas is recorded.

289

Chapter 25

Maximum surgical blood ordering schedule

Table 25.8 Maximum surgical blood order schedule

(general surgery).

This is a table of elective surgical procedures that lists the number of units of blood routinely crossmatched for each. The number of units allocated takes into account the likelihood of the need for transfusion and the response time for receiving blood following an immediate spin crossmatch or electronic issue. An MSBOS reduces the workload of unnecessary crossmatching and issuing of blood, and can improve stock management and wastage. The successful implementation of an MSBOS depends upon all parties agreeing to the tariff, the education of junior staff, the confidence of senior staff that there is a robust system for accessing blood promptly when there is unexpected blood loss and the ability to override the tariff when there are reasons to indicate that greater blood loss will occur. A tariff is constructed by: • analysing each surgical procedure in terms of the C : T ratio; • managing procedures with a C : T ratio greater than 2, i.e. a low probability of transfusion, with a group and screen, and issuing blood only when there is a need for transfusion; and • allocating an agreed number of units for procedures with a C : T ratio of less than 2. In recipients with red cell alloantibodies, consideration should be given to the time taken to acquire and crossmatch antigen-negative units. An overall C : T ratio of 1.5 for elective surgery is achievable when the laboratory is centrally issuing blood. However, lower ratios would be possible with remote electronic issue in theatre suites. An example of an MSBOS is provided in Table 25.8.

Pretransfusion compatibility testing This testing comprises: • determination of the ABO and D group of the recipient; • a screen for red cell alloantibodies reactive at 37°C in the plasma of the recipient; • a check for previous records or duplicate records, and comparison of current with historical 290

Operation

Units crossmatched or group and screen (G & S)

Adrenalectomy Colectomy Cholecystectomy Gastrostomy, ileostomy, colostomy Gastrectomy (partial) Liver biopsy Mastectomy Oesophagectomy Pancreatectomy Parathyroidectomy Partial hepatectomy Splenectomy Thyroidectomy Vagotomy Bile duct stricture repair

3 2 G&S G&S G&S G&S G&S 4 4 G&S 6 2 G&S G&S 3

findings (these elements comprise a group and screen); • identification of the specificity of any alloantibody detected in the antibody screen; • selection of blood of an appropriate blood group or extended phenotype; • a serological or electronic crossmatch; • labelling of the blood with the recipient’s identifying information. Detection of red cell antigen–antibody reactions

The phenotyping (‘blood grouping’) of red cells and the detection of red cell alloantibodies depend upon interpretations of serological interactions between red cell antigens and antibodies. Various serological methods and test systems are available to demonstrate these interactions, and these must be optimized in order to obtain the appropriate sensitivity and specificity for their intended clinical use. Failure to follow the instructions provided by reagent manufacturers can lead to incorrect conclusions. Test methods have been developed to allow the detection of antibodies of different isotypes. Antibodies that have specificities for red cell antigens

Blood transfusion in hospitals

are usually IgG or IgM. IgM antibodies are pentameric molecules that can cross-link between antigens on adjacent cells, thus causing direct agglutination of red cells. Conversely, IgG antibodies are monomeric, and although they are divalent, the distance between the Fab regions on a single IgG molecule is in general insufficient to allow direct agglutination to take place, as there are stronger intercellular repulsive forces between red cells at these distances. Methods such as the indirect antiglobulin test (IAT) (which uses a secondary antibody; see Fig. 25.2) or the enzyme method (which uses proteolytic enzymes such as papain to cleave negatively charged, hydrophilic residues from the red cell membrane) must therefore be used to detect most IgG red cell antibodies. Test systems for the detection of serological reactions can be classified into three broad categories. Liquid-phase systems

Liquid-phase systems rely on the visualization of haemagglutination reactions in individual glass/plastic tubes or 96-well microplates. The presence or absence of agglutinated red cells distin-

guishes positive and negative reactions, allowing the grading of reaction avidity according to the strength of haemagglutination. These liquid-phase systems are more commonly used for blood grouping than for antibody screening, as IgM blood grouping reagents can be used in simple, rapid, direct agglutination methods. IAT methods using red cells suspended in a low ionic strength solution remain the gold standard for the detection of clinically significant red cell alloantibodies, although these methods require meticulous attention to procedure, in particular during the washing phase to remove unbound IgG. Column-agglutination systems

The introduction of column-agglutination systems during the last decade has resulted in a very significant change to routine laboratory practice in the UK. One of these systems, first described by Lapierre in 1990, uses a plastic card containing six channels, each of which contains a mixture of Sephadex and Sephacryl gels. This gel mixture is formulated to allow the passage of unagglutinated cells but not of agglutinated red cells. Positive reactions are therefore distinguished by agglutinates at

Red cells incubated with serum containing IgG antibody

Unbound antibody washed away. Anti-human globulin added to precipitate cells

Fig. 25.2 Indirect antiglobulin test.

291

Chapter 25

or near the top of the gel column and negative reactions appear as buttons of red cells at the bottom (Fig. 25.3). A similar column-agglutination method involving a glass microbead density barrier in place of a gel is also available. Reagent (IgM) antibody can be incorporated into the gel or bead columns, allowing phenotyping to be undertaken simply by the addition of test cells to the top of the column. Similarly, the IAT can be performed in columns containing antiglobulin reagent to which plasma and reagent red cells are added. Because plasma proteins are less dense than the gel, a washing phase is not needed. This property, and the relative stability of the reaction end point, give column agglutination methods a degree of simplicity and reliability not achieved by other methods. Solid-phase sytems

These systems are based on 96-well microplates and provide another alternative to the tube or column IAT. Although differing in detail, all these methods achieve a positive reaction end point that is characterized by a monolayer of red cells across the surface of the microplate well. A discrete button of red cells at the bottom of the well indicates a negative reaction (Fig. 25.4). Solid-phase systems suffer from the disadvantage that they require carefully standardized centrifugation and

washing steps; however, unlike the situation with liquid-phase test systems, fully automated equipment allows these steps to be performed safely and consistently without operator intervention. Reduction of error in pretransfusion compatibility testing

An analysis of cases in the UK during 1996–2002 where blood components had been incorrectly transfused showed that laboratory errors were implicated in 28% of incidents. Many of these laboratory errors were due to the transposition of samples or to simple human error in the setting up or interpretation of tests. Provided that the correct laboratory identifier (as a barcode) is placed on the patient’s sample, these errors can be avoided by using a fully automated system that is interfaced to the blood transfusion laboratory computer. The basic features of a fully automated (‘walk away’) system should include: • trays or carousels to stack samples; • automated liquid handling and other robotic operations; • devices to ensure that positive sample identification is maintained; • clot sensor and liquid level alarms; • an optical device to record reaction patterns; and • a comprehensive system management software

Fig. 25.3 Column-agglutination

technology. Positive results are seen in the first and last columns.

292

Blood transfusion in hospitals

Fig. 25.4 Solid-phase technology.

that interprets reaction patterns and flags discrepant results. Automated systems can utilize solid-phase/microplate or column-agglutination technology. Where full automation is not available, or cannot be used for the whole process (as is currently the case with antibody identification), steps must be taken to minimize the occurrence of error or minimize its impact. A high standard of training, participation in internal and external quality assessment schemes, and strict adherence to validated documented procedures are among the measures that reduce the occurrence of human error. The most critically important procedure in the blood transfusion laboratory, the determination of the ABO group, should be performed by two people (except in urgent situations) if there is no record of a grouping result from a previous sample. Similarly, the determination of a D group should be performed in duplicate, in the absence of full automation. ABO and D grouping

• The patient’s red cells should be tested against monoclonal anti-A and anti-B grouping reagents. • The patient’s serum/plasma should be tested against A1 and B reagent red cells, except in neonates. • The expected reaction patterns in ABO grouping are illustrated in Table 25.9.

Table 25.9 ABO grouping patterns.

Group

Anti-A

Anti-B

A1 cells

B cells

O cells

A B O AB

+++ +++

+++ +++

+++ +++ -

+++ +++ -

-

• The patient’s red cells should be tested with an IgM monoclonal anti-D reagent, which does not detect the ‘partial D’ group, DVI. • ABO and D groups must be repeated when a discrepancy (anomaly) is found. Repeats should be performed using a fresh suspension of washed cells. An autocontrol should be included. Antibody screening

• The IAT performed at 37°C is the best method available for the detection of red cell antibodies of clinical importance. It is simple (especially when using a column-agglutination system), has an appropriate level of sensitivity and has a high degree of specificity. • The patient’s serum/plasma should be tested against two or more ‘screening cells’ using the IAT. • The reagent red cells used for screening should between them express antigens reactive with all clinically significant antibodies; ideally the 293

Chapter 25

phenotypes R1R1 or R1wR1 and R2R2 should be represented in the screening cell set. It is recommended that the following phenotypes should also be represented: Jk(a+b–), Jk(a–b+), S+s–, S–s+, Fy(a+b–), Fy(a–b+). • Antibody screening is the most reliable and sensitive method of detecting a clinically significant antibody, since stronger reactions may be obtained with cells having homozygous expression of the antigen and the red cells are preserved in a medium to minimize loss of antigens during the storage period. • Antibody screening performed in advance of the requirement for transfusion also provides the laboratory with time to identify the specificity of the antibody and, when clinically significant, to select antigen-negative units for crossmatching. Antibody identification

• When an antibody has been detected in the screening test, the specificity should be determined by testing the patient’s serum/plasma against a panel of reagent red cells of known phenotypes. • In addition to the IAT, other methods (e.g. using enzyme-treated red cells) may be helpful, particularly when a mixture of antibodies is present. • The specificity of the antibody can be determined when the serum/plasma is reactive with at least two examples of red cells bearing the antigen and non-reactive with at least two examples of red cells lacking the antigen. • When one antibody specificity has been determined, it is essential that additional clinically significant antibodies are not overlooked. Multiple antibodies can only be confirmed by choosing red cells that are antigen negative for the recognized specificity but positive for other antigens to which clinically significant antibodies may arise. Autoantibodies

These may be suspected when the patient’s serum/plasma reacts with all cells used in the reverse ABO group or with all cells in the antibody identification panel including the patient’s own red cells. Not all autoantibodies give rise to haemolysis. 294

Serological investigations should focus on obtaining the correct ABO and RhD group of the patient and on excluding the presence of underlying alloantibodies. Cold-type autoimmune haemolytic anaemia

• The red cells should be washed at 37°C for performing the direct antiglobulin test (DAT), which will usually be strongly positive due to coating with C3d. • Underlying alloantibodies can be excluded by using cells and serum separately warmed to 37°C. Warm-type autoimmune haemolytic anaemia

• The red cells will usually have a positive DAT due to coating with IgG with or without complement. Rarely, the red cells may be coated with IgA or IgM and IgG. • Underlying alloantibodies may be detected following the removal of autoantibodies from the patient’s serum. This may be achieved by either absorbing the serum with the patient’s own red cells (e.g. using a combination of papain and dithiothreitol ‘ZZAP’ to elute the autoantibody and enzyme treat the red cells) or, if the patient has been recently transfused, with red cells of similar phenotype (if already known) or with two or more examples of red cells of known phenotypes. Selection of red cells for transfusion

• Red cells of the same ABO and D group as the patient should be selected, except in a lifethreatening situation before the patient has been grouped. In this case, group O should be used; if the patient is a premenopausal female, group O, D negative should be used. Group-specific units should be provided as soon as the patient’s group is known. • The selection of blood for patients with red cell alloantibodies is summarized in Table 25.10. However, in life-threatening situations, the immediate need for red cell transfusion may necessitate the use of incompatible units. • Units for fetal or neonatal exchange transfu-

Blood transfusion in hospitals Table 25.10 Recommendations for the selection of blood for patients with red cell alloantibodies.

Typical examples

Procedure

Antibodies that could be considered clinically significant

Anti-D, -C, -c, -E, -e Anti-K, -k Anti-Jka, -Jkb Anti-S, -s, -U Anti-Fya, -Fyb

Select ABO-compatible, antigen-negative blood for serological crossmatching

Antibodies directed against antigens with an incidence of < 10%, and where the antibody is often not clinically significant

Anti-Cw Anti-Kpa Anti-Lua Anti-Wra (anti-Di3)

Select ABO-compatible blood for serological crossmatching

Antibodies primarily reactive below 37°C, and never or only very rarely clinically significant

Anti-A1 Anti-N Anti-P1 Anti-Lea, -Leb, -Lea+b Anti-HI (in A1 and A1B patients)

Select ABO-compatible blood for serological crossmatching, performed strictly at 37°C

Antibodies sometimes reactive at 37°C and clinically significant

Anti-M

If reactive at 37°C, select ABO-compatible, antigen- negative blood for serological crossmatching If unreactive at 37°C, select ABO-compatible blood for serological crossmatching, performed strictly at 37°C

Other antibodies active by IAT at 37°C

Many specificities

Seek advice from blood centre

IAT, indirect antiglobulin test.

sions should be selected to be compatible with the maternal serum/plasma. • Premenopausal females should ideally receive K-negative red cells, and R1R1 units if they are c-negative. • Patients with a lifelong dependency on red cell support should receive red cells matched for Rh antigens and K (see Chapter 9). • Recipients of ABO- and D-incompatible allogeneic haemopoietic stem cell grafts will need to be transfused with red cells of the donor group in the case of a minor ABO mismatch, or group O in the case of a combined ABO mismatch. D-negative red cells should also be selected for D-positive recipients of a D-negative donation (see Chapter 9). Serological crossmatch

Crossmatching techniques have been simplified in recent years, and only the immediate spin crossmatch and the IAT crossmatch remain in common use.

The IAT crossmatch can be abolished when antibody screening is performed with screening cells that share apparent homozygous expression of common antigens capable of stimulating clinically significant antibodies, and the patient’s serum/plasma has never been found to contain clinically significant antibodies. An exception to this rule should be made for patients who have received an ABO-incompatible solid organ transplant and who may develop an IgG anti-A or antiB produced by passenger lymphocytes. Several retrospective and prospective studies have shown that there is negligible risk in omitting the IAT crossmatch. Although up to 0.2% IAT crossmatches may reveal an unpredicted incompatibility, few of these transfusions result in haemolysis. Antibodies directed against low-frequency antigens may be missed but the majority of these are naturally occurring and do not cause patient morbidity. If the IAT crossmatch is omitted, there must be some check included to detect ABO incompati295

Chapter 25

bility. The immediate spin crossmatch (i.e. agglutination in saline following a 2–5 min incubation) is a serological check that can be used. However, this technique is fallible when the patient has low levels of anti-A or anti-B and, unless ethylenediamine tetra-acetic acid (EDTA) saline is used, falsenegative results may also arise as a result of steric hindering of agglutination by C1. False-positive crossmatches arising from rouleaux or cold agglutinins not detected in the antibody screen have the potential to delay the issuing of compatible units. The limitations of the immediate spin crossmatch have heralded the acceptance of electronic issue as an alternative method of preventing the release of ABO-incompatible units of blood. Electronic issue

Electronic issue should only be used for detecting ABO incompatibility between the donor unit and the patient sample that was submitted for pretransfusion testing. There are several essential requirements for adopting this approach, which are common to the various professional standards. • The computer contains logic to prevent the assignment and release of ABO-incompatible blood. • No clinically significant antibodies are detected in the recipient’s serum and there is no record of previous detection of such antibodies. • There are concordant results of at least two determinations of the recipient’s ABO and D groups on file, one of which is from a current sample. • Critical elements of the system (application software, readers and interfaces) have been validated on-site and there are mechanisms to verify the correct entry of data prior to release of blood, such as the use of barcode identifiers to enter information when it cannot be automatically transferred. Fully automated blood grouping and antibody screening, although not a requirement in current UK guidelines, is strongly recommended. Electronic issue has been routinely practised in Sweden for over 10 years, during which time one error has been noted due to an incorrectly labelled unit of blood, which was supplied by a small noncomputerized blood centre. 296

Electronic issue has several potential advantages: • reduced technical workload; • rapid availability of blood; • improved blood stock management and reduced wastage; • less handling of biohazardous material; • elimination of unwanted false-positive results in the immediate spin crossmatch; and • ability to issue blood electronically at remote sites, using trained non-laboratory staff. This last characteristic has allowed the development of networked electronic blood release systems. When the patient details are entered, the system checks that the criteria for electronic issue are fulfilled and lists the compatible units available in the remote site blood refrigerator. The barcodes of the unit selected are scanned into the computer and, if ABO and D compatible, a compatibility label is printed that is attached to the unit and rescanned. This step generates a second label, or compatibility form, which is signed by the clinical staff at the time of the transfusion.

Summary The transfusion process is unique for several reasons: • it links one sector of the community (the donors) with another (the patient) in an altruistic, potentially life-saving activity; • it links many grades of staff across a healthcare organization; and • for many patients, there is still no substitute for human-derived blood components. Prescribers of blood components have a moral obligation to the donors to ensure that the donations are used appropriately. Prescribers of blood components also have a duty of care to their patients to ensure that the benefits of the transfusion outweigh the risks. There are many different members of staff involved in the hospital transfusion process, which provides too many opportunities for human error to prevail. Investment in a quality infrastructure, computerization and automation is essential to prevent errors in this process.

Blood transfusion in hospitals

Further reading Audet AM, Greenfield S, Field M. Medical practice guidelines: current activities and future directions. Ann Intern Med 1990; 113: 709–14. Department of Health. Better Blood Transfusion. HSC 2002/009. London: HMSO, 2002. Dzik WH, Corwin H, Goodnough LT et al. Patient safety and blood transfusion: new solutions. Transfus Med Rev 2003; 17: 169–80. Eisenstaedt RS. Modifying physicians’ transfusion practice. Transfus Med Rev 1997; 11: 27–37. Handbook of Transfusion Medicine, 3rd edn. London: The Stationary Office, 2001 (www.transfusionguidelines.org). Heddle NM, O’Hoski P, Singer J, McBride JA, Ali MA, Kelton JG. A prospective study to determine the safety of omitting the antiglobulin crossmatch from pretransfusion testing. Br J Haematol 1992; 81: 579–84. Jensen NJ, Crosson JT. An automated system for bedside verification of the match between patient identification and blood unit identification. Transfusion 1996; 36: 216–21. Judd WJ. Requirements for the electronic crossmatch. Vox Sang 1998; 74 (Suppl. 2): 409–17. Mollison PL, Engelfriet CP, Contreras M. Detection of the

reaction between red cell antigens and antibodies. In: Blood Transfusion in Clinical Medicine. Oxford: Blackwell Science, 1997: 242–76. Research Unit of the Royal College of Physicians. Audit Measures for Good Practice in Blood Transfusion Medicine. London: RCP Publications, 1995. Serious Hazards of Transfusion. Annual Report 2002–2003. Manchester: SHOT Office, 2004 (www.shotuk.org). Turner CL, Casbard AC, Murphy MF. Barcode technology: its role in increasing the safety of blood transfusion. Transfusion 2003; 43: 1200–9.

Guidelines British Committee for Standards in Haematology. Guidelines for compatibility procedures in blood transfusion laboratories. Transfus Med 2004; 14: 59–73. British Committee for Standards in Haematology. The administration of blood and blood components and the management of transfused patients. Transfus Med 1999; 9: 227–38. British Committee for Standards in Haematology. Guidelines for blood bank computing. Transfus Med 2000; 10: 307–14.

297

Chapter 26

Autologous transfusion Dafydd Thomas

Autologous transfusion is the collection of the patient’s own blood for reinfusion. This can be by prior collection either in the weeks before or immediately prior to surgery. Alternatively the blood loss during or after surgery can be collected and reinfused. These methods can be summarised as: • intraoperative cell salvage (ICS); • postoperative cell salvage (PCS), which may be washed or unwashed on reinfusion; • predeposit autologous donation (PAD), taken and stored in the weeks before surgery; • acute normovolaemic haemodilution (ANH), taken just before surgery and returned following the operation.

Reasons to consider autologous transfusion Clinical transfusion practice includes an attempt to reduce the risks involved in blood transfusion. This may be achieved in part with extensive testing of the blood to decrease the risk of transfusiontransmitted infection. Improvements in the safety of blood over the last 20 years (Fig. 26.1) has ensured there continues to be considerable demand for allogeneic blood. Blood conservation strategies should be adopted to minimize the use of allogeneic blood by withholding transfusion until strictly clinically necessary, and employing techniques such as autologous transfusion. In some situations autologous transfusion is definitely indicated, such as patients with very rare blood groups or complex red cell antibodies for whom it is difficult to find compatible blood. Autologous transfusion should be also used instead of, or to supplement, allogeneic blood in situations where it 298

has been shown to be effective and safe. It has been suggested that more than 20% of surgical demand can be met by autologous transfusion. There are certain procedures that can be undertaken with virtually no allogeneic blood use, allowing conservation of supplies for areas of medicine where there are currently few alternatives to allogeneic blood, such as haematological oncology. The provision of comprehensive testing of donor blood and other safety measures have led to a significant increase in cost. The escalating costs, and fears over a continued adequate supply of donor blood have led to a greater interest in autologous transfusion. Considerable experience has been gained in the various methods of autologous transfusion elsewhere in the world particularly in North America. The autologous transfusion procedures which seem to offer the most cost-efficient and clinically effective method in the current situation are ICS and PCS.

Blood conservation strategies Avoidance of unnecessary blood transfusion includes optimizing the preoperative haemoglobin (Hb) concentration, minimizing blood loss, and being clear about when and how much to transfuse (see Chaper 6). Autologous transfusion options are at their most effective when combined with these other strategies (Table 26.1). Before surgery: optimizing Hb and haemostasis

This involves assessing a patient in advance of surgery and taking steps to reduce the patient’s requirements for transfusion. If a patient is

Autologous transfusion

Risk of infection per unit transfusion

1/100

HCV HBV HIV

1/1000

1/2000

1/10 000 1/58 000 to 1/149 000

1/100 000

1/872 000 to 1/1.7 ¥ 106

Nucleic acid testing for HCV/HIV

Testing for p24 antigen

Screening for HCV-antibody

Screening for HIV-antibody Surrogate screening for non-A, non-B hepatitis

of blood over the last 20 years. HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus. (From Goodnough et al. 2003 with permission.)

1/1.4 ¥ 106 to 1/2.4 ¥ 104

19 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 01

Fig. 26.1 Improvements in the safety

Donor screening criteria changed

1/1 000 000

Years

Table 26.1 An approach to blood conservation and the reduction of risk associated with blood transfusion in patients having

elective surgery. Some principles apply to other situations where blood transfusion is being considered. • • • • • •

Check the blood count well in advance of surgery and correct any treatable anaemia Ask about any drugs the patient is taking and consider if any should be stopped Check antibody status and blood group so group-specific blood can be used in an emergency Consider whether patient has a hereditary or acquired bleeding tendency and investigate/treat as appropriate During surgery consider ways to reduce bleeding, e.g. the use of lasers or the use of aprotinin in cardiac surgery Postoperatively, consider whether blood transfusion is clinically indicated (transfusion trigger) and, if it is, consider how many units are required to achieve the desired Hb (transfusion target) • If operation would normally require blood transfusion, consider the option of autologous blood transfusion.The options are listed below Technique

Situations in which it might be considered

Intraoperative cell salvage

Any patient (generally without infection or malignancy in the operative field) with blood loss > 0.5 L; especially suitable for massive blood loss Patients with postoperative blood loss from a clean site Suitably fit patients expected to lose a moderate amount of blood (1–2.5 L) Suitably fit patients expected to lose more than 20% blood volume (~ 2 L)

Postoperative cell salvage Predeposit autologous transfusion Acute normovolaemic haemodilution

anaemic, it is important to consider the reasons and address the underlying cause, e.g. iron deficiency may be due to a gastrointestinal malignancy. If a patient with iron deficiency anaemia is started on iron, the Hb can be expected to rise by about 1 g/dL per week. It is therefore important to check the blood count sufficiently far in advance of surgery to allow time for treatment to be given if required. Improved preparation of patients for elective surgery may ensure that correctable anaemia is treated prior to surgery. Patients presenting with a normal or near-normal Hb will

require transfusion at a later stage or may even avoid blood transfusion altogether. Recent evidence suggests that the newer preparations of intravenous iron can be given with fewer adverse reactions than were associated with these preparations in the past. Intravenous iron is almost immediately available for red cell production. Research is beining undertaken to determine whether administration of intravenous iron as late as the preoperative day can improve red cell production in response to surgical anaemia and thus decrease the use of donor blood. 299

Chapter 26

It is also important to consider if there are any patient factors which might cause excessive blood loss during surgery and which can be corrected in advance. The following list outlines some of the preoperative measures that may need attention. • Drugs: if patients are on aspirin or clopidogrel, it is usually advisable to stop them 7 days before surgery (except if the patient is at high risk of suffering a myocardial infarction). Patients on warfarin can often stop a few days before surgery. If it is imperative to continue anticoagulation, the patient may be changed to heparin to cover the surgical period. • Inherited disorders of coagulation: it is important to take a bleeding history when the patient is being seen prior to surgery. Screening tests and specific treatment may be required (see Chapter 9). • Bleeding diathesis: important to consider in patients with renal or liver disease. Agents such as desmopressin (DDAVP) or tranexamic acid may decrease the bleeding tendency temporarily (see Chapter 7). Expert advice from a haematologist should be sought. During surgery: reduction in blood loss

With advancing surgical and anaesthetic techniques, blood loss in many operations has fallen significantly. Use of lasers and microsurgical techniques have had a huge impact on blood usage. • Maintenance of normothermia allows optimum coagulation perioperatively and has been shown to decrease blood loss. • Cardiac surgery: patients on bypass are at particular risk of bleeding. Aprotinin, an inhibitor of fibrinolysis, has been shown to reduce blood loss, particularly in high-risk operations, and does not appear to increase the reinfarction rate (see Chapter 7). During/after surgery: when to transfuse

No blood transfusion is without risks, but equally the administration of blood may be life-saving. In making the decision to transfuse the balance of risks must be considered for each individual patient. Factors influencing the decision to transfuse include: 300

• Hb concentration; • the patient’s life expectancy, i.e. age/prognosis (many of the adverse effects of transfusion are delayed, e.g. transfusion-transmitted infections often have a long latent period before the patient becomes symptomatic); and • clinical judgement about the patient’s ability to tolerate anaemia including the presence of other factors such as cardiac and respiratory disease and sepsis.

Transfusion triggers Data from patients who refuse blood on religious grounds or who live in parts of the world where blood is scarce or dangerous have helped our understanding of the effects of different levels of Hb. In otherwise healthy patients the following transfusion triggers for stable anaemia might be considered: • <4 g/dL: transfuse unless fit and Hb rising; • 4–7 g/dL: transfusion usually necessary; • 7–10 g/dL: transfusion not usually necessary; • >10 g/dL: transfusion rarely required. A study of patients in intensive care showed that less severely ill patients (Acute Physiology and Chronic Health Evaluation II score £20) and patients under 55 years actually had a survival advantage if the Hb was maintained between 7 and 9 g/dL rather than between 10 and 12 g/dL. For patients with clinically significant cardiac disease the mortality was similar in both groups. For otherwise fit patients with a previously normal Hb who are actively bleeding the following guidelines might be applied. • Blood loss <15% blood volume: give fluids, no need to transfuse. • Blood loss 15–30% blood volume: consider transfusion. • Blood loss 30–40% blood volume: transfusion usually necessary. • Blood loss >40% blood volume: transfusion indicated. Note that blood volume is about 70 mL/kg in adults, and that 20% of blood volume is approximately 1 L. For patients with a short life expectancy or those

Autologous transfusion

with chronic anaemia and impaired red cell production, the main trigger for transfusion should be the patient’s symptoms.

Transfusion targets In addition to considering when to transfuse, a target Hb should be established for each clinical scenario using the best data available (see above and Chapter 6), and it is also important to consider how many units to give. When a patient is actively bleeding, replacement of red cells should be guided by an estimate of blood loss. A guide to how many units are required to achieve the target is shown in Table 26.2. Oneunit transfusions have previously been discouraged. However, Table 26.2 shows that it might be reasonable to give a single unit to a small elderly woman who is symptomatic with an Hb of 7 g/dL to bring it up to just under 9 g/dL. The transfusion of blood just because it has been made available for the patient should be avoided. If blood is not used, it can be returned to the blood bank and used for another patient.

tion of a single unit of red cells may be enough to raise a patient’s Hb to avoid donor blood transfusion. After much debate, red cell salvage now seems to offer the most cost-effective method of autologous transfusion. Future issues in blood supply and demand combined with the discovery of other blood-borne diseases may change this view. A full description of all the autologous methods has therefore been retained in this chapter. All autologous blood must be clearly labelled and be distinct from allogeneic blood. An example of an autologous blood label is shown in Fig. 26.2; they are more easily identified if printed a different colour to the ones used for allogeneic blood. Perioperative cell salvage Principle

During surgical operations when blood loss is expected, blood can be collected, processed and then returned to the patient. This can be done SWANSEA NHS TRUST AUTOLOGOUS TRANSFUSION HOSPITAL NO.

Techniques for providing autologous blood The measures for avoidance of blood transfusion described above, when combined with some form of autologous technique, can maintain the Hb at a clinically acceptable level, sufficient to avoid the transfusion of allogeneic blood. Even the collec-

Table 26.2 Guide to number of units required to achieve the

NAME

DOB OPERATOR

SIGNATURE

DATE

TIME

EXPIRY DATE/TIME

‘target’ haemoglobin (Hb).

Amount of Hb in 1 unit of red cells Example: volume bled 450 mL ¥ average Hb concentration 13 g/dL = 58 g/unit

HOSPITAL NO.

Weight (kg) 43 57

71

3L 1.9

5L 1.2

AFFIX IN TRANSFUSION RECORD DOB

NAME LD. CHECKED BY

Blood volume (70 mL/kg in adults) Increase in Hb after one unit transfusion (g/dL)

4L 1.6

ANH

WITNESS PHE DONATION

WASHED

DATE UNWASHED

Fig. 26.2 Autologous colour-coded labels.

301

Chapter 26

either intraoperatively or postoperatively depending on the type of operation. This process can be cost-effective even when even small volumes of blood (i.e. more than 500 mL) are collected. The amount salvaged not only decreases the use of allogeneic blood but in many instances completely removes the need for allogeneic blood transfusion. Intraoperative cell salvage

ICS involves the collection and reinfusion of red cells lost during surgery. This may be performed as follows. • Single-unit reinfusion devices (contain anticoagulant, unless defibrinated blood is being collected): these are simple and cheap for low-volume losses. • Continuous reinfusion of unprocessed blood: may be used in conjunction with cardiac bypass but is not of proven benefit and may be associated with risk of haemolysis. • Reinfusion of processed blood (discussed below).

There are a number of machines available which wash red cells by centrifugation and resuspend them in saline (an example is shown in Plate 26.1). Blood is aspirated from the wound site and mixed with anticoagulant (Fig. 26.3) before it enters the reservoir of the machine (Plate 26.2) (Plates 26.1 and 26.2, shown in colour between pp. 304 and 305). The cycle can be either run automatically or controlled manually. In general about 75% of red cells can be recovered for reinfusion back into the patient. The machines can deliver the equivalent of 10 units of blood per hour. Advantages of ICS • Considerable reduction in allogeneic blood usage in cases where blood loss is large (>2 L). Suitable operations might include open heart surgery, liver transplantation and ruptured ectopic pregnancy. • Available to all patients having appropriate surgery regardless of medical fitness. • In some situations of uncontrolled blood loss it may be life-saving.

Anticoagulant

Blood from site of operation

Reservoir Reinfusion bag

Wash solution

Pump

Waste bag

Centrifugal bowl

302

Fig. 26.3 Complete cell saver set-up.

Autologous transfusion

• Unlike other techniques, ICS can be used selectively in cases where the actual, rather than the predicted, blood loss is high. • Blood can be collected in the reservoir and the decision to use the machine and harness can be deferred until it is clear that the blood loss is sufficient to warrant processing. • Unlike PAD and ANH, in which blood is stored outside the body, cell salvage may be accepted by Jehovah’s Witnesses, provided the collected blood remains in continuity with the patient via tubing which is connected to the patient’s intravenous cannula and hence the patient’s circulation. Disadvantages/risks of ICS • The reinfusion of haemolysed blood is unlikely, providing the wash process is undertaken correctly. Currently available machines operate an automatic washing process, and a sensor monitors the effluent from the wash cycle, continuing the wash process until the liquid being discarded is completely clear, suggesting removal of all free Hb and fragmented red cells. • There have been no recent deaths associated with air embolism as a result of improved design and greater awareness of such potential problems. Deaths were associated with air embolism in the earlier machines and collected blood should not be used with pressurized reinfusion. • It does not recover all the blood lost so allogeneic blood is often required if major haemorrhage has occurred. • It requires a capital outlay and trained operators, so ICS can only be used in hospitals with suitable operations to become cost-effective. As the cost of allogeneic red cells continues to rise with the introduction of safety measures such as universal leucocyte depletion of blood and increasingly sensitive and expensive microbiology testing, cell salvage has become cost-effective for many hospitals. • It is very important to follow agreed standard operating procedures and to document all stages of the process to avoid mistakes being made. Indications Operations where expected blood loss is likely to be in excess of 500 mL. Even when blood loss is

unpredictable, the collection of operative blood loss may be worthwhile. Providing this blood is anticoagulated, it can be processed and reinfused if sufficient volumes are collected. The processing kits are separately packaged so only the collection jar is wasted if small volumes are salvaged with a decision not to wash. It is cost-effective providing 1 unit of packed red cells is produced and used. Relative contraindications There are a number of situations where the use of cell salvage has been discouraged. However, in the presence of massive haemorrhage, ICS may be considered as the possibility of hypovolaemic shock and perhaps death becomes part of the risk–benefit equation. • Potential for aspiration of malignant cells: although leucocyte filters may remove the majority of malignant cells, and small numbers may not be clinically significant compared with the numbers that enter the circulation during surgery, some would advocate the use of gamma-irradiation in this setting but this is logistically difficult to arrange. • Presence of infection: although the balance of risk depends on the clinical urgency for salvaged blood, antibiotics may be added to the anticoagulant solution used and given parenterally to the patient to minimize any bacteraemia that occurs. • Presence of ascitic or amniotic fluid in the operative field, which may cause embolism/disseminated intravascular coagulation. Filtering of the salvaged blood via a leucocyte filter removes all lamellar bodies and decreases the risk of amniotic fluid embolism. • Sickle cell disease: cells may sickle in the low oxygen tension of the machine. • Where topical clotting agents have been used, e.g. fibrin glue, or iodine has been used to wash out the abdomen: in practice these contaminants haemolyse red cells. Even if these agents are collected they are washed out during the centrifugal process because of their low molecular weight relative to the red cells, and therefore not reinfused. Postoperative cell salvage

Postoperative recovery of blood involves the 303

Chapter 26

collection of blood from surgical drains followed by reinfusion with or without processing. The blood recovered is dilute, partially haemolysed and defibrinated and may contain high levels of cytokines. If the collected wound drainage blood is simply reinfused, some centres limit the quantity reinfused. Other centres recommend that all blood is washed and resuspended in saline. This can either be done with the apheresis machines used in the main theatre suite or with the newer and more compact processing machines that wash collected blood by the patient’s bedside (Plate 26.3, shown in colour between pp. 304 and 305). Its most common application is the collection of blood after the tourniquet is removed at the end of knee surgery. In this setting it may help avoid the need for allogeneic blood in the majority of patients undergoing such procedures. Acute normovolaemic haemodilution Principle

Whole blood is taken into standard blood bags containing citrate anticoagulant, either just before or during induction of anaesthesia. The volume removed is simultaneously replaced with crystalloid or colloid so that the patient remains normovolaemic. The amount of blood withdrawn depends on the target haematocrit and can be calculated using standard formulas. By this process the Hb that remains is diluted so that if the patient then bleeds a given volume, the total quantity of Hb lost is not as great as if the concentration had been higher. At the end of the procedure, or before if necessary, the blood removed at the start is returned to the patient. ANH is only really beneficial if significant haemodilution is achieved and the blood loss is large, over 20% of the total blood volume. Safety

In the hands of experienced anaesthetists, ANH has been shown to be safe when performed on patients who are fit and closely monitored. It requires the establishment of, and adherence to,

304

standard operating procedures. As microbiology testing is not performed, the units should be treated as high risk and labelled ‘untested blood for autologous use only’. Units should remain with the patient and not be put in a blood refrigerator where they could be used for the wrong person. Any unused autologous blood should be disposed of as hazardous waste. All stages of the process should be carefully documented. As for all autologous transfusions, any serious events should be reported to the hospital transfusion committee and to the Serious Hazards of Transfusion (SHOT) scheme. Efficacy

Studies of ANH have generally been small or retrospective and no conclusive reduction in the use of donor blood has been demonstrated (see Chapter 6). Although it seems to be less effective at reducing the need for red cell transfusion, studies have not looked at other potential benefits. This blood may supply coagulation factors and platelets which have not been diluted within the intravascular compartment during surgery or been exposed to the stress response. At the UK conference ‘Autologous transfusion: 3 years on. What is new? What has happened?’ in November 1998, it was concluded that ‘randomised controlled trials are required before this technique can be widely recommended’. The few studies published on the topic since that meeting have not provided strong support for the technique. Indications

Guidelines from the British Committee for Standards in Haematology (BCSH) for perioperative haemodilution suggest than ANH can be considered when: • the likely surgical blood loss exceeds 20% of the patient’s blood volume; • preoperative Hb is more than 11 g/dL; • patients do not have severe myocardial disease ( patients >45 years should be assessed with caution).

(a)

(b)

Mix thoroughly and incubate at 22°C for 30 min

Add 50 mL FITC labelled anti-IgG or IgM

1000

64

1:

(i)

(ii)

10 1

1

Hist Region

Count

2

B % +ve 1.9 C x Ch 100

90.6 10 000 190 10 000

1

10 100 log F1-1 (FITC)

Mean X

Mean Y

PK PK PK Pos X Pos Y Cnt

27.3

7.33

21

3.29 0.262

64 48

32

6.6

B

HPCV FPCV

10 1.15 51.7 4344 0.000 106.2

64

Anti-HPA -1a (NAIT)

2:A

48

C

32 16 0 0.1

1 10 100 1000 log F1-1 (FITC)

1000

133

2.3 0.10

C

0 0.1

B

0 0.1

1000

AB serum

2:A

16

C

32

Count

Count

%

A autoa

48

(iv)

10 100 ss log

1

64

2:A

16

0.1 0.1

(iii)

Stop reaction Read in flow cytometer or wash x 2 in PBS/BSA and read microscopically

48 Count

FS log

100

Wash x 4 in PBS/BSA

Count

(c)

Mix thoroughly and incubate at 37°C for 30 min

Add 50 mL serum

50 mL platelet suspension in PBS

(v)

Plate 5.1 Indirect platelet immunofluorescence test (PIFT).

(a) Outline of assay. (b) Results of microscopic analysis of PIFT showing a strongly positive reaction. (c) Results of flow cytometric analysis of PIFT: (i) the platelet population is identified from forward/side scatter characteristics; (ii) the gated population is analysed for fluorescence intensity;

B

Anti-HPA -1a (PTP)

2:A C

32 B

16 0 0.1

1 10 100 1000 log F1-1 (FITC) (vi)

1

10

1

10

log F1-1 (FITC)

(iii) results may be expressed as percentage positive cells within a region (Region B, ‘% + ve’) or as mean channel fluorescence (Region C, ‘x Ch’); (iv)–(vi) plots of fluorescence intensity versus number of events for (iv) a negative sample, (v) a sample containing a weak anti-HPA1a and (vi) a potent anti-HPA-1a.

1

Plate 8.1 Acid elution technique (Kleihauer test) for

haemoglobin F containing cells; the blood specimen was taken from a postpartum woman and shows that a fetomaternal haemorrhage had occurred. A single stained fetal cell is seen against a background of ghosts of maternal cells. From Bain (1995).

Plate 20.2 Section through the brain of a patient with variant CJD with immunohistochemical staining for PrP demonstrating abnormal accumulation of PrPSc throughout the brain. (Reproduced with the permission of Professor James Ironside.)

2

Plate 20.1 Section through the brain of a patient with variant CJD demonstrating spongiform degeneration of neuronal tissue and a florid amyloid plaque (centre). (Reproduced with the permission of Professor James Ironside.)

Plate 20.3 Section through the lymphoid tissue of a patient with variant CJD with immunohistochemical staining for PrP demonstrating abnormal accumulation of PrPSc in follicular dendritic cells. (Reproduced with the permission of Professor James Ironside.)

Plate 26.1 A centrifuge bowl within an apheresis machine

Plate 26.2 Collection reservoir which may be used either

showing the dense red cell layer towards the outside of the bowl and separation from the buffy layer and plasma.

operatively or postoperatively to collect the spilt blood or wound drainage.

Plate 26.3 Equipment is now available that can wash

salvaged blood in the ward environment.

3

R1

R2

a)

Region Statistics (a and b) (Gate: No Gate) Region Events % Gated R1 27491 96.46 R2 27674 97.10

b) Quadrant Statistics (c) (Gate: G2) Quad Events % Gated UL 10 0.04 UR 434 1.57 LL 561 2.03 LR 26669 96.37

c)

d)

Plate 32.1 Haemopoietic stem cell (CD34 cell) +

enumeration in a mobilized apheresis sample using International Society for Hematotherapy and Graft Engineering guidelines. Fluorochrome-conjugated antiCD45 antibodies (fluorescein isothiocyanate, FITC) and anti-CD34 antibodies (phycoerythrin, PE) are used in this flow cytometric assay. Plot 1 (a) shows the initial gating of all cells with exclusion of visual debris (platelets, cell fragments), or R1. Plot 2 (b) sets CD45-gating (R2) to include only leucocytes for further analysis. Plot 3 (c) shows gating on cells from R2 with addition of anti-CD34-PE. The

4

Quadrant Statistics (d) (Gate: G2) Quad Events % Gated UL 354 1.28 UR 65 0.23 LL 15101 54.57 LR 12154 43.92 arrow indicates cells of interest in the upper right (UR) quadrant (CD45+/CD34+ cells = 1.57% of population). This plot serves as quality control in support of the final plot. Plot 4 (d) showing side scatter versus CD34 positivity with gating on R2 is the final plot, and CD34+ cell enumeration is reported as 1.28%; arrow at upper left (UL) quadrant points towards cells exhibiting low side scatter and CD34 positivity. (Courtesy of Dr W. Jaszcz and M. KraftWeisjahn, Fairview-University Medical Center and University of Minnesota.)

Autologous transfusion

Hypervolaemic haemodilution

A variation of ANH, which is not strictly autologous transfusion, is known as hypervolaemic haemodilution. Rather than removing blood preoperatively and replacing it with fluid, the fluid is given without removing any blood. This serves to dilute the Hb concentration with a reduction in the total amount of Hb lost, as described above. At the end of the procedure the patient is made normovolaemic either as a result of the operative blood loss or by diuresis. This technique is not widely practised. It is not without risks and its efficacy remains to be established. Preoperative autologous donation Principle

The principle behind PAD is that units of blood are collected, usually at weekly intervals, in the 4–5 weeks before surgery, during which time the patient will make up the blood lost. Advantages

This technique has been shown to reduce exposure to allogeneic blood: in one series of 116 adolescent patients undergoing spinal surgery, it was reduced from 60 to 11%. Disadvantages Increased risk of transfusion While PAD reduces the number of allogeneic units transfused, it may in fact increase the likelihood of being transfused, with the inherent risk this poses. This is because by taking several units of blood preoperatively, the preoperative Hb is often lower than it would otherwise have been. Autologous blood may also be perceived as safer and therefore be given more readily, but any transfusion carries risks. Predeposited autologous blood is just as likely as allogeneic to be associated with administrative errors, fluid overload or bacterial sepsis. Allergic or febrile reactions have also been reported. The effect of non-leucocyte-depleted blood whether allogeneic or autologous may still

suffer the storage lesion and result in immunomodulation in the recipient. Risk of donation The other additional source of morbidity/mortality is in the donation process itself. One North American study looked at more than 4 million whole-blood donations over a 10-month period from July 1993 and found that the risk of hospital admission following PAD was 1 in 16 783 (more than 10 times the rate for homologous donors). The reason was most commonly a severe vasovagal reaction but angina or trauma due to the venepuncture also occurred. Wastage One of the biggest problems associated with PAD is predicting how much blood an individual patient will lose during surgery and hence what the blood requirements will be. Even if autologous blood is only collected in those who would normally be crossmatched according to a maximum surgical blood-ordering schedule (MSBOS) (see Chapter 25), 30–50% will be wasted. This means that for more than one-third of patients, the resources used and the time, travelling and discomfort involved will have been wasted. Expense With the high wastage rates, PAD is an expensive option. In the USA, Medicare and some private insurers will not reimburse the costs. Looking at cost-effectiveness models from the USA, the cost of PAD per quality-adjusted life-year ranges from $230 000 in a total hip replacement to over $23 million in a transurethral resection of the prostate. These compare with less than $50 000 for most commonly accepted medical interventions. The advantages and disadvantages of PAD are summarized in Table 26.3. Practicalities Patient selection In order to minimize the risks of donation and reduce the wastage of autologous blood, guidelines have been drawn up for the selection of

305

Chapter 26 Table 26.3 Advantages and disadvantages of predeposit

autologous blood donation. Advantages Reduced exposure to allogeneic blood reducing the risks of: Transfusion-transmitted infection Transfusion-associated graft-versus-host disease Immunization to red cell/platelet and HLA antigens Provides compatible blood for patients with complex red cell antibodies or antibodies to common red cell antigens Supplements the blood supply but uses more resources and staff time to collect per unit transfused than homologous blood Disadvantages Reduces preoperative Hb, thus increasing the risk of receiving a blood transfusion Does not abolish the risks of: Bacterial contamination ABO-mismatched blood being given as a result of administrative errors Some febrile reactions Fluid overload Is associated with risks due to the donation of several units of blood 30–50% of units are wasted, exposing many patients to unnecessary morbidity and the healthcare system to unnecessary expense

• In children under 8 years old or weighing less than 25 kg, PAD is technically difficult and rarely justified; 8–16 year olds can be considered if the child is willing and the parents or guardian gives consent. The guidelines recommend that donations should be collected in a hospital in close collaboration with a paediatrician. • Exercise caution with patients under 50 kg: may need small-volume packs (250-mL packs are available) to avoid taking more than 12% of blood volume. Other requirements • The patient must have suitable venous access. • The operation would normally require blood to be crossmatched (according to MSBOS). • The operation date is fixed. • There is sufficient time to donate the required number of units: the last unit should be taken 7–10 days before the planned surgery. The minimum safe interval before surgery is 72 h, which allows the blood volume to be replenished. The process

suitable patients (see the BCSH ‘Guidelines for autologous transfusion’, part 1). In summary the medical exclusion criteria for preoperative donation are as follows. • Active bacterial infection. • Positive results for human immunodeficiency virus (HIV) or hepatitis C virus (HCV). • Epilepsy. • Prolonged or frequent angina, left main stem disease, significant aortic stenosis or cyanotic heart disease. Other patients with cardiac disease can be considered subject to assessment by a cardiologist. • Caution with patients on b-blockers or angiotensin-converting enzyme inhibitors: consider isovolaemic replacement with crystalloids. • Uncontrolled hypertension. • Pregnancy, especially if impaired placental flow or intrauterine growth retardation, due to possible harm to the fetus. • Patients with a previous history of prolonged faint after blood donation. • Hb below 11 g/dL at the start or below 10 g/dL for subsequent donations; 306

As outlined above, the use of PAD should be confined to suitable patients in whom the risk of taking autologous blood is smaller than the risk associated with the use of allogeneic blood. The request may come from patient or clinician. In either case the patient should be given a full explanation of the risks and benefits (preferably with an information leaflet). The referring physician should be satisfied that the patient is medically fit and meets the criteria for PAD. The discussion should include the following. • The risks and benefits of PAD. • The requirement for microbiology testing for HIV, HBV and HCV (to reduce the risk if the unit was inadvertently transfused to another). • The possibility that the patient will not need a transfusion and the units will be wasted. In the UK, unlike some other countries, units are not ‘crossed over’ for other patients’ use since many autologous donors would not meet the required criteria. • The possibility that more blood will be required than has been collected so the patient may need additional, allogeneic blood.

Autologous transfusion

• The possibility that some or all of the units collected may not be usable, e.g. if the bag had a leak or the patient develops an infection soon after donation. If a patient wants to proceed, appointments are made, usually at weekly intervals. The first appointment should be less than 5 weeks before the operation date as units can only be stored for a maximum of 35 days. The last one should be at least 72 h (preferably a week) before the planned surgery to allow time for plasma and the majority of the Hb to be replenished. It has been shown that oral iron improves the yield, even in iron-replete patients, so patients are advised to take iron. Patients who are iron deficient should start this in advance of the first appointment. Erythropoietin is licensed for use in PAD but is expensive and may be associated with an increased risk of thrombosis, and is therefore not generally recommended. The use of intravenous iron is likely to increase due to the improved safety profile of the iron sucrose preparations. Whether the units are collected in a hospital or a transfusion centre, it is vital that strict standard operating procedures are drawn up and followed to ensure that any risk of administrative error is minimized. Once the units are collected they should be clearly labelled with the patient’s details and the label signed by the patient. The units are then transferred to the hospital blood bank where they should be kept securely segregated from the stocks of donor blood. Prior to surgery a further sample from the patient should be sent and the autologous units crossmatched against it as an added safety check. This sample is also required in case additional allogeneic blood is needed. It is vital that there is good communication between all the staff involved. These include the referring clinician, the transfusion centre, the hospital blood bank, junior staff requesting the blood and the doctor who prescribes it. One of the most common errors in autologous transfusion is to give allogeneic blood instead of the predeposited units, often because it was not realized that autologous units had been collected.

Blood substitutes The use of blood substitutes is discussed in Chapter 30.

Summary The place of the various forms of autologous transfusion should be considered as part of a strategy for minimizing the risk associated with transfusion for patients having surgery, where blood loss is expected. • All transfusions, whether allogeneic or autologous blood, carry a risk. • Planning and appropriate treatment in advance of or during surgery may reduce transfusion requirements. • Before transfusing a patient always consider whether it is necessary and if so how many units are required. • ICS and PCS are considered the most costeffective methods of autologous transfusion.

Further reading Birkmeyer JD, Goodnough LT, Aubuchon JP, Noordsij PG, Litenberg B. The cost effectiveness of preoperative autologous blood donation for total hip and knee replacement. Transfusion 1993; 33: 544–51. British Committee for Standards in Haematology. Guidelines for implementation of a maximum surgical blood order schedule. Clin Lab Haematol 1990; 12: 321–7. British Committee for Standards in Haematology. Guidelines for autologous transfusion. 1. Preoperative autologous donation. Transfus Med 1993; 3: 307–17. British Committee for Standards in Haematology. Guidelines for autologous transfusion. 2. Perioperative haemodilution and cell salvage. Br J Anaesth 1997; 78: 768–71. Calman KC. Cancer: science and society and the communication of risk. Br Med J 1996; 313: 799–802. Consensus statement. Autologous Transfusion: 3 years on. What is new? What has happened? Transfus Med 1999; 9: 285–6. Goodnough LT, Brecher ME, Kanter MH, AuBuchon JP. Transfusion medicine. Part 1. N Engl J Med 1999; 340: 438–47.

307

Chapter 26 Goodnough LT, Brecher ME, Kanter MH, AuBuchon JP. Transfusion medicine. Part 2. N Engl J Med 1999; 340: 525–33. Goodnough LT, Shander A, Brecher ME. Transfusion medicine: looking to the future. Lancet 2003; 361: 161–9. Hébert PC, Wells G, Blajchman MA et al. A multicentre, randomised controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999; 340: 409–17. Newman JH, Bowers M, Murphy J. The clinical advantages of autologous transfusion. A randomised controlled study after knee replacement. J Bone Joint Surg 1997; 79B: 630–2.

308

Popovsky MA, Whitaker B, Arnold NL. Severe outcomes of allogeneic and autologous blood donation: frequency and characterization. Transfusion 1995; 35: 734–7. Royal College of Physicians of Edinburgh. Consensus Conference on Autologous Transfusion. Transfusion 1996; 36: 625–67. Thomas D, Wareham K, Cohen D, Hutchings H. Autologous blood transfusion in total knee replacement surgery. Br J Anaesth 2001: 86: 669–73. Vanderlinde ES, Heal JM, Blumberg N. Clinical review: autologous transfusion. Br Med J 2002; 321: 772–5.

Chapter 27

Tissue banking Deirdre Fehily and Ruth M. Warwick

The banking of cord blood, as described in Chapter 28, has been a natural progression for the UK blood services, where much of the expertise and infrastructure required already existed for the banking of blood. Less obvious, perhaps, is the parallel that exists between the requirements for safe blood banking and the safe banking of other human tissues such as bone, skin and heart valves, which are used for orthopaedic procedures such as revision hip surgery, burns and heart valve replacement respectively. Although the tissues themselves may be very different from blood, the principles applied to blood and cord blood banking are equally relevant to the banking of these tissues. In recognition of the UK blood services’ expertise and infrastructure, the National Blood Service in England has greatly increased its involvement in tissue banking over the last 10 years. Blood centres in other European countries, notably France and Spain, also contribute significantly to the provision of tissue banking services. There are some features of tissue banking, however, that make it distinctly different from blood or cord blood banking, and these must be addressed by any organization undertaking tissue banking. • In many cases, tissue donors are necessarily deceased. Donor selection, consent, testing and counselling issues all present different challenges in this context. • Processing is by necessity very ‘open’, and therefore the requirements for processing facilities are much more stringent than for blood. Methods of processing also vary greatly and tissues can be supplied for transplant in either a viable or non-viable state. • The medical director of a blood service tissue

bank is normally a haematologist, who will not be expert in the clinical applications of tissue transplantation in the surgical specialities of orthopaedics, plastic surgery, and cardiac surgery for example. Consequently, the clinical dialogue between the tissue bank and the user hospital is different to that which exists for blood and may rely on the existence of a medical advisory committee with clinical experts from the relevant disciplines. While tissue banking within the blood services has grown significantly, the service is also provided by many hospital departments and particularly in the USA by independent organizations some of which operate for profit. Where cadaveric tissues are banked, there is always some degree of collaboration with organ transplantation programmes and in some cases organ and tissue donation is coordinated in a fully integrated way, though this is unusual in the UK. Hospital-based tissue banks usually concentrate on the banking of a single tissue and are directed by a surgeon who is also a user of that tissue. This latter type of arrangement gives the benefit of very good clinical feedback and motivation to maximize collection but has the disadvantage of a lack of expertise in donor selection and good manufacturing practice (GMP). Guidance is available from a number of sources on the quality standards that should be applied in tissue banking. Table 27.1 includes references to the most important of these documents. Legally binding regulation is in place in the USA, where the Food and Drug Administration (FDA) has published a number of relevant rules (also listed in Table 27.1). In the European Union, legally binding regulation will soon be in place. A new Directive, adopted in 2004, will require all 309

Chapter 27 Table 27.1 Pertinent laws and guidance in the field of tissue

Consent

banking. The Human Tissue Act 1961 (UK) Human Organ Transplants Act 1989 (UK) Anatomy Act 1984 (UK) Coroners Act 1988 (UK) Human Tissue Ethical and Legal Issues published by the Nuffield Council on Bioethics 1995 Committee on Microbiological Safety of Blood and Tissues for Transplantation, Department of Health: guidance on the microbiological safety of human tissues and organs used in transplantation (NHS Executive,August 2000 A Code of Practice for the Diagnosis of Brain Stem Death (including Guidelines for the Identification and Management of Potential Organ and Tissue Donors) (UK Department of Health, March 1998) Rules and Guidance for Pharmaceutical Manufacturers (EEC Orange Guide) (London, HMSO, 1997) Standards for tissue banking, British Association of Tissue Banks (Transfus Med 1996; 6: 155–8) Standards for tissue banking,American Association of Tissue Banks (1998) ‘Guidelines for tissue banking’ (section 4, added in 1999) in Guidelines for the Blood Transfusion Services in the United Kingdom (3rd edn., 1996) FDA regulations governing human tissue intended for transplantation. Title 21 Code of Federal Regulations Part 1270 FDA Guidance for the Industry on Validation at www.fda.gov/cver/guidelines.htm The Belgian Tissue Banking Law 1988 The Spanish Tissue Banking Law 1996 The French Tissue Banking Law 1994 Ethical Aspects of Human Tissue Banking. Opinion of the European Group on Ethics in Science and New Technologies to the European Commission (21 July 1998) A Code of Practice for Tissue Banks Providing Tissues of Human Origin for Therapeutic Purposes (UK Department of Health, 2001) Guide to safety and quality assurance for organs, tissues and cells (Council of Europe, 2002)

member states to have inspection and accreditation systems in place by June 2006 which ensure that all banks providing these services comply with an agreed set of standards. This chapter aims to identify the key considerations for a blood centre embarking on the banking of human tissues.

310

Patients undergoing surgery for joint replacement or heart transplant can donate bone and heart valves respectively. In the former cases, written consent for all the relevant aspects of tissue donation and testing should be obtained in advance of tissue retrieval and be witnessed, and the process of obtaining consent for donation should be entirely separate from the process of obtaining the patient’s consent for surgery. The legal requirements for obtaining permission for the retrieval of tissues after death vary from country to country. However, even where ‘opting out’ or ‘presumed consent’ systems are operated, it is considered best practice to confirm that relatives do not object to the donation. In the UK the Human Tissue Act 1961 makes it clear that the requirement in these cases is to establish a ‘lack of objection’ from the deceased’s family, rather than ‘consent’, and, once established, the donation can proceed even in the absence of a donor card. The UK legal framework for consent to donation is currently under review. Consent should cover the following areas: • the intended clinical use of the donation; • virological testing at the time of donation (and, for living donors, again at least 6 months later) including specific mention of human immunodeficiency virus (HIV); • review of medical records held elsewhere, if necessary; • receiving counselling in the event of positive virology markers in living donors or cadaveric donors, if they are deemed to have clinical significance for surviving relatives; and • the use of the tissue for research and development, if it proves unsuitable for clinical use.

Donor selection Living donors

Where patients can be interviewed face to face, the selection process can be very similar to blood donation but should also include a review of the patient’s hospital notes.

Tissue banking

Cadaveric donors

The primary source of donor selection information for cadaveric donors is the interview with the donor’s relative(s). The interviewee should be the person who knew the potential donor best, even if they are not the next of kin. This should be established at the beginning of the interview. For cadaveric donors, additional information should be sought from the family doctor as an added security measure required in the absence of a face-toface interview with the donor. Additionally, where a postmortem has been performed, the results should be reviewed as part of the donor selection process. Interviews with donors or their families should include enquiries as detailed in Table 27.2.

Donor testing Testing of cadaveric and living donors closely follows the testing undertaken for blood donation. Living donors

In many countries, including the UK, there is a requirement to quarantine living tissue donations and to obtain a further blood sample from the donor at least 180 days following their initial donation. Virology testing at donation and after quarantine should be as for blood donors. Cadaveric tissue donors

The quality and nature of blood samples removed from deceased donors varies considerably due to autolysis and haemolysis and there is documented observation of a high rate of false positivity in antibody assays and a significant rate of inhibition in nucleic acid technology (NAT) assays. It is important to standardize, as much as possible, the site and method of sample collection and to minimize the time period between death and blood sampling. Consideration should also be given to the validity of the sample that is tested. A sample that is taken close to an intravenous administration or central line may be diluted even if the donor has

not received a significant amount of fluids. Wherever possible, antemortem blood samples should be taken, as long as they can be reliably identified. Recent work suggests that newer DNA extraction techniques can be adapted and applied to remove the inhibition caused by cadaveric samples in NAT assays. It appears that NAT may therefore provide the most secure method for mandatory marker testing in cadaveric donors if applied in the appropriate way. An additional consideration with cadaver donors is whether transfusion of blood or other fluids in the antemortem period, particularly where the donor has also lost blood, may have resulted in a plasma dilution effect which might render virology tests unreliable. It is essential therefore to record all fluids administered in the 48 h prior to death. An estimation of any plasma dilution effect can be calculated. It is generally accepted that a blood sample that is more than 50% dilute should not be considered valid. An algorithm can be applied in the calculation of plasma dilution for cadaveric donors (Table 27.3). In the context of hepatitis B surface antigen (HBsAg) testing, high rates of non-specific reactivity are recognized in viral screening tests undertaken on cadaver blood samples. American Association of Tissue Bank guidance requires the discard of donations associated with such samples. In the UK, where confirmatory tests clearly indicate the absence of infection, tissues derived from cadaveric donors whose blood samples are repeatably reactive in an HBsAg screening test may be utilized, according to the Guidelines for the Blood Transfusion Services in the UK. Clarification of the significance of the repeatably reactive tests can be undertaken using the following approach. • Samples must lack anti-hepatitis B core antigen (HBc) using a validated assay for blood donors or, when available, an assay validated in the context of cadaveric testing. • The sample must not demonstrate evidence of neutralization of HBsAg with a specific antibody. • Testing must be undertaken in a designated diagnostic laboratory with proven expertise in the context of blood donor testing and cadaveric donor testing.

311

Chapter 27 Table 27.2 Enquiries to be made of tissue donors or their families.

Enquiry category

Information sought and outcome

Notes, examples and special circumstances

General past medical history

Malignancy specifically excludes

Exceptions are cured in situ cancer of the cervix and basal cell carcinoma. Primary brain tumours should be excluded unless benign nature is confirmed histologically. This is due to the risk that a solitary metastasis might be mistaken for a primary. Diagnostic procedures during life may increase the chance of extracranial metastases by breaching the blood–brain barrier For example sarcoidosis, Crohn’s disease and ulcerative colitis have some features in common with some infectious disorders Parkinson’s disease and multiple sclerosis are common diseases in this category Examples include systemic bacterial infections such as fulminant pneumonia (although small foci of infection associated with ventilation, even if currently being treated, may not exclude the donor), septicaemia, acute myocarditis and active tuberculosis Examples include rheumatoid arthritis, systemic lupus erythematosus and polyarteritis nodosa.These may affect the quality of the tissues. May be treated with drugs that may affect the quality of the tissues. May be treated with immunosuppressants that may affect validity of test results.Are of unknown aetiology, possibly with an infectious trigger

Diseases of unknown aetiology are reason to exclude Diseases of neurodegenerative aetiology specifically exclude Diseases of known infectious origin usually exclude

Multisystem autoimmune diseases exclude donation

Medication

Enquiries primarily aimed at identifying underlying disease that may make the donor ineligible

Long-term steroid therapy can affect the quality of skin and bone and immunosuppression may render antibody-based tests invalid

Hepatitis and HIV transmission risk

Risks due to acupuncture, tattooing, ear or body piercing, etc. in the previous year. Receipt of an organ or tissue transplant

Living donors with these risks can be accepted although the retest should be timed to ensure that it is conducted at least 1 year after the risk event. Consideration should be given to adding an anti-HBc test at the recall. Corneal recipients excluded from donation

CJD

Enquiries should elicit any family history of CJD and any brain or spinal surgery before 1992. Hormone treatment for infertility or growth before 1985

Details and dates of brain or spinal surgery should be recorded and further investigations made with the hospital concerned

Travel (malaria and Chagas’ disease)

The rules for history-taking, acceptance and malarial antibody testing of blood donors should be applied equally to tissue donors

It is not clear whether any risk of malaria transmission remains in non-viable tissues. Cornea banks do not exclude donations on the basis of malarial risk

Tissue-specific medical history

Depending on the tissues to be donated, enquiries should be made to exclude donors on the basis of a medical history which may imply that the quality of the specific tissue is compromised

For example, previous eye surgery in eye donors, or previous hip surgery in femoral head donors

Recent history

Enquiries should establish circumstances surrounding the death, including whether a hospital or coroner’s postmortem is to be performed

CJD, Creutzfeldt–Jakob disease.

312

Tissue banking Table 27.3 Algorithm for the

calculation of plasma dilution.

Interval prior to sampling

Volume infused (mL)

Per cent retained

Volume retained (mL)

Crystalloid infused > 24 h 2–24 h 1–2 h <1h Total crystalloid retained =

… … … …

0 25 50 75

None … … …

… … …

100 (blood) 50 (colloid) 100

… … …

Blood/colloid infused 24–48 h 0–24 h Total blood/colloid retained = Estimated total blood volume % Haemodilution =

70 mL/kg body weight

Crystalloid retained + blood/colloid retained ¥ 100

In addition to testing for the mandatory markers, bacteriology testing of the tissues themselves is vital. Policies and procedures must address issues including sampling and culture techniques, which must be validated to detect pathogenic bacteria including spore-forming types such as Clostridium. Testing of the donor for variant Creutzfeldt– Jakob disease (vCJD) or classical CJD has not yet been validated for the latent period.

Tissue procurement Living donors

By necessity, living donations are retrieved during surgery by the operating team. Clear written instructions and staff training should be provided by the tissue bank for tissue collection, with regular auditing to ensure compliance with procedures. A critical aspect of the retrieval is the identification of the donor, the donation and the associated blood sample. The use of barcoded donation number labels for donations and samples greatly increases the security of this step. It is essential that operating theatres collaborating in living donor tissue retrieval are supplied with kits that are easily used in a sterile field. It is advisable that a written document is in place

blood volume

which clarifies the responsibilities of the retrieving hospital and the tissue bank. Cadaveric donors

The donor must be positively identified by means of a wristband, toe-tag or by the mortuary staff. The appearance of the donor’s body must tally with the description of the donor and the circumstances of death, e.g. age, gender, ethnicity. Where tissues are to be processed in the tissue bank, with a decontamination or terminal sterilization step, it is common practice for the retrieval to be conducted in a mortuary. However, sterile instruments should be used and a local sterile field created. Before the retrieval commences, a thorough examination of the donor body should be conducted and recorded, to be included as part of the donor selection assessment. The examination should include detection and recording of: • tattoos; • jaundice; • evidence of intravenous drug use; • skin abnormalities; • body piercing; • open wounds or signs of infection; • scars; • intravenous cannula sites; and • operation incision sites. 313

Chapter 27

The team should also record the site and time of blood sampling, if they are taking the blood sample. The sample should be taken at a site which is distant from the placement of any intravenous lines. In general, tissues should be retrieved within 24 h of death. It is preferable for the retrieval to be undertaken prior to any autopsy as long as circumstances allow. An important aspect of the retrieval is the careful reconstruction of the donor body. Extendible plastic prostheses can be used to replace large bones.

Tissue processing If tissues have been retrieved in an operating theatre and validated bacteriological testing reveals absence of bacterial or fungal contamination, they may be frozen and transplanted without further processing (e.g. femoral heads). If, however, there is evidence of bacterial contamination or the tissues have been retrieved in a mortuary, they should be further processed. In the case of a multiorgan, multitissue donor in the USA, from whom HIV was transmitted by organs and frozen bone grafts, there was no transmission by bone that had been processed by cleaning, shaping or morsellizing and freeze-drying, even though it had not been subjected to terminal sterilization. Processing reduces the risk of disease transmission by: • the removal of blood and marrow; • reducing infectious contamination by chemical and physical means. Pooling of cadaveric donations during process-

ing is not permitted by standards in Europe or the USA. The UK standards do permit the pooling of bone donations from living donors which have been tested at donation and again after 6 months’ quarantine, although this is not permitted in the USA or by the European Association of Tissue Banking standards and is under review in the UK. Processing of bone and processed bone are shown in Figs 27.1 and 27.2. Processing facilities

The standard of processing facilities required depends on whether: • the tissue is terminally sterilized in its final packaging; or • further product manipulation is required after the decontamination or sterilization step. In the former case, the Orange Guide (see Further reading) can be interpreted as advising the need for a class C environment. In the latter case, an environment of class A with class B background is necessary (Table 27.4 lists the maximum particulate levels permissible for these clean-room categories, while Table 27.5 demonstrates how these European clean-room categories broadly relate to other international standards). To achieve these environmental standards, processing facilities must have: • highly filtered air and must operate under positive pressure; • air change rates of at least 20/h; • weekly environmental monitoring; • regular particle counting and air change rate monitoring;

Table 27.4 Air classification system for

Maximum permitted number of particles per m3 equal to or above At rest

manufacture of sterile medicinal products.

In operation

Grade

0.5 mm

5 mm

0.5 mm

5 mm

A B C D

3 500 3 500 350 000 3 500 000

0 0 2 000 20 000

3 500 350 000 3 500 0000 Not defined

0 2 000 20 000 Not defined

314

Tissue banking

Fig. 27.1 Processing morsellized bone.

• smooth and washable floors, walls, ceilings and work surfaces; and • clothing that complies with the guidance given in the Orange Guide, or other equivalent standard.

main processes in use (and the products’ clinical applications) are summarized in Table 27.6.

Supply and tracking of tissues Processing, products and their uses

The processing methodology applied varies depending on the type of tissue concerned. The

Most tissue banks currently supply tissues direct to theatre departments. It is not possible for tissue banks to take responsibility for the ultimate fate of 315

Chapter 27

Fig. 27.2 Freeze-dried struts and chips.

Table 27.5 Comparison of European,

EC GMP for Medicinal Products 1998 (Annex 1) EEC Orange Guide 2002 grade

US Federal Standards 209 D 1989 class

ISO 14644

A, B C D

100 10 000 100 000

5 7 8

American and ISO classifications.

Table 27.6 Tissue processing, tissue products and their uses.

Tissue

Usual processing methods

Main products

Major uses

Unprocessed bone allografts

Freezing to -40°C or lower; no further processing

Frozen femoral heads

Impaction grafting in joint revision surgery

Non-viable bone allografts

Cutting, shaping or grinding Bone marrow depletion Freezing at -40°C or lower Freeze drying Terminal sterilization (usually g-irradiation)

Whole or half femoral heads or femoral slices Cancellous or corticocancellous ground bone Cancellous cubes, pegs, cortical rings and struts

Impaction grafting in joint revision surgery Cortical struts are applied as ‘onlay’ grafts to the outside of the femur and wired in place to provide extra support to a weakened or damaged femur Cortical rings from the femoral shaft of younger donors can be used in spinal surgery to replace removed discs

Demineralized bone allografts

Cutting and grinding Bone marrow depletion Demineralization by acid treatment Freeze drying Sterilization by ethylene oxide or electron beam

Demineralized bone powder Demineralized cortical sheets

Demineralization increases the osteoinductive capacity of bone by rendering the relevant cytokines (bone morphogenic proteins) more available Used primarily in dental and maxillofacial surgery

316

Tissue banking Table 27.6 Continued

Tissue

Usual processing methods

Main products

Major uses

Massive bone allografts

Aseptically retrieved in theatre and unprocessed or surface cleaned Frozen at -40°C or lower Usually sterilized with g-irradiation

Proximal and distal femurs Proximal and distal tibias Proximal and distal humerus, etc.

Used to replace large sections of bone that have been removed (osteosarcoma, trauma or very severe damage caused by joint replacement) The articular surfaces must be removed and replaced with a prosthesis unless they have been cryopreserved (see osteochondral allografts)

Osteochondral allografts

Surface decontamination with antibiotic solutions Frozen at a controlled rate with articular surfaces soaked in DMSO

Proximal and distal femurs, etc.

Used as massive allografts above but without the need to replace articulating surfaces with prostheses

Tendon allografts

Cutting and shaping Soft tissue and bone marrow removal Decontamination with antibiotics or ethanol Freezing or freeze drying Sterilization with g-irradiation

Whole, half or specifically shaped patella tendons Achilles and semitendinosus tendons

Primarily used in sports injury knee reconstructions (anterior and posterior cruciate) Biomechanical strength is of particular importance

Cardiovascular allografts

Decontaminated in an antibiotic solution Cryopreserved with DMSO Competency tested during processing (NB: some banks keep ‘homovital’ valves in culture medium for up to 1 month)

Cryopreserved pulmonary valves, aortic valves, mitral valves, non-valved conduits

Used to replace diseased or damaged valves

Skin allografts

Trimming and cutting Preservation in high concentration glycerol (non-viable) Decontamination in an antibiotic cocktail and cryopreservation with glycerol or DMSO

Glycerolized non-viable skin Cryopreserved skin

Majority use in the UK is applied as a temporary dressing in patients with large-area burns It can subsequently be used as the substrate for the application of cultured autologous epidermis

Corneas

Stored in culture medium for up to 1 month

Viable corneal allografts

Used to replace diseased or damaged corneas

Amniotic membrane

Cut in pieces Frozen to –40°C or lower with glycerol Cryopreserved

Frozen amniotic membrane Grafts Cryopreserved amniotic membrane grafts

Used in ophthalmic surgery for the reconstruction of the ocular surface following chemical, thermal or surgical trauma or severe scarring following primary or secondary pterygium or other syndromes Potentially also useful as a burn or skin ulcer dressing

every graft once delivered to a hospital, though many supply the hospital with a recipient record to be completed for each graft and returned to the bank. The users should always be advised to:

• keep a log of tissue received and used; • record any allograft unit numbers in the patient’s notes; and • inform the tissue bank immediately of any 317

Chapter 27

adverse reaction that might be attributable to the tissue graft. In most cases tissues are supplied for specific cases and stocks are not held in theatres, the notable exception to this being freeze-dried bone. Substantially more experience in the storage and recording of human tissue for clinical use is available in the hospital blood bank. If hospital haematology departments are also willing to receive, store and distribute tissues in the future, then full product barcoding will be required to enable the use of existing computer systems. Such a step would greatly increase the level of security and traceability of tissues once received in a hospital.

Ethical issues in tissue banking There are many challenging ethical issues in this field. Two that have been much debated in the UK are the supply of tissues for research and the supply of recipient information to donor families. Supply of tissues for research

There is an increasing need for human tissues in research and development within tissue banks, academic institutions, hospital departments and the commercial sector. This need creates ethical and organizational difficulties that tissue banks must address. Some guidance has been provided in the UK by the Nuffield Council on Bioethics and also by the European Commission’s Group on Ethics in Science and New Technologies and a recent consensus paper from the USA, which has been adopted in the UK. Provision of recipient information to donor families

The tradition within the organ transplantation field is to give quite detailed information about recipients to donor families. As many organ transplant coordinators refer donors to tissue banks, they sometimes expect to receive the same type of information about tissue recipients and to pass this to the donor family. This raises difficult issues, particularly relating to confidentiality, and is not consistent with practice in blood banking. Any tissue 318

bank retrieving cadaveric tissue donations needs to have a clear policy on this issue.

Summary The banking of non-blood tissues is increasing within blood services where expertise in donor selection, donor testing and GMP is being utilized for the banking of bone, tendons, heart valves and skin. Living donors can donate bone and heart valves during joint replacement or heart transplant surgery, respectively. All other types of tissue donation must be made after death. For cadaveric donors, a more thorough medical and behavioural history is recorded to compensate for the lack of a face-to-face donor interview. Additional information should be sought from the donor’s GP and the postmortem report (where applicable). Great care should be taken in the collection and acceptance of test results where tests were performed on cadaveric blood samples. Tissue processing is necessarily open and usually involves decontamination or terminal sterilization. Facilities for tissue processing should be designed to achieve class C for tissues destined for terminal sterilization and class A, with class B background, for the manipulation of tissues in the absence of a terminal sterilization step but following chemical or antibiotic decontamination (see Tables 27.4 and 27.5). The most important application of bone allografting is in impaction grafting during revision joint replacement surgery; this requires large volumes of cancellous or corticocancellous bone. Bone grafting is also used in spinal surgery and tumour and trauma surgery. Allograft tendons are primarily used in knee ligament reconstruction. Skin grafts are used primarily as temporary dressings on patients with major burns.

Further reading Anonymous. Guidelines for preventing transmission of human immunodeficiency virus through transplantation of human tissue and organs. MMWR 1994; 43: RR-8. Eastlund T. Infectious disease transmission through tissue transplantation: reducing the risk through donor selection. J Transpl Co-ord 1991; 1: 23–30.

Tissue banking Gie GA, Linder L, Ling RS et al. Impacted cancellous allografts and cement for revision total hip arthroplasty. J Bone Joint Surg 1993; 75B: 14–21. Guidelines for the Blood Transfusion Service in the United Kingdom, 4th edn. Norwich: The Stationery Office, 2000. Kearney J. Quality issues in skin banking. Burns 1998; 24: 299–305. Ling RSM, Timperley AJ, Linder L. Histology of cancellous grafting in the femur. J Bone Joint Surg 1993; 75B: 693–6.

Padley DJ, Lucas SB, Saldanha J. Elimination of false negative HCV RNA results by removal of inhibitors in cadaver donor blood specimens. Transplantation 2003; 76: 432–4. Simonds RJ, Holmberg SD, Hurwitz RL et al. Transmission of human immunodeficiency virus type 1 from a seronegative organ and tissue donor. N Engl J Med 1992; 326: 726–32. Warwick RM, Eastlund T, Fehily D. Role of the blood transfusion service in tissue banking. Vox Sang 1996; 71: 71–7.

319

Chapter 28

Cord blood banking Ruth M. Warwick, Sue Armitage and Deirdre Fehily

Rationale for cord blood banking A variety of malignant or severe genetic diseases of the bone marrow (BM) or immune system can be treated by BM transplantation (BMT). About onethird of patients can find a match within their own family; for the remainder, searches of registries of unrelated BM donors are required. Even so, many patients with leukaemia are unable to undergo BMT because of the lack of a suitable donor or because their disease cannot wait for the time required to find a matched BM donor, which on average takes 4 months. In practice, many patients, particularly those from the ethnic minorities, cannot find a suitable match. Cord blood (CB) transplantation offers a therapeutic option for these patients, and the following sections deal with unrelated CB banking. In addition, donations from related individuals may be banked. These directed donations are also discussed. Benefits of CB banking

The advantages of CB are now well recognized and these are described below.

found from CB banks for many patients for whom there is no chance of finding a match from other sources. Availability

Generally considered a waste product, CB is abundantly and readily available. Once stored, CB can be provided on demand, reducing the morbidity and death in patients associated with the long procurement times for BM. Absence of risk for the CB donor

There is no risk to the donor in collecting CB from the placenta. Low incidence of viral carriage

CB has a lower incidence of viral infection than that found in adult BM donors, particularly for cytomegalovirus (CMV) which can cause profound illness in BMT patients, as well as Epstein–Barr and other viruses. Recipient tolerance of mismatch

Ethnic targeting

Ethnic minority groups that are markedly underrepresented in BM donor registries can be targeted for CB collection. For example, the London Cord Blood Bank’s policy to focus on collection from ethnic minority communities has resulted in approximately 40% of the donations deriving from ethnic minorities compared with only about 2% in British Bone Marrow Registry (BBMR) donors (Table 28.1). Consequently matches can be 320

Data suggests that CB recipients tolerate human leucocyte antigen (HLA) mismatches better than when BM is used. A lower incidence of graftversus-host disease (GVHD) is reported in CB transplantation for both children and adults, contributing to reduced morbidity compared with BMT in appropriate patients. This relative tolerance reduces the tissue-matching requirement and enables larger numbers of recipients to find a donor, resulting in the need for fewer CB dona-

Cord blood banking

tions to be banked than the number of donors needed for an effective BM registry. Disadvantages of CB as a source of haematopoietic stem cells

• Only one donation can be obtained, after which the donor cannot be contacted again. Table 28.1 Ethnic background of London Cord Blood Bank

(LCBB) donors and bone marrow donors from the British Bone Marrow Registry (BBMR) North London Blood Centre.

Ethnic background European Caucasian Indian Subcontinent African/Afro-Caribbean Oriental Mixed Other

• The volume obtained is limited, with many banks reporting a median of 80 mL compared with 1–1.2 L or more of BM that can be harvested from an adult. • This limits the cell dose available, as reflected by the longer engraftment time compared with BMT, and is a major limiting factor particularly for adults. • The donor may have a genetic disease of the marrow or immune system, which is not apparent at the time of donation, but could be transmissible by transplantation.

BBMR (n = 107 000) (%)

LCBB (n = 4987) (%)

CB banking and the UK National Blood Service

97.5 0.7 0.7 0.1

57.7 21.0 6.2 1.0 7.7 6.5

There is much expertise and an appropriate infrastructure within blood centres in England to facilitate CB banking, as shown in Table 28.2.

1.0

Table 28.2 Skills and experience of

English blood centres applicable to tissue and cord blood banking.

Skills

Examples

Public support and accountability

Skill in public education with a broad-based public information system History of success in donor recruitment

Medical infrastructure

Donor selection and counselling, expert advisory committees (e.g. Standing Advisory Committee for Donor Care) Comprehensive quality system Donor testing expertise through the Standing Advisory Committee for Transfusion-transmitted Infections

Good manufacturing practice

Donor testing using automated virology testing and reporting to avoid transcription errors Temperature-controlled and monitored storage Regulatory compliance Culture of stringent record keeping

Operational infrastructure

Transportation network Cryopreservation facilities Clinical and financial interface with hospitals 24-hour, 7-day operation Secure national IT systems using barcoding for all banking functions from the donor to the finished product

321

Chapter 28

Unrelated CB banking

Table 28.3 Consent for cord blood banking.

Ethics and consent issues

Collection and storage of cord blood for transplantation into unrelated individuals in the UK and abroad Examination of the mother’s and infant’s relevant medical notes and dialogue with relevant clinical professionals Permission for microbiological testing, including for HIV, and for the donor to be counselled in the event of results relevant to their health Storage of samples for future testing Storage of personal information Research and development use if the donation is unsuitable for clinical use

The ethical issues associated with CB banking are complex and include the following. • ‘Ownership’ of CB and issues of consent, with varying views on who ‘owns’ the cord: the neonate, the mother, the CB bank or the community. • Commercial interests in the field of autologous CB banking, with patent applications relating to all aspects of CB banking and transplantation recently refused in Europe, Japan and the USA. • Controversy exists regarding the most appropriate method of collection. With in utero collection there is less delay, and it has been suggested that this might both increase the collection volume and reduce the incidence of clotted collections. Issues of privacy and intrusion complicate collection at this time for the mother in the delivery suite and by a potential conflict of interest on the part of the professionals caring for the mother–infant pair, if they are also required to undertake the collection. • Intrusion into the confidentiality and convenience of the donor, especially with regard to infectious and genetic testing, must be balanced against safety of the recipient. • In the context of directed collections, antenatal tissue typing and screening for genetic diseases could be undertaken on potential donors, and result in the elective abortion of fetuses with genetic disease or those which are not a tissue type match for the intended recipient. • The content of the written consent given by unrelated donating mothers should include all the aspects detailed in Table 28.3. Operation of CB banking

The banking of CB is conducted in different ways at different centres, involving many steps in the process. Figure 28.1 details the operational process at the London Cord Blood Bank as an example. Close collaboration and clear definition of roles and responsibilities between the bank and the participating obstetric units is essential, whether the obstetric unit staff participate in the collection and donor interview process or these 322

steps are carried out independently by CB bank staff. The majority of CB banks work to the FACTNETCORD international standards for CB banking. A number of sources provide guidance regarding quality standards that should be applied to CB banking. Many of the documents listed in Table 27.1 (Chapter 27) are also relevant to CB banks. In the European Union legally binding regulation is being developed: a new Directive was adopted in 2004 which will require all member states to have inspection and accreditation systems in place by 2006 to ensure that all banks providing these services comply with an agreed set of standards. The Food and Drug Administration (FDA) in the USA are proposing new regulations for good tissue practice. Tissue typing and search mechanisms

On release from quarantine when all donor and donation data have been reviewed, CB units are made available for search through national and international registries. There are specific steps in the selection and release of units for banking, search and subsequent transplantation. CB banks vary in their procedures and an example from the London Cord Blood Bank is provided in Fig. 28.2.

Reducing the risk of transplant-transmitted disease and infection Strategies are needed to minimize the risk of infectious and genetic disease transmission by CB trans-

Cord blood banking The CB banking process at the London Cord Blood Bank Donor recruitment Information provided to mothers during antenatal period in the form of leaflets, videos, posters and presentations at parentcraft classes Collection Placenta passed to member of LCBB staff Umbilical cord cleaned CB collected Initial donor interview Mother provides written consent, brief medical, behavioural, ethnic and travel history. Maternal blood samples Processing of donation Volume reduction – removal of plasma & red cells Cryopreservation with 10% DMSO Controlled rate freezing Storage of donations Over wrapped Stored in liquid nitrogen Temperature monitored continuously Follow up donor interview Donor mother telephoned 8 weeks after delivery Provides detailed medical, genetic & family history Provides details of postnatal health of mother & baby

Collections <40 ml discarded

If mother refuses collection is discarded. If mother gives consent, maternal blood samples taken for: ∑ mandatory microbiology screening plus CMV ∑ DNA/plasma archive Samples removed from collection for ∑ cell counting & viability ∑ mandatory microbiology screening ∑ ABO/Rh/Kell typing ∑ HLA –A, B, DR typing ∑ bacteriology screening ∑ archiving

Correspondence with healthcare professionals, if necessary

Donor/donation clearance for banking Clinical information collated with haemoglobinopathy screening reports Donor files reviewed by LCBB physician and quality. Additional information may be sought Fig. 28.1 Cord blood banking process at the London Cord Blood Bank (LCBB). CMV, cytomegalovirus; DMSO, dimethyl

sulfoxide.

plantation. Volunteer BM donors would probably manifest any genetic disease at the time they volunteer. This may not be the case for the infant donor of CB, hence the need for specific screening for haematological abnormalities including thalassaemia and sickle cell disease, particularly in donors from ethnic groups known to be at risk for these conditions. The London Cord Blood Bank strategies to reduce the risk of disease transmission by CB are detailed in Table 28.4.

Donor testing and the need for a maternal and CB analyte archive

• Where neonatal haemoglobinopathy data are not available and the ethnic background demonstrates the need for further investigation, stored CB DNA can be used for testing for the common haemoglobinopathies. • In the UK mandatory tests for blood donors currently include serological testing for human immunodeficiency virus (HIV) type 1 and 2, hepatitis C virus (HCV), human T-cell leukaemia virus (HTLV) I and II, hepatitis B surface antigen 323

Chapter 28 Tissue typing and search mechanisms HLA Typing for banking Banked units typed for HLA-A, B & DR

Medium resolution (antigen split) level using sequence specific oligonucleotide probing and sequence specific priming Sequence based typing may be used to resolve ambiguities.

Registration of cord blood donations HLA type & NCC registered with international CB and BM registries

Search TC search registries for patient matches

Preliminary information Initial report sent to TC on request for further information

Potential selection of CB unit Confirmatory & high resolution HLA typing Confirmatory & additional microbiology screening

Confirmed selection Confirm identity and functional viability of CB unit using a sample stored as an integral attachment to the freezing bag

Report volume cell content microbiology status of mother and CB unit bacteriology status of CB unit ethnic background sex

Maternal sample HLA typed Confirmatory testing for microbiology with increased sensitivity and wider range of organisms as requested by TC e.g. NAT for HIV, HCV, CMV

STR analysis in parallel with original CB sample CFU analysis

Fig. 28.2 Tissue typing and search mechanisms. BM, bone marrow; CB, cord blood; CFU, colony-forming unit; CMV,

cytomegalovirus; HCV, hepatitis C virus; NAT, nucleic acid technology; NCC, nucleated cell count; STR, short tandem repeat; TC, transplant centre.

(HBsAg), and a test for syphilis plus nucleic acid technology (NAT) for HCV. Additional testing may be undertaken at the time of selection of a unit for transplantation to include the requirements of the transplant centre, e.g. toxoplasmosis, Epstein– Barr virus. • Extra mandatory tests may have been stipulated in the period since collection or by a non-UK country. • Delay between donation and selection may permit the use of the latest testing technology. • Additional sensitive test methods, i.e. NAT for HIV, can be used as an alternative to recall and resampling of the donor to narrow the serological window period. • All CB units are screened by NAT for CMV 324

using an archive sample of CB; less than 0.3% of infants would be expected to be infectious for CMV. Retesting of the donor: should this be undertaken?

It has been suggested that the donor should be recalled and retested 6 months after donation. There are alternative strategies to reduce the risks. The advantages and disadvantages are summarized in Table 28.5.

Related CB banking The advantages and disadvantages of the use of CB

Cord blood banking Table 28.4 Ensuring safety of cord blood (CB).

Procedure

Purpose

Donor interview, by a trained member of staff including medical, behavioural and ethnic history

Donor selection to reduce the risk of transmission of either genetic or infectious disease

Neonatal haemoglobinopathy results from regional screening programme

To identify haemoglobin abnormalities not recognizable at birth, e.g. sickle cell disease

Cytogenetic reports from regional services

To receive reports of identified genetic abnormality

Contact with other health professionals including the general practitioner if required (mothers are also invited to report any medical complications occurring after donation)

To receive information about infectious, genetic or malignant disease occurring or identified after donation

Thorough cleansing of the umbilical cord prior to aseptic collection Dedicated collection room Closed system processing where possible or use of controlled environment and laminar flow hoods Aerobic and anaerobic bacteriology testing of final CB product

Minimizing bacterial contamination

Microbiological screening of the mother and cord Additional testing using high-sensitivity techniques at the time of selection of CB unit for transplant

Reduce microbiological transmission risk Mandatory microbiological markers

Overwrapping of the unit to be frozen

Prevent cross-contamination of CB in storage

over BM in the related setting where a child has a disease that may require transplantation in the future are given in Table 28.6. The National Blood Service operational and policy details of the directed programme are given in Fig. 28.3.

Summary of clinical results of CB transplantation The first CB transplant was in 1972: multiple CBs, transfused to a patient treated for leukaemia, resulted in a temporary graft. A successful sibling CB transplant for Fanconi’s anaemia using modern conditioning therapy in 1988 paved the way for other sibling transplants and then for unrelated transplants. It has been estimated that over 2500 related and unrelated CB transplants have been undertaken worldwide. Results of more than 1000 unrelated CB transplants, in both children and adults, have been published. About 1 L of BM is usually harvested, but with CB only one-tenth of that volume is generally

Table 28.5 Resampling the donor.

Advantages Detect viral window-period of donors Disadvantages The experience of large CB banks has shown that significant numbers of mothers do not reattend unless there is a statutory requirement to do so. Retesting of BM donors after harvesting, freezing and quarantine period is not undertaken for a similar reason as well as for time constraints Potential wastage of a valuable clinical resource and the loss of units where the donor is not resampled would reduce the availability of units in which considerable investment has been made The chance of a seronegative but HIV-infected CB is minute compared to BMT-related mortality or the risk of not having a donor The cost of recall and resampling could instead be invested in greater numbers of collections Donor archives can be used to retest using more sensitive techniques at the time of final selection of a unit for transplantation without recalling the donor BMT, bone marrow transplantation; CB, cord blood.

325

Chapter 28

Referral Clinician caring for potential recipient makes a written referral

Referrals are not accepted from patients or families

Permission of obstetrician Information and consent forms sent to obstetrician to obtain permission for mother to have collection undertaken during delivery

Counseling of mother Carried out by referring obstetrician

Need for mandatory microbiology screening If tests are positive then collection will not take place Details of collection in utero Successful collections cannot be guaranteed Written consent

Collection NBS facilitates collection

Liaison with midwifery staff providing equipment, disposables, detailed instructions and training Telephone support Arrange necessary tests Cryopreserve & store CB

Fig. 28.3 Directed CB banking. NBS, National Blood Service.

Table 28.6 Advantages and disadvantages of directed

related cord blood (CB) banking. Advantages No risk to the mother or baby In the case of urgent transplants, avoids the delay involved of waiting for a matched sibling donor to grow large enough to donate BM Avoidance of liability. A clinician who does not refer a case for directed CB collection could be held liable for the omission if an untoward event prevents that individual from being a BM donor or if that individual suffered an adverse reaction to subsequent BM harvest Disadvantages Cost of collecting units that have a 1 in 4 chance of being an HLA match and of holding units that may never be used, e.g. units held for the potential recipient who has good-risk ALL and may never require transplantation No categorical advantage in transplant outcome to using related CB over BM has been demonstrated Wastage associated with the storage of units that are not a tissue-type match or which derive from an individual affected by the same genetic disease as the intended recipient ALL, acute lymphoblastic leukaemia; BM, bone marrow.

326

available. Because success varies with the amount of material transplanted (although the absolute threshold dose for successful engraftment has not been established) the majority of CB recipients have so far been children in whom a larger dose, in terms of cells transplanted compared with body weight, can be achieved. However, adult patients have been successfully engrafted with CB, including a 90-kg patient. In the largest series published to date, of 562 patients who received unrelated grafts from a single CB bank, and of whom 18% were over 18 years of age, 61% had survived by day 100. In this series, the overall incidence of transplantrelated events other than relapse correlated with a number of factors, including the recipient’s diagnosis, age, dose of cells and the matching achieved. A pattern is emerging of disease types and patient selection criteria that will permit the identification of those patients in whom this type of therapy has a poor outcome and those in whom successful outcomes are very much more likely. These various series have demonstrated the efficacy of CB transplantation in some patient groups, particularly small children, including where only

Cord blood banking

partial matching has been possible between donor and recipient, offering a feasible alternative to BM for patients with both malignant and nonmalignant disease. Unrelated CB transplants may give good results in children with leukaemia if they are at standard risk and transplanted in remission and also in children with inborn errors of metabolism or immune deficiencies. This is despite the fact that published survival rates reflect early work, where there was a predominance of poor-risk cases, as is often seen in early trials of innovative therapies. Improvements in techniques are constantly being introduced, for example a new method for thawing and washing donations increases the speed of myeloid engraftment.

The future A small number of publications have compared the outcomes of unrelated BM and CB transplants. Bar the logistic and biological differences, data suggests that for children with acute leukaemia results are comparable between BM and CB. Several strategies to overcome the low cell dose issues are being investigated: ex vivo expansion, multiunit transplants, infusion of mesenchymal stem cells or transplant after non-myeloblative preparative regimens. There are hopes that laboratory expansion of CB stem cells may lead to their increased use for transplantation in adults and to gene therapy. Recent consideration has been given to performing stem cell transplants in utero, and CB may be the appropriate material to use. In the mean time, CB can offer a potential source of haematopoietic stem cells for children for whom there are no alternatives. CB may also offer a repository of non-haemopoietic stem cells for a variety of indications, although work in this field is still at an early phase.

Summary CB banking is ideally undertaken in the context of a blood transfusion service. It offers the opportunity for finding the equivalent of a BM match for children, particularly for those from the ethnic minorities. The risks of transmissible disease specific to the use of CB are genetic diseases, undetected at birth, and infectious diseases; careful donor selection and testing strategies can limit these.

Further reading Armitage S. Collection and processing of cord blood units. In: Cohen SBA, Gluckman E, Rubinstein P, Madrigal JA, eds. Cord Blood Characteristics: Role in Stem Cell Transplantation. London: Dunitz, 1999; 129–48. Gluckman E. Hematopoietic stem cell transplants using UCB (editorial). N Engl J Med 2001; 344: 1860–1. Grewal SS, Barker JN, Davies SM, Wagner JE. Unrelated donor hematopoietic cell transplantation: marrow or UCB? Blood 2003; 101: 4233–44. Laughlin MJ, Barker J, Bambach B et al. Hematopoietic engraftment and survival in adult recipients of UCB from unrelated donors. N Engl J Med 2001; 344: 1815–22. Reed W, Smith R, Dekovic F et al. Comprehensive banking of sibling donor cord blood for children with malignant and nonmalignant disease. Blood 2003; 101: 351–7. Rocha V, Cornish J, Sievers EL et al. Comparison of outcomes of unrelated BM and UCB transplants in children with acute leukemia. Blood 2001; 97: 2962–71. Rubinstein P, Carrier C, Scaradavou A et al. Outcomes among 562 recipients of placental blood transplants from unrelated donors. N Engl J Med 1998; 339: 1565–77. Sugarman J, Reisner EG and working group. Consensus statement: ethical issues in umbilical cord blood banking. J Am Med Assoc 1997; 278: 938–43. Warwick RM, Barbara JA. Cord blood donor selection: reducing the risk of disease transmission. In: Cohen SBA, Gluckman E, Rubinstein P, Madrigal JA, eds. Cord Blood Characteristics: Role in Stem Cell Transplantation. London: Dunitz, 1999; 149–67.

327

Chapter 29

Therapeutic apheresis Tim B. Wallington and David J. Unsworth

The word apheresis is derived from the Greek meaning ‘a withdrawal’. Therapeutic apheresis is the treatment of disease by removal from the circulation of cells, plasma or both. The rationale for this is that it will remove or reduce, with consequent improvement, a substance or substances implicated in the pathology of the disease being treated. The feasibility of this simple concept was first demonstrated in 1914. Its wide application needed the development of technology capable of safely removing large volumes of plasma and cells from the circulation and safe replacement fluids. The introduction of plastic blood pack systems in the 1960s allowed a safe manual approach, first shown to be effective in the management of hyperviscosity syndromes. This was soon followed by the development of blood cell separators capable of removing larger volumes safely and quickly. These machines have now achieved a high degree of sophistication, with computer control adding to their safety and technical capability. Since 1960, apheresis has been advocated for a whole range of diseases. The justification for this ranges from randomized controlled trials to the rationalization that no other effective therapy is available and that the disease or complication concerned might be mediated by circulating antibodies, immune complexes or blood cells. Because apheresis has real risks for the patient, is costly and may provide little if any benefit, it is important to understand the strength of the rationale for its use before undertaking treatment. It is equally important to understand that apheresis is usually only a part of a patient’s treatment plan. In most instances, apheresis represents an acute intervention that will be effective only for a few days at most unless accompanied by treatment that 328

suppresses the underlying cause. For example, effective immunosuppression is required if an autoantibody-driven disease is being treated. Apheresis is often used to allow time for that treatment to become effective. It follows that there should be agreed criteria for the number of procedures needed in a given clinical situation.

Cell separators A number of machines are manufactured to carry out therapeutic apheresis procedures. Those most commonly used are listed in Table 29.1. Filtration systems will only perform plasmapheresis, whereas centrifugal systems will also perform leucopheresis and red cell exchange. Two-stage filtration systems are used that conserve smaller plasma molecules such as albumin. They are also designated cascade or systems. They are not popular in European or North American practice. Continuous-flow centrifugation systems are based on a disposable harness in which the blood components are separated in a spinning hollow belt or in compartmentalized chambers. The compartmentalized system involves two sequential steps of separation. In the first chamber, red cells are separated; in the second, platelets and leucocytes are separated from plasma. Machines retain the unwanted blood component for disposal, returning all the others to the patient. In the case of plasmapheresis, retained plasma is replaced with appropriate volume replacement fluids. Replacement is needed through the procedure to maintain fluid balance and is automatically controlled by some machines. For practical purposes, this technology can be regarded as a ‘black box’. Modern

Therapeutic apheresis Table 29.1 Cell separators commonly

Centrifugal systems

used in therapy.

Circular belt Two-stage separation Latham bowl

Filtration systems

COBE Spectra apheresis system* Fresenius AS 104* Baxter Fenwall CS3000* Baxter Fenwall Amicus* Haemonetics V50 pheresis system COBE Centry TPE system*

* Continuous-flow centrifugation.

continuous-flow cell separators are computer controlled, the machines being provided with software tailored to each procedure. An important practical advantage of continuous-flow centrifugation is low extracorporeal circuit volume, not more than 150 mL. For some patients a disadvantage is the need for two points of access to the circulation, one from which blood is drawn and another to which it is returned. Intermittent-flow machines only require one point of access. Blood enters a spinning bowl shaped so that plasma exits first as the bowl fills and overflows into a collection bag. The bowl is filled during the draw cycle, which is complete when it is full of packed red cells. The red cells are then returned to the patient; thus the flow of blood to and from the patient is intermittent. In particular with plasmapheresis, the maximum extracorporeal volume can be large. It is dependent on the capacity of the bowl and the haematocrit of the patient. If the patient’s haematocrit is low, large volumes may be removed in the draw cycle and the machine operator must monitor the patient’s fluid balance carefully. Replacement fluid is normally returned during the return cycle but the draw cycle may be interrupted and fluid replaced if the total volume that will be removed is likely to produce circulatory problems. Machines are also made which exploit affinity columns and remove a single blood component. These have a narrow but specific clinical application. Examples are the dextran sulphate absorption system (Liposorber LA) used for low-density lipoprotein (LDL) apheresis and protein A columns (Prosorba) which bind IgG and IgG-containing immune complexes. Photopheresis is similarly a focused procedure for which purpose-built machinery has been developed. It involves the administration of a photosensitizing substance,

psoralen, which is absorbed by leucocytes. In an extracorporeal circuit leucocytes are exposed to ultraviolet light and then returned to the body. They are only available through tertiary referral to specialized units and are most used in European clinical practice. Well-rehearsed operating procedures and machine safety features are of utmost importance to patient safety during therapeutic procedures. Firstly, patients must be passed as fit for the planned procedure following careful assessment by an experienced physician. Well-trained and experienced machine operators are essential. Added to this, machines have safety features built into them. These should include: • a manual override system to allow operator intervention; • a blood flow monitor, both draw and return, so that the patient’s vein does not collapse nor is blood pumped back when the apheresis cannula is obstructed or dislodged; • in-line air detection to stop the return pump if air enters the circuit; • blood filter to remove any aggregated material that might be reinfused; • an anticoagulant flow monitor with a visible means of monitoring flow; • a transmembrane pressure monitor on filtration devices; and • a system for detecting haemolysed red cells on filtration devices.

Technical aspects Successful treatment by therapeutic apheresis is dependent on: • familiarity both with the procedure and the machine used for it; and 329

Chapter 29

• good access to the venous circulation for both the removal of blood and return of separated components and replacement fluid. Blood can be drawn and returned to forearm veins, especially when a number of days separate procedures. In acutely ill patients requiring a series of daily treatments a rigid double-lumen catheter of at least 16 gauge should be placed in a major vein, usually the femoral or subclavian. Most procedures involve plasma exchange (plasmapheresis). Although not obligatory, flow rates are required in practice that will allow the removal of at least 40 mL/kg body weight of plasma over a period of about 2 h, as determined by the patient’s haematocrit. The main replacement fluid used is human albumin solution (HAS) 4.5%, with usually one-third of the volume as normal saline depending on the patient’s albumin status. Care must be taken not to overload patients with replacement fluid. Some procedures require replacement using fresh frozen plasma (FFP) or cryosupernatant, e.g. for thrombotic thrombocytopenic purpura (TTP), and the blood group compatibility requirements of these components must be remembered. A series of three consecutive daily treatments will reduce clotting factor concentrations to levels where bleeding is a danger in susceptible patients, for example those who have had a recent (within 3 days) renal biopsy. In this case a therapeutic dose of FFP (10–15 mL/kg) should be included in the replacement fluid as the last to be infused. Alternative colloid fluids have been used instead of HAS but are not satisfactory unless exchanges are infrequent.

Complications of plasmapheresis Plasmapheresis has various complications. These occur in up to 10% of procedures; most are mild but rare deaths directly associated with the procedure are recorded. Complications may be divided into machine problems, problems relating to venous access, type of replacement fluids and anticoagulant. • Machine problems are unusual: all systems monitor for air and access pressure, so air emboli 330

have been eliminated and access problems are promptly recognized. Failure of the machine can result in red cell loss of up to 350 mL of blood. • The more serious complications reported in recent surveys are from central catheter placement. Pneumothorax is the commonest and there can be internal bleeding. Catheter problems should always be considered if a patient shows clinical evidence of hypovolaemia, develops chest pain or becomes breathless while undergoing plasmapheresis. • Reactions to replacement fluids are uncommon but can be significant. Reactions to HAS are now rare as the preparations contain lower amounts of significant contaminants than previously, especially of vasoactive kinins. HAS essentially carries no risk of infection as it is pasteurized and it does not increase the citrate return. Dilution of coagulation factors has been mentioned. FFP poses the risk of blood-borne infection (although virus-inactivated products are now available) and allergic reactions to the infused plasma. It also contributes to the citrate load as it contains approximately 14% citrate anticoagulant by volume. • Adverse effects of the citrate anticoagulant almost universally used are particularly common. These result from hypocalcaemia and can include paraesthesiae (particularly perioral), abdominal cramps and, rarely, cardiac dysrhythmias and seizures. Citrate toxicity is usually managed easily by slowing the rate of return and providing extra calcium either orally or occasionally intravenously. Patients with renal failure who are receiving large amounts of citrate during plasma exchange may develop a profound metabolic alkalosis. Patients receiving repeated treatments over a long period of time can lose significant quantities of calcium.

Treatment aspects There are some disorders in which treatment by plasma exchange is supported by high-category evidence such as randomized trials (Table 29.2, column 1), some where it is scientifically rational and the broad consensus of experience is that it is useful (Table 29.2, column 2), some where it is

Therapeutic apheresis Table 29.2 Evidence base for therapeutic apheresis.

Supported by randomized trials* Guillain–Barré syndrome Chronic inflammatory polyneuropathy Peripheral neuropathy associated with monoclonal gammopathy Thrombotic thrombocytopenic purpura Focal crescentic glomerulonephritis with systemic involvement (vasculitis)

Consensus treatment can be helpful

Suggestion that treatment may be helpful

Refuted by randomized trials

Goodpasture’s syndrome Myasthenia gravis Hyperviscosity syndrome Cryoglobulinaemia Rheumatoid vasculitis Cutaneous vasculitis Persistent HELLP syndrome Familial hypercholesterolaemia

Pemphigus vulgaris Autoimmune cytopenias Acute myeloma kidney Acute multiple sclerosis unresponsive to steroids Primary antiphospholipid syndrome FSGS particularly recurring in transplant Anti-Ro/La in pregnancy Systemic vasculitis Cold agglutinin disease

Systemic lupus erythematosus Rheumatoid arthritis (exception possibly protein A columns in selected cases)

* For references see Further reading. FSGS, focal segmental glomerulosclerosis; HELLP, haemolysis, elevated liver enzymes and low platelet count.

speculative, i.e. may be of help but high-category evidence is lacking (Table 29.2, column 3), and some where lack of efficacy has been demonstrated (Table 29.2, column 4). This account concentrates on conditions in columns 1 and 2 and some of potential immediate importance in column 3, leaving the others to the referenced reviews. A summary of positive indications and the management of plasmapheresis is shown in Table 29.3.

Neurological diseases Guillain–Barré syndrome

Three large randomized trials reported in the 1980s including more than 500 patients provided the evidence of the value of plasmapheresis in the treatment of Guillain–Barré syndrome (GBS). These trials showed that patients treated with plasmapheresis improved more rapidly, especially if treatment was started within 1 week of the onset of symptoms. Patients are unlikely to benefit if symptoms have been present for more than 4 weeks. Relapses in symptoms can respond to further plasmapheresis and can be managed as though the disease is new. Subsequently there have been further trials and a recent review of the Neuromuscular Disease Group Register confirms the original findings and guidelines for treatment derived from them.

Usually treatment consists of a course of five to six single plasma volume exchanges, administered over a 10-day period, the first three on consecutive days. The measurable milestones used in the trials, such as the duration of mechanical ventilation, time taken to walk around the hospital bed, and so on, are all achieved some time after plasmapheresis has been completed. However, a proportion of patients improve dramatically early in the course of treatment. This is most likely if treatment is started early. Further plasmapheresis is not usually required once it is clear that the patient is improving. Equally there is no evidence that continuing treatment beyond the initial 10-day period is useful unless there has been improvement followed by clear relapse. Treatment with high-dose intravenous immunoglobulin (see Chapter 12) is also effective. When directly compared with plasmapheresis it is equally effective. Immunoglobulin is usually the first choice for treatment as it is more easily administered and is associated with fewer complications. Plasmapheresis is particularly risky in patients with involvement of the autonomic nervous system, quite common in GBS. There is no trial evidence, although in patients given one treatment (i.e. plasmapheresis) but who continue to deteriorate after allowing reasonable time for improvement (around 25 days), experience suggests it is 331

Chapter 29 Table 29.3 Summary of positive indications for, and management of, plasmapheresis.

Disease

Indications for plasmapheresis

Plasmapheresis management

Acute Guillain–Barré syndrome

Failed to respond to IVIg or severe rapidly progressive disease*

Plasmapheresis ¥5 over 7 days Thereafter according to response

Chronic inflammatory demyelinating polyneuropathy

Failed to respond to IVIg or severe rapidly progressive disease†

Plasmapheresis ¥3 per week for 2–3 weeks

Polyneuropathies associated with paraproteinaemia

Some will respond

Plasmapheresis ¥5 over 14 days Thereafter according to response

Myasthenia gravis

Urgency of response or failure to respond to medical treatment‡

Plasmapheresis ¥5 over 7–10 days Thereafter according to response

Myeloma

Symptomatic hyperviscosity (the measured PV is not a good indicator of need for plasmapheresis)

1–3 plasmaphereses performed over 1–3 days (guide improvement of symptoms) followed by systemic therapy

Myeloma (usually light-chain) with renal failure

Acute renal failure (due to light-chain toxicity, causes eliminated)

Plasmapheresis ¥5 over 7–10 days combined with chemotherapy

Waldenstrom’s macroglobulinaemia

Symptomatic hyperviscosity (the measured PV is not a good indicator of need for plasmapheresis)

1–3 plasmaphereses over 1–3 days (guide improvement of symptoms) followed by systemic therapy Maintenance may be appropriate if the frequency of treatments is not greater than once per month

Thrombotic thrombocytopenic purpura

Neurological symptoms/signs Thombocytopenia Renal failure Haemolysis§

Daily plasmapheresis against FFP/cryosupernatant until stable red cell fragmentation Platelet count and LDH are good objective monitors

Myeloproliferative and lymphoproliferative disorders

Leucostasis with deteriorating neurology, respiratory function

Intensive programme of leucopheresis Patient prescribed hydroxyurea

Sickle-cell disease

Red cell exchange for acute chest syndrome and other life-threatening crises

Single blood volume exchange

Goodpasture’s syndrome

Absolute indication for pulmonary haemorrhage

Intensive plasmapheresis, daily until clinical problem resolving

Acute renal failure where creatinine > 600 mmol/L unlikely to respond Cryoglobulinaemia

Systemic vasculitis Multiple organ failure Cerebral symptoms

Daily for 3 days (2 further treatments on alternative days may be required), then review. Monitor cryoglobulin through treatment Immune suppression

Rapidly progressive glomerulonephritis (pauciimmune)

Acute renal failure, dialysis dependent or clearly progressing to dependency

Daily for 3 days, 2 further treatments on alternative days, then review Immune suppression

Familial hypercholesterolaemia

Rare indication Individual cases should be discussed with specialist

Long-term maintenance plasmapheresis every 2– 6 weeks.Affinity columns are more effective

* Intravenous immunoglobulin (IVIg) is usually the treatment of first choice. It is common practice to consider plasmapheresis for patients who have not responded to immunoglobulin within 25 days of starting that treatment. A patient who has responded and then relapsed should have the same treatment repeated. † IVIg is usually used first and plasmapheresis next if there is no response. Maintenance is with therapy that has worked. Plasmapheresis each 2 weeks as a guide ‡ Severely ill patients, e.g. ventilator-dependent or in preparation for thymectomy. Rarely if inadequate response to standard treatments; no trial evidence to support. § If the diagnosis is firm and renal function is compromised or neurological signs/symptoms are present, plasmapheresis is indicated as a matter of urgency.

332

Therapeutic apheresis

worth trying the other treatment (i.e. intravenous immunoglobulin). Miller–Fisher syndrome, characterized by the triad of ataxia, ophthalmoplegia and areflexia of acute onset, is thought to be a variant of GBS and can be usefully treated with plasmapheresis. Chronic inflammatory demyelinating polyneuropathy

There is trial evidence to support the use of plasmapheresis in chronic inflammatory demyelinating polyneuropathy. Patients who respond do so within 2–3 weeks when treated with three plasmaphereses per week. Equally patients can respond to high-dose intravenous immunoglobulin. Where one treatment does not work the other is worth trying as the trial evidence is that a subgroup of patients respond to plasmapheresis and this only overlaps partially with the group responding to immunoglobulin. If plasmapheresis succeeds, it may also be used for maintenance treatment. The regimen is empirical, just sufficient to maintain improvement; usually an interval of 10–14 days between exchanges will suffice. Peripheral neuropathy associated with paraproteinaemia

The peripheral neuropathies associated with Waldenström’s macroglobulinaemia, myeloma, amyloidosis and monoclonal gammopathies of unknown significance (MGUS) have also been treated with plasmapheresis on the presumption that the paraprotein has a direct toxic effect on nerve function. A double-blind crossover trial provides convincing evidence for justifying treatment in certain patients with MGUS and severe peripheral neuropathy. Patients treated with plasmapheresis did better than sham-treated control subjects, and patients who were crossed over from the sham treatment also improved. Similar data supporting the use of plasmapheresis in the other paraproteinopathies associated with neuropathy are not available.

Myasthenia gravis

In myasthenia gravis, plasmapheresis has such a clear therapeutic effect that the treatment is accepted by consensus without the evidence of a controlled trial. Its place is as an emergency measure to treat patients whose life is threatened by respiratory failure or swallowing difficulties. It is also used to prepare patients for thymectomy. In common with other autoimmune disorders in which an antibody is removed, this tends to rebound, often to higher concentrations, in the 2–3 weeks after plasmapheresis has been completed. In treating myasthenia gravis, plasmapheresis must always be accompanied by an appropriate immunosuppressive regimen if it is to be of longterm benefit. There is no evidence supportive of plasmapheresis as long-term treatment for myasthenia gravis. Multiple sclerosis

This is a condition where desperate treatment measures can seem reasonable and plasmapheresis has been tried. The weight of evidence from trials is that it is not effective, although there is evidence from a double-blind crossover trial that a regimen of seven alternate-day treatments can produce substantial improvement in patients experiencing catastrophic acute inflammatory demyelination unresponsive to steroids. Blood diseases

Blood diseases helped by apheresis include the following: • where cells obstruct vascular flow; • where proteins block flow through viscosity or cryoprecipitation (when there can also be an inflammatory element); • where there is antibody-mediated destruction of the formed elements of the blood; or • the thrombotic microangiopathies. Myeloproliferative and lymphoproliferative disorders

Leucostasis is a function of cell number and cell 333

Chapter 29

type. Myeloid cells, because of their size, are more likely to cause stasis than an equivalent number of lymphoid cells. Unless pulmonary or cerebral leucostasis is severe enough for progressive lifethreatening clinical deterioration, treatment with hydroxyurea in the case of chronic myeloid leukaemia will usually reduce the cell count sufficiently by 24 h, obviating the need for apheresis. If measurable improvement within 4–8 h is required, then leucopheresis is the treatment of choice. Studies have shown that photopheresis offers a longer median survival rate for patients with cutaneous T-cell lymphoma than standard chemotherapy. Hyperviscosity syndromes

Clinically evident and progressive hyperviscosity syndrome is a medical emergency requiring urgent plasmapheresis to lower the concentration of the responsible paraprotein. IgM, the largest immunoglobulin and nearly 100% intravascular, is most likely to cause hyperviscosity. IgA and IgG3 tend to aggregate and are more likely than the other isotypes or subclasses to be associated with hyperviscosity. Two or three treatments will usually alleviate symptoms long enough for chemotherapy to take effect. These patients are often severely anaemic. They should not be transfused until the viscosity has been lowered, as a rise in haematocrit can precipitate a serious worsening of their symptoms. Cryoglobulinaemia

Plasmapheresis can be life-saving in the few cases where cryoglobulins are associated with a fulminant clinical picture. The temperature at which the cryoglobulin precipitates is a major factor in its pathogenic potential and is usually concentration dependent; thus just two or three consecutive exchanges can have a major clinical impact. Return fluids should always be warmed. At the same time the cause of the cryoglobulinaemia must be determined and definitive chemotherapy instituted if appropriate.

334

Myeloma

There is some evidence that acute-onset renal failure caused by myeloma proteins can be improved by lowering the concentration of paraprotein, but this is not firm. If this treatment option is considered, other commoner causes of acute renal failure in myeloma should have been excluded. A multicentre trial of this approach is presently in progress. Autoimmune cytopenia

Despite the role that antibody and immune complexes play in these disorders, there are no well-controlled studies supporting the use of plasmapheresis. More effective alternative treatments are available. In cold agglutinin disease, which is usually IgM based, plasmapheresis is theoretically more likely to help but is rarely in reality an essential part of treatment. TTP and haemolytic–uraemic syndrome

Management of TTP routinely requires either infusion of FFP or, more often, plasmapheresis using FFP or cryosupernatant plasma as the replacement fluid. Instituting adequate treatment is an emergency. A recent direct comparison of plasma infusion versus plasmapheresis with FFP replacement showed a better response to plasmapheresis, which is now the treatment of choice and has resulted in a reduction in mortality from 85 to 15%. Cryosupernatant plasma contains fewer multimers of von Willebrand factor, which may play a part in the pathogenesis of TTP. Similarly, solvent–detergenttreated FFP, which is pooled prior to virus inactivation treatment, contains fewer mulitimers and provokes fewer reactions than FFP, and may well turn out to be the replacement fluid of choice. Logic favours these replacement fluids over singleunit FFP but as yet there is a lack of supportive clinical evidence. Daily plasma exchange is needed for at least 2 days after the platelet count has been normal. A recent study suggests that if enough FFP is given alone (patients received an average of 27.5 mL/kg) it works as well as plasmapheresis but troublesome complications, particularly

Therapeutic apheresis

associated with fluid overload, were significantly commoner in the FFP-only group. Case reports suggest that patients with severe pre-eclampsia, HELLP syndrome (haemolysis, elevated liver enzymes and low platelet count) or both may benefit from plasmapheresis with plasma replacement if they fail to improve after delivery. Sickle cell disease

In patients with sickle cell disease, red cell exchange is used in acute chest syndrome and other life-threatening crises, priapism, and in some instances as preparation for surgery. It may also break a cycle of closely spaced painful crises. The goal is to achieve sickle cell concentrations of less than 30%, with a haematocrit of no more than 46%. Because of the high rate of alloimmunization in sickle cell patients, the use of red cells phenotypically matched as closely as possible with those of the recipient is recommended (see Chapters 9 and 24). Renal diseases

Plasmapheresis has been used widely in the treatment of acute renal diseases but there are currently relatively few true indications for its use. Goodpasture’s syndrome (antiglomerular basement membrane antibody disease)

A place for plasmapheresis in the treatment of this form of rapidly progressive glomerulonephritis (RPGN) was first established in 1974. As with other autoantibody-mediated disorders, success is dependent on a vigorous concomitant immunosuppressive regimen, typically prednisolone (60 mg daily) with cyclophosphamide (3 mg/kg daily). Certain ground rules have been learnt with experience: • plasmapheresis is not indicated for all patients; and • renal function is rarely improved if the creatinine is greater than 60 mmol/L at diagnosis and unless there is another indication, such as pulmonary haemorrhage, plasmapheresis may not help.

Although this condition generally progresses rapidly to renal failure, some patients have mild focal disease and a good prognosis and plasmapheresis is not indicated. Pulmonary haemorrhage requires urgent plasmapheresis. Patients in the initial days of immunosuppression, before the disease is fully under control, should be monitored for occult lung haemorrhage and treated if it occurs. Daily treatments for 10–14 days are usually required. It is important not to overload the patient with replacement fluids, return volumes must be carefully calculated and the patient left a little ‘dry’ (e.g. by 200–300 mL). Fluid overload will provoke pulmonary haemorrhage. A combination of intense plasmapheresis, which removes antibody, and leucocyte-depleting immunosuppression makes these patients very vulnerable to infection. This can be life-threatening and should be anticipated in the patient’s treatment plan. Pauci-immune RPGN

This occurs as part of Wegener’s granulomatosis, microscopic polyarteritis and as a primary disorder of the kidney. Patients usually carry antibodies to neutrophil cytoplasm (ANCA). Trial results are mixed but there is reasonable evidence that plasmapheresis has a place in patients with acute renal failure who are dialysis dependent. Patients need early treatment, starting as soon as the diagnosis is firm, with daily treatments for 3 days and then on alternate days thereafter for up to 2 weeks, until improvement is evident. Concomitant immunosuppression is also necessary, as it is in Goodpasture’s disease. Focal segmental glomerulosclerosis

The observation that this condition commonly recurs after renal transplantation suggests that a serum factor may be involved and that plasmapheresis should help. No controlled trials have been reported but there have been encouraging results particularly in secondary disease following transplantation if plasmapheresis is introduced as soon as proteinuria is detected. A prospective controlled trial is needed to confirm efficacy but 335

Chapter 29

apheresis units are increasingly asked to participate in the treatment of this difficult problem. Connective tissue diseases

Plasmapheresis has little proven utility in the treatment of connective tissue diseases other than in the overlap with renal disease in pauci-immune RPGN. However, it has been much used so some discussion is warranted here. Systemic vasculitis

There is no consistent evidence supporting the use of plasmapheresis in the treatment of systemic vasculitis. However, where the inflammatory process in the vessel wall appears to be driven by immune complexes, as for example in rheumatoid vasculitis and certain cases of cryoglobulinaemia, there is a consensus that plasmapheresis may provide time for standard medical treatment with cyclophosphamide and steroids to work. Exchanges during the acute phase of the illness (daily for 3 days and alternate days thereafter) may halt its progress. Improvement depends on immunosuppression, which takes 7–10 days to take effect. Treatment is reserved for fulminating cases. It is also used in the other systemic vasculitides when they are life-threatening; pulmonary haemorrhage is particularly dangerous, but with less hard evidence in support. Systemic lupus erythematosus

A number of studies, most recently those of the Lupus Plasmapheresis Study Group, which were large and statistically powerful, have not demonstrated benefit. Occasionally in very ill patients plasmapheresis seems justified on the same basis as in systemic vasculitis (see above) and no doubt will continue to be used in that way. Rheumatoid arthritis

There is clear evidence to show that plasmapheresis has no place in the treatment of uncomplicated rheumatoid arthritis. As mentioned, it may have a

336

place as part of the treatment strategy in fulminating cases of rheumatoid vasculitis and those rare cases complicated by cryoglobulinaemia or hyperviscosity syndrome. A recent trial suggests that passing plasma through protein A columns in an extracorporeal circuit can benefit patients with severe progressive disease resistant to multiple drug therapy. Other autoantibody-based disorders

There is anecdotal evidence for therapeutic impact from plasmapheresis in a number of other autoantibody-based disorders, and some have already been discussed. There are others where it is not the treatment of choice but might be considered in specific cases resistant to standard treatment. Pemphigus vulgaris, the primary antiphospholipid syndrome, especially to treat recurrent pregnancy loss and pregnancy complicated by anti-Ro/La, are examples, but intravenous immunoglobulin is a more practical option especially in pregnancy. Familial hypercholesterolaemia

Homozygous familial hypercholesterolaemia is a potentially fatal condition that often responds poorly to treatment with drugs and diet. Plasma exchange can remove the LDLs which accumulate in this disorder. Improved survival was first demonstrated in a study reported in 1991 with two weekly plasma exchanges. Subsequently, procedures more selective in the removal of LDLs, dextran sulphate/cellulose absorption columns and on-line heparin precipitation, have been introduced along with more intensive regimens and are now regarded as the best approach. Referral for specialist advice is required.

Further reading Artero ML, Sharma R, Savin VJ et al. Plasmapheresis reduces proteinuria and serum capacity to injure glomeruli in patients with recurrent glomerulosclerosis. Am J Kidney Dis 1994; 23: 574–81. Coppo P, Busse A, Charrier S et al. High-dose plasma infusion versus plasma exchange as early treatment of

Therapeutic apheresis thrombotic thromocytopenic purpura/haemolytic uraemic sundrome. Medicine (Baltimore) 2003; 82: 27–38. Dyck P, Daube J, O’Brien P et al. Plasma exchange in chronic inflammatory demyelinating polyradiculopathy. N Engl J Med 1986; 314: 461–5. Dyck P, Low PA, Windebank AJ et al. Plasma exchange in polyneuropathy associated with monoclonal gammopathy of undetermined significance. N Engl J Med 1991; 325: 1482–6. Guillain–Barré Syndrome Study Group. Plasmapheresis in acute Guillain–Barré syndrome. Neurology 1985; 34: 1096–104. Korach JM, Berger P, Giraud C, Le Perff-Desman C, Chillet P. Role of replacement fluids in the immediate complications of plasma exchange. Intensive Care Med 1998; 24: 452–8. Leblond PF, Rock G, Herbert CA. The use of plasma as a replacement fluid in plasma exchange. Transfusion 1998; 38: 834–8. Levy JB, Pusey CD. Still a role for plasma exchange in rapidly progressive glomerulonephritis? J Nephrol 1997; 10: 7–13.

Lewis EJ, Hunsicker LG, Lan SP, Rohde RD, Lachin JM. A controlled trial of plasmapheresis in severe lupus nephritis. N Engl J Med 1992; 326: 1373–9. Rock G, Buskard N. Therapeutic plasmapheresis. Curr Opin Hematol 1996; 3: 504–9. Rock GA, Shumak K, Buskard NA et al. and the Canadian Apheresis Study Group. Comparison of plasma exchange with plasma infusion in the treatment of thrombotic thrombocytopenic purpura. N Engl J Med 1991; 325: 393–7. Smith JW, Weistein R, Hillyer KL. Therapeutic apheresis: a summary of current indication categories endorsed by the AABB and the American Society for Apheresis. Transfusion 2003; 43: 820–2. Thompson GR, Kitano Y. The role of low density lipoprotein apheresis in the treatment of familial hypercholesterolemia. Ther Apheresis 1997; 1: 3–16. Weinshenker BG, O’Brien PC, Petterson TM et al. A randomized trial of plasma exchange in acute central nervous system inflammatory demyelinating disease. Ann Neurol 1999; 46: 878–86. Winters JL, Pineda AA. New directions in plasma exchange. Curr Opin Hematol 2003; 10: 424–8.

337

Part 5

Developments in transfusion medicine

Chapter 30

Blood substitutes Chris V. Prowse and David J. Roberts

Collecting and fractionating human blood for medical use is an expensive and time-consuming process. Large donor panels must be recruited and tested to maintain a constant supply of safe, phenotyped cellular and protein fractions of whole blood. Collection and processing of blood are complex procedures. Moreover, blood transfusion carries risks and has significant, and in some cases unavoidable, adverse effects. There are obvious attractions to the potential replacement of transfusion of cellular components with alternative products that do not have the same dependence on a readily available blood donor population, can be treated to reduce infectious and non-infectious risks, do not require crossmatching, and have a less restrictive shelf-life than the current red cell and platelet components provided by transfusion services (Table 30.1). Such products would be of particular interest in battlefield and emergency situations and the armed services have been a major funder of research in this field. Despite this and research programmes that stretch back to the earlier half of this century there are, as yet, no licensed products in this field, other than one haemoglobin solution in South Africa. An alternative approach of ‘virtual blood substitutes’ to achieve the desired effect without transfusion is described below. In broad terms there are three categories of blood substitute under development: • products that are still based on the use of donorderived blood cells (human or animal); • synthetic products that achieve the same end point by mirroring the function of the natural product or by novel mechanisms; • ‘virtual’ blood substitutes (Table 30.2),

using growth factors to stimulate endogenous haemopoiesis or drugs to secure haemostasis. The outstanding ‘virtual’ blood substitutes are the haemopoietic growth factors that can stimulate production of red cells and platelets and mobilize white cells and stem cells. Increasing the effectiveness of circulating platelets using 1deamino-8-D -arginine vasopressin (DDAVP) or recombinant factor VIIa, inhibiting fibrinolysis by tranexamic acid or e-aminocaproic acid, or securing haemostasis by the use of fibrin sealant are well-established methods of reducing bleeding and through avoiding red cell and/or platelet transfusion are classic ‘virtual’ blood substitutes. This chapter discusses the ‘real’ red cell and platelet substitutes in development. The virtual blood substitutes are covered in Chapters 6, 7 and 31. Understanding the potential role of blood substitutes and the practical and theoretical obstacles to their introduction into clinical practice provides illuminating lessons about the physiology of blood and modern biotechnology.

Red cell substitutes Modified haemoglobin-based blood substitutes

Red blood cells have a number of functions beyond oxygen and carbon dioxide transport, including: • modulation of oxygen delivery under conditions of low pH and/or high PCO 2 (the Bohr effect); • encapsulation of haemoglobin to prolong circulating half-life; • modulation of vascular tone via effects on nitric oxide (NO) concentrations; • reduction of methaemoglobin. 341

Chapter 30 Table 30.1 Potential ‘real’ blood substitutes.

Table 30.2 Virtual blood substitutes.

Red blood cells Cross-linked haemoglobin tetramers Recombinant haemoglobin tetramers Polymerized haemoglobin Conjugated haemoglobins Encapsulated haemoglobin Perfluorocarbons

Red blood cells Erythropoietin

Platelets Freeze-dried platelets Infusible platelet membranes Fibrinogen-coated microspheres Peptide-coated red cells Glycoprotein receptor-carrying liposomes Megakaryocytes In vitro expansion of megakaryocytes White blood cells In vitro generation of antiviral and antitumour cytotoxic lymphocytes In vitro generation of dendritic cells Stem cells In vitro expansion of stem cells

These functions depend on a complex and elegant interplay between the haemoglobin molecule, red cell enzymes, the internal milieu and the red cell membrane. Perhaps, not surprisingly, the higherorder functions of the red cell have proved difficult to mimic in artificial components. Early attempts to transfuse purified unmodified haemoglobin did show that oxygen-carrying capacity could be restored. However, transfusion of unmodified haemoglobin causes a number of problems as follows. • Isolated tetramers are unstable and dissociate to globin dimers and monomers. As the tetramers dissociate, the allosteric cooperativity and the modulation of oxygen affinity by bound 2,3diphosphogylcerate (2,3-DPG) are lost, giving a reduced oxygen-carrying capacity. The P50 (partial pressure of oxygen at which haemoglobin is halfsaturated with oxygen) is reduced from 26 mmHg to less than 10 mmHg (Fig. 30.1a). This may be less critical than previously thought as it has recently been shown that molecular size may be more critical than P50 for delivery of oxygen to peripheral tissues. 342

Leucocytes Antibiotics, antiviral and antifungal agents Active immunization G-CSF and GM-CSF Platelets Thrombopoietin Pegylated recombinant human megakaryocyte growth and development factor (MGDF) Interleukin 11 Thrombopoietin mimetics: microbial peptides Haemostatic and pharmacological agents Aprotinin DDAVP e-Aminocaproic acid and tranexamic acid Recombinant coagulation factor VIIa Fibrin sealants DDAVP, 1-deamino-8-D -arginine vasopressin; G-CSF, granulocyte colony-stimulating factor; GM-CSF, granulocyte–macrophage colonystimulating factor.

• Globin chains, and to some extent tetramers, are filtered by the kidneys and precipitate in the renal tubules, causing renal dysfunction. • Isolated tetramers transit the vascular endothelium and by reducing NO availability in the extravascular compartment cause vasoconstriction and oesophageal spasm. Several modifications have been made to free haemoglobin tetramers to overcome these problems. Currently, several second-generation red cell substitutes, including intramolecularly crosslinked haemoglobin, conjugated haemoglobin and polymerized haemoglobin, are in clinical trials and the third generation of substitutes of artificial red blood cells is under development and at the stage of animal trials (for summary see Table 30.3). In the USA, the Food and Drug Administration have stated that they will only consider licensing red cell substitutes for three indications: • regional perfusion, e.g. percutaneous transcoronary angioplasty, enhancing radiation therapy of tumours; • acute haemorraghic shock; • for use in the perioperative period.

Blood substitutes

(a) Normal vein

Normal artery

100 Myoglobin

Saturation (%)

80

60

40

Haemoglobin: pH 7.6 ( P CO2 = 25) pH 7.4 ( P CO2 = 40) pH 7.2 ( P CO2 = 61)

20

Intramolecularly cross-linked haemoglobin Di-aspirin cross-linked haemoglobin

The cross-linking of haemoglobin tetramers with bis-(3,5-dibromosalicyl) fumarate yields di-aspirin cross-linked haemoglobins, with a high P50 for good oxygen delivery (e.g. Hemassist). However, haemoglobin tetramers still cause significant smooth muscle spasm, leading to oesophageal spasm and increases in blood pressure. Phase III trials of this product in both trauma and perioperative settings have demonstrated that it can reduce allogeneic transfusion by 19%, but the product has now been withdrawn due to an excess death rate in the trauma trial.

0 0

20

(b)

O2 content mL/dL

25

40 60 P O2 (mmHg)

80

100

Blood 14 g/dL

20

sion

mul

e FOB

15

P

Hgb 7 g/dL

As a result, the majority of clinical trials have been in the latter two fields, for which avoidance of allogeneic transfusion is accepted as a surrogate end point if mortality is not used. In such trials, a haemoglobin ‘trigger’ level usually determines whom to transfuse. It is of interest to remember that a recent systematic review has demonstrated that reducing trigger levels can itself reduce allogeneic transfusion by up to 40%.

10 Fluosol-DA

5

Recombinant haemoglobin

0 0

100

200

300

400

500

P O2 (mmHg) Fig. 30.1 (a) Oxygen affinity of haemoglobin tetramers and

monomers. Oxygen dissociation curve of myoglobin or dissociated haemoglobin monomers compared with that of haemoglobin at three pH values. PO 2 and PCO2 , partial pressure of oxygen and carbon dickide. (b) Oxygen affinity of perfluorocarbons. Comparison of oxygen-carrying capacity of whole blood, haemoglobin solution, and fluorocarbons. Whole blood with a haemoglobin content of 14 g/dL possesses an arterial oxygen content of 20 mL/dL at a PO 2 of 100 mmHg. In contrast, fluorocarbon emulsions carry less oxygen at a given partial pressure of oxygen. A 90% perfluoro-octylbromide (PFOB) emulsion can carry 10 mL of oxygen at a PO 2 of 300 mmHg. Perfluorodecalin (Fluosol-DA 20), which used early emulsification technology to achieve a 20% fluorocarbon emulsion, can only carry 2–3 mL/dL of oxygen at a PO 2 of 300 mmHg.

Large-scale production of recombinant haemoglobin in Escherichia coli and yeast has been established by Somatogen, who were purchased by Baxter in 1998. Using recombinant DNA technology the a-globin chains were fused to yield an undissociable ‘tetramer’. It was also possible to engineer haemoglobin molecules to reduce NO affinity. Baxter has recently announced withdrawal from this development. Human haemoglobin has also been produced in transgenic pigs, but it proved difficult to separate from the endogenous porcine protein. Polymerized haemoglobin

Haemoglobin may be cross-linked by bifunctional chemicals to form polymers or haemoglobin molecules can be directly linked to a high-molecularweight non-protein carrier. In either form renal

343

Chapter 30 Table 30.3 Red cell substitutes under trial or development.

Product/company

Current status†

Intramolecularly cross-linked haemoglobin Di-aspirin cross-linked haemoglobin Recombinant haemoglobin Polynitroxylated haemoglobin tetramers Sebacoyl-linked haemoglobin tetramers

Hemassist, Baxter Healthcare (USA) Optro/rHb2.0, Somatogen Inc. with Baxter (USA) Hemozyme, SynZyme (USA) OxyVita IPBL Pharmaceuticals

Failed phase III trials Shelved Preclinical Preclinical

Polymerized haemoglobin Glutaraldehyde cross-linked haemoglobin Glutaraldehyde cross-linked bovine haemoglobin* O-Raffinose cross-linked haemoglobin

Polyheme, Northfield (USA) Hemopure, Biopure (USA) Hemolink, Hemosol (Canada)

In phase III trials Completed phase III trials Failed phase III trials

PHP,Apex Bioscience (USA) Enzon (USA) Hemospan, Sangart (USA) PolyHb-SOD-CAT, McGill University

In phase III trials In phase Ib/II trials In phase II trials Preclinical

Hemotech, HemoBiotech Inc.

Preclinical

Terumo (Japan), US Navy

Preclinical

Oxygent,Alliance (USA) Adverse effects concern Oxycyte, Synthetic Blood/International Inc.

In phase III trials Phase II orthopaedic surgery trial

Conjugated haemoglobin Polyoxyethylene–haemoglobin Polyethylene glycol–bovine haemoglobin Polyethylene glycol–human haemoglobin Bovine haemoglobin polymer containing superoxide dismutase and catalase Covalent complex bovine haemoglobin with GSSG, adenosine and ATP Encapsulated haemoglobin Liposome-encapsulated haemoglobin Perfluorocarbons Synthetic Perflubron/emulsifer

* Licensed in South Africa for anaemia therapy. GSSG, oxidized glutathione. † Current status is described in clinical trial phase: I is volunteer safety study, II is pilot patient safety and efficacy study, III is pivotal patient efficacy study.

filtration and smooth muscle dysfunction may be reduced. The oxygen-carrying capacity, reduced by the loss of 2,3-DPG binding, may be restored by other modifications. Three forms of cross-linked polymerized haemoglobin are on trial. Glutaraldehyde cross-linked haemoglobin

Human haemoglobin has been cross linked with glutaraldehyde and pyridoxal phosphate added to the 2,3-DPG pocket to increase P50 (Polyheme, Northfield, Inc.). Clinical trials mainly in trauma patients show that the product is safe and efficacious, reducing transfusion in the first 3 days of hospitalization. Reduction in mortality has also been demonstrated, particularly in patients refus344

ing standard transfusion on religious grounds. These trials used transfusions up to the equivalent of 20 units of red blood cells without serious ill effects. A pivotal trial of this product in 500 at-thescene trauma patients is planned. The second polymerized product is a glutaraldehyde cross-linked bovine haemoglobin (Hemopure). This product is licensed in South Africa and a similar product is already licensed for canine use. Phase III trials have been undertaken in various types of elective surgery using up to 10 units showing, for example, a reduction of allogeneic transfusion of 27% in vascular surgery. An unpublished trial of nearly 700 patients in orthopaedic surgery forms the main basis for a licence application currently under consideration in the USA.

Blood substitutes

O-Raffinose cross-linked haemoglobin

The third form of polymerized haemoglobin is one with oxidized O-raffinose cross-linking, which produces a haemoglobin polymer with a high P50. However, the product contains biologically significant amounts of cross-linked haemoglobin tetramers, which can and do cause smooth muscle spasm in the gastrointestinal tract. Trials showing a reduced use of allogeneic blood in orthopaedic and cardiac surgery have been published, but pivotal trials in both indications have recently been curtailed due to an excess rate of myocardial infarction in the patients in the cardiac trial. Conjugated haemoglobin

Polymeric haemoglobin may also be made by cross-linking haemoglobin not to itself but to highmolecular-weight polyoxyethylene (PHP, Apex Bioscience) or to polyethylene glycol (PEG-Hb, Enzon Inc.; Hemospan, Sangart Inc.). These methods increase the half-life of the preparations and reduce NO-mediated vasoactivity. The Apex Bioscience and Enzon products have been at trial in sepsis and to improve solid tumour radiation therapy. Hemospan is unusual in having a deliberately low P50 to prevent the release of oxygen until the haemoglobin reaches the capillaries, and phase I trials have shown it lacks the vasoactivity of most other preparations that can result in smooth muscle spasm. The product is now entering phase II trials in orthopaedic surgery. Other adaptations under development include cross-linking superoxide dismutase and catalase directly to polymerized haemoglobin to reduce oxygen radical formation and subsequent reperfusion injury. Artificial red blood cells: encapsulated haemoglobins

The third generation of haemoglobin-based red cell substitutes would be artificial red blood cells. Chang and colleagues pioneered artificial red blood cells using lipid bilayers. The modern formulations have used phospholipid vesicles (0.2 mm in diameter) with sialic acid analogues added to the membranes to reduce clearance by the reticuloendothelial system. Further improvements to

microencapsulated haemoglobin under investigation are: • inclusion of catalase and superoxide dismutase to reduce oxygen radical and methaemoglobin formation; and • use of biodegradable polylactides and polyglycolides in artificial membranes to increase haemoglobin concentration to 15 g/dL in small nanometer diameter vesicles. These third-generation haemoglobin substitutes are at the early stage of animal trials. It seems possible that artificial erythrocytes may mimic some of the complex higher-order functions of ‘real’ red cells in the not too distant future.

Clinical use of haemoglobin-based red cell substitutes

The real dangers of ‘natural’ red blood cells are low and the safety of substitutes has to be proven in large-scale trials if they are to be accepted for everyday use. In the last 5 years two major developments, using di-aspirin and O-raffinose crosslinked haemoglobin (Hemassist and Hemolink), have failed trials at the last hurdle. A glutaraldehyde cross-linked haemoglobin (Polyheme) remains on trial in trauma patients and the equivalent bovine product (Hemopure) awaits a licensing decision for use in elective surgery in the USA. Haemoglobin polymerized by cross-linking to polyethylene glycol (Hemospan) uses a promising novel approach and is in preliminary clinical trials. Apart from Hemopure, all these products are still reliant on standard blood donations and their potential advantages will have to outweigh their additional marginal cost.

Perfluorocarbons Principle

Liquid perfluorocarbons (PFCs) are synthetic hydrocarbons in which most of the hydrogen atoms have been substituted by fluorine atoms. The low intermolecular attractions result in a high capacity to dissolve gases such that the oxygen content of a PFC is up to 20 times that of water. These chemicals have inherent limitations, 345

Chapter 30

including a short intravascular half-life (<12 h), insolubility in water requiring emulsification with surfactants, and limited oxygen-carrying capacity (the amount of oxygen carried is directly proportional to the inspired oxygen concentration; see Fig. 30.1b). This requires that patients breathe oxygen-rich air, limiting their use to operating rooms and intensive care settings. First-generation fluorocarbons

Fluosol-DA 20, an emulsion of 20% perfluorodecalin, is the only oxygen-carrying volume expander licensed in the USA. It was initially hoped it would gain widespread use but trials showed no efficacy in patients who refused blood transfusions. The only indication for which it has been approved is percutaneous transluminal coronary angioplasty, although some trials showed no benefit in combination with tissue plasminogen activator (tPA) over tPA alone. The inherent limitations of this perfluorocarbon are compounded by the adverse effects, which include: • marked uptake by the reticuloendothelial system; • disruption of pulmonary surfactant leading to ventilation–perfusion defects in the lungs; and • complement activation resulting in anaphylaxis. This product is no longer easily available. Oxygent

This second-generation PFC, based on perfluorooctylbromide (PFOB), has been trialled by Alliance Pharmaceutical Company, and contains egg-yolk phospholipids as emulsifier. This composition confers several advantages over previous products, including: • greater oxygen-carrying capacity (see Fig. 30.1b); • reduced or absent complement activation; • reduced interference with pulmonary surfactants; and • improved stability and shelf-life. Trials have been performed in a number of perioperative settings, most notably in cardiac surgery in conjunction with acute normovolaemic haemodilution (ANH). Such trials have shown a 346

delay in the time to reach the trigger levels for allogeneic transfusion, but a large pivotal trial was recently suspended due to concerns about excess rate of stroke. This has now been ascribed to overenthusiastic ANH rather than the use of the PFC, and the company is now hoping to recommence trials. Adverse effects of flushing and flu-like symptoms and delayed fever, headaches and nausea, as a result of macrophage activation, and a transient thrombocytopenia occur in some patients and may limit clinical applications. The small size of PFCs (~0.2 mm) has suggested that they may improve oxygenation in ischaemic or infarcted tissues or increase oxygenation in tumours and so enhance sensitization to radiotherapy or chemotherapy. Other second-generation PFCs, such as Oxycyte, are also at an earlier stage of development.

Platelet substitutes Platelet concentrates are widely used in the management of thrombocytopenia and abnormal platelet function. These products have allowed the development of chemotherapy regimens that cause prolonged absence of platelet production and have made extracorporeal bypass a safe routine procedure. However, both the supply and use of fresh platelets pose particular problems due to storage being limited to 5 days as a result of gradual loss of efficacy and the risk of bacterial contamination. Supply also requires the maintenance of large wellcharacterized donor panels and specialized centres for apheresis procurement. Repeated platelet transfusions are frequently accompanied by the development of antiplatelet antibodies, usually directed against major histocompatibility complex (MHC) class I antigens or against other platelet surface antigens. Artificial platelet substitutes hold the promise of avoiding these logistic, technical and medical problems and so achieving cheaper, safer and more readily available therapy for thrombocytopenia. However, as for red cells, replacement of the natural product has not been straightforward. Attempts to replace platelets can again be divided into ‘real’ and ‘virtual’ platelet substitutes. Virtual platelet substitutes range from improved clinical

Blood substitutes

guidelines and their implementation (Chapter 6), through drugs that may reduce blood loss (Chapter 7) to compounds that stimulate platelet production (Chapter 31). Although not strictly speaking a platelet substitute, the development of pathogen reduction technologies for platelets may eliminate bacteria, viruses and leucocytes from this product, so reducing transfusion-transmitted infection, febrile non-haemolytic transfusion reactions and transfusion-associated graft-versus-host disease. Substitutes for platelets have not yet been licensed but several products are under development (see Table 30.1). The most promising are summarized below. Platelet membrane preparations

In the search for an alternative to fresh platelet concentrates, freeze-dried platelets were initially shown to be superior to frozen and thawed platelets in tests of haemostasis in vitro. Freezedried platelets were subsequently shown to be as effective as stored platelets in vitro and to provide haemostasis in thrombocytopenic animals. Clinical evaluation is planned. Compared with platelet concentrates, freeze-dried platelets have the apparent advantage of reduced viral and bacterial load as a result of paraformaldehyde treatment. However, they have some disadvantages such as: • they must be made from fresh platelets; and • they may still stimulate an alloimmune response. Infusible platelet membranes are derived from stored platelets as membrane fragments which seem to promote haemostasis without causing thrombosis in animals. They are the only platelet substitute to have undergone clinical trial where they were shown (in 1999), in a small number of patients, to be effective in individuals refractory to standard platelet transfusion studies. The advantages of infusible plasma membranes over platelet concentrates include: • reduced viral and bacterial load; • reduced expression of human leucocyte antigen (HLA) class I antigens; and • may be made from outdated platelets.

Synthetic platelets

Beyond the manipulation of platelet membranes, the search for a useful substitute for platelet concentrates has led to a totally synthetic approach. Microspheres of human albumin coated with human fibrinogen (SynthocytesTM, ThrombospheresTM) reduce bleeding time and acute blood loss in thrombocytopenic animals. They have no immediate toxicity in rodents or primates. Fibrinogen-coated microspheres would have the advantages of: • sterility; • production independent of platelet concentrates; and • absence of HLA class I and platelet surface alloantigens. Interestingly, these microspheres appear to promote the formation of a platelet plug by interacting with residual normal platelets (Fig. 30.2). It seems likely that both lyophilized platelet and infusible plasma membranes may also function in a similar manner. Liposomes with inserted platelet receptors are also under investigation as a platelet alternative. The efficacy of lyophilized platelets, infusible platelet membranes and fibrinogen-coated microspheres in the prophylaxis of bleeding in severely thrombocytopenic patients will require careful evaluation, and there has been little progress in this field since the first edition of this book. More immediate applications for these platelet substitutes may be in improving haemostasis where the platelet count is moderately reduced and as alternative or adjuvant therapy where patients have become refractory to platelet transfusions through alloimmunization.

Summary Real red blood and platelet substitutes have yet to reach the clinic. Simple substitutes lack the more complex and important function of whole cells. Nevertheless, modification of haemoglobin by chemical technology has provided products that have been successful or unsuccessful in phase III clinical trials, with one now awaiting a license 347

Chapter 30

Fig. 30.2 Artificial platelet substitutes (Synthocytes™). Electron micrograph showing the interaction of Synthocytes and normal platelets on a collagen surface.

decision in the USA. The real benefits and safety of these substitutes remain to be shown in clinical practice. One PFC has also reached the stage of pivotal clinical trial. Progress with platelet substitutes has been much less apparent. Synthetic microspheres that provide platelet-like activity may be free of viral contamination and polymorphic molecules, but would seem unlikely to be as effective as fresh platelets with the possible exception of treating haemorrhage, for example in patients with immune platelet refractoriness with no compatible donors. So, while non-toxic substitutes with reasonable biological activity are likely to be available, it is far from clear whether they will replace cells derived from donors for the majority of clinical uses. At the risk of making speculative assessments, it seems more likely that real blood substitutes will find small niche applications and that virtual blood 348

substitutes and improved prescribing will reduce the use of donor-derived products.

Further reading Blajchman MA. Novel platelet products, substitutes and alternatives. Transfus Clin Biol 2001; 8: 267–71. Kitaguchi T, Murata M, Iijima K, Kamide K, Imagawa T, Ikeda Y. Characterization of liposomes carrying von Willebrand factor-binding domain of platelet glycoprotein Ibalpha: a potential substitute for platelet transfusion. Biochem Biophys Res Commun 1999; 261: 784–9. Levi M, Friederrich PW, Middleton S et al. Fibrinogencoated albumin microcapsules reduce bleeding in severely thrombocytopenic rabbits. Nat Med 1999; 5: 107–11. McCarthy MR, Vandegriff KD, Winslow RM. The role of facilitated diffusion in oxygen transport by cell-free hemoglobins: implications for the design of hemoglobinbased oxygen carriers. Biophys Chem 2001; 92: 103–17.

Blood substitutes Prowse CV. Alternatives to human blood and blood resources. Vox Sang 1998; 74 (Suppl. 2): 21–8. Reid TJ. Hb-based oxygen carriers: are we there yet? Transfusion 2003; 43: 280–7. Scigliano E, Enright H, Telen M et al. Infusible platelet membrane for the control of bleeding in thrombocytopenic patients. Blood 1997; 90 (Suppl. 1): 267a (abstract 1170).

Spahn DR. Artificial oxygen carriers: status 2002. Vox Sang 2002; 83 (Suppl. 1): 281–5. Vandegriff KD, Malavalli A, Wooldridge J, Lohman J, Winslow RM. MP4, a new nonvasoactive PEG-Hb conjugate.Transfusion 2003; 43: 509–16. Winslow RM. Blood substitutes: refocusing an elusive goal. Br J Haematol 2000; 111: 387–96.

349

Chapter 31

Cytokines in transfusion practice Derwood H. Pamphilon

Introduction Cytokines are soluble or membrane-bound factors that stimulate the growth of cells and tissues, including cells of the lymphoid and haemopoietic systems. They belong to a number of cytokine superfamilies and act as lineage-specific, multilineage, stem cell or accessory factors (Table 31.1). Many cytokines are produced using recombinant DNA technology and are available for clinical use in vivo or in ex vivo systems. There are two principal ways in which cytokines are relevant to transfusion medicine. 1 They may be used to shorten the periods of cytopenia that occur in a number of disease states. For example, the anaemia of chronic renal failure may be treated with erythropoietin (EPO), and patients with neutropenia following chemotherapy for cancer can be treated with granulocyte colony-stimulating factor (G-CSF) to accelerate the return of neutrophils to the bloodstream. Thus, cytokines are used to avoid the need for transfusions of red cells, platelets and perhaps also granulocytes. This helps to minimize the risks associated with transfusion of blood components such as microbiological contamination, acute lung injury and others discussed in previous chapters. 2 They may be used to facilitate the generation of dendritic cells and cytotoxic T lymphocytes (CTLs) that are used in the immunotherapy of cancer and viral infections. The use of cytokines to mobilize peripheral blood stem cells (PBSC) into the bloodstream also reduces the requirement for transfusion since PBSC are associated with a shorter engraftment time than bone marrow stem cells (see Chapter 33). The use of cytokine-generated immunothera350

peutic cells such as dendritic cells does not impact directly on routine aspects of transfusion practice. However, it has become clear that blood centres have an important role to play in utilizing their expertise in good manufacturing practice for the production of such cells. This chapter is devoted mainly to the use of cytokines to reduce transfusions of donor blood. In particular it concentrates on those cytokines most commonly used in clinical practice, i.e. EPO for the treatment of anaemia, G-CSF for direct treatment of neutropenia and for mobilization of PBSC, and thrombopoietin (TPO) to shorten periods of thrombocytopenia. There are a number of ways in which cytokines could reduce periods of cytopenia. • In patients: by direct stimulation of endogenous blood stem cell production. • In patients: by indirect stimulation of proliferation of blood cell progenitors harvested by apheresis and cultured ex vivo prior to reinfusion. • From volunteer donors: by stimulating the growth and differentiation of harvested progenitors cultured ex vivo, sometimes referred to as blood cell ‘farming’. • From volunteer donors: by stimulating the numbers of circulating blood cells of a particular lineage, e.g. TPO to increase the platelet count in a donor who then donates these by apheresis.

Erythropoiesis Red cell development occurs during a 2-week period. The earliest erythroid precursor is the burst forming unit, erythroid (BFU-E). This is small and without specific cytological features. It expresses

7p15

Interleukin 6 Interleukin 11

IL-11

19q13.3–13.4

2q13

13q12–13

Interleukin 1

Flk2/Flt3 ligand

FL

12q2–24

5q23–31

5q23–31

5q31

Synergistic with SCF, IL-3 on primitive progenitors Synergistic with IL-3, SCF. Myeloma cell growth factor Also synergistic with SCF and IL-3

Synergistic with a range of cytokines on primitive and committed haemopoietic progenitors

Has synergistic effects on primitive precursor cells with IL-3, IL-6, IL-11

Acts on all CFCs in vitro

Acts on all lineages at colony-forming cell (CFC) stage

Eosinophil stimulator

Stimulates platelet production

Stimulates erythropoiesis

7q11–22

3q27–28

Monocyte stimulation

Produces granulocyte colony-forming units in in vitro cultures

Action

1p13–21

17q11.2–21

Chromosome

IL-1D IL-1E IL-6

Steel factor, stem cell factor, kit ligand

SCF

Granulocyte– macrophage colonystimulating factor

Interleukin 5

IL-5

GM-CSF*

Thrombopoietin

TPO

Multi-CSF

Monocyte colonystimulating factor Erythropoietin

M-CSF (also known as CSF-1) EPO*

IL-3

Granulocyte colonystimulating factor

G-CSF*

Name(s)

* Growth factors licensed for clinical use in the UK. BMT, bone marrow transplantation; PBSC, peripheral blood stem cells.

Accessory or synergistic factors

Stem cell factors

Multilineage factors

Lineage-specific factors

Factor

Table 31.1 Cytokine superfamilies.

Trialled as megakaryocyte stimulator, but not clinically effective Megakaryocyte stimulator Licensed in USA for post-chemotherapy thrombocytopenia

None as yet

Clinical trials bedevilled with adverse events due to mast cell activation in vivo (asthma, etc.) Likely use will be in vitro in haemopoietic stem cell expansion protocols Likely to be confined to in vitro cell expansion

Actions on mast cells and basophils produce problematic adverse effects Likely value in in vitro culture applications in future Mobilization of PBSC Similar applications to G-CSF but more adverse events Possible role in management of fungal infections and wound healing

Anaemia of renal failure Some effect in anaemia of malignancy As adjunct to autologous transfusion regimens Platelet collection May speed platelet recovery after chemotherapy/BMT Immunogenicity of one TPO analogue has slowed clinical development Research applications only

Mobilization of peripheral blood stem cells Increases neutrophil counts after chemotherapy and BMT Granulocyte collection (see Chapter 9) No clinical applications

Application/clinical use

Cytokines in transfusion

351

Chapter 31

the CD34 antigen found on most haemopoietic stem cells. Maturation occurs into the colonyforming unit, erythroid (CFU-E), which has deeply basophilic cytoplasm and represents about day 7 of development, immediately prior to the time of haemoglobinization. Following this, cell size reduces and the nucleus is extruded. The cytokines interleukin (IL)-3 and stem cell factor (SCF) are intimately associated with the first 7–9 days of erythroid differentiation. The predominant erythropoietic cytokine is EPO, which acts on erythroid progenitors from day 2 onwards and is the only cytokine capable of differentiating BFU-E. EPO is a glycoprotein growth hormone largely produced by the peritubular capillary endothelium of the kidney in response to reduced oxygen content in the renal arterial circulation. Some EPO is also produced by hepatocytes. It binds to receptors on the surface of erythroid progenitors. EPO is available in the UK as short-acting epoetin alfa or beta (Eprex, Janssen-Cilag; NeoRecormon, Roche) respectively and the long-acting darbepoetin alfa (Aranesp, Amgen); the latter is achieved by modifying the glycosylation of the molecule. Clinical indications for EPO therapy

EPO is given to prevent red cell transfusion in patients with: • chronic renal failure; • human immunodeficiency virus (HIV)-infected patients treated with zidovudine; • cancer patients on chemotherapy; • prior to surgery to increase the procurement of autologous blood and so reduce exposure to allogeneic transfusions. In chronic renal failure, the rate of rise of the haematocrit is dependent on time, the dose of EPO given and individual patient variation. For example, an EPO preparation given at a dose of 50–150 units/kg caused transfusion independence in more than 95% of patients and many qualityof-life parameters were significantly improved. EPO preparations are indicated and licensed for the treatment of anaemia related to therapy with zidovudine in HIV-infected patients. Studies show that it is effective in reducing the transfusion requirement and increasing the haematocrit in 352

patients with low endogenous serum EPO levels compared with patients who receive placebo. An additional benefit is that patients may maintain a satisfactory haemoglobin without the need for a significant reduction in their zidovudine dose. EPO has been used to treat patients with anaemia secondary to malignant diseases such as multiple myeloma and in patients with myelodysplastic syndromes. It also may reduce the anaemia and transfusion requirement in patients who receive chemotherapy for cancer. Better oxygen delivery to tumours may have a radiosensitizing effect. There is evidence that EPO preparations help to maintain a more stable haemoglobin, avoiding the peaks and troughs that occur in patients who receive chemotherapy and supportive transfusions alone. This may be associated with better quality-of-life parameters. Patients scheduled for major elective surgery, such as hip replacement, may be treated with EPO to increase the haematocrit and allow more intensive scheduling of autologous blood donations. This is discussed in greater detail in Chapters 6 and 26. Patients usually receive iron to prevent depletion of iron stores following EPO therapy. This reduces the number of units of allogeneic blood required and may avoid the requirement for allogeneic blood transfusion. Risks of EPO therapy

• Iron deficiency due to rapid expansion in the number of erythroid cells. • Increased blood pressure. • Increased blood viscosity. • Pure red cell aplasia: this is an important although uncommon adverse effect. It has occurred largely with the Eprex preparation. It appears that the administration of EPO causes the formation of anti-EPO antibodies, which crossreact with the recipient’s own EPO and lead to destruction of erythropoietic cells. Ex vivo culture of erythroid progenitors

At the present time, the large-scale culture of stem cells to produce red cells suitable for transfusion is not practical, although it is possible to differentiate

Cytokines in transfusion

mature red cells from precursors in the laboratory. However, there is considerable interest in the culture of red cells from progenitor cells donated by patients with rare red cell phenotypes. This would represent a major advance in the transfusion support of patients with disorders such as sickle cell disease. These patients frequently have phenotypes unusual in Caucasian blood donor populations and form multiple red cell alloantibodies.

Thrombopoiesis The search for the major haemopoietic growth factor responsible for maintaining normal platelet levels in the blood had continued for many years before the isolation of TPO in 1994. A number of other cytokines are important in platelet development. IL-3, G-CSF, granulocyte–macrophage (GM)-CSF and SCF act at the progenitor cell stage. IL-6 acts late in megakaryocyte maturation and TPO together with IL-11 stimulate all stages of megakaryocyte development. TPO is also known as c-Mp1 ligand and megakaryoctye growth and development factor (MGDF). IL-3 and IL-6 seem unlikely to have a clinically useful effect because of excessive toxicity and insufficient efficacy. A summary of the influence of cytokines on platelet development is shown in Fig. 31.1.

Thrombopoietin

TPO is a glycoprotein of molecular mass 30 kDa. The gene encoding it is located on chromosome 3q27. The amino-terminal residues have 21% sequence identity and 46% overall sequence similarity with human EPO and it is this domain that binds to c-Mpl (the TPO receptor). Recombinant human TPO (rTPO) is a full-length polypeptide while rMGDF, also used in clinical studies, is a truncated protein containing the receptor-binding region modified by addition of polyethylene glycol (PEG). TPO is primarily synthesized in the liver, while lesser amounts are seen in the kidneys, brain and testes. Circulating TPO levels are inversely related to platelet mass. Platelets contain an avid TPO receptor that removes it from the circulation; therefore normal or high platelet levels prevent the action of TPO on the bone marrow by binding to circulating receptors. This is clinically relevant because platelet transfusions may blunt the recovery of megakaryocytes. In addition, reduced clearance of TPO by abnormal platelets may account for the high platelet counts seen in myeloproliferative syndromes such as essential thrombocythaemia. In bone marrow failure states, e.g. aplastic anaemia, TPO levels are high whereas in idiopathic thrombocytopenic purpura they are low or normal, since TPO is rapidly removed from the circulation by platelets with a short circulation span.

Megakaryocytopoiesis Committed Pluripotent progenitor stem cell cell

Mature mk Platelets

Immature mk

IL-11 IL- 3 Fig. 31.1 Influence of cytokines on

platelet development. G-CSF, granulocyte colony-stimulating factor; GM-CSF, granulocyte– macrophage colony-stimulating factor; IL, interleukin; SCF, stem cell factor.

Thrombopoietin

SCF

G-CSF GM-CSF

IL-6

353

Chapter 31

Risks of TPO

• • • • •

Thrombocytosis. Thrombosis. Marrow fibrosis. Veno-occlusive disease. Interaction with other growth factors. Thrombocytopenia has been observed in some individuals treated with recombinant PEGMGDF. In some cases, thrombocytopenia was severe, causing easy bruising and heavy menses. In the bone marrow there was a marked reduction in megakaryocytes and all patients had antibody to PEG-MGDF that cross-reacted with endogenous TPO and neutralized its biological activity. Thus TPO administration may cause the production of anti-TPO antibodies, causing immunemediated thrombocytopenia. In one case, pancytopenia was reported in a patient who developed neutralizing antibodies to TPO after PEGMGDF administration. This complication has significantly slowed the development of TPO as a therapeutic agent. Clinical trials of TPO Solid tumour chemotherapy Both rTPO and PEG-MGDF have been administered with few immediate adverse effects. If given prior to chemotherapy, there was marked stimulation of platelet production. Subsequent studies showed that the duration of thrombocytopenia was shorter and the nadir of platelet count higher in patients given TPO compared with control during a first cycle of chemotherapy. Stem cell transplantation In bone marrow autografting the duration of thrombocytopenia is shortened when TPO is administered. The benefit of TPO is less clear in PBSC transplantation since the duration of thrombocytopenia is much shorter. TPO has not been found to be of benefit in patients with delayed haemopoietic recovery or engraftment failure. TPO given in conjunction with other cytokines such as G-CSF may increase the mobilization of stem cells into the peripheral blood.

354

Use in clinical haematology and oncology A recent study investigating the use of PEG-MGDF in patients with acute myeloid leukaemia undergoing chemotherapy failed to show any significant differences in the number of days of platelet transfusions between either the treatment group or those patients receiving placebo. Moreover, in patients given a schedule of PEG-MGDF from day –6 to day +6 of chemotherapy, the granulocyte count recovery was delayed compared with those receiving placebo. This and other studies emphasize the need for careful consideration of dose and scheduling of TPO in each clinical situation. At present, TPO therapy is not routine and further trials are needed to establish whether it is a costeffective therapeutic option. Transfusion medicine TPO has been given to apheresis donors to elevate the platelet count and increase the number of platelets that might be harvested in a single apheresis procedure. Studies carried out in the USA have reported an increased yield of platelets following TPO administration resulting in higher increments in transfused patients. However caution should be exercised when giving cytokines to donors because of adverse effects (see above), particularly severe thrombocytopenia, that have occurred in volunteer donors given PEG-MGDF. Secondly, TPO might be used ex vivo in the culture of platelets, so-called platelet ‘farming’, using healthy stem cells derived from volunteer donors. This is an area of interest but has not yet produced a platelet product suitable for clinical use. In addition, studies have investigated the use of TPO and other cytokines to stimulate the production of intermediate-stage megakaryocyte progenitors from stem cells collected after cytokine (e.g. GCSF) administration. A portion of the collected product was cultured to increase the number of megakaryocyte progenitors. When these were infused into the recipients, there was a shortening of thrombocytopenia by 1–2 days. Another approach that has been utilized is to collect platelets by apheresis in the ‘rebound’ phase when the blood count recovers after chemotherapy. At this time, TPO levels are high. These may then

Cytokines in transfusion

be cryopreserved and transfused when thrombocytopenia develops after further cycles of chemotherapy. Interleukin 11

Although a number of other recombinant cytokines have stimulatory effects on platelet production, most either have no beneficial effects in vivo or they have had serious adverse effects. IL-11 has passed clinical trials and is approved for use by the Food and Drug Administration in the USA for secondary prophylaxis against thrombocytopenia following chemotherapy. It has a number of adverse effects, including fluid retention, arrhythmias and visual disturbance. Roughly 1% of patients develop antibodies to it.

Granulopoiesis The production of mature granulocytes from stem cells is controlled by SCF, IL-3, GM-CSF and GCSF. G-CSF is a glycoprotein encoded for by a gene on chromosome 17. Recombinant G-CSF is indicated for the treatment of neutropenia after chemotherapy and stem cell transplantation, mobilization of PBSC, severe congenital neutropenia, cyclic and idiopathic neutropenia when there is a history of severe infection, and persistent neutropenia in advanced HIV infection. It may be given by subcutaneous injection or intravenous infusion, stimulating: • granulocytic precursors to produce mature granulocytes; • more primitive precursors, so that CD34+ cells are released into the bloodstream. Clinical uses of G-CSF

• Shortening post-chemotherapy neutropenia. • Granulocyte collection from normal donors. • Ex vivo culture of committed granulocyte progenitors (similar in concept to megakaryocyte progenitors, see above) for in vivo infusion after myeloablative therapy. • Stem cell mobilization.

Risks of G-CSF

G-CSF is a non-toxic agent which, as a naturally occurring hormone, is present in health at low levels. Common adverse effects include: • bone pain; • headache; • myalgia; • fatigue; • nausea and vomiting; • minor alterations in blood chemistry. The first three of these occur in up to 80% of healthy people given G-CSF and can be treated with regular paracetamol. Leucocytosis may be marked and can increase blood viscosity, causing leucostasis. When it is given to healthy people the dose should be omitted if the neutrophil count exceeds 80 ¥ 109/L. There is a theoretical risk of stimulating the growth of leukaemic cells in patients with myeloid malignancy. In patients with severe congenital neutropenia treated with G-CSF a proportion go on to develop myelodysplasia and acute myeloid leukaemia. However this is associated with the development of cytogenetic abnormalities and occurs also in patients who do not receive cytokines. No case of myeloid malignancy has been reported in normal individuals who have received G-CSF for stem cell mobilization. Granulocyte collection

Patients receiving chemotherapy and stem cell transplantation for cancer often have prolonged periods of neutropenia and are susceptible to lifethreatening bacterial and fungal infections. Granulocytes collected by apheresis of healthy donors or made from buffy coat preparations have been used to prevent or treat such infections. However, the dose administered is usually only about 5–10 ¥ 109/L. Moreover, increments in the patient’s granulocyte count are rare and doubt has been cast on the clinical efficacy of such an approach. Corticosteroids elevate the peripheral granulocyte count and increase by two to four times the dose of granulocytes that can be collected. Priming donors with a single dose of G-CSF increases the 355

Chapter 31

average dose collected to 50–100 ¥ 109/L and often good count increments are seen. There is now considerable interest in this approach but further prospective randomized studies of both prophylactic and therapeutic therapy are required before definite recommendations can be made. In the mean time it is reasonable to give granulocytes collected by apheresis from mobilized donors to patients with severe neutropenia and bacterial or fungal infections unresponsive to 2 days or more of appropriate antimicrobial therapy. G-CSF is given to donors as a single dose, alone or plus a single dose of dexamethasone 12–24 h prior to apheresis. Stem cell collection from healthy volunteer donors

Roughly half of all allogeneic stem cell transplants from sibling donors use mobilized PBSC. There are in excess of 8 million donors on the Bone Marrow Donors Worldwide (BMDW) database at the present time and many registries now sanction the use of G-CSF for stem cell collection by apheresis. Ordinarily, the cytokine is given for 4 or 5 days with careful donor monitoring and then one or two apheresis procedures are carried out (see Chapter 29 for further details).

Lymphopoiesis The cornerstone of this is the production in culture systems or in vivo of ‘educated’ CTLs that kill either tumour or virus-infected cells. This can be

356

done for example by taking peripheral blood from a patient and culturing the mononuclear cell fraction to generate dendritic cells that are effective in presenting antigen to T cells. These are then cultured (‘pulsed’) in the laboratory with some form of tumour antigen and incubated with T cells from the patient. There is proliferation of cells in response to antigen recognition, with production of tumour-specific CTLs. These can then be reinfused to the patient. Alternatively, the dendritic cells may be primed with antigen and injected, causing the proliferation of CTLs in vivo. Growth factors for dendritic cells include SCF, IL-4, GMCSF, anti-CD40 ligand and transforming growth factor-b. The principal cytokine that drives T-cell growth ex vivo is IL-2 (see Chapter 33 for more details of immunotherapeutic approaches).

Further reading Eagleton HJ, Littlewood TJ. Colony stimulating factors for chemotherapy induced febrile neutropenia. Update on the clinical use and misuse of erythropoietin. Curr Hematol Rep 2003; 2: 109–15. Kaushansky K. Use of thrombopoietic growth factors in acute leukemia. Leukaemia 2000; 14: 505–8. Kuter DJ, Begley CG. Recombinant human thrombopoietin: basic biology and evaluation of clinical studies. Blood 2002; 100: 3457–69. Lyman G, Castro LG, Djulbegovic B. Cochrane Database Systemic Review 2003; 3: CD003039. Macdougall IC. Erythropoietin and renal failure. Curr Hematol Rep 2003; 2: 459–64.

Chapter 32

Haemopoietic stem cell processing and storage David H. McKenna and Mary E. Clay

Over the past several years haemopoietic stem cell (HSC) transplantation has become an increasingly viable option in the treatment of a growing number of malignant and non-malignant diseases. Essential to the success of such transplants are reliable methods of HSC processing, cryopreservation and storage. The goal of this chapter is to familiarize the reader with these methods while introducing related topics within the field of cell and tissue engineering. The following items are discussed: • sources of HSCs; • processing methods; • quality control testing techniques; • cryopreservation and storage; • quality assurance and good manufacturing practice; • regulations and standards.

Sources of HSCs There are three sources of HSCs for clinical transplant: bone marrow, peripheral blood and umbilical cord blood (Table 32.1). Bone marrow

The traditional source of HSCs is bone marrow (BM). Use of BM has decreased over recent years as other sources of HSCs with definite advantages have become available. Despite this decline, however, a role for BM in HSC transplantation does still exist. Target dosage (typically 2–4 ¥ 108 nucleated cells/kg recipient body weight) is efficiently collected in a single procedure. The harvesting procedure involves aspiration of BM

(10–15 mL/kg recipient weight) at the posterior iliac crest under full general, spinal or epidural anaesthesia. Following filtration (typically sequential in-line filters of decreasing filter pore size) the collection is free of bone spicules and other debris and ready for any necessary further processing. Peripheral blood

Studies performed over 30 years ago showed HSCs to be in the peripheral blood (PB) at very low concentrations. The subsequent discoveries of haemopoietic growth factors (granulocyte– macrophage colony-stimulating factor, GM-CSF, and granulocyte colony-stimulating factor, G-CSF) and their role in mobilization of HSCs from BM led to development of apheresis strategies to collect HSCs from the mononuclear cell fraction of blood. PB has become the major source of HSCs at many transplant centres in both the autologous and allogeneic setting. The PB collection can be less problematic for the donor, as there is no need for anaesthesia or hospitalization. Venous access (peripheral or central) is utilized in a procedure that lasts roughly 4 h. Higher numbers of HSCs, relative to typical BM harvests, can be collected, facilitating engraftment. Many centres use rising peripheral white blood cell count and/or peripheral CD34+ cell enumeration (usually 10–12 CD34+ cells/mL) to determine the most efficient collection plan (i.e. date, process volume or length of collection). Using this strategy, most healthy G-CSF-stimulated allogeneic donors need just one collection to reach the target dose. In the autologous setting more collections are typically required. The same apheresis strategies (without growth 357

Chapter 32 Table 32.1 Sources of haemopoietic stem cells (HSCs). (Modified from Smith 2000.)

Source

Characteristics

Bone marrow

Original source of HSCs; decreasing in use Used predominantly in allografting (occasional autograft procurement for HSC back-up) Requires operating room harvesting procedure to collect 10–15 mL/kg recipient weight (nucleated cell target dose typically 2–4 ¥ 108/kg) Advantages include need for only one procedure for full collection; a possible advantage is the relatively lower T-cell content

Peripheral blood

Widely used in the autologous setting; surpassing bone marrow as the primary allogeneic HSC source in many/most transplant centres Requires mobilization of stem cells in the patient/donor with either chemotherapy, haemopoietic growth factors, or both (CD34+ cell target dose typically 5 ¥ 106/kg) Advantages include collection without anaesthesia/hospitalization, more rapid engraftment and, in the autologous transplant setting, possibly lower tumour cell contamination

Umbilical cord blood

Collected from the placenta (in utero or ex utero) following delivery Stored either for eventual use by the family or placed into a bank for transplant into unrelated patient Minimum CD34+ cell dose: 1.7–2.0 ¥ 105/kg Advantages include decreased incidence of graft-versus-host disease, decreased search time, and reduced histocompatibility matching requirements

factor stimulation) may be utilized to harvest PB mononuclear cells for use in the processing/engineering of a variety of other cellular therapies. These therapies, e.g. donor lymphocyte infusion, natural killer cells, dendritic cells, and antigen (i.e. cytomegalovirus, Epstein–Barr virus, tumour, etc.)-specific T-cell immunotherapies, may serve as adjunctive or supportive therapy post-HSC transplant or as therapies altogether independent of HSC transplant. Donor lymphocyte infusions (DLI) commonly serve as post-allogeneic HSC transplant therapy for the prevention of leukemia relapse or loss of engraftment. DLI are collected from the original HSC donor. Upon completion of the apheresis procedure, an aliquot from the mononuclear cell product is drawn for lymphocyte quantitation. Dosage is based on lymphocyte content and is calculated using automated or manual differential or flow cytometric analysis (CD3+ cell content), as determined by the institutional protocol. Other than an occasional volume reduction for a small recipient, there is no further processing necessary in the manufacturing of a DLI. 358

Umbilical cord blood

Umbilical cord blood (UCB) is the final source of HSCs for transplantation. Since the first clinical transplantation in 1988 into a child with Fanconi’s anemia, use of UCB for haemopoietic reconstitution has steadily increased. Once regarded as biological waste, UCB has been demonstrated to contain HSCs with higher proliferative and selfrenewal capacity than those of BM and PB. Due to dose limitations, however, most patients deemed appropriate for UCB transplantation are children or small adults. Several strategies have been initiated to overcome limitations of dose, including transplantation of two or more units and ex vivo expansion. UCB is collected by obstetricians or dedicated staff before (in utero) or following delivery (ex utero) of the placenta. A typical UCB collection measures 50–200 mL. Bacterial contamination rates, which historically approached 10–20%, have decreased markedly since collection techniques have improved. Most collections now are by ‘closed-system’ involving umbilical vein cannulation or venepuncture and direct drainage into a

Stem cell processing and storage Table 32.2 Routine haemopoietic stem

cell (HSC) processing procedures. (Modified from Law 2000.)

Procedure

Application

Volume reduction (plasma depletion)

Reduction of incompatible plasma (minor ABO mismatch) Prevention of volume overload in recipient Concentration of cells for cryopreservation

Red blood cell depletion

Reduction of incompatible red cells (major ABO or other antigen mismatch) Maximization of storage space Limitation of infusion of lysed red cells and free haemoglobin (cryopreserved products)

Buffy coat preparation

Maximization of storage space Debulking of red cells prior to further manipulation

Thawing, washing and filtration

Preparation of HSC products prior to infusion (see text)

plastic blood bag. The unit then undergoes further processing and cryopreservation and is stored in liquid nitrogen for eventual use by the family (autologous or related allogeneic) or an unrelated recipient. In addition to higher potency relative to BM and PB, UCB has several more advantages over these other sources of HSCs. Decreased incidence of graft-versus-host disease and reduced histocompatibility matching requirements are attributable to the naive immunological state of the lymphocytes within the UCB. It follows that patients with rare HLA types are often successful in finding a suitable graft when other sources are not acceptable. Finally, search time is generally decreased, as units are banked and already HLA typed.

Processing methods HSC processing involves both routine methods and more specialized complex methods. The routine methods utilize concepts and equipment well known to the blood banking profession. Cellular components and plasma are separated based on physical properties (e.g. size, density) using various reagents (e.g. hydroxyethyl starch for red blood cell sedimentation), centrifugationbased instruments (e.g. Spectra Apheresis System and COBE 2991 Cell Processor, both from Gambro.BCT), and plasma expressors (Table

32.2). The more specialized methods involve concepts, reagents and instruments unique to cell processing/cellular engineering (Table 32.3).

Routine methods Volume reduction

Volume reduction, or plasma depletion, is accomplished with centrifugation and is commonly performed to reduce incompatible plasma (antibody load) in the case of minor ABO-mismatched allografts (BM and PB). Despite plasma compatibility, volume reduction may be necessary in the transplantation of patients with current or potential issues of fluid balance/overload (e.g. pediatric and small adult patients, renal/cardiac failure patients). Volume reduction may also be employed to reduce product volume or increase cell concentration for purposes of cryopreservation. Red blood cell depletion

Sedimenting agents (e.g. hydroxyethyl starch, HES) facilitate removal of red blood cells. Red blood cell depletion is necessary with major ABOincompatible and other clinically relevant red blood cell antigen (e.g. Kell, Kidd)-incompatible BM allografts to prevent haemolytic transfusion reactions. Furthermore, in the case of cryopreserved HSCs, red cell depletion prior to freezing 359

Chapter 32 Table 32.3 Specialized haemopoietic stem cell (HSC) processing procedures. (Modified from Law 2000.)

Procedure

Application

Counterflow centrifugal elutriation

Separation of cell populations by cell size and density Uses include T-cell depletion of HSC grafts and monocyte enrichment for further manipulation

Cell selection systems

Positive and negative cell selection Uses include HSC (CD34+ or CD133+ cells) enrichment with consequent T-cell depletion and a variety of other cell type enrichments and depletions based on cell surface antigen expression (see text)

Cell expansion

Expansion of HSCs and progenitors in an effort to enhance engraftment and long-term outcome Utilizes culture systems designed to mimic in vivo microenvironment

Other

T-cell depletion and HSC graft purging by monoclonal antibody-based methods and/or physical characteristics HSC graft purging by pharmacological agents

limits the amount of lysed red cell fragments and free haemoglobin infused. Red blood cell depletion additionally accomplishes reduction of overall volume, which is useful when storage space is a concern (e.g. UCB banking). Red cell depletion is not necessary when PB is the source of HSCs; apheresis collections usually result in less than 20 mL of red blood cells due to efficiency of instrumentation. Buffy coat preparation

Buffy coat concentration of BM involves centrifugation and harvesting of the white blood cell fraction. When the red blood cell volume is large enough this procedure can be performed with an apheresis or cell washing device using semi-automated processing; HSC loss is minimal with such systems. In instances where red cell volume is too low for machine processing, manual centrifugation suffices. Buffy coat preparation can be used for volume reduction for cryopreservation or as an initial red cell debulking step subsequent to further, more complex procedures (e.g. CD34+ cell selection). Thawing of cryopreserved HSCs

The final processing step prior to infusion of HSCs that have been cryopreserved is the thaw procedure. Depending on institutional policy, this proce360

dure may take place in the laboratory or on the patient care unit. The thaw procedure for all HSCs, regardless of source, is similar. Although quite simple, proper execution is essential, as frozen plastic containers may be prone to breakage for a variety of reasons. The product should be carefully removed from the storage container and inspected to evaluate the integrity of the bag. Following label verification of product identity by two technologists, the unit should be gently placed and tightly sealed within a clean or sterile plastic bag and submerged in a 37ºC water bath. The thaw should be performed relatively quickly to prevent recrystallization and consequent cell damage/ death. Gentle kneading of the contents helps to accelerate the process. If a leak is discovered, the site of the break should be determined, and a haemostat should be used to prevent loss of the product. The contents of a broken or leaking bag should be aseptically diverted into a transfer bag, and an aliquot should be sent for culture. Although it is now common practice to deplete a UCB unit of red cells prior to cryopreservation, many institutions continue to perform a post-thaw wash step. Washing serves to remove the minimal lysed red cells, haemoglobin, and dimethyl sulfoxide (DMSO). Many institutions base their UCB processing methodology, including the wash procedure, on that originally described by Dr Pablo Rubinstein of the New York Placental Blood Program. Between institutions there may be slight

Stem cell processing and storage

modifications of this simple procedure with regard to centrifugation (i.e. rpm, duration of spin, etc.), concentration of solutions used for dilution/resuspension, and so on. Briefly, the thaw involves slow sequential addition of a wash solution (e.g. 10% dextran, then 5% albumin), transfer into a bag of appropriate size for centrifugation, and resuspension of cell pellet(s) before delivery to the patient care unit for infusion. Filtration of HSCs

The issue of filtration of HSCs with a standard blood filter (170 mm) deserves brief mention. As noted, BM typically undergoes sequential filtration in the operating room or in the laboratory to remove aggregates/debris. However, opinions regarding use of standard blood filters prior to or upon infusion of HSCs from PB or UCB do vary from institution to institution. The decision to use a standard blood filter rests with the individual cell processing laboratory and/or transplant centre. It is recommended, of course, that the laboratory validate their filtration process prior to making filtration a standard policy. Specialized methods

Specialized cell processing methods generally serve to optimize product purity and potency beyond levels obtained through routine methods and, as mentioned previously, use unique reagents and instrumentation. Elutriation

Counterflow centrifugal elutriation is a specialized method that separates cell populations based on two physical characteristics: size and density (sedimentation coefficient). A centrifuge is used to separate cell populations of an HSC product based on density. Fluid/media is passed through the chamber housing the cells in the direction opposite (counterflow) to the centrifugal force. Adjusting the flow rate of the fluid/media and/or the speed of centrifugation to enable counterflow rate to balance centrifugal force allows for alignment of cells based on sedimentation coefficient. A given

cell population can then be diverted as a fraction of the initial product. Beckman Coulter has a few high-performance centrifuge systems that have elutriation applications, and Gambro.BCT is currently developing a system called the Elutra. T-cell depletion has been the traditional application of counterflow centrifugal elutriation, as lymphocytes are readily removed (2–3log depletion) from an allogeneic HSC collection. Using this method a specific dose of T cells can be given to a patient as a means to reduce graft-versus-host disease while maintaining engraftment and graftversus-malignancy effect. Although T-cell depletion is the primary application of elutriation, great potential for further applications exists, including monocyte enrichment for dendritic cell generation and subsequent tumour vaccine production. Cell selection systems

Cell selection systems incorporating monoclonal antibody-based technologies that target cell surface antigens (Isolex 300i Magnetic Cell Selection System, Baxter Healthcare Corp. and CliniMACS system, Miltenyi Biotec) have become the method of cell depletion/enrichment at many institutions. These specialized immunomagnetic methods involve isolation of the cell type of interest by either positive selection (target cells retained) or negative selection (target cells depleted). Cell products manufactured by these methods attain an unequalled level of purity. The Isolex 300i is limited to positive selection of HSCs (CD34+ cells). Selection is accomplished in four steps. Murine-derived anti-CD34 monoclonal antibody (primary antibody) solution is mixed with cells in suspension. Following washing of unbound antibody, sheep anti-mouse IgG (secondary antibody)-coated paramagnetic polystyrene beads (Dynabeads M-450) are added, and CD34+ cell–bead rosette complexes are formed. A magnetic field is then applied to the chamber allowing CD34+ cell–bead complexes to be separated from the rest of the suspension. Finally, a releasing agent, an octapeptide that acts through competitive displacement, is added to separate beads/antibodies from the CD34+ cells. Selection of CD34+ cells with the Isolex 300i achieves an approximately 85% 361

Chapter 32

purity of enriched cells and an approximately 50% CD34+ cell yield with a simultaneous, indirect T-cell depletion (3–4log). Miltenyi Biotec technology involves direct capture of target cells. Specific monoclonal antibodies are coupled to 50-nm ferromagnetic particles, comprising the so-called MicroBeads. Magnetically labelled target cells are retained in the process as the cell suspension passes through a column in which a magnetic field is generated. Unlabelled cells pass through the column and are collected in a negative fraction bag. Target cells are then released from the column by removing the magnetic field from the column, allowing passage of the cells into another collection bag. Antibodycoated beads remain on the cell surface with positively selected cell products manufactured on the CliniMACS, as there is no releasing mechanism. In the case of HSCs (CD34+ cells), purity of enriched cells and CD34+ cell yield are greater than 90% and about 70%, respectively. Consequent T-cell depletion is similar to that of the Isolex system (3–4log). The Miltenyi CliniMACS system, in contrast to the Isolex 300i, is not limited to CD34+ cell selection. Clinical-grade reagents allow for positive and negative selection of HSCs (CD34+ or CD133+) and monocytes (CD14+). Several other clinicalgrade reagents (e.g. anti-CD3, anti-CD19, antiCD56) in development will make cell selection of B- and T-cell subpopulations and natural killer cells and other cell types possible. With the broadening of this technology, applications to cell therapy/cellular engineering appear limitless. Cell expansion

Because cell dose (nucleated cell, CD34+ cell and colony forming cell) and patient outcome correlate positively, much effort has been focused on ex vivo expansion of HSCs and progenitors. It is thought that successful expansion would enhance haemopoietic engraftment while reducing transfusion dependence, risk of infection and duration of hospitalization. Further, cell expansion could be exploited to increase the donor pool and support a variety of clinical applications, such as HSC autograft purging and harvesting of various 362

cell populations for genetic modification and immunotherapy generation. Most expansion strategies have utilized mobilized PB apheresis collections that have undergone CD34+ cell selection. However, UCB has been the focus more recently due to the higher proliferative and self-renewal capacity of HSCs from this source. Constituents of the cytokine cocktails used to promote expansion have varied, with stem cell factor, flt-3 ligand and thrombopoietin likely being the most integral to true stem cell expansion. Media have differed as well, with most consisting of human serum or albumin and/or a variety of culture media (e.g. Iscove’s modified Dulbecco’s medium, Amgen defined medium). Culture duration has ranged from as few as 4 days to as many as 12 days. Most of the HSC expansion clinical trials have involved patients with breast cancer undergoing autologous HSC transplantation following highdose chemotherapy. Trial designs have varied, with some involving exclusive use of ex vivo expanded cells (BM and PB) and others involving transplantation of ex vivo expanded cells along with unmanipulated cells of a dose alone usually adequate for engraftment (BM, PB and UCB). Results have been promising, particularly with a handful of recent trials involving mobilized PB. The few trials using ex vivo expanded UCB, as a whole, have shown less promise. Progress to date has been indicative of successful expansion of short-term repopulating cells, with decreases in severity and duration of cytopenias and resultant decreases in neutropenic fever, transfusion dependence and duration of hospital stay. Studies have yet to determine whether long-term repopulating cells, more primitive HSCs that maintain multilineage differentiation capacity and self-renewal, can be successfully expanded to actually improve the overall rate of engraftment and survival. Current efforts are aimed at recreating the in vivo microenvironment (i.e. stroma, extracellular matrix, and haemopoietic and nonhaemopoietic cytokines, etc.) to allow for expansion of true long-term repopulating cells. Additional cell enrichment/depletion techniques

Other techniques for T-cell depletion include

Stem cell processing and storage

monoclonal antibody-based technologies that target surface antigens on T cells, e.g. OKT-3 (antiCD3), Campath-1H (anti-CD52), and strategies that exploit T-cell surface characteristics (e.g. soybean agglutinin/erythrocyte rosette depletion), or both. Complement incubation following antibody binding and coupling of monoclonal antibodies to various molecules (e.g. immunotoxins) have been used to consistently achieve T-cell depletion of 2–3log. While effective, most of these T-cell depletion techniques have been replaced by the more efficient and less cumbersome elutriation and cell selection systems. Approaches in addition to HSC enrichment by cell selection systems, and similar to those mentioned for T-cell depletion, have been attempted in an effort to deplete, or purge, autologous BM and PB HSC grafts of tumour cells. Physical separation (based on size, density, etc.), direct exposure of grafts to pharmacological agents (e.g. 4hydroperoxycyclophosphamide) and monoclonal antibody-based technologies have all been investigated. However, the physical and pharmacological, or chemical, methods lack specificity for malignant cells; as a consequence, HSCs are damaged and lost as evidenced in the laboratory by poor clonogenic assay results and clinically by delayed engraftment. Alternatively, monoclonal antibody-based technologies hold great promise as a means of purging autografts, as they offer the advantage of target cell specificity. Malignant cells are depleted while HSCs remain present and functional. Continued recognition of various cell surface antigens and marked improvements in antibody production over the past 5–10 years have propelled antibodybased strategies to the forefront of autologous HSC purging. Monoclonal antibodies may be used alone, e.g. Campath-1H (anti-CD52), or as chemotherapy- or radioisotope-conjugated forms, e.g. HuM195 (antiCD33) coupled with calicheamicin, anti-CD45 coupled with 131I. While initial approaches have focused on in vitro methods, more recent approaches have included in vivo techniques aimed at eliminating tumour cells prior to HSC collection, e.g. rituximab (anti-CD20) for B-cell malignancies. Although some clinical trials have been promising, the true benefits of HSC

graft purging have yet to be determined. Purging (in vivo or ex vivo) in combination with posttransplant immunotherapy may be the solution in the effort to prevent disease relapse.

Quality control testing techniques Quality control (QC) testing in the clinical cell therapy laboratory serves two purposes: to determine the suitability and safety of the cellular product for the individual patient and to monitor overall laboratory practices. The level of QC testing performed is dependent on the complexity of the manufacturing process. Testing is aimed at characterizing the safety, purity, identity, potency and stability of the cellular product and may include simple tests (e.g. Gram stain) and/or complex tests (e.g. quantitative polymerase chain reaction). Here we present five standard QC tests for HSC products: cell count and differential, CD34+ cell enumeration, viability and clonogenic assays, and sterility testing (Table 32.4 for summary). Cell count and differential

Nucleated cell content and the cell differential may be determined with a haematology analyser. Haematology analysers are capable of simultaneously determining several cellular physicochemical characteristics. The majority of these instruments rely on electrical impedance to determine cell size. As cells pass through an aperture located between two electrodes, a change in the electric current occurs, producing a voltage pulse proportional to cell size. Conductivity by high-frequency electromagnetic energy elicits details of chemical composition and nuclear characteristics, and laser scattering characteristics reveal cell shape/surface and cytoplasmic granularity. Because haematology analysers are designed for the purpose of characterizing normal (unstimulated) PB, difficulties may arise with analysis of HSCs. For this reason, following determination of nucleated cell content with an analyser, a manual differential may be performed to further characterize HSCs and to quantify mononuclear cell content. 363

Chapter 32 Table 32.4 Quality control testing of

Test

Method(s)

Cell counts CD34+ cell enumeration Viability assay

Haematology analyser, manual differential Flow cytometry (single or dual platform) Dye exclusion (light microscopy), fluorescence microscopy, flow cytometry CFU (most common in clinical laboratory), LTC-IC Aerobic/anaerobic culture

Clonogenic assay Sterility testing

haemopoietic stem cell products.

CFU, colony-forming unit; LTC-IC, long-term culture-initiating cell.

CD34+ cell enumeration

It has been known for quite some time that higher CD34+ cell doses lead to greater likelihood of engraftment and successful clinical outcome in HSC transplantation. Although other markers of HSCs (e.g. CD133, aldehyde dehydrogenase) do exist, the CD34 antigen is currently the most commonly used HSC marker in the clinical laboratory. It is a 115-kDa glycoprotein that is present on HSCs, as well as some malignant cells of haemopoietic derivation, vascular endothelium, embryonic fibroblasts and a few unusual tumours. Despite its evident importance as an HSC marker, the function of CD34 has not yet been elucidated. The International Society for Hematotherapy and Graft Engineering (now the International Society for Cellular Therapy, or ISCT) developed the Guidelines for CD34+ Cell Determination by Flow Cytometry in 1995. The guidelines were initially drafted for enumeration of HSCs in peripheral blood by a dual-platform method (i.e. a two-instrument method employing a haematology analyser for determination of total nucleated cell count and a flow cytometer for calculation of per cent CD34+ cells). They proposed a sequential gating strategy to focus on cells exhibiting low side scatter (low complexity/granularity) and dim CD45 (leucocyte common antigen) staining along with CD34 positivity. Many institutions use this strategy or a modification of this strategy for enumeration of HSCs (Plate 32.1, shown in colour between pp. 304 and 305). In more recent years single-platform methods (i.e. flow cytometry alone), which utilize a known concentration of anti-CD34-coated fluorescent beads, have been 364

introduced. Single-platform methods allow for determination of absolute CD34+ cell counts. These methods appear to be particularly helpful when analysing UCB, the HSC source with characteristically the lowest concentration of CD34+ cells and a preponderance of non-viable cells, cellular debris and nucleated red blood cells. Viability assays

Viability of HSC products may be measured by dye exclusion using vital dyes (e.g. trypan blue) or with fluorescent stains (e.g. acridine orange/propidium iodide). Scoring is determined by light and fluorescence microscopy, respectively. Dye exclusion with trypan blue is widely used despite difficulties including variable staining and presence of staining artefact and high background (red cells, debris). The acridine orange/propidium iodide method is more reliable, with several advantages over dye exclusion using trypan blue. Acridine orange binds to nucleic acids of viable cells and fluoresces green; alternatively, propidium iodide binds to nucleic acids of non-viable cells and fluoresces orange. Use of fluorescence (dark-field) microscopy precludes background interference. Less variable staining and simultaneous visualization of both viable and non-viable cells allows for greater readability and more accurate interpretation. Although the acridine orange/propidium iodide method has proven to be superior to that of trypan blue, both methods measure only nucleated cell viability and therefore are not necessarily informative of the condition of the HSCs. For this reason, the most relevant viability assay utilizes flow

Stem cell processing and storage

cytometry with dyes such as 7-amino-actinomycin D or propidium iodide. With proper gating, viability of the HSC population (i.e. CD34+ cells) may be determined. Clonogenic assays

Although CD34+ cell content has become the accepted surrogate marker for graft potency, clonogenic assays remain the only truly functional QC tests. Therefore, a role for these assays in the evaluation of graft adequacy does still exist. Colony-forming unit (CFU) and long-term cultureinitiating cell (LTC-IC) assays are the most commonly performed clonogenic assays, the CFU assay being most practical in the clinical laboratory. Most clinical laboratories limit analysis to CFU-granulocyte, macrophage (GM), as multiple studies have shown correlation between engraftment and dose of CFU-GM. CFU assays require expertise and unique equipment (e.g. CO2 incubator, dissecting microscope), materials and reagents (e.g. culture media). A 14–16 day incubation (37ºC/5% CO2) limits its usefulness, though with proper planning timely analysis of cryopreserved products may be accomplished. Many laboratories limit performance of CFU assays to monthly QC testing. However, routine testing of cryopreserved HSC products, particularly UCB, may prove useful. Sterility testing

A small volume of the HSC product is removed for sterility testing. Aerobic and anaerobic cultures are set up. Microbial contamination may occur at any point from collection (e.g. donor bacteraemia, lack of aseptic technique) through processing. Therefore, to ensure efficient detection of contamination, the aliquot is removed from the final product shortly before infusion. For products undergoing cryopreservation, an additional sample is taken for microbial culture just prior to freezing. Of course, only testing systems approved by regulatory bodies should be used. If an HSC product is contaminated, immediate notification should be given to the patient’s physician. Factors to be weighed by the physician and the medical director of the clini-

cal cell therapy laboratory regarding use of the product include condition of the patient and identity, virulence and antibiotic sensitivities of the organism(s).

Cryopreservation and storage Although HSCs may be processed and infused within a few hours of collection, components collected for autologous or allogeneic HSC transplantation may need to be transported to processing centres for short-term (a few hours to a few days) or long-term (weeks to years) storage. Cryopreservation is not required for short-term storage of HSCs, but it is the most appropriate process for the long-term storage of HSCs regardless of their source. Fundamental to preserving HSC integrity and viability during the freezing process is the utilization of proper cryoprotective solutions, procedures and freezing rates that prevent intracellular ice crystal formation and cellular dehydration. Cryopreservation techniques that are optimal for HSCs will not preserve mature blood cells such as granulocytes, platelets and red cells that are often present in HSC components. Since damaged cells may cause infusion-related toxicities, HSC cryopreservation can be enhanced by the prefreeze removal of the mature blood elements or setting concentration limits. Although HSC cell concentrations used by different programmes and protocols vary widely, they tend to focus on practical issues such as the need to freeze more than one bag or minimize the total volume of components stored. HSC cryopreservation has generally required (i) DMSO as an intracellular cryoprotectant, with or without extracellular cryoprotectants such as HES or human serum albumin; (ii) cooling at 1–3∞C/min; and (iii) storage at –80∞C or colder. Although DMSO is a reasonably well-tolerated agent, it does have the potential for HSC and clinical toxicity, thus supporting conservative usage procedures. Traditional cryopreservation techniques have used 10% DMSO, but there have been multiple reports showing that it is possible to cryopreserve HSCs using a combination of 6% HES and 5% DMSO. 365

Chapter 32

The cryopreservation process requires the use of special bags constructed with material capable of withstanding the temperature range of freezing and thawing and either controlled-rate or non-controlled-rate freezing methods. Computerassisted controlled-rate freezing chambers allow a progressive temperature reduction of 1∞C/min through the liquid–solid phase change (which occurs at about –8∞C) to –60∞C and then a 3∞C/min reduction to –80 or –100∞C followed by storage in a mechanical freezer at –80∞C or –135∞C or below –120∞C in either liquid or vapour phase of nitrogen. Liquid nitrogen freezers have the advantage of maintaining a very consistent storage temperature (–196∞C), but they can serve as reservoirs of infectious agents. Therefore, centres storing HSC components in liquid nitrogen must have procedures to prevent cross-contamination between stored products. Vapour-phase storage reduces the risk of crosscontamination but specific procedures, materials and systems must be employed to minimize the temperature gradient that can form within these freezers (i.e. –100 to –190∞C). Several investigators have established that HSC products can be cryopreserved with the use of a single-step non-controlled-rate freezing procedure and stored at –80∞C for extended periods of time without compromising engraftment potential. This technique is less costly and does not require investment in programmable freezers and liquid nitrogen filler tanks or freezers. Although –80∞C freezers are easier to maintain, they are prone to mechanical malfunction and significant temperature fluctuations, especially when the door is opened. Most HSC components stored at –80∞C are cryopreserved with a mixture of DMSO and HES. All storage freezers should have (i) continuous temperature monitoring and recording, (ii) an alarm system for detection of temperature range deviations or equipment failure, (iii) a racking and inventory system, and (iv) restricted access. Although cryopreservation results in an immediate loss of some portion of the HSCs, the loss is not progressive over time if the storage conditions are appropriate. Several studies have now shown that the duration of HSC storage may be indefinite if 366

adequate storage temperatures are maintained and appropriate cryopreservation techniques used.

Quality assurance and good manufacturing practices Quality assurance (QA) is the sum of activities planned and performed to provide confidence that all systems that influence product quality are reliable and functioning as expected. The QA, or quality, programme defines the policies and environment necessary to attain minimum quality and safety standards. The basic components of a programme include standard operating procedures (SOPs), documentation/record-keeping and traceability requirements, personnel qualifications and training including a continuing education programme, building and facilities/equipment QC, process control, auditing and investigation, and error and accident system/management (see Table 32.5 for a summary of elements of a quality system). Regulatory authorities worldwide, realizing the importance of QA, have placed major emphasis on the establishment of an effective quality programme. It is expected that the HSC collection/processing centre’s quality programme will ensure a laboratory’s compliance with regulations. At a cellular therapy/processing laboratory, the quality programme is the means by which regulaTable 32.5 Elements of a quality system.

Quality plan Quality assessments/audits Documentation/record-keeping Facility design/maintenance Process control Validation Personnel training Equipment Materials management Receipt/storage/distribution Labelling Safety/safety training Deviations Adverse events

Stem cell processing and storage

tory requirements, such as good manufacturing practices (GMPs), are instituted. GMPs are scientifically sound methods or procedures that are followed and documented throughout product manufacturing. Initially developed for enforcement in the pharmaceutical industry, GMPs have been applied to cellular therapies to minimize lotto-lot variation, allowing for guarantee of safety, purity and potency. The GMP programme is a QA programme and, as such, includes requirements and specifications for production facilities, staff, supplies and reagents, equipment, laboratory controls, finished product controls, and records and reports. GMPs are not dissimilar from the ISO 9000 approach, and facilities that achieve ISO 9000 certification should also comply with GMP requirements.

Regulations and standards The scope and complexity of cellular therapies have grown markedly over the past few years. The current worldwide regulatory atmosphere reflects this growth in science and likewise is evolving rapidly. Several regulatory initiatives are underway in Europe. The European Commission is designing a directive to ensure the quality and safety of human tissues and cells. The Directive, known formally as the ‘Directive on setting standards of quality and safety for the donation, procurement, testing, processing, storage, and distribution of human tissues and cells’, is slated for implementation in 2005–06. It is based upon a quality systems approach with additional emphasis on information reporting (i.e. adverse events), standardization (i.e. ISBT 128 uniform coding system) and record retention. The Joint Accreditation Committee of ISCT Europe and the European Group for Blood and Marrow Transplantation (EBMT), or JACIE, was created in 1997 and approved the first JACIE Standards for Hematopoietic Progenitor Cell Collection, Processing and Transplantation in 1998. The second edition of the JACIE Standards became available in May 2003. Several centres have attained, or are in the process of attaining, accreditation. It is anticipated that health authori-

ties throughout Europe will recommend that the JACIE standards serve as general guidance for regulatory compliance. In the UK, stem cell collection and processing centres are moving towards compliance with the European Directive, and several centres have indicated interest in JACIE accreditation. The UK has its own accreditation programme, administered by the Medicines and Healthcare Regulatory Agency (MHRA) on behalf of the Department of Health. A Code of Practice for Tissue Banks Providing Tissues of Human Origin for Therapeutic Purposes specifies the requirements for UK centres that store and/or process human tissue (including HSCs) for therapy. It is based upon a quality management system (ISO 9000) framework and GMP principles, and it is currently the mandatory regulatory compliance and accreditation document for HSC collection and processing facilities. The Department of Health programme became effective in 2001, and both hospital and National Blood Service collection/processing centres must be accredited to provide services/products to the National Health Services Trusts. The UK Blood Transfusion Services and the National Institute of Biological Standards and Control provide a further regulatory framework with the Guidelines for the Blood Transfusion Services in the United Kingdom. The guidelines (also known as the ‘Red Book’), now in its sixth edition, provide guidance and, where appropriate, specifications and standards for the production of materials for therapeutic and diagnostic use. National Blood Service centres are required to adhere to the guidelines, whereas hospital-based centres are not. In the USA, the Food and Drug Administration has established a tiered, risk-based system to regulate cellular therapies. The framework for this system will consist of the existing GMPs (discussed briefly in the previous section) supplemented with additional regulations currently in proposed status, entitled good tissue practices (GTPs). The concept of GTPs is similar to that of GMPs, though the employed scientifically sound practices have a purpose somewhat limited to prevention of circumstances that increase the risk of introduction, transmission and spread of communicable 367

Chapter 32

disease. Although GTPs are comprehensive and clearly share several elements with GMPs, their focus is narrow by comparison, and therefore they will not supersede GMPs. However, because GMP requirements do not contain specific provisions for prevention of spread of communicable disease, GTPs will serve to supplement GMPs. Minimally manipulated products will be required to adhere to the GTPs, while products undergoing more than minimal manipulation will be held to the GTPs as well as the more rigorous GMPs. Two professional organizations, the Foundation for the Accreditation of Cellular Therapy and the American Association of Blood Banks, have written standards to serve as guidance for compliance with existing and proposed HSC regulations.

Further reading A Code of Practice for Tissue Banks Providing Tissues of Human Origin for Therapeutic Purposes. London: Department of Health, 2001. Bock TA. Assay systems for hematopoietic stem and progenitor cells. Stem Cells 1997; 15 (Suppl. 1): 185–95. Burger SR. Current regulatory issues in cell and tissue therapy. Cytotherapy 2003; 5: 289–98. Devine SM, Lazarus HM, Emerson SG. Clinical application of hematopoietic progenitor cell expansion: current status and future prospects. Bone Marrow Transplant 2003; 31: 241–52. James V, ed. Guidelines for the Blood Transfusion Services in the United Kingdom, 6th edn. UK: TSO, 2002. Joint Accreditation Committee of the International Society for Cellular Therapy Europe and the European Group for Blood and Marrow Transplantation (JACIE). JACIE Standards for Hematopoietic Progenitor Cell Collection, Processing and Transplantation, 2nd edn. Europe: JACIE, 2003.

368

Krause DS, Fackler MJ, Civin CI, May WS. CD34: structure, biology, and clinical utility. Blood 1996; 87: 1–13. Law P. Graft processing, storage, and infusion. In: Ball ED, Lister J, Law P, eds. Hematopoietic Stem Cell Therapy. Philadelphia: Churchill Livingstone, 2000: 312–13. McCullough J. Quality assurance and good manufacturing practices for processing hematopoietic progenitor cells. J Hematother 1995; 4: 493–501. Malek SN, Flinn IW. Incorporating monoclonal antibodies in blood and marrow transplantation. Semin Oncol 2003; 30: 520–30. Rowley SD. Current good manufacturing practices: application to the processing of hematopoietic cell components. Cytotherapy 2000; 2: 59–62. Smith BR. Basic biology of hematopoietic progenitor cell transplantation. In: Snyder EL, Haley NR, eds. Hematopoietic Progenitor Cells: A Primer for Medical Professionals. Bethesda, MA, AABB Press, 2000: 8. Sutherland DR, Anderson L, Keeney M, Nayar R, Chin-Yee I. The ISHAGE guidelines for CD34+ cell determination by flow cytometry. J Hematother 1996; 5: 213–26. US Food and Drug Administration. Current good tissue practice for manufacturers of human cellular and tissuebased products: inspection and enforcement; proposed rule. Fed Reg 2001; 66: 1508–59. Verfaillie CM. Hematopoietic stem cells for transplantation. Nat Immunol 2002; 3: 314–17.

Suggested textbooks Ball ED, Lister J, Law P, eds. Hematopoietic Stem Cell Therapy. Philadelphia: Churchill Livingstone, 2000. Snyder EL, Haley NR, eds. Hematopoietic Progenitor Cells: A Primer for Medical Professionals. Bethesda, MA: AABB Press, 2000. Blume KG, Forman SJ, Appelbaum FR, eds. Thomas’ Hematopoietic Cell Transplantation. Oxford: Blackwell 2004.

Chapter 33

Haemopoietic stem cell transplantation and immunotherapy Ian M. Franklin

The lymphohaemopoietic system is derived from a pluripotent stem cell and separates at an early stage into cells that give rise to the lymphoid tissues and lymphocytes, and those that produce myeloid cells. Stem cells are under the influence of cytokines and growth factors, the contributions of which have been clarified gradually over the past 15 years. The development of lymphoid tissues and lymphocytes will not be further covered in this section. The purpose of the myeloid cell precursors is to produce those blood cells which, in the main, lead to end-stage non-dividing cells which have a finite life. This lifespan may be very brief (a few hours for granulocytes) or some months (120 days for erythrocytes). Exceptions to this include some very long-lived tissue macrophages and dendritic cells which are derived from myeloid cells of the monocyte lineage. Development of the myeloid cells — erythrocytes, granulocytes, basophils, eosinophils and platelets — is achieved through differentiation into the different lineages and maturation of the cells as they develop into mature end-stage cells. These two processes of differentiation and maturation lead ultimately to a sufficient supply of cells of each type in response to a normal steady-state environment and also the ability of the bone marrow to respond to stresses such as hypoxia, anaemia, infection or bleeding. The development of bone marrow tissue is dependent upon the appropriate bone marrow microenvironment which, other than in pathological states, prevents haemopoiesis taking place in other sites. The appropriate development of haemopoietic cells requires the combination of: • a three-dimensional bone marrow microenvironment;

• haemopoietic growth factors delivered to cells both locally and in the blood circulation; and • haemopoietic stem cells. The number of cells required to maintain the normal levels of each cell type in the blood is truly astronomical, and yet normal levels of blood cells are controlled, in health, within a very narrow range.

Haemopoietic growth factors Haemopoietic growth factors are mainly glycoproteins and include: • colony-stimulating factors, such as granulocyte colony-stimulating factor (G-CSF); • some of the interleukins (IL), e.g. IL-3, IL-6; and • other haemopoietic growth factors such as thrombopoietin (TPO) and erythropoietin (EPO). In addition to their molecular characteristics they may be considered in four different functional groups: • lineage-specific factors; • multilineage factors; • stem cell factors; and • accessory or synergistic factors. There is considerable redundancy in haematological ontogeny, with the exception of erythroid precursors, which appear in the main to respond only to EPO. Inhibitors of haemopoiesis have also been described, but despite much theoretical promise as stem cell protectors during anticancer chemotherapy these hopes have not yet been realized.

369

Chapter 33

Impact of haemopoietic growth factors on patient care

The ability of physicians to control haemopoiesis is a compelling goal. However, the real impact of myeloid growth factor therapy in the management of malignant diseases has been relatively limited. In this regard, only EPO, G-CSF and granulocyte–macrophage (GM)-CSF have made a significant contribution. The current lack of a safe and reliable platelet growth factor is a great disappointment. The major exception has been the mobilization and collection of peripheral blood stem cells (PBSC) (see below). This has been simplified and made a routine procedure since the introduction of G-CSF and GM-CSF. Myeloid growth factors such as G-CSF and GMCSF shorten the period of neutropenia following chemotherapy for malignant disease when defined as a neutrophil count of 0.5 or 1 ¥ 109/L. However, reducing the duration of very profound neutropenia (neutrophil count < 0.2 ¥ 109/L) has not been so easy to confirm. A number of studies have shown modest reductions of length of hospital stay with the use of G-CSF after autologous PBSC transplants. Such reductions are probably clinically worthwhile and cost-effective but do not make a difference between being able to offer patients a transplant or not, and neither is there any detectable impact on mortality. In essence, haemopoietic growth factors can work only when there is a sufficient target subpopulation to be stimulated: they cannot create cells from nowhere. Equally the kinetics by which haemopoietic growth factors operate is dependent upon the development of the target cell population. Thus with neutrophil production in a functioning bone marrow the effect of G-CSF and GM-CSF is seen within a matter of hours since these cells are being turned over rapidly. Increased production can then be achieved very quickly. Conversely, haemopoietic growth factors that stimulate platelet production, such as IL-11 and TPO, may take between 5 and 11 days to show any effect because of the time it takes for platelets to mature from megakaryocytes.

370

Haemopoietic stem cell transplants Haemopoietic stem cell transplants, or blood and marrow transplants (BMT) as they are conventionally known, must be viewed in the context of the management of patients with, in the main, haematological malignancies. In most instances a BMT is part of the management, often used when either conventional dose therapy has failed or is expected to have a high likelihood of failure. The failure of primary therapy as evidenced by relapsed disease is a clear-cut end point. The perceived high risk of future relapse is a more subjective judgement in many cases. Objective evidence for high risk of relapse might be: • adverse chromosome abnormalities such as Philadelphia chromosome in chronic myeloid leukaemia (CML) or acute lymphoblastic leukaemia (ALL), chromosome 4;11 translocations in ALL, or chromosome 13 deletions in myeloma; • failure to enter remission promptly with induction chemotherapy, or; • perhaps bulky disease, or a high white blood cell count at presentation. Bone marrow cells are exquisitely sensitive to chemotherapy and radiotherapy. The recognition that toxic doses of radiation could kill bone marrow function permanently while other organs either recovered or were largely unaffected led to the realization that a BMT might be feasible. Thomas, working in Seattle, USA, was the first to show that BMT between human leucocyte antigen (HLA)-matched siblings was feasible. Initially the function of the pretransplant chemoradiotherapy conditioning of the patient was perceived as providing ‘space’ for the incoming cells to engraft as well as to kill any residual cancer cells. The recognition that allogeneic transplants also produce an immune-mediated graft-versus-leukaemia (GVL) effect, by observing improved disease-free survival in patients with chronic graft-versus-host disease (GVHD), was the first step towards the current awareness that allogeneic BMT is a combination of the effect of the chemotherapy and/or radiotherapy given before transplant and an immune-mediated effect against the leukaemia or other malignancy. Approximately 10 years ago it

Stem cell transplantation and immunotherapy

became clear that certain patients, especially those with CML who relapsed after an allogeneic transplant, could be induced back into full molecular remission by infusions of immune-competent lymphocytes from the original donor. This provided more direct proof of the GVL effect but also began the modern era of transplantation. BMT today can be viewed as much more of an immunotherapy, in which the transplant is a major component of the disease control. The advent of reduced-intensity conditioning (RIC) transplants (sometimes known as ‘mini-transplants’ or even ‘transplant-lite’!) has further emphasized the immunotherapy component of the BMT, in which RIC is insufficient to eradicate bone marrow cells but in which immune tolerance is induced to permit full engraftment of the incoming donor transplant, which almost always uses PBSC-derived marrow tissue. The purpose of stem cell transplantation is: • to enable intensification of chemotherapy and radiotherapy such that toxicity to the bone marrow is no longer an important factor in determining outcome; • to ensure complete engraftment of the donor marrow through immunosuppression of the host, thus permitting tolerance to develop; • to promote, in allogeneic transplants, a GVL effect. In conventional myeloablative transplants the immediate function of the stem cell transplant might be seen as a rescue from the effects of chemoradiotherapy. Once the patient has recovered, the function of the graft is in addition to mount a potent GVL effect to maintain control of the minimal residual disease that may still be present. In recent years it has become clear that it is possible to separate GVL from graft-versus-host reactions in allogeneic stem cell transplants by adding planned doses of donor lymphocytes back to the patient in the months following the transplant. In many cases it is possible to show a GVL effect in the absence of GVHD. Sources of stem cells

• Allogeneic stem cells: provided from another individual, most commonly a sibling. There is an increasing use of alternative donors, mainly unre-

lated adults from donor registries but also cord blood-derived cells. • Syngeneic cells: from an identical twin. • Autologous: from the patient. If the patient has no sibling who is suitable as a donor, then occasionally it is fruitful to explore the extended family, particularly if parents share HLA haplotypes, whether or not they are consanguineous. Increasingly, however, unrelated transplants are being performed, using national and international registries of volunteer donors, and the toxicity and results of these procedures is improving steadily. The use of umbilical cord blood, both directed (i.e. from a sibling when an existing child is already known to have the potential to benefit from such a transplant) or from unrelated donor cord blood banks is increasing worldwide, albeit slowly and still, in the main, for individuals with low body weight (< 50 kg). More recently, successful engraftment in over 70% of adults in a series of over 40 patients has suggested that there may be potential of this stem-cell source in larger humans. The relative merits of these different sources are considered in Table 33.1. In addition, it is possible to obtain bone marrow-derived cells either directly or from the peripheral blood following G-CSF administration. • Bone marrow harvesting. Classically, to obtain haemopoietic stem cells for autologous or allogeneic transplants, procedures involved bone marrow aspiration from both posterior iliac crests (under general anaesthesia) in order to collect a minimum of 2 ¥ 108/kg nucleated cells wherever possible. This provides reliable engraftment after transplant. • PBSC mobilization and collection. In the last decade marrow harvesting has been largely replaced for autologous transplants by the use of GCSF or GM-CSF mobilized PBSC and such transplants have been gaining popularity in allogeneic practice as well. At present it is likely that PBSC allografts produce more chronic GVHD but not acute GVHD compared with BMT in siblings. There are no data available at present for unrelated donor transplants using PBSC.

371

Chapter 33 Table 33.1 Comparison of sources of stem cells.

Bone marrow Availability Allogeneic

Peripheral blood stem cells; G-CSF mobilized

Cord blood

Need matched sibling donor (1 in 3 chance) or unrelated donor from registry

As for bone marrow; may not be available from unrelated donors from all registries or countries

Depends on tissue type

Autologous

Available if fit for anaesthesia Not of use in: Aplastic anaemia Myelodysplasia Inherited disorder Much prior chemotherapy

Need good venous access or central line Prior chemotherapy may prevent good collections

Not relevant

HLA matching

Siblings usually straightforward Unrelated panel searching and typing can take time and need molecular typing

As for bone marrow

‘Off the shelf’ capability May not need such stringent matching but more data needed

Engraftment

Reliable for siblings if achieve > 2 ¥ 108 nucleated cells/kg recipient body weight Neutrophils > 1 ¥ 109/L by day 21, platelets > 20 ¥ 109/L by day 30, unrelated donors need > 3.7 ¥ 108 nucleated cells/kg recipient body weight for sturdy graft and good outcome

Doses of > 4 ¥ 106 CD34+ cells/kg recipient body weight will give reliable prompt engraftment slightly faster than bone marrow Neutrophils > 1 ¥ 109/L by day 17, platelets > 20 ¥ 109/L by day 21

Engraftment is slow and uncertain, and at least 28 days may be needed for both neutrophil and platelet recovery Uncertain if one cord blood donation can engraft an adult > 75 kg reliably

Graft-versus-host disease (GVHD) Acute Standard-risk cases will get acute GVHD in about 65%, with about 25% getting severe disease Greater risk in unrelated donor transplants

Appears to be no different from bone marrow in sibling transplants

Since cord blood transplants mainly used in children, incidence is low, either because children known to have low incidence of GVHD or naive T cells in cord blood

Chronic

Risk is increased to 50–60%

Not known for sure as yet

Some 20–30% of patients will develop some form of chronic GVHD In about 10% this will be severe and debilitating

G-CSF, granulocyte colony-stimulating factor.

Indications for haemopoietic stem cell transplants

The detailed indications for intensive therapy requiring haemopoietic stem cell rescue or support change over time. Current indications as considered appropriate by the European Group for Blood and Marrow Transplantation (EBMT) are regularly updated and provide a broad consensus of what the transplant community in Europe considers reasonable. The transplant data have 372

been analysed from the major transplanting countries contributing to the EBMT database. This has enabled a hierarchy of indications to be prepared based on actual transplant activity. That is, it should reflect what transplant physicians do and not merely what they say they would do. Naturally, the indications agreed by transplanters are likely to be at the limits of the procedure. Some indications may be greeted with less

Stem cell transplantation and immunotherapy

enthusiasm by more conservative physicians. Also, as trial data emerge, enthusiasm and indications may change. Recent studies looking at the role of high-dose chemotherapy and autologous BMT in breast cancer have not shown a significant benefit for BMT. Although not conclusive, it is clear that further trials are needed in breast cancer before the role of intensive therapy with BMT can be clarified, if indeed there is any role. Most clinicians would agree that the following groups are highly appropriate for BMT. Allogeneic BMT

This is indicated in the following conditions. • CML in first chronic phase with a matched sibling or unrelated donor. The introduction of the specific tyrosine kinase/bcr/abl inhibitor imatinib (Glivec) has provided improved responses for many patients with CML, but is still not believed to be a curative agent. • Acute leukaemias in second complete remission. • Very severe aplastic anaemia in children and young adults. • Some high-risk acute leukaemias in first complete remission. Relative indications for allogeneic BMT or indications appropriate for clinical trial would be: • acute leukaemias in first complete remission in adults; • multiple myeloma; and • myelodysplastic syndromes in children or young adults. Autologous transplantation has less clear-cut indications, most remaining appropriate for clinical trials that compare BMT with less intensive approaches. Clear-cut indications based on randomized clinical trial data include: • high-grade non-Hodgkin’s lymphoma in second complete remission or in first relapse with responsive disease; and • multiple myeloma. Patients with Hodgkin’s disease who have relapsed after at least two different treatments, e.g. radiotherapy and chemotherapy, have improved disease-free survival after autograft but overall survival may not be so different since some patients can be ‘salvaged’ by transplant

after further relapses following conventional chemotherapy. Complications of transplantation

Patients who are being considered for any form of blood or marrow transplant must be given full information about the procedure prior to their agreement to proceed. Although procedures such as allogeneic BMT for CML and severe aplastic anaemia have transformed the prognosis such that the majority of patients with these diagnoses will survive, all transplant procedures carry major risks of mortality, morbidity and later ‘collateral damage’. Some of these risks will remain lifelong. The chronology of the major complications of allogeneic blood or marrow transplant is shown in Fig. 33.1. Regimen-related toxicity

This refers to the immediate toxic effects of the radiotherapy or chemotherapy used for the transplant. With the exception of some current experimental procedures, all transplants have used intensive therapy that will ablate (kill off) the patient’s own bone marrow. Other organs are also damaged, especially the gut, with severe mucositis a major problem. Less commonly, liver, heart, lungs and kidneys may suffer transient or even permanent damage. Careful pretransplant assessment of each patient is essential. The use of RIC transplants is intended to minimize this toxicity and enable patients to receive transplants who might be unfit for full chemotherapy and radiotherapy conditioning. Rejection

Rejection is an immune-mediated event in which the pretransplant conditioning and immunosuppression are insufficient to prevent recognition by the recipient of HLA on the donor cells. It only occurs in allogeneic transplants, although graft failure can occur in autografts. Graft failure is due to inadequate numbers of stem cells in the transplant and/or pre-existing damage to the marrow microenvironment. 373

Chapter 33

Before transplant After transplant Alloimmunization GVHD

Chronic GVHD

Acute GVHD

Pneumocystis Bacteria Infection

Fungi Viruses

High risk: continuing while GVHD or poor graft function

CMV Zoster Infertility Central line problems: thrombosis and infection

Central line removed

Day + 50

Day + 100

Risk factors for rejection include the following. • HLA incompatibility between patient and donor, especially in unrelated donor transplants. • Low numbers of stem cells infused (< 3.7 ¥ 108 nucleated cells per kilogram body weight in unrelated BMT). • Prior sensitization of the patient to HLA or other marrow cell antigens. This is particularly problematic in patients with severe aplastic anaemia and can be prevented by using leucocytedepleted blood components from presentation onwards. The kinetics of engraftment of bone marrow and PBSC is shown in Fig. 33.2.

Fig. 33.1 Relationship of the major

complications of BMT to the time before and/or after transplant. Such a diagram can give only a broad view of the most common periods of relatively uncomplicated transplants. Those using an unrelated or other alternative donor will have greater risks of graft failure, graft-versus-host disease (GVHD) and thus continuing likelihood of infections. CMV, cytomegalovirus; HSV, herpes simplex virus; Zoster, varicella-zoster virus.

HSV

after the transplant. Acute GVHD is characterized by involvement of the following. • Skin: from an erythematous sunburn-like rash to a blistering exfoliative erythroderma. • Liver: typically the bile ducts are attacked and an obstructive jaundice-type picture develops. Milder forms may lead to elevated transaminases and cause considerable difficulties with diagnosis. Liver biopsy is advised. • Gastrointestinal tract: classically a profuse watery diarrhoea develops, bloody in the most severe cases. Upper gastrointestinal upset is not uncommon, with nausea and sickness. Rectal biopsy or upper gastrointestinal endoscopy is required for diagnosis.

Graft-versus-host disease

GVHD is caused by immune-competent T lymphocytes in the transplant recognizing antigens in the patient as foreign. This then leads to cytokine release, which increases and perpetuates the response. Despite immunosuppression of the patient to prevent GVHD, e.g. using cyclosporin plus methotrexate or corticosteroids, more than half of patients receiving allogeneic transplants will develop acute GVHD in the first 100 days 374

Relapse

Despite the intensive preparation for transplant, a significant proportion of patients will suffer recurrent disease after transplant (Fig. 33.3). Risk factors for acute leukaemia recurrence include the folowing. • Patient not in remission at time of transplant, especially if disease is resistant to conventional therapy.

Stem cell transplantation and immunotherapy (a)

PBSC infusion

120

Red cell transfusions

Hb (g/L)

100 80 60 40 20 0 300

myeloid leukaemia who received a peripheral blood stem cell (PBSC) allogeneic transplant from a sibling using reduced-intensity conditioning. The blood count falls following the chemotherapy (fludarabine and melphalan) and recovers quickly once engraftment begins. The patient had minimal acute graft-versus-host disease and the marrow graft has remained robust. Molecular evidence of chronic myeloid leukaemia was detected some 9 months after the transplant and was treated with low-dose donor lymphocyte infusion. The patient is alive and disease-free 6 months later. Hb, haemoglobin; WBC, white blood cells.

200 150 100 50 0 12 10

WBC (¥109/L)

Fig. 33.2 (a) A patient with chronic

Platelets (¥109/L)

Platelet transfusions 250

8 6 4 2 0 –20

–10

• Patient is beyond first complete remission, i.e. has already relapsed once after chemotherapy even if now in remission. • Autologous marrow or PBSC transplant (no GVL). • No GVHD after allogeneic transplant. Although GVL can be separated from GVHD in some situations, the presence of GVHD, especially chronic GVHD persisting more than 100 days after transplant, is associated with a potent GVL effect. However, overall survival may still be worse due to the toxic effects of GVHD.

0

10 20 30 Days before and after transplant

40

50

60

Infectious complications

The immune system of the transplant recipient must be suppressed to allow the graft to be accepted, and antitumour therapy such as total body irradiation (TBI) ensures that the patient has minimal immune function at the time of the transplant. Even RIC transplants have this risk, because they utilize intensive immunosuppression in order to ensure that the transplant is not rejected. The agents used (fludarabine and anti-Tcell or pan-lymphoid monoclonal antibodies) provide prolonged and profound immune deficiency. Haemopoietic recovery takes at least 2–3 375

Chapter 33

(b) Bone marrow infusion 125

Hb (g/L)

Red cell infusions

100

75 –10

0

10

20

30

40

50

60

Day

1000 Platelet infusions

WBC (x 109/L)

100

WBC (x 109/L)

Platelets (x 109/L)

10

1

0.1 –10

0

10

20

30 Day

376

40

50

60

Fig. 33.2 (b) A patient with severe aplastic anaemia secondary to hepatitis. Blood counts are low before the transplant, reflecting the aplasia and are relatively slow to recover. Severe acute graft-versus-host disease occurred but was controlled eventually with corticosteroids and antilymphocyte globulin. The patient is alive and well 4 years later. Hb, haemoglobin; WBC, white blood cells.

Stem cell transplantation and immunotherapy

Probability %

100

Fig. 33.3 Probability of survival after

allogeneic transplants for chronic myeloid leukaemia in chronic phase by donor type and disease duration, 1994–99.

80

HLA-identical sibling, <1 year from diagnosis (n = 2876)

60

HLA-identical sibling, ≥1 year (n = 1391) Unrelated, <1 year (n = 613) Unrelated, ≥1 year (n = 936)

40 20 P = 0.0001 0 0

weeks, but recovery of neutrophils is only part of the reconstitution of the immune system that must occur for full recovery. BMT-related immune problems may be divided conveniently into three phases as described below. Immune deficiency is compounded at any point after BMT by the presence of active GVHD. Immediate post-BMT phase This phase is characterized by neutropenia as well as lymphopenia and hypogammaglobulinaemia. During this period the patient is managed as follows. • Protective isolation: filtered air to reduce fungal contamination is especially important. • Prophylactic antifungal, antiviral and antibacterial therapy is routine. • Pre-emptive use of therapeutic antimicrobials: broad-spectrum antibacterials are used at the first sign of fever, followed by antifungal treatment empirically in the absence of prompt resolution. • Intravenous immunoglobulin may be used in some patient groups. • Prophylactic neutrophil infusions are not used routinely. Trials are needed urgently but are very difficult to design and deliver. Early post-engraftment period The patient will now have some marrow function and, if GVHD is absent or controlled, may be able to leave hospital. Although patients having autologous transplants rarely have major problems

1

2

3 Years

4

5

6

after this time, vigilance is necessary. Allogeneic BMT recipients remain at risk of the following. • Bacterial infections related to central lines. • Fungal infection. • Cytomegalovirus (CMV): most units will monitor for emerging CMV using a polymerase chain reaction (PCR)-based test and/or use a regimen of prophylactic antiviral drugs (usually ganciclovir). Such strategies are very effective and CMV is becoming a much less important cause of mortality after allogeneic BMT. • Other viruses, especially respiratory syncytial virus and other herpesviruses such as herpes zoster. • Toxoplasma. Later problems Patients who have active GVHD requiring immunosuppressive therapy will continue to have impaired immunity to pathogens, and most patients who have received unrelated donor transplants will have detectable abnormalities of the immune system. However, by 3 years after the transplant almost all patients not still on immunosuppression will have virtually normal immunity. • Vaccination has a role to play but those patients still on immunosuppression will not respond optimally. • Continued prophylaxis and vigilance are required. • Hyposplenic cover: patients who have had TBI are hyposplenic and must receive vaccine against organisms such as pneumococcus, 377

Chapter 33

meningococcus and Haemophilus influenzae B, as well as receiving lifelong chemoprophylaxis, e.g. with amoxicillin. Late effects It might be thought that having endured a lifethreatening primary disease, followed by the immediate risks of BMT outlined above, that survivors might be entitled to a respite from further problems. Unfortunately, a litany of potential problems can and do occur, including the following: • cataracts (TBI only); • endocrine problems (worse with TBI); • hypothyroidism; • growth retardation in children (especially TBI with or without steroids); • infertility; • sexual dysfunction; • second malignancies (this remains a risk even 20 years after the transplant); • transfusion-transmitted viruses, e.g. hepatitis C; and • iron overload and liver dysfunction from red cell transfusions. These problems mean that BMT recipients require lifelong follow-up at a centre familiar with the range of late complications and with a sufficiently large practice to ensure that emerging problems are identified promptly. Outcome

A discussion of the results of BMT for the wide range of indications now accepted is beyond the scope of this chapter. Current results from the International Bone Marrow Transplant Registry (IBMTR) are shown in Table 33.2. Chronic myeloid leukaemia

The best results are for patients with CML and this is especially so for young patients transplanted early after diagnosis. Candidates for allogeneic transplantation should not receive interferon-a since this may reduce the overall survival. In large single centres such as in Seattle, about 80% of patients with such early-phase CML will survive 378

long term with a matched sibling or unrelated donor. Results from registries are less good, as shown in Fig. 33.3, the probability of survival being 64% at 3 years for transplants reported to IBMTR. For patients receiving an unrelated matched transplant, probability of survival reduces to 47% at 3 years. At present, it is unclear what role or effect imatinib will have in the outcome of CML transplants. Patients with a matched sibling donor are currently advised to have an allogeneic transplant within 1 year of diagnosis, or earlier if any adverse features are present. Acute leukaemia

For acute leukaemias, the precise details of each case are needed before outcome can be determined. Prognostic features in ALL include the following. • Age over 16 years is inferior and over 40 years worse still for results of conventional treatment and transplants. • White blood cell count (WBC) at diagnosis: WBC more than 20 ¥ 109/L is less good than lower counts. Some 45% of transplants reported to IBMTR have a count greater than 35 ¥ 109/L. • Philadelphia chromosome-positive leukaemia: very few such cases will be cured without an allogeneic transplant. Results for all patient groups at 3 years for leukaemia-free survival (LFS) in ALL were: • 54% in first complete remission (CR1) (relapse risk 25%); • 40% for second or subsequent complete remission (relapse risk 46%); and • 20% for patients transplanted not in remission (relapse risk 68%). Similar results are found for acute myeloid leukaemia (AML), and again the specific type of AML can be crucial in determining outcome from conventional therapy. It is this, plus the confidence the BMT unit has in a successful outcome, which will then determine how strongly to recommend BMT in first remission. Overall IBMTR results for allogeneic BMT in AML show: • 59% LFS at 3 years when transplanted in CR1 (relapse risk 24%);

Acute myeloid leukaemia

Acute lymphoblastic leukaemia

Allogeneic identical sibling

Chronic myeloid leukaemia (all ages)

Autologous

Allogeneic unrelated donor

Allogeneic sibling

Autologous (> 20 years)

Allogeneic unrelated donor

Allogeneic sibling

Allogeneic unrelated donor

Transplant type

Disease

Table 33.2 Outcomes for bone marrow transplantation.*

CR1 CR2 or other remission Not in remission CR1 CR2 or other remission Not in remission CR1 CR2 or other remission Not in remission

CR1 (> 20 years) (£ 20) CR2 or other remission (> 20 years) (£ 20) Not in remission(> 20 years) (£ 20) CR1 (> 20 years) (£ 20) CR2 or other remission (> 20 years) (£ 20) Not in remission(> 20 years) (£ 20) CR1 CR2 or other remission Not in remission

CP1 (first chronic phase) Acc. phase Blast phase CP1 (first chronic phase) Acc. phase Blast phase

Disease stage

1912 524 1067 202 252 387 729 265 177

741 455 312 778 363 163 176 233 171 652 191 153 95 59 39

3528 750 249 1240 460 174

Number of transplants

59 ± 2 45 ± 5 23 ± 3 36 ± 8 25 ± 7 10 ± 4 51 ± 5 41 ± 8 22 ± 8

49 ± 5 61 ± 6 32 ± 7 52 ± 4 15 ± 5 27 ± 8 37 ± 11 48 ± 8 25 ± 8 38 ± 4 13 ± 6 19 ± 8 51 ± 14 15 ± 15 9±9

64 ± 2 45 ± 5 30 ± 7 50 ± 3 28 ± 5 20 ± 7

Survival at 3 years post transplant (%)

56 ± 3 36 ± 7 18 ± 4 36 ± 8 25 ± 7 5±4 39 ± 8 34 ± 9 20 ± 8

44 ± 5 58 ± 7 29 ± 7 46 ± 6 13 ± 5 24 ± 9 22 ± 20 48 ± 8 25 ± 8 33 ± 6 NA 17 ± 8 46 ± 15 NA NA

60 ± 3 33 ± 12 26 ± 8 45 ± 4 25 ± 6 20 ± 7

Survival at 5 years post transplant (%)

Stem cell transplantation and immunotherapy

379

380

Non-Hodgkin’s lymphoma (NHL) Follicular NHL

All ages; 90% > 20 years. All ages; 77% > 20 years

Hodgkin’s disease

Severe aplastic anaemia

Allogeneic sibling

Myelodysplasia (all ages; 85% > 20 years)

Allogeneic identical sibling

Autologous

Allogeneic sibling Other related or unrelated

Autologous

Other related or unrelated

Allogeneic sibling

Allogeneic unrelated donor

Transplant type

Disease

Table 33.2 Continued

CR1 CR2 CR3 or greater Relapse no. 1 Relapse beyond no. 1 Primary treatment failure Relapse no. 1 Primary treatment failure

CR1 CR2 CR3 or greater Relapse no. 1 Relapse beyond no. 1 Primary treatment failure Advanced disease assumed Advanced disease assumed

Age £ 20 years Age > 20 years. Age £ 20 years Age > 20 years.

Refractory anaemia (RA) RA with excess blasts (RAEB) RAEB in transformation (RAEBT) RA RAEB RAEBT

Disease stage

118 189 45 476 223 252 108 100

106 375 96 912 347 364 172 34

679 636 244 118

229 296 279 87 116 124

Number of transplants

84 ± 9 71 ± 9 67 ± 17 70 ± 5 60 ± 8 64 ± 7 69 ± 10 62 ± 11

87 ± 9 76 ± 6 72 ± 10 65 ± 4 58 ± 7 53 ± 7 33 ± 9 34 ± 17

78 ± 3 68 ± 4 52 ± 7 32 ± 10

53 ± 8 39 ± 7 41 ± 7 31 ± 13 34 ± 9 27 ± 9

Survival at 3 years post transplant (%)

79 ± 13 71 ± 9 51 ± 24 59 ± 11 54 ± 11 42 ± 20 59 ± 20 62 ± 11

87 ± 9 64 ± 10 67 ± 12 54 ± 7 43 ± 12 48 ± 8 25 ± 12 34 ± 17

77 ± 4 66 ± 4 50 ± 8 27 ± 13

50 ± 9 27 ± 11 40 ± 8 27 ± 13 34 ± 9 15 ± 14

Survival at 5 years post transplant (%)

Chapter 33

Allogeneic sibling Other related or unrelated

Fanconi’s anaemia (all ages; 98% < 30 years)

Diagnosis to transplant £ 18 months Diagnosis to transplant > 18 months Diagnosis to transplant £ 18 months Diagnosis to transplant > 18 months Diagnosis to transplant £ 18 months Diagnosis to transplant > 18 months

CR1 Relapse no. 1 Primary treatment failure CR1 Relapse no. 1 Primary treatment failure

CR1 CR2 CR3 or greater Relapse no. 1 Relapse beyond no. #1 Primary treatment failure Relapse no. 1 Primary treatment failure

175 111

526 31

2690 835 564 211 52 33

53 30 24 49 26 27

259 435 67 1027 305 597 94 99

76 ± 8 28 ± 10

81 ± 3 67 ± 18

60 ± 3 48 ± 5 46 ± 5 41 ± 8 32 ± 14 11 ± 10

65 ± 16 35 ± 22 42 ± 22 (30 months) 55 ± 16 14 ± 14 49 ± 20

66 ± 7 53 ± 6 38 ± 15 44 ± 4 31 ± 6 49 ± 5 33 ± 12 22 ± 10

76 ± 8 NA

81 ± 3 NA

40 ± 7 25 ± 9 42 ± 5 29 ± 10 26 ± 16 NA

65 ± 16 35 ± 22 NA 55 ± 16 NA 49 ± 20

53 ± 11 45 ± 8 32 ± 16 37 ± 5 30 ± 8 41 ± 7 33 ± 12 22 ± 10

* Results from the International Bone Marrow Transplant Registry (IBMTR) (allogeneic) and the American Autologous BMT Registry (ABMTR) (autologous) for adult patients over 20 years old (except where stated). The data presented here were obtained from the Statistical Center of the International Bone Marrow Transplant Registry and the Autologous Blood and Marrow Transplant Registry. The analysis has not been reviewed or approved by the Advisory Committees of the IBMTR/ABMTR. Transplants took place between 1994 and 1999 and reflect all those registered with the IBMTR and the ABMTR. Figures are given for survival at 3 and 5 years, where available. This is not disease-free survival, although most patients with acute leukaemia, myelodysplasia, aplastic anaemia and chronic myeloid leukaemia are likely to be disease-free, if alive.The reliability of the data is clearly greatest where the greatest numbers of transplants have been performed.These results do not reflect the emerging trend for reduced-intensity conditioned transplants. CR, complete remission (CR1, first complete remission); CP, chronic phase of chronic myeloid leukaemia.

Allogeneic sibling Other related or unrelated

Other related or unrelated

Allogeneic sibling

Autologous

Allogeneic identical sibling

Autologous

Allogeneic identical sibling

Autologous

Thalassaemia (all ages; 93% < 20 years)

Multiple myeloma

Lymphoblastic NHL

Diffuse large-cell NHL

Stem cell transplantation and immunotherapy

381

Chapter 33 Table 33.3 Classification of indications

Allogeneic haemopoietic stem cell transplant

Autologous haemopoietic stem cell transplant

CV < 50; very high degree of consensus

AML CR1 AML other than CR1 ALL CR1 ALL other than CR1 CML CP1 CML other than CP1 Myelodysplasia Non-Hodgkin’s lymphoma

Multiple myeloma Hodgkin’s disease Non-Hodgkin’s lymphoma

CV 50–80; some variation in practice between BMT units/nations

Multiple myeloma Chronic lymphocytic leukaemia

AML CR1 AML other than CR1

CV > 80; little consensus as to evidence in support of indication. Clinical trials highly appropriate in these conditions

Hodgkin’s disease

ALL CR1 ALL other than CR1 CML CP1 CML other than CP1 Myelodysplasia Chronic lymphocytic leukaemia

Degree of consensus

for blood and marrow transplant according to BMT unit practice in Europe in 2001.

All transplant procedures are arduous, even though mortality has fallen over the past years. In addition, the use of allogeneic donors causes major problems with immune reconstitution such that few patients over 60 years old would be considered for sibling transplants, or for unrelated procedures over 50 years of age. Reduced-intensity conditioned transplants are increasing these age thresholds but no mature data are available for long-term outcome. For autografting some groups have extended the limit to 75 years and the author has experience up to 68 years. Fitness of the patient and the likelihood of benefit are the most important considerations. ALL, acute lymphoblastic leukaemia; AML, acute myeloid leukaemia; CML, chronic myeloid leukaemia; CR, complete remission (CR1, first complete remission); CP, chronic phase of CML; CV, coefficient of variance.

• 35% LFS in second or subsequent complete remission (relapse risk 45%); and • 26% LFS for patients in relapse at the time of allogeneic BMT (relapse risk 57%). Results for other common indications are shown in Table 33.3. Registry data are of great importance but cannot replace careful assessment of individual patients. Such individual factors may serve to improve or worsen the risks for a particular case, e.g. coexistent disease or toxic effects of prior chemotherapy. Also, registries report only mature data, usually with a minimum of 3 years follow-up. Since RIC transplants are a recent development, experience with these is not yet 382

reflected in registry data, and primary research publications and meeting reports must form the basis of current outcomes. Regulatory aspects of haemopoietic stem cell transplantation

Concern over the appropriate indications for transplants and an increased awareness that the support services needed for BMT require careful quality assurance has led to an increasingly stringent regulatory environment. Some of these support services include: • stem cell cryopreservation;

Stem cell transplantation and immunotherapy

• stem cell manipulation, e.g. CD34 selection, stem cell expansion; and • molecular and cellular diagnostic facilities. The number of teams performing transplants in Europe increased from eight in 1973 to almost 100 in 1983. Over 340 teams reported data to the EBMT group in 1995, and 479 in 2001, when 16 555 transplants were performed. Some of the less complex transplants are no longer the domain of specialist units but can be performed safely and appropriately in smaller centres. However, safety of patients remains the first concern and this has led to the current regulatory environment, most of which remains managed by professional groups and is essentially voluntary. The European Union is likely to develop its own guidance in the near future, which will be statutory. At present the regulatory structures in Europe and the UK, and their status, are as follows. • Joint Accreditation Committee of EBMT and International Society for Cellular Therapy (ISCT) Europe (JACIE). Voluntary/professional guidelines. Accreditation in the UK will be administered by the British Society of Blood and Marrow Transplantation (BSBMT). Implementation not yet complete. • European Directive 85/374/EEC (1985). Statutory in relation to product liability in general. • UK Department of Health guidance notes on the collection, storage and infusion of bone marrow and stem cells (1997): advisory. • Medicines and Healthcare Products Regulatory Agency (MHRA). Accreditation of UK establishments involved in tissue banking, including haemopoietic stem cells. In legal terms a voluntary process, but the intention is that this will be an essential requirement in the UK. • British Committee for Standards in Haematology (BCSH). Professional guidelines on stem cell collecting, processing and storage as well as separate guidance on the appropriate clinical facilities for the care of patients with severe bone marrow failure. • European Union Directive on Tissues. Currently being drafted, this will include stem cell regulation. It is likely to enter EU law in 2004, and to come into force in individual member states in 2006.

It is clear that if clinicians wish to ensure that they are operating a safe service they will need to show adherence to clear quality systems including: • attention to the quality of raw materials, premises, equipment, reagents and storage; • each product, e.g. CD34-selected stem cells, must have an appropriately defined specification set; • safety and efficacy of each technical procedure must be validated and documented; • standard operating procedures (SOPs) should be written for each procedure; • evidence of the competence, training and proficiency of each staff member must be available; • internal and external quality assurance systems must be in place; • quality audits should be performed by inspectors external to the BMT service; • each stem cell donation or product must be uniquely identifiable, and the harvest procedure, time, date and nature of processing documented; • records must be kept for at least 11 years or until the patient/donor has reached 21 years of age, whichever is the longer. Although these quality measures may often appear pedantic at best or even restrictive of innovation at worst, they are essential if, over time, the confidence of donors, patients and the public in the safe conduct of haemopoietic stem cell transplants is to be secured and maintained.

Immunotherapy The idea that the immune system might be exploited to prevent recurrence of malignant disease or even treat it in the first place has been present for a long time. Initially, studies involved crude preparations of bacteria to provide immune stimuli. Landmarks include: • pyogenic bacteria (erysipelas) used by William Coley in New York in 1893 to stimulate antitumour responses; • use of bacille Calmette–Guérin (BCG) as an anticancer ‘vaccine’ by Holmgren in Sweden (1935); • Mathe and colleagues showed a beneficial effect of BCG against ALL in 1969, although 383

Chapter 33

Resting lymphocyte

1

Activated lymphocyte

2 3 4 Tumour cell

Apoptosis Fig. 33.4 Interaction between tumour cells and

lymphocytes. The tumour cell expresses antigens essential for recognition by T lymphocytes. Adhesion molecules (1) maintain cell-to-cell contact while major histocompatibility complex (MHC) structures present peptide antigens that may be tumour specific (2). If costimulatory molecules such as B7 (3) are present as well, the T lymphocyte will become

subsequent randomized trials failed to show any benefit; and • South-west Oncology Group in the USA showed clear benefit of local BCG in bladder cancer in 1980. It is only in the past 5–10 years, however, that essential knowledge of the details of the immune response, and how it may be enhanced or abrogated, has emerged. In the 1980s the identification first of IL-2 and its occasional effects against malignant melanoma, despite frequent major toxicity, showed that there might be a role for more specific immune targeting. Adding lymphokine (IL-2)-activated killer cells or isolating tumourinfiltrating lymphocytes on the basis that they may have a specific effect against tumours again showed anecdotal promise but has not entered the mainstream of therapy. IL-2 does have a clear, albeit inconsistent, effect against renal cell cancers, although the mechanisms for this and the other non-specific cellular therapies are not known. More recently, a greatly improved understanding of the activation and stimulation of cells in 384

activated and immunity will develop. In the absence of B7, however, the T lymphocyte will either die by apoptosis or become anergic (inactive but alive) and there will be tolerance towards the antigen. Tumour cells may produce soluble cytokines or express other molecules (4) that can either block activation or induce apoptosis.

the cellular immune response has triggered much research. This, combined with an explosive increase in knowledge of antigen-presenting cells such as dendritic cells, means that protocols being developed now have the benefit of immunological logic rather than clinical empiricism. It is now understood that immunity against foreign antigens or cells is mediated, at least partially, by a combination of surface antigens having to be all in place on both the target tumour cell and the effector cells of the immune system. In addition the most potent effects are seen when foreign antigens have been processed by follicular dendritic cells. The necessary antigens are shown in Fig. 33.4. The combination of T-cell receptors, adhesion molecules and the costimulatory molecules B7.1 (CD80) and B7.2 (CD86) are all required for effective signalling to trigger a cytotoxic T-cell response. Firstly, two areas that have established the role of cellular immunotherapy will be considered. These include GVL and the use of cytotoxic T cells to treat post-BMT lymphomas. Secondly,

Stem cell transplantation and immunotherapy

the future potential of immunotherapy will be summarized. Graft versus leukaemia

During the 1980s there was a vogue for BMT procedures in which the T lymphocytes were removed in order to prevent GVHD. It was recognized in the late 1980s that these transplants were associated, particularly in CML, with a very high incidence of relapsed disease, especially if the patient developed no GVHD at all. Although a relationship between GVHD and reduced relapse had been noted earlier, this was the first occasion on which there was clear evidence of a GVL effect being mediated via GVHD. This is caused by alloreactive T lymphocytes in a BMT identifying foreign antigens in the host. Subsequent to these data from Tcell depleted transplants, it became apparent that there was a difference in relapse rate with other forms of GVHD prophylaxis. For example, with cyclosporin and methotrexate there was a higher incidence of relapse after transplant than with either cyclosporin or methotrexate alone, presumably because of the reduced GVHD seen with the dual therapy. Overall survival was not impaired, however, since GVHD remains the most important adverse event predicting survival after allogeneic BMT.

further clinical, cytogenetic and molecular remission by infusions of donor lymphocytes collected from the original transplant donor. However, this treatment was complicated by the following. • Severe GVHD. This could be avoided by giving lower and graded doses of donor T lymphocytes. In the majority of patients 1 ¥ 107/kg will induce complete molecular and haematological remission without a significant risk of GVHD or graft failure. • Failure of the graft leading to marrow aplasia. Prevented by monitoring the patient more carefully so that when donor lymphocyte infusions (DLI) are given early in relapse, before a significant proportion of the transplant has been rejected and replaced by leukaemia, they can be successful without the risk of bone marrow failure. Thus, by using DLI early in relapse and giving a specific dose of cells, it is possible to separate GVL from a graft-versus-host response. In addition to CML, DLI has been tried in numerous diseases with varying success. In multiple myeloma there is good evidence of reinduction of remission, although the number of cases remains small. Also it appears to be necessary to use rather more lymphocytes, e.g. 1 ¥ 108/kg. A scheme for immunotherapy in myeloma is shown in Fig. 33.5. In other diseases, such as acute leukaemias and myelodysplasia, the impact of DLI has been less great.

Post-BMT monitoring

With the advent of PCR it was possible to monitor leukaemic clones using sensitive molecular techniques. It became apparent that the loss of the molecular marker of malignancy in CML and some cases of ALL occurred gradually after transplant and not immediately. This suggested that there was some continued immune surveillance in addition to the initial impact of the radiotherapy and highdose chemotherapy used to prepare patients for BMT. Donor lymphocyte infusions

In the early 1990s it was shown that those patients with CML who suffered recurrence of their leukaemia after transplant could be induced into a

Cytotoxic T-cell therapy

Another form of potent GVL or antilymphoma is that demonstrated by Brenner and his colleagues in treating Epstein–Barr virus (EBV)-related lymphoproliferative disorder (LPD) in children who had received unrelated donor BMT. Intensive immunosuppression was given to ensure the graft was not rejected. This led to increased risk of reactivation of EBV, which triggered LPD. If untreated this progressed into an aggressive lymphoma. EBV causes glandular fever as well as Burkitt’s lymphoma in other patient groups. Although in the very early stages EBV LPD can be treated with antiviral therapy (aciclovir), once established it is likely to progress and respond poorly to conventional antilymphoma chemother385

Chapter 33

Dendritic cells

VEGF Muc-1

Plasma cells

Fas L TGF-b

Lymphocytes

TGF-b

IL-1b, IL-6

Fibronectin, collagen

TGF-b

Stromal cells Fig. 33.5 Immunosuppression in multiple myeloma. In

myeloma, a number of strategies have been identified that enable the plasma cells to prevent their killing by cytotoxic T lymphocytes. They produce transforming growth factor (TGF)-b, which stimulates stromal cells to produce more matrix materials (fibronectin, collagen) to provide a microenvironment conducive to plasma cells. The plasma cells employ three strategies to deflect or kill T lymphocytes.

apy. However, by isolating T cells and exposing them to EBV in vitro it was possible to generate clonal cytotoxic T cells that would recognize EBV antigens. These T cells were grown in the laboratory and then infused into the patients. In the main this treatment has been highly effective in both the prophylactic and therapeutic management of EBV LPD. These data show that by presenting tumourrelated antigens, it is possible to generate sufficient antitumour effect to induce remissions. It is of interest that the one disease that has been treated effectively with immunotherapy is a virus-driven malignancy. In the past few years more progress has been made in developing strategies for using cytotoxic T cells against viruses in the BMT setting than for antitumour indications. This reflects the fact that viruses possess foreign antigens not pos386

(1) Muc-1 antigen expression may cause apoptosis of T cells. (2) Fas L (Fas ligand) binding to Fas on T cells will also lead to apoptosis. (3) TGF-b secretion prevents T lymphocytes responding appropriately to interleukin (IL)-2 and they fail to increase in numbers sufficiently to kill plasma cells. Vascular epithelial growth factor (VEGF) may prevent ingress of dendritic cells and so prevent antigen presentation to T cells.

sessed by the human patient, and the relative lack of progress in antitumour immunotherapy over the past 5 years, beyond basic DLI approaches, suggests that this is a complex area difficult to overcome. Future of immunotherapy

Approaches to immunotherapy may be either passive or active. Passive immunotherapy

This involves the use of monoclonal antibodies. Studies in the 1980s using mouse monoclonal antibodies produced some encouraging results in Bcell malignancies but immunity to mouse protein prevented prolonged effects. Currently, antibodies

Stem cell transplantation and immunotherapy

engineered to be mainly human in origin have overcome this.

conjoined immunoglobulin idiotype component ensures specificity against the B-cell malignancy.

CD20 Antibodies to CD20 present on B lymphocytes have proven of value in the management of lowgrade B-cell malignancies.

Tumour cell vaccination This may involve: • apoptotic cells; • fresh tumour cells irradiated; or • apoptotic and/or fresh tumour cells fed to dendritic cells to process tumour antigens.

CD40 Such antibodies hold out great promise for treating more aggressive tumours. In mice, not only can lymphomas expressing CD40 be successfully treated by antibodies against CD40, but such immunotherapy appears to induce immunity such that the mice are resistant to subsequent challenge by tumour cell injections. Active immunotherapy

Tumour cell antigens, i.e. the proteins, may be used as the target for vaccination. More effect seems to be obtained with DNA vaccines, in which the DNA encoding the target gene or protein is used as the vaccine. Cancer vaccines have been reviewed in detail (see Further reading). Examples of antigens to target using immunotherapy approaches are: • virus antigens, e.g. EBV LPD; • mutated proto-oncogenes, e.g. p53; • mucins, e.g. Muc-1 in multiple myeloma and breast cancer; • idiotypic proteins, e.g. for B cells, have been used to immunize a sibling donor against the myeloma paraprotein of the recipient; • oncofetal antigens such as a-fetoprotein; and • products of chromosomal translocations such as t14;18 and bcr/abl.

Dendritic cell therapy Dendritic cells are essential for processing antigen to the immune system and many malignant conditions suppress the normal immune system and prevent antitumour effects. By taking dendritic cells out of the body and exposing them to tumour antigens (so-called pulsing), dendritic cells carrying specific antitumour antigens may be reinfused into the patient and stimulate T-cell responses in vivo. Cytotoxic T-cell therapy Such cells can be isolated and developed as performed by Brenner’s group for EBV LPD. One major problem is that tumour cells have developed specific strategies for avoiding control by the immune system of the host. Thus an understanding of how cancer cells avoid destruction is needed before detailed immunotherapy protocols can be developed. For example, in multiple myeloma it is known that antigens such as Muc-1 and Fas, and secretion of the cytokine transforming growth factor-b can adversely influence T cells to enable the tumour cells to proliferate. A current possible schema for this is shown in Fig. 33.6.

Conclusion DNA vaccines These show great promise since they can be more easily engineered than peptide antigens for vaccination and composite antigens can be created. An idiotype motif has been constructed together with fragment C of tetanus toxoid to immunize patients against their B-cell malignancy. The tetanus toxoid acts as a ‘danger signal’ or recall antigen, since most individuals have been immunized previously against tetanus toxoid, and the

At present there are still no clearly developed indications or proven strategies for cellular immunotherapy other than for generating GVL against CML (see Fig. 33.6) (and possibly myeloma) and for the prevention and treatment of EBV LPD. The last 5 years have been disappointing in failing to deliver any of the above approaches other than in single-centre studies or small-scale pilot trials. The exploitation of our 387

Chapter 33

Diagnosis

Disease control first chronic phase (CPI)

Collect HSC T lymphocytes Dendritic cells

Further cytoreductive therapy (? autograft)

Preparation of bcr/abl directed Cytotoxic T cells Dendritic cells pulsated with bcr/abl DNA vaccine

Minimal residual disease (molecular evidence only)

Complete remission Bone marrow and molecular Cure Molecular relapse (a)

Diagnosis

Collect HSC T lymphocytes Dendritic cells

Possible future routine use to prevent infection and maintain remission

Disease control first chronic phase (CPI)

HLA-compatible donor (sibling or unrelated)

Allogeneic transplant

Haemopoietic stem cells

Complete remission Bone marrow and molecular

Peripheral blood lymphocytes

Complete remission Bone marrow and molecular

Post-transplant virus infections CMV EBV +/– lymphoma (b)

388

Molecular relapse

Fig. 33.6 Future immunotherapy

strategies in chronic myeloid leukaemia (CML). Imatinib has transformed the initial treatment of CML, but may not be curative. From experience with allogeneic BMT it seems likely that some form of graftversus-leukaemia effect will be needed to maintain molecular and clinical remission. Ideally, this might involve autologous cells (dendritic cells and T lymphocytes) with no risk of graftversus-host disease. Patients with an allogeneic donor may still benefit from a transplant, particularly if immunotherapy enables the use of reduced-intensity conditioning. CMV, cytomegalovirus; EBV, Epstein–Barr virus; HSC, haemopoietic stem cells.

Stem cell transplantation and immunotherapy

increased knowledge of the immune system has not been easy. At the moment, the routine role of immunotherapy in cancer seems further away than a few years ago. However, its huge potential means that further effort and trials are clearly justified. The most likely application appears to be in a more refined approach to the use of DLI after transplant, and the advent of RIC transplants may permit the extension of haemopoietic stem cell transplants to more non-malignant diseases in which immune modulation may have a role. Meanwhile, what is the long-term future for BMT generally? These transplants can save life in many patients with incurable leukaemias and lymphomas. Patients who survive the first 3 years are likely to enjoy long-term survival, although life expectancy does not return to normal. RIC transplants will extend the benefits to more patients who might have been unfit to undergo the rigours of a myeloablative procedure. Whether the extension of RIC transplants to a wider range of malignant diseases is wise will become apparent in the next few years. Certainly, GVL effects are most potent in the context of minimal residual disease in slow-growing malignancies, such as CML and myeloma, in which there also happen to be specific tumour antigens (bcr/abl products and paraprotein). DLI-induced GVL struggles to compete with rapidly emerging relapses of acute leukaemias. Additional approaches are needed to deal with those patients whose primary disease is poorly responsive to current chemoradiotherapy.

Dazzi F, Goldman JM. Adoptive immunotherapy following allogeneic bone marrow transplantation. Annu Rev Med 1998; 49: 329–40. Goldman JM, Schmitz N, Niethammer D, Gratwohl A. Allogeneic and autologous transplantation for haematological diseases, solid tumours and immune disorders: current practice in Europe in 1998. Accreditation Sub-Committee of the European Group for Blood and Marrow Transplantation. Bone Marrow Transplant 1998; 21: 1–7. Gratwohl A, Baldomero H, Passweg J, Frassoni F, Neuderweisser D, Schmidz, N, Urbano-Ispizua A. Hematopoietic stem cell transplantation for hematological malignancies in Europe. Leukaemia 2003; 17: 941–59. Greten TF, Jaffee EM. Cancer vaccines. J Clin Oncol 1999; 17: 1047–60. Kolb HJ, Holler E. Hematopoietic transplantation: state of the art. Stem Cells 1997; 15 (Suppl. 1): 151–7; discussion 158. Laupacis A, Fergusson D. Erythropoietin to minimize perioperative blood transfusion: a systematic review of randomized trials. The International Study of Perioperative Transfusion (ISPOT) Investigators. Transfus Med 1998; 8: 309–17. Slavin S. Immunotherapy of cancer with alloreactive lymphocytes. Lancet Oncol 2001; 2: 491–8. Socie G, Stone JV, Wingard JR, Weisdorf D, HersleeDowney PJ, et al. Long-term survival and late deaths after allogeneic bone marrow transplantation. N Engl J Med 1999; 341: 14–21. Stovek J, Joseph A, Espine G, Dawson MA, Douek MC, Sullivan KM et al. Immunity of patients surviving 20 to 30 years after allogeneic or syngeneic bone marrow transplantation. Blood 2001; 98: 3505–12. Thomas ED. Does BMT confer a normal life span? N Engl J Med 1999; 341: 50–1.

Further reading

Web resources

Barker JN Weisdorf DJ, Defor TE, Blazar BR, Miller JS, Wagner JE. Rapid and complete donor chimerism in adult recipients of unrelated donor umbilical cord blood transplantation after reduced-intensity conditioning. Blood 2003; 102: 1915–19. Barrett AJ, van Rhee F. Graft-vs.-leukaemia. Baillières Clin Haematol 1997; 10: 337–55.

http://www.ibmtr.org International Bone Marrow Transplant Registry (IBMTR) and the Autologous Blood and Marrow Transplant Registry (ABMTR). Website of the International Allogeneic database and the Autologous database for North and South America. http://www.ebmt.org/ European Group for Blood and Marrow Transplantation (EBMT).

389

Chapter 34

Gene therapy Colin G. Steward and Marina Cavazzana-Calvo

The ability to cure disease by genetic manipulation is one of the most fascinating concepts in modern medicine. Although widely discussed in the early 1970s, ‘gene therapy’ underwent an explosion of popularity in the early 1990s. This was fuelled by improved methods of blood cell purification/ expansion, the development of more sophisticated vectors (the vehicles used to transport genes into cells) and the birth of the biotechnology industry. By 2003, more than 640 gene therapy protocols had been sanctioned, and more than 3500 patients recruited. This chapter concentrates on basic principles, methods of gene delivery and the major clinical benefits reported so far in monogenic diseases of the immune system. However, a considerable section is also devoted to potential uses in cancer therapy, since this has come to be the major focus of clinical trials and financial investment.

Basic concepts and definitions Somatic gene therapy

Research is concentrated on genetic manipulation of somatic cells, since manipulation of gametes is unethical and is prohibited in all countries. Transfection and transduction

The process of introducing extraneous genetic material into a cell is termed ‘transfection’ and may be performed either in the laboratory (in vitro) or by direct injection into tissue or blood (in vivo). The gene being transferred is termed the ‘transgene’. Once it has been introduced by a viral vector, the cell is said to be ‘transduced’. Genes may be trans390

ferred into cells using either chemical methods (transfection) or a wide variety of physical or viral vectors, summarized in Tables 34.1 and 34.2. Methods of gene transfer

Vectors are the vehicles used to carry genes into cells and are usually modified viruses. Although no perfect vectors exist, an ideal vector system for clinical purposes needs to have the following characteristics. • Highly efficient, transducing a large proportion of target cells • Result in stable integration so that the therapy is long-lasting. This can only be achieved by vectors, e.g. retroviruses and adeno-associated viruses (AAV), which insert their genetic material permanently into that of the host, so that it is replicated in all cell progeny. • Either transfect specific organs or cell types (targeted therapy) or transfect cells indiscriminately but ensure that the gene product is only expressed in cells requiring a therapeutic effect (tissuespecific or disease-specific expression). This necessitates the identification of physiological gene control mechanisms such as the locus control region (LCR) that promotes globin gene expression in red blood cells exclusively. In vitro techniques allow manipulation of target cells and include the following. • Cytokine stimulation of haemopoietic progenitor cells, in order to stimulate cell proliferation and improve retroviral transduction rate. • Selection of successfully transduced cells: the commonest methods involve introducing either a neomycin resistance (neoR) gene (which imparts resistance to successfully transduced cells when

Gene therapy Table 34.1 Major viral vectors used in gene therapy experiments.

Virus type

Advantages

Disadvantages

Retrovirus

Permanent integration

Small capacity for gene inserts (6–7 kb) Low titre (106–107 virus particles/mL), necessitating large volumes of stock Potentially pathogenic and carcinogenic Can only infect cycling cells

Adenovirus

Capacity for large gene inserts (7–36 kb) Can be prepared at high titre (1011–1012/mL) Infect non-dividing cells

Immunogenic, causing inflammatory reaction Lost on cell division because not permanently inserted into DNA

Adeno-associated virus

Insert preferentially on chromosome 19, a ‘safe’ area of DNA Non-pathogenic Can be prepared at high titre (106–1012 virus particles/mL)

May be poor for infecting HSC Low infection efficiency Small capacity for gene inserts (2–4.5 kb) Requires helper adenovirus for infection

Herpes virus

High potential as vectors for CNS-directed gene therapy Capacity for large gene inserts (10–100 kb)

Potentially pathogenic

Lentivirus

Permanent integration Able to divide non-dividing cells

Major concerns about safety

CNS, central nervous system; HSC, haemopoietic stem cell.

Table 34.2 Physical agents used in gene therapy experiments.

Category

Advantages

Disadvantages

Naked DNA

Directly targeted injection (e.g. to muscle via ‘gene gun’)

Many tissues inaccessible Low transfection efficiency Expression lost if cells divide

Liposomal DNA

Non-pathogenic Capable of transferring large genes

Low transfection efficiency Expression lost if cells divide

DNA–protein conjugates

Tissue targeting possible

May be degraded in circulation Low transfection efficiency Expression lost if cells divide

Oligonucleotides

Cheap and easily prepared Non-toxic Can target specific genes

Transient effect Non-specific binding Poor efficiency May be degraded in circulation

Ribozymes

Potential for highly specific RNA cleavage/repair Catalytic: each particle can perform multiple reactions

Developmental: may be problems of delivery, degradation, expression or specificity

391

Chapter 34

grown in the presence of a neomycin analogue called G418) or a gene for a fluorescence-activated cell sorter (FACS)-selectable marker (e.g. CD24 or green fluorescent protein). Selected cells may then be further expanded. In vivo techniques tried have included: • naked DNA coated onto gold particles and fired into tissue via a ‘gene gun’ (an inefficient method of gene transfer); • injected/inhaled adenoviral or AAV vectors; • protein–DNA conjugates, i.e. DNA coupled to proteins for which specific cellular receptors exist, in order to allow tissue targeting. Viral vectors

The major characteristics of the more frequently used viral vector systems are summarized in Table 34.1. Oncoretroviruses

The most widely used retroviral vectors have been those based on the Moloney murine leukaemia virus (MLV). The core structure of a retrovirus is shown in Fig. 34.1. These allow permanent integration of therapeutic genes and infect a wide variety of cell types. Depending on the envelope type of the virus it may only be able to infect cells of the species that it was originally isolated from or identified in (termed ‘ecotropic’) or to more broadly infect mammalian cells (‘amphotropic’).

LTR

gag

pol

env

LTR

Packaging signal (psi)

Fig. 34.1 The core structure of a retrovirus: the genome

comprises gag, pol and env genes (encoding core proteins, reverse transcriptase and viral envelope proteins, respectively) sandwiched between two long terminal repeat (LTR) elements. These genes can synthesize an empty virion but require a packaging signal (psi) to insert a copy of the viral genome, thus creating a virus capable of selfreplication.

392

The MLV-derived vector can only integrate successfully into the genome of cells which pass through mitosis. They can also only accommodate small genes (maximum size of incorporated DNA is 8 kb). Larger genes can still be utilized in their complementary DNA (cDNA) form, where the non-coding sequences (introns) have been spliced out, leaving only the coding segments (exons). This explains the use of cDNA rather than whole genomic DNA in most protocols. Problems are that (i) intronic sequences may contain control elements critical for transcriptional regulation and (ii) the limited capacity precludes the inclusion of large stretches of upstream and downstream DNA which may also contain regulatory sequences. Therapeutic viral particles (retroviruses containing a therapeutic gene) are synthesized in modified murine fibroblast cells, termed a packaging cell line. In the simplest form of packaging cell line (‘first generation’), the cells contain viral genome from which the packaging signal has been deleted. Instead this signal is ligated to a plasmid containing a therapeutic gene and transfected into packaging cells. A successfully transfected packaging cell will then release into overlying supernatant culture fluid viral particles containing the therapeutic gene, packaging signal and long terminal repeat (LTR) regions (which contain retroviral promoter elements) but no other replication-competent retroviral genes. These particles are capable of one round only of infection, thereby inserting the therapeutic gene randomly into host DNA. Retroviral vectors appear prone to silencing in humans at a much greater rate than in animals. This results in rapid loss of expression, although transgene can be shown to persist by genetic analysis such as polymerase chain reaction (PCR). Processes which may explain silencing include the following. • Methylation of promoter sites. • Position effects imparted by chromosomal sequences at their integration site: these may be reduced by introducing chromatin insulator elements (e.g. cHS4). • Immune responses against vector or transgene sequences: silencing occurs more slowly when immunosuppressant drugs are given.

Gene therapy

Lentiviruses

The most significant hope for widespread effective application of gene therapy resides with this group of vectors derived from human or animal immunodeficiency viruses. Lentiviral vectors have the following characteristics. • They can infect both dividing and non-dividing cells as their ‘preintegration complex’ (viral ‘shell’) can get through an intact nuclear membrane. Cell cycling is therefore not needed for infection of cells, allowing the possibility of in vivo gene therapy and avoiding differentiation or other deleterious effects on target cells by in vitro culture. • They show wide infectivity of non-dividing or terminally differentiated cells, including neurones, haemopoietic stem cells (HSC), muscle and liver cells. • They are associated with a low incidence of immune responses against vector. • Target specificity is provided by protruding membrane proteins in the lipid coat of the virus; these include gp120 which provides the CD4 T-cell specificity of human immunodeficiency virus (HIV). In the future it may be possible to engineer vectors specific for target cell type. • The cell specificity of HIV can be greatly broadened by exchanging the gene encoding gp120 for genes encoding other glycoproteins, e.g. G glycoproteins from vesicular stomatitis virus. This is called ‘pseudotyping’ the vector. As one example, a vector pseudotyped with glycoprotein from RD114 feline endogenous virus is relatively resistant to inactivation induced by human complement and has shown augmented transduction of human primary blood lymphocytes and CD34+ cells. • Lentiviral vectors can be designed as ‘selfinactivating’ vectors by making deletions in the LTR of the virus. These inactivate the LTR promoter and eliminate the production of vector RNA, with transcription being driven from an exogenous viral or cellular promoter that is inserted into the lentiviral vector. There is interest in two other classes of lentiviral vectors. • Non-human lentiviral vectors: examples are vectors derived from feline immunodeficiency virus (present in 2–20% of cats but never known

to have infected humans) and equine infectious anaemia virus. These may be less likely to elicit immune responses but have so far not been well developed for human applications. • Spumaviruses (foamy viruses): these are poorly characterized due to lack of association with human diseases, but can often be isolated from primate bone marrow cultures. They appear to have a larger packaging capacity, can infect more cell types and are better at transducing nondividing cells than MLV vectors. Adenoviruses

The major attractions of adenoviruses are as follows: • a high propensity to infect respiratory epithelium, neurones and hepatic cells, all major targets for gene therapy; • the ability to infect both dividing and quiescent cells; • the ability to be prepared in high titre, reducing the volume of stock solutions and glassware needed for transfection procedures; • large capacity, carrying up to 36 kb of transgene. A major drawback is that DNA is not integrated but held in the form of ‘episomes’, which are lost when the cell divides. This necessitates repeated treatments with a highly immunogenic virus. Adeno-associated virus

Major attractions of this vector are: • the propensity to integrate at a specific site on chromosome 19, i.e. 19q13 (an area lacking known tumour-suppressor genes); • the ability to transduce non-dividing cells; • the fact that it is naturally replication deficient (requiring coinfection with a helper virus such as adenovirus for productive replication); • they are less immunogenic than standard adenoviral vectors (although humoral responses may occur to capsid proteins). These features make AAV vectors a popular choice for in vivo gene therapy experiments. There are promising animal studies to suggest that infection of liver or muscle could act as a source of 393

Chapter 34

coagulation factors but so far no beneficial results in humans. However, it is not clear that human CD34 cells have appropriate surface receptors for AAV infection and enthusiasm is tempered by the small capacity (< 5 kb) of these vectors. Historically, the production of these vectors has relied on coinfection of helper cell lines with adenoviruses, requiring subsequent separation from the more immunogenic adenovirus. This has been overcome by cotransfection of producer lines with adenovirus helper genes. The problems of capacity and poor tropism for human haemopoietic cells may be improved by creation of vectors with chimeric capsids containing adenovirus type 35 fibres. Herpes viral vectors

The particular attractions of herpes-based vectors are: • their predilection for infecting nervous tissue; and • large capacity for inserted transgene (30–40 kb). Non-viral vectors (see Table 34.2) Cationic liposomes

• A plasmid carrying a therapeutic gene is linked by charge to the liposome surface. • Attractive for gene therapy in that there is widespread experience with liposomal drug formulations, they can be given repeatedly and the process is non-infectious, reducing regulatory problems. • Subject to degradation in vivo and gene expression is only transient. • Gene transfer efficiency is low, although it may be possible to improve this in the future, for example by coating the liposomes with polyethylene glycol (PEG) to slow clearance from the blood, and by conjugating them with antibodies to receptors on specific cells or viral cell fusion proteins. • Liposomes are typically 100 nm in size, but nanoliposomes (25 nm) coated with positively charged peptides are small enough to enter pores in the nuclear membrane.

394

DNA–protein conjugates

DNA is coupled via a linker arm to a protein recognized by a cell surface receptor. A typical example is transferrin–polylysine–DNA, where the transferrin will bind to receptors on hepatocytes. These vectors share many of the problems of liposomal vectors. Oligonucleotides and antisense approaches

Approaches that can neutralize the process of gene expression are of major interest in the gene therapy of cancer and HIV. Examples include the following. • Oligonucleotides: DNA fragments, typically 15–25 bases in length, of complementary sequence to either the normal, sense mRNA transcript (antisense therapy) or specific genomic DNA sequences (‘antigene’ therapy). Problems include nonspecificity of target binding, the requirement for large doses of injected oligonucleotides in order to attain adequate intracellular concentrations and degradation of oligonucleotides in the circulation (although this can be reduced by chemical modification). • Ribozymes: these are short segments of RNA which form complementary base pairing with mRNA and cleave the target in a highly specific manner. They then dissociate and repeat the process on other mRNA strands in a catalytic manner. Ribozyme technology is still in the early phase of development, but may even allow targeted repair of specific DNA mutations. • Small interfering RNA molecules: small (21–23 nucleotides) double-stranded RNA molecules were first identified in petunias but have become recognized as a powerful potential tool for gene therapy by post-transcriptional gene silencing in humans. One application may be in combating HIV infection. Targeted gene repair

Although homologous recombination can be performed in vitro where cells can be cloned and expanded, this is not possible on a therapeutic scale at present. However, there is growing interest

Gene therapy

in the use of ribozymes and chimeric DNA/RNA oligonucleotides to repair specific mutations in DNA.

Transducing haemopoietic stem cells Gene therapy using HSC has been greatly hampered by the: • inability to identify primitive stem cells with precision; • predominant dormant state of these cells (oncoretroviral transduction requires cells to proceed through the S phase of the cell cycle); • low expression of retroviral receptors on their cell surface. These problems have forced scientists to use cytokine cocktails to stimulate cell proliferation. These appear to differentiate progenitor cells, reducing the duration of the therapeutic effect, and succeed typically in transducing less than 1% of true primate/human stem cells (although this has been sufficient to correct several human immunodeficiency diseases). Scientists are now engineering vectors to contain proteins which bind specifically to haemopoietic cells.

Conditioning therapy In rare instances, spontaneous genetic mutation may revert the mutated gene in a cell to a normal genotype. If this confers a genetic advantage to that cell over diseased counterparts, the reverted cell will proliferate more successfully and produce progressive disease reversion. This phenomenon has been observed in several forms of severe combined immunodeficiency (SCID), Wiskott–Aldrich syndrome and Fanconi’s anemia and provides compelling evidence that gene therapy should be effective, even if only a minority of cells are transduced successfully. Similar evidence comes from bone marrow transplantation (BMT) where patients often engraft with a mixture of donor and recipient cells (‘mixed chimerism’) but eventually become 100% donor chimerae (e.g. thalassaemia, Fanconi’s

anaemia, SCID). This implies competitive advantage for normal (donor) cells over diseased (recipient) cells. After gene therapy such competitive advantage would be expected to favour progressive engraftment/expansion of gene-modified cells and avoid the necessity for chemoradiotherapy. However, this advantage typically appears to operate at the level of differentiated cells rather than the stem cell. In the absence of a selective advantage for the transduced cells, this can be achieved artificially by using conditioning therapy such as low-dose cyclophosmamide or busulfan.

Risks of gene therapy Immune responses

A recurring theme in the literature on gene therapy is the problem of immune responses hindering therapy. These may take two forms: Responses to the therapeutic protein

It is well recognized that tolerance to self proteins is established by 14 weeks in utero. Any new protein expressed after this time is perceived as foreign and cells expressing it become the target for an immune response. Transgene products can induce both humoral and cell-mediated responses, and the subsequent loss of transduced cells is well documented. Responses to the vector

Immune responses against viral vectors have also proved to be a major problem. These are a major problem in protocols using adenoviruses, especially where repeated administrations are required. In one trial of inhaled gene therapy for cystic fibrosis using the cystic fibrosis transmembrane conductance regulator gene (CFTR) this resulted in severe pneumonitis. The risks were most tragically demonstrated in 1999 by the death of an 18year-old American teenager, Jesse Gelsinger, who developed multiorgan failure after an adenoviral vector injection into the hepatic artery for ornithine decarboxylase deficiency. 395

Chapter 34

(a) Packaging cell Promoter

gag

Promoter

env

pol

Empty retrovirus particles Fig. 34.2 (a) ‘Third-generation’

(b) 1 Packaging cell transfected with retroviral plasmid DNA containing therapeutic gene

2 Retroviral transcript produced containing therapeutic gene

LTR

Therapeutic gene

LTR

PSI Promoter

gag

Promoter

env

pol

3 Packaged into empty retroviral particle

4 Therapeutic retroviruses released into supernatant

Risks associated with retroviral vectors Risk during production

In first-generation packaging cell lines, a single genetic recombination event will result in formation of replication-competent wild-type ‘helper’ virus, i.e. MLV. In an early gene therapy trial in rhesus monkeys, helper virus contamination resulted in some animals developing T-cell nonHodgkin’s lymphoma. This risk can be reduced by further splitting the retroviral genome on the chromosomes of packaging cells, a process exemplified in Fig. 34.2. Gene therapy regulatory authorities insist on scrupulous testing for helper virus in retroviral supernatant before clinical use. 396

packaging cell line for production of retroviral vectors: a mouse fibroblast cell line has been transfected with separate plasmids containing gag/pol and env genes, respectively, with 3¢LTR removed. Multiple recombination events are now required to produce live retrovirus containing gag, pol and env genes and an LTR. (b) The therapeutic gene, together with a packaging signal, is introduced on a separate plasmid and viral particles containing this gene are then produced and released into the supernatant. Although thought to be an unlikely event, formation of retroviruses capable of self-replication has been documented from such cells and meticulous testing for ‘helper’ virus therefore remains essential.

Risks during vector production and following administration

In vivo recombination events between retroviral vectors and endogenous retroviral sequences already present in human DNA could result in the formation of replication-competent retroviruses. In the case of lentiviral vectors based on HIV this could result in devastating new diseases, especially since a new virus would probably lack the CD4+ Tcell specificity of HIV. Self-inactivating vectors with modifications of the lentiviral LTR may overcome this risk. Random incorporation of therapeutic genes could disrupt potentially dangerous genes such as tumour-suppressor genes or activate protooncogenes. This has occurred in 2 of 14 children

Gene therapy

3 years after they received gene therapy for Xlinked SCID. In both cases the vector had inserted within or upstream of the LMO2 proto-oncogene, resulting in a T-cell lymphoproliferative disease that required chemotherapy for control. Selfinactivating lentiviral vectors are thought to reduce the risk of insertional mutagenesis.

Gene therapeutic approaches to cancer The past decade has seen rapid advances in the understanding of tumour immunology, particularly of mechanisms of immune evasion and the (relatively) specific antigens expressed by some tumours. Tumour antigen profiling is likely to advance at an even more rapid pace as DNA array profiling becomes more routine. Coupled with greater recognition of tumour-suppressor genes and transforming oncogenes, this has led to major interest in genetic therapeutic approaches to cancer. As a result, 80% of experimental protocols are directed to cancer rather than to the cure of monogenic diseases. It seems likely that combinations of immunological and genetic approaches with conventional therapy will be the next major step in improving survival. The particular role of these newer therapies may lie in eliminating the last vestiges of cancer, which have been reduced to a state of ‘minimal residual disease’ by conventional therapy, and in preventing relapse. Clinical application of many of these techniques is hampered by the fact that mouse models often falsely predict responses which do not then occur in humans, and because of our inability to target all of the cells of a tumour. Transducing all cells within a cancer may not be necessary because of a ‘bystander effect’, the death of untransduced cells surrounding a tumour cell that is dying as a consequence of genetic manipulation. This may result from intercellular transfer of harmful cytokines or prodrugs (see below) or due to immunological reactions, although the mechanisms are not well understood.

Gene marking

The integration of foreign ‘marker’ DNA sequences into cells in vitro allows their fate to be followed after reinfusion by PCR amplification of gene sequences or FACS detection of expressed proteins. This has allowed demonstration of the source of relapse when this follows autologous bone marrow transplantation for acute and chronic myeloid leukaemia and tracking of Epstein–Barr virus (EBV)-specific cytotoxic lymphocytes given to abrogate post-transplant lymphoproliferative disease (EBV-LPD). Suicide genes

This term is used to describe genes that convert non-toxic prodrugs into cytotoxic agents. The most widely employed has been thymidine kinase from herpes simplex virus (HSV-Tk). This phosphorylates the antiviral drug ganciclovir, producing a metabolite that interferes with DNA synthesis on subsequent cell division and leads to cell death. For example, HSV-Tk transduction of T cells administered following allogeneic BMT (for control of EBV-LPD or leukaemic relapse) has enabled reversal of acute graft-versus-host disease by administration of ganciclovir in some patients. However, there are various disadvantages to this approach: • the prolonged transduction processes required currently may cause loss of function in the target lymphocytes; and • transfected cells may be lost or silenced due to immune responses directed against HSV-Tk. There are many other candidate suicide gene/prodrug combinations, including cytosine deaminase (which converts 5-fluorocytosine to 5-fluorouracil) and P450-2B1 (which converts cyclophosphamide to the active metabolite 4hydroperoxycyclophosphamide). It is essential to target the suicide gene only to those cells that are to be killed. For this it is possible to take advantage of the enhanced mitotic rate of tumour cells or, in brain tumours, the fact that normal neurones do not divide. An alternative technique of tumour targeting is to put the suicide gene under the control of a promoter that is only 397

Chapter 34

active in tumour cells, e.g carcinoembryonic antigen in colorectal carcinomas. This is termed ‘gene-directed enzyme prodrug therapy’. Increasing the drug resistance of normal haemopoietic progenitors

It is possible to confer resistance to cytotoxic drugs by transducing them with genes such as MDR-1, O6-alkyl DNA transferase, dihydrofolate reductase or aldehyde dehydrogenase. If normal bone marrow cells are transfected with these genes and returned to the patient before intensive chemotherapy, this could potentially shorten the subsequent period of pancytopenia or allow dose intensification. The principal problem with this elegant approach is that the threshold of major toxicity for other organs is close to that of bone marrow, limiting the potential for dose intensification. Immunotherapeutic approaches

Tumours appear to impair immunological responsiveness by generating tumour antigen-specific tolerance and by inducing global immunosuppression. This may be explained in part by factors such as: • downregulation of expression of tumour antigens, major histocompatibility complex (MHC) class I or costimulatory molecules or progressive selection of tumour cells expressing low amounts of these proteins; • downregulation of the z-chain of the T-cell receptor (TCR) CD3 complex (a key protein in Tcell signalling) on tumour infiltrating lymphocytes; • production of immunosuppressive cytokines, such as transforming growth factor-b and interleukin (IL)-10; • induction of lymphocyte apoptosis by tumour cells using the Fas/Fas ligand (Fas L) pathway. Yet many tumours express potential novel protein targets (e.g. immunoglobulin and TCR idiotypes, fusion proteins secondary to chromosomal translocation) and tumour antigens at high concentration. This has encouraged a number of developments based either on: • preparing tumour-specific cytotoxic T cells ex 398

vivo (e.g. for trials in controlling disseminated melanoma or EBV-LPD); or • enhancing the immunogenicity of tumour cells, thus creating a ‘tumour vaccine’. Tumour vaccines

Tumour vaccines have been created by transducing tumour cells with genes encoding proteins involved in antigen presentation or in eliciting inflammatory reactions. After return to the animal these sometimes stimulate immunological antitumour responses that eradicate the injected cells and may even cause the death of metastatic non-transduced tumours or confer resistance to subsequent tumour injections. Major interest is currently focusing on the identification, characterization, expansion and reinjection of dendritic cells pulsed with tumour antigens or peptides. In humans, efforts have principally been concentrated on those tumours which show occasional spontaneous regression or which express fetal antigens not normally expressed after birth. These include melanoma, renal cell carcinoma and neuroblastoma. The commonest approach has been to transduce tumour cells with cytokine genes, such as those encoding IL-2 or granulocyte– macrophage colony-stimulating factor (GM-CSF), thereby making them the target of an inflammatory response. Some encouraging responses have been seen. Tumour infiltrating lymphocytes

There is much evidence that the lymphocytes contained within tumours are either themselves suppressed or actively suppressive of immune responses. Examples include: • an increased frequency of a regulatory phenotype (Treg: CD4+CD25+); • high production of suppressive cytokines (e.g. IL-4 and IL-10); and • reduced TCR z expression. These characteristics can be partially reversed by culture and transfection with proinflammatory cytokines such as IL-2. Homing and tumour responses to gene-modified cells have been largely disappointing.

Gene therapy

Manipulation of tumour-suppressor genes and oncogenes

Some of the earliest interest centred on supplementing tumours with mutated tumoursuppressor genes with normal (wild-type) gene copies, e.g. injection of wild-type p53 directly into bronchial carcinomas or gliomas, or into the peritoneal cavity of women with disseminated ovarian carcinoma. A major limitation of this approach is the necessity to genetically modify a large proportion of the cells in tumour nodules, which often contain large necrotic elements and have a compromised blood supply. Typically, effects are only seen close to the injection site. Where dominant oncogenes (e.g. myc, ras) are implicated, research is concentrated on antisense technologies (see above). Modifying angiogenesis

Tumours above 2 mm in diameter are critically dependent (for adequate supply of oxygen and nutrients) on the development of new blood vessels from an existing vascular network. This process, called angiogenesis, depends on a balance of cytokines responsible for stimulating and suppressing the growth of blood vessels. Angiogenesis presents an attractive target for interfering with tumour growth and spread, either by: • developing drugs that suppress endothelial cell responses to tumour-derived growth factors; or • directly inhibiting the pro-angiogenic activities of other cell types in tumours. Candidates include the angiogenesis inhibitors, endostatin and linomide, and a 50-kDa proteolytic fragment of fibrinogen.

Progress in gene therapy for monogenic haematological disorders Gene therapy was initially largely targeted towards inherited single-gene disorders, especially those governed by small, unregulated and easily transferable genes. This has resulted in detailed, highly publicized experiments concerning compar-

atively rare disorders, for example SCID in its Xlinked form (SCID-X1) or secondary to adenosine deaminase (ADA) deficiency (ADA-SCID), and cystic fibrosis. However, there is now encouraging animal work on commoner disorders involving large genes with complex control mechanisms (e.g. haemoglobinopathies). It should be remembered that truly successful gene therapy of such conditions depends on the following caveats. • It is only possible for recessively inherited diseases. In dominant diseases the abnormal proteins encoded are essentially toxic. Successful approaches to dominant diseases will therefore require widespread blockade of RNA production, a much more difficult task than gene supplementation. • It needs permanent gene integration for lasting cure without repeated rounds of therapy. • It needs transgene expression without eliciting immune responses. Some workers have proposed continuing immunosuppression after gene therapy in order to reduce this risk. However, this carries long-term risks and is not viable. • It may be difficult in diseases where a high proportion of cells must be modified, e.g. congenital erythropoietic porphyria, where a majority of red cell precursors would need to be modified to prevent the excess production of toxic porphyrins (there is no evidence of selective advantage for normal red cell precursors over diseased cells). Important considerations will be the number and type of cells that need to be transduced for success. Although the difficulty of transducing HSC has already been discussed, some haematological diseases may be curable by transfecting liver or muscle cells, or by the use of implantable reservoirs of gene-corrected cells. If this improved factor VIII in a haemophiliac by 5–10%, this could dramatically improve lifestyle. Unfortunately, no study has managed to show so far that this threshold is reachable. Immunodeficiency diseases

Immunodeficiency diseases represent ideal targets for gene therapy for the following reasons: • they are monogenic; 399

Chapter 34

• responsible genes have been identified in a number of conditions; • they are very severe diseases, being lethal within the first year in absence of any treatment; • symptoms can be largely alleviated or abolished by BMT from matched siblings without the use of conditioning chemotherapy; • full chimerism is not required for cure; • normal donor cells (especially T lymphocytes) carry competitive advantage over diseased cells; • immune response to vectors and therapeutic genes are less likely. ADA-SCID

ADA is responsible for the detoxification of metabolites in the purine salvage pathway. Deficiency causes accumulation of deoxyATP, which inhibits cell division and causes death by apoptosis. This results in early block of lymphoid differentiation and decreased survival of mature T cells and early lymphoid precursors, resulting in SCID. The most severe symptoms of this disease can be ameliorated by use of infused enzyme (modified with PEG in order to enhance its persistence in the circulation). This allows patients lacking an appropriate donor for BMT to be kept in reasonable health while attempting gene therapy for long-term disease correction. The first administration of human gene therapy was to a 4-year-old girl with ADA-SCID on 14 September 1990 by W. French Anderson and colleagues at the National Institutes of Health, Bethesda, USA. Since that time, five clinical trials have been conducted in this condition. All cells administered have been transduced in vitro by MLV retroviral vectors. In the early studies T cells were transduced but in more recent studies CD34+ HSC from marrow or cord blood have been used. Major findings are as follows. • Results were generally poor in children maintained on PEG-ADA following gene therapy. This is probably because accumulation of toxic metabolites may offer a selective advantage to cells that produce adequate vector-derived ADA and this advantage is lost in the presence of a therapeutic level of PEG-ADA.

400

• Transduced T cells have lasted for more than 12 years even after in vitro activation to induce their cycling. Therefore, at least in these cells, the MLV promoter can drive transgene expression for long periods without gene silencing. • The best results were seen in children who received transduced bone marrow HSC and who were given mild myeloablation with busulphan before receiving modified cells. Sustained HSC engraftment was followed by differentiation into multiple lineages, increased lymphocyte counts, improved immune function and reduction in toxic metabolites. • However, even immunodeficient patients were able to mount immune responses to both the retroviral envelope and the fetal calf serum used in transfection procedures. SCID-X1

• Accounts for 50–60% of all cases of SCID. • Results from mutation of the gene encoding the common g-chain subunit of the haemopoietic cytokine receptor family (IL-2, IL-4, IL-7, IL-9, IL15 and IL-21). • Between 1999 and 2002, 14 children received HSC retroviral gene therapy in Paris or London. • Thirteen patients developed substantial or complete correction of T- and B-cell function (although intriguingly reconstitution of natural killer cell function has mostly been poor in the long term). These patients are alive and well without any additional treatment. • Two children developed uncontrolled Tlymphoproliferation due to insertional mutagenesis inducing hyperexpression of the LMO2 protooncogene. They have been treated using acute lymphoblastic leukemia chemotherapy protocols and are in complete remission. Chronic granulomatous disease

Chronic granulomatous disease (CGD) results from mutations in any of four genes encoding essential subunits of respiratory burst NADPH oxidase, the enzyme complex required for the production of reactive oxygen intermediates in phago-

Gene therapy

cytes. The absence of oxidants results in recurrent bacterial and fungal infections and development of inflammatory granulomas. The potential for successful gene therapy in CGD is highlighted by the finding that female carriers with only 5–10% NBT-positive cells (due to highly skewed X-inactivation) are usually healthy and the knowledge that post-BMT patients with mixed chimerism and as little as 30% donor neutrophils remain well. Mouse models of CGD transplanted with bone marrow cells transduced with gp91phox (the gene defective in X-linked CGD) after lethal irradiation show neutrophil expression at 10% and superoxide production at approximately one-third that of wild-type neutrophils. Human phase I clinical studies in CGD patients have yet to produce clinically beneficial numbers of corrected neutrophils for significant periods. Current trials are focusing on techniques that allow higher levels of gene transfer efficiency (50–80%) into CD34+ haemopoietic progenitor cells and use of conditioning chemotherapy. Haemoglobinopathies

b-Thalassaemia and sickle cell diseases should be amenable to gene therapy if normal b-globin or anti-sickling genes, respectively, could be expressed at sufficiently high levels in the red blood cell lineage. An attractive approach for sickle cell disease is introduction of a g-globin gene, since g-globin is a much stronger inhibitor of HbS polymerization than b-globin. Unfortunately, progress in developing gene therapy for these disorders has been slow due to: • the large size of the a- and b-globin genes; • poor transduction efficiency of HSC by oncoretroviral vectors; • complex mechanisms which control globin gene expression, notably the existence of a specific LCR; • instability of b-globin/LCR retroviral vectors partly caused by unwanted splicing of retroviral RNA before incorporation into virions; • gene expression from retroviral vectors is position dependent, being influenced by elements in surrounding chromatin.

Progress has come with the following observations. • The use of lentiviral vectors incorporating RNA export elements that allow incorporation of more genetic material and larger LCR segments; these approaches have allowed dramatic improvement in mouse b-thalassemia and SCID models. • Identification of one amino acid residue from the g-globin gene that appears to be responsible for most of the anti-sickling properties, and incorporation of this in a b-globin gene variant. In mice this has produced erythroid-specific accumulation of anti-sickling protein in most red blood cells. Haemophilia

Haemophilia A and B make attractive candidates for gene therapy because of the dramatic improvements in quality of life which result from relatively small increments in the levels of clotting factors (to 5% or above). Clinical effect would not be dependent on transducing HSC and might be achieved by targeting either liver or muscle. Progress so far in these diseases includes the following observations. • Promising demonstrations of effect in dog models of haemophilia A and B using portal venous or multiple intramuscular injections and AAV vectors. Factor levels up to 2–4% of normal have been seen for up to 14 months after the gene therapy procedure. • Evidence of effective gene transfer also using high-capacity adenoviral or lentiviral vectors. • Phase I and II studies in humans have showed only limited efficacy (with plasma levels less than 1% of normal) but no significant toxicities.

Gene therapy for HIV infection Potential techniques of gene therapy for HIV infection include the following. • Use of antisense constructs or ribozymes which can negate HIV gene expression. In vitro these are capable of substantial reduction, though not cessation, of HIV replication. • Infection of CD4+ cells with HIV-based vectors

401

Chapter 34

containing HSV-Tk under the control of the HIV LTR promoter. If these cells subsequently become infected with wild-type HIV, HSV-Tk is upregulated and renders the cell sensitive to killing by ganciclovir. • Expression of small inactivating RNA species from vectors. This could interfere with production of proteins involved in HIV-1 infection (e.g. CCR5 coreceptor) and so reduce the rate of lymphocyte infection.

Vascular diseases In occlusive vascular disease stimulation of angiogenesis is being investigated widely. • The safety and tolerability of therapeutic angiogenesis by gene transfer has been demonstrated in phase I clinical trials. • Evidence of efficacy from early phase II studies of angiogenic gene therapy for ischaemic myocardial and limb disease is limited. • An alternative strategy to the use of transgenes encoding angiogenic growth factors is therapy based on transcription factors such as hypoxia inducible factor-1a, which can regulate the expression of multiple angiogenic genes. • Large, randomized, placebo-controlled phase II and III trials will be required to ascertain the value of therapeutic angiogenesis for ischaemic cardiovascular disease.

Future directions There are two major areas of challenge if gene therapy is to become a routine component of haematological medicine. Firstly, scientists need to strive yet further to produce efficient nonimmunogenic vectors capable of carrying large pieces of genetic material into primitive HSC. Ideally these vectors should be sufficiently safe for direct injection into the bloodstream or organs.

402

This will depend critically on a more thorough understanding of the mechanisms of gene integration and control. Secondly, regulators have the unenviable task of deciding what constitutes acceptable risk in the USA for HIV patients.

Further reading Brenner S, Malech HL. Current developments in the design of onco-retrovirus and lentivirus vector systems for hematopoietic cell gene therapy. Biochim Biophys Acta 2003; 1640: 1–24. Ciceri F, Bordignon C. Suicide-gene-transduced donor Tcells for controlled graft-versus-host disease and graftversus-tumor. Int J Hematol 2002; 76: 305–9. Hacein-Bey-Abina S, Le Deist F, Carlier F et al. Sustained correction of X-linked severe combined immunodeficiency by ex vivo gene therapy. N Engl J Med 2002; 346: 1185–93. Hacein-Bey-Abina S, Von Kalle C, Schmidt M et al. LMO2associated clonal T cell proliferation in two patients after gene therapy for SCID-X1. Science 2003; 302: 415–19. Howe S, Thrasher AJ. Gene therapy for inherited immunodeficiencies. Curr Hematol Rep 2003; 2: 328–34. Humrich J, Jenne L. Viral vectors for dendritic cell-based immunotherapy. Curr Top Microbiol Immunol 2003; 276: 241–59. Long MB, Jones JP III, Sullenger BA, Byun J. Ribozymemediated revision of RNA and DNA. J Clin Invest 2003; 112: 312–18. Persons DA, Nienhuis AW. Gene therapy for the haemoglobin disorders. Curr Hematol Rep 2003; 2: 348–55. Richter J, Karlsson S. Clinical gene therapy in hematology: past and future. Int J Hematol 2001; 73: 162–9. St George JA. Gene therapy progress and prospects: adenoviral vectors. Gene Ther 2003; 10: 1135–41. Thomas CE, Ehrhardt A, Kay MA. Progress and problems with the use of viral vectors for gene therapy. Nat Rev Genet 2003; 4: 346–58. VandenDriessche T, Collen D, Chuah MK. Gene therapy for the haemophilias. J Thromb Haemost 2003; 1: 1550–8. Wall NR, Shi Y. Small RNA: can RNA interference be exploited for therapy? Lancet 2003; 362: 1401–3.

Chapter 35

Recombinant antibodies and other proteins Marion Scott

Introduction Many potentially useful human proteins for therapeutic, diagnostic and research use are expressed in the body at very low concentrations, and it is difficult, if not impossible, to isolate them by conventional biochemical methods. Other proteins, such as antibodies of a particular specificity, are difficult to purify from a complex mixture of very similar proteins. However, once the gene encoding a protein has been cloned and sequenced, it becomes possible to express the protein at high concentrations, using virally derived expression vectors designed to produce full-length proteins at high levels in various different in vitro culture ‘host’ cell systems. Some blood proteins, such as the clotting factors to treat haemophilia, have been efficiently purified by fractionation of pooled human plasma, but have been shown to have the potential of transmitting diseases, such as human immunodeficiency virus (HIV) and hepatitis C virus (HCV). The cloning and expression of these proteins has led to the availability of recombinant clotting factors for the treatment of haemophilia, with reduced risk of infection. As the recombinant clotting factors are grown in vitro, there is also the advantage of an unlimited supply of constant guaranteed product. Similar drivers have led researchers to try to develop recombinant replacements for specific immunoglobulins currently fractionated from high-titre blood donations, such as anti-D and anti-varicella-zoster virus (VZV). Some concern has been expressed about the safety of such recombinant products, as they may potentially contain viruses or other infectious agents arising from the host cells used to express the protein, or the culture medium components

used to grow the host cells. Increasing awareness of the risks from pooled polyclonal blood products have been heightened by the variant Creutzfeldt– Jakob disease (vCJD) crisis in the UK. Are the potential risks from such biotechnology products any worse than the risks from blood products derived from pooled human plasma? Apart from cloning and expressing such naturally occurring proteins, it is possible using recombinant DNA technology to produce modified forms of the proteins that do not occur naturally, that might have desired therapeutic effects or diagnostic advantages. Indeed, the technology has been used to produce totally novel recombinant proteins for particular applications.

General methods for recombinant protein expression Choice of the host cell system to use for recombinant protein expression relies on several factors. Bacterial expression systems, such as Escherichia coli are the cheapest, simplest and most effective, but cannot be used for many types of human proteins that require eukaryotic posttranslational modifications for biological activity, e.g. glycosylation. Most of the enzymes used in recombinant DNA technology, e.g. many restriction enzymes, DNA polymerases, DNA ligases, polynucleotide kinase and reverse transcriptases, are now produced as recombinant proteins themselves. These enzymes are readily produced in E. coli. The ready availability of these recombinant proteins has paved the way for the routine use of recombinant DNA technology in every biological discipline. In addition, 403

Chapter 35 Table 35.1 Production systems for recombinant mammalian proteins.

System

Cost

Production time scale

Scale-up capacity

Product quality

Glycosylation

Contamination risks

Bacteria Yeast Plants

Low Medium Low

Short Medium Long

High High High

Low Medium High

None Incorrect Some differences

Insect cells Mammalian cells Transgenic animals

Medium High High

Medium Long Very long

High Low Low

Medium Very high Very high

Incorrect Correct Correct

Endotoxins Low risk Low risk, but environmental concerns Low risk Animal viruses Animal viruses

some simple eukaryotic proteins, such as protein hormones that are medically important, e.g. insulin, growth hormone, granulocyte colonystimulating factor (G-CSF), can be produced in bacteria and in sufficient quantities for use as therapeutic agents. Toxicity to the host organism can be a problem if a gene is overexpressed at too high a level, so it is important to control levels of expression. Strong promotors, which can be easily regulated, are used to overcome the problems of toxicity and yet still maintain high levels of expression and protein production. Two promoter systems are in common use for expressing proteins in E. coli, the lac complex and the T7 late promoter system. Prokaryotes lack the enzymes that catalyse many of the post-translational modifications found on eukaryotic proteins. Proteins produced in prokaryotes may not be folded properly and/or can be insoluble, forming inclusion bodies. It can be very difficult to resolubilize proteins found in inclusion bodies and restore biological activity. Research is currently underway to produce genetically modified strains of yeast (Pichia pastoris) that have human glycosylation pathways. Insect cells have also been used for recombinant protein expression, using baculoviral vectors. Again, there are issues about the correct folding and glycosylation of mammalian proteins in these systems, and attempts have been made to produce insect cells that have been transfected with mammalian glycosylation enzymes. Other workers have expressed recombinant proteins in plants and plant cell cultures, but this again 404

raises issues about correct folding and glycosylation of proteins, and introduces further concerns about environmental containment of genetically modified crops. Transgenic animals have also been produced, with targeted production of recombinant proteins in milk. A comparison of different production systems for recombinant proteins is shown in Table 35.1. For many types of human proteins, expression in a mammalian system is the best option, as this is the approach most likely to yield soluble, biologically active proteins, although it is considerably more expensive than expression in E. coli, yeast or insect cells. Cell lines commonly used are NS0 (mouse myeloma), CHO (chinese hamster ovary) and COS-7 (African green monkey fibroblast). A number of techniques have been developed for rapid one-stage purification of recombinant proteins. Epitope tags are short amino acid sequences for which commercial monoclonal antibodies are available, and can be placed anywhere within the protein where it will not disrupt the protein’s function. It is also common to create fusion proteins, i.e. to create a single open reading frame that encodes a well-characterized protein such as glutathione S-transferase (GST) together with the sequence of the protein of interest. The most popular tag systems for the purification of recombinant proteins in mammalian expression systems are GST, His6-tag, myc and FLAG. When the tag protein is produced, the protein of interest is produced as well, as one fusion or chimeric protein. Fusion proteins are useful because they enable rapid purification by affinity chromatography, and

Recombinant antibodies Restriction site Promoter

Tag

2 Restriction endonuclease, ligate with insert

Gene

Antibiotic resistance gene 1 Expression vector 3 Express fusion protein Fig. 35.1 Production of recombinant

fusion proteins.

the fused tag can be removed after purification using a specific protease. Some workers have used IgG Fc domain tags, and used protein-A and protein-G affinity matrices. Plasmids (expression vectors) used for expression commonly contain a viral promoter sequence, an antibiotic resistance gene, a fusion tag sequence and a restriction endonuclease site for insertion of the coding sequence of interest (Fig. 35.1). cDNA coding for the protein sequence of interest is normally derived by reverse trancriptase polymerase chain reaction (RT-PCR) from cells expressing the protein, using sequence-specific primers to amplify the region required. This cDNA is then inserted into the expression vector, and used to transfect a mammalian cell line. Growth in medium containing the antibiotic to which the vector codes resistance results in selection of transfected cells only. Production of the fusion protein can then be detected using antibodies to the fusion tag sequence, and the fusion protein purified and characterized. Some expression vectors do not insert into the host cell nuclear material and give rise to transient expression. Other vectors insert into the host cell DNA and give rise to stable expression. Integration into the host genome is random, such that it is worth screening different clones of infected cells, as some may show higher levels of expression than others. Different clones also vary in their stability of secretion of the protein, such

protein

Tag

that several high-producing clones should be kept in culture over a period of several months and secretion monitored. Once a high-producing stable clone has been isolated, master and working cell banks of the clone should be frozen in liquid nitrogen, and each batch of the recombinant protein grown from one vial of the working cell bank. This process ensures that every batch of the recombinant protein produced is identical.

Recombinant antibodies Limitations of rodent monoclonal antibodies

Conventional monoclonal antibody technology uses immunization of mice or rats with antigen to yield hyperimmunized spleen cells, which are then fused with non-secreting myeloma cell lines to yield hybridoma cell lines that can be grown in vitro to produce monoclonal antibodies. Effectively, the fusion process inserts the DNA from the spleen cells into the myeloma cells. While many such conventional monoclonal antibodies were very successfully developed into diagnostic reagents (such as the high-avidity anti-A and antiB now used routinely worldwide for blood grouping), it was not possible to produce antibodies of certain specificities in rodents, and attempts to use rodent monoclonal antibodies in humans as therapeutics rapidly ran into problems, as the recipients developed a strong human anti-rodent response, 405

Chapter 35 Table 35.2 Success rates for

1980–82 1983–85 1986–88 1989–91 1992–94 1995–97 1998–2000 All Murine Chimeric Humanized

Total monoclonal antibodies

Discontinued

Approved

Completed (%)

Success (%)

2 9 33 34 41 33 34 186 49 23 59

1 8 29 29 23 12 2 104 34 13 15

1 0 2 2 5 0 0 10 1 4 5

100 89 94 91 68 36 6 61 71 74 34

50 0 6 6 18 0 0 9 3 24 25

monoclonal antibodies entering clinical trials.

which rapidly cleared the antibodies from the body.

pipeline (see Tables 35.2 and 35.3 and Figs 35.2 and 35.3).

Humanizing rodent monoclonals

Human recombinant antibodies

The early promise of monoclonal antibodies as therapeutics was not realized, and many became disillusioned with the concept of the ‘magic bullet’. The success rate of rodent monoclonal antibodies that entered clinical trials was only 9% over the 20 years from 1980 to 2000, and only 10 were approved as products out of 186 entering trials (Table 35.2). However, when the possibility of making recombinant antibodies became available in 1986, things rapidly changed. Using recombinant DNA technology, it was possible to replace the mouse constant domains of antibodies with corresponding human domains, and express these chimeric recombinant immunoglobulin molecules in myeloma cell lines. There was far less human anti-mouse response to such antibodies, such that 4 of 23 chimeric antibodies entering clinical trials have now been approved as products (Tables 35.2 and 35.3; Figs 35.2 and 35.3). Further engineering work also allowed the replacement of the framework regions of the mouse variable domains with human framework regions, resulting in virtually fully humanized antibodies. Such antibodies are proving to be very successful, with five already approved as products and many others in the

Human circulating B cells can be selected from immune individuals and transformed into cell lines that can be grown in culture by transformation with Epstein–Barr virus (EBV). The cDNA coding for the antibodies can then be derived from these cells using RT-PCR, ligated into expression vectors and expressed in a suitable mammalian host cell line. Alternatively, phage display technology can be used. Bacteriophage that infect E. coli are modified such that they carry the cDNA encoding for antibody variable domains, while at the same time they express the antibody protein on their surface. This permits in vitro selection of antibodies of the required specificity, and then expansion in E. coli. RT-PCR is used to amplify all the heavy and light chain variable domains in a buffy coat sample. PCR is then used to assemble these randomly into VH and VL pairs, by inclusion of DNA encoding for a flexible linker chain between the heavy and light chain domains. A ‘tag’ sequence is also included, normally c-myc, to aid detection and purification. These linked heavy and light chain domains are known as single-chain Fv (scFv) (Figs 35.4 and 35.5). The scFv constructs are then

406

Recombinant antibodies Table 35.3 Therapeutic monoclonal

antibodies approved by the Food and Drug Administration, USA.

Name

Specificity

Trade name

Company

Type

Approval

Muronomab Abciximab Rituximab Daclizumab Basiliximab Palivizumab Infliximab Trastuzumab Gemtuzumab Alemtuzumab

CD3 GPIIb/IIIa CD20 CD25 CD25 RSV TNF-a HER-2 CD33 CD52

Orthoclone ReoPro Rituxan Zenapax Simulect Synagis Remicade Hercetin Mylotarg Campath

Ortho Centocor Genentech Hoffman La Roche Novartis MedImmune Centocor Genentech Wyeth-Ayerst Milennium/ILEX

Murine Chimeric Chimeric Humanized Chimeric Humanized Chimeric Humanized Humanized Humanized

1986 1994 1997 1997 1998 1998 1998 1998 2000 2001

RSV, respiratory syncytial virus;TNF, tumour necrosis factor.

12 10 8

Murine Chimeric Humanized

6 4 2

00 20

98 19

96 19

94 19

90

88

86

92 19

19

19

19

84 19

82 19

19

Fig. 35.2 Number of monoclonal

80

0

antibodies entering clinical trials.

ligated into a phage display vector. The scFv domain is ligated into the vector next to regions that code for the PIII phage coat protein. The recombinant phage then express the scFv protein alongside their PIII coat protein at the tip of the phage (Fig. 35.6). Phage libraries can be panned against antigens, and those phage selected that are displaying scFv bind to the antigen. Selected phage are eluted from the antigen, expanded by culture in E. coli, and then repanned against antigen. Selected human scFv can then be removed from the phage vector and can be ligated to cloned human IgG constant domains to express full-length human recombinant antibody molecules. One large advantage of this approach is that antibodies can be derived from phage display libraries made

from non-immunized individuals, and that normally restricted antibodies (e.g. anti-self) can be derived. Human recombinant monoclonal anti-D

Despite the overall success in the production of rodent monoclonal antibodies to human ABO blood group antigens, no such monoclonal antibodies have been produced to the Rh antigens. Analyses of immunoglobulin gene usage in human monoclonal anti-D have shown that it is very restricted: it is possible that the rodent strains used lacked the appropriate immunoglobulin genes. It may also be that Rh antigen processing and presentation in rodents is different to that in humans. 407

Chapter 35

PCR RNA

PCR assembly V H VH VH VL VH

VL

VH

VL

VH

VL

VL 108 B lymphocytes

VL V-gene repertoires

VL VL

Clone108 clones scFv phage library

scFv-Gene repertoires

Fig. 35.4 Generation of scFv phage libraries. PCR,

polymerase chain reaction Mouse

Chimeric

Human

Humanized

Fig. 35.3 Chimeric/humanized antibodies.

It is also possible that the D antigen is very similar to a rodent antigen, such that rodents do not respond because the lymphocytes recognizing the common structure have been deleted from their repertoire. Various different approaches have been developed to produce human monoclonal antibodies specific for RhD. Early work used the immortalization of human B cells by infection with EBV. However, although it is relatively easy to establish lymphoblastoid cell lines producing specific antibody in this way, antibody production is commonly lost on expansion of the culture. Similarly, specific antibody production 408

is often lost during cloning. Repeated selection for cells producing antibody using antigen (by rosetting with RhD-positive red cells) is required to produce stable cell lines. Improvements in the stability of human cell lines have been achieved by back-crossing human anti-D-secreting EBV lines to a mouse–human heterohybridoma line or to a mouse myeloma line. Use of these approaches has enabled the production of a large number of blood group-specific human anti-Rh monoclonal antibodies. More recently, various molecular biology techniques have been used to produce recombinant human antibodies with Rh specificity. Single-chain human Fv specific for Rh antigens have been produced by panning phage display libraries of scFv derived from non-immune donors. Fab fragments with RhD specificity have been produced by panning Fab libraries from a hyperimmune donor, and the Fab fragments have been converted to fulllength immunoglobulin molecules by cloning the variable regions into expression vectors containing genomic DNA encoding the immunoglobulin constant regions. The resultant IgG1 constructs have been successfully expressed in CHO cells. DNA coding for anti-D in lymphoblastoid cell lines and heterohybridomas has been isolated, modified and expressed in rodent myeloma cell lines. Another group has expressed anti-D DNA in a baculovirus–insect cell expression system. Candidate monoclonal anti-Ds for immunoprophylaxis are selected, firstly, on their ability to bind to the RhD antigen via the Fv part of the molecule

Recombinant antibodies myc tag

link VH

VL

30 kDa

VL

Ag

NH2

Fig. 35.5 Structure of scFv.

myc

p lll coat protein

VH

VL

p vlll coat protein DNA

Fig. 35.6 scFv displayed on phage surface.

and, secondly, on their ability to interact with Fc receptors via the Fc part of the molecule to bring about immunomodulation. The exact mechanism of immunosuppression by anti-D is not known, but it is clear that it involves interaction of anti-D with Fc receptors. To be effective, prophylactic antibody must be capable of not only binding to

COOH

VH

the RhD antigen on the red cells via its Fv regions, but also interacting with the effector cells of the immune system via its Fc region. Selection of recombinant monoclonal anti-D for therapeutic use therefore depends not only on the antigen specificity and avidity of the monoclonal antibody but also its functional activity in interacting with effector cells. The exact mechanism of immunosuppression by anti-D is not known, but it is clear that it involves interaction of anti-D with Fc receptors. To suppress immunization, IgG-coated red blood cells need to be rapidly cleared from the maternal circulation and localized in the spleen. It has been suggested that D antigen-specific B cells in the spleen are then deactivated by the simultaneous binding of the Fc region of the anti-D to FcgRIIb together with binding of the B-cell receptor to the D antigen. Interactions of anti-D with FcgRI, FcgRIIb and FcgRIIIa may thus all be required for effective immunosuppression. IgG monoclonal anti-D antibodies have been evaluated in various in vitro systems to test how effective the antibodies are at interacting with immune system effector cells. Each assay tests efficacy at binding to different Fc receptors. Rosette formation of sensitized cells with monocytes and phagocytes, adherence of sensitized cells to monocyte monolayers, and chemiluminescent measurements of the oxidative burst caused when monocytes react with sensitized red cells are all in vitro measures of interaction with FcgRI. Measure409

Chapter 35

ments of antibody-dependent cellular cytotoxicity (ADCC) by radiolabelled chromium release from natural killer cells measures interaction with FcgRIIIa. It is not clear at present how well performance in these various in vitro assays will predict in vivo efficacy. Some of the functional activities may be dependent on glycosylation of the antibodies, and therefore dependent upon the host cell line used for expression of recombinant antibodies. For example, expression of an anti-D in NSO cells produced a recombinant antibody with significantly lower ADCC activity than the same antibody expressed in CHO cells. Compliance with regulatory requirements

Regulatory bodies in the USA and Europe have guidelines for the manufacture and production of recombinant antibodies for therapeutic use. These have been harmonized under the auspices of the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH). Most of the guidelines are aimed at ensuring the microbiological safety of biotechnology products derived from cell lines of human or animal origin. The risk of viral contamination is a feature common to all biotechnology products produced from cell lines. Such contamination could have serious clinical consequences and can arise from the contamination of the cell sources themselves or from adventitious introduction of virus during production. To date, however, biotechnology products derived from cell lines have not been implicated in the transmission of viruses. Nevertheless, it is expected that the safety of these products with regard to viral contamination can be reasonably assured only by the application of a virus-testing programme and assessment of virus removal and inactivation achieved by the manufacturing process. Three principal, complementary approaches have evolved to control potential viral contamination of biotechnology products. 1 Selecting and testing cell lines and other raw materials, including media components, for the absence of undesirable viruses which may be infectious and/or pathogenic for humans.

410

2 Assessing the capacity of the production processes to clear infectious viruses. 3 Testing the product at appropriate steps for the absence of contaminating infectious viruses. The current regulations state that ‘where a known human pathogen is identified, the product may be acceptable only under exceptional circumstances’. This statement may mean that production of monoclonal anti-D directly from EBV-transformed human cells may not be acceptable to the regulatory authorities. However, technology is now readily available to express antibody DNA from EBV-transformed cells in other cell types, and the guidelines further state that Cell lines such as Chinese hamster ovary (CHO), C127, baby hamster kidney (BHK) and murine myeloma cell lines have frequently been used as substrates for drug production with no reported safety problems related to viral contamination of the products. For these cell lines in which the endogenous particles have been extensively characterized and clearance has been demonstrated, it is not usually necessary to assay for the presence of the non-infectious particles in purified bulk product. From the guidelines it would appear that expression of anti-D monoclonal antibodies in CHO cells would be the easiest and safest method of production. Clinical trials

In the first clinical trial of monoclonal anti-D, the IgG1 antibody used (UCHD4) did not promote phagocytosis by purified monocytes and did not induce accelerated clearance of Dpositive cells. In a later trial, two monoclonal anti-D antibodies, BRAD-3 and BRAD-5, were selected because of their high activity in in vitro functional assays, high avidity and specificity for the immunodominant epitope region of the RhD antigen. Initial studies in D-negative male volunteers showed expected half-lives and pharmacokinetics after injection. Further studies on the antibodies administered with 51 Cr-labelled D-positive red cells demonstrated

Recombinant antibodies

accelerated red cell clearance in all subjects and provided preliminary evidence for protection from immunization. It is clear from these clinical trials to date that recombinant anti-D has the potential to replace polyclonal prophylactic anti-D. There is a case for universal antenatal prophylaxis if sufficient supplies of anti-D are available. How quickly recombinant anti-D becomes available will be largely determined by commercial investment and regulatory procedures. There have been concerns about the safety of biotechnology products in general, and trying to develop a biotechnology product for routine administration to healthy, young, pregnant women has been seen as a high commercial risk, that some are not prepared to take. Cost of litigation, if something went wrong, would be massive. As recombinant antibody products become more established, with a good safety record, such concerns will hopefully diminish.

model. We are currently engineering human chimeric versions of these mouse antibodies to progress this work into clinical trials for the potential treatment of vCJD. Anti-HCV recombinant antibodies

Two mouse monoclonal antibodies that beween them cover virtually all strains of HCV have been produced by conventional monoclonal technology, by immunizing mice with synthetic peptides corresponding to known neutralization sites on the virus. The antibodies neutralize the virus in vitro. Human chimeric versions of these antibodies are now being constructed before proceeding to clinical trials. Anti-HPA-1a recombinant antibodies

In a similar fashion to anti-D, recombinant antiVZV antibodies have been produced by DNA technology. Human cell lines were derived by EBV transformation of selected B cells from immunized individuals. cDNA coding for the antibodies has been expressed in CHO cells. Antibodies have been shown to neutralize VZV in vitro, and studies are currently underway in a guinea-pig model of the disease prior to clinical trials.

scFv specific for the human platelet antigen HPA1a have been derived from a phage display library prepared from an individual with antibodies to HPA-1a. The scFv has been ligated to scFv specific for the RhD antigen on red blood cells, and the novel bispecific recombinant antibody can be used in a mixed passive haemagglutination test for the HPA-1a antigen on platelets. The scFv has also been expressed as a full-length human IgG antibody by ligation to the constant domains of human IgG1, and this antibody has been used either fluorescently labelled or enzyme labelled in other diagnostic tests for the HPA-1a antigen on platelets.

Anti-prion recombinant antibodies

‘Null’ recombinant antibodies

A range of monoclonal antibodies to human recombinant prion proteins has been produced by immunizing prion knockout mice. Selected antibodies have been developed into a diagnostic test for bovine spongiform encephalopathy using homogenized bovine brain post-mortem. Work is currently underway to try to increase the sensitivity of the assay to make it suitable for screening human blood for vCJD. It has been shown that these mouse monoclonal antibodies can prevent the spread of vCJD prion disease in a mouse

Using site-directed mutagenesis, the Fc domains of human IgG antibodies have been mutated to have as little biological function as possible. Recombinant anti-D and anti-HPA-1a antibodies have been produced with this ‘null’ Fc region. In vitro studies have shown that these ‘null’ antibodies can effectively compete with clinically significant antibodies, and prevent them causing immune destruction of red cells and platelets respectively. A clinical trial in male volunteers showed that the ‘null’ anti-D protected D-positive red cells

Human recombinant anti-varicella-zoster immunoglobulin

411

Chapter 35

from clearance by anti-D in vivo. The aim is to see if the ‘null’ anti-HPA-1a antibody can be administered to HPA-1a-negative pregnant women who are carrying HPA-1a-positive fetuses and prevent neonatal alloimmune thrombocytopenia by crossing the placenta and competing with maternal anti-HPA-1a that can cause destruction of fetal platelets. Secretory IgG

Intravenous immunoglobulin is a pooled blood production used to treat patients with primary immune deficiency throughout their lives. Whereas the product does protect from fatal infection, many patients still suffer from recurrent infections of the mucosal surfaces, presumably because intravenous immunoglobulin does not provide primary immune defence at these surfaces. Experiments have shown that antibodies to the polymeric immunoglobulin receptor (normally responsible for the transport of polymeric IgA and IgM to the mucosal surfaces) can trigger transport across epithelial cells. scFv specific to the receptor have been produced by panning phage display libraries and also shown to trigger transport. Bispecific Fv constructs are being produced which have specificity for the receptor and the Fc domain of IgG. In vitro studies have shown that such a bispecific recombinant antibody molecule can promote transport of IgG across epithelial cells in culture. A mouse model is being developed to study the efficacy of this approach in vivo. Recombinant phenotyping reagents

Many monoclonal human IgG antibodies have been produced to blood group antigens, but these require use in enzyme or antiglobulin techniques that are not suited to high-throughput automated blood grouping machines. By cloning the variable regions of these antibodies, it is possible to ligate them to the constant domains of human IgM antibodies, and express hybrid recombinant molecules in myeloma cells. These antibodies are highly potent because they are hexameric rather than pentameric structures, and they combine the high affinity of the affinity-matured IgG antibodies with 412

the polymeric structure of primitive hexameric IgM. They are very potent direct agglutinins that can readily be used in automated blood grouping machines.

Recombinant antigens Detection and identification of clinically significant blood group, platelet and granulocyte antibodies currently relies on the availability of high-quality antibody screening and identification cells that cover all clinically significant antigens, and carry them in combinations such that the specificity of antibodies can be deduced. The quality of these panels of cells is critical, and their variability has been shown in UK National External Quality Assessment Scheme exercises to be the main cause of error in the detection and identification of antibodies. Quantitation of antibodies during pregnancy is carried out using titration or autoanalyser technology, both of which show high levels of variation, such that it is difficult to set levels at which clinical action is required. Some studies have used enzyme-linked immunosorbent assay techniques to try to also determine the subclass of the antibody, in case that might aid clinical judgement, but this is not widely used. Most of the relevant antigens have been sequenced and cloned. Some have been inserted in expression vectors and expressed in the membranes of in vitro cultured cells, e.g. expression of the Rh protein in K562 human erythroleukaemia cells. For some antigens it is possible to amplify just the extracellular domain of the protein that carries the antigen, ligate this to a fusion partner protein or tag (such as FLAG), and express a soluble recombinant protein that carries antigenic activity. This has been demonstrated for the Kell, Lutheran, Duffy, MNSs and Cartwright red cell antigens and HPA-1a and HPA-1b platelet antigens. More work is underway to produce further antigens. The target of this work is to be able to produce microarrays of recombinant antigens which could then be used for high-throughput antibody screening, identification, subclass determination and quantitation of antibodies in transfusion recipients

Recombinant antibodies

and pregnant women. Such recombinant antigen microarrays have already been developed and used to study autoantibodies.

Recombinant cytokines Erythropoietin (EPO) and thrombopoietin (TPO) have both been produced as recombinant proteins and licensed for therapeutic use for the treatment of anaemias and thrombocytopenias respectively. Improvements have been made to recombinant EPO by engineering different glycosylation of the recombinant protein. Darbepoetin-alfa has two amino acid substitutions engineered into the native molecule, which result in a hyperglycosylated molecule. It has a prolonged half-life in plasma and increased biological activity. Similarly, recombinant TPO is available as a molecule with identical amino acid sequence to naturally occurring TPO, or as a pegylated recombinant megakaryocyte growth and development factor, which is a nonglycosylated molecule consisting of the first 163 amino acids of TPO and is coupled to polyethylene glycol. Several studies have reported red cell aplasia due to autoantibodies made to the recombinant erythropoietins administered, and thrombocytopenia due to autoantibodies induced by administration of recombinant TPO. Such development of neutralizing antibodies to endogenous cytokines after administration of recombinant growth factors occurs rarely. Mechanisms for the immune responses are not currently understood, nor is it clear whether there is a higher incidence of antibodies to the modified recombinant growth factors compared with the ‘native’ recombinant factors. The availability of recombinant cytokines and growth factors has advanced research on understanding the processes of erythropoiesis and thrombopoiesis. It is now possible to isolate stem cells from cord blood and direct their differentiation in vitro to produce mature red cells or megakaryocytes. This introduces the possibility of engineering and growing ‘designer’ blood cells in vitro for multitransfused patients for whom it is difficult to resource compatible blood.

Recombinant clotting factors Recombinant clotting factors have been successfully used for the treatment of haemophilia for several years. Recombinant protein technology has virtually eliminated transmissible disease risk from these products, such that recombinant products are the products of choice for haemophiliacs. Other non-infectious complications, including inhibitor formation, remain a concern. There is no evidence to date that the recombinant clotting factors induce a higher level of inhibitor formation than fractionated plasma products.

Recombinant haemoglobin Recombinant protein technology is also being used to engineer haemoglobin variants to act as red cell stroma-free oxygen-carrying solutions. The only recombinant haemoglobin subjected to clinical trials is expressed in E. coli and is a modified human haemoglobin tetramer cross-linked with a glycine bridge between the a-subunits.

Conclusions Recombinant protein technology has rapidly advanced over the last 20 years, and we are now starting to see the routine use of recombinant proteins in transfusion medicine. Recombinant proteins will probably totally replace clotting factors and specific immunoglobulins that are currently produced from fractionated pooled plasma. However, it is unlikely that recombinant products will replace intravenous immunoglobulin or albumin. Intravenous immunoglobulin works because of its broad specificity: it would be very difficult/impossible to mimic this successfully with a recombinant product. However, a recombinant product may be able to improve the efficacy of intravenous immunoglobulin by targeting part of it to the secretions to provide primary immune defence. Albumin could be produced as a recombinant protein, but this is unlikely to be economically viable compared with the ease of production 413

Chapter 35

from plasma. Only evidence of disease transmission by plasma-derived albumin could drive the production of recombinant albumin. Further specific recombinant immunoglobulins are being produced that are not currently available as blood products (anti-HCV and anti-vCJD) and the efficacy of these needs to be investigated in clinical trials. Haemopoietic growth factors have also been produced by recombinant technology and used to treat anaemia and thrombocytopenia. Blood group antigens are now available as recombinant molecules, such that we may no longer need to use red cells, platelets and granulocytes for antibody screening, identification and quantitation. Various recombinant haemoglobin molecules have been produced and these are currently undergoing clinical trials. They may be useful as blood replacements in particular surgical situations but are unlikely to replace most applications of blood transfusion. The most likely application of recombinant protein technology that may one day replace donated blood for transfusion is the use of recombinant growth factors to grow ‘designer’ blood cells from stem cells in vitro.

Summary Any protein that has a known DNA sequence can be expressed as a recombinant protein. cDNA is inserted into a virally derived expression vector and transfected into host cells. Recombinant proteins can be expressed in bacteria, yeasts, plants, insect cells, mammalian cells or in transgenic animals. Novel proteins that do not occur in nature can be engineered. Recombinant antibodies

414

have been produced and may replace specific immunoglobulins currently derived from fractionated plasma. Novel bispecific and null antibody molecules have the potential to form novel therapeutics. Recombinant antigens, growth factors, clotting factors and haemoglobins have also been produced.

Further reading Bretthauer RK. Genetic engineering of Pichia pastoris to humanize N-glycosylation of proteins. Trends Biotechnol 2003; 21: 459–62. Hesse F, Wagner R. Developments and improvements in the manufacturing of human therapeutics with mammalian cell cultures. Trends Biotechnol 2000; 18: 173–80. Kato T, Miyasaki H. Therapeutically induced autoantibodies in patients treated with recombinant hematopoietic growth factors: a brief summary. Curr Pharm Des 2003; 9: 1129–32. Ma JK-C, Drake PMW, Christou P. The production of recombinant pharmaceutical proteins in plants. Nat Rev Genet 2003; 4: 794–805. Macdougall IC. Erythropoietin and kidney failure. Curr Hematol Rep 2003; 2: 459–64. Robinson WH, DiGennaro C, Hueber W et al. Autoantigen microarrays for multiplex characterisation of autoantibody responses. Nat Med 2002; 8: 295–301. Schlesinger KW, Ragni MV. Safety of the new generation recombinant factor concentrates. Expert Opin Drug Saf 2002; 1: 213–23. Scott ML. Monoclonal anti-D for immunoprophylaxis. Vox Sang 2001; 81: 213–18. Watkins NA, Ouwehand WH. Introduction to antibody engineering and phage display. Vox Sang 2000; 78: 72–9. Winslow RM. Alternative oxygen therapeutics: products, status of clinical trials and future prospects. Curr Hematol Rep 2003; 2: 503–10.

Chapter 36

Blood transfusion in a global context David Roberts, Jean-Pierre Allain, Alan Kitchen, Stephen Field and Imelda Bates

Introduction 17% of the world’s population has access to 60% of the global blood supply Inequality in the provision of ‘safe blood’ around the world mirrors the unequal distribution of almost all other resources crucial for effective health services or indeed for health itself. Unfortunately, in many countries, providing safe blood is made more difficult by lack of donors and the high frequency of transfusion-transmissible infections. At the same time, the problems posed by the poor supply of safe blood are compounded by the frequent need for urgent life-saving transfusions in childbirth and in children with malaria. The purpose of this chapter is not to guide those developing transfusion services in less affluent countries but to inform a wider audience of the problems faced in the development of effective transfusion services in these countries. A secondary aim is to stimulate some debate and analysis of the problems faced by transfusion services globally. Finally, a short chapter must be selective and our choice of topics and examples and their solutions reflects our own experience in SouthEast Asia and tropical Africa south of the Sahara (or sub-Saharan Africa). Blood safety

An unsafe blood supply is costly in both human and economic terms. Transfusion of infected blood not only causes direct morbidity and mortality in the recipients, but also has an economic and emotional impact on their families and communities and undermines confidence in modern healthcare.

Those who become infected through blood transfusion are infectious to others and contribute a significant secondary wave of iatrogenic infections. Investment in safe supplies of blood is therefore a cost-effective investment for every country, even those with few resources. The World Health Organization (WHO) has identified four key objectives of all strategies used to ensure that blood is safe for transfusion. • Establish a coordinated national blood transfusion service that can provide adequate and timely supplies of safe blood for all patients in need. • Collect blood only from voluntary nonremunerated blood donors from low-risk populations and use stringent donor selection procedures. • Screen all blood for transfusion-transmissible infections and have standardized procedures in place for grouping and compatibility testing. • Reduce unnecessary transfusions through the appropriate clinical use of blood, including the use of intravenous replacement fluids and other simple alternatives to transfusion, wherever possible. WHO also emphasizes that effective quality assurance should be in place for all aspects of the transfusion process, from donor recruitment and selection, through infection screening, blood grouping and blood storage to administration to the patients and clinical monitoring for adverse reactions. It is axiomatic that transfusion medicine is a distinct and multidisciplinary sector of the health service and should be incorporated into all national health plans. It is therefore the responsibility of governments to develop policies and legislation that will facilitate the development of a national transfusion service and ensure that the 415

Chapter 36

blood transfusion process and its associated quality assurance programmes are of a high standard. WHO has simply and succinctly recommended the structure of national blood transfusion services. It suggests that at the national level the transfusion service should have a medical director, an advisory committee and clear national transfusion policies and strategies with the appropriate statutory instruments to ensure the national coordination and standardization of blood testing, processing and distribution. However, less than 70 out of the 191 member states meet WHO’s recommendations for a national blood programme. In Africa, in 2002 WHO estimated that among the 46 member states in the African continent, only 14 had a national blood policy and just six had a policy to specifically encourage and develop a system of voluntary non-remunerated donation. It is worthwhile reflecting on why the development of national transfusion services has been delayed. One reason may be that the emphasis on primary healthcare over the last 25 years has diffused interest in hospital-based curative medicine. A second reason may have been the high cost of blood transfusion in relation to disposable income and healthcare budgets. The average annual income in sub-Saharan Africa is in the range of $400–1000 and a unit of blood costing $10–20 is an expensive commodity in relation to the annual per-capita budget for healthcare in these countries. Nevertheless, blood transfusion for severe malarial anaemia and severe haemorrhage can be lifesaving and here the cost of transfusion is well within the generally accepted cost–benefit range for health interventions in poorer countries of $1 per year of life saved. However, it is also true that many of these countries cannot afford sustainable basic healthcare systems, let alone develop a national transfusion service, which supplies blood according to the standards established by affluent countries, where a unit of blood now costs in the region of $200. To prepare enough safe blood in a sustainable fashion, African countries need to develop their own ways to produce affordable safe blood. Uncritical adoption of external advice and models

416

may lead to unsustainable and inappropriate solutions. What then are the transfusion services in Africa and what are the consequences for the supply of safe blood? Organization of transfusion services in sub-Saharan Africa

African countries have developed a variety of systems to try to achieve a sustainable safe blood supply. These vary from large, modern, national blood centres to locally organized donor programmes for isolated district healthcare facilities. A minority of countries have invested significant resources in transfusion services, often with financial support and advisers from European governments, USAID or non-governmental organizations (NGOs), including Red Cross, Red Crescent, Family Health International and the Safe Blood for Africa Foundation. In these countries there has been a commitment to establishing centralized systems based on a model of sustainability and effectiveness, similar to that used in wealthy nations. These centres typically collect over 10 000 units a year, use automated equipment and produce some components. Blood donor recruitment and screening and processing of donated blood are carried out in specifically designed premises away from the hospitals where blood is transfused. The cost of a unit of blood produced by these centres is of the order of $40 per bag. However, as we have seen, many countries in sub-Saharan Africa do not operate a centralized transfusion service. Each hospital recruits donors for its own patients and processes the blood for transfusion itself. These hospitals often handle less than 1000 units a year and experience difficulties in standardization, quality assurance and in maintaining supplies of high-quality reagents. Recruiting voluntary donors from the community is complex and expensive and depends on regular education programs, venesection teams, vehicles and cold storage. Because of these difficulties, the majority of donors in poorer countries are ‘replacement’, not volunteer, donors, although many local initiatives exist to create small local panels of tested volunteers who may be called

Global blood transfusion

upon to donate at short notice. In some regional schemes, anti-human immunodeficiency virus (HIV), hepatitis B surface antigen (HBsAg) and anti-hepatitis C virus (HCV) are screened before donation with high-performance rapid tests so those blood bags are not wasted. Furthermore deferred donors can be identified, informed and counselled, consequently decreasing the prevalence of viral markers in repeat volunteer donors. In 2002, in Africa as a whole, WHO estimated that over 60% of blood originated from replacement/family donors. In sub-Saharan Africa the proportion of blood derived from replacement donors is certainly higher. These replacement donors should be family members, but are too often professional donors paid directly by the family, who are asked to provide blood for their relatives in the hospital. Cultural taboos and lack of education about donating blood (e.g. ‘men will become impotent if they donate blood’; ‘HIV can be caught from the blood bag needle’) makes relatives reluctant to donate so they may choose to purchase blood from ‘professional’ donors. It is worth noting that similar problems faced widespread acceptance of blood donation when Percy Oliver and Geoffrey Keynes began to establish the first blood banks of volunteer donors in London over 70 years ago. Nevertheless, the local systems allow many patients to survive serious illness and are often maintained by dedicated staff in difficult circumstances. However, even with the best input from local staff these district services experience problems of supply, safety and cost of blood. Supply of blood

Patients in poorer countries usually present late in the course of their disease and the delays and lack of stored blood inherent in the replacement donor system mean that patients may die before a blood transfusion can be organized. By the time a donor has been found, screened and venesected, and the blood transfused into the patient, several hours or even days can elapse. A survey of the blood supplied by a dedicated district service in East Africa showed that the average delay in sourcing blood

for children with severe malarial anaemia was 6 h. Anecdotal evidence suggests that in some areas, and occasionally in many areas, blood may not be available at all. Finally, locally based services make it difficult to separate blood, even into simple fractions such as red cells and plasma, in order to provide specific components if needed. Safety of blood

Local blood transfusion systems encounter many problems, almost always centred around lack of funding, including inconsistency and high prices of screening tests, breakdown of the cold chain mostly related to frequent power cuts, and poorly trained staff. Blood frequently has to be collected in small hospital-based units often with no dedicated staff or specifically allocated budget. The WHO review estimated in 2000 that 25% of the blood in sub-Saharan Africa was untested for antiHIV and that blood transfusion was the origin of 5–10% of new HIV infections. HBsAg was screened in 50% of donors or donations and only 19% were screened for anti-HCV. Most countries in sub-Saharan Africa do not screen for human Tcell leukaemia virus (HTLV) since the prevalence is low (<2%). Syphilis and malaria may cause less of a problem for blood safety than viral infections. Although the risk of acquiring syphilis from infected blood is low, most blood banks in sub-Saharan Africa do screen for Treponema pallidum. Fresh blood is potentially infectious for syphilis but storage at 4°C can inactivate the bacterium. Malaria can be transmitted by transfusion and has an incubation period of 7–50 days. In areas of low or no malaria transmission, screening for the parasite is important, as recipients are likely to have no immunity. In countries where malaria is highly endemic, the prevalence of malaria parasites in donor blood is often very high (up to 50%) and excluding donors with low-grade parasitaemia is impracticable. Here, preventive treatment of young children receiving transfusion with antimalarial drugs is considerably more cost-effective than blood screening.

417

Chapter 36

Cost of local blood supply

Maintaining a district blood supply may appear simple but without the economies of scale achieved in large-scale programmes is inefficient and may be a considerable drain on scarce local resources. When a transfusion service is provided by individual hospitals it places an enormous burden on laboratory resources. One survey showed that in a typical district hospital in Southern Africa, the overall cost of the transfusion service, including consumables, proportional amounts for capital equipment, staff time and overheads, was 36% of total laboratory costs and so each unit of whole blood cost the laboratory approximately $20 to collect and process. The cost of a national service would be the greater with the additional costs of local education programmes, dedicated venesection teams, vehicles and cold storage. In addition, a national service would have to solve the very real practical problems of maintaining regular distribution of sufficient quantities of blood. However, a national service would release the time of valuable skilled staff in the district hospital and also allow the crucial standardization of testing and processing essential for quality control. Clinical use of blood

In Europe the majority of transfusions are planned or accompany elective surgery. In contrast, most transfusions in sub-Saharan Africa are given for life-threatening emergencies, most often anaemia in children or pregnant women and haemorrhage following childbirth or trauma. It is rare for blood to be used for any other reason except in the largest of hospitals, usually in capital cities. Transfusions are administered to children predominantly for malaria-related anaemia and can undoubtedly reduce the mortality of children with severe anaemia (typically having haemoglobin below 4 or 5 g/dL and symptoms of decompensation). Even in areas of high HIV prevalence, young children generally have a relatively low risk of becoming naturally infected with HIV and potentially have a long life expectancy. Pregnant women are the second most common recipients of blood, particularly for haemorrhage emergencies. Signifi418

cant quantities of blood are also used in trauma, surgery and general medicine. There are neither systematic reviews nor international guidelines covering the use of blood in these specific contexts and few audits of blood use. The scope for improving clinical practice is unquantified but probably substantial.

Putting the WHO objectives into practice: improving the supply, safety and use of blood in sub-Saharan Africa Some countries have used external funds to establish an integrated national service but few have been able to make the transition to a sustainable national transfusion service in the absence of external funding. Moreover, in several countries, external funding for 10 years or more has failed to develop a functioning national transfusion service and these failures have led some funders to withdraw grants to national transfusion services. Therefore, in many areas, transfusion services have to be optimized within the existing general hospital budget. Whatever sums are available, the specific, often interconnected problems surrounding the supply, safety, cost and use of blood must be addressed. Improving the blood supply and the safety of the donor pool

Careful donor selection is crucial not only to improve the supply of blood but also to reduce transfusion-transmitted infection risk. The selection of volunteer donors from lower-risk populations is considered the most effective approach and considerable effort has been devoted to promoting voluntary repeat donations. However, in most parts of Africa, replacement donors are the main resource. They are typically young males in the high-risk bracket of HIV infection or other sexually transmitted viruses. As the availability of replacement donors is limited, the most effective way to improve the availability and safety of blood is to recruit volunteer donors primarily among secondary school students with median age ranging between 16 and 20 years. They are younger, have a

Global blood transfusion

greater proportion of females and are 5–10 times less likely to be infected with HIV than replacement donors. Several strategies have been devised to encourage repeat voluntary donors and so reduce the risk of virus carriage. In Zimbabwe, Pledge 25 Club, a programme using education and incentive to attracting school students to give blood 25 times, has been successful. Similar, less ambitious schemes, for example a ‘Club 5’, could also be effective. The WHO slogan of ‘Safe blood starts with me’ has also resulted in successful educational programmes around the world. These schemes can be complemented by strategies to recruit donors from faith-based organizations or by collaborating with radio stations to organize and promote blood donations. The success or otherwise of well-organized requests for blood donors remains to be tested in many countries where national calls for donors have not been made in the absence of centralized blood transfusion services. Other novel solutions are being studied. It may be feasible to use placental blood as an accessory source of blood to transfuse small children in malarious areas. The placenta containing this blood is normally discarded after delivery. However, the high haematocrit and easy availability may make it suitable for small-volume emergency transfusions if blood can be collected free of bacterial contamination. In the long term, artificial blood substitutes could provide a useful alternative to allogeneic blood transfusion during acute illnesses. However, they remain expensive and have only limited approval pending further randomized trials of their safety and efficacy (see Chapter 30). Improving screening for blood-transmitted infections

Test sensitivity is critical in the face of high prevalence rates for HIV, hepatitis B virus (HBV) and HCV (Table 36.1). The overall prevalence of HIV antibody in sub-Saharan Africa ranges between 0.5 and 16%. In donors, it tends to remain below 5% in West Africa, below 10% in East and Central Africa and above 10% in Southern Africa. Chronic HBV prevalence, indicated by the pres-

ence of circulating HBsAg, ranges between 5 and 25% of the population including blood donors. This high prevalence is due to (vertical) transmission at birth or (horizontal) infection in infancy and the virtual absence of national vaccination programmes. Infection after the age of 10 is uncommon. HBsAg is more prevalent in West Africa (10–25%) than in East or Central Africa (5–10%); the lowest prevalence is found in Southern Africa (5% or less). Antibody to HCV is not routinely screened for in many parts of Africa but the prevalence of this infection ranges between 0.5 and 3% and reaches 10–15% in Egypt. The prevalence may be high locally, suggesting the importance of specific factors such as various types of injections or past diagnostic or vaccination campaigns contributing to spread the infection (see Table 36.1). These high prevalence rates pose a very substantial danger and a major logistical and technical challenge to those trying to provide safe blood. Even with the best available testing procedures, an HIV prevalence rate of 10% would imply a residual risk of HIV transmission of 1 in 3000 through the failure to detect seronegative but latent HIV infection in the window period. The residual risk of HBV infection remains substantial because of donations containing undetected low levels of HBsAg or occult HBV DNA. This risk remains high for children below age 10 but the problem is at least in part mitigated by the very high prevalence of adult recipients carrying HBV markers (60–90%). Precise estimates of the residual risk of HIV, HCV and HBV transmission are 1 in 2600, 1 in 1500 and 1 in 330, respectively, when using enzyme immunoassay (EIA) screening. The actual present residual risk of viral transmission by transfusion may be much higher. This risk has been assessed for HIV. In studies conducted in Kenya, Zambia and the Democratic Republic of the Congo, the risk of HIV transmission by transfusion was estimated between 1 and 3%, related in part to prevalence but also to test performance, storage conditions and staff training. These figures reinforce the value of a stable donor pool, as the prevalence rate may be reduced by 90% in repeat donors, and as ever emphasize the 419

Chapter 36 Table 36.1 Prevalence of transfusion-transmissible agents in sub-Saharan African blood donors.

Prevalence (%) Country

Year collected

Anti-HIV

HBsAg

Anti-HCV

HTLV

Benin Botswana Cameroon Ghana Kenya Malawi Nigeria RDC*† Republic of South Africa* Tanzania Togo Uganda Zambia Zimbabwe

1998 2000 1994–98 1998–2002 1995–98 2000 1992–98 1998 2001 1998 1995–2000 2000 1991–95 1997

0.5–3 10 4.1–5.8 1.7–3.8 4.5–3.0 10.7 3.9–5.4 6.4 4.5 8.7

12 5 10–16 15 4.2–3.9 8.1 15–20 9.2 5 11

1.4–2.3 1 1.6 1.7–8.4 1.5–1.8 6.8 12.3 4.3 0.5 8–10.3 3.3

0.3–5.4

3.9–5.4 8–16 8.8

6.5 2.5–15.4

0.5

Syphilis

33.5

13.5

2.5 0.7

0 1.8

Malaria

19–41

12.7

0.1

* Donors of African origin. † République Démocratique du Congo.

need for careful consideration of the clinical use of blood. Many countries are unable to spend money on test kits, or simply do not have the money to do so. Sometimes, even when money is available, the test kits may change regularly where each purchase requires an individual tender. The techniques used for screening must be considered carefully to ensure effective screening of the particular donor population and the skills of the staff involved. Nucleic acid testing is highly effective but is not affordable and less sophisticated methods must be optimized or novel solutions sought. WHO has established a systematic evaluation of both EIA and rapid tests to guide developing countries in their choice of reagents. These evaluations include test costs. Many rapid tests for anti-HIV and HBsAg, fewer for anti-HCV, are available but sensitivity and specificity, ease of use and cost vary greatly. Some of these tests are performed in one single step, with results obtained in 10–20 min using whole blood, plasma or serum samples. The best assays have sensitivity similar to EIA for antiHIV, detect 1 ng/mL of HBsAg, and have more

420

than 95% sensitivity and more than 98% specificity for anti-HCV. New approaches adapted to local situations appear promising. In small blood banks, the expensive microtitre plate systems used after donation can be replaced by cheaper, more costeffective high-performance rapid tests performed before or after donation. Predonation testing provides the advantages of reducing material waste and easy on-site communication with deferred donors who otherwise could not be reached. However, there are fears that predonation testing may reduce the willingness of donors to come forward and there is no consensus regarding predonation testing at present. To increase the level of blood safety, governments, hospitals, WHO and aid agencies need to show flexibility and consider the benefits of multiple strategies adapted to local needs rather than directly importing rigid models designed for totally different populations, staff and resources. Some new technology to reduce the problem of transfusion-transmitted infections is on the horizon. Rapid immunochemical and nucleic acid

Global blood transfusion

dipsticks are being developed for blood-borne pathogens and may cut the cost of predonation and post-donation testing to one-tenth of present costs. Clearly, this and other inexpensive and effective testing technology as well as pathogen inactivation techniques directed towards the needs of developing countries should become a major target of external support. Reducing the cost of transfusion services

The challenge for Africa is that safe blood should be available for health services and individuals even when resources are extremely limited. The majority of a blood unit cost originates from imported goods such as the blood bags and grouping and screening assays. Staff costs are a relatively small proportion of the overall costs because salaries are low and because negligible resources are put into staff training, supervision and auditing mechanisms. According to published studies, a unit of blood may cost $10–40; even $10 is not affordable by most families in sub-Saharan Africa, if a user fee is applied. Because transfusion is such an expensive service, the costs have to be subsidized, usually by aid packages and through external agencies, or occasionally through governments. Resources for transfusion services are often therefore vulnerable to fickle political and non-sustainable fluctuations. Developing systems that rely more on local resources means that in the long term they may be more dynamic, productive and sustainable, although it is becoming clear that many long-term transfusion projects are still nowhere near to becoming sustainable even 10 years from instigation. There are several examples of effective costsavings in transfusion services from sub-Saharan Africa: • using cheap but effective rapid tests that do not require equipment; • diagnostic companies reducing prices for resource-poor countries; • avoiding additional costs to intermediaries and limiting blood bag waste by predonation screening where appropriate.

Confirmatory testing is an expensive process and WHO has advocated that confirmation of reactive samples with alternative screening assays rather than expensive, highly specific, confirmatory assays is adequate in regions where prevalence is high. This approach will contribute to cost reduction. It might also be debated that a test specificity in excess of 99.8% is sufficient to avoid the need for confirmation. Certainly, much more research is needed comparing the costeffectiveness of various strategies to supply safe blood to patients in poor countries. Improving the clinical use of blood: guidelines for transfusion practice

The use of simple guidelines can reduce unnecessary transfusions and many institutions in subSaharan Africa and Asia have developed guidelines to promote rational use of blood transfusions and blood products. The scope for improvement in clinical practice is great. For example, strict enforcement of a transfusion protocol in a Malawian hospital reduced the number of transfusions by 75% without any adverse effect on mortality. The principles underlying most transfusion guidelines are similar and combine a clinical assessment of whether the patient is developing complications of inadequate oxygenation with measurement of their haemoglobin (as a marker of intracellular oxygen concentration). In subSaharan countries, the recommended haemoglobin threshold for transfusions is often well below that which would be accepted in more wealthy countries. In the USA, anaesthetists suggest that transfusions are almost always indicated when the haemoglobin level is less than 6 g/dL, whereas in many African countries transfusions are recommended for children at haemoglobin levels less than 4 g/dL, providing there are no other clinical complications. Ensuring that the transfusion guidelines are implemented is extremely difficult for poorer countries without formal monitoring and auditing systems. This is particularly problematic if the quality of haemoglobin measurements is not assured. Studies from a number of countries have

421

Chapter 36

shown that if clinicians do not have confidence in haemoglobin results, they will rely entirely on clinical judgement to guide transfusion practice and this can lead to significant numbers of inappropriate transfusions. In a typical district hospital in Africa the cost of providing a unit of blood is approximately 40 times the cost of a qualityassured haemoglobin test. Investment in improving low-cost haemoglobin testing has the potential for significant cost savings downstream in the much more expensive transfusion process as well as reducing the risk of transfusion-related infections.

Conclusion: the future of blood transfusion in a global context Fulfilling the first WHO objective of establishing ‘a coordinated national blood transfusion service that can provide adequate and timely supplies of safe blood for all patients in need’ has proved to be very difficult in many countries even given substantial external funding. Nevertheless, some countries have made progress and have recently established national transfusion services. Certainly the future of transfusion in the poorer countries depends on developing such coordinated national transfusion services, with the resources to collect, test, process and supply safe blood and reducing unnecessary transfusions through the appropriate clinical use of blood and products. Increased blood supply depends on the recruitment of voluntary non-remunerated donors. The examples and discussions in this chapter have centred on Africa but the same considerations apply to many of the poorer countries in Asia and Latin America. Here, there are wide variations in resources available for healthcare not only between but also within countries. In all countries, increased blood supply depends on the recruitment of volunteer donors and this should become a priority for policy development and resource allocation. A reliable and expanded donor pool will not only provide a life-saving therapy but also improve safety. Resources must be made available by governments to ensure that the essential supplies are available, including blood bags, grouping reagents and test kits. Labo422

ratory and blood bank management systems also need to be improved to ensure effective testing and processing, and the maintenance of the cold chain. Hospitals and other health facilities could cooperate to directly purchase cheap high-quality tests adapted to their needs. There is currently a feeling of guarded optimism about the future of blood safety in developing countries. The recent increase in allocation of resources for the prevention of HIV across the world, including the investment by governments of wealthy countries and contributions from international and private agencies, have begun to recognize the importance of transfusion safety. Parallel to the price reduction for antiviral drugs, the cost of screening tests supplied to developing countries has also decreased. The high cost of anti-HCV will soon be reduced when the patent expires. Methods of pathogen inactivation applicable to whole blood are being developed and these could, in one step, reduce or eliminate the risks of viral, bacterial and parasitic infections. More effective and efficient methods for testing blood are to be welcomed. The real challenge will be to integrate improvements in the supply and safety of blood in sustainable coordinated national transfusion services.

Further reading Allain J-P, Candotti D, Soldan K et al. The risk of hepatitis B virus infection by transfusion in Kumasi, Ghana. Blood 2003; 101: 2419–25. Bates I, Mundy C, Pendame R et al. Use of clinical judgement to guide administration of blood transfusions in Malawi. Trans R Soc Trop Med Hyg 2001; 95: 510–12. Beal RW, Bontick M, Fransen L, eds. Safe Blood in Developing Countries. Brussels: EEC AIDS task force, 1992: 11–12. Consten EC, van der Meer JT, de Wolf F, Heij HA, Henny PC, van Lanschot JJ. Risk of iatrogenic human immunodeficiency virus infection through transfusion of blood tested by inappropriately stored or expired rapid antibody assays in a Zambian hospital. Transfusion 1997; 37: 930–4. English M, Ahmed M, Ngando C, Berkley J, Ross A. Blood transfusion for severe anaemia in children in a Kenyan hospital. Lancet 2002; 359: 494–5.

Global blood transfusion Hassall O, Bedu-Addo G, Adarkwa M, Danso K, Bates I. Umbilical-cord blood for transfusion in children with severe anaemia in under-resourced countries. Lancet 2003; 361: 678–9. Lackritz E, Campbell C, Ruebush T. Effect of blood transfusion on survival among children in a Kenyan hospital. Lancet 1992; 340: 524–8.

Wellcome Trust. Rapid test for hidden disease. http://www.wellcome.ac.uk/en/1/awtprerel1203n310. html World Health Organization. Blood safety for too few. Press release WHO/25, 7 April 2000.

423

Chapter 37

The design of interventional trials in transfusion medicine Paul C. Hébert, Alan Tinmouth and Dean Fergusson

Introduction Randomized controlled clinical trials (RCTs) have evolved to become the ‘gold standard’ clinical research design used to distinguish the risks and benefits of therapeutic interventions. In 1948, for the first time a controlled clinical trial made use of random allocation, a control group and blinding. Additional principles guiding the design of RCTs were first elaborated by Sir Austin Bradford-Hill in the 1960s. Many important questions regarding the use of blood products and alternatives such as blood conservation therapies have not been the subject of well-designed and executed RCTs. Consequently, clinicians frequently base their therapeutic decisions on suboptimal levels of clinical evidence, including observational studies, poorly controlled clinical trials or laboratory studies, and personal experience or observations, which are not evidence based. There are a number of plausible reasons why there are so few large clinical trials in transfusion medicine: • transfusion medicine has historically been a laboratory-based specialty with research focused on the product; • a dearth of clinical epidemiologists and clinical trialists interested in transfusion medicine; • difficulty in obtaining funding for research of a supportive as opposed to a curative therapy; • many of the products have been in standard use for years;

We wish to thank our students, teachers and colleagues who contributed many of the ideas outlined in this manuscript.

424

• there are few industry partners willing to invest in large clinical trials given that products are already in wide use. It could also be postulated that unique difficulties in the field may have impeded the development of important clinical studies. In this chapter, we outline some of the methodological issues central to the development and conduct of interventional trials and RCTs in transfusion medicine.

What is unique about transfusion medicine? There are a number of unique difficulties in transfusion medicine that require consideration in the development of clinical research. First and foremost, transfusion medicine as a discipline is largely based upon the provision of supportive interventions in the treatment of a wide variety of acute illnesses as opposed to disciplines such as cardiology or oncology where interventions are evaluated in well-defined diseases. Furthermore, studies are conducted by trialists trained within a specific discipline (e.g. oncologists, cardiac surgeons). In comparison, blood products are indicated in response to conditions such as anaemia or coagulopathy induced by medications or disease entities. The supportive nature of transfusions leads to consideration of outcomes that may not be directly clinically relevant to the underlying disease process. In addition, most benefits and risks of care would not be attributed to supportive interventions. Conditions that require blood products, such as anaemia and coagulopathies, occur in a broad range of diseases. This raises significant difficulties in designing studies and setting a research agenda.

Interventional trials in transfusion medicine

If one tries to evaluate a transfusion intervention in many diseases, then larger sample sizes and more robust outcomes are required to account for the variation within the patient population. Alternatively, many smaller trials in targeted populations may also be considered. However, this strategy will limit the generalizability of the study as the conclusions may not be applicable to patient groups outside the studied target population. A further concern is the complex biological nature of blood products. For instance, red cell concentrates are prepared using a variety of techniques and storage media, and intravenous immunoglobulins are manufactured by different companies using different processes. This leads investigators to consider whether studies evaluating a transfusion intervention should only be done with one product or preparation, or with many different preparations. Indeed, there may be unforeseen or unexpected clinical consequences due to different approaches to the preparation of blood products. In the planning of studies, one must carefully consider whether products are sufficiently similar to consider including in the same study. Regulatory concerns within or between jurisdictions may also impact on the choice of products included in the study. As discussed in subsequent sections, regulations may not permit the conduct of randomized trials for studies assessing the clinical consequences of different products or testing procedures. Under such constraints, quasi-experimental designs such as before-and-after studies or time-series analyses should be considered. One of the remaining unique aspects of conducting clinical research in transfusion medicine is that transfusions are often incorporated into complex care paths or within therapeutic algorithms of care. The evaluation of transfusions with many other interventions and diagnostic tests increases the complexity of any clinical evaluation.

Types of studies To ascertain the effectiveness of an intervention, the RCT remains the preferred study design as it minimizes the most important biases if properly

conceived and executed. Despite being the gold standard, there are often practical, legal, financial and ethical limitations to the use of clinical trials. For instance, exposing subjects to undesirable and dangerous interventions such as cigarettes and toxins would not be permitted in an RCT. While many of these limitations have been well described, one unique obstacle encountered in transfusion medicine is the conduct of an RCT when an intervention is universally implemented, such as a new processing method or testing procedure for the entire blood supply. By universally implementing an intervention such as universal prestorage leucocyte reduction or universal polymerase chain reaction (PCR) testing for hepatitis C, an RCT becomes impossible within that population. If an RCT is not possible, other study designs including quasi-experimental and observational designs should be considered. Observational studies

Two types of observational designs are often considered in clinical research, case–control studies and cohort or prognostic studies. A case–control study refers to a study where one identifies a group of individuals with an outcome and another group of individuals who would be considered at risk of developing the outcome. Once both groups have been identified, investigators usually seek to identify potential risk factors in both the group with the outcome and the controls. This classic epidemiological design is ideally suited to the investigation of rare diseases and the identification of potential aetiological or risk factors. In transfusion medicine, case–control studies would be ideally suited for the initial study of rare conditions such as transfusion-related acute lung injury (TRALI) and the association between blood transfusion and variant Creutzfeldt–Jakob disease (vCJD). By definition, this study design is always retrospective in nature. Cases, controls and potential risk factors are identified from historical records or past events. By comparing 46 cases of patients with TRALI (cases) and 225 randomly selected transfusion recipients without TRALI (controls), Siliman et al. were able to identify that certain diagnoses (haematological 425

Chapter 37

malignancies and cardiac disease) and the age of the platelets were associated with TRALI. In a smaller subset of cases and controls, the implicated units also had greater neutrophil priming activity as compared with controls. While these results demonstrate an association between neutrophil priming activity, which increases in older platelets, and TRALI reactions, the case–control design does not allow causation to be determined. Despite some of the potential advantages of this study design, it is also difficult to do well and is fraught with potential biases. A systematic review of case–control studies attempting to establish the association between blood transfusion and vCJD demonstrated that blood transfusion had a protective effect. Such a protective effect makes little sense and is probably the result of some biased approach to the sampling of controls. The second observational design choice is a cohort study. In this type of study, individuals are identified well in advance of developing a disease and followed forward in time. Ideally, information on potential risk factors would be gathered from patients throughout the period of observation until the occurrence of an outcome or the end of the study. If subjects are identified well in advance of the development of a disease, then comparing individuals who have a given risk factor with individuals who do not may provide important clues to the aetiology of a disease or health state. It may also lead to a better understanding of the course of the disease and its incidence. If patients are identified and followed once a disease has developed, then this design may also provide invaluable prognostic information. Cohort studies are most powerful when all eligible individuals are identified early, followed prospectively and without any losses to follow-up. A number of cohort studies have examined the relationship between anaemia, red cell transfusion and outcomes such as hospital mortality. These studies illustrate both positive and negative attributes of cohort studies. A retrospective study conducted by Carson and colleagues evaluated the relationship between increasing degrees of anaemia, the presence of ischaemic heart disease and mortality rates. In 1958 Jehovah’s Witness patients, the adjusted odds of death increased from 426

2.3 (95% confidence interval (CI) 1.4–4.0) to 12.3 (95% CI 2.5–62.1) as preoperative haemoglobin concentrations declined from 10.0–10.9 g/dL to 6.0–6.9 g/dL in patients with cardiac disease as compared with patients without cardiac disease. This study shows a clear relationship between increasing anaemia and death. In comparison, the risks of anaemia and benefits of transfusions may be quite complex. This interdependence is often referred to as confounding by indication. Observational studies in transfusion medicine have attempted to compare clinical outcomes at varying haemoglobin concentrations in transfused and non-transfused patients in various clinical settings. Previous studies have major limitations including confounding by indication. Wu et al. retrospectively studied Medicare records of 78 974 patients older than age 65 who were hospitalized with a primary diagnosis of acute myocardial infarction. The authors then categorized patients according to their admission haematocrit. Although anaemia defined in the study as a haematocrit of less than 39% was present in nearly half the patients, only 4.7% (3680 of 78 974) patients received a red cell transfusion. Lower admission haematocrit values were associated with increased 30-day mortality, with a mortality rate approaching 50% among patients with a haematocrit of 27% or lower who did not receive a red cell transfusion. Even though the study was published in a prestigious journal, there were several major limitations making it difficult to draw useful conclusions: • confounding by indication may have been a major concern; • admission haematocrits rather than haematocrits prior to transfusion were used in the analysis; • the timing and number of transfusions were unknown; • few adjustments for other sources of confounding were undertaken because of the retrospective nature of the study; • inability to adjust for important but unmeasured variables. One of the positive aspects of the study by Wu and colleagues was their approach to analysis. They did not presume that the risks of anaemia

Interventional trials in transfusion medicine

and transfusion would be similar at all degrees of anaemia. This was one of the most important limitations of a recent study by Vincent and colleagues. These authors completed a prospective observational cross-sectional study involving 3534 patients admitted to 146 western European intensive care units (ICUs) during a 2-week period in November 1999; 37% of patients received a red cell transfusion during their ICU admission, with the overall transfusion rate increasing to 41.6% over a 28-day period. For those patients who were transfused, the mean pretransfusion haemoglobin concentration was 8.4 ± 1.3 g/dL. In an effort to control for confounding created by illness severity and the need for transfusion, these investigators employed a strategy of matching transfused and non-transfused patients based on their propensity to receive a transfusion, thereby defining two wellbalanced groups (516 patients in each group) to determine the influence of red cell transfusions on mortality. Using this approach, the associated risk of death was increased instead of decreased by 33% for patients who received a transfusion compared with similar patients who did not receive blood. However, as pointed out in the accompanying editorial, the results may have differed if the propensity scores were derived separately for categories of pretransfusion haemoglobin concentrations (e.g. <8, 8–10 and >10 g/dL) instead of haemoglobin concentrations at ICU admission. For example, if one were to consider groups of patients with a pretransfusion haemoglobin concentration below 6.0 g/dL, it is unlikely that the observed 33% increase in mortality would similarly hold true or blood transfusion would never have been recommended. The use of cohort studies in the evaluation of a universally implemented intervention such as prestorage leucocyte reduction requires that subjects either be sampled over a period of time prior to and after the implementation of the programme (a ‘before-and-after’ or interrupted time-series study) or that sampling occur among subjects who received leucocyte-reduced blood products and another population that did not receive such products (standardized incidence study). In this type of study, a standardized incidence ratio is calculated by comparing (standardizing) the incidence of an

outcome in a defined exposed population with that of another population. In the standardization procedure, care is taken to adjust for important confounders. Using universal leucocyte reduction as an example, the incidence of nosocomial infection in Canadian patients receiving a transfusion could be compared to a Japanese population of transfused patients receiving non-leucocyte-reduced blood products. In comparison, a before-and-after study design measures the frequency of an outcome in a specified population during a period of time when the exposure is absent followed by a measurement in the same population during a period of time where exposure is present. Consecutive periods before and after the implementation of a treatment are often compared. One type of before-and-after study is the interrupted timeseries design that proposes to make determinations of an outcome at multiple time points, rather than only one, before and after the implementation of an intervention. Well-executed case–control studies may provide clues about the aetiology or risk factors associated with the development of a disease or a complication. A cohort study may provide the best estimate of incidence, prognosis and risks associated with the development of a disease or its complications. Both designs provide weak inferences regarding specific therapeutic interventions because many forms of bias and confounding remain even after complex multivariable analysis. Before-andafter studies and time-series analysis, both quasi-experimental designs, may provide some inferences regarding clinical consequences attributed to the implementation of a universal programme when a randomized trial is not possible. Inherent in both case–control and cohort studies is the inability to determine causality between a risk factor or treatment, and a specific outcome.

Randomized controlled trials Overall design approaches for RCTs

Clinicians, hospital administrators and policymakers should always seek to identify the best evidence for decision-making. Researchers should aspire to conduct the highest-quality studies. For 427

Chapter 37

therapeutic interventions, there is little debate that this should be an RCT. However, there should be an awareness that randomized trials may be complex. The question being addressed, the many choices and compromises made by the investigators pertaining to different study manoeuvres, such as the selection of patients and centres, may affect inferences made from the results of the trial. In this section, a conceptual framework is provided for randomized trials that should assist providers and consumers of clinical research. The ideal RCT establishes whether therapeutic interventions work, and determines the overall benefits and risks of each alternative in predefined patient populations. This is accomplished by minimizing the influence of chance, bias and confounding through appropriate methodology. In addition, the ideal RCT should attempt to fulfill its objectives with the fewest patients possible (often termed ‘statistical efficiency’). Unfortunately, these objectives are often in direct conflict rather than complementary. More importantly, economic considerations often limit our ability to fulfill all these objectives. For instance, by maximizing the efficiency of a study, investigators might sacrifice their ability to draw conclusions in clinically important subgroups because of inadequate sample size. The most important consequence of these conflicting objectives is that choices made in the design of RCTs must focus on whether an intervention works or whether it results in more good than harm for patients. Trials that attempt to determine therapeutic efficacy address the question ‘Will the therapy work under ideal conditions?’ Trials attempting to determine therapeutic effectiveness address the question ‘Will the therapy do more good than harm under usual practice conditions in all patients who are offered the intervention?’ Clearly, both questions will yield useful information for health practitioners. Efficacy is often established first, then the intervention is evaluated for its effectiveness. In pivotal RCTs used in the final phase of obtaining regulatory approval (phase III trials), pharmaceutical companies primarily wish to demonstrate that their product has proven efficacy; rarely are attempts made to demonstrate therapeutic effectiveness. The design characteristics of efficacy and effec428

tiveness trials tend to differ considerably (Table 37.1). As a consequence of design choices, inferences and threats to the validity of effectiveness and efficacy trials are different. Therefore, one of the first steps in planning an RCT is to determine which of these two design approaches will best reflect the primary study question. Efficacy trials often opt for restricted eligibility, rigorous treatment protocols, and disease-specific outcomes responsive to the potential benefits of the experimental intervention. By using this approach efficacy studies attempt to maximize internal validity, defined as the extent to which the experimental findings represent the true effect in study participants. As a consequence, this design approach will often lack the ability to maximize external validity, defined as the extent to which the experimental findings in the study represent the true effect in the target. Hence, there is often a trade-off between the two forms of validity in any one study. As an example of an efficacy trial, Rivers and colleagues undertook an RCT in which they randomly allocated 263 patients with early sepsis and septic shock to receive either goal-directed therapy using a monitor of continuous central venous saturation in one group versus standard care in the other arm of the trial. Both groups received fluids, vasopressors such as noradrenaline (norepinephrine), inotropic agents such as dobutamine, and red cell transfusions according to a strict clinical algorithm. In addition to the clinical algorithm, the goal-directed therapy arm was required to maintain mixed venous saturation greater than 70%. Saturations below 70% suggest an ongoing oxygen debt and shock. In the first 6 h of care in the emergency department, the experimental group received more fluids (4981 mL vs. 3499 mL, P < 0.001), more inotropic agents (13.7% vs. 0.8% of patients, P < 0.001) and more red cell transfusions (64.1% vs. 18.5% of patients, P < 0.001). As a result of the multiple interventions including red cells, in-hospital mortality was decreased from 46.5 to 30.5% (P < 0.009). In this efficacy study, many of the study manoeuvres were tightly controlled. Specifically, the trial was conducted in a single tertiary centre, by a small number of experts, in a well-defined patient population using elaborate treatment algorithms in

Interventional trials in transfusion medicine Table 37.1 Comparison of study characteristics using either an efficacy or an effectiveness approach when designing a study.

Study characteristics

Efficacy trial

Effectiveness trial

Research question

Will the intervention work under ideal conditions?

Will the intervention result in more good than harm under usual practice conditions?

Setting

Restricted to specialized centres

Open to all institutions

Patient selection

Selected, well-defined patients

A wider range of patients identified using broad eligibility criteria

Study design

Smaller RCT using stringent rules

Larger multicentre RCT using simpler rules

Baseline assessment

Elaborate and detailed

Simple and clinician friendly

Intervention

Tightly controlled Optimal therapy under optimal study conditions

Less controlled Therapy administered by investigators using accepted approaches

Treatment protocols

Rigorous and detailed Compliance

Very general Non-compliance tolerated

End points

Disease-related Related to biological effect Surrogate end points

Patient-related such as all-cause mortality or quality of life

Analysis

By treatment received Non-compliers removed

Intention-to-treat All patients included

Elaborate Detailed and rigorous

Minimal and simple Minimal

Data management Data collection Data monitoring*

* Data monitoring refers to the review of source documents and adjudication/verification of outcomes.

both the experimental and standard of care groups. This efficacy approach may be contrasted with large cardiovascular trials such as the International Study of Infarct Survival trials in acute myocardial infarction which enrolled thousands of patients. One of the major shortcomings of effectiveness trials is the limited data collection and the limited control imposed on most aspects of the study design, thereby increasing biological variability, minimizing information on biological mechanisms and curtailing the possibility of understanding negative results or the influence of cointerventions and confounding on study outcomes. At this juncture, there are no published RCTs in transfusion medicine that have opted for a large simple trial design approach. There is, however, a recently published large randomized trial of 7000 critically ill patients that compares 4% albumin to saline in a wide variety of patients. This trial found that mortality did not differ between the two groups. There was no evidence of

harm by albumin, but the lack of benefit of albumin makes the use of albumin, which is expensive, hard to justify in this group of patients. Many trials opted for a hybrid approach between large simple trials and tightly controlled clinical studies. The Transfusion Requirements in Critical Care trial provides such an example. This was an 838-patient trial that randomly allocated patients to either a restrictive or a liberal transfusion strategy. The study was conducted in 25 clinical centres, enrolled patients using broad eligibility criteria, followed simple treatment strategies for the administration of red cells and ascertained mortality rates and rates of organ failure. RCT design alternatives

Once investigators have chosen whether an efficacy, effectiveness or a hybrid approach will best answer the research question, there are several design options that may be considered (Table 429

Chapter 37 Table 37.2 Types of randomized clinical trial (RCT) designs.

Type of RCT

Description

Advantage

Disadvantage

Two group parallel

Patients randomized to one of two groups

Simplest approach Widely used and accepted

Limited to simple comparisons

Factorial (2 ¥ 2)

Patients randomized to one of four groups: therapy a, therapy b, therapy a + b or control

Combinations of therapy may be compared

Larger sample size required More complex design

Sequential study

Pairs of outcomes continuously compared in patients randomized to one of two therapies

Ongoing evaluations of therapy

Limited uses (efficacy only) Not a well-accepted approach Sample size unpredictable

Two-period crossover

Patients allocated to one of two therapies, then receive other therapy in second treatment period

Smaller sample required

Limited to reversible outcomes Major concern with carry-over effect

N of study

Single patient sequentially and repeatedly receives a therapy and a placebo

Optimal method of determining if a therapy is beneficial to given patients

Results not generalizable Difficult in unstable patients Very labour intensive

Cluster design

Groups of patients are randomized

Ideal for programme or guideline evaluation

Less well accepted in clinical practice Difficult to implement when large variability between clusters

37.2). A two-group parallel design is the most common of RCT design choices. In this design, patients are randomly allocated to one of two therapeutic interventions and followed forward in time. It is the simplest to plan, implement, analyse and, most importantly, interpret. Therefore, a parallel group design is the most frequently adopted choice of RCT design. Parallel group designs may also be used to independently compare three or more treatments. The use of factorial designs may also be considered when a number of therapies are being evaluated in combinations. For instance, in a twoby-two (2 ¥ 2) factorial design, two interventions are tested both alone and in combination, and compared with a control group (usually a placebo). This means that investigators can efficiently test two interventions with only marginal increases in sample size. In addition, the benefits of treatment combinations can be evaluated in a controlled manner. This design is most useful when interactions are either very strong or non-existent. Thus, before embarking on a large, more complex 430

factorial study, investigators should expect either strong additive or synergistic effects from combined therapy or none at all. Prospective investigators should realize that detecting interactions is also more difficult and requires much larger sample size as compared with comparison of either therapy with a placebo. Factorial designs have been used very successfully to evaluate thrombolytic therapy in combination with an antiplatelet agent in acute myocardial infarction and unstable angina. Factorial designs imply concurrent comparisons between at least two therapies. It is also possible to implement a design that compares interventions sequentially. For example, two therapies in the early treatment of a disease could be compared, followed by the evaluation of a second intervention in the late phase of care several days later. The authors are not aware of a study in transfusion medicine that has made use of a factorial design. An example of a factorial trial would be to randomize patients to an algorithm of care versus standard care in addition to either a conservative

Interventional trials in transfusion medicine

or liberal transfusion threshold. Traditionally, the factorial design is used to answer two unrelated study questions. Both the simple parallel group design and a factorial design are designed using classical or frequentist statistical approaches, where the sample size is fixed according to pre-established assumptions (anticipated outcomes in treatment and control group, power and significance levels) prior to the commencement of enrolment. There are other experimental designs that are more responsive to patient outcomes as the study progresses. Sequential designs use frequentist statistical methods to set boundaries for significance levels which consider the increasing number of comparisons and sample size throughout the study. True sequential studies randomly allocate patients to receive one of two therapies. Pairs of patients are then sequentially compared. The study is terminated as soon as one of the significance boundaries is crossed. One of the major concerns with this design may be its inability to conceal the randomization process and the uncertainty of not knowing the exact sample size in advance. From this methodology, several biostatisticians have developed methods of performing interim analyses in large clinical trials, referred to as group sequential methods. A Bayesian statistical approach offers an alternative methodology. In a Bayesian analysis, previous beliefs about the effectiveness of a therapy are combined with the observed data from the trial to provide a new revised set of likely values for the effectiveness of the therapy. This approach allows for repeated or continuous monitoring of study results as patients accrue. As a result, predetermined sample sizes are not required and enrolment continues until the results meet predetermined significance levels. This can allow for increased trial efficiency as studies will not enroll additional patients unnecessarily or terminate the study prematurely. A Bayesian approach has also been advocated for interim analyses of large clinical trials. Another RCT design option particularly amenable to an efficacy evaluation is a two-period crossover study in which patients are used as their own controls. In a two-period crossover trial, patients are randomized to one of two therapies

for a fixed period of time and then proceed to receive the other therapy in a second comparable interval. Minimizing ‘between-subject’ variability in this manner makes significant gains in efficiency. Crossover studies are therefore best suited to relatively stable conditions (stability is required during the study), interventions with rapid onset of action and a very short half-life (the biological effect must disappear prior to the second treatment period), and rapidly modifiable end points such as haemodynamic and respiratory measures. An example of a crossover trial in transfusion medicine would be the evaluation of a modified red cell product (e.g. bacterially inactivated or pegylated red cells) in patients with transfusion-dependent congenital anaemias. The time to next transfusion or red cell survival could be measured in patients who receive, in a random order, standard red cell transfusions and modified red cell transfusions for fixed periods of time. An appropriate washout time between the two interventions is required to ensure there is no contamination of the modified red cells during the period of standard transfusions. All designs discussed so far have described the evaluation of interventions for individual patients. However, it is sometimes necessary to evaluate therapies, protocols, guidelines or treatment programmes for groups of individuals. Using this design, groups such as ICUs, wards, hospitals, physician practice, often referred to as clusters, are randomized to receive alternative interventions. Cluster design may be the most appropriate design for evaluating interventions such as the implementation of educational interventions, care paths and approaches to auditing transfusion practices. Wilson and colleagues undertook a systematic review of behavioural interventions, such as education and audit/feedback aimed at reducing inappropriate transfusion. The authors identified nine studies that evaluated prospective audits (n = 2), education (n = 4), the use of an algorithm, retrospective audit, and patient-specific decision support. The systematic review suggested that behavioural interventions may be useful in improving transfusion practice. The authors also noted that the quality of the primary studies was extremely poor. Indeed, there was only one study that made use of a cluster design. In this one study, 431

Chapter 37 Table 37.3 Considerations in

Choice of design

determining which design approach to implement in transfusion trials.

Criteria to consider

Favouring efficacy

Favouring effectiveness

Evidence Importance of the question Feasibility

Limited evidence Rare and less serious Not demonstrated

Risks

Unknown or significant consequences Limited or unknown benefits

Efficacy well documented Common and serious problem Adequate accrual and confirmed feasibility Minimal or acceptable risks given benefits Significant benefits anticipated

Benefits

there were only two matched pairs. This would usually be insufficient to draw any strong inferences. All other studies were before-and-after studies, which are susceptible to selection bias from secular changes in practice from one time period to the next. One of the major concerns is the possibility of large variations between clusters that may make it difficult to detect actual differences between therapies. Selecting a study population

In transfusion medicine, most blood products are currently used in a wide variety of diseases and conditions. The choice of study population will invariably depend on the study question, the underlying hypothesis and on a number of other factors. The choice of a hypothesis that will address either therapeutic efficacy or effectiveness will have a substantial impact on the selection of the study population. Specifically, in choosing an efficacy approach, investigators usually perform the study in a well-defined patient population where the intervention has the highest probability of demonstrating an effect. This may be done by narrowly defining the patient population through the use of restrictive eligibility criteria and disease definitions, as well as selecting specialized centres with clinical expertise in the field. Choosing a narrowly defined study population will decrease overall variability attributed to patient selection but may have adverse consequences such as hampering patient recruitment and jeopardizing the generalizability of study results. When defining the 432

eligibility criteria for an effectiveness trial, investigators should consider utilizing more liberal criteria in a wide range of clinical settings. Thus, as the study is being designed, medical or surgical critically ill patients with a broad range of primary diagnoses (or underlying conditions) from a range of tertiary care centres might be considered for enrolment in the study. On the spectrum between highly selected patients (efficacy) and large patient population (effectiveness), investigators should consider a number of factors in making the decision (Table 37.3). The spectrum of biological activity of the intervention is an important consideration. For instance, a narrow spectrum of biological activity should translate into restricted eligibility while a broad spectrum of biological activity should yield more liberal eligibility criteria. Eligibility may also be restricted through the selection of study centres. In efficacy trials, highly specialized units should be sought while studies evaluating the effectiveness of an intervention would require the inclusion of a large number of non-specialized centres. In practice, investigators should first focus on therapeutic efficacy and seek to first study the intervention in high-risk and/or well-defined patient populations initially. Selecting outcomes (Table 37.4)

In most clinical trials, the clinical investigative team should consider a number of potential outcomes, both fatal and non-fatal. An outcome is defined as a measurement (e.g. haematocrit) or an

Interventional trials in transfusion medicine Table 37.4 Guides to the choice of outcome measure in a

randomized clinical trial. Is the outcome causally related to the consequences of the disease? Is the outcome clinically relevant to the healthcare providers and/or patients? Has the validity of the outcome (for complex outcomes such as scoring systems or composite outcomes) been established? Is the outcome easily and accurately determined? Is the outcome responsive to changes in a patient’s condition? Is the outcome measure potentially able to discriminate between patients who benefit from a therapy and patients in the control group?

event (e.g. death) potentially modified following the implementation of an intervention. If all are given equal consideration, concerns arise about multiple comparisons and interpretation of a study with heterogeneous findings. Thus it is important to choose a primary outcome that will determine an intervention’s therapeutic success or failure, as well as secondary outcomes that will provide supportive evidence in secondary analyses. As a corollary, a predefined hierarchy implies that the investigators believe that clinically or statistically important differences in secondary outcomes, in the absence of important changes in the primary outcome, will not be interpreted as strong evidence of therapeutic benefit. The primary outcome is also essential in determining the sample size requirements in a clinical trial. Thus, once a decision has been made to determine either therapeutic efficacy or effectiveness (or possibly a combined approach), the second task facing investigators is ranking outcomes as primary and secondary. The choice of study outcome is one of the most important design considerations to be made by investigators. However, there are a number of factors that should be considered prior to selection of an outcome. The primary outcomes should be considered clinically important and easily ascertained. By fulfilling these two criteria, the investigator will have a much greater chance of influencing clinical practice once a study has been completed and published. Outcomes should also measure what they are supposed to measure (validity) and be precise and

reproducible. An outcome must be able to detect a clinically important true positive or negative change in the patient’s condition following a therapy. In a recent study evaluating erythropoeitin-a in 1302 critically ill patients from 65 American centres, Corwin and colleagues, after consultation with the Food and Drug Administration (FDA) and the pharmaceutical sponsor, opted to compare the proportion of patients who received at least one red cell unit once randomized. The study documented a 10% decrease in the proportion of patients avoiding a transfusion (60% to 50%, P < 0.05). They were unable to detect differences in any clinically important outcomes beyond transfusions. As a consequence, erythropoeitin has not received approval for use in ICUs by the FDA and has not been widely adopted into clinical practice in critical care. Using mortality as an outcome, the sample size in a clinical trial comparing two therapies is based on the baseline event rate, the expected incremental benefit or difference, the level of significance (a) and the power to detect differences (1 – b). Establishing the incremental benefit of a new therapy is vitally important because of the enormous sample size repercussions. A sample size calculation for an RCT requires that the investigators establish the minimum therapeutic effect detectable within the trial. This difference in outcomes between interventions is referred to as the minimally important difference (MID) or minimal clinically important difference (MCID). The MID is essentially establishing the level of discrimination in the study population who are exposed to the interventions given acceptable levels of type I (finding a difference when one does not truly exist) and type II (not finding a difference when one truly exists) errors and the baseline event rate. Too often, investigators calculate a sample size based on very large and unrealistic expected differences in outcomes. To determine a plausible effect size, investigators should ask themselves the following questions. • What difference or incremental benefit can be realistically expected of the experimental therapy? (Anticipated biological effect of therapy.) • Are the required number of patients available to participate in the clinical trial? (Feasibility.) • How much of a survival benefit, given the added 433

Chapter 37

costs and expected adverse effects of therapy, would be required for clinicians, patients and administrators to adopt a new therapy? (Overall benefit of therapy.) As a concrete example, assuming that a given study population has an expected mortality rate of 25% in the standard therapy group while the experimental therapy is expected to decrease mortality by an absolute difference of 12.5% (a 50% relative risk reduction), the total number of patients required would be approximately 250. Most therapies used in the ICU would not be expected to decrease mortality so dramatically. More realistic expectations may be in the range of a 5% absolute decrease (a 20% relative risk reduction) which would require a total sample size of 2200 patients respectively if the baseline mortality was 25%. Investigators need to consider whether an absolute incremental benefit in the range of 5–10% is attainable using the experimental therapy. If not, another more discriminating outcome should be sought.

Conclusion In this chapter on interventional studies in transfusion medicine, several major RCT design characteristics have been discussed. Study design issues of special interest to health professionals interested in transfusion medicine have been outlined. Suggestions when planning an RCT in transfusion medicine are provided in Table 37.5. Observational and quasi-experimental studies may provide invaluable information in transfusion medicine. Although RCTs provide the most unbiased and accurate assessment of the efficacy and effectiveness of therapeutic and preventive interventions, they remain challenging and expensive to conduct. As more research groups form to address unanswered therapeutic questions in transfusion medicine, investigators will invariably better understand the strengths and limitations of different RCT design characteristics.

Further reading Table 37.5 Suggestions when planning a randomized

clinical trial (RCT) in transfusion medicine. Explicitly determine whether you are primarily interested in establishing therapeutic efficacy or effectiveness Whenever possible, undertake an RCT as part of a broader research programme If the study intervention is complex (or risky) or if other aspects of study feasibility are questionable, a pilot study should be considered Whenever possible, investigators should use simple rather than complex designs (two-group parallel design vs. factorial design) The study population should be tailored to the intervention Ideally, the study intervention and treatment protocols should not aim to substantially modify or affect usual clinical practice Given the complexity of RCTs, data collection should aim to clearly describe the study population, describe cointerventions and all major study outcomes In choosing primary study end points, investigators should focus on patient-oriented outcomes rather than surrogate or biological markers If you are planning a seminal RCT, you may only have one chance to get it right. When making compromises, always opt to answer questions that most clinicians consider most important In establishing the minimally important difference, select a potentially achievable benefit

434

Campbell DT, Stanley JC. Experimental and Quasi-experimental Designs for Research. Chicago: Rand McNally College Publishing Company, 1966. Carson JL, Duff A, Poses RM et al. Effects of anaemia and cardiovascular disease on surgical mortality and morbidity. Lancet 1996; 348: 1055–60. Corwin HL, Gettinger A, Pearl RG et al. Efficacy of recombinant human erythropoietin in critically ill patients: a randomized controlled trial. J Am Med Assoc 2002; 288: 2827–35. Friedman LM, Furberg CD, Demets DL. Fundamentals of Clinical Trials, 3rd edn. St Louis: Mosby-Year Book, 1996. Guyatt GH, Sackett DL, Cook DJ. Users’ guides to the medical literature II. How to use an article about therapy or prevention. A. Are the results of the study valid? J Am Med Assoc 1993; 270: 2598–601. Hébert PC, Fergusson DA. Red blood cell transfusions in critically ill patients. J Am Med Assoc 2002; 288: 1525–6. Hébert PC, Wells G, Blajchman MA et al. and the Transfusion Requirements in Critical Care investigators for the Canadian Critical Care Trials Group. Transfusion Requirements in Critical Care: A multicentre randomized controlled clinical trial. N Engl J Med 1999; 340: 409–17.

Interventional trials in transfusion medicine The SAFE Study Invetigators. A comparison of albumin and saline for fluid resuscitation. New Engl J Med 2004; 350: 2247–56. Sackett DL. Bias in analytic research. J Chron Dis 1979; 32: 51–63. Sackett DL. The competing objectives of randomized trials. N Engl J Med 1980; 303: 1059–60. Sackett DL, Gent M. Controversy in counting and attributing events in clinical trials. N Engl J Med 1979; 301: 1410–12. Sackett DL, Haynes RB, Guyatt GH, Tugwell P. Clinical Epidemiology: A Basic Science for Clinical Medicine, 2nd edn. Boston/Toronto/London: Little, Brown and Company, 1991. Silliman CC, Boshkov LK, Mehdizadehkashi Z et al. Transfusion-related acute lung injury: epidemiology and

a prospective analysis of etiologic factors. Blood 2003; 101: 454–62. Vincent JL, Baron J-F, Reinhart K et al. Anemia and blood transfusion in critically ill patients. J Am Med Assoc 2002; 288: 1499–507. Wilson K, Macdougall L, Fergusson D, Graham I, Tinmouth A, Hébert PC. The effectiveness of interventions to reduce physician’s levels of inappropriate transfusion: what can be learned from a systematic review of the literature. Transfusion 2002; 42: 1224–9. Wu WC, Rathore SS, Wang Y, Radford MJ, Krumholz HM. Blood transfusion in elderly patient’s with acute myocardial infarction. N Engl J Med 2001; 345: 1230–6.

435

Chapter 38

Getting the most out of the evidence for transfusion medicine Simon J. Stanworth, Susan J. Brunskill and Chris J. Hyde What is meant by evidence-based medicine? Evidence-based medicine (EBM) is a relatively recent phrase which Sackett describes as ‘the integration of best research evidence with clinical expertise and patient values’ (see Further reading A). Proponents of EBM have particularly highlighted the nature of the evidence that is used to make clinical decisions, i.e. where is it from, how believable is it, how relevant is it to my patient and can it be supported by other data? However, evidence is only one of the factors driving clinical decision-making, and in some clinical situations the judgement of the clinician will determine that the available evidence for a specific clinical problem is not relevant. EBM should not be thought of as just about obtaining and evaluating clinical research evidence; it is also a means by which effective strategies for self-learning can be applied, aimed at continuously improving clinical performance. The focus of this chapter will be a discussion of some of the core elements of the practice of EBM. Particular reference will be made to the appraisal of all identified clinical research evidence in a systematic and transparent way. Although these points are initially described in a general way, they apply to transfusion medicine as much as to any other branch of medicine.

What is meant by optimal clinical evidence? Optimal evidence is the best evidence available to answer a question. Since the first randomized controlled trial (RCT) was published comparing two different treatment plans for pulmonary tuberculosis in 1948, this form of study has been generally 436

regarded as the ‘gold standard’ design for providing evidence of the effectiveness of an intervention; in other words, to establish a cause-and-effect relationship for this intervention. This is because if the process of randomization is undertaken correctly, the differences observed between the groups of patients randomized in the trial should be attributable to the intervention being studied and not to other confounding factors related to the patients, study setting or quality of care. RCTs are not without their difficulties. • Randomized trials are costly to undertake, and a number of logistic problems can arise if these studies are conducted at multiple centres (which is often necessary if large numbers of participants need to be enrolled). • Small studies, although easier to develop, may overestimate any observed intervention effect, and may place too much emphasis on those outcomes with more striking results. • Small studies may be designed to look for unreasonably large differences in the effects of an intervention (which they will never be able to show because of the size of the study). • Randomized trials with negative results may never be fully reported, or only found in abstract form (publication bias). • Intervention effects can be overgeneralized and misapplied to different and unstudied patient populations. • RCTs are not suited to investigating adverse effects of an intervention, or to studies of prognostic factors, prevalence or diagnostic criteria. These points illustrate the need for an important phase in the appraisal of the study, that of the quality of the trial report, and this will be revisited later in this chapter.

Evidence for transfusion medicine

Observational studies Observational studies, whether prospective or retrospective in nature, may demonstrate an association between intervention and outcome. But it is often difficult to be sure that this association does not reflect the effects of unknown confounding factors. It is recognized that the influence of confounding factors and biased participant selection can dramatically distort the accuracy of the study findings in observational studies but space precludes further description. The best way to be sure that the observed differences in the results from a non-controlled study are not due to confounding factors is to design a trial with true random allocation of participants. (It is acknowledged that other forms of experimental study design exist, e.g. before and after studies.) However, this does not mean that findings from well-designed observational studies should be disregarded. Such study designs can be very effective in establishing or confirming effects of large size. Interpretation is more difficult when the observed effects are smaller. As mentioned, clinical questions assessing possible aetiology or monitoring adverse effects may be more suited to observational studies.

Appraisal of primary research evidence for its validity and usefulness One component of EBM is the critical appraisal of evidence generated from a study. Published RCTs should provide sufficient detail to allow the reader to make an independent assessment of the trial’s strength and weaknesses. Guidelines and checklists have been designed to help a reader to complete an assessment of the findings in RCTs. As shown in Table 38.1, the key components of this appraisal process for clinical trials relate to the methodology of the study (the participants, interventions and comparators, the outcomes and the methods used in the randomization process, the sample size, blinding) and the presentation of the results (numbers analysed/evaluated, the role of chance, i.e. confidence intervals). Poor study methodology and presentation of results will lead to bias at selection, bias at detection and bias due

Table 38.1 Key components of the critical appraisal process

for clinical trials. (From the Critical Appraisal Skills Programme worksheets, Milton Keynes Primary Care Trust, 2002) Did the study ask a focused question? Was the allocation of participants to the study arms appropriate? Were the study staff and participants unaware (blind) to the treatment allocation? Were all the participants who entered the study accounted for within the results? Were all the participants followed up and data collected in the same way? Was the study sample size big enough to minimize any play of chance that may occur? How are the results presented and what is the main result? How precise are the results? Were all the important outcomes for this patient population considered? Can the results be applied to practice/different populations?

to unacceptable attrition losses and result in a risk of inaccurate conclusions being drawn by readers. Appraisal