Blood Safety and SURVEILLANCE PDF

BLOOD
SAFETY
and
SURVEILLANCE
edited by
Jeanne V. Linden
Wadsworth Center
New York State Department of Health
Albany, New York
Celso Bianco
New York Blood Center
New York, New York
Marcel Dekker, Inc. New York Basel

TM
Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved.

ISBN: 0-8247-0263-8
This book is printed on acid-free paper.
Headquarters
Marcel Dekker, Inc.
270 Madison Avenue, New York, NY 10016
tel: 212-696-9000; fax: 212-685-4540
Eastern Hemisphere Distribution

Marcel Dekker AG
Hutgasse 4, Postfach 812, CH-4001 Basel, Switzerland
tel: 41-61-261-8482; fax: 41-61-261-8896
World Wide Web

http://www.dekker.com
The publisher offers discounts on this book when ordered in bulk quantities. For more
information, write to Special Sales/Professional Marketing at the headquarters address
above.
Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved.
Neither this book nor any part may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopying, microfilming, and recording, or
by any information storage and retrieval system, without permission in writing from the
publisher.
Current printing (last digit):

10 9 8 7 6 5 4 3 2 1
PRINTED IN THE UNITED STATES OF AMERICA

Foreword
Foreword
The blood contains the soul . . .
Such was the claim of the Chinese in the year 1000 B.C.! Since that time, the
importance of blood has largely moved from the spiritual to the medicinal arena.
Over the centuries, belief in the therapeutic benefits of blood has evolved in a
dramatic manner. At one time in the Roman empire, gladiators drank the blood
of their fallen opponents to reinvigorate themselves. Elsewhere, others sought
rebirth from being showered with the blood of a sacrificed bull.
The first blood transfusion is credited to Richard Lower, who transfused the
blood of one dog into another in 1665 (1). In July 1667, Jean Dennis, a young
French physician, performed the first known human transfusion using lamb blood,
preceding Lowers similar attempts by only a few months:
The project of causing the blood of an healthy animal to pass into the veins
of one diseased, having been conceived about ten years ago, in the illustrious
Society of Virtuosi which assembles at your houses and your goodness having
received M. Emmeriz and my self, very favorably at such times as we have
performd to entertain you either with discourse concerning it, or the sight of
some not inconsiderable effects of it (2).
Needless to say, these experimentshowever apparently successful in the

short termled unavoidably to the death of the recipients. Eventually, the French
Faculty of Medicine, as well as the Royal Society in London, criminalized blood
transfusion.
iii
iv Foreword
Although other blood transfusion experiments were done in the eighteenth

and nineteenth centuries, the first true scientific breakthrough occurred in 1900
with the discovery by Karl Landsteiner of the A, B, and O blood groups (3).
From then on, progress was fast and remarkable, aided by the exigencies of World
Wars I and II as well as by the work of organizations such as the Red Cross.
Scores of discoveries occurred during subsequent decades, leading to improve-
ment of blood transfusion and to its complete clinical acceptance. Indeed, millions
of persons have benefited from blood transfusions.
However, the question remains: Is blood transfusion free of risk? Of course,
the issue is not the safety of the procedure itself, but that of the blood supply.
Are there sometimes reactions to the blood that is transfused? Is the blood a
vector to transmit diseases? Unfortunately, despite all the considerable advances
in blood banking and blood safety, the answer to these questions is yes.
This volume, edited by Jeanne V. Linden and Celso Bianco, provides
up-to-date information about the various risks of transfused blood, with a focus
on immunological complications and on transmissible disease risks. But it does
more, because it describes strategies of minimizing these risks. The editors have
assembled a roster of contributors who are well-recognized experts in their fields.
As we enter a new millennium, patients all over the world will be the
beneficiaries of a routine procedureblood transfusionand it should be as safe
as possible. That is what this book is about. The health professionals involved in
the practice of blood or blood product transfusions will benefit from this volume,
but the patients will benefit even more if the strategies described are followed!
Claude Lenfant, M.D.

Bethesda, Maryland
REFERENCES
1. Lower R. The success of the experiment of transfusing the blood of one animal to
another. Philos Trans Royal Soc London 1666; 1:352.
2. Dennis J. A letter concerning a new way of curing sundry diseases by transfusion of
blood. Philos Trans Royal Soc London 1667; 2:489504.
3. Landsteiner J. Uber agglutinationserscheinungen normallen menschlichen blutes.
Wien Klin Wochenschr 1901; 14:11321134.
Preface
Preface
Human immunodeficiency virus (HIV) infection has become the most feared
complication of blood transfusion. Fear of acquired immunodeficiency syndrome
(AIDS) has triggered a series of radical changes in transfusion medicine. These
changes have focused mostly on the transmission of viruses. Other infectious and
noninfectious complications have been less well appreciated. One of the major
goals of this book is to examine transfusion risks in a broader context, and to
present some strategies to minimize these risks. A special effort was dedicated to
areas that have received less attention in recent years, particularly errors and
accidents, and the role of public health agencies as expressed by those in charge
of monitoring and regulation. Between 1990 and 1999, as risks of transmission
of HIV and hepatitis became relatively minuscule, there were 168 reports to FDA
of fatal hemolytic reactions due to ABO-incompatible transfusion through error.
This is an average of 17 cases a year, far exceeding deaths from transfusion-
associated infectious agents. The book also attempts to provide physicians and
other health care providers with the information necessary for appropriate coun-
seling of patients who may need, or who have received, blood transfusions and
for counseling of blood donors with abnormal screening test results.
The first chapter examines the contributions that blood donor screening
procedures make to transfusion safety. The remainder of the book is organized in
related sections to facilitate reference. Immunological complications discussed
include hemolytic transfusion reactions, other types of reactions, alloimmuniza-
tion, graft-versus-host disease, transfusion-related acute lung injury, and immuno-
modulation resulting from allogeneic transfusion. Also included is an overview
of transfusion-related errors and methods to prevent errors that may lead to
v
vi Preface
transfusion of incompatible blood. The following section describes various infec-

tion-related complications resulting from infectious agents that 1) are present at
least transiently in the circulation of donors, 2) survive the blood collection and
storage conditions in at least one component, and 3) are infectious to human
recipients when infused. Some agents pose risk mostly to immunocompromised
patients, who constitute an increasing proportion of recipients of blood and blood
products today because of HIV infection, transplants, and chemotherapy for
malignancies.
Diseases that pose transfusion risks in various parts of the world are
included because the book is intended for a worldwide audience, and also because
some of these diseases could potentially become important in North America.
An overview of infectious disease risks sets the stage for the remaining chapters
in this section, including a description of current screening for transmissible
disease markers, bacterial contamination, viruses transmissible by transfusion
(hepatitis viruses, human T-lymphotropic virus, cytomegalovirus, and others), and
other infectious agents, including parasites and prions.
The last section focuses on methods employed to reduce risks of transmis-
sion of disease by transfusion of blood and blood products. Contents include a
summary of current surveillance efforts in the United States, alternatives to
allogeneic blood transfusion, leukoreduction, viral inactivation, and red blood
cell substitutes. They also review the contributions of quality programs, profes-
sional standards, and federal regulatory oversight to blood safety. A chapter
presenting a cost-effectiveness analysis of various risk reduction strategies com-
pletes this section.
Efforts have been made to present the text in a user-friendly fashion for
easy reference. Tables have been presented whenever possible. A comprehensive
index facilitates the location of pertinent material. Reference lists are not intended
to be comprehensive, but to present relevant, recent references, including review
articles that could be consulted for further detail. A zero-risk blood supply is not
possible. However, we hope that this book will provide information to put risks
in perspective and encourage employment of risk reduction strategies.
We wish to thank all the authors for their efforts. They are a diverse group
of experts in their fields. We believe their aggregate contributions have resulted
in a well-rounded single resource for a multifaceted and highly complex subject.
Jeanne V. Linden
Celso Bianco
Contents
Contents
Foreword Claude Lenfant iii

Preface v
Contributors xi
I. Introduction
1. Impact of Blood Donor Screening Procedures on

Transfusion Safety 1
Steven Kleinman
II. Immunological Complications
2. Hemolytic Transfusion Reactions 47

Elizabeth J. Kicklighter and Harvey G. Klein
3. Other Reactions and Alloimmunization 71
Christopher J. Gresens and Paul V. Holland
4. Managing Error for System Improvement 87
Harold S. Kaplan and James B. Battles
5. Graft-Versus-Host Disease 109
Iain J. Webb and Kenneth C. Anderson
6. Transfusion-Related Acute Lung Injury 125
Mark A. Popovsky
vii
viii Contents
7. Clinical Effects of the Immunomodulation

Associated with Allogeneic Blood Transfusions 139
Jos O. Bordin and Morris A. Blajchman
III. Infectious Complications
A. General
8. Overview of Infectious Disease 159

Roger Y. Dodd
9. Blood Donor Screening and Supplemental Testing:
Principles, Procedures, and Consequences 185
Jay E. Valinsky
B. Bacterial Infection
10. Bacterial Contamination 221

Laura L. Spinelli and Mark E. Brecher
C. Viral Infections
11. HIV and Blood Transfusion 251

Celso Bianco and Maria Rios
12. Hepatitis Viruses and Blood Transfusion 279
Alfred M. Prince
13. Human T-Cell Leukemia Virus 295
Maria Rios and Celso Bianco
14. Transfusion-Acquired Cytomegalovirus Infection:
Approaching Resolution 315
Gary E. Tegtmeier
D. Other
15. Creutzfeldt-Jakob Disease, Other Human Transmissible

Spongiform Encephalopathies, and Transfusion of
Blood and Blood Products 335
Celso Bianco
16. The Protozoan ParasitesMalaria and Chagas Disease 355
Silvano Wendel
Contents ix
17. Tickborne Infections 399

Ritchard G. Cable and Jonathan Trouern-Trend
IV. Strategies to Reduce Risk
18. Surveillance for Transfusion-Transmitted

Infectious Diseases 423
Mary E. Chamberland and Rima F. Khabbaz
19. Alternatives to Allogeneic Blood and Strategies to
Avoid Transfusion 447
Lawrence T. Goodnough
20. Leukoreduction 463
Anne B. McDonald and Walter H. Dzik
21. Viral Inactivation 479
Bernard Horowitz and Ehud Ben-Hur
22. The Role of Quality in Blood Safety 497
Lucia M. Berte and David E. Nevalainen
23. Cost-Effectiveness Analysis of Risk-Reduction
Strategies 521
James P. AuBuchon
24. Red Blood Cell Substitutes 543
Zbigniew M. Szczepiorkowski and Christopher P. Stowell
25. Professional Standards and Voluntary Accreditation 569
Jay E. Menitove and Hillary V. Schaeffler
26. The Role of Federal Regulation in Blood Safety 579
Jay S. Epstein and Mary Gustafson
Index 607
Contributors
Contributors
Kenneth C. Anderson, M.D. Department of Adult Oncology, Dana-Farber

Cancer Institute, Medical Director, Kraft Family Blood Donor Center, and
Professor of Medicine, Harvard Medical School, Boston, Massachusetts
James P. AuBuchon, M.D. Professor of Pathology and Medicine, Dartmouth-

Hitchcock Medical Center, Lebanon, New Hampshire
James B. Battles, Ph.D. Associate Professor, Office of Medical Education,

University of Texas, Southwestern Medical Center, Dallas, Texas
Ehud Ben-Hur, Ph.D. Consultant in Photomedicine, New York, New York
Lucia M. Berte, M.A., MT(ASCP)SBB, DLM, CQA(ASQ),CQMgr. Quality

Systems Consultant, Elmhurst, Illinois
Celso Bianco, M.D. Vice President, Medical Affairs, New York Blood Center,
New York, New York*
Morris A. Blajchman, M.D., F.R.C.P.(C.) Professor, Department of Pathology

and Medicine, McMaster University, and Hamilton Centre Canadian Red Cross
Society, Hamilton, Ontario, Canada
*Current affiliation: Executive Vice President, Americas Blood Centers, Washington, D.C.
xi
xii Contributors
Jos O. Bordin, M.D., Ph.D. Associate Professor, Hematology and Transfusion

Medicine Services, Escola Paulista de Medicina, So Paulo, Brazil
Mark E. Brecher, M.D. Director, Transfusion Medicine Service, and Professor,

Pathology and Laboratory Medicine, University of North Carolina, Chapel Hill,
North Carolina
Ritchard G. Cable, M.D. Medical Director, American Red Cross Blood

ServicesConnecticut Region, Farmington, Connecticut
Mary E. Chamberland, M.D., M.P.H. Assistant Director for Blood Safety,

Division of Viral and Rickettsial Diseases, Centers for Disease Control and
Prevention, Atlanta, Georgia
Roger Y. Dodd, Ph.D. Head, Transmissible Diseases Department, Holland

Laboratory, American Red Cross, Rockville, Maryland
Walter H. Dzik, M.D. Director, Blood Bank and Tissue Typing Labora-
tory, Beth Israel Deaconess Hospital, and Harvard Medical School, Boston,
Massachusetts
Jay S. Epstein, M.D. Director, Office of Blood Research and Review, Center
for Biologics Evaluation and Research, U.S. Food and Drug Administration,
Rockville, Maryland
Lawrence T. Goodnough, M.D. Professor of Medicine and Pathology, Depart-

ments of Pathology and Internal Medicine, Washington University School of
Medicine, St. Louis, Missouri
Christopher J. Gresens, M.D. Associate Medical Director, Sacramento Medi-

cal Foundation Blood Centers, Sacramento, California
Mary Gustafson, M.S., MT(ASCP)SBB Director, Division of Blood Applica-

tions, Office of Blood Research and Review, Center for Biologics Evaluation and
Research, U.S. Food and Drug Administration, Rockville, Maryland
Paul V. Holland, M.D. Medical Director and CEO, Sacramento Medical Foun-
dation Blood Centers, Sacramento, California
Bernard Horowitz, Ph.D. VITEX (V.I. Technologies, Inc.), New York,

New York
Contributors xiii
Harold S. Kaplan, M.D. Professor of Clinical Pathology and Director, Trans-

fusion Medicine, New York-Presbyterian Hospital, New York, New York
Rima F. Khabbaz, M.D. Deputy Director, Division of Viral and Rickettsial

Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia
Elizabeth J. Kicklighter, M.D. Warren G. Magnuson Clinical Center, National

Institutes of Health, Bethesda, Maryland
Harvey G. Klein, M.D. Chief, Department of Transfusion Medicine, Warren G.

Magnuson Clinical Center, National Institutes of Health, Bethesda, Maryland
Steven Kleinman, M.D. Clinical Professor, Department of Pathology, Univer-

sity of British Columbia, Vancouver, British Columbia, Canada
Jeanne V. Linden, M.D., M.P.H. Director, Blood and Tissue Resources, New
York State Department of Health, Wadsworth Center, Empire State Plaza, Albany,
New York
Anne B. McDonald, M.D. Fellow, Transfusion Medicine, New England Dea-

coness Hospital, Boston, Massachusetts
Jay E. Menitove, M.D. Executive Director and Medical Director, Community

Blood Center of Greater Kansas City, Kansas City, Missouri
David E. Nevalainen, Ph.D., MT(ASCP), CQA(ASQ) Quality Consultant,

Health Care, Baileys Harbor, Wisconsin
Mark A. Popovsky, M.D. Executive Director and Chief Medical Officer,

American Red Cross Blood ServicesNew England Region; Associate Clinical
Professor of Pathology, Harvard Medical School; and Beth Israel Hospital,
Boston, Massachusetts
Alfred M. Prince, M.D. Member/Head, Laboratory of Virology, Lindsley F.

Kimball Research Institute, New York Blood Center, New York, New York
Maria Rios, Ph.D. Manager and Research Scientist, Scientific and Technolog-
ical Services Department, New York Blood Center, New York, New York
Hillary V. Schaeffler Director, Standards and International Affairs, American

Association of Blood Banks, Bethesda, Maryland
xiv Contributors
Laura L. Spinelli, M.D. Fellow in Hematopathology, Department of Pathology,

University of North Carolina, Chapel Hill, North Carolina
Christopher P. Stowell, M.D., Ph.D. Director, Blood Transfusion Service,

Department of Pathology, Harvard Medical School and Massachusetts General
Hospital, Boston, Massachusetts
Zbigniew M. Szczepiorkowski, M.D., Ph.D. Associate Director, Blood Trans-

fusion Service, Department of Pathology, Harvard Medical School and Massa-
chusetts General Hospital, Boston, Massachusetts
Gary E. Tegtmeier, Ph.D. Director, Viral Testing Laboratories, Community

Blood Center of Greater Kansas City, Kansas City, Missouri
Jonathan Trouern-Trend Epidemiology and Surveillance Program, American

Red Cross ARCNET, Farmington, Connecticut
Jay E. Valinsky, Ph.D. Vice President, Laboratory and Technical Services,

New York Blood Center, New York, New York
Iain J. Webb, M.D. Medical Director, Cell Manipulation and Gene Transfer
Laboratories, Dana-Farber Cancer Institute, and Instructor in Medicine, Harvard
Medical School, Boston, Massachusetts
Silvano Wendel, M.D. Medical Director, Blood Bank, Hospital Sirio Libanes,
So Paulo, Brazil
1
Impact of Blood Donor Screening
Procedures on Transfusion Safety
Steven Kleinman
University of British Columbia, Vancouver, British Columbia, Canada
I. INTRODUCTION
A major goal of transfusion medicine practice in the last decade has been to
reduce the risk of transfusion-transmitted infection to as low a level as possible.
In order to approach the desired level of zero risk from transfused allogeneic
blood, multiple layers of safety are needed. Methods utilized in attempting to
maximize safety from donated allogeneic units include donor-selection criteria,
donor-screening procedures, confidential unit exclusion (CUE) and telephone
callback procedures, laboratory testing, and modification of the blood unit after
collection, either by leukocyte removal or physicochemical procedures for patho-
gen inactivation.
This chapter will focus on donor selection and screening procedures, CUE
and callback procedures, product recall, and recipient notification in the event that
a potentially infectious unit has been transfused.
II. IMPLEMENTATION OF DONOR-SCREENING

PROCEDURES
When it is first suspected that an infectious agent may be transmitted by
transfusion, the initial response is to develop donor deferral policies using
epidemiological data derived from high-risk populations (1). After this initial step,
subsequent review and debate often leads to the nationwide implementation of a
1
2 Kleinman
particular donor-screening procedure following recommendations of the Food

and Drug Administration (FDA) and/or the American Association of Blood Banks
(AABB) (2,3). Because of the current regulatory, political, and legal climate, a
screening procedure is rarely rescinded even if there is no good evidence for
its effectiveness. Since screening procedures tend to become permanent, it is
important to carefully evaluate their expected utility by pilot studies in blood
donor populations prior to widespread implementation. If this is not done ade-
quately, the result will be the addition of procedures that do little or nothing to
increase transfusion safety but will result in deferral of safe donors. In recent
years, some procedures have been implemented without any pilot data, whereas
in other cases evaluations of potential new donor interview questions have been
performed (4,5).
For diseases for which there are no routine laboratory tests (e.g.,
Creutzfeldt-Jacob disease, Chagas disease, babesiosis), donor-screening proce-
dures are the only method available for increasing transfusion safety. In contrast,
for those infectious diseases for which routine blood donor testing is performed,
the importance of donor-screening procedures relates to their ability to identify
some potentially infectious donors during the infectious seronegative window
period (6).
While there are distinct advantages to eliminating donors with possible
risks of disease transmission, there are also negative consequences of this
approach. Because donor deferral criteria often have poor specificity, a large
number of safe donors will be deferred (6,7). While this dilemma has been well
recognized, there is no available algorithm to decide whether a given procedure
or screening question should be implemented. In reality, such decisions attempt
to balance the following factors: the extent of disease transmission, the severity
of the consequences of such disease to the recipient, the expected sensitivity
(effectiveness) of the procedure, and the number of donors lost by implementation
of the procedure.
III. METHODOLOGICAL APPROACHES FOR

EVALUATING EFFICACY
It would seem logical that better compliance with existing donor deferral criteria
or the implementation of new donor-screening questions would improve the
safety of the blood supply. However, this conclusion is based upon the often
unproven assumption that a deferral criterion has been demonstrated to enhance
transfusion safety. Most donor questions have been implemented without effec-
tive measurements of their yield. Therefore, if a deferral criterion is inappropriate
or outdated, more effective application of this criterion will not enhance transfu-
sion safety.
Blood Donor Screening and Transfusion Safety 3
The ideal way to study the efficacy of a new donor-screening procedure

would be to perform infectious disease screening on those donors deferred by the
procedure comparing their rate of seropositivity to that of acceptable donors.
To my knowledge, this approach has never been utilized for studying health
history questions, almost certainly due to logistic considerations. These include
obtaining an informed consent from the deferred donor for infectious disease
testing and performing a separate phlebotomy to obtain a venous sample from
individuals who are not eligible to donate a unit of blood.
A second approach for evaluating screening procedures is to measure the
rate of donor deferral following the implementation of new or modified donor-
screening procedures. In this approach, it is inferred that an increased number of
deferred donors indicates the interdiction of an increased number of infectious
donors. However, measured changes in the donor deferral rate may be misleading;
even questions with high specificity will result in the identification of many more
false-positive than true-positive donors due to very low prevalence of transfusion-
transmitted infections (7). Therefore, an increased deferral rate may actually be a
measure of the relative cost of the intervention to the blood donor collection
system (i.e., deferral of otherwise acceptable donors), rather than an indication of
increased safety.
A third approach has been used recently by investigators from the Retro-
virus Epidemiology Donor Study (REDS) (8,9). They mailed an anonymous
survey questionnaire to recent successful donors for the purpose of determining
the percentage of such donors who would admit to behavioral infectious disease risk
factors that they had previously denied at the time of their donation. While these data
establish the rate of failure to obtain accurate answers to screening questions (see
below), they do not provide information as to how many of these donors were
infected and/or capable of transmitting infection through their donated unit.
IV. MECHANISMS FOR SELECTING SAFE DONORS

AND SAFE UNITS FOR TRANSFUSION
Outlined in Table 1 are types of donor-screening and product-selection proce-
dures, exclusive of laboratory testing, implemented to enhance transfusion safety.
These procedures are discussed below with regard to theoretical rationale, as well
as past, present, and future applications. When available, data to support the use
of the procedures are given.
A. Selection by Donor Group or Donor Site

Because of the known spread of hepatitis between individuals inhabiting partic-
ular types of institutions, blood collection agencies do not collect blood at homes
4 Kleinman
Table 1 Blood Safety Procedures (Excluding Laboratory Testing)
Timing Procedure
Predonation 1. Exclusion of donor groups/donor sites

2. Elimination of donation incentives
3. Donor education
At donation site prior to donation 4. Self-exclusion in response to written material
5. Health history interview
6. Donor deferral registrya
7. Confidential unit exclusion (CUE)
Postdonation 8. Telephone call-back
9. Product retrieval
10. Recipient notification
aThis mechanism can be used either to disqualify the donor at the collection site or to discard the
collected unit prior to release from the blood center.
for the mentally retarded or prisons (10,11). In order to decrease the risk of human
immunodeficiency virus (HIV) transmission, blood collection agencies do not
collect from groups or institutions that are known to be comprised of male
homosexuals (12). Initially, these policies were instituted as primary protective
safety measures prior to the development of laboratory testing for the infectious
agents of hepatitis and acquired immunodeficiency syndrome (AIDS). However,
despite routine laboratory testing for these agents, these policies still remain
prudent in order to eliminate donors from settings where there is a high risk of
acute seronegative infection.
B. Impact of Donor Incentives

The primary example of a donor incentive associated with disease transmission
is the payment of cash in exchange for blood donation. Because of this, FDA, in
1978, instituted regulations that required the labeling of a blood unit as either
volunteer or paid (13), reading, in part, as follows: A paid donor is a person who
receives monetary payment for a blood donation. A volunteer donor is a person
who does not receive monetary payment for a blood donation. Benefits, such as
time off from work, membership in blood assurance programs, and cancellation
of nonreplacement fees that are not readily convertible to cash, do not constitute
monetary payment within the meaning of this paragraph. At about the same time
that FDA established its regulation, several states passed legislation virtually
prohibiting the use of commercial donor blood (13). The FDA regulations sparked
extensive discussions about the appropriate definition of a volunteer blood donor,
which are still unresolved 20 years later (1416).
These regulatory and statutory requirements were instituted in response to

data indicating that commercial blood donors were more likely than volunteer
blood donors to transmit hepatitis to recipients. This conclusion was supported by
several lines of evidence: higher rates of hepatitis B surface antigen (HBsAg)
positivity in commercial donors, higher rates of hepatitis B and non-A, non-B
hepatitis in recipients of paid versus volunteer donor blood, and, most defini-
tively, documentation in a well-studied cohort of transfusion recipients that
the elimination of commercial blood resulted in substantially fewer cases of
posttransfusion hepatitis (1720). An analysis of the transition from commercial
donor to volunteer blood in New Mexico indicated that virtually 100% of the
commercial donor population ceased donating with the removal of the cash
incentive (21), indicating that this incentive was apparently the sole motivator for
donation in this large group of commercial donors.
While paid blood doors are now very uncommon, the U.S. plasma industry
still depends on commercial donors. It is of interest that seropositivity rates
among this donor population for newly implemented blood screening tests
[human T-cell lymphotrophic virus, type I (HTLV-I) and hepatitis C virus (HCV)]
are markedly higher than among volunteer blood donors (22,23), mimicking the
situation identified in the 1970s with regard to commercial blood donors and
HBsAg testing. It thus appears that a significant percentage of commercial donors
have relatively high rates of infection with multiple infectious disease agents,
raising the possibility that rates for as-yet-undiscovered transmissible agents
would also be higher.
Although the increased risk of blood contamination from commercial
donors has been documented, it has also been recognized that it is not the actual
act of payment that makes commercial donors unsafe; rather, it is due to the
financial incentive attracting donors from a population (i.e., of a lower socio-
economic class) with a higher rate of infectious disease (24,25). Two recent
reports have indicated that populations of paid donors drawn from different
segments of society than were the commercial donors of the 1960s and 1970s
were no less safe than volunteer donors (26,27). In one study of carefully screened
repeat-paid plateletpheresis donors in the Midwest, the infectious disease marker
rate was equivalent to that of the same blood centers volunteer blood donor
population, suggesting that both groups have equivalent safety (27). However,
extrapolating these results to other populations of paid donors is not appropriate
given the use of special donor-selection criteria at the study blood center. These
criteria required a previous whole blood donation, an orientation session, and an
8-week waiting period before donors were accepted into the paid plateletpheresis
program (28).
An incentive would be expected to be detrimental to blood safety if it
resulted in the recruitment of a population that has a higher risk of transmitting
infectious disease than does the general blood donor population. If such an
6 Kleinman
unsafe population of donors was to be recruited, the combined application of

donor-screening procedures and laboratory testing would be unable to lower the
risk to that of the general donor population for two reasons. First, a significant
percentage of donors infected with HIV, human T-cell lymphotropic virus
(HTLV), or hepatitis C virus (HCV) will not admit to a recognized risk factor
even after careful postdonation interviews, thereby escaping detection by donor-
screening procedures (2933). The second factor relates to donors who would
have admitted to risk factors at the donation interview in the absence of incentives
but who would knowingly give untruthful answers to interview questions when
motivated by the desire to obtain the offered donor incentive (3436).
Currently there are limited data to assess whether the use of donor incen-
tives other than direct payment of cash (e.g., time off from work, gifts, a monetary
credit for blood nonreplacement fees, participation in a group assurance plan, or
working towards recognition as a multigallon donor) have an influence on blood
safety (37). REDS investigators conducted a small anonymous pilot mail survey,
which asked donors if they had received any one of 10 different incentives at the
time of their most recent donation and correlated these responses with behavioral
risk factors for infectious disease (38). They found that donors who reported
receiving a gift or extra time off from work had marginally higher levels of
behavioral risk; however, these were no higher than risk levels associated with
particular demographic characteristics such as 25- to 44-year age group, male
gender, or first-time donor status. In postdonation interviews of donors identified
as HIV seropositive, Centers for Disease Control and Prevention (CDC) investi-
gators established that the incentive of extra time off from work was a major
motivational factor for donation in 3.4% of such donors, and receipt of a gift was
a major motivator in 2.5% (39). However, they also found that the desire to help
the community was the primary motivation for donation for most of the HIV-
seropositive donors. A recent case report of an HIV seroconverting donor from
the Midwest established that this person continued to donate because of his
employers policy of granting 48 hours time off from work if a successful blood
donation was completed (40). In aggregate, these recent data suggest that the use
of donor incentives do not have a large impact on the safety of donated blood.
Nevertheless, they also suggest that further study of the impact of extra time off
and/or gifts on transfusion safety should be undertaken. Consequently, REDS has
recently completed a new large-scale donor survey that included detailed ques-
tions about donor incentives.
Several studies have suggested that incentives are more effective in attract-
ing blood donation by persons who have never previously donated than in
influencing repeat donations in previous donors (37). If this proves to be the case,
it has implications for how we think about the use of incentives. If an incentive
attracts a large percentage of first-time donors from a different population group
than the usual first-time donors, one might hypothesize that transfusion safety
could be compromised. However, if a particular incentive is used to solicit more

frequent donations from repeat donors (e.g., recognition of gallon donation
status), it may be less likely to be detrimental to transfusion safety. It has also
been suggested that an incentive might have a lesser impact on transfusion safety
if it is provided to all persons who present at the donation site, whether or not
they complete a successful donation (41).
The appropriateness of specific donor-recruitment techniques and the defi-
nition of acceptable donor incentives have been addressed in a recent policy
statement by the American Association of Blood Banks (AABB). The AABB
statement suggests criteria that individual blood centers can use to help determine
whether a specific incentive might have an adverse impact on transfusion safety.
The U.S. National Marrow Donor Registry and, internationally, the Council of
Europe have also recently clarified their definitions of voluntary donation (42,43).
C. Impact of HIV Test Seeking

An additional mechanism of safeguarding the blood supply by removing a rather
unique donor incentive for a particular high-risk population was the establishment
of alternative testing sites for HIV antibody in 1985. In order to prevent donation
for the purpose of obtaining HIV test results, federal and state governments
established alternative test sites at which free and anonymous anti-HIV testing
could be obtained (44). Data from such programs indicated that HIV-seropositive
rates were significantly higher than those found at blood centers (45). Although
it is uncertain how many alternative test site clients would have donated blood
if there had been no other way to obtain HIV test results, at least two studies have
indicated that some percentage would have done so (45,46).
The concern about HIV test seeking is still pertinent a decade after HIV
donor screening was first begun. Postdonation interviews have indicated that the
motivation for donation in 1529% of HIV-seropositive donors was to obtain their
HIV antibody test result (3436,47). Furthermore, of donors who responded to
the REDS anonymous mail survey in 1993, 6.0% indicated that they had donated
blood some time in the past to obtain an HIV antibody test result; moreover, 3.2%
stated that they had done so within the past 12 months (48). The implementation
of HIV-1 p24 antigen testing in 1996 raised further concerns that this more
sensitive HIV screening test might encourage an even greater amount of HIV test
seeking in prospective donors (the so-called magnet effect) (49). However, an
analysis of HIV seropositivity rates in donors following the first 6 months of HIV
p24 antigen testing showed no significant change when compared to a 6-month
interval prior to such testing, indicating that a magnet effect, if it occurred at all,
was not observed to affect transfusion safety (50).
It is clear that some blood donors are donating to receive the perceived
benefit of free and confidential HIV testing. Blood centers currently provide
8 Kleinman
information to blood donors about the inappropriateness of HIV test seeking, and
some centers specifically ask donors whether they are donating to get an HIV test.
Nevertheless, it appears that more effective ways are needed to educate prospec-
tive donors that donating to obtain HIV antibody test results may endanger
transfusion recipients.
D. Donor Education and Donor Self-Exclusion

In 1983, with the recognition that AIDS could be transmitted by blood transfu-
sion, specific groups were identified as being at high risk of AIDS (e.g., male
homosexuals with multiple partners and intravenous drug users who shared
needles) (1,6,51). Through the use of the media and through discussion with gay
community leaders, blood centers made extensive efforts to inform the public,
prior to arriving at the donation site, that individuals with such risk factors should
not donate blood. In some locations, this community education was augmented
by distribution of written information concerning donor eligibility at the time of
donor recruitment.
Blood donors were provided with written material at the collection site,
which included a description of risk factors (or risk groups) for AIDS and for
other transfusion-transmissible agents (52); donors were instructed that if they
belonged to a risk group or had a risk factor, they should not proceed with the
donation process.
The process of a donor evaluating his or her own eligibility has been termed
self-exclusion. Data obtained from 1983 through 1985 indicate that self-exclusion
was effective in decreasing the risk of AIDS transmission. A survey of gay men
showed that many who had donated prior to the recognition of the AIDS epidemic
subsequently ceased their donation behavior (53). Data from New York City,
Los Angeles, and San Francisco indicated that there were decreased donations
from the populations at highest risk for transmitting HIV (7,5456).
Since 1983, educational materials have been modified periodically to
reflect the latest knowledge of transfusion-transmissible disease and have been
required reading for each prospective donor at the donation site.
E. Health History Interview

1. General Considerations
History-based donor screening was originally established to reduce the risk of
transfusion-transmitted hepatitis. In 1983, with the recognition of the possible
transmission of HIV by transfusion, the health history interview assumed new
importance and became more complex and probing than it had previously
been (1,51). The trend of adding medically sophisticated questions and/or so-
cially sensitive behavioral questions to the donor interview has continued to the
present day.
Health history interview questions designed to defer donors with increased
risk of infectious disease transmission can be broadly classified into several
categories. Table 2 indicates the diseases for which the various categories of
questions have been applied based upon known epidemiological risk factors.
Several methods have been utilized to conduct health history interviews:
these include the oral approach, the self-administered written approach, and
the combined approach. In the oral approach, a donor interviewer asks the
donor all health history questions and records the responses on the donor card.
In the self-administered approach, the donor checks off yes or no answers on the
donor card after reading each question. In the combined approach, the donor
may self-administer some questions and orally respond to others, or the
donor may self-administer all questions and then respond a second time to
selected questions orally presented by the donor interviewer. Regardless of the
approach utilized, if the donor gives answers that may potentially affect his or her
eligibility, the donor interviewer needs to elicit further information and to
document this on the donor history form.
Until recently, there was little standardization of donor questioning in
different blood centers in the United States. In 1991, a consortium of California
blood centers reported on their 3-year effort to design and implement a uniform
medical donor history card (57). The process used to generate the card was one
of consensus building among the medical directors of the various blood centers
with the final questions being approved by the Center for Biologic Evaluation and
Research (CBER) of the FDA. Limitations of this approach were an inability to
pretest the questions in donor populations prior to their implementation and the
reliance on medical directors rather than social scientists or communication
Table 2 Health History Interview Questions
Category Disease prevented
Medical history of a specific disease AIDS, hepatitis, malaria, Chagas

disease, babesiosis, CJD
Medical symptoms compatible with a specific AIDS, bacteremia, viremia
disease
Blood exposure by needlestick injury or blood AIDS, hepatitis
transfusion
Medical treatment Creutzfeldt-Jakob disease
Sexual or drug use activities of donor or sexual AIDS, HTLV-I/II, hepatitis
partner(s)
Previous residence in or visit to endemic area Malaria
10 Kleinman
experts to design questions. More recently, the AABB developed an FDA-

approved uniform donor history card. This is preapproved by FDA for use by any
blood collection center in the United States, provided that it is used without
alteration of any of the existing questions (58). There are no data to evaluate
whether the use of this uniform card has resulted in enhanced transfusion safety.
As donor interview questions have become more standardized, it has
become apparent that uniformity is lacking in the standard operating procedures
used when a donor answers a screening question with a response that requires
further evaluation. Recently, two blood centers have reported that they have
standardized their response to over 240 donor eligibility conditions (59). Since
the ultimate decision as to whether to accept a particular donor rests on medical
judgment, uniformity in evaluating a donors medical history is inherently diffi-
cult. In problematic situations, it is sometimes easier for staff to accept rather than
reject a blood donor. Rejection may make the donor feel bad, may make the donor
anxious about his or her infectious disease status, and, in the case of directed
donors, may be perceived by the donor as a denial of a right to donate (60,61).
Obviously, it is incumbent upon nursing and medical staff to make the proper
safety decisions and not to be unduly influenced by these factors.
2. Review of Data
In addition to asking the right questions during the donor interview, it is also
necessary to obtain the right answers. Interviews with HIV- and HCV-seropositive
donors conducted after these donors have been notified of their test results show
that a high percentage of such donors will admit to a history of risky sexual
behavior or past intravenous drug use that they had denied prior to donating
(30,62). Recently, these data from seropositive donors have been complemented
by observations in a large geographically diverse donor population (9). In 1993,
REDS investigators mailed a 53-question optical scan format questionnaire to
50,162 allogeneic blood donors who successfully donated blood within the
previous 12 months at one of five participating REDS blood centers. Donors
were randomly sampled with the exception of oversampling of those younger
than 26 and those in minority ethnic and racial groups to compensate for their
lower proportional representation in the donor population. The questionnaire
contained items related to demographics, donation history, comprehension of
written donation literature, the use of CUE or callback procedures, sexual history,
injection drug use history, other history related to HIV risk, HIV test seeking,
donors knowledge about AIDS, and donors knowledge about donor eligibility
criteria. Multipart questions were used to determine the time intervals (corre-
sponding to intervals used in blood donor screening) in which events occurred:
these included ever, since 1977, within the past 12 months, and within the past
3 months. Questions pertaining to sexual behavior and injection drug use
were preceded by a short statement explaining their purpose and the need for
truthful answers and giving the respondent permission to not answer any objec-
tionable questions.
Of 50,162 donors sampled, 34,726 (69.2%) responded and 98.1% of the
respondents answered all of the risk questions. The data were analyzed by using
sampling weights to adjust for differential sampling and/or response rates among
different demographic groups. A total of 1.9% of respondents reported at least one
behavioral risk that should have resulted in donor deferral. In 0.4%, this risk
occurred in the prior 3 months, a time frame compatible with acute seronegative
window period infection. These data demonstrate a low level of behavioral risk
that was not eliminated by donor questioning, laboratory testing, or CUE and
callback procedures. Given the strong psychological denial donors may have
about HIV risk behaviors or illicit activities, there is probably an inherent
limitation to the sensitivity that can be achieved by behavioral questioning (34).
Nevertheless, these data indicate that continued efforts to improve the sensitivity
of behavioral screening appear to be warranted (63).
3. Practical Aspects
Table 3 lists some of the important parameters that can be examined to potentially
increase the quality of the answers given in the donor health history interview.
Because the health history interview involves questions of a delicate nature,
it is important to ensure that the interview is conducted in a private and
confidential fashion. This can be difficult in the blood mobile setting. Two studies
have documented that the majority of donors felt that privacy during the history
interview was adequate (34,64). A study conducted by the American Red Cross
with 4651 donors at a variety of blood drives found that 74% of donors perceived
their privacy to be adequate at the health history station; this increased to 94%
when mobile visual partitions (standing screens) were used (64). These privacy
screens appeared to make a useful contribution to the donors perception of
privacy. Auditory privacy (the inability to hear another donors health history
interview) was considered as good to excellent by 92% of donors and increased
slightly to 97% when a device was used to mask extraneous noise. The second
Table 3 Issues in Optimizing Donor Health

History Screening
1. Privacy and confidentiality of health history
2. Donors comprehension of written material
3. Adequacy of interpreters
4. Interviewer training and competency
5. Time constraints
12 Kleinman
study revealed that subsequent to donor notification, 31% of HIV-seropositive

donors stated they had felt a lack of privacy during the health history inter-
view, with 20% stating that they would have changed their answers to interview
questions if they had been in a more private situation (34). While the data from
these studies are somewhat encouraging, it is also disturbing that a small
percentage of routine donors and a greater percentage of HIV-seropositive donors
in these two studies stated that privacy was inadequate; such perceptions may
affect donor responses to sensitive sexual or drug use behavior questions during
the donor interview. Therefore, it remains important to continue to improve the
perception of privacy at the health history station.
One approach to the health history interview that could address donors
privacy concerns and thereby might result in eliciting more truthful risk behavior
information is the use of a computer interactive interview. In one study of this
technique in 272 donors, the subjects stated that they had a feeling of increased
privacy compared to the usual interview format (65). In addition, the rate of
deferrals for HIV-related symptoms or risk behaviors was significantly increased
when compared to the usual interview format. Unfortunately, these data must be
interpreted cautiously because of inherent limitations in the study design. Because
of FDA requirements, the group given the computer interview was also screened
by standard health history screening techniques. In addition, unlike the health
history interview, the computer interview was conducted anonymously, without
the donor providing any identifying information that could link their responses to
a permanent record.
Despite these preliminary data and initial enthusiasm for the computer
interactive approach, additional studies of computerized screening have not been
published (65a). This most likely reflects the difficulty in conducting such studies
in an FDA-regulated environment.
A prospective donors ability to understand predonation written materials
can often pose a problem. In the past, this material has often been written at
an educational level that required a high degree of reading comprehension.
More recently, attempts have been made to rewrite such material at a lower grade
level. A second solution to this problem has been to place less reliance on the use
of written material. To this end, the AABB and FDA have recommended that in
the case of HIV risk factors, a verbal discussion occur between the interviewer
and the donor (66,67). In addition to presenting HIV risk factor information to
donors with limited written comprehension skills, this verbal discussion also
serves the purpose of emphasizing the significance and importance of this
information to all donors.
Another potential problem occurs when the prospective donor does not
speak English well or at all. In this setting, it is important to have the question-
naire administered in the donors native language. Multilingual staff are useful in
this regard; otherwise, an interpreter will be needed to assist in the interview.
The use of an interpreter poses problems if the interpreter is either a relative or a

fellow employee. Since many HIV risk questions are of an extremely personal
nature, this creates a situation in which the donor may not give a truthful history
to the interpreter. Currently there is no consensus as to how to solve this problem;
one possible solution is not to accept such donors unless an independent inter-
preter not known by the donor can be used (6).
Factors such as the adequacy of training for donor room personnel, the level
of accreditation of such staff, and the level of staff competency have received
increased attention in the last several years as a result of blood centers need to
comply with the FDAs Good Manufacturing Practices (cGMP) regulations.
In the interview setting, a broad definition of competency would extend be-
yond the simple ability to follow SOP and would include the ability to build
rapport with the donor so as to be able to effectively solicit information during
the interview. It is plausible that the attitude of interviewers with regard to sexual
behavior may have an impact on their ability to elicit truthful sexual behavior
histories; however, there are no data in the blood donor setting to address
this issue.
The goal of maximizing donor throughput at the collection site, which is
important both for productivity and donor convenience, may also affect the
quality of the donor interview. Despite the pressures to process donors quickly, it
is important that the donor be given sufficient time to think carefully about
answers to interview question, particularly since it may be necessary to recall
behaviors from many years ago (68).
Some critics have charged that the donor interview should be renamed the
donor interrogation and that the regulatory emphasis on cGMP has led to the focus
on donors as suppliers of raw material rather than as human beings with a
complex motivational structure (61,69). As the health history has continued to get
longer and more obtrusive, concerns have been expressed that donors may not be
capable of recalling all pertinent significant events (68). A related criticism has
been that by asking donors questions about events in the remote past (since 1977),
donors become less focused on recent risk behaviors, which are more relevant to
the possibility of a donation occurring during the infectious seronegative window.
Suggestions have been made that repeat donors not undergo the same broad-based
health history interview that is required of first-time donors (61,69).
In summary, based upon current knowledge of logistical difficulties, each
blood collection organization must carefully examine its own operations and
devise procedures that maximize the ability to obtain correct answers from
donors. In addition, there is a national need for educational materials and the
health history questionnaire to be validated across all demographic strata of blood
donors using scientifically proven methods (9,65a). Input from behavioral scien-
tists will be needed to refine this validation approach and to design appropriate
research protocols.
14 Kleinman
V. DONOR SCREENING FOR SPECIFIC TRANSMISSIBLE

AGENTS OR CONDITIONS
A. Human Immunodeficiency Virus
1. Current Criteria
A full-scale revision of FDA recommendations for screening prospective donors
for prevention of HIV transmission was issued in April 1992 (66). These recom-
mendations state that prospective donors:
1. Must receive both oral and written information concerning HIV (AIDS)
risk factors and the potential for HIV transmission through donated
blood. This information should include informing the donor that in
early HIV infection a donor may be infectious and capable of transmit-
ting HIV despite a negative test for HIV antibody.
2. Should be given specific information as to how they can obtain an HIV
antibody test at a site other than the donor center.
3. Should be told that all units of donated blood will be tested for HIV
antibody, and, if positive, the donor will be notified of the test result
and his or her name placed on a donor deferral registry.
4. Should be advised not to donate if symptoms that may be compatible
with HIV infection are present. These symptoms include persistent
fevers, night sweats, unexplained weight loss, persistent cough or
shortness of breath, persistent diarrhea, swollen lymph nodes that
persist for longer than one month, presence of whitish oral lesions, or
presence of bluish purple spots on the skin or in the mouth.
5. Should be asked direct health history questions about behaviors that put
them at risk for HIV. Preferably these questions should be asked orally.
The specific information to be obtained from donors during the donor interview
includes answers to the following:
Have you ever had clinical or laboratory evidence of AIDS or HIV
infection?
(For Men) Have you had sex with another man even once since 1977?
Have you ever injected intravenous drugs?
Have you engaged in sex in exchange for money or drugs since 1977?
Have you received clotting factor concentrates for hemophilia or other
clotting disorders?
If a donor answers yes to any of the above questions, he or she is permanently
deferred as a blood donor.
The donor should also be asked questions regarding behaviors during the
previous 12-month interval. A yes answer to any of these questions will lead to a
temporary deferral which is removed 12 months after the last potential exposure.
These questions are:
In the past 12 months, have you had sex with a person who has HIV
infection or AIDS?
In the past 12 months, have you had sex with a person who currently or
previously used intravenous drugs?
(For Women) In the past 12 months, have you had sex with a man who has
had sex with another man (i.e., a man who is bisexual)?
In the past 12 months, have you had sex with a prostitute?
In the past 12 months, have you had sex with a person receiving clotting
factor concentrates?
In the past 12 months, have you had syphilis or gonorrhea?
In the past 12 months, have you had a blood transfusion?
In the past 12 months, have you had an accidental needlestick injury or a
blood splash to a mucous membrane or to nonintact skin?
At the conclusion of the medical history, donors must sign a consent that
specifically states that they understand that they should not donate blood if they
are at risk for HIV infection.
Two additional changes to HIV screening procedures have been imple-
mented since these comprehensive recommendations in 1992. In 1995, FDA
recommended that individuals who are inmates of correctional institutions and
individuals who have been incarcerated for more than 72 consecutive hours
during the previous 12 months be deferred for 12 months from their last date of
incarceration (70). FDA stated that this information could be provided to donors
through the predonation written material and did not indicate that this question
was required as part of the health history interview. In 1996, FDA recommended
that three questions be added to the direct questions on high-risk behavior to
exclude donors who are at increased risk for HIV-1 group O infection (71). One of
these questions relates to birth in or residence in Cameroon or surrounding West
African countries where HIV-1 group O infection has been identified. The others
refer to blood transfusion or medical treatment received in those countries and to
sexual contact with anyone who was born in or lived in those countries since
1977. These questions are similar in format to the type of geographic exclusion
questions for HIV-2 exposure that were made obsolete by the introduction of
HIV-2 antibody screening in 1992 (72).
2. Rationale for the Current Criteria

When donor-screening procedures for preventing AIDS were first implemented
in the early to mid-1980s, most blood centers in the United States did not ask
donors direct questions about sexual orientation or behavior (i.e., a question from
16 Kleinman
the nurse to the donor such as Have you ever had sex with another male?) (6).
At the time there were no available data to indicate that more explicit questioning
might result in deferral of an increased number of high-risk persons; indeed, one
hypothesis was that direct questioning about a persons sexual orientation might
lead to untruthful answers because of concerns about maintaining the confidenti-
ality of information (6,12). A more direct approach also posed the possibility of
embarrassing or offending potential blood donors.
In the late 1980s, the donor interview process was reevaluated based upon
information obtained from interviews with HIV-seropositive donors and the
recognition of a much more open societal attitude towards discussing issues of
sexual orientation and sexual behavior. A retrospective study demonstrated that
asking donors direct oral questions about sexual behavior resulted in a fivefold
increase in HIV-related deferrals from 1.46 to 7.31 donors per month in one blood
center (73). A second blood center conducted a survey concerning donors
attitudes toward being asked such explicit sexual behavior questions. Of 1204
regular blood donors who responded, over 90% endorsed the use of these
questions, while only 1% felt that the questions might cause them to stop donating
(73). Subsequently, a prospectively designed study was performed in two U.S.
blood centers to address the effectiveness of donor-screening procedures, includ-
ing revised donor information brochures, the use of an AIDS information video,
and the use of a set of behavior-oriented direct questions asked orally by the
health historian. This study strongly suggested that the use of direct, behavior-
oriented oral questions led to a statistically significant increase in donors deferred
for HIV risk behavior (63). A second study, performed by the American Red
Cross, compared deferral rates in three groups, each of approximately 4000
donors, using direct behavior-oriented questions, indirect comprehension ques-
tions, and more generalized written interview questions; this study reached a
similar conclusion with regard to the benefit of direct oral questioning (74).
The current FDA and AABB recommendations suggest, but do not require, that
direct questions be asked orally (66,67).
With the implementation of anti-HIV testing in 1985, anti-HIV-positive
blood donors were identified, notified of their test results, and interviewed.
These interviews revealed that although many HIV-seropositive donors were
men who had sex with other men, they did not consider themselves to be at
risk for HIV infection (75). It became apparent that a person might respond
very differently if asked whether he were a member of a specific group (i.e.,
a homosexual man) as opposed to being asked if he had ever engaged in a
specific behavior (i.e., having sex with another male). The FDA therefore recom-
mended in September 1985 that any male who has had sex with another
male since 1977 should not donate blood (76). The recommendation emphasized
that individuals who had had only a single homosexual experience should refrain
from donating.
Three additional studies, summarized in Table 4, have examined the reasons

why HIV-seropositive donors with known risk factors donated blood (3436).
The CDC Multicenter Study included 512 donors (304 of whom had HIV risk
factors) from 20 blood centers; the American Red Cross (ARC) Multicenter Study
included 80 donors with HIV risk factors from three geographic areas; and the
National Institutes of Health (NIH) study included 98 donors with HIV risk
factors from Washington, DC. These studies show a similar profile of motivations
for donation by HIV-seropositive donors. One major reason for donating was the
donors self-assessment that they were not at risk for HIV infection; CDC data
indicated that many of these donors were in psychological denial about their HIV
risk. A second motivation for donation was social or peer pressure to donate; in
the CDC Study this pressure came from an employer or fellow employees, rather
than from the blood center, in 75% of cases. Another group of donors was at least
partially motivated to donate in order to obtain their HIV antibody test results.
Some persons who donated indicated that they misunderstood the accuracy of the
HIV antibody tests; some donors concluded that their donations were safe because
they had previously tested HIV antibody-negative, while other donors felt they
would not be jeopardizing recipients since, if they were infectious, the HIV
antibody test would inevitably be positive and their blood would not be utilized.
Data from these studies contributed to a revision of donor questioning to include
more specific HIV risk questions; these changes are reflected in current FDA
recommendations and in the AABBs uniform donor history card (58,66).
A preponderance of data has demonstrated that the theoretical possibility of
long-term persistent HIV infection in the absence of detectable HIV antibody,
HIV antigen, or clinical symptoms does not exist or is exceedingly rate (7781).
Table 4 Motivations for Donation by HIV-1 AntibodyPositive Blood Donors
Motivation (%)
CDC ARC
Reasons multicentera multicenter NIH
Denial of risk NA 61 26
Social pressure 27 29 15
Wanted to get test results 15 29 26
Previous HIV-seronegative test 9 NA NA
Positive test would stop blood from being transfused 10 NA 14
Misunderstood the pamphlet 7 NA 6
Other 32 NA 13
NA = Data not available in this format.
aSome donors gave more than one reason.
18 Kleinman
These data have demonstrated that seroconversion for HIV is highly likely to
occur within 6 months of HIV exposure (81) and that HIV nucleic acid cannot be
detected in HIV antibody-negative individuals from high-risk groups who are at
risk for latent HIV infection. Thus, HIV risk behaviors that can be defined as
ending at a specific point in time (i.e., sex with a particular person who
demonstrated HIV risk behaviors, an accidental exposure to blood, a blood
transfusion), should only defer a prospective donor until serological testing can
definitively prove that the individual is free of HIV infection. In order to allow
for outliers and to therefore be absolutely certain that the required time for
seroconversion has elapsed, a 12-month deferral interval for these behaviors has
been selected.
Other behaviors, such as past intravenous drug use or male-to-male sex
even once since 1977, still require a lifetime deferral. The unproven rationale for
a permanent deferral for a remote past history of male-to-male sex is that this
behavior represents a lifestyle choice, which may not be limited to a defined time
interval. To my knowledge, there are no data to support this contention and it
seems scientifically inappropriate to have differing deferral intervals for potential
homosexual and heterosexual exposure to HIV. With regard to intravenous drug
users, a permanent deferral seems appropriate give the higher seroprevalence of
most known infectious agents in this population and the concern that a past
injection drug user is more likely to be a carrier of other unidentified transfusion
transmissible agents.
As heterosexual transmission of HIV infection becomes more common in
the United States (82), it is to be anticipated that such a shift will be reflected in
the demographics of HIV-seropositive blood donors. Thus far, however, no
dramatic changes have occurred. In a CDC study of HIV-seropositive donors from
20 blood centers, the number of seropositive donors who acquired their HIV
infection by heterosexual exposure did not increase significantly when 1990
1991 data were compared to data from 19881989 (30). American Red Cross data
indicate that the rate of HIV-seropositive female donors has remained constant
from 1988 to 1992; however, a slightly decreasing rate of HIV-seropositive male
donors has resulted in an increased percentage of females among all HIV-
seropositive donors (83).
The risk of heterosexual spread of HIV has been addressed by several items
in the 1992 FDA recommendations (66). Several potential heterosexual risk
exposures are a cause for 12-month deferrals. These include sex with an intrave-
nous drug user, sex with a prostitute, sex with a bisexual man, and sex with a
person using clotting factor concentrates. In the early years of the AIDS epidemic,
donors who had immigrated to the United States from a country (e.g., Haiti and
sub-Saharan Africa countries) in which heterosexual activity played a major role
in transmission of HIV infection were permanently deferred (1,6). Questions
pertaining to geographic exclusion were continued until 1992 for sub-Saharan
Africa as a safeguard against transmitting HIV-2, an HIV strain more prevalent

in this region. With the implementation of HIV-2 antibody testing, deferral
policies based on geographic factors were no longer necessary and were discon-
tinued until 1996 when similar concerns about the inability of HIV screening tests
to detect HIV-1 group O infection prompted their reinstatement for Cameroon and
adjoining countries (71,72,84). In 1992, FDA added a history of syphilis or
gonorrhea to the list of 12-month deferrals (66). The rationale for this policy was
the assumption that acquisition of either of these sexually transmitted diseases
would indicate that a prospective donor had an increased risk of acquisition of
HIV through heterosexual activity.
A recent study evaluated the impact of a proposed change in deferral criteria
for heterosexual exposure to HIV (30). This study analyzed the responses of
HIV-seropositive donors with regard to their number of sexual partners in the last
year and during their lifetime; furthermore, similar response data were cited from
surveys of noninfected donors. The authors concluded that the implementation of
their proposed deferral criteria would be very nonspecific, leading to deferral
of high percentages of noninfected donors while providing only a marginal
increase in blood safety.
It is well established that HIV is not transmitted by casual contact; there-
fore, persons who have had close contact with individuals who are anti-HIV-
positive or who have AIDS should not be deferred as blood donors (85). If a
person admits to living in the same household as an HIV-infected person, the
potential for sexual or body fluid contact with that individual should be deter-
mined; if such contact has not occurred, the donor should not be deferred.
3. Hearsay Evidence of HIV Risk

On occasion, donor interviewers may acquire information that is difficult to
evaluate because it has been obtained from a third party, such as a spouse, a sex
partner, or a fellow worker (hearsay evidence) (6). For example, a third party may
assert that a particular donor is at risk for HIV even though the donor has
responded that he is not at risk. In this situation, the blood center will need to
decide whether to use the collected unit and whether to defer the donor from
future donations. The AABB has recommended that blood centers have written
procedures indicating how such information will be assessed for validity (86).
In such a case, the blood center is faced with fulfilling two obligations: protecting
the safety of a potential transfusion recipient and avoiding falsely placing a donor
on a computerized deferral list based on a third-party accusation. One possible
approach to this dilemma is for a blood center physician or a senior nursing staff
member to reinterview the donor. It can be explained to the donor that additional
information has prompted the need to reassess the donors eligibility. The physi-
cian can emphasize the importance of truthful history in protecting recipient
20 Kleinman
safety and can review HIV risk factor information with the donor. The physician
can obtain answers to explicit HIV risk questions from the donor. If the donor
admits to risk, he can be deferred. On the other hand, if the donor denies all HIV
risk, the physician will need to make a decision as to whether to believe the donor
or the third party. Unfortunately, there are no precise guidelines as to how such a
decision should be made. My belief is that when the third-party evidence is not
convincing and when the donor insists they have been truthful, the correct
procedure is to advise the donor that he is eligible to make future donations.
The donor will be reinterviewed and retested for HIV antibody on subsequent
donations, thereby providing some degree of confirmation of their safety.
One blood center reported their one-year experience with a protocol for
contacting the donor after third-party information was received (87). Eleven such
reports were received, 10 donors were interviewed, and 7 donors were determined
to be acceptable for future donation. The investigators stated that all contacted
donors were open, cooperative, and understanding of the blood centers position.
They concluded that the policy of confronting donors with third-party information
was effective in resolving these difficult donor suitability decisions.
4. Surveillance and Monitoring

At least three areas of continued monitoring are necessary to ensure that HIV-
related donor deferral criteria are optimally effective for preventing potentially
HIV-infected persons from donating; these are continued observation of HIV-
seropositive donor demographics, continued assessment of the risk factors and
motivation for donation of HIV-infected donors, and evaluation of whether donor
history questions are successful in eliminating persons with known HIV risk
factors. The first two of these items are undergoing continued assessment by
the multicenter CDC HIV Seropositive Donor Study. The third item, the effec-
tiveness of donor history questions, is being addressed by another anonymous
survey recently conducted by the NHLBI-sponsored Retrovirus Donor Epidemi-
ology Study.
B. Hepatitis
Federal guidelines for preventing transfusion-transmitted hepatitis were estab-
lished decades ago in the Code of Federal Regulations (CFR) (3). The current
regulations require the following deferral policies: (a) donors with a history of
viral hepatitis are permanently deferred; (b) donors with a history of close contact
with someone who has viral hepatitis are deferred for 12 months following their
last potential exposure; and (c) donors who have received a blood transfusion are
deferred for 12 months. [Although the CFR indicates a 6-month deferral for these
categories, subsequent FDA memos to blood establishments changed this deferral
to 12 months (88). This change was based upon establishing consistency with
HIV deferral time intervals as well as the observed lag period to HCV seroconver-
sion following HCV exposure using first-generation HCV tests (89)]. In addition,
a person testing positive for HBsAg or known to have previously tested positive
for HBsAg is also permanently deferred as a blood donor.
Recently, FDA clarified that a history of viral hepatitis applies only to
clinical disease and that deferral is not required when the donors history is based
solely on a positive serological test result (i.e., anti-HBc or anti-HBs) that
indicates past exposure to HBV (90). Furthermore, the permanent deferral re-
quirement for a history of viral hepatitis no longer applies to an episode of viral
hepatitis occurring before the age of 10. (91). This recent change in FDA policy
is based upon epidemiological evidence that viral hepatitis in childhood, occur-
ring prior to the onset of sexual activity, is almost exclusively due to infection
with the hepatitis A virus (HAV) (92). Since HAV infection does not induce a
chronic carrier state, it is evident that persons who acquired HAV in the remote
past do not pose a hazard to transfusion recipients (93). Such reasoning raises the
issue of whether donors with a history of HAV at any age should be acceptable
as blood donors (94,95). Although medical knowledge supports the safety of
donations from these individuals, such a policy would be of limited value since
it would be difficult to prove definitively that a particular past episode of viral
hepatitis was due to HAV. Because of this problem in establishing accurate
diagnosis, it is still required that donors who present with a stated history of HAV
above age 10 be deferred.
In the past, blood collection agencies deferred all donors with a history of
jaundice (96). Currently the policy at most blood centers is to question the donor
concerning the etiology of the jaundice and to accept those donors whose history
of jaundice is related to the neonatal period or to obstructive biliary tract disease.
Other donor deferral criteria for potential hepatitis risk are based on the
known parenteral routes of spread of HBV and HCV. Persons who have ever
injected intravenous drugs by needle have long been deferred as blood donors due
to hepatitis risk; the importance of this criterion has been reemphasized with the
advent of the HIV epidemic.
Donors with a history of tattoos are deferred for 12 months because of the
possibility of hepatitis spread by contaminated needles (2). Donors who have had
needle exposure through ear piercing, skin piercing, acupuncture, or electrolysis
should also be considered for a possible 12-month deferral (2,6). In such cases it
should be evaluated whether these procedures were performed utilizing dispos-
able sterile needles as occurs in a professional setting; if so, deferral is not
necessary. If it cannot be verified that sterile technique was followed, then
consideration should be given to deferring such donors for 12 months.
Donors receiving hepatitis B immunoglobulin (HBIG) should be deferred
for 12 months following such administration because of the underlying hepatitis
22 Kleinman
B exposure risk. On the other hand, donors receiving hepatitis B vaccine do not
need to be deferred unless the vaccine was given for a recent exposure to HBV.
According to FDA regulations, donors who have had close contact with a
patient with acute viral hepatitis should be deferred for 12 months (88). Clearly,
sexual contact is included in this definition; it is less clear how to define
nonsexual close contact. One commonly used definition is the sharing of house-
hold, kitchen, or toilet facilities, as would occur with living in the same household
(6). In the case of HBV, this definition appears reasonable, in that it has been
demonstrated that HBV can rarely be transmitted from an acutely infected patient
to a household contact, probably through nonsexual contact with body fluids (97).
Data for nonsexual household transmission of HCV are more equivocal, and it is
unclear whether nonsexual contacts of such patients pose a risk to recipients (98).
Nevertheless, FDA requires that such prospective donors must be deferred.
On the other hand, persons who occasionally eat a meal or visit a patient who
has viral hepatitis may have little hepatitis risk; medical judgment should be
utilized to determine whether a particular donor with this type of history should
be deferred.
FDA requirements do not formally apply to sexual or close contacts of
asymptomatic or symptomatic chronic carriers of HBV and HCV. Given the high
rates of sexual and body fluid transmission of HBV, the same 12-month deferral
criterion should be applied to contacts of chronic HBV carriers. With regard to
HCV, data suggest that sexual transmission occurs with a low enough frequency
to make it safe to accept sexual partners of HCV carriers (99). Deferral policies
for close, nonsexual contacts of HCV carriers are more problematic given the
scarcity of reproducible clinical data; current data do not support the need to defer
such contacts (98).
C. Parasitic Diseases
Transfusion-transmitted malaria is common in some parts of the world but is rare
in the United States, occurring at an estimated rate of 0.25 cases per million
donated units for 19721988 (100,101). Policies for preventing such transmis-
sions rely on donor questioning during the health history interview.
The deferral criteria for malaria risk were revised by FDA in 1994 (102).
Travelers to a malaria endemic area are deferred for one year after their return to
the United States (provided they have not had malarial symptoms), whereas
immigrants from or residents of malarial endemic countries are deferred for 3
years after their departure from the endemic country. Donors with a history of
malaria are deferred for 3 years after becoming asymptomatic. Previous deferral
criteria in place from 1974 to 1994 had required that travelers to a malaria
endemic area be deferred for 3 years if they had received antimalarial prophylaxis
and for 6 months if they had not, because it was believed that prophylaxis might
extend the latency period (3,101). The revised deferral criteria no longer use the
receipt of chemoprophylaxis as a determinant of the length of deferral.
Travelers to malarial endemic areas are deferred for one year because, if
infected, nonimmune persons will develop symptoms within this time frame.
The 3-year deferral period for immigrants from such areas is based on the premise
that such individuals may have partial tolerance to malarial parasites, thereby
resulting in the delay of malarial symptoms in an infected donor beyond one year
(3,100,101). The 3-year deferral for persons with a history of malaria is based on
the general consensus that Plasmodium falciparum organisms will be cleared
within 2 years and that P. vivax or P. ovale will be cleared within 3 years of the
resolution of symptoms in the vast majority of cases. This latter policy represents
a compromise between prevention of transmission and acceptable levels of donor
deferral, in that it is well known that a P. malariae chronic carrier state may persist
for decades, resulting in transfusion transmission many years after the resolution
of symptoms (103,104).
The CDC recently reported three cases of transfusion-transmitted malaria,
two of which were fatal (104a). In these two cases, the donors did not accurately
report their malaria risk information during the donor interview. These cases have
prompted an FDA review of malarial donor interview questions.
Chagas disease has been commonly transmitted by transfusion in Latin
America, but has only rarely been transmitted in this fashion in the United States
(105). Research studies conducted in the United States have attempted to identify
risk factors that might be associated with previous exposure to the causative
organism, Trypanosoma cruzi; these risk factors include birth in a country
endemic for Chagas disease, extended time spent in an endemic area, and lower
socioeconomic level (106). One institution with a large Hispanic donor popula-
tion has included Chagas risk questioning as part of its donor health history
interview and has performed investigational testing for T. cruzi antibody on
donors with affirmative answers prior to releasing blood for transfusion (5).
However, because of the low risk of documented transfusion-transmitted T. cruzi
infection in North America (four cases) and the low specificity of these proposed
questions, AABB-required health history questioning is confined to asking donors
whether they have ever had Chagas disease (2).
Babesiosis is a rare, malaria-like illness caused by a protozoan parasite,
Babesia microti, which invades human erythrocytes. Human transmission occurs
as a result of tick bites, most commonly in the summer months in particular
regions of the United States. Transfusion-transmitted babesiosis, although rare,
has been reported as a consequence of blood donation by asymptomatic carriers
(107,108). Strategies for eliminating transfusion risk by restricting donations
from endemic areas in the summer months or by questioning donors as to a
history of tick bites have been considered but have been judged to be largely
ineffective, although some centers nonetheless do avoid blood collection in highly
24 Kleinman
endemic areas during the summer (108,109). Currently, AABB-required health

history questioning is confined to asking donors whether they have ever had
babesiosis (110).
D. Creutzfeldt-Jacob Disease
Creutzfeldt-Jacob disease (CJD) is a rare, fatal, degenerative neurological disease
with a long asymptomatic latent period (111). The etiological agent is thought by
most experts to be a prion, although some favor a viral etiology (112). CJD has
been transmitted from human to human by the transplantation of dura mater, the
injection of pituitary-derived human growth hormone, and the reuse of EEG
electrodes; a single case of transmission by a corneal transplant has also been
reported (113). There are no reported cases of transmission by blood transfusion,
and epidemiological studies have not provided any evidence for such transmis-
sion (114,115). Nevertheless, because of the long incubation phase of the disease
(as demonstrated from growth hormone transmissions) and the inability of
conventional sterilization methods to inactivate the organism, the theoretical risk
of CJD transmission from asymptomatic donors to recipients of blood compo-
nents or plasma derivatives cannot be ruled out (116). Policies designed to prevent
this theoretical transfusion transmission of CJD have been adopted with regard to
donor deferral, product recall, and, in some cases, recipient notification (117120).
The initial concern for blood transfusion safety arose from the fact that
some persons who received therapeutic injections of cadaver-derived pituitary
human growth hormone developed CJD, thus raising the possibility that other
recipients of this hormonal product may have become asymptomatic chronic
carriers of the CJD agent (121,122). To eliminate possible risk to transfusion
recipients, FDA issued a 1987 recommendation that all prospective donors
be asked if they have ever received human growth hormone (117,118,122).
If the growth hormone was pituitary derived (i.e., the injection was given before
the availability of recombinant growth hormone in 1985), the donor was indefi-
nitely deferred.
In 1995 and 1996, FDA issued additional recommendations to further lower
the theoretical risk of transfusion-transmitted CJD by deferral of donors and recall
of previously acceptable donations (119,120). These recommendations included
the addition of questions about the receipt of a dura mater transplant and a family
history of CJD in a blood relative. Donors who respond affirmatively are
indefinitely deferred. A 9-month experience with CJD questioning at one large
blood center demonstrated that approximately one in 10,000 donors were deferred
for possible CJD risk, with 83% of the deferrals due to the donor having one blood
relative with CJD (123).
Since 90% of CJD cases are sporadic rather than familial, it should be
expected that a family history of one blood relative with CJD would be most
likely due to the sporadic occurrence of the disease. Therefore, in 1996, FDA
policy for product recall for CJD was revised to indicate that product recall of
previously transfused single donor product and manufactured pooled plasma
derivatives was not necessary if a donor gave a history of only one family member
with CJD. Rather, the criteria for product recall would require two blood relatives
with CJD or genetic testing that was diagnostic of the familial variant (120).
In 1998, based on a review of international data, FDA amended its product recall
guidance to exclude recall of plasma-derivative products for donors with a history
of classical CJD or risk factors for classical CJD (120a).
E. New Variant Creutzfeldt-Jacob Disease

New variant Creutzfeldt-Jacob disease (nvCJD) is a fatal, degenerative neurolog-
ical disease newly discovered in the United Kingdom in 1996 (120b). To date,
approximately 12 cases occur annually in the United Kingdom and only two cases
have been detected outside the United Kingdom (120c). The etiological agent of
nvCJD is the same agent that causes bovine spongiform encephalopathy (BSE)
(120d). The spread of the agent from cattle to humans and the detection of the
nvCJD prion in lymphoid tissue has raised concerns that nvCJD could be trans-
mitted by blood transfusion (120c120e). No cases of transfusion-transmitted
nvCJD have been reported, but the observation period has been too short to draw
any firm conclusions. In 1998 and 1999, FDA advisory committees and public
policy committees in Canada debated proposals to defer donors who had traveled
to the United Kingdom as a means of reducing the theoretical risk of transmission
of nvCJD through transfusion (120f). In June 1999, FDA announced a donor
deferral policy calling for deferral of donors who had spent more than 6 months
in the United Kingdom between 1981 and 1996.
F. Bacterial Disease
Another potential fatal complication of transfusion is bacterial infection (124).
Bacteria can be introduced into the donor unit at the time of collection if the donor
is bacteremic, if the skin site is not properly decontaminated, if there is an
undetected abscess adjacent to the phelbotomy site, or through introduction of a
small skin plug in the phlebotomy needle (124126). Examination of the skin at
the venipuncture site is conducted before phlebotomy, and strict requirements for
assuring the sterility of the site are adhered to by the phlebotomist (2,127). Blood
is not collected from persons who are febrile at the time of donation, who state
that they do not feel well, or who are taking systemic antibiotics (2). In the past,
blood centers have temporarily deferred donors with a history of dental proce-
dures for up to 72 hours because of the high frequency of postprocedure
bacteremia. These criteria have been liberalized with the recognition that many
26 Kleinman
other types of trauma to mucous membranes not evaluated at the time of donation
may also cause bacteremia (128). It has been recommended that donors need only
be deferred for 2472 hours after particular traumatic types of dental procedures,
such as root canals and tooth extractions; some blood centers no longer inquire
about recent dental procedures, and no such question is present on the AABB
uniform donor history card (58,124).
Unlike other infectious diseases, the risk of bacterial infection applies also
to candidates for autologous donation, as a result of the ability of gram-negative
rods to multiply at refrigerator temperatures and secrete endotoxin into the blood
bag (129). For this reason, autologous donors who are taking antibiotics or who
give a history of recent or concurrent medical procedures are evaluated for the
possibility of bacteremia and deferred accordingly.
G. Other Infectious Disease Considerations

Donors are asked whether they have had a previous history of syphilis or
gonorrhea in the past 12 months as a means of decreasing the risk of HIV
transmission. There is no requirement to ask donors about other types of venereal
disease such as herpes simplex, genital warts, or chlamydia.
Donors who have received live, attenuated viral vaccines are deferred for
2 weeks (except for rubeola and varicella, which are 4-week deferrals). Donors
receiving toxoid or killed viral vaccines should not be deferred (2).
Over the last decade, experience with transfusion-transmitted viruses such
as HIV and HCV has clearly demonstrated that in the absence of laboratory
testing, asymptomatic donors with chronic, latent viral infection may endanger
recipients (1,6). Concern for this type of circumstance may lead to deferral of
donors with unusual histories (i.e., chronic fatigue syndrome in quiescence,
treated Dengue fever), despite the fact that there is no definitive evidence of their
potential infectivity.
H. Donors with a History of Malignancy

Donors with a history of malignancy pose a theoretical risk to recipients;
however, no cases of transfusion-transmitted malignancy have been reported.
Since many transfusion recipients are immunosuppressed, it may be theoretically
possible that malignant cells circulating in a donors blood could engraft and
multiply in a recipient, provided there was a sufficient degree of genetic match-
ing. In order to decrease the possibility of this occurrence, donors are questioned
about a history of cancer. In most blood centers, a donor with a history of a solid
organ tumor will be deferred and will be eligible to donate only if he or she has
been symptom-free and considered to be clinically cured for a defined time
period, usually 5 or 10 years. Donors with a history of hematological malignancy
are still permanently deferred. Donors with specific malignancies that have been
fully excised are not deferred (e.g., basal cell cancer of skin, cervical carcinoma
in situ), since the tumor is known to be low grade and not capable of hematoge-
nous spread.
VI. ADDITIONAL MECHANISMS FOR SELECTING

SAFE DONORS OR SAFE UNITS
A. Donor Deferral Registries
Donor deferral registries began as an effort to decrease the risk of transfusion-
associated hepatitis at a time when other, more specific methods were not
available. More recently their use has been extended to include donors who might
transmit HIV and other infections (3,130,131).
These registries are computer, microfiche, or manual files of names and
identifying information from donors who have been deferred for specified rea-
sons. Individual blood centers have established inhouse (local) donor deferral
registries to comply with a section of the Code of Federal Regulations that states
that persons who have ever tested positive for HBsAg or anti-HIV should no
longer be accepted as blood donors (3,131). The rationale of donor deferral
registries is that individuals with a previous positive test result may revert to
a negative test result at a later date despite the fact that they could still trans-
mit disease. This situation could occur because of either a biological phenomenon
or a testing error. Donor deferral registries also include the names of donors
who have previously volunteered information (i.e., history of intravenous drug
use, hepatitis, or high-risk AIDS activity) that should have permanently excluded
them as blood donors. Donors who volunteer such information at one visit
may be tempted to withhold the information at a subsequent visit, especially
if their self-assessment is that the history ought not to have precluded them as
blood donors.
The donor deferral registry can be used in one of two ways. In most blood
centers in recent years, the donors name is checked against the deferral registry
prior to donation; no blood is collected if the donor is listed on the registry.
This method is highly desirable in that it avoids the unnecessary phlebotomy of
ineligible donors, and it prevents the potential for an error leading to the release
of the ineligible donor unit. The alternate method of complying with the require-
ments for a donor deferral registry was to collect the unit from the donor and to
subsequently, at a central location, check the donor deferral registry prior to
placing the unit into inventory. The unit was quarantined and destroyed if the
donors name was found on the registry.
Since donors may not always donate to the same blood collection agency,
some states have created statewide donor deferral registries that include the
28 Kleinman
names of donors deferred at any blood collection agency within the state (130).
The American Red Cross uses a national system in which donors who are entered
into specific categories of the deferral registry at an individual Red Cross region
also have their names included in a national registry (130,131). Because of donor
confidentiality concerns, donor deferral registries that extend beyond an individ-
ual blood center do not list the reason for donor entry into the registry.
Over the past 10 years, the size and complexity of donor deferral registries
have increased enormously. For example, the national component of the Ameri-
can Red Cross donor deferral registry contains over 20 separate deferral catego-
ries and over 300,000 donor names (130,131). It has been estimated that there
may be up to 3 million entries in donor deferral registries throughout the United
States (130). The problems associated with managing such expanded deferral
registries have also increased dramatically, creating a major source of difficulty
in adhering to FDA regulations. These regulations and blood center standard
operating procedures require that if a donor is listed in certain permanent deferral
categories on a donor deferral registry, additional donations from that individual
should not be distributed for transfusion. Difficulties in complying with this
requirement have resulted in numerous product recalls of subsequent units
donated by individuals who should have been placed on donor deferral registries
according to standard operating procedures. To my knowledge, there have been
no documented instances of these subsequently transfused units (which tested
infectious disease negative) causing adverse outcomes in recipients. These oper-
ational data suggest that donor deferral registries may not contribute significantly
to the enhancement of transfusion safety, given the other layers of safety that are
built into the system. A study to evaluate the usefulness of a regional donor
deferral registry demonstrated that 0.41% of donors who donated to a hospital-
based blood bank were later placed in that institutions donor deferral registry and
yet also donated to a regional blood center in the same geographic location (132).
These data can be interpreted to indicate that if donor deferral registries are
believed to be of value, their extension to regional registries may offer some
additional benefit.
Problems encountered in managing donor deferral registries include diffi-
culty in obtaining accurate information to uniquely identify a donor, the existence
of multiple records for the same donor due to conflicting information obtained on
separate donations, the need to move donors from one category of the registry to
another, and the fact that some information leads to permanent deferral while
other data only result in deferral if the phenomenon [e.g., a positive anti-hepatitis
B core (anti-HBc) test] occurs on two occasions (130).
Maintenance and continuous use of national or statewide donor deferral
registries is time-consuming, logistically complex, and expensive. Unfortunately
it has remained difficult to assess whether such massive efforts afford significant
increases in safety to the blood transfusion recipient.
B. Confidential Unit Exclusion

The procedure known as confidential unit exclusion (CUE) was introduced at
many blood centers following a 1986 FDA recommendation stating that at the
time of donation donors should be offered a procedure by which they could
designate confidentially whether or not their blood should be transfused to others
(133). The rationale for CUE was to provide an opportunity for those donors who
felt pressured to donate (from peers, fellow employees, employers, etc.) at a
workplace or community blood drive to indicate that their blood should not be
transfused. The CUE procedure works as follows: after the health history inter-
view but prior to donation of the unit, each blood donor indicates whether his or
her unit should or should not be used; if a donor neglects to make any indication,
the unit will not be used.
Numerous small studies have attempted to evaluate the sensitivity, speci-
ficity, or predictive value of the CUE procedure (134141). Assessments of the
efficacy of CUE have focused primarily on the frequency of its use by HIV-
seropositive donors; these data have been somewhat conflicting and have led to
differing conclusions about the value of the CUE procedure. Furthermore, those
studies in which follow-up interviews were performed with donors who chose the
CUE option have consistently established that the majority of donors selecting
this option did so either as a result of misunderstanding or error (138142). These
latter findings have prompted suggestions that the CUE procedure be modified to
make it more understandable to blood donors, thereby improving its specificity
(135,138). Several modifications to the CUE procedure have occurred since 1986.
Many centers have switched from the use of a manual ballot separated from the
donor history card to the use of a peel-off barricaded sticker that the donor affixes
to the history card. There have also been simplifications in the wording of the
CUE form, clearer written and oral instructions as to the purpose of CUE, and
clearer instructions as to how to correctly complete the process.
In some blood centers, donors who have chosen CUE are placed on the
donor deferral registry, whereas at other centers they are not. In some centers
donors are recontacted and provided with a mechanism such as a donor interview
to reestablish their ability to donate (143).
The rate of discard of blood as a result of the CUE procedure has decreased
from 7 per thousand donations in the early years (134,144,145) to 24 per
thousand in more contemporary reports (141,144,145). It is probable that some
of this decrease has resulted from improvements in the CUE procedure. Other
factors that may also have contributed are fewer persons with HIV risk presenting
at the donation site or a greater number of deferrals of such individuals during
the health history interview.
Recently, two large studies have attempted to reevaluate whether CUE is a
useful procedure (144,145). In addition to looking at whether HIV-seropositive
30 Kleinman
donors excluded their blood from transfusion, these two studies were able
to analyze whether donors who were demonstrated to seroconvert for HIV
(i.e., HIV-seronegative donation followed by HIV-seropositive donation) used
the CUE option on their preseroconversion sample. This is the most relevant
group for evaluation of CUEs effect on transfusion safety, since, unlike HIV-
seropositive units, these units would otherwise be transfused to recipients if the
CUE procedure were not in place (137,144). Investigators from the CDC com-
piled data on the use of CUE at the time of the preseroconversion donation of 322
HIV seroconverting donors at 40 blood centers from January 1987 through
December 1990 (144). They found that 3.4% of such donations were excluded;
however, 9 of the 11 exclusions came from one center (New York Blood Center),
which has consistently shown results for the CUE procedure that differ from the
remainder of the United States (54,135). If these data are eliminated, the data from
the remaining 39 centers indicate that 2 of 246, or 0.8%, of HIV-seroconverting
donors used this option. REDS analyzed the CUE process in 1.5 million donations
made at five blood centers in 1991 and 1992. These investigators found that 8%
of 169 HIV-seropositive donations were excluded and subsequently found that
6% (2 of 33) of HIV-seroconverting donors used this option on their pre-
seroconversion unit (145,146). Given the very low rate of HIV transmission by
transfusion and the low rate at which such seroconverting donors may choose the
CUE option, it has been estimated that this procedure may interdict one additional
HIV infectious unit per 7.14 million blood donations (146).
Despite its projected ineffectiveness in decreasing HIV risk from transfu-
sion, many still advocate retaining CUE as part of the donation process based on
other sources of data. Several studies have demonstrated that donors who use
CUE have higher rates of seropositivity for many infectious disease markers
(140,144,145,147) and that some individuals who use CUE admit to HIV risk
factors (139,142). The previously described anonymous mailed REDS survey
correlated the donors stated use of CUE (or telephone callback) with admitted
behavioral risk factors for infectious disease (9). This study found that all
behavioral risks, with the exception of injection drug use and history of transfu-
sion in the past year, were reported statistically significantly more frequently in
donors who used CUE. In aggregate, the relative prevalence of any reported risk
behavior in donors who used CUE was 7.6-fold greater than in those who did not
and was 9.4-fold greater for risk behaviors in the prior 3 months. However, CUE
was not used by the majority of donors who reported behavioral risk (low
sensitivity) and was used more frequently in donors without behavioral risk
(low specificity).
In 1992, FDA analyzed the then available data on CUE sensitivity and
specificity and stated that the CUE procedure was no longer mandatory, and its
use was left to the discretion of each individual blood center (66). Currently some
blood centers who previously used the CUE procedure have discontinued it.
In summary, CUE has poor sensitivity and specificity but nevertheless may
prevent the transfusion of a very small number of units that cause infection
in recipients.
C. Telephone Callback and Postdonation

Information Reports
A further safeguard intended to increase transfusion safety is the mechanism of
telephone callback (1). At the time of blood donation, donors receive instructions
that they may call the blood center sometime after the donation to report
additional pertinent medical history information. Telephone callbacks fall into
two prominent categories. Donors may call back to report the development of an
acute illness, such as fever, upper respiratory tract infection, or a gastrointestinal
disorder, occurring from several hours to several days postdonation. Donors may
also call back to indicate risk factors for HIV or other infectious diseases that
might not have been disclosed at the time of donation.
In a 2-year series (19931994) at one large American Red Cross center, 199
(40%) of 492 reports of postdonation information were obtained through the
mechanism of telephone callback (148). Postdonation information reports oc-
curred at a rate of approximately 0.033% of donations. Classification of the
postdonation reports revealed that 30% were due to acute illness, 25% to possible
hepatitis risk, and 29% to possible HIV risk.
A Canadian blood center evaluated the occurrence of bacteremia in blood
units drawn from donors who called back to report diarrhea, vomiting, fever,
severe pharyngitis, or diagnosed streptococcal pharyngitis within 7 days of
donation (149). All 187 components (from 99 such donors) cultured for bacteria
were negative. These negative data had 95% confidence intervals of 4% in red
cells and plasma and 18% in platelet concentrates. These limited data indicate
that the risks to recipients are small (if any) if such donated units are transfused
prior to receiving call back information from the donor.
An 8-year experience with telephone callback and its potential impact on
HIV transmission has been reported from San Francisco (150). The blood center
averaged 24 donor-initiated callbacks per year in which donors revealed HIV-
related risk factors that they had not disclosed at the time of donation. Despite
these risk factors, all such donated units tested HIV seronegative. On follow-up
investigation, however, it was found that two donors subsequently seroconverted
for HIV. These data suggest that telephone callback may have had a beneficial
effect on eliminating infectious HIV window period units; however, because
of the small number of observations, it is not possible to evaluate the extent of
this impact.
Blood centers must have established policies for managing postdonation
telephone callback information. Usually this information will be received by
32 Kleinman
nursing or other staff, who will complete required documentation. A useful

additional step is for a medical director to subsequently review these data and to
decide whether withdrawal of the product is necessary. One large blood center
has reported its experience that 52% of such postdonation reports did not require
product withdrawal (151). In general, if postdonation information caused the
donor to be deferred, blood centers will quarantine and destroy any components
that have not yet been shipped and will transmit such information, usually by
letter, to the hospital transfusion services that have received components from the
donation. The transfusion service will retrieve any such indate components and
either return them to the blood center or destroy them.
Postdonation information reports that affect the eligibility evaluation of the
donor must be reported by licensed blood establishments to the FDA (152).
In FDAs summary of error and accident reports for fiscal year 1996, 5620 reports
of postdonation information were received, 1137 (20%) of which resulted from
telephone callbacks from the donor to the blood center (153). These postdonation
reports accounted for 51% of the error and accident reports received by FDA from
blood establishments.
D. Market Withdrawal, Product Retrieval, and

Recipient Notification
Information obtained through telephone callback is usually received within days
of a given donation. In contrast to telephone callback, there are other circum-
stances leading to market withdrawal (e.g., a request to remove a blood compo-
nent from active inventory) in which information is not received by the blood
center until weeks, months, or even years after the donation of the unit. These
circumstances include a change in donor suitability requirements that retrospec-
tively disqualify previous donations from a given donor or a record review
leading to the discovery of improperly processed or released units.
Unfortunately, there are no clear-cut guidelines as to how the hospital
transfusion service should handle the varied and complex situations that may arise
following telephone callbackdriven or other market withdrawals. In such cases,
the hospital transfusion service medical director must decide whether to inform
the recipients physician that a recent blood transfusion has a higher than usual
risk of transmitting infection. The hospital transfusion service medical director
can use medical judgment to resolve each specific case but must follow a written
standard operating procedure for evaluating and documenting market withdrawal
information received from the blood supplier. The information provided by the
blood center to the transfusion service physician is often insufficient to either
adequately assess risk to the recipient or structure an informed notification
message to the recipients physician. In such instances, the transfusion service
physician should consider telephoning the blood center to obtain additional
pertinent donor information that may assist with the recipient notification deci-
sion or message. In some cases, it may be helpful to the transfusion center
physician if donor center personnel are able to recall the donor in a timely manner
and obtain further clarifying history or pertinent serological testing.
Consider the case of a unit transfused from a donor who later reports HIV
risk factors. If such information is forwarded to a recipient, it would be expected
to cause anxiety. Furthermore, since HIV antibody testing will not be able to
definitively resolve the situation for several weeks to several months, this anxiety
may not be easily alleviated. In my opinion, the best way to handle such a
situation is for a blood center physician or designee to recontact the donor who
volunteered the information and attempt to get a specific and accurate history and,
if possible, a follow-up HIV test. This may allow the hospital transfusion service
medical director to better estimate the likelihood of risk to the recipient. Using
this information, the decision as to when and whether to notify the recipients
physician and/or the recipient and the details of the specific notification message
can be determined.
The American Association of Blood Banks has issued an Association
Bulletin to assist transfusion services in developing policies for physician and
recipient notification in cases in which the transfused component was obtained
from a donor who later revealed risk factors for or had a clinical diagnosis of
Creutzfeldt-Jacob disease (154). The situation with regard to these recipients is
problematic, since the possibility of transfusion transmission is only theoretical
and there is no diagnostic testing available for the recipient subsequent to
notification. The AABB states that the ultimate responsibility for deciding
to withhold information from a recipient rests with the recipients physician and
not with the medical director of the transfusion service or with the institution.
The AABB suggests that each hospital convene a committee, such as an Ethics
Committee or an Institutional Review Board, that could develop criteria for
informing or not informing patients of possible transfusion exposure to CJD.
These criteria would allow physicians at the institution to individualize decisions
for particular patients.
VII. RECIPIENT NOTIFICATION (LOOKBACK) PROGRAMS

Lookback is the term used for a program of notifying large groups of recipients
of their risk of having been exposed to an infectious agent at the time of a previous
transfusion (155). Targeted lookback is a program to identify recipients of prior
units donated by specific donors who have subsequently been identified as
infected with a specific agent (e.g., HIV). It involves tracing of previous donation
and product shipment records by blood centers and tracing of transfusion records
by hospital transfusion services in order to identify the individuals who received
34 Kleinman
these specific blood components. These recipients, if still living and locatable,
can then be notified of their potential risk, usually by their physician. Laboratory
testing can be performed to determine whether infection occurred.
A generalized lookback program (also termed universal lookback) is a
program in which all recipients transfused within a designated time frame are
informed of their potential risk of infection from their transfusion. This program
can be implemented either through a public education campaign using the media
or communications to physicians or through direct mailings to recipients trans-
fused during the given time frame. This latter approach requires identification of
such recipients through hospital record searches.
The decision to perform targeted or generalized lookback programs for a
specific transmissible agent must consider multiple factors. Traditional public
health concerns, such as the yield and cost of the procedures, must be evaluated.
Public health authorities also need to consider lookback programs in the context
of more widespread screening programs to detect persons infected by routes other
than transfusion. In addition, an effective lookback program should ensure the
adequacy of diagnostic testing services, the adequacy of communication and
counseling resources for identified recipients, and medical follow-up for these
individuals. Another factor that may influence decisions concerning lookback
programs is the ethical premise that an individual has the right to know informa-
tion that might affect his or her future health. Additionally, liability concerns may
also influence institutional decision making.
In conjunction with considering these public health, ethical, and legal
issues, I believe that the answers to a critical set of medical questions are
important when formulating policies with regard to lookback for any transmissi-
ble agent. These questions are:
1. Is there evidence that the agent is transmissible by a particular blood
component?
2. Is there a diagnostic test to determine if infection has occurred in the
recipient?
3. Are there known modes of secondary transmission which can be
interrupted by the recipient having knowledge of his infection?
4. Are there medical means to monitor and assess the progression of the
disease in the recipient?
5. Is there treatment available?
These can be more concisely summarized into two broad considerations:
halting secondary spread of infection and the potential for intervening, or at least
monitoring, the natural history of the disease.
Testing of donated blood for anti-HIV, anti-HTLV, and anti-HCV has given
rise to the identification of blood donors who are HIV, HTLV, or HCV infected
and who have given previous, transfusable donations prior to the implementation
of blood screening. These donors may have been infectious at the time of their
previous donation; therefore, recipients of these previous donations are at in-
creased risk of acquiring these infections. Currently, FDA requires targeted
lookback to be performed for donors identified as HIV antibody positive and has
established a rule requiring the transfusing institution to assume ultimate respon-
sibility for notification of the recipient or next of kin (66,156). A more limited
time frame of lookback, i.e., units transfused in the prior 3 months, is required
for donors identified as HIV p24 antigen positive (157). Although FDA does not
require targeted lookback for HTLV infection, the AABB has established such a
requirement (2,158). After several years of debate, FDA issued guidance requiring
that a targeted lookback program be conducted for recipients of blood products
from HCV antibodypositive donors. The specific requirements as to which
recipients must be informed are complex because of the evolution of HCV
screening and confirmatory tests (158). All transfusion recipients prior to 1992
are also part of a generalized lookback programs for HCV under the direction of
the U.S. Public Health Service (159).
VIII. CONCLUSION
Optimal donor-screening procedures represent a balance between maximizing
safety for both recipient and donor and minimizing the unnecessary deferral of
safe blood donors. Given the importance of recipient safety, decisions about
donor-screening policies tend to favor the use of less specific procedures in an
effort to enhance transfusion safety. In some cases, such enhancement can be
demonstrated, while in other cases it is inferred from indirect evidence and in still
other cases data may be completely lacking. Donor-screening policies can be
evaluated scientifically, providing that the inherent limitations of the methodol-
ogy used in this type of operational research are recognized. Additional attention
to quality assurance of donor-screening procedures is warranted due to the
difficult logistical situations that exist in some blood collection settings. Decisions
regarding product retrieval and recipient notification are complex and require
consideration of multiple factors.
REFERENCES
1. Kleinman S. Donor screening procedures and their role in enhancing transfusion
safety. In: Smith D, Dodd RY, eds. Transfusion-Transmitted Infections. Chicago:
American Society of Clinical Pathologists, 1991:207.
2. Klein HG, ed. Standards for Blood Banks and Transfusion Services. 17th ed.
Bethesda, MD: American Association of Blood Banks, 1996.
3. U.S. Department of Health and Human Services, Food and Drug Administration.
36 Kleinman
The Code of Federal Regulations. Washington, DC: U.S. Government Printing

Office, 1996. (Revised annually.)
4. Grossman BJ, Kollins P, Lau PM, et al. Screening blood donors for gastrointestinal
illness: a strategy to eliminate carriers of Yersinia enterocolitica. Transfusion 1991,
31:500.
5. Appleman MP, Shulman IA, Saxena S, Kirchhoff LV. Use of a questionnaire to
identify potential blood donors at risk for infection with Trypansoma cruzi. Trans-
fusion 1993; 33:61.
6. Kleinman S. Donor selection and screening procedures. In: Nance SJ, ed. Blood
Safety, Current Challenges. Bethesda, MD: American Association of Blood Banks,
1992:169.
7. Kaplan HS, Kleinman SH. AIDS: blood donor studies and screening methods. In:
Barker LF, Dodd RY, eds. Infection, Immunity and Blood Transfusion. New York:
Alan R. Liss, 1985: 297.
8. Zuck TF, Thomson RA, Schreiber GB, Gilcher RO, Kleinman SH, et al. The
Retrovirus Epidemiology Donor Study (REDS): rationale and methods. Transfusion
1995; 35:944.
9. Williams AE, Thomson RA, Schreiber GB, et al. Estimates of infectious disease risk
factors in US blood donors. JAMA 1997; 277:967.
10. Schafer IA, Mosley JW. A study of viral hepatitis in a penal institution. Ann Intern
Med 1958; 149:1162.
11. Krugman S, Friedman H, Latimer C. Hepatitis A and B: serologic survey of various
population groups. Am J Med Sci 1978; 275:249.
12. American Association of Blood Banks, American Red Cross, Council of Commu-
nity Blood Centers: Joint statement on acquired immune deficiency syndrome
related to transfusion, January 13, 1983.
13. Barker LF. International forum: Which criteria must be fulfilled for a donation or a
donor to be considered voluntary? Vox Sang 1978; 34:363.
14. Morris JP. International forum: Which criteria must be fulfilled for a donation or a
door to be considered voluntary? Vox Sang 1978; 34:369.
15. Beal RW, van Aken VG. Gift or goods? Vox Sang 1992; 63:1.
16. American Association of Blood Banks: Donor Incentives. Association Bulletin 94-6,
October 1994.
17. Cherubin CE, Prince AM. Serum hepatitis specific antigen (SH) in commercial and
volunteer sources of blood. Transfusion 1971; 11:25.
18. Szmuness W, Prince AM, Brotman B, Hirsch RL. Hepatitis B antigen and antibody
in blood donors: an epidemiologic study. J Infect Dis 1973; 127:17.
19. Allen JG, Dawson D, Saymore WA, et al. Blood transfusion and serum hepatitis:
use of monochloroacetate as an antibacterial agent in plasma. Ann Surg 1959;
150:455.
20. Alter HJ, Holland PV, Purcell RH, et al. Post transfusion hepatitis after exclusion of
commercial and hepatitis B antigen positive donors. Ann Intern Med 1972; 77:691.
21. Surgenor DM, Cerveny JF. A study of the conversion from paid to altruistic blood
donors in New Mexico. Transfusion 1978; 18:54.
22. Canavaggio M, Leckie G, Allain JP, et al. The prevalence of antibody to HTLV-I/II
in United States plasma donors and in United States and French hemophiliacs.
Transfusion 1990; 30:780.
23. Dawson GJ, Lesniewski RR, Stewart JL, et al. Detection of antibodies to hepatitis
C virus in U.S. blood donors. J Clin Microbiol 1991; 29:551.
24. Mosley JW, Galambos JT. Viral hepatitis. In: Schiff L, ed. Disease of the Liver.
Philadelphia: JB Lippincott, 1975:500.
25. Kahn RA. Donor screening to prevent posttransfusion hepatitis. In: Keating LJ,
Silvergleid AJ, eds. Hepatitis. Washington, DC: American Association of Blood
Banks, 1981:99.
26. Taswell HA. Directed, paid and self donors. In: Clark GM, ed. Competition in Blood
Services. Arlington, VA: American Association of Blood Banks, 1987:137.
27. Strauss RG, Ludwig GA, Smith MV, et al. Concurrent comparison of the safety of
paid cytapheresis and volunteer whole-blood donors. Transfusion 1994; 34:116.
28. Huestis DW, Taswell HF. Donors and dollars (editorial). Transfusion 1994; 34:96.
29. Petersen LR, Doll LS. Human immunodeficiency virus type 1-infected blood
donors: epidemiologic, laboratory and donation characteristics. Transfusion 1991;
31:698.
30. Petersen LR, Doll LS, White CR, et al. Heterosexually acquired human im-
munodeficiency virus infection and the United States blood supply: considerations
for screening of potential blood donors. Transfusion 1993; 33:552.
31. Operskalski EA, Schiff ER, Kleinman SH, et al. Epidemiologic background of blood
donors with antibody to human T-cell lymphotropic virus. Transfusion 1989;
29:746.
32. Lee HH, Swanson P, Rosenblatt JD, et al. Relative prevalence and risk factors of
HTLV-I and HTLV-II infection in US blood donors. Lancet 1991; 337:1435.
33. Kolho EK, Krusius T. Risk factors for hepatitis C virus antibody positivity in blood
donors in a low-risk country. Vox Sang 1992; 63:192.
34. Doll LS, Petersen LR, White CR, et al. Human immunodeficiency virus type
1-infected blood donors: behavioral characteristics and reasons for donation. Trans-
fusion 1991; 31:704.
35. Leitman SF, Klein HG, Melpolder JJ, et al. Clinical implications of positive tests for
antibodies to human immunodeficiency virus type 1 in asymptomatic blood donors.
N Engl J Med 1989; 321:917.
36. Williams A, Kleinman S, Lamberson H, et al. Assessment of the demographic and
motivational characteristics of HIV seropositive blood donors. Paper presented at
the III International Conference on AIDS, Washington, DC, June 1987.
37. Mayo DJ. Evaluating donor recruitment strategies. Transfusion 1992; 32:797.
38. Williams AE, Watanabe K, Hansma D, et al. Prevalence and risk relationships of
blood donation incentives. Abstracts of FDA Workshop on incentives for volunteer
(nonrenumerated) donors. Bethesda, MD, September 1996.
39. Kleinman S. Donor incentives and motivation for donation: data from the CDC
sponsored multicenter study of HIV seropositive blood donors. Abstracts of FDA
Workshop on incentives for volunteer (nonrenumerated) donors. Bethesda, MD,
September 1996.
40. McFarland JG, Aster RH, Buggy BP. Paid time off for blood donationan incentive
to be less than truthful? Transfusion 1997; 37:447.
38 Kleinman
41. Kleinman SH, Shapiro A. The agreement to donate blood: Presenting information
to blood donors and obtaining consent. In: Smith DM, Carlson KB, eds. Current
Scientific/Ethical Dilemmas in Blood Banking. Arlington, VA: American Associa-
tion of Blood Banks, 1987:29.
42. Director of Recruitment Development, National Marrow Donor Program. Policy
regarding donor incentives at time of recruitment. Memorandum to donor center &
recruitment group coordinators and donor center medical directors, November 22,
1993.
43. Council of Europe. Guide to the Preparation Use and Quality Assurance of Blood
Components. Strasbourg: Council of Europe Press, 1992:12.
44. Mason JO. Alternative sites for screening blood for antibodies to AIDS virus. N Engl
J Med 1985; 313:1157.
45. Forstein M, Page PL, Coburn TJ. Alternative sites for screening blood for antibodies
to AIDS virus. N Engl J Med 1985; 313:1158.
46. Snyder AJ, Vergeront JM. Safeguarding the blood supply by providing opportunities
for anonymous HIV testing. N Engl J Med 1988; 319:374.
47. Lackritz EM, Kennedy MB, Doll LS, et al. Risk behaviors and test seeking among
HIV-1 positive blood donors. Transfusion 1995; 35(suppl):42S.
48. Williams AE, Thomson RA, Horton JA, Kleinman SH. Characterization of active
blood donors who recently donated blood primarily to receive an HIV test. Trans-
fusion 1995; 35(suppl):42S.
49. Korelitz JJ, Busch MP, Williams AE. Antigen testing for human immunodeficiency
virus (HIV) and the magnet effect: Will the benefit of a new HIV test be offset by
the numbers of higher-risk, test seeking donors attracted to blood centers? Transfu-
sion 1996; 36:203.
50. Stramer SL, Salemi BL, Sievert WS, et al. Evaluation of US blood donations testing
positive for HIV-1 p24 antigen. Transfusion 1996; 36(suppl):38S.
51. Dodd RY. Donor screening for HIV in the United States. In: Madhok R, Forbes CD,
Evatt BL, eds. Blood, Blood Products and AIDS. London: Chapman & Hall,
1987:143.
52. Director, Office of Biologics Research and Review, Food and Drug Administration.
Recommendations to decrease the risk of transmitting acquired immune deficiency
syndrome (AIDS) from blood donors. Memorandum to all establishments collecting
human blood for transfusion, March 24, 1983.
53. Seage III GR, Barry MA, Landers S, et al. Patterns of blood donations among
individuals at risk for AIDS, 1984. Am J Public Health 1988; 78:576.
54. Pindyk J, Waldman A, Zang E, et al. Measures to decrease the risk of acquired
immunodeficiency syndrome transmission by blood transfusion. Transfusion 1985;
25:3.
55. Perkins HA, Samson S, Busch MP. How well has self-exclusion worked? Transfu-
sion 1988; 28:601.
56. Busch MP, Young MJ, Samson SM, et al. Risk of human immunodeficiency virus
transmission by blood transfusions prior to the implementation of HIV antibody
screening in the San Francisco Bay Area. Transfusion 1991; 31:4.
57. Kolins J, Silvergleid AJ. Creating a uniform donor medical history questionnaire.
58. American Association of Blood Banks. Uniform donor history questionnaire. Asso-
ciation Bulletin #95-5, Bethesda, MD, June 2, 1995.
59. Sutor LJ, Sussman H, McCraw L, Fortenberry D. Regional standardization of the
blood donor medical history screening. Transfusion 1996; 36(suppl):88S.
60. Piliavan JA. Temporary deferral and donor return. Transfusion 1987; 27:199.
61. Sayers MH. Duties to donors. Transfusion 1992; 32:465.
62. Maclennan S, Barbara JA, Hewitt P, et al. Screening blood donations for HCV.
Lancet 1992; 339:131.
63. Mayo DJ, Rose AM, Matchett SE, et al. Screening potential blood donors at risk for
human immunodeficiency virus. Transfusion 1991; 31:466.
64. Kline L, Friedman LI, Dempesy D, et al. Assessment of blood donor privacy during
health history interviews. Transfusion 1996; 36:456.
65. Locke SE, Kowaloff HB, Hoff RG, et al. Computer-based interview for screening
blood donors for risk of HIV transmission. JAMA 1992; 268:1301.
65a. Kleinman S, Williams AE. Donor selection procedures: Is it possible to improve
them. Trans Med Rev 1998, 12:288.
66. Director, Center for Biologics Evaluation and Research. Recommendations for the
prevention of human immunodeficiency virus (HIV) transmission by blood and
blood products. Memorandum to all registered blood establishments, April 23, 1992.
67. American Association of Blood Banks. Statement and recommendations of the
American Association of Blood Banks regarding donor history questions. Memo-
randum to institutional members, July 29, 1989.
68. Mosley JW. Who should be our blood donors? Transfusion 1991; 31:684.
69. Westphal RG. Donors and the United States blood supply. Transfusion 1997;
37:237.
70. Director, Center for Biologics Evaluation and Research, Food and Drug Adminis-
tration. Recommendations for the deferral of current and recent inmates of correc-
tional institutions as donors of whole blood, blood components, source leukocytes,
and source plasma. Memorandum to all registered blood establishments, June 8,
1995.
tration. Interim recommendations for deferral of donors at increased risk for HIV-1
group O infection. Memorandum to all registered blood and plasma establishments,
December 11, 1996.
72. OBrien RT, George JR, Holberg SD. Human immunodeficiency virus type 2
infection in the United States: epidemiology, diagnosis, and public health im-
plications. JAMA 1992; 267:2775.
73. Silvergleid AJ, Leparc GF, Schmidt PJ. Impact of explicit questions about high-risk
activities on donor attitudes and donor referral patterns-results in two community
blood centers. Transfusion 1989; 29:362.
74. Gimble JG, Friedman LI. Effects of oral donor questioning about high-risk behav-
iors for human immunodeficiency virus infection. Transfusion 1992; 32:446.
75. Schorr JB, Berkowitz A, Cumming PD, et al. Prevalence of HTLV-III antibody in
American blood donors. N Engl J Med 1985; 313:384.
76. Director, Office of Biologics Research and Review, Food and Drug Administration.
Revised definition of high risk groups with respect to acquired immunodeficiency
40 Kleinman
syndrome (AIDS) transmission from blood and plasma donors. Memorandum to all
registered blood establishments, September 3, 1985.
77. Sheppard HW, Dondero D, Arnon J, Winkelstein Jr W. An evaluation of the
polymerase chain reaction in HIV-1 seronegative men. J AIDS 1991; 4:819.
78. Eble BE, Busch MP, Khayam-Bashi H, et al. Resolution of infection status of
HIV-seroindeterminate and high-risk seronegative individuals using PCR and virus-
culture: absence of persistent silent HIV-1 infection in a high-prevalence area.
79. Jackson JB. Human immunodeficiency virus (HIV) indeterminate Western blots and
latent HIV infection. Transfusion 1992; 32:497.
80. Centers for Disease Control and Prevention. Persistent lack of detectable HIV-1
antibody in a person with HIV infectionUtah, 1995. MMWR 1996; 45:181.
81. Busch MP, Satten GA. Time course of viremia and antibody seroconversion
following primary HIV infection: implications for management of exposed health
care workers. Am J Med 1997; 102(suppl 5B):117.
82. Brookmeyer R. Reconstruction and future trends of the AIDS epidemic in the United
States. Science 1991; 253:37.
83. U.S. Department of Health and Human Services, Public Health Service, Centers for
Disease Control and Prevention. National HIV Serosurveillance Summary. Results
Through 1992. Vol. 3, Atlanta, GA, 1993.
84. Schable C, Zekeng L, Pau CP, et al. Sensitivity of United States HIV antibody tests
for detection of HIV-1 group O infections. Lancet 1994; 334:1333.
85. Friedland GH, Saltzman BR, Rogers MF, et al. Lack of transmission of HTLV-
III/LAV infection to household contacts of patients with AIDS or AIDS related
complex with oral candidiasis. N Engl J Med 1986; 314:344.
86. American Association of Blood Banks. Management of hearsay information about
blood donors. Association Bulletin #93-4, Bethesda, MD, December 2, 1993.
87. Elliot D, Craig L, Price T. Resolving hearsay information received on a blood donor.
Transfusion 1995; 35(suppl):71S.
tration. Revised recommendations for testing whole blood, blood components,
source plasma and source leukocytes for antibody to hepatitis C virus encoded
antigen (anti-HCV). Memorandum to all registered blood establishments, April 23,
1992.
89. Alter HJ, Purcell RH, Shih JW, et al. Detection of antibody to hepatitis C virus in
prospective followed transfusion recipients with acute and chronic non-A, non-B
hepatitis. N Engl J Med 1989; 321:1494.
tration. Donor suitability related to laboratory testing for viral hepatitis and a history
of viral hepatitis. Memorandum to all registered blood establishments, December
22, 1993.
tration. Exemptions to permit persons with a history of viral hepatitis before the age
of eleven years to serve as donors of whole blood and plasma: alternative proce-
dures, 21 CFR 640.120. Memorandum to all registered blood establishments, April
23, 1992.
92. Trepo C. International forum: Should donors with a history of jaundice still be
rejected? Vox Sang 1981; 41:110.
93. Sheretz RJ, Russell BA, Reuman PD. Transmission of hepatitis A by transfusion of
blood products. Arch Intern Med 1984; 144:1579.
94. Alter HJ. Discussion: transfusion associated hepatitis. In: Polesky H, Walker RH,
eds. Safety in Transfusion Practices. Skokie, IL: College of American Pathologists,
1980:32.
95. Polesky HF, Hanson M. Tests for viral hepatitis markers in blood donors. In: Polesky
HF, Walker RH, eds. Safety in Transfusion Practices. Skokie, IL: College of
American Pathologists, 1980:17.
96. Aach RD. International forum: Should donors with a history of jaundice still be
rejected? Vox Sang 1981; 41:110.
97. Perillo RP, Gleb L, Campbell C, et al. Hepatitis B antigen, DNA polymerase activity,
and infection of household contacts with hepatitis B virus. Gastroenterology 1979;
76:1319.
98. Alter MJ. The detection, transmission, and outcome of hepatitis C virus infection.
Infect Agents Dis 1993; 2:155.
99. Seeff, LB, Alter HJ. Spousal transmission of the hepatitis C virus? Ann Intern Med
1994; 120:807.
100. Guerrero IC, Weniger BC, Schultz MG. Transfusion malaria in the United States,
19721981. Ann Intern Med 1983; 99:221.
101. Nahlen BL, Lobel HO, Cannon SE, Campbell CC. Reassessment of blood donor
selection criteria for United States travelers to malarious areas. Transfusion 1991;
31:798.
tration. Recommendations for deferral of donors for malaria risk. Memorandum to
all registered blood establishments, July 26, 1994.
103. Sazama K. Prevention of transfusion-transmitted malaria: Is it time to revisit the
standards? Transfusion 1991; 31:786.
104. Bruce-Chwatt LJ. International Forum: Which are the appropriate modifications of
existing regulations designed to prevent transmission of malaria by blood transfu-
sion, in view of the increasing frequency of travel to endemic areas? Vox Sang 1987;
52:138.
104a. Centers for Disease Control and Prevention. Transfusion-transmitted malariaMis-
souri and Pennsylvania MMWR 1999; 48:253.
105. Wendal S, Gonzaga AL. Chagas disease and blood transfusion: a new world
problem? Vox Sang 1993; 64:1.
106. Leiby DA, Yund J, Read EJ, et al. Risk factors for Trypanosoma cruzi infection in
seropositive blood donors. Transfusion 1996; 36(suppl):57S.
107. Mintz ED, Anderson JF, Cable RG, et al. Transfusion-transmitted babesiosis: a case
report from a new endemic area. Transfusion 1991; 31:365.
108. Gerber MA, Shapiro ED, Krause PJ, et al. The risk of acquiring Lyme disease or
babesiosis from a blood transfusion. J Infect Dis 1994; 170:231.
109. Popovsky MA, Lindberg LE, Syrek AL, Page PL. Prevalence of babesia antibody
in a selected blood donor population. Transfusion 1988; 28:59.
110. American Association of Blood Banks Bulletin. Recommendations regarding babe-
42 Kleinman
siosis and potential transmission of blood transfusion. American Association of

Blood Banks, Arlington, VA, July 29, 1989.
111. Brown P, Gibbs CJ, et al. Human spongiform encephalopathy: the NIH series of 300
cases of experimentally transmitted disease. Ann Neurol 1994; 44:513.
112. DeArmond SJ, Pruisner SB. Etiology and pathogenesis of prion diseases. Am J
Pathol 1995; 146:785.
113. Brown P, Preece MA, Will RG. Friendly fire in medicine: hormones, homografts,
and Creutzfeldt-Jacob disease. Lancet 1992; 340:24.
114. Esmonde TFG, Will RG, Slattery JM, et al. CJD and blood transfusion. Lancet 1993;
341:205.
115. Heye N, Hensen S, Muller N. CJD and blood transfusion. Lancet 1994; 343:298.
116. Brown P. Can Creutzfeldt-Jacob disease be transmitted by transfusion. Curr Opin
Hematol 1995; 2:472.
117. Director, Office of Biologics Research and Review, Food and Drug Administration;
Deferral of donors who have received human pituitary derived growth hormone.
Memorandum to all registered blood establishments, November 25, 1987.
118. Acting Director, Office of Blood Research and Review, Center for Biologics
Evaluation and Research, Food and Drug Administration. Deferral of blood and
plasma donors based on medications. Memorandum to all registered blood and
plasma establishments, July 28, 1993.
tration. Precautionary measures to further reduce the possible risk of transmission
of Creutzfeldt-Jacob disease by blood and blood products. Memorandum to all
registered blood and plasma establishments, August 8, 1995.
tration. Revised precautionary measures to further reduce the possible risk of
transmission of Creutzfeldt-Jacob disease by blood and blood products. Memoran-
dum to all registered blood and plasma establishments and all establishments
engaged in manufacturing plasma derivatives, December 11, 1996.
120a. Center for Biologics Evaluation and Research, Food and Drug Administration.
Change to the Guidance entitled Revised precautionary measures to further reduce
the possible risk of transmission of Creutzfeldt-Jacob disease by blood and blood
products. September 1998.
120b. Will RG, Ironside JW, Zeidler M, et al. A new variant of Creutzfeldt-Jacob disease
in the UK. Lancet 1996; 347:921925.
120c. Murphy MF. New variant Creutzfeldt-Jacob disease (nvCJD): the risk of transmis-
sion by blood transfusion and the potential benefit of leukocyte-reduction of blood
components. Transfusion Med Rev 1999; 13:7583.
120d. Hill AF, Debruslais M, Joiner S, et al. The same prion strain causes vCJD and BSE.
Nature 1997; 389:448450.
120e. Hill AF, Butterworth RJ, Joiner S, et al. Investigation of variant Creutzfeldt-Jacob
disease and other human prion diseases with tonsil biopsy samples. Lancet 1999;
353:183189.
120f. Hoey J, Giulivi A, Todkill AM. New variant Creutzfeldt-Jakob disease and the
blood supply: Is it time to face the music? (editorial). Can Med Assoc J 1998;
159(6):669670.
121. Fradkin JE, Schonberger LB, Mills JL, et al. CJD in pituitary growth hormone
recipients in the United States. JAMA 1991; 265:880.
122. Holland PV. Why a new standard to prevent Creutzfeld-Jacob disease? Transfusion
1988; 28:293.
123. Kessler D, Bianco C. Donor deferrals related to Creutzfeldt-Jacob disease. Transfu-
sion 1996; 36(suppl):57S.
124. Goldman M, Blajchman MA. Blood product-associated bacterial sepsis. Trans Med
Rev 1991; 5:73.
125. Blajchman MA, Ali AM. Bacteria in the blood supply: an overlooked issue in
transfusion medicine. In: Nance SJ, ed. Blood Safety: Current Challenges. Bethesda,
MD: American Association of Blood Banks, 1992: 213.
126. Anderson KC, Lew MA, Gorgone BC, et al. Transfusion-related sepsis after
prolonged platelet storage. Am J Med 1986; 81:405.
127. American Association of Blood Banks. Bacterial contamination of blood products.
Association Bulletin #96-6, Bethesda, MD, August 7, 1996.
128. Ness PM, Perkins HA. Transient bacteremia after dental procedures and other minor
manipulations. Transfusion 1980; 20:82.
129. Richards C, Kolins J, Trindade CD. Autologous transfusion-transmitted Yersinia
enterocolitica. JAMA 1992; 268:1541.
130. Sherwood WC. Donor deferral registries. Transfusion Med Rev 1993; VII:121.
131. Grossman BJ, Springer KM. Blood donor deferral registries: highlights of a confer-
ence. Transfusion 1992; 32:868.
132. Grewal ID, Domen RE, Hirschler NV. The value of shared donor deferral registries.
133. Department of Health and Human Services, Food and Drug Administration. Addi-
tional recommendations for reducing further the number of units of blood and
plasma donated for transfusion or for further manufacture by persons at increased
risk of HTLV-III/LAV infection. Memorandum to all registered blood establish-
ments, October 30, 1986.
134. Ciavetta J, Nusbacher J, Wall A. Donor self-exclusion patterns and human im-
munodeficiency virus antibody test results over a twelve month period. Transfusion
1989; 29:81.
135. Loiacono BR, Carter GR, Carter CS, et al. Efficacy of various methods of confiden-
tial unit exclusion in identifying potentially infectious blood donations. Transfusion
1989; 29:823.
136. Busch MP, Perkins HA, Holland PV, et al. Questionable efficacy of confidential unit
exclusion (letter). Transfusion 1990; 30:668.
137. Petersen LR, Busch MP. Confidential unit exclusion: How should it be evaluated?
138. Kean CA, Hsueh Y, Querin JJ, et al. A study of confidential unit exclusion.
139. Wolles S, Galel S. Value of confidential unit exclusion. Transfusion 1993;
33(suppl):80S.
140. Kessler D, Valinsky JE, Bianco C. Sensitivity and Specificity of confidential unit
exclusion (CUE)-Does it work? Transfusion 1993; 33(suppl):35S.
141. Menitove JE, Lewandowski C, Ashworth LW, et al. Confidential unit exclusion
44 Kleinman
process continues to identify donors with an increased frequency of HIV seroposi-

tivity. Transfusion 1991; 31:69S.
142. Kleinman S, Crawley P. An assessment of HIV related donor screening procedures.
143. Weitekamp LA, Meyer TL. Second chance for self deferring donors. Transfusion
1995; 35(suppl):72S.
144. Petersen LR, Lackritz E, Lewis WF, et al. The effectiveness of the confidential unit
exclusion option. Transfusion 1994; 34:865.
145. Korelitz JJ, Williams AE, Busch MP, et al. Demographic characteristics and
prevalence of serologic markers among donors who use the confidential unit
exclusion process: The Retrovirus Epidemiology Donor Study. Transfusion 1994;
34:870.
146. Kleinman SH, Busch MP, Korelitz JJ, Schreiber GB. The incidence/window period
model and its use to assess the risk of transfusion-transmitted HIV and HCV
infection. Trans Med Rev 1997; 11:155.
147. Nusbacher J, Chiavetta J, Naiman R, et al. Evaluation of a confidential method of
excluding blood donors exposed to human immunodeficiency virus: studies on
hepatitis and cytomegalovirus markers. Transfusion 1987; 27:207.
148. Kimball P, Popovsky MA. Analysis of post-donation information: need for im-
proved pre-donation education. Transfusion 1995; 35(suppl):34S.
149. Goldman M, Long A, Roy G, et al. Incidence of positive bacterial cultures after
donor call-back (letter). Transfusion 1996; 36:1035.
150. Samson SA, Edmiston RK, Busch MP, Perkins HA: How well has donor call-back
worked? Transfusion 1993; 33(suppl):35S.
151. Keelan LT, Grindon AJ. Units saved by investigation of post-donation information.
tration. Guidance regarding post donation information reports. Memorandum to all
registered blood and plasma establishments, December 10, 1993.
153. Center for Biologics Evaluation and Research, Division of Inspections and Surveil-
lance. Error and Accident Reports-Summary for FY-96, April, 1997.
154. American Association of Blood Banks. Guidance on notifying recipients of blood
components from a donor who subsequently was diagnosed as having CJD and/or
CJD risk factors. Association Bulletin 96-4, May 3, 1996.
155. Busch MP. Lets look at human immunodeficiency virus lookback before leaping
into hepatitis C virus lookback. Transfusion 1991; 31:655.
156. Department of Health and Human Services. Current good manufacturing prac-
tices for blood and blood components; notification of consignees receiving blood
and blood components at increased risk for transmitting HIV. Fed Reg 1996;
61:47413.
tration. Recommendations for donor screening with a licensed test for HIV-1 antigen
Memorandum to all registered blood and plasma establishments, August 8, 1995.
158. Center for Biologics Evaluation and Research. Food and Drug Administration.
Current Good Manufacturing Practice for Blood and Blood Components: (1) Quar-
antine and disposition of prior collections from donors with repeatedly reactive
screening tests for Hepatitis C virus (HCV); (2) Supplemental testing and the
notification of consignees and transfusion recipients of donor test results for anti-
body to HCV (anti-HCV). Draft Guidance for industry, Rockville, MD, June 1999.
159. U.S. Public Health Service. Briefing document on public health service options for
the identification of hepatitis C virus infection among prior transfusion recipients,
Washington, D.C., March 28, 1996.
2
Hemolytic Transfusion Reactions
Elizabeth J. Kicklighter and Harvey G. Klein

National Institutes of Health, Bethesda, Maryland
I. HISTORICAL PERSPECTIVES
The concept of blood transfusion began with the description of the circulation of
blood by William Harvey in 1628. By the late 1660s, Jean Denis in France and
Richard Lowery in England were performing the first documented blood transfu-
sions into humans. Both used animals as the source of blood (1). In view of what
is now known about antispecies antibodies, it is not surprising that the first
description of a hemolytic transfusion reaction (HTR) was recorded by Denis in
1668 after giving a second infusion of calfs blood to a patient. Some of the same
symptoms described with this reactiondiaphoresis, back pain, tachycardia,
dyspnea, gastrointestinal distress, and dark urineremain classic findings for the
severe HTRs seen today (2). Deniss patient died after a third transfusion attempt,
and blood transfusions were subsequently banned by the French and British
medical societies (3).
In 1818, James Blundell rekindled interest in blood transfusion for use
primarily in the treatment of postpartum hemorrhage. He was the first to transfuse
blood from one human to another and to discourage the use of animals as donors
(1). The mortality rate of his patients was extremely high. However, because
Blundells patients were very ill, the role of incompatible blood and hemolytic
reactions in their clinical course is difficult to evaluate. Nevertheless, indiscrim-
inate use of blood transfusions by others during this period resulted in unaccept-
able rates of severe hemolytic reactions (4).
The modern era of blood transfusion began with Karl Landsteiners dis-
cover of the ABO blood group system in 1900. Initially, few people understood
47
48 Kicklighter and Klein
the significance of Landsteiners reports. In 1907, Richard Weil and Reuben

Ottenburg were among the first to recognize that if red cells bearing foreign
ABO antigens are infused into a recipient with the corresponding antibodies
(isohemagglutinins), severe acute hemolysis could result. Ottenburg became the
first to perform ABO typing of patient and donor prior to transfusion (5).
Unfortunately, clinicians often refused to comply with requests for pretransfusion
laboratory specimens, giving Ottenburg many opportunities to observe the ex-
tremes of immune-mediated hemolysis that occur as a result of ABO incompati-
bility. In 1921, Unger hastened the acceptance of pretransfusion testing by
publishing a series of cases with serious acute HTRs that could have been avoided
by honoring pretransfusion testing demonstrating ABO incompatibility (6). Be-
cause few red cell antigens outside of the ABO system can be directly aggluti-
nated by human antibodies, no new blood group systems were found for more
than 25 years.
In 1945, the development of the antiglobulin test by Coombs et al. revolu-
tionized the field of immunohematology. The indirect and direct antiglobulin tests
(IAT and DAT, respectively) are the primary means by which antibodies to red
cell antigens are detected. The ability to detect antibody on red cells provided a
critical tool for recognizing immune-mediated HTRs. Of the 254 classified red
cell antigens and 23 established blood group systems (7), many were first detected
by an antiglobulin test performed on patients with HTRs or accelerated destruc-
tion of transfused red cells.
A better understanding of the immune response to foreign red cell antigens
and rigorous pretransfusion testing have made transfusion of blood remarkably
safe. Still, 0.53% of all transfusions result in some adverse event (8) (Table 1).
The majority of adverse effects are relatively innocuous. Of the life-threatening
reactions, the general public tends to focus on transfusion-transmitted infections.
Nevertheless, the risk of immune-mediated hemolysis often necessitates extensive
serological testing to evaluate compatibility between donor and recipient and
arguably may be the most important factor limiting the availability of blood.
Acute HTRs remain the most common cause of immediate life-threatening
complications associated with blood transfusion.
II. CLASSIFICATION OF TRANSFUSION-RELATED

HEMOLYSIS
An HTR is defined as the immune destruction of red cells mediated by an
antibody directed against the corresponding red cell antigen despite the best
efforts to provide compatible cells (9). HTRs are categorized as being acute or
delayed, as well as intravascular or extravascular. The manifestations, differential
diagnosis, and management can vary significantly among these groups.
Hemolytic Transfusion Reactions 49
Table 1 Classification of Transfusion Reactions

Acute
Immune Mediated
Acute hemolytic transfusion reaction (AHTR)
Transfusion-related acute lung injury
Febrile nonhemolytic transfusion reaction
Urticarial reaction
Anaphylactic
Non-Immune Mediated
Nonimmune hemolysis
Bacterial contamination
Volume overload
Metabolic
Embolic
Delayed
Immune Mediated
Delayed hemolytic transfusion reaction
(DHTR)
Posttransfusion purpura
Graft-versus-host disease
NonImmune Mediated
Transfusion-transmitted infection
Iron overload
However, not all incidents of transfusion-related hemolysis are HTRs, nor

are they all mediated through immunological pathways. Several nonimmune
mechanisms can result in transfusion-related hemolysis (2). In addition, not all
hemolysis seen during or after blood transfusion results from the blood infu-
sion itself. Hemolysis that is temporally related but not etiologically related to
transfusion can be termed a pseudo-HTR. Therefore, the history, careful
analysis of events, and laboratory evaluation of hemolysis during or after a
transfusion all play critical roles in the immediate and subsequent management
of suspected HTRs.
A. Immune-Mediated Versus
NonImmune-Mediated Hemolysis
An HTR occurs when a blood recipient has or subsequently develops antibodies
directed against antigens found on transfused red cells, resulting in accelerated
destruction of those cells. The recipient may mount an immune response if the
antigens on the infused cells are recognized as foreign. Antigenic stimulation may
occur as a result of blood transfusion, transplantation, or pregnancy. Pregnancy

presents a much lower immunogenic challenge than does blood transfusion,
probably because the number of foreign antigens is limited to those of the father
and because the number of cells entering the mothers circulation is often too
small to initiate a primary response (10). However, immune-mediated HTRs are
almost three times more common in females than in males because of prior
sensitization during pregnancy (9). The susceptibility to alloimmunization as a
result of transfusion appears to be the same in males and females. The risk of
forming an alloantibody to a (non-ABO) red cell antigen after the infusion of
one unit of red cells has been estimated to be 1% (11). The increased frequency
of HTRs reported with age appears simply to reflect the increased frequency of
transfusion with age (9).
There is a wide spectrum of potential immune responses to allogeneic red
cells. The range of outcomes includes immediate intravascular hemolysis; de-
layed extravascular hemolysis characterized by red cell alloantibodies, which
promote phagocytosis by macrophages; seroconversion without obvious hemoly-
sis; or, at the far end of the spectrum, no detectable immune response. Acute
intravascular hemolysis usually results from preformed complement-fixing anti-
bodies. The best known examples are the ABO naturally occurring
isohemagglutinins, which occur in the absence of exposure to allogeneic red cells.
Delayed hemolysis characteristically results from a secondary or anamnestic
response to a previously encountered antigen. Seroconversion in this context is
defined as the detection of red cell alloantibodies after antigenic stimulation.
Seroconversion without evidence of clinical hemolysis is characteristic of a
primary immune response. In some cases there may have been subclinical
hemolysis that resulted in clearing of the sensitizing red cells. No detectable
immune response is defined as an absence of red cell alloantibodies in individuals
who have had repeated exposure to red cell alloantigens (3).
To complicate matters further, the clinical significance of red cell al-
loantibodies varies widely. Antibodies to some red cell antigens never result in
hemolysis, but most antibodies vary significantly in their clinical activity and may
result in severe hemolysis in one patient while resulting in delayed, subclinical,
or no hemolysis in another patient. Anti-A and anti-B usually cause immediate,
life-threatening hemolysis, but case reports document patients without symptoms,
even after transfusion of an entire unit of ABO-incompatible blood. Antibodies
to red cell antigens in the Rh, Kell, Kidd, or Duffy systems may cause extremely
serious hemolysis in one patient but only shortened red cell survival with no
significant clinical symptoms in another (6). The detection of a red cell antibody
does not mean that hemolysis will occur. While in vitro hemolysis and a broad
thermal amplitude often suggest a clinically significant antibody, no single
laboratory test or combination of tests allows one to determine the specific in vivo
activity of an alloantibody. As a rule of thumb, antibodies that react in vitro at
temperatures of 30C and above should be respected, whereas most antibodies

that react only at colder temperatures, even when present in high titers, are
clinically unimportant.
The immunogenicity of red cell antigens varies widely. Immune responses
to the Rh(D) red cell antigen have been studied most carefully. The D antigen is
among the most immunogenic and clinically significant of the red cell allo-
antigens. Sixty to 80% of D-negative individuals will produce anti-D after
repeated immunizations (12). Anti-D is capable of causing severe HTRs. The dose
of D-positive cells required for primary immunization can vary from less than
1 mL to as much as 200 mL (12). The time of appearance of anti-D after primary
immunization varies among individuals. Low levels of anti-D may first be
detected at about 4 weeks (10), or anti-D may not be detected by standard
serologic techniques for up to 5 months after exposure to the D antigen (13).
There have been few systematic studies of immune responses to alloantigens
other than D. In general, D-negative subjects who form anti-D are capable of
forming other red cell alloantibodies, whereas those who do not form anti-D
seldom form any red cell alloantibodies (10).
In contrast to active immunization, which results in the formation of red
cell alloantibodies by the recipient, transfusion of plasma-containing components
can result in passive transfer of donor antibodies to the recipient. If the passively
acquired antibody is directed against an antigen on recipient red cells or other
transfused red cells in the recipient, accelerated destruction of the antigen-positive
cells could occur. Because donors are screened for the presence of antibodies to
common red cell antigens, these reactions are unusual. Two situations that might
result in passive hemolysis are (a) emergency transfusion of group O whole blood
to nongroup O patients, or (b) use of plasma from donors who have a red cell
alloantibody that is directed toward a low-incidence antigen not present on
routine screening cells.
While immune-mediated hemolysis is the most common mechanism re-
sponsible for severe HTRs, nonimmune-mediated destruction of red cells
must always be considered in the differential diagnosis of hemolytic reactions
(Table 2). Nonimmune-mediated hemolysis typically occurs before or during the
infusion process and is due to improper component preparation, storage, or
infusion. Some hemolysis may also be seen with the infusion of aged cells.
Nonimmune-mediated hemolysis can produce signs and symptoms that
mimic those of immune-mediated hemolysis, such as hemoglobinuria and ab-
sence of the expected increment in hemoglobin, but the most severe symptoms
and complications associated with immune-mediated hemolysis are usually ab-
sent (2). A notable exception is hemolysis resulting from bacterial contamination
of the blood unit, which often presents with fever and rigors that lead to shock
before the entire unit is transfused. Fortunately, with current donor screening,
Table 2 Mechanisms of Nonimmune Hemolysis of Donor Cells

Transfusion of aged cells
Thermal hemolysis (overheating, freezing)
Osmotic hemolysis (inadequate deglycerolization, administration with hypotonic
solutions or drugs)
Mechanical hemolysis (improper infusion devices, catheters, or needles)
Bacterial/parasitic contamination
Hemolysis due to congenital defects (G6PD deficiency, sickle trait)
aseptic methods of collection and storage, and meticulous handling of blood, the
frequency of such reactions appears to be low.
Other mechanisms of nonimmune-mediated hemolysis include thermal,
mechanical, osmotic, and toxic damage to red cells from improper storage or
handling (2). Either overheating or freezing without an appropriate cryopreserva-
tive can result in significant hemolysis. Appropriate thermal monitoring should
be performed at every stage of blood collection, transport, storage, and infusion.
Blood should never be stored in an unmonitored refrigerator or warmed with a
device that has not been certified for this purpose. Transfusion of previously
frozen but inadequately deglycerolized red cells can result in osmotic hemolysis.
Osmotic hemolysis can also result from a simultaneous infusion of red cells and
a hypotonic solution. Mechanical trauma can lyse transfused red cells, especially
during infusion. For this reason, only pumps and other equipment that have been
properly validated for infusion of blood and blood components should be used.
Nonimmune-mediated hemolysis may occur in vivo posttransfusion,
from congenital defects of transfused red cells, such as glucose-6-phosphate
dehydrogenase (G6PD) deficiency, or from contamination with malarial parasites.
The latter is common in the developing world, but rare in donors screened in the
United States. Nonimmune-mediated hemolysis can occur with the infusion of
both allogeneic and autologous blood products. Since there are no screening tests
for nonimmune-mediated hemolysis that might occur in red cell units during
storage, visual inspection of each unit for hemolysis prior to issue and prior to
infusion is essential.
B. Acute Versus Delayed Transfusion Reactions

Acute hemolytic transfusion reactions (AHTR) are due to the immune-mediated
destruction of red cells that occurs during or within hours of the infusion of blood.
An AHTR represents immune-mediated hemolysis due to an antibody present in
the recipient at the time of the infusion that reacts with donor red cells. Depending
on the nature of the antibody and its ability to bind complement, the reaction can
result in intravascular or extravascular hemolysis.
Delayed hemolytic transfusion reactions (DHTR) are the result of immune-
mediated hemolysis that occurs days to weeks after the transfusion but is a direct
result of the transfusion. There is usually an anamnestic antibody response in
which antibody titers rise to detectable levels after transfusion of red cells
that bear an alloantigen to which the recipient has been previously sensitized.
Most DHTRs result in extravascular hemolysis. However, DHTRs can be associ-
ated with intravascular hemolysis, leading to life-threatening complications.
C. Intravascular Versus Extravascular Hemolysis

Intravascular hemolysis is the lysis of red cell membranes within the lumen of
the blood vessel, with release of hemoglobin into the plasma. Immune-mediated
intravascular hemolysis can occur if the antibody is capable of activating the
complete complement cascade. Intravascular hemolysis is frequently rapid, oc-
curring with only a few milliliters of blood, and can lead to shock, acute
renal failure, bleeding due to disseminated intravascular coagulation (DIC), and
death (14).
Extravascular hemolysis occurs when red cells are phagocytized by macro-
phages at the reticuloendothelial system (RES). Immune-mediated extravascular
hemolysis occurs with alloantibodies that do not bind complement or that activate
only a portion of the complement pathway. Most red cell antibodies will promote
the removal of red cells via the RES, unlike anti-A and anti-B, which are most
efficient at causing intravascular hemolysis. Extravascular hemolysis is usually
not clinically severe, as the rate of hemolysis is usually much slower. Often,
hemolytic reactions are a combination of intravascular and extravascular hemol-
ysis. The clinical manifestations and therapy depend on the mechanism that
predominates (14).
D. Pseudo-Hemolytic Transfusion Reactions

Finally, things are not always as they seem. Hemolysis that occurs during or after
a transfusion may not be related to the transfusion (Table 3). For example,
medications administered in temporal proximity to a transfusion may cause
hemolysis of the recipients own blood by several immune mechanisms or
through a nonimmune mechanism such as hypotonicity (15). A patient with
underlying G6PD deficiency or sickle cell trait might develop hemolysis, espe-
cially in the operative setting, which could mimic a transfusion reaction. Hemol-
ysis may also result from an underlying primary infection in the patient, such as
malaria, clostridial infection, or infectious mononucleosis, which is unrelated to
transfusion. Hemolysis of red cells can be due to mechanical trauma from
Table 3 Mechanisms of Hemolysis from Causes Other than Transfusion:

Pseudo-Hemolytic Transfusion Reactions
Immune Mediated
Autoimmune hemolytic anemia
Drug-induced anemia: various immune mechanisms
Nonimmune Mediated
Osmotic damage (hypotonic solutions, drugs, bladder irrigation)
Congenital red cell abnormalities in recipient (sickle cell disease, sickle trait,
G6PD deficiency)
Infection-induced hemolysis (infection in recipient unrelated to transfusion: malaria,
clostridial infection, etc.)
Microangiopathic hemolysis (disseminated intravascular coagulation, thrombotic
thrombocytopenic purpura/hemolytic uremic syndrome)
Mechanical hemolysis (valvular and arterial prostheses, extracorporeal circulation)
Reabsorption of blood from internal hemorrhage
valvular or arterial protheses, DIC, or thrombotic thrombocytopenic purpura-

hemolytic uremic syndrome. In addition, the picture of hemolysis, especially
delayed hemolysis, can be mimicked by the reabsorption of blood after an internal
hemorrhage or with a large hematoma (9).
Two clinical situations merit particular attention: (a) distinguishing between
a sickle cell crisis from a DHTR, and (b) distinguishing alloimmune hemolysis
from autoimmune hemolytic anemia, where the hemolysis is related to the
underlying disease process. Both situations present a difficult, but important,
diagnostic and management problem. These examples emphasize the importance
of obtaining a meticulous clinical history, physical examination, and pertinent
laboratory data when an HTR is suspected.
III. THE ACUTE HEMOLYTIC TRANSFUSION REACTION

AHTRs almost invariably represent immune-mediated hemolysis occurring as the
result of circulating red cell antibody in the recipient that is directed toward a red
cell antigen carried on the transfused red cells. The antibody may be naturally
occurring or one to which the recipient has been previously sensitized, resulting
in the rapid onset of hemolysis. The symptoms occur during or within hours of
the infusion. The most serious reactions usually result from ABO incompatibility,
but incompatibility in the Rh, Kell, Duffy, and Kidd systems exceed ABO blood
type mismatch in frequency as the cause of acute hemolysis (9).
The frequency of AHTRs appears to be declining (16). However, estimates
of AHTR frequency vary depending on the level of awareness of transfusion-
related complications in the persons performing the transfusions, as well as on

the presence of underlying conditions in recipients that might mask the signs and
symptoms of a hemolytic reaction. The estimated frequency of AHTRs is approx-
imately 1 of 25,000 red cell units (17), while fatal HTRs occur in approximately
1 of 600,000 red cell units transfused (18).
In 1975, the U.S. Food and Drug Administration mandated the reporting of
deaths associated with blood collection and transfusion. In 1990, Sazama re-
viewed these data for a 10-year period (19761985). The most commonly
reported cause of death was acute immune-mediated hemolysis, accounting for
158 of 355 deaths (51%). Acute nonimmune-mediated hemolysis accounted for
6 deaths, while delayed immune-mediated hemolysis was responsible for 26
deaths during this same period. ABO incompatibility clearly caused 131 of the
158 fatalities (83%). Of the remaining 27 deaths due to AHTR, most were
suspected to be ABO related, but appropriate documentation was lacking. Only 9
of the 158 deaths due to AHTR were definitely the result of incompatibilities in
other red cell antigen systems (19).
Case reviews indicate that fatal AHTRs have occurred as a result of human
error, most commonly an error in identification at the time of blood infusion
(77/158 or 49%) (Table 4). These errors tend to be associated with inadequate
training of the personnel administering the blood in the proper identification
procedures (19). In a review in New York State, Linden et al. reported the risk of
a transfusion error as 1 in 12,000 red cell units transfused, with a 1 in 33,000 risk
of an ABO- incompatible transfusion, and a risk of 1 in 600,000 for a fatal
transfusion error. Forty-three percent of these errors resulted from failure to
properly identify the patient or the unit prior to transfusion, and 11% resulted
from identification errors at the time of recipient sample collection. While the
majority (58%) of these errors occurred outside of the blood bank, the blood bank
alone was responsible for 25% of errors; combined errors of the blood bank and
other hospital services contributed to 17% of errors (18). AHTRs tend to occur in
urgent situations that require large amounts of blood, such as during surgery, in
the emergency department, or in the intensive care unit. But they can occur in any
setting if proper operating procedure for handling and identifying blood products
is not followed.
Table 4 Most Common Errors Resulting in Death

Blood given to wrong patient
Errors in obtaining and labeling pretransfusion specimen from patient
Serological mistakes
Clerical errors in laboratory
Wrong blood issued from laboratory
A. Clinical Presentation and Differential Diagnosis

Because the earliest signs and symptoms of hemolysis are nonspecific, all
personnel involved with ordering and administering blood products must main-
tain a high level of suspicion (Table 5). Prompt recognition and appropriate
management of a transfusion reaction may prevent a death (16,20). No pathogno-
monic signs or symptoms clearly differentiate an AHTR from other acute trans-
fusion reactions. The clinical presentation may vary widely because signs and
symptoms depend on a number of different factors, including the antibody
specificity, the quantity of antigen on the transfused cells, the quantity of
incompatible cells infused, the immunoglobulin class and subclass of the anti-
body, the antibody titer, the thermal amplitude of the antibody, the ability of the
antibody to activate complement, as well as the clinical condition of the patient
(2). Almost every symptom experienced during or after the infusion of red cells,
except for isolated urticaria and/or simple pruritis, should raise the possibility of
hemolysis and should be properly evaluated.
Classically, fever, defined as an increase in body temperature of greater than
one degree Celsius, is considered the most common initial manifestation of
immune hemolysis, whether acute or delayed (9). In approximately 50% of the
cases, fever will be accompanied by chills (21). A patient often experiences a
vague uneasiness and pain or discomfort at the infusion site during the early
stages of an AHTR. Generalized flushing, nausea and other gastrointestinal
symptoms, dyspnea, chest pain, back pain, and headache or lightheadedness occur
less commonly. The most severe transfusion reactions may begin with hypoten-
sion and rapidly progress to shock. Hemolysis may result in hemoglobinemia and
hemoglobinuria. Evidence of renal dysfunction, including oliguria or anuria, may
be part of the clinical presentation in more severe reactions. Rarely, a patient
may develop generalized oozing or fulminant bleeding as a result of DIC.
Table 5 Signs and Symptoms of Acute Hemolytic Transfusion Reaction

Fever, chills
Flushing, diaphoresis
Localized pain (infusion site, chest, back)
Anxiety, agitation, feeling of impending doom
Dyspnea, tachycardia
Gastrointestinal symptoms (nausea, emesis, diarrhea, abdominal pain)
Hemoglobinemia, hemoglobinuria
Hypertension, hypotension, shock
Oliguria, anuria
Generalized bleeding
Cardiac arrest
If the patient is unconscious, as when under anesthesia, after trauma, or in

shock, or if the patient is an infant or a very young child, the diagnosis can be
particularly difficult. Early prodromal symptoms may be absent. Hypotension,
hemoglobinuria, shock, or a bleeding diathesis may be the first and only indica-
tion of an HTR (2). It is generally accepted that hemolytic reactions are un-
derrecognized and underreported in these settings. These patients may be given
additional units of incompatible blood prior to the realization that an HTR has
occurred (3).
The clinical spectrum of an AHTR can vary from immediately life threat-
ening to subclinical hemolysis. The severity of a reaction cannot be predicted by
early signs and symptoms. Rarely, a patient may hemolyze an entire unit of blood
and yet show no adverse clinical manifestations. However, any patient transfused
with a unit of blood that is found to be incompatible because of an antibody
associated with severe intravascular hemolysis must be monitored as if life-
threatening hemolysis is occurring.
B. Pathophysiology of Potential Complications

The most feared complications of acute immune-mediated intravascular HTRs are
hypotension progressing to shock, DIC with diffuse bleeding, and renal failure
leading to oliguria and anuria (3). The morbidity and mortality rates of AHTR
appear to be directly related to the occurrence of renal failure or DIC (21). It is
important to understand the pathophysiology of these most serious complications
so that the most effective intervention can be provided. Activation of the coagu-
lation cascade and alteration of the vasomotor tone are among the most damaging
consequences of these reactions and, if not recognized early and treated effec-
tively, can result in a downward spiral of clinical events culminating in death.
Activation of the complement system by antigen-antibody complexes leads
to the production of anaphylatoxins (C3a and C5a), causing degranulation of mast
cells with the release of histamine and serotonin. These vasoactive amines are
thought to increase vascular permeability and cause bronchial and intestinal
smooth muscle contraction as well as release of lysosomal enzymes from neutro-
phils. In addition, C5a can act directly on capillaries to induce vasodilation,
further worsening hypotension (10). Immune-mediated activation of complement
and Factor XII (Hageman factor) can result in activation of the coagulation
cascade and, if unchecked, can result in DIC with uncontrolled bleeding.
Thromboplastic substances from lysed red cells probably only potentiate the
intravascular coagulation. Nonimmune causes of hemolysis are not usually asso-
ciated with DIC (2).
Activation of Hageman factor can also lead to the production and activation
of bradykinin. Bradykinin causes increased capillary permeability and dilation of
arterioles, resulting in hypotension, flushing, and localized pain. Another import-
ant group of intracellular mediators, catecholamines, are released in response to

hypotension and in response to immune mechanisms. Catecholamine activation
contributes to localized pain and can cause tachycardia and gastrointestinal
symptoms such as nausea and vomiting (22).
The combined effects of systemic hypotension, renal vasoconstriction, and
clotting within the renal vasculature often lead to renal ischemia. Depending on
the degree and duration of ischemia, the effect may be a transient decrease in
renal function, temporary acute tubular necrosis, or permanent renal failure
secondary to bilateral renal cortical necrosis (3). In animal studies, free hemo-
globin is not directly toxic to the kidneys except in the presence of dehydration
and, by itself, is not thought to have a significant role in renal dysfunction (2).
Recently, there has been increased interest in the potential role of the
cellular immune system during transfusion reactions. A variety of cytokines
[tumor necrosis factor, interleukin (IL)-1, IL-6, IL-8, and monocyte chemoattract-
ant protein] have been implicated as mediators of many of the systemic effects
produced by immune-mediated intravascular hemolysis (23). New research into
cytokines as inflammatory mediators in sepsis may lead to new therapeutic
approaches for HTRs, because many of the same pathways are involved during
intravascular hemolysis.
C. Management and Appropriate Follow-Up

An AHTR is a medical emergency. The highest morbidity and greatest risk of
fatal reactions are with ABO-incompatible transfusions. Severe reactions have
been reported with the infusion of as little as 520 mL of ABO-incompatible
blood (21). Fatal reactions have been reported with as little as 30 mL (19).
For this reason, whenever an AHTR is suspected the blood transfusion must be
stopped immediately (Table 6). The unit of blood as well as the administration
tubing should be removed down to the needle hub, maintaining intravenous
access, but assuring that the patient receives no more of the component, including
the 1020 mL that might be present in the tubing. Because the most common
cause of fatal AHTR is an identification error, a thorough check of labels on the
unit should be done and the patients identification verified. Personnel adminis-
Table 6 Initial Management of Acute Hemolytic Transfusion Reaction

Stop transfusion at needle hub; maintain intravenous access
Notify primary care physician
Repeat identification of patient and unit
Initiate support therapy to maintain blood pressure and renal function
Notify the transfusion department; initiate a transfusion-reaction workup
tering the blood should notify the patients primary physician immediately.
Treatment should be initiated as required clinically. The transfusion service
should be notified immediately and a transfusion reaction workup initiated.
Prompt blood bank evaluation can eliminate a potential second AHTR.
If there has been an identification error that involves mislabeled test samples or
misidentified blood components, a second patient could be in danger of receiving
a transfusion of incompatible blood. A patient who receives a random unit of
blood will have a one-in-three chance of a major ABO incompatibility (pa-
tients serum contains an ABO antibody to an antigen on the donors red cells)
that may result in a life-threatening transfusion reaction (10), and of these, about
a tenth are associated with a fatal outcome (24).
Fortunately, severe, acute life-threatening HTRs are rare. It has not been
possible to determine the most effective modes of therapy by means of controlled
studies. Therapy is supportive and must be directed towards correcting or
minimizing the pathophysiological events just described. Close monitoring and
aggressive support in an intensive care setting is indicated for all patients with
significant hemolysis.
Hypotension must be treated aggressively. Since the renal failure that may
accompany HTRs is thought to result from renal ischemia, therapy should be
directed toward the prevention of systemic hypotension and the maintenance of
renal cortical blood flow. If hypotension or shock can be adequately treated, renal
perfusion can usually be maintained and renal failure prevented (20).
One of the worst prognostic signs for the patient with an acute immune-
mediated HTR is the development of fulminant DIC. Fortunately, in most cases
the initiating immune hemolysis is short-lived, and few patients progress to DIC
(3). The optimal therapy for DIC precipitated by a transfusion reaction remains
controversial. Factors to consider when treating DIC include the severity of the
reaction, the antibody involved, the amount of incompatible blood infused, and
the patients clinical status. The infusion of fresh frozen plasma, cryoprecipitate,
and platelets may be necessary to stop hemorrhage and correct the coagulopathy.
The most severe HTRsthose that result in DICare associated with the
infusion of greater than 200 mL of ABO-incompatible blood. In the treatment of
severe AHTRs, Goldfinger has recommended the prophylactic use of heparin to
prevent DIC (22). However the efficacy of heparin in this setting is unknown.
Many of these patients may already have a source of active bleeding for which
the initial blood transfusion was ordered.
The response to all therapy should be closely monitored with frequent
laboratory and clinical assessment, and subsequent therapy should be guided by
both. The most useful measurements for ongoing hemolysis include hemoglobin
concentration, fractionated bilirubin, and lactic dehydrogenase (LDH); standard
measurements of coagulation and renal function are also indicated. The blood
bank can perform serological studies to determine whether an offending antibody
is present, identify the specificity of the antibody, monitor the removal of

antigen-positive transfused red cells, and provide antigen-negative blood should
further transfusions be necessary.
IV. THE DELAYED HEMOLYTIC TRANSFUSION REACTION

The definition of a DHTR is accelerated immune-mediated destruction of trans-
fused red cells after a period of time, during which there is production of an
alloantibody in the recipient to an alloantigen carried on the transfused cells.
The first case of a delayed hemolytic reaction as a result of blood incompatibility
was described by Boorman et al. in 1946 (25). In 1957, Fudenberg and Allen
reported a series of cases and clearly established that, in a previously sensitized
patient, a hemolytic reaction can occur even if pretransfusion testing fails to
identify an alloantibody or incompatibility with the donor red cells. Thus, DHTRs
became recognized as a distinct clinical entity (26).
The true incidence of DHTRs remains unknown. A significant number of
these reactions go undetected. The accurate detection of a DHTR is largely
determined by two factors: (a) the level of clinical awareness of this potential
complication among the medical staff caring for patients, and (b) the sensitivity
of the laboratory tests used for diagnosis. From 1974 to 1977, 35% of the DHTRs
at the Mayo Clinic were recognized in the laboratory but were not suspected
clinically. Signs or symptoms indicative of an HTR had not been appreciated (27).
In three successive series from the Mayo Clinic, the reported increased frequency
of DHTRs [196473: 1/11,650 units (28); 197477:1/4,000 units (27); 1978
80:1/1,500 units (29)] was attributed to increased clinical awareness combined
with more sensitive methods of antibody detection.
It is even more difficult to assess mortality data for DHTRs. Most of the
literature still consists of only individual case reports. Twenty-six deaths directly
attributed to a DHTR were reported to the FDA from 1976 to 1985. Nearly all
reports of fatal DHTRs involved patients with multiple antibodies (19). The ma-
jority of patients who died were extremely ill prior to the DHTR, making it very
difficult to know to what extent the transfusion reaction contributed to their death.
It is possible for red cell alloantibodies formed during a primary immune
response to lyse the transfused cells that initiated the response and result in a
DHTR. However, most DHTRs result from a secondary or anamnestic immune
response in patients previously sensitized by transfusion or pregnancy; the patient
typically has no detectable antibodies prior to transfusion, but rapidly develop
high titers of antibody after transfusion. Antibodies most often implicated are
directed against antigens of the MNS, Rh, Kell, Kidd, or Duffy blood group
systems. Because some antibodies rapidly decline to undetectable levels, only to
reappear rapidly with subsequent stimulation, knowledge of prior red cell al-
loimmunization is essential to preventing the DHTR.
A. Clinical Presentation and Differential Diagnosis

Unlike the dramatic clinical presentations of the classic AHTR, most DHTRs are
completely asymptomatic or associated with very mild symptoms resulting from
the gradual destruction of sensitized donor red cells. Many reactions are noticed
only because of an unexplained failure to maintain the patients posttransfusion
hemoglobin level or because of low-grade fever and/or mild jaundice with an
elevated unconjugated bilirubin. As with AHTR, fever is the most frequent
symptom (Table 7). The triad of fever, anemia, and a history of a recent blood
transfusion should alert one to the possibility of a DHTR (28). These patients
often have other potential causes of fever and anemia, and in the days to weeks
after a transfusion, medical staff may fail to connect the prior transfusion with the
current symptoms. In addition, there may be no overt symptoms, or the symptoms
may be very mild, occurring after discharge, and not even noticed or reported
by patients.
Such reactions usually come to light when a subsequent request for trans-
fusion reveals the presence of a new red cell alloantibody and/or a newly positive
DAT. The maximal rate of red cell destruction seems to occur between the 4th
and 13th days, although the signs and symptoms are most commonly encountered
on about the 7th day (10). The insidious presentation is typical of extravascular
hemolysis. Hemoglobinuria and hemoglobinemia are seldom present (21). Some
DHTRs with extravascular hemolysis are accompanied by oliguria, but acute
renal failure rarely develops in the absence of intravascular hemolysis (30). Very
infrequently, DHTRs may result in intravascular hemolysis and present with the
devastating complications usually associated with AHTRs, such as hypotension,
renal failure, and DIC.
The consideration of a DHTR is particularly important in transfused
patients with sickle cell anemia. A DHTR occurring after a red cell exchange
Table 7 Expected Onset of Clinical and Laboratory Manifestations of a

Delayed Hemolytic Transfusion Reaction
Positive direct antiglobulin test: 23 days posttransfusion
Spherocytes: 34 days posttransfusion
Free antibody detectable by indirect antiglobulin test: 57 days posttransfusion
Anemia: 57 days postttransfusion
Jaundice: 57 days posttransfusion
Hemoglobinuria: 57 days posttransfusion
transfusion may be very difficult to distinguish from a sickle cell crisis. These
patients may become acutely ill as a result of a DHTR, with symptoms including
vaso-occlusive crisis, bone infarction, renal insufficiency, and profound anemia.
Initially, both DAT and IAT may be negative as a result of the clearance of
sensitized red cells. Elevated bilirubin lactate dehydrogenase (LDH), and urinary
hemoglobin, as well as other indications of hemolysis, may be present in both
conditions. If DHTR is not considered and the symptom complex is attributed to
a sickle cell crisis, these patients could continue to be transfused with antigen-
positive blood (31,32).
When fever without hemolysis occurs days to weeks after a transfusion,
transmission of an infectious agent or transfusion-associated graft-versus-host
disease should be considered (21). If delayed intravascular hemolysis of both
donor and recipient red cells is occurring, transfusion-induced malaria or babesi-
osis should be considered (33). The nonimmune-mediated hemolysis of trans-
fused G6PD-deficient red cells can also present with mild transient hemolysis
several days after their infusion, accompanied by mild jaundice and a two- to
threefold increase in bilirubin and LDH (3).
B. Management and Appropriate Follow-Up

In cases of predominately extravascular hemolysis with a slowly decreasing
hemoglobin concentration and no other sequelae, adequate hydration and cautious
clinical observation with careful monitoring of the hemoglobin and renal function
are usually sufficient (14). If additional transfusions become necessary, the patient
will require antigen-negative blood. In patients with significant intravascular
hemolysis and more severe symptoms, management is supportive and should
parallel that for the patient with a severe acute immune-mediated hemolysis.
It is also imperative that the clinical staff and patient be instructed in red
cell alloantibody status and understand the importance of honoring these anti-
bodies for all future transfusions, even when they can no longer be detected by
pretransfusion screening tests. The patient must always be notified in writing
and given personal documentation indicating the presence and specificity of a
clinically significant red cell alloantibody. Widespread accessibility of blood
bank records between medical facilities will further reduce the risk of a subse-
quent DHTR.
V. LABORATORY EVALUATION OF THE SUSPECTED

HEMOLYTIC TRANSFUSION REACTION
When an AHTR is suspected, the initial evaluation to confirm the presence of
hemolysis and determine its etiology and clinical significance is ordinarily
performed by the blood bank. Specimens sent to the blood bank should include
a coagulated and anticoagulated tube of blood, along with all suspected units, IV
tubing, and infusion sets. The patients blood samples must be drawn carefully to
avoid artifactual hemolysis. Symptoms that occur several hours after a transfusion
may be observed during the infusion of a subsequent unit in a patient receiving
multiple transfusions. The blood bank will perform a serological evaluation on all
suspected units.
The blood banks initial evaluation will include three steps: a clerical check,
an evaluation of hemolysis, and an evaluation for any evidence of serological
incompatibility. The clerical check will confirm the identity of the patients
sample and of the blood product(s), ensuring that there was no misidentification
that could have resulted in the patients receiving an incompatible unit.
Posttransfusion plasma will be examined for evidence of hemolysis. Intra-
vascular hemolysis of as little as 5 mL of blood usually raises the plasma
hemoglobin concentration to >50 mg/dL. This level of free plasma hemoglobin
should be rapidly and easily detectable by gross visual inspection of the
plasma/serum layer of a centrifuged blood specimen (6). Free hemoglobin is
usually cleared from the plasma in 512 hours. If the posttransfusion blood
sample is taken several hours after the transfusion, when the plasma could
possibly be cleared of hemoglobin, an assay for the presence of methemalbumin
might be useful (10). A posttransfusion urine sample can also be evaluated for
any evidence of hemoglobinuria. It is important that the urine is tested for free
hemoglobin and distinguished from hematuria and myoglobinuria.
Pre- and posttransfusion DATs are performed to check for serological
incompatibility (8). A positive posttransfusion DAT with a mixed-field appear-
ance is indicative of alloantibody coating antigen-positive transfused red cells.
Newly detected antibodies coating transfused cells in this setting are virtually
diagnostic of an immune basis for the hemolysis. Antibody may be demonstrable
in the patients serum using the IAT. However, the DAT may be positive days
before free antibody can be detected. It may be possible to identify the allo-
antibody by eluting it from the red cells.
When an AHTR is suspected after the three initial checks, additional
laboratory tests to determine the cause of the reaction may include confirmation
of ABO and Rh(D) types on pre- and posttransfusion samples as well as the
unit(s), repeat crossmatch, and other serological testing to detect alloantibodies
and/or incompatibility with donor units. If no clerical error has occurred, the
plasma is not grossly hemolyzed, and the DAT is negative or unchanged, an acute
HTR is extremely unlikely. However, if rapid hemolysis of all antibody-coated
red cells has occurred with rapid clearance of the free hemoglobin from the
plasma, or if antibody-coated red cells are removed rapidly from the circulation
by the reticuloendothelial system, immune-mediated AHTR is possible with a
negative DAT (21).
If initial test results are unclear, additional confirmatory tests may be useful
to verify an AHTR. One should also determine whether a patient achieved the
expected hemoglobin rise after the transfusion: in a 70 kg adult, the hemoglobin
should increase by 1 g/dL per unit of red cells when measured 15 minutes after
the transfusion (34).
Serum bilirubin and LDH increase in both intravascular and extravascular
hemolysis. Rising unconjugated bilirubin may be detectable as early as one hour
postreaction, with peak levels occurring in 46 hours and disappearing in 24
hours if bilirubin excretion is normal. A serum haptoglobin level may be helpful
in suspected hemolysis, but a precipitous decrease in haptoglobin level tends to
occur very early and is not as reliable as hemoglobinemia. Visible hemoglobine-
mia develops after haptoglobin depletion, therefore little is gained by measuring
haptoglobin when hemolysis is already visible. The usefulness of a haptoglobin
level is also limited by the wide range of its normal values. Nonimmune-
mediated hemolysis will also lower haptoglobin concentrations.
When an immune-mediated DHTR has occurred, spherocytes observed on
the peripheral blood smear may be an early indication of red cell destruction (14).
A blood bank evaluation should be initiated in any case of suspected DHTR. If the
evaluation reveals that a positive DAT has developed, demonstrating the presence
of complement and/or antibody-coated red cells or the presence of previously
undetected red cell antibodies associated with a rapid disappearance of transfused
donor cells, then the diagnosis of a DHTR is made (21).
Characteristically, the DAT becomes positive a few days after transfusion
and remains positive until the incompatible red cells are eliminated. Typically,
antibody becomes detectable 47 days after transfusion and reaches a peak value
1015 days after transfusion (10). Antibody eluted from the red cells may help
identify the alloantibody when the antibody titer is low and difficult to detect in
the serum.
VI. PRETRANSFUSION COMPATIBILITY TESTING

The goal of pretransfusion compatibility testing is to allow the selection of red
cells that will circulate for an acceptable period and that will have a low risk for
adverse effects (Figure 1). First and foremost is accurate and reliable ABO and
Rh(D) typing of recipient and donor (35). In addition, pretransfusion compatibil-
ity tests are designed to detect the presence of potentially hemolytic red cell
alloantibodies by use of both an antibody screen (patients serum against a known
panel of red cells) and the major crossmatch (patients serum with donor red
cells). The antibody screen detects alloantibodies to common red cell antigens
that could go undetected in a major crossmatch because of the expression of a
limited number of red cell antigen sites, as with heterozygous genes, or weakened
Figure 1 Pretransfusion compatibility testing.
expression of an antigen. On the other hand, the major crossmatch confers the
ability to detect alloantibodies to uncommon antigens, which might not be present
on the standard panel of screening cells but are present on the donor red cells that
have been selected for the particular patient. The antibody screen and major cross
match complement one another. The FDA requires that 18 red cell antigens
representing the most clinically significant of the blood group systems be present
on reagent red cells used for antibody screening. In addition, most other high-
frequency antigens are present on reagent red cells (20). Only nine antibodies
reactive to antigens not usually contained on commercial prepared red cell panel
cells have been the reported cause of HTRs. A critical review of these reports
reveals inadequate documentation in many instances (6). Under special condi-
tions, an electronic or computer crossmatch may substitute for serological
testing (36).
No in vitro test is capable of predicting the clinical significance of a red
cell antibody. However, several serological characteristics of red cell alloanti-
bodies correlate well with clinical consequences and form the basis for most of
our clinical judgment. The two serological characteristics that have proved most
helpful in predicting clinical significance are the antibody specificity, which is
determined by demonstrating reactivity with the corresponding red cell antigen,
and the antibodys in vitro reactivity at 3037C (37). Some antibodies are
invariably associated with hemolytic events, while others have never been
reported to cause untoward reactions in any patient. As previously discussed, the
severity of the hemolysis caused by these antibodies varies dramatically. Experi-
ence gained from transfusing patients in life-threatening situations when compat-
ible, antigen-negative blood supplies are unavailable or exhausted has illustrated
that anticipated hemolysis may not occur or may be minimal (38). This situation
is most likely to occur in the setting of a patient who has multiple red cell
alloantibodies. Although every effort should be made to transfuse these patients
with antigen-negative blood, in the emergency setting it may be necessary to
transfuse blood that is positive for the antigen(s) thought to be the least clinically
significant. In these situations, small aliquots should be infused slowly, with close
clinical supervision and careful monitoring for any evidence of hemolysis. In less
urgent situations where a negative crossmatch is difficult or impossible to obtain
and the red cell antibodies are of questionable clinical significance, in vivo
compatibility testing has been applied (15).
Just as the presence of an antibody does not necessarily mean that hemoly-
sis will occur, the absence of an antibody does not guarantee safety. Although
antibody screening and crossmatching with a recently obtained blood sample
identify most red cell alloantibodies, state-of-the-art serological techniques do not
identify all cases of incompatibility. Previously identified alloantibodies that may
be clinically significant should always be honored. Many clinically significant
antibodies become undetectable over time, only to reappear after transfusion of
red cells bearing the corresponding antigen. In recent studies by Ramsey et al.,
3035% of clinically significant red cell alloantibodies become undetectable on
follow-up screening within one year (with anti-Kidd undetectable in 59% of
follow-up screens and anti-C undetectable in 45% of follow-up screens), and
nearly 50% of all clinically significant alloantibodies became undetectable after
10 or more years (39).
Rarely, there will be cases where all evidence indicates immune-mediated
accelerated destruction of red cells in the absence of demonstrable antibodies
(40). An HTR may occur, possibly repeatedly, without serological identification

of an alloantibody by standard screening tests. This may become a serious clinical
dilemma when additional transfusions are necessary. In such cases, more sensitive
assays and more detailed testing may be required. It may be possible to confirm
the identity of the incompatible antigen by radiolabeling a small aliquot of red
cells expressing the suspect antigen and performing red cell survival studies in
the patient (40). It is still assumed that the hemolysis in these cases is immune
mediated even without detectable antibody, just as cases of autoimmune hemo-
lytic anemia have been described in the absence of a positive DAT (many of these
cases have been shown to have IgG on the red cell surface, which is undetectable
by conventional techniques) (6).
VII. PREVENTIVE MEASURES

Numerous and sophisticated serological techniques can be applied to donor red
cells and recipient serum in an effort to detect an incompatibility. However, the
array of laboratory methods must be put in the proper perspective. Ironically,
further perfecting the serological aspects of compatibility testing would prevent
only a small number of the rarely fatal HTRs. Most deaths related to blood
transfusion remain a consequence of giving incompatible blood to a patient as a
result of human error, not of insufficiently sensitive laboratory screening tech-
niques. Most of these erroneous blood administrations result from identification
errors that occur outside of the blood bank.
Educating and training all involved personnel in the safe and correct
handling of laboratory specimens and blood components remain the most critical
steps in preventing death from blood transfusion. Errors are usually made during
times of stress and are compounded by inadequate training.
Emergency transfusion of previously unhospitalized patients can be com-
plicated by special problems in patient identification and accurate labeling of
blood specimens. Even in extremely urgent clinical situations, a properly labeled
patient sample must be drawn prior to any transfusions to allow appropriate
testing for future transfusions and to serve as a pretransfusion reference in the
event of a transfusion reaction. The blood bank cannot rely on old records or
typing results from other institutions.
Clear and detailed standard operating procedure for handling laboratory
specimens and blood components should be carefully developed and fully under-
stood by well-trained clinical and laboratory personnel. The proper procedures
should always be followed. Identification errors are most likely to occur when the
blood is hung for transfusion, but they may occur at any point in the chain of
sample collection, testing, blood preparation, storage, issue, and transfusion. It is
important to remember that even with the use of autologous blood, the risk still
exists of human error leading to transfusion of an incorrect unit of blood (16).

Computer-assisted automation shows promise as a mechanism to reduce the
chance of human error at several points in the transfusion process, including
automated patient and donor identification, automated crossmatch, and automated
release of blood. Interest is increasing in automated tracking systems that monitor
identification from initial phlebotomy until the blood is transfused (41). Systems
technology may prove to be the method of choice in preventing many of the
so-called clerical errors that resist the most persistent training.
ACKNOWLEDGMENT
Much thanks to Ms. Jo Procter, M. Ed., MT, (ASCP) SBB, of the Transfusion
Medicine Department for her careful reading and thoughtful assistance with the
manuscript in draft.
REFERENCES
1. Wick MR. Blood transfusion. An historic and current perspective. Am J Clin Pathol
1989; 91(suppl):S2S6.
2. Beauregard P, Blajchman MA. Hemolytic and pseudo-hemolytic transfusion re-
actions: an overview of the hemolytic transfusion reactions and the clinical conditions
that mimic them. Transfusion Med Rev 1994; 8:184199.
3. Brecher ME, Taswell HF. Hemolytic transfusion reactions. In: Rossi EC, Simon TL,
Moss GS, eds. Principles of Transfusion Medicine. Baltimore: Williams and Wilkins,
1996:747763.
4. Oberman HA. The history of transfusion medicine. In: Petz LD, Swisher SN,
Kleinman S, et al., eds. Clinical Practice of Transfusion Medicine. New York:
Churchill Livingstone, Inc., 1996:1132.
5. Oberman HA. The crossmatch. A brief historical perspective. Transfusion 1981;
21:645651.
6. Shulman IA, Petz LD. Red cell compatibility testing: clinical significance and
laboratory methods. In: Petz LD, Swisher SN, Kleinman S, et al., eds. Clinical
Practice of Transfusion Medicine. New York: Churchill Livingstone, 1996:199244.
7. Daniels GL, Anstee DJ, Cartron JP, et al. Blood group terminology 1995. ISBT
working party on terminology for red cell surface antigens. Vox Sang 1995; 69:265
279.
8. Jeter EK, Spivey MA. Noninfectious complications of blood transfusion.
Hematol/Oncol Clin North Am 1995; 9:187204.
9. Pineda AA, Brzica SM Jr, Taswell HP. Hemolytic transfusion reaction. Recent
experience in a large blood bank. Mayo Clin Proc 1978; 53:387390.
10. Mollison PL, Engelfriet CP, Contreras M. Blood Transfusion in Clinical Medicine.
Oxford: Blackwell, 1993.
11. Lostumbro MM, Holland PV, Schmidt PJ. Isoimmunization after multiple trans-
fusions. N Engl J Med 1966; 275:141144.
12. Silberstein L, Spitalnik SL. Blood group antigens and antibodies. In: Rossi EC,
Simon TL, Moss GS, eds. Principles of Transfusion Medicine. Baltimore: Williams
and Wilkins, 1996:6787.
13. Pollack W, Ascari WQ, Crispen JF, et al. Studies on Rh prophylaxis. Rh immune
prophylaxis after transfusion with Rh-positive blood. Transfusion 1971; 11:340344.
14. Baron B, Baron J. Blood products and plasmapheresis. In: Hall JB, Schmidt GA,
Wood LDH, eds. Principles of Critical Care. New York: McGraw-Hill, 1992:
15. Silvergleid AJ, Wells RF, Hafleigh EB, et al. Compatibility test using 51-chromium-
labeled red blood cells in crossmatch positive patients. Transfusion 1978; 18:814.
16. Linden JV, Kaplan HS. Transfusion errors: causes and effects. Transfusion Med Rev
1994; 8:169183.
17. Walker RH. Special report: transfusion risks. Am J Clin Pathol 1987; 88:374378.
18. Linden JV, Paul B, Dressier KP. A report of 104 transfusion errors in New York state.
Transfusion 1992; 32:601606.
19. Sazama K. Reports of 355 transfusion-associated deaths: 1976 through 1985. Trans-
fusion 1990; 30:583590.
20. Schulman IA. Safety in transfusion practices. Red cell compatibility testing issues.
Clin Lab Med 1992; 12:685699.
21. Jenner PW, Holland PV. Diagnosis and management of transfusion reactions. In: Petz
LD, Swisher SN, Kleinman S, et al., eds. Clinical Practice of Transfusion Medicine.
New York: Churchill Livingstone, 1996:905927.
22. Goldfinger D. Acute hemolytic transfusion reactions: a fresh look at pathogenesis and
considerations regarding therapy. Transfusion 1977; 17:985987.
23. Davenport RD, Kunkel SL. Cytokine roles in hemolytic and nonhemolytic trans-
fusion reactions. Transfusion Med Rev 1994; 8:157168.
24. Murphy WG, McClelland DBL. Deceptively low morbidity from failure to practice
safe blood transfusion: an analysis of serious blood transfusion errors. Vox Sang
1989; 57:5962.
25. Boorman KE, Dodd BE, Loutit JF, et al. Some results of transfusion of blood to
recipients with cold agglutinins. Br Med J 1946; 1:751.
26. Fudenberg H, Allen FH. Transfusion reactions in the absence of demonstrable
incompatibility. N Engl J Med 1957; 256:11801184.
27. Moore BS, Taswell HF, Pineda AA, et al. Delayed hemolytic transfusion reactions.
Evidence of the need for an improved pretransfusion compatibility test. Am J Clin
Pathol 1980; 74:9497.
28. Pineda AA, Taswell HF, Brzica SM Jr. Delayed hemolytic transfusion reaction. An
immunologic hazard of blood transfusion. Transfusion 1978; 18:17.
29. Taswell HF, Pineda AA, Moore SB. Hemolytic transfusion reactions: frequency and
clinical and laboratory aspects. In: Bell CA, ed. A Seminar on Immune-Mediated Red
Cell Destruction. Washington, DC: American Association of Blood Banks, 1981:71
92.
30. Holland PV, Wallerstein RO. Delayed hemolytic transfusion reaction with acute renal
failure. J Am Med Assoc 1968; 204:10071008.
31. Diamond WJ, Brown FL, Bitterman P, et al. Delayed hemolytic transfusion reaction
presenting as sickle cell crisis. Ann Intern Med 1980; 93:231233.
32. Milner PF, Squires JE, Larison PJ, et al. Posttransfusion crises in sickle cell anemia:
role of delayed hemolytic reactions to transfusion. South Med J 1985; 78:14621468.
33. Smith RP, Evans AT, Popovsky M, et al. Transfusion-acquired babesiosis and failure
of antibiotic treatment. J Am Med Assoc 1986; 256:27262727.
34. Wiesen AR, Hospenthal DR, Byrd JC, et al. Equilibrium of hemoglobin concentration
after transfusion in medical inpatients not actively bleeding. Ann Intern Med 1994;
121:278280.
35. Schulman IA. Controversies in red cell compatibility testing. In: Nance SJ, ed.
Immune Destruction of Red Blood Cells. Arlington, VA: American Association of
Blood Banks, 1989:171199.
36. Butch SH, Judd WJ, Steiner EA, et al. Electronic verification of donor-recipient
compatibility: the computer crossmatch. Transfusion 1994; 34:105109.
37. Garratty G. Factors affecting the pathogenicity of red cell auto- and alloantibodies.
In: Nance SJ, ed. Immune Destruction of Red Blood Cells. Arlington, VA: American
Association of Blood Banks, 1989, pp 109157.
38. Ramsey G, Cornell FW, Hahn L, et al. Red cell antibody problems in 1000 liver
transplants. Transfusion 1987; 27:552.
39. Ramsey G, Smietana SJ. Long-term follow-up testing of red cell alloantibodies.
Transfusion 1994; 34:122124.
40. Harrison CR, Hayes TC, Trow LL, Benedetto AR. Intravascular hemolytic trans-
fusion reaction without detectable antibodies: a case report and review of the
literature. Vox Sang 1986; 51:96101.
41. Jensen NJ, Crosson JT. An automated system for bedside verification of the match
between patient identification and blood unit identification. Transfusion 1996;
36:216221.
3
Other Reactions and
Alloimmunization
Christopher J. Gresens and Paul V. Holland
Sacramento Medical Foundation Blood Centers, Sacramento, California
I. INTRODUCTION
Adverse reactions to transfusions are remarkably heterogeneous. Depending upon

both the patients condition and the blood component transfused, the patient may,
on occasion, experience any of a variety of transfusion reactions. Some are
precipitated by the reaction of endogenous factors (e.g., recipient antibodies) with
corresponding antigens in the transfused components. Others may be caused
by exogenous substances, such as cytokines, bacterial organisms (and/or their
toxins), excess iron, donor antibodies, or even the citrate used as a blood
anticoagulant. And, still, some untoward effects of transfusion are related to
physiological processes as simple as volume overload.
It is our intent to familiarize the reader with some of the more common,
but not usually life-threatening, transfusion reactions, delving into their (a) pre-
sentation; (b) etiology, pathophysiology, and important mediators; (c) fre-
quency; (d) therapy; and (e) prevention. Special emphasis will be placed upon
febrile, nonhemolytic reactions, allergic (both urticarial and anaphylactic) reac-
tions, and transfusion-related alloimmunization. We will also briefly discuss
transfusion-associated volume and iron overload, hypothermia and arrhythmias,
citrate toxicity, air emboli, and isolated hypotensive platelet reactions. The
generally more serious types of transfusion reactions, such as hemolytic transfu-
sion reactions (HTRs), septic transfusions reactions (STRs), transfusion-related
acute lung injury (TRALI), transfusion-associated graft-versus-host disease
71
72 Gresens and Holland
(TA-GVHD), and posttransfusion purpura are discussed in detail elsewhere in this

book and will therefore not be further covered in this chapter.
II. FEBRILE, NONHEMOLYTIC TRANSFUSION REACTIONS

By most definitions, any isolated temperature increase of greater than or equal to
one degree Celsius in association with a transfusion, when there is no other
reasonable explanation for the fever, is said to be a febrile, nonhemolytic
transfusion reaction (FNHTR). The diagnosis is one of exclusion; therefore, other
possible causesboth extraneous (e.g., neutropenic fevers) and related (e.g., an
HTR or STR)must first be ruled out. The temperature rise will generally begin
anytime from early on in the transfusion to 1 or 2 hours after it has been
completed. The FNHTR is one of the most common forms of acute transfusion
reaction, occurring in association with approximately 1% of red blood cell (RBC)
transfusions and anywhere from 5 to 30% of platelet transfusions (1,2).
Two major causes of FNHTRs have been recognized. The first, identified
in the late 1950s (3), is the interaction between recipient antibodies (leuko-
agglutinins) and donor leukocytes (4). The leukoagglutinins result from prior
transfusions and/or pregnancies, and very often have human leukocyte antigen
(HLA) specificity. It is widely accepted that RBC components, when leuko-
reduced such that they contain fewer than 5 108 leukocytes, are associated with
fewer FNHTRs than are their nonleukoreduced counterparts (5,6). Investigators
have proposed that the leukoagglutinins interact with donor leukocytes and
somehow cause the release of pyrogens (from patient macrophages and/or donor
leukocytes), producing fever and chills (7).
While the antibody model accounts for many FNHTRs, there appears to be
at least one more mechanism. This was demonstrated in the early 1990s, when it
was shown that bedside leukoreduction had limited efficacy in the prevention of
FNHTRs associated with platelet transfusions. Heddle et al. (2) and Muylle et al.
(8) demonstrated that there appears to be a veritable cytokine shower (9), which
can be linked to many of these febrile reactions. They showed that the age of the
platelet component as well as the leukocyte count correlated with the probability
that a febrile reaction will occur. After a nonleukoreduced platelet unit has been
stored for 5 days at room temperature, the plasma contains significantly increased
concentrations of interleukin-1 (IL-1), tumor necrosis factor-alpha (TNF-),
interleukin-6 (IL-6), and interleukin-8 (IL-8) (2,8,10,11). All these cytokines have
been associated with febrile responses and can freely pass through filters (12).
Corresponding levels of cytokines in prestorage leukoreduced platelet units
remain at fairly constant levels throughout the storage period as a result of the
reduction of the contained leukocytes (10). Not surprisingly, these units are
associated with significantly reduced FNHTR rates. For example, Muylle et al.
Other Reactions and Alloimmunization 73
observed that traditional platelet concentrates (PCs) were associated with a 9.3%
incidence of transfusion reactions (primarily rigors and/or fevers), whereas PCs
from which the buffy coats were removed prior to storage were associated with
an incidence of only 2.7% (p = 0.007) (13).
Fortunately, FNHTRs are almost never serious, although they can, in some
cases, be very uncomfortable to the patient. The fever usually responds to
antipyretics, such as acetaminophen. As with other acute transfusion reactions,
the transfusion should be discontinued and a work-up performed to rule out the
more serious types of reactions for which fever may be the harbinger (e.g., acute
HTRs, STRs, or TRALI). Most current literature states that the transfusion should
not be restarted with the same unit (14). In certain situations, however, such as
when the reaction is very mild, rapidly responds to antipyretics, and is definitively
shown to be unrelated to the aforementioned serious causes, some believe that
restarting the transfusion with the same unit may be acceptable (15).
The best means of preventing FNHTRs is to leukoreduce the cellular blood
component that is to be transfused. As mentioned previously, leukoreduction of
RBC units to levels of less than or equal to 5 108 leukocytes/unit eliminates
most FNHTRs (16), regardless of whether the reduction is done before or after
storage. Unfortunately, however, while poststorage leukoreduction (e.g., at the
bedside or in the lab immediately prior to issue) is reasonably effective at
preventing FNHTRs associated with RBC transfusions, it is less effective with
those due to platelet transfusions. This is because the aforementioned pyrogens
(IL-1, TNF-, etc.) reach much higher levels in nonprestorage-leukoreduced
platelet units than in similarly prepared RBC units, probably largely owing to the
increased storage temperature for platelets. Prestorage leukoreduction has an
obvious advantage here, in that it eliminates most of the leukocytes that would
otherwise produce pyrogens in vitro.
These days, plateletpheresis units are often collected in a manner that
automatically produces leukoreduced units and are therefore much less likely to
cause FNHTRs. Some blood centers have utilized this approach to set up a
virtually 100% leukoreduced plateletpheresis inventory, as the logistics needed to
accomplish this are not overly complicated. Prestorage leukoreduction of RBC
units and platelet concentrates, on the other hand, is a cumbersome process that
can be difficult to perform on every unit, and often necessitates the establishment
of a separate component inventory. For this reason, it has not usually been done
(until recently) on a very large scale. However, as the U.S. healthcare system
appears to be moving toward near-100% leukoreduced blood component inven-
tories, more and more blood collection facilities are gearing up their prestorage
leukoreduction processes for RBC and platelet concentrate units.
Generally, a patient will not receive leukoreduced cellular components
expressly for the prevention of FNHTRs unless he or she has had two such
documented reactions (assuming that the patient is not already receiving
leukoreduced cellular components for another reason). This is because most

patients who have one FNHTR will not have another one, despite receiving
nonleukoreduced units (17). As a final note, one more means of reducing the risk
of FNHTRs during platelet transfusions is to wash the product immediately prior
to issue. Theoretically, this approach seems very appealing, as washing effectively
removed donor plasma, as well as the cytokines contained within. In practical
situations, however, washing should rarely be needed, except, perhaps, for the
patient who is exquisitely sensitive to even the small levels of cytokines present
in prestorage-leukoreduced platelet units.
III. ALLERGIC TRANSFUSION REACTIONS

A. Urticarial Reactions
The urticarial transfusion reactions generally manifest with some combination of
localized erythema, hives, and pruritis and are unaccompanied by fever or other
adverse findings. They are thought to result from the presence of soluble
allergens in the donor plasma. The frequency of these reactions is said to be
around 1% (17); however, this may be a falsely low value because of under-
reporting. The accepted method of treatment is to interrupt the transfusion,
administer an antihistamine (such as diphenhydramine), and wait for the symp-
toms to subside, at which point, if the reaction were a relatively mild one, the
transfusion could be restarted. Conversely, if the symptoms were more severe
(e.g., affecting a large portion of the patients body surface area), the same unit
should not be restarted.
Many allergic reactions can be prevented by the pretransfusion administra-
tion of an antihistamine. Rarely, the concomitant use of a corticosteroid (e.g.,
hydrocortisone) may be warranted. And in very unusual cases, such as when a
patient has a history of repeated and/or severe urticarial reactions, it may be
necessary to utilize only washed or deglycerolized red blood cells (18) and
washed or resuspended platelets (19,20) for future transfusions.
B. Anaphylactic Transfusion Reactions

Anaphylaxis, fortunately a very rare reaction, may occur after the infusion of only
a few milliliters of a blood component and can cause mortality or significant
morbidity unless treated promptly (Table 1). Typically, some combination of
coughing, bronchospasm, respiratory distress, vascular instability, nausea, ab-
dominal cramps, vomiting, diarrhea, shock, and/or loss of consciousness will
manifest. There is usually an absence of fever (helping to distinguish this reaction
from acute HTRs, TRALI, and STRs). Also, several of the aforementioned
findings will usually occur simultaneously, thereby allowing the differentiation of
Table 1 Some Proposed Causes of

Anaphylactic Transfusion Reactions
Donor IgA
Plasticizers/Sterilizing agents
Passively transferred donor IgE
Passively transferred donor allergens/drugs
HLA-incompatible platelet transfusions
Use of leukocyte reduction filters
Undefined plasma-soluble antigens
an anaphylactic reaction from the purported hypotensive platelet reaction (see

Sec. X).
Many anaphylactic reactions are believed to occur as a result of the reaction
of donor IgA with anti-IgA antibodies (class IgE) produced by IgA-deficient
recipients. Approximately 1 in 700800 people is IgA-deficient (18), but, as
shown by Sandler et al. (who looked at 97 asymptomatic IgA-deficient individu-
als), only approximately 22% of IgA-deficient people will make anti-IgA anti-
bodies (21). Other possible causes for anaphylactic transfusion reactions include
various undefined plasma-soluble antigens, plasticizers and sterilizing agents
used during blood bag production (22), passive transfer of donor IgE antibody
(23,24), passive transfer of donor allergens (25), HLA-incompatible platelet
transfusions (26), and even the use of leukocyte reduction filters (27).
Fortunately, the incidence of anaphylactic transfusion reactions is very
low, with two different series putting it at 1 in 20,000 and 1 in 47,000 units
transfused, respectively (28,29). Still, the morbidity, with resultant increased
hospital stays, associated with this form of transfusion reaction is not inconsider-
able. And although most patients do recover completely, a few do not. For exam-
ple, Sazama found anaphylaxis to be implicated in 8 of 256 non-AIDS/hepatitis
transfusionrelated deaths that occurred between 1976 and 1985. Interestingly,
anti-IgA was demonstrated in only 1 of the 8 cases (30).
When an anaphylactic transfusion reaction is suspected, the transfusion
should be discontinued immediately, and the intravenous line kept open (e.g., with
saline) in order to treat any ensuing hypotension. What follows is an example of
the typical treatment for an adult patient who is having an anaphylactic transfu-
sion reactions: For mild-to-moderate reactions, 0.30.5 mg of epinephrine (0.3
0.5 mL of a 1:1000 solution) can be given subcutaneously every 1530 minutes,
as needed, for a maximum of three doses. For severe reactions, 0.10.5 mg
epinephrine (15 mL of a 1:10,000 solution) should be slowly given intravenously
and repeated every 515 minutes as needed. Sublingual or endotracheal epineph-
rine administration may be used when an IV line is not available. Intravenous
aminophylline is often administered to treat bronchospasm (loading dose of
6 mg/kg, followed by 0.51 mg/kg/h). Volume expansion, using crystalloids, is
also recommended for the treatment of hypotension. A vasopressor, such as

dopamine, may be necessary if the hypotension does not respond to volume
expansion alone. Intravenous hydrocortisone, at a dose of 500 mg every 6 hours,
should be given for severe reactions; it is important to remember, however, that
this is not the first drug to administer, as it takes 612 hours to reach maximal
effect and probably plays a greater role in preventing the redevelopment of
anaphylaxis. Antihistamines, such as diphenhydramine hydrochloride, while hav-
ing little value in the treatment of acute symptoms, may reduce the duration of
the reaction and keep it from redeveloping. And, finally, airway protection and
oxygen administration are sometimes needed. After the diagnosis of anaphylaxis
has been made, the patient should be observed for at least 6 hours, assuming he
or she makes a rapid and complete recovery (31). Also, the patient should be
provided a Medi-Alert bracelet documenting his or her susceptibility to these
reactions to prevent a reoccurrence.
Pretransfusion specimens from any patient who has had a suspected ana-
phylactic reaction should be analyzed for the presence of serum IgA. If the pa-
tient is IgA-deficient, it is important to document the presence or absence
of anti-IgA antibody in his or her serum. When such patients have evidence of
anti-IgA antibodies, prevention of subsequent anaphylactic reactions is accom-
plished by transfusing only IgA-deficient blood components. In the case of RBC
transfusions, it is often possible to remove sufficient quantities of IgA via either
deglycerolization or extensive washing of the units. When platelets are needed,
various washing protocols have been established; however, all cause some
degree of platelet activation and loss, and some may not remove sufficient
quantities of IgA to prevent subsequent reactions. Fortunately, IgA-deficient units
may be obtained through the American Association of Blood Banks (AABB) Rare
Donor Registry and other sources. It is important to note though that, because of
the precious nature of such components, these units should be used only when
absolutely necessary.
IV. TRANSFUSION-RELATED ALLOIMMUNIZATION

Sensitization to RBC antigens following transfusions occurs commonly. After
excluding the ABO system, the most immunogenic of the RBC antigens is the D
antigen, followed, in descending order, by K, E, Fya, and Jka (32). Rosse et al.
demonstrated a direct correlation between the number of RBC transfusions and
the percentage of sickle cell disease (SCD) patients sensitized (33). Hoeltge et al.
formed a similar conclusion based upon their retrospective study looking at
RBC alloantibodies in 159,262 transfused patients with varying diagnoses (34).
Lostumbo et al. determined that the risk of alloimmunization is additive, at a rate
of approximately 1 per cent per unit of transfused blood (35). This cumula-
tive RBC alloimmunization risk is of particular importance to frequently trans-

fused patients. Coles et al. observed that of multiply transfused SCD and
thalassemia patients, 23 of 99 (23%) and 4 of 39 (10%), respectively, became
alloimmunized (36). Others have found similarly high rates of RBC allo-
immunization, with the incidence in SCD patients ranging from 8% in pediatric
populations to 50% in multiply transfused adults (37). This translates into
significantly higher rates of delayed hemolytic transfusion reactions in these
patient populations, the results of which, particularly in patients who receive RBC
exchange transfusions (e.g., SCD patients), can sometimes be devastating (38,39).
Another very important form of alloimmunization is that which occurs
against antigens in the HLA system. These generally occur in multiply transfused
and/or multiparous individuals and, in the transplant setting, can be minimized
through leukoreduction techniques (40). Approximately 3070% of patients
who are exposed to multiple, unmodified (e.g., nonleukoreduced or ultraviolet
Btreated) RBC and platelet transfusions will become alloimmunized to HLA
antigens (41). The consequences of HLA alloimmunization can include FNHTRs
(see earlier), platelet transfusion refractoriness (when antibodies against HLA
class I antigens are involved), and a reduced likelihood for successful solid organ
and bone marrow transplants. The Trial to Reduce Alloimmunization to Platelets
(TRAP) looked at approximately 600 acute myelogenous leukemia patients,
comparing alloimmunization rates in patients who received unmodified, pooled
platelet concentrates (controls) versus rates in patients enrolled in three different
experimental arms. The three arms separated patients into those who received
only leukoreduced platelet concentrates (filtered) versus those who received only
leukoreduced plateletpheresis units (filtered) versus those who received only
ultraviolet Birradiated, pooled platelet concentrates. All patients in the study
received leukoreduced (filtered) RBC units. Rates of HLA alloimmunization and
refractoriness in the control group were approximately 45 and 13%, respectively.
Alloimmunization and refractoriness rates for all three of the experimental arms
were similar (and significantly less than those for the control group), at 1721
and 35%, respectively (i.e., regardless of the type of platelet treatment) (40).
Transfusion-associated alloimmunization may also occur against platelet-
specific antigens. Antibodies have been observed against antigens in each of the
five human platelet antigen systems (HPA-1 through 5) and are the cause of
neonatal alloimmune thrombocytopenia and posttransfusion purpura. They can
also, in some cases, lead to platelet transfusion refractoriness, although less
commonly than is seen with antibodies to the HLA system (42). Neutrophil anti-
gen alloimmunization also occurs, with the resultant antibodies rarely leading
to neonatal alloimmune granulocytopenia (less than 1% incidence). These anti-
bodies are also causal in the development of FNHTRs; however, their role
is considered to be far less important than that of HLA antibodies (42,43).
Neutrophil-specific antibodies have also been shown to cause TRALI (44);
however, their relative importance in the pathogenesis of TRALI, vis--vis HLA

antibodies, is not entirely known (45).
V. TRANSFUSION-ASSOCIATED CIRCULATORY
OVERLOAD
Rapid increases in blood volume are poorly tolerated in many patients with
cardiac, pulmonary, or renal failure, as well as in chronically anemic patients
with expanded plasma volumes and in very young or old patients (4,46). Mani-
festations of transfusion-associated circulatory overload (TACO) include dys-
pnea, rapid increases in systolic blood pressure, coughing, orthopnea, severe
headache, and peripheral edema. The reaction is not so much related to the blood
component itself as it is to the volume infused. In other words, the problem is
primarily a physiological one.
The incidence of TACO is difficult to determine, largely because of the
problem of underreporting. In a series from the Mayo Clinic, Popovsky and
Taswell (47) found that when the formal diagnosis of TACO was made by the
clinicians alone, only 1 in 3168 patients transfused with RBCs was reported as
being affected (as observed by a retrospective review looking at all transfusions
for a 7-year period). When they then set up a more structured consultation service,
that rate increased to 1 in 708. A recent study of elderly orthopedic surgery
patients revealed that over 1% of this population developed TACO. Not surpris-
ingly, the patients who developed TACO were older (mean of 84 vs. 77 years)
than those who did not, and each was in positive fluid balance (mean of 2480
mL) prior to administration of the offending transfusions (46). A final note
regarding the importance of TACO as a cause of patient morbidity and mortality
is that of 355 transfusion-associated deaths that occurred between 1976 and 1985,
39 were caused by acute pulmonary injury. This was the third most common cause
of death after acute hemolysis (158 deaths) and non-A, non-B hepatitis (42
deaths). Anaphylaxis accounted for 8 of the acute pulmonary injury-related
deaths, with the remaining 31 (9% of total deaths) caused by the acute onset of
pulmonary edema or respiratory insufficiency (30). While no distinction could be
made between TACO and TRALI as the cause of these, it is likely that at least
some (and perhaps many) were due to TACO.
Treatment of TACO is to stop the transfusion, sit the patient up as much as
possible, and give a diuretic (e.g., 40 mg intravenous furosemide for a typical
adult) and oxygen. In rare cases, such as when marked pulmonary edema occurs,
phlebotomy of 200400 mL whole blood, with the concomitant slow transfusion
of packed RBCs, may be useful (this is actually faster than plasmapheresis).
Prevention may be accomplished by transfusing the at-risk patient with very
small volumes of the required component, as slowly as possible. Generally, the
transfusion rate should not exceed 1 mL per kilogram body weight per hour
(18,48). Also, consideration should be given to administering a diuretic prior to
the transfusion of an at-risk individual. In rare cases, partial RBC exchange
(i.e., removing whole blood and replacing with packed RBCs) may be utilized in
order to reduce the patients plasma volume while increasing his or her hemo-
globin concentration.
VI. TRANSFUSION-INDUCED IRON OVERLOAD

A single unit of RBCs contains approximately 200250 mg of iron. Most
transfused patients never receive sufficient quantities of RBCs to cause iron
overload (Table 2). However, some individuals, such as those who require chronic
transfusion (e.g., thalassemic patients), may, over time, receive literally hun-
dreds of RBC transfusions, placing them at risk for developing secondary
hemochromatosis. This can potentially lead to significant organ damage, with the
end result in some patients being dysfunction of the liver, heart, kidneys, and/or
multiple other organs. There is a multifaceted approach to prevention. First, only
the minimum required number of RBC units should be transfused. Second, when
possible, pharmacological means should be used to reduce the need for RBC
transfusions. For instance, erythropoietin may be used to treat anemia in chronic
renal failure and zidovudine-treated HIV-infected patients. Third, patients who
are expected to require long-term chronic transfusions, such as those with
thalassemia, and some sickle cell anemia patients, should be treated with an
iron-chelating agent (e.g., desferoxamine). And, fourth, in certain instances (e.g.,
some sickle cell anemia and paroxysmal nocturnal hemoglobinuria patients), the
use of RBC exchange procedures is warranted, as this approach serves both to
introduce new, healthy RBCs into the patient and to remove the patients
pathological RBCs that would otherwise be rapidly hemolyzed and broken down
into additional unwanted storage iron (49). One more approach that has been used
in a few facilities is to transfuse young RBCs (neocytes), so as to provide more
longer-lived RBCs, thus minimizing the number of required transfusions (50).
Table 2 Prevention of Transfusion-Induced Iron Overload

1. Minimization of RBC transfusions
2. Pharmacological therapy (e.g., erythropoietin), when appropriate
3. Iron chelation agents (e.g., desferoxamine)
4. RBC exchange (in some cases)
5. Neocyte transfusions (if available)
VII. HYPOTHERMIA AND ARRHYTHMIAS RELATED

TO TRANSFUSIONS
The rapid transfusion of large volumes of cold blood may cause ventricular
arrhythmias, especially in a patient who is predisposed to arrhythmias or whose
central venous catheter (CVC) is positioned in close proximity to the cardiac
conduction system. Various means for treatment and prevention exist. These
include pulling the CVC slightly back, reducing the rate of infusion (not always
possible), using ultra-efficient blood warmers that are capable of rapidly warming
(without overheating) large volumes of blood, and adding warmed saline directly
to the RBC component prior to transfusion.
Another cause of arrhythmias is the transfusion of hyperkalemic RBC units.
Over time, the red cell storage lesion contributes to increasingly elevated levels
of potassium in the supernatant plasma/anticoagulant/preservative. The leakage
rate is further increased by irradiation of the component. Although hyperkalemia
is rarely a problem for most transfusion recipients (because of rapid uptake of
extracellular potassium by RBCs after transfusion, as well as dilution in the
circulation), it can, on occasion, cause morbidity, and even mortality, particularly
in premature infants and newborns, massively transfused patients, and acidotic
patients (51,52). Cardiac manifestation of hyperkalemia can include bradycardia
(which can progress to asystole), complete heart block, and ventricular fibrilla-
tion (73). Generally, no special prophylactic approach is needed to prevent these
reactions; however, when transfusing relatively large volumes of RBCs to sus-
ceptible patients, many authorities recommend the use of fresh (<710 days old)
packed RBCs, or, if fresh components are unavailable, saline-washed packed
RBCs (51).
VIII. TRANSFUSION-ASSOCIATED CITRATE TOXICITY

When large volumes of blood are transfused rapidly (e.g., at rates exceeding
100 mL/min), the result may be increased plasma citrate levels in the recipient.
These may produce transient hypocalcemia, with resultant perioral paresthesia,
chills, nausea, vomiting, muscular twitching, chest pressure, and, finally, tetany
(especially if there is concomitant hyperkalemia). Treatment and prevention are
aimed at either slowing the rate of transfusion (not always possible) or, in certain
instances, administering intravenous calcium solutions. When the latter method
is used, it is important to (a) avoid infusing so much calcium as to cause iatrogenic
hypercalcemia; and (b) administer the calcium through a line other than the one
through which the transfusion is being administered (so as to prevent the calcium
from causing the blood to clot in the line).
IX. TRANSFUSION-ASSOCIATED AIR EMBOLISM

Air emboli related to conventional transfusion are virtually never seen anymore
now that glass bottles are no longer used for the storage of blood components,
although it has been reported in association with intraoperatively or post-
operatively recovered blood (54). However, when blood is transfused under
pressure while using an open system, it is possible for air to enter the system and
cause significant morbidity, and possibly mortality. Likewise, if air enters while
containers or blood administration sets are being changed, an air embolism may
result. Symptoms include coughing, dyspnea, chest pain, and shock. Treatment is
to place the patient on his left side with his head down. If this is not effective,
intracardiac aspiration (to remove the air bubble) may be necessary.
X. HYPOTENSIVE PLATELET TRANSFUSION REACTIONS

In 1993 the AABB was alerted by one of its members that several clusters of
severe hypotensive reactions had been associated with the transfusions of leuko-
cyte-reduced (filtered) platelets. These appeared to be unrelated to the anaphylac-
tic reactions described in Section III.B of this chapter. Hume et al. surveyed all
AABB institutional members and looked in detail at 17 reactions that had similar
findings (i.e., primarily characterized by hypotension and occurring in association
with platelet transfusions). Some 88% of these reactions occurred within one hour
of beginning the transfusions; 82% were associated with respiratory distress; 82%
rapidly resolved after stopping the transfusion; and 88% of the implicated platelet
units had been given through leukocyte reduction filters (55). While this uncon-
trolled study lends some support for the existence of hypotensive reactions to
platelet transfusions, it would also be important to look for these reactions
during/after transfusions of other components before ascribing them to platelet
transfusions alone (i.e., no detailed search for hypotensive reactions related to,
say, RBC or plasma transfusions has yet been made). Also, other possible causes
for these reactions (unrelated to transfusion) must be ruled out. For instance, 2 of
the 17 hypotensive reactions in this study occurred in patients who were taking
angiotensin-converting enzyme (ACE) inhibitors, which are known to be associ-
ated with hypotensive reactions during therapeutic apheresis with albumin re-
placement (56). Moore, in his editorial accompanying Humes article, therefore
asks, Do we know enough about all medications taken by all of these patients
to be sure that the concomitant administration of some other medications or
combination of medications might not have an even stronger association (than
ACE inhibitors) with hypotensive reactions? (57). In other words, clinical and
basic science research must be performed before we can say whether or not
hypotensive platelet transfusion reactions truly exist, or are merely the by-
Table 3 Treatment of Some Adverse Consequences of Transfusions
Type of reaction Recommended treatment
FNHTRs Stop transfusion; administer acetaminophen 5001000 mg

PO; do not restart transfusion with same unit
Allergicurticarial Stop transfusion; administer diphenhydramine hydro-
chloride 50 mg PO/IM/IV; rarely, hydrocortisone will also
be needed; if reaction is mild and symptoms subside,
transfusion may be restarted
Allergicanaphylactic Stop transfusion; depending upon severity of anaphylaxis,
administer epinephrine, aminophylline, fluid and vaso-
pressor support (e.g., colloid/crystalloid and dopamine);
hydrocortisone, diphenhydramine, oxygen, and airway
protection (see text for details) may also be necessary.
Circulatory overload Stop transufsion; sit patient up as much as possible and
administer diuretic (e.g., furosemide 40 mg PO/IV) and
oxygen; phlebotomy, with concomitant slow transfusion of
packed RBCs (or partial RBC exchange), may rarely be
necessary
Citrate toxicity Slow down or temporarily stop transfusion; oral or
parenteral calcium salts (e.g., calcium gluconate 1 g IV)
may occasionally be needed
FNHTRs = Febrile, nonhemolytic transfusion reactions.
products of other types of transfusion reactions, or are simply caused by

confounding factors.
XI. TRANSFUSION-ASSOCIATED RED EYE SYNDROME

In 1997/1998, a series of unusual reactions was reported to the Centers for
Disease Control and Prevention (CDC) (58). Collectively, they were referred
to as transfusion-associated red-eye syndrome. The name derives from the
severe conjunctival erythema and/or conjunctival hemorrhage that all affected
patients experienced, as well as the fact that all the reactions occurred within
24 hours of red blood cell transfusion. The first case was reported in late
1997. As of early 1998, 14 states had reported a total of 106 similar reactions in
74 patients.
CDC looked closely at a subset of 49 of these reactions involving 38
patients in Michigan, Oregon, and Washington. The patients ranged in age from
28 to 84 years (median age: 59 years). Twenty-two (58%) were male, and all were
diagnosed previously with an hematological or oncologic illness. All of the
reactions manifested between one and 24 hours of the initiation of transfusion

(median: 20 hours) and were characterized by: severe conjunctival erythema
and/or conjunctival hemorrhage (100%); eye pain (62%); headache (25%); peri-
orbital edema (23%); arthralgias (19%); nausea (15%); dyspnea (6%); and rash
(6%). Time to resolution ranged from 2 to 21 days (median: 5 days). Intriguingly,
all 38 patients had been transfused with leukocyte-reduced red blood cells.
Moreover, it was discovered that, in 45 of 46 reactions for which information was
available, the prestorage system used to leukoreduce the (presumably) offending
red blood cell units was the same. The mechanism by which these reactions
occurs has not been deduced; however, prime suspects include an allergic
response to an unknown allergen inor a toxic reaction to a chemical or material
used to producethe collection-filtration system (Table 3).
REFERENCES
1. 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; 33:794797.
2. Heddle NM, Klama L, Singer J, Richards C, Fedak P, Walker I, Kelton JG. The role
of plasma from platelet concentrates in transfusion reactions. N Engl J Med 1994;
331:625628.
3. Payne R. The association of febrile transfusion reactions with leuko-agglutinins. Vox
Sang 1957; 2:233241.
4. Litty C. A review: Transfusion reactions. Immunohematology 1996; 12:7279.
5. Meryman HT, Hornblower M. The preparation of red cells depleted of leukocytes.
Transfusion 1986; 26:101106.
6. Mangano MM, Chambers LA, Kruskall MS. Limited efficacy of leukopoor platelets
for prevention of febrile transfusion reactions. Am J Clin Pathol 1991; 95:733738.
7. Dzik WH. Is the febrile response to transfusion due to donor or recipient cytokines?
8. Muylle L, Joos M, Wouters E, De Bock R, Peetermans ME. Increased tumor necrosis
factor alpha (TNF-alpha), interleukin 1 (IL-1), and interleukin 6 (IL-6) levels in the
plasma of stored platelet concentrates: relationship between TNF-alpha and IL-6
levels and febrile transfusion reactions. Transfusion 1993; 33:195199.
9. Ferrara JLM. The febrile platelet transfusion reaction: a cytokine shower. Transfusion
1995; 35:8990.
10. Aye MT, Palmer DS, Giulivi A, Hashemi S. Effect of filtration of platelet concentrates
on the accumulation of cytokines and platelet release factors during storage. Trans-
fusion 1995; 35:117124.
11. Stack G, Snyder EL. Cytokine generation in stored platelet concentrates. Transfusion
1994; 34:2025.
12. Federowicz I, Barrett BB, Anderson JW, Urashima M, Popovsky MA, Anderson KC.
A characterization of reactions after transfusion of cellular blood components that are
white cell reduced before storage. Transfusion 1996; 36:2128.
13. Muylle L, Wouters E, Peetermans ME. Febrile reactions to platelet transfusion: the
effect of increased interleukin 6 levels in concentrates prepared by the platelet-rich
plasma method. Transfusion 1996; 36:886890.
14. Widmann FK. Controversies in transfusion medicine: Should a febrile transfusion
reaction occasion the return of the blood component to the blood bank? Pro.
Transfusion 1994; 34:356358.
15. Oberman HA. Controversies in transfusion medicine: Should a febrile transfusion
response occasion the return of the blood component to the blood bank? Con.
Transfusion 1994; 34:353355.
16. Ryden SW, Oberman HA. Compatibility of common intravenous solutions with CPD
blood. Transfusion 1975; 15:250255.
17. Kevy S, Schmidt PJ, McGinniss MH, Workman WG. Febrile nonhemolytic trans-
fusion reactions and the limited role of leukoagglutinins in their etiology. Transfusion
1962; 2:716.
18. Vengelen-Tyler V, ed. Noninfectious complications of blood transfusion. In: Techni-
cal Manual. 12th ed. Bethesda, MD: American Association of Blood Banks,
1996:543562.
19. Vesilind GW, Simpson MB, Shifman MA, Colman RE, Kao KJ. Evaluation of a
centrifugal blood cell processor for washing platelet concentrates. Transfusion 1988;
28:4650.
20. Silvergleid AJ, Hafleigh EB, Harabin MA, Wolf RM, Grumet FC. Clinical value of
washed-platelet concentrates in patients with non-hemolytic transfusion reactions.
21. Sandler SG, Eckrich R, Malamut D, Mallory D. Hemagglutination assays for the
diagnosis and prevention of IgA anaphylactic transfusion reactions. Blood 1994;
84:20312035.
22. Poothullil J, Shimizu A, Day RP, Dolovich J. Anaphylaxis from the product(s) of
ethylene oxide gas. Ann Intern Med 1975; 82:5860.
23. Branch DR, Gifford G. Allergic reaction to transfused cephalothin antibody. JAMA
1979; 241:495496.
24. Stern A, van Hage-Hamsten M, Sondell K, Johansson SGO. Is allergy screening of
blood donors necessary? Vox Sang 1995; 69:114119.
25. Michel J, Sharon R. Non-hemolytic adverse reaction after transfusion of a blood unit
containing penicillin. Br Med J 1980; 1:152153.
26. Take H, Tamura J, Sawamura M, Murakami H, Naruse T, Tsuchiya, Miyawaki S,
Hirabayashi H. Severe anaphylactic transfusion reaction associated with HLA-
incompatible platelets. Br J Haematol 1993; 83:673674.
27. Sano H, Koga Y, Hamasaki K, Furuyama H, Itami N. Anaphylaxis associated with
white-cell reduction filter. Lancet 1996; 347:1053.
28. Bjerrum OJ, Jersild C. Class-specific anti-IgA associated with severe anaphylactic
transfusion reactions in a patient with pernicious anemia. Vox Sang 1971; 21:411424.
29. Pineda A, Taswell HF. Transfusion reactions associated with anti-IgA antibodies:
Report of four cases and review of the literature. Transfusion 1975; 15:1015.
fusion 1990; 30:583590.
31. Lippman M, Rumley W. Medical emergencies. In: Dunagan WC, Ridner ML, eds.
Manual of Medical Therapeutic. 26th ed. Boston: Little, Brown and Company,
1989:482503.
32. Shirey RS, Ness PM. Alloimmunization to blood group antigens. In: Anderson KC,
Ness PM, eds. Scientific Basis of Transfusion Medicine: Implications for Clinical
Practice. Philadelphia: W.B. Saunders Company, 1994:507516.
33. Rosse WF, Gallagher D, Kinney TR, Castro O, Dosik H, Moohr J, Wang W, Levy
PS, and the Cooperative Study of Sickle Cell Disease. Transfusion and alloimmuniza-
tion in sickle cell disease. Blood 1990; 76:14311437.
34. Hoeltge GA, Domen RE, Rybicki LA, Schaffer PA. Multiple red cell transfusion and
alloimmunizationexperience with 6996 antibodies detected in a total of 159,262
patients from 1985 to 1993. Arch Pathol Lab Med 1995; 119:4245.
35. Lostumbo MM, Holland PV, Schmidt PJ. Isoimmunization after multiple trans-
fusions. N Engl J Med 1966; 275:141144.
36. Coles SM, Klein HG, Holland PV. Alloimmunization in two multitransfused patient
populations. Transfusion 1981; 21:462466.
37. Wayne AS, Kevy SV, Nathan DG. Transfusion management of sickle cell disease.
Blood 1993; 81:11091123.
38. Diamond WJ, Brown FL, Bitterman P, Klein HG, Davey RJ, Winslow RM. Delayed
hemolytic transfusion reaction presenting as sickle-cell crisis. Ann Intern Med 1980;
93:231234.
39. Milner PF, Squires JE, Larison PJ, Charles WT, Krauss JS. Posttransfusion crisis in
sickle cell anemia: role of delayed hemolytic reactions to transfusion. Southern Med
J 1985; 78:14621469.
40. The Trial to Reduce Alloimmunization to Platelets Study Group. Leukocyte reduction
and ultraviolet B irradiation of platelets to prevent alloimmunization and refractori-
ness to platelet transfusions. N Engl J Med 1997; 337:18611869.
41. Williamson LM, Wimperis JZ, Williamson P, Copplestone JA, Gooi HC, Morgenstern
GR, Norfolk DR. Bedside filtration of blood products in the prevention of HLA
alloimmunizationA prospective randomized study. Blood 1994; 83:30283035.
42. von dem Borne AEGK, Simsek S, van der Schoot CE, Goldschmeding R. Platelet
and neutrophil alloantigens: their nature and role in immune-mediated cytopenias. In:
Garratty G, ed. Immunobiology of Transfusion Medicine. New York: Marcel Dekker
Inc., 1994:149171.
43. Puig N, de Haas M, Kleijer M, Montoro JA, Perez A, Villalba JV, Gomez I, von dem
Borne AEJK. Isoimmune neonatal neutropenia caused by FcgammaRIIIb antibodies
in a Spanish child. Transfusion 1995; 35:683687.
44. Van Buren NL, Stroncek DF, Clay ME, McCullough J, Dalmasso AP. Transfusion-
related acute lung injury caused by an NB2 granulocyte-specific antibody in a patient
with thrombotic thrombocytopenic purpura. Transfusion 1990; 30:4245.
45. Popovsky MA, Chaplin HC, Moore SB. Transfusion-related acute lung injury: a
neglected, serious complication of hemotherapy. Transfusion 1992; 32:589591.
46. Popovsky MA, Audet AM, Andrzejewski Jr. C. Transfusion-associated circulatory
overload in orthopedic surgery patients: a multi-institutional study. Immunohemato-
logy 1996; 12:8789.
47. Popovsky MA, Taswell HF. Circulatory overload: an underdiagnosed consequence of
transfusion. Transfusion 1985; 25:469.
48. Marriott HL, Kekwick A. Volume and rate in blood transfusion for the relief of
anemia. Br Med J 1040; 1:10431046.
49. Kim HC, Dugan NP, Silber JH, Martin MB, Schwartz E, Ohene-Frempong K, Cohen
AR. Erythrocytapheresis therapy to reduce iron overload in chronically transfused
patients with sickle cell disease. Blood 1994; 83:11361142.
50. Propper RD, Button LN, Nathan DG. New approaches to the transfusion management
of thalassemia. Blood 1980; 55:5560.
51. Hall TL, Barnes A, Miller JR, Bethencourt DN, Nestor L. Neonatal mortality
following transfusion of red cells with high plasma potassium levels. Transfusion
1993; 33:606609.
52. Brown KA, Bissonnette B, MacDonald M, Poon AO. Hyperkalemia during massive
blood transfusion in paediatric craniofacial surgery. Can J Anaesth 1990; 37:401408.
53. Kohan DE. Fluid and electrolyte management. In: Dunagan WC, Ridner ML, eds.
Manual of Medical Therapeutics. 26th ed. Boston: Little, Brown and Company,
1989:5271.
54. Linden JV, Kaplan HS, Murphy MT. Fatal air embolism due to postoperative blood
recovery. Anesthesia Analgesia 1997, 84:422426.
55. Hume HA, Popovsky MA, Benson K, Glassman AB, Hines D, Oberman HA,
Pisciotto PT, Anderson KC. Hypotensive reactions: a previously uncharacterized
complication of platelet transfusion? Transfusion 1996; 36:904909.
56. Owen HG, Brecher ME. Atypical reactions associated with use of angiotensin-
converting enzyme inhibitors and apheresis. Transfusion 1994; 34:891894.
57. Moore AB. Hypotensive reactions: Are they a new phenomenon? Are they related
solely to transfusion of platelets? Does filtration of components play a role? Trans-
fusion 1996; 36:852853.
58. Adverse ocular reactions following transfusionUnited States, 19971998. MMWR
1998; 47:4950.
4
Managing Error for
System Improvement
Harold S. Kaplan
New York-Presbyterian Hospital, New York, New York
James B. Battles
University of Texas, Southwestern Medical Center, Dallas, Texas
The safety of blood transfusion has long been an area of concern and study (1).
However, with the intense public and legislative scrutiny following the discovery
of the potential for human immunodeficiency virus (HIV) transmission by blood
transfusion, there has been an exceptional increase in the investigative and
regulatory interest directed at a broad range of blood safety issues.
In an effort to minimize undesirable variations in the manufacturing
aspect of transfusion medicine, i.e., the recruitment, collection, testing, and
processing of units of donated blood, there is an increased regulatory emphasis
on rigorous process control coupled with training to protocols of standard
operating procedures (2). This approach, contained within the Current Good
Manufacturing Practices (CGMPs), clearly contributes to more reliable processes.
However, even with the long-established CGMPs in the biopharmaceutical man-
ufacturing industry, there is a documented persistence of human error even with
strict adherence to CGMPs (3). Similar problems are seen with the process of
transfusion to the intended recipient.
87
88 Kaplan and Battles
I. METHODS TO IDENTIFY ERROR

There are a number of methods to identify and reduce the contribution of human
error to the risks of transfusion.
A. Observation/Audit
One approach has been direct observation by skilled observers in the actual
operating environment. This is a well-established way to address human error in
its actual setting (4). Shulman and coworkers (5) developed a multidisciplinary
team approach to quality assurance and improvement directed at reducing pa-
tient identification errors by improving compliance with standard procedures.
The program involved periodic concurrent audits, which included direct observa-
tion of procedures. With feedback of deviations from protocol and active educa-
tional efforts, adherence to patient identification protocols improved gradually
from 50% during a pilot study to nearly 100% by the 125th audit (5). Although
the authors found in-service education to be effective in increasing compliance
with protocol and reducing the risk of error, observation may itself alter the
circumstances studied, as may observer error and limitation in controlling all
relevant variables. In addition, the enduring effect of such improvement may be
difficult to maintain given employee turnover and the need for sustained inten-
sive effort.
B. Accident Analysis
The second approach to error identification and prevention is the analysis of
accident data. This approach has been an important source of information used
by the National Transportation Safety Board (NTSB) in aviation (4). However,
hindsight bias and incomplete data lead to distortion. Additionally, although
human error has often been identified as a cause of an accident, little insight has
been provided as to why an error occurred.
Despite these limitations, analysis of accident data has been an important
source of information in blood transfusion. In 1975, the Food and Drug Admin-
istration (FDA) first mandated the reporting of transfusion-associated fatalities,
and by 1980, the Health Care Finance Administration agreed to investigate these
fatalities in greater depth. The file of these reports is accessible under the Freedom
of Information Act. This database has been extensively studied by a number of
investigators (610).
Mummert and Tourault (11) reviewed 150 transfusion-associated fatalities
reported to the FDA from 1990 to 1992. They concluded that nearly one third of
these fatalities could have been prevented by adherence to proper procedure.
Interestingly, a failure to follow procedures is also responsible for one third of
Managing Error for System Improvement 89
major air carrier accidents. However, as pointed out by Nagle (4), even with
categorization of error data, if it is not known why someone failed to follow
standard procedures, development of an effective preventive strategy remains
problematic. In this regard, Nagle has stressed the need for a model of human
error to be used in conjunction with error data collection and classification.
Mummert and Tourault also noted that improper transfusion of ABO-
incompatible red blood cells due to error continues to be a primary cause of
preventable death. They also reported that failure to identify a reaction in progress
contributed to many of the fatalities. Again, in analogy with other error-critical
areas, such as nuclear power and aviation, delay in detecting a problem, im-
properly identifying its cause, and delay in implementing corrective actions are
recognized as critical issues (12).
The management of error to limit adverse outcomes or unplanned effects is
now recognized to be of fundamental importance in system design and training
in error-critical activities, particularly since the Three Mile Island accident (12).
Although the importance of this has also been appreciated in transfusion medi-
cine, its proper emphasis has not been carried through to all critical operational
areas. In some cases analyzed by Mummert and Tourault (11), signs or symptoms
were treated, but the transfusion was not identified as the cause and was
continued. These authors also reported that in several cases signs such as
hemoglobinuria were noted without being recognized as a transfusion reaction.
In an extensive analysis of fatalities reported to the FDA over a 10-year
period, Sazama (10) reviewed 355 reports and studied 256 (99 did not involve
transfusion, 68 related to transfusion-associated hepatitis, 3 related to transfusion-
associated acquired immune deficiency syndrome, and one report recorded could
not be located). Acute hemolytic transfusion reaction resulting from ABO incom-
patibility accounted for the majority of fatalities. Sazama found one third of all
deaths and two thirds of incompatible red cell transfusions to be attributable to
error and therefore preventable. She estimated that 124 fatal ABO errors occurred
in approximately 100,000,000 transfusions, or 1/800,000 units. Sazama also
reported that 5 of the 26 hepatitis B deaths resulted from error, with three
seropositive units labeled properly but released in error, and two seropositive
units labeled improperly as negative. Additionally, of 12 reported donor deaths,
one related to an O-positive plasma donor who received A red cells from another
donor. Sazama determined that errors leading to fatality were most often mana-
gerial or system errors, rather than isolated human error.
In a 2-year period from 1990 through 1991, transfusion-related incidents
reported to the New York State Department of Health (NYSDOH) also included
errors that as precursor events could have but did not necessarily result in harm
(13). The reported incidence of administration of the wrong unit of blood, or of
blood given to the wrong recipient, was estimated at 1/12,000 units transfused.
Of these mistakes, ABO-incompatible red cell transfusions had an estimated
incidence of 1/33,000. Three of the ABO-incompatible transfusions resulted in

death, yielding a fatal transfusion error rate of 1/600,000 units (13).
The importance of management and system contributions to human error
identified in the 10-year study of FDA reports (10) and in the NYSDOHs
demonstration of the importance of benign precursor events (50 times greater
occurrence rate than errors resulting in fatal accidents) (13) parallel the experi-
ence in other error-critical fields (4).
C. Simulation
A third method of error analysis for development of prevention strategies is the
study of error in the laboratory and by simulation. This approach has also proven
useful since laboratory circumstances allow simplification and control of con-
founding variables. Simplification, however, may itself be an important shortcom-
ing in understanding inherently complex situations (4).
The laboratory approach has not been generally applied to transfusion error
analysis, although the announced introduction of simulated benign errors into
transfusion operations has been an effective means for increasing error detection.
Taswell et al. (14) demonstrated that by modifying work to demand staff attention
in looking for known introduced errors and by providing positive feedback when
they were found, the blood bank not only achieved an increased detection of the
introduced errors, but also increased the detection of real, previously undetected
errors from 4 in the first 3 months to 73 in the final 3 months of the study.
D. Record Review/Chart Audit

A fourth approach to identifying errors is to review records, including donor and
laboratory records and patient charts. The review of such records has been the
most traditional means of performing quality assurance checks and documenting
patient outcome. The chart or record is a documentation of actions performed or
missing information. This auditing of charts or records against predetermined
criteria can be a valuable method of identifying errors and near-miss events.
Classen and colleagues (15,16) have successfully used a sophisticated automated
hospital information and record system to identify adverse drug events that
would have otherwise gone unreported. The limitation of record review and chart
audit is determined by whether information necessary to detect an error is part of
the record.
E. Event Reporting
A fifth approach to compiling information for the study of error is the event
report, including self-reporting as exemplified by the Aviation Safety Reporting
System (ASRS) operated by NASA for the Federal Aviation Administration

(FAA) (17). Commercial airline accidents are very rare events, with approxi-
mately one accident (defined as aircraft hull loss or by one or more fatalities) per
million departures by U.S. carriers (4). Despite this excellent record, human error
is identified as the causal factor in more than one half of all airline accidents and
as much as 90% in general aviation (4). The ASRS has been operational for
almost two decades. More than 250,000 reports (18) have been archived, ana-
lyzed, and made available for research and study by interested professionals as
well as by regulatory and investigative bodies. An advisory group representing
the major U.S. aviation operations organizations oversees and monitors the
ASRS. This a no-fault, confidential, voluntary, self-reporting system, in which
pilots and controllers report noncalamitous mistakes, including caught or trapped
errors, occurrences/situations. The ASRS philosophy recognized that post hoc
analysis (which is what the study of incident reports necessarily is) cannot prove
causationit can only observe a significant association of certain possibly causal
factors with a class of occurrence (4).
Confidentiality and immunity from prosecution for noncriminal acts are
important features of this or any system intended to capture operational data.
There is no disincentive to report, and this approach also optimizes access to
information from the incident reporters themselves. The no-fault confidential
nature of this system has led to an increased frankness in reporting and complete-
ness of invaluable data (15). This eminently successful voluntary reporting system
has provided a rich database from which valuable improvements in safety have
been made possible (15).
One drawback of the voluntary reporting system is the variability of
reporting by different individuals. This sampling variability makes any quantifi-
cation problematic and leads to more reliable qualitative than quantitative data.
There is a better idea of what errors occur, but relatively little is known about
their frequency.
The preponderance of benign error in transfusion medicine leads to a
deceptively low morbidity from failure to follow safe practice (19). This obscures
the frequency and potential of these errors for hemolytic reaction and increased
risk of disease transmission. Even though such errors are an important source of
information for improving transfusion safety, there has been no means for their
systematic collection and analysis. The study of the benign or caught errors
(i.e., potential errors trapped by the system) could provide a rich database for
improving the safety of the blood supply (20). As demonstrated in studies of
safety incidents in commercial aviation, these near-miss events are very similar
to those associated with full-blown disasters. Since there are many more near
misses than there are events with major adverse outcomes, there is a need to
collect near-miss event data in addition to accident data (21). The relationship
between near misses and accidents has been compared to a pyramid or iceberg
(22), which represents a continuum from the rare visible accident to the much
more frequent near misses. Linden et al.s (13) data from New York State can
illustrate this relationship with the iceberg in Figure 1. The fatalities from
hemolytic reactions due to ABO-incompatible blood transfusions are the visible
tip of the iceberg, the number of adverse reactions reported as being just below
the waterline, the number of incorrect units transfused by nurses as next, and the
unreported near misses as unknown.
A fundamental aspect of any event-reporting system is the identification or
detection of events that occur within an organization. Zapt and Reason (23)
indicate that error detection is the first step in error management. If an error
is not detected, it cannot be managed. They go on to point out that, from an
organizational point of view, it is very important that the error detection rate
Figure 1 The iceberg model of accidents and near-miss events.

is high; errors that are not detected for a long time could have disastrous
consequences. Thus, the goal of error management should be to increase error
detection or reporting rate. The number of events reported is an indicator of
an organizations detection sensitivity level (DSL). High reporting rates indi-
cate a high DSL; few reported events indicate a low DSL. A low DSL might
be considered an indicator of an inadequate error-detection and reporting ap-
proach. While the DSL should remain high, the event severity level (ESL) of the
incidents detected should decrease over time as corrective actions are im-
plemented (24).
In order to achieve a high DSL, an organization must remove any im-
pediments to reporting an event. Confidential no-fault reporting is one of the
best ways to encourage event reporting. Since the purpose of the event-reporting
system is to learn about how systems are operating, there must be separation
between event reporting and employee performance assessment approaches.
Research literature indicates that organizational encouragement such as confiden-
tial no-fault reporting significantly increases error detection and reporting (24).
Table 1 is a review of the strengths and limitations of the five error analysis
methods previously discussed.
II. CLASSIFICATION OF EVENTS

Regardless of the method that one chooses to identify error, there is a need to
classify them once they are discovered. Just what are the elements of events, both
near misses and major misadventures, that should be studied? In essence all
events are nested within the context of what happened, where in the process it
occurred, when it happened, and who was involved in the event. Although it is
necessary to investigate all of these elements of an event for complete understand-
ing, most existing approaches to event analysis concentrate on describing what
happened. Little information is found as the causes or description of why the event
occurred. A review of 200 cases obtained from reports submitted to the FDA
found an almost total lack of information regarding the root causes of the events.
The lack of in-depth investigation of the root causes of transfusion events could
indicate that the corrective actions that are taken may be inappropriate or
unrelated to the actual cause of the event (24).
A. Causes of Errors
Reason (12) has identified two major categories of failures or errors that occur in
complex systems: active and latent. The distinction hinges on both who initiated
the failure and how long it took to have an adverse effect (25,26). Active failures
are committed by people in direct contact with the human/system interface, i.e.,
Table 1 Methods for Identifying Errors
Method Advantage Limitations
Observation/Audit Accurate method to identify per- Labor intensive and limited

formance of individuals com- lasting value with employ-
pared to standards ee turnover
Accident analysis Used to determine error rates and Limited by hindsight and de-
understanding major misadven- lays in investigation; often
tures and is required to be per- lack of information as to
formed when such events occur causes
Simulation An accurate way to assess team Difficult to create and can
and individual performance only simulate some aspects
and decision making; has the of the transfusion process
ability to introduce known
errors or events to be
identified and managed
Record review/ Standard method of documenting Only as good as the informa-
Chart audit outcomes and many processes; tion recorded; always after
normal activity which can pro- the fact
vide valuable information
Event reporting Captures near-miss events and Subject to underreporting be-
deviations as perceived by indi- cause individuals do not
viduals; gives perspective of feel comfortable reporting
individuals involved errors
health professionals. The consequences of these failures are usually readily

apparent almost immediately. It is the active failure that we most often associate
with human error. Latent failures are the delayed-action consequences of techni-
cal design or organizational issues and decisions. These latent failures often are
initiated at the upper levels of an organization. Accidents or major misadventures
with adverse outcomes occur when latent errors or system considerations com-
bine with an active human error, as illustrated in Figure 2. Error researchers stress
the importance of examining both human or active failures as well as the
underlying latent or system failures.
B. Human Active Failures

As discussed previously, active or human failures are most commonly associated
with human error. These errors are associated with the individuals who are at what
Reason (12) calls the sharp end of the system. In the case of transfusion
medicine, these are the people responsible for collecting, processing, and trans-
Figure 2 The relationship between latent and active failures and misadventures.
fusing blood products. We can moderate human fallibility, but we can never
eliminate it completely. Active errors are tied to human cognitive processes
and associated behaviors. Rasmussen (27,28) gives us a useful taxonomy for
identifying and classifying these different types of human behavior underlying
human errors:
Skill-based behavior refers to routine tasks requiring little or no conscious
attention during task execution.
Rule-based behavior refers to familiar procedures applied to frequent
decision-making situations.
Knowledge-based behavior refers to problem-solving activities such as
when one is confronted with new situations for which no readily avail-
able standard solutions exist.
1. Skill-Based Behavior
Most health professionals operate in a skill-based behavioral mode for all of the
routine tasks that are carried out, from drawing blood to the transfusion process
itself. These highly skilled activities become routine and can be executed without
conscious thought. It is like riding a bicycleonce you learn you can do it
automatically. Driving a car is another examplewe can drive while listening to
the radio or talking to a passenger. It is because these skills are so often used that
one can perform with a very high level of reliability. We operate almost as if
we were on automatic pilot, performing virtually without thinking about what
we are doing, at almost an unconscious level. However, there are opportuni-
ties for failures when one operates in the skill-based mode. If we are distracted
or something interrupts the smooth flow of a skill-based routine, a skill-based
failure can occurcalled a slip. An example of a slip is being distracted by
someone in the parking lot and inadvertently locking the keys in the car or
arriving home only to discover that you forgot to stop at the store and buy milk
on the way home. Slips are unintended errors caused by either not doing some
part of the routine, an error of omission, or doing something that one should not
have done, like a repeating step in the process twice, which would be a failure of
commission. Slips cannot effectively be remediated by retraining an individual.
It is a waste of time because the individual already knows how to perform the
task at a very high level of accuracy and retraining is often insulting and
ineffective. Counseling employees to be more careful is equally ineffective as a
means of remediation. However, slips can be prevented by redesign of equipment
or procedures so that it is harder to make a slip. For example, feedback mecha-
nisms can be designed into the process that give clues to the individual as soon
as they have made a slip. Job aids such as a template for reviewing donor records
that highlight omissions or inconsistencies can help to prevent slips in the
skill-based mode.
2. Rule-Based Behavior
Rule-based behavior occurs at the conscious level within the context of the
situation that exists. Rule-based behavior involves recognizing the situation, then
selecting the proper routine or protocol that is called for in a given situation.
For example, if we are driving and come to a stop sign, we must decide what
rules to apply in this situation. If it is a two-way stop, then there is a given set of
traffic rules to follow, but if it is a four-way stop then there are different rules of
the road that we must follow. Rule-based behavior is an if-then condition. Failure
in rule-based behavior can occur at the different stages in this conscious decision
and action process. These failures are often referred to as mistakes. A mistake can
occur under two conditions: selecting the wrong rule for a given situation, or
selecting the correct rule but carrying it out incorrectly. Rule-based failures can
occur when someone carries out a procedure that they are not qualified to
perform. Another type of rule-based failure is inadequately assessing or verifying
the situation, which can result in a mistake in selecting the correct rule. An ex-
ample of verification would be failing to check the patients identity before

transfusion. While most rule-based failures are unintended, some are not. In some
instances an individual can consciously choose to apply a different rule or carry
out a task differently than is proscribed by standard operating procedures.
This type of action is a violation. Violations can be either routine, a work around
of an inadequate procedure, or a cultural artefact of the organizationi.e.,
everyone does it this way. Routine violations often occur when procedures are
changed and individuals continue to apply the old procedure. In rare cases an
individual may choose to freelance and carry out tasks in a manner contrary to
standard procedures. Rule-based failures are subject to remediation through
training in many instances. In addition they can be reinforced through clearly
written procedures and protocols and job aids. Rule-based failures of verification
can be prevented in some cases by redesigning the task to place a forcing function,
like the blood lock that prevents a transfusion from being started until the patients
identify has been matched to the unit.
3. Knowledge-Based Behavior
Knowledge-based behavior involves solving unique problems or selecting a
plan of action in a new or unfamiliar setting. Knowledge-based behavior most
often occurs with new employees. They do not have the depth of experience to
operate in the skill-based mode or to draw from past experience to select the
appropriate rule or protocol to carry out a task or to solve a problem. Learners
such as medical residents and other trainees often operate in the knowledge-based
mode because the number of unique or new situations for them is significant
compared to the experienced individual or expert. Experienced individuals only
rarely operate in the knowledge-based mode. Thus, the expert and the novice are
likely to make different types of errors. The expert is most likely to make a slip
or an occasional rule-based error, while the novice is most likely to have
knowledge-based failures. If is possible for the expert to encounter unique
conditions and be placed in a situation where they can be subject to knowledge-
based failures. The procedures, equipment, and situation are all slightly different,
and their set of skills and rules may be inadequate for the new situation.
Examples of such conditions would be expert pilots moving from one type of
aircraft to another. The skills and rules used in operating a Boeing 737 are not the
same as for a Boeing 757. While it may not take an expert long to become
familiar with a new setting, there is a need for orientation and knowledge
transfer from the previous setting to the new one. This is why it is good practice
to have individuals recertified or credentialed when moving to a new job or
assuming new responsibilities.
Figure 3 summarizes the three levels of human behavior associated with
such common activity as driving a car.
Figure 3 Rasmussens model of human behavior.
C. Latent Failures
While we may never totally eliminate human or active errors, we can eliminate
the technical or organizational aspects that might set the person up for an active
failure. Latent or system failures include both technical and organizational
aspects. The technical aspects associated with latent failure include such things
as the design of equipment and software, the construction of facilities, including
mobiles, and materials. One aspect of organizational failures stems from normal
management considerations including the structure of the organization, planning
and scheduling, forecasting, budgeting, and allocating resources. The policies and
procedures in place in an organization can also be a source of latent failure, as
are the orientation, training, and selection of employees. The informal but very
real culture of an organization can be another source of latent failure. These latent
failures have the potential of setting up the individuals for failure. The adverse
safety consequences of normal technical and organizational decisions may lie
dormant for a very long time. Reason (12) has referred to latent error as
organizational pathogens, which wait to combine with the right active human
failure to have an adverse consequence.
It is difficult to recognize a latent error before the fact and to predict how
it will present itself in the future, since there are so many possible outcomes of
technical and management decisions. However, since an event represents a fixed
outcome, one is often able to identify the responsible latent error by reasoning
backward from the event. Once latent failures are recognized, they can be
diagnosed and corrected before they combine with an active error to produce a
bad outcome.
III. ASSIGNING BLAME

One of the major problems with the management of error is detecting that an error
has occurred. This is particularly true if one tries to move from dealing with only
the rare visible events to major adverse consequences. Perper (29) has noted that
there is substantial underreporting of medical misadventures, which is in part
attributable to the very strong tendency in the health care field to blame the
individual or individuals associated with an active failure, the individuals at the
sharp end of an error chain. This produces a climate where individuals are
reluctant to report events where there may be an adverse consequence to report-
ing, i.e., losing ones license to practice. As Reason (25) has pointed out, blaming
people is universal, natural, and emotionally satisfying. In addition, in a litigious
environment it is easier to blame the person involved. In fact, the willingness of
individuals to accept blame for their actions is almost universal among health-care
professionals.
Health professionals all share one common professional focus, and that is
to take personal responsibility for applying their particular skills to solve the
patients health problems either directly or indirectly. This sense of personal
responsibility and accountability for patient care is ancient, stemming in the
Hippocratic oath that physicians recite to this day on graduation. Nurses are at
the sharp end in the transfusion process; as Curtain (30) has stated, in the end,
nurses are the patients last line of defense against system errors. Curtain goes
on to relate a statement made by one of her instructors in nursing school: and
you, you alone, stand responsible for it (giving the correct medication)before
God, the patient and the state board of nursing. Since the consequencecausing
harm or failing to protect the patienthas such potentially grave outcomes,
including death, a physician or nurse can lose his or her license to practice and,
in rare cases, may face criminal charges. In the health care field the norm is to
expect perfect performance at all times; anything less could be considered as
being careless or negligent. This sense of perfectionism has become a professional
norm and has been codified in law.
There are some obvious unwanted consequences of perfectionism in terms
of overall safety and event reporting. One consequence is a personal sense of guilt
when patient harm has occurred. This guilt focuses on ones own professional
behavior and diminishes ones ability to find underlying system or latent causes
for an error. Within the organization there is a desire to focus event investigations
on finding the guilty party(s) who were negligent and committed the error.
The system searches for those negligent or less than perfect health professionals
to eliminate them and, in so doing, reduce error from the system. Human resource
policies encourage or mandate recording the number of errors or adverse events
within an employees personnel record. Such policies can lead to denial that errors
have occurred. If an error is caught and is not revealed and if no harm was caused
to the patient, then it is rationalized that it did not matter and need not be revealed.
Both conditions tend to result in underreporting of events. Despite the illusion of
perfect performance, Paget (31) points out that health care is error-ridden, inexact,
uncertain, and is practiced on the human body.
Reason (25) points out that the blaming individuals leads to ineffective
countermeasures: disciplinary action, exhortations to be more careful, retraining,
and writing new procedures to proscribe those actions implicated in some recent
event. He goes on to point out that these measures can have an impact at the outset
of some necessary safety program, but they have little or no value when applied
to a well-qualified and highly motivated work force. Figure 4 illustrates the issue
of assigning blame.
Figure 4 Blaming traditions and consequences.

IV. DEVELOPING A SAFETY CULTURE

To avoid the blaming of individuals and its negative consequences, both Berwick
(32) and Lucas (33,34) have emphasized the importance of creating an environ-
ment within an organization where events can be reported in an open and free
manner. It is essential that there be no adverse consequences applied to those
submitting reports of events. In order for this to be accomplished, there should be
a decoupling of discovery error from individual employee performance. Error
management efforts should be directed at learning how the system actually
operates as opposed to how management thinks it is operating. Therefore, a
separation between event reporting and employee assessment should be made.
Everyone in the organization should be encouraged to report events that have the
potential for adverse outcome for products or patient/donor safety. It is important
that feedback be provided to blood center or hospital employees on any changes
that result from events reported. Such information is essential for continued
reporting and for employees to feel that they have a degree of ownership in
the system.
If there is truly a concerted effort to create a positive safety culture in the
organization and to have individuals capture all events including near-miss
events, then a significant increase in the number of known events is likely to
occur. In a medical setting in which there is a strong, positive safety culture and
employees are encouraged to voluntarily report events, there may be as much as
a 10-fold increase in reporting (35). The authors found that when an event
reporting system was established within the blood bank at a large indigent care
public hospital, this same 10-fold increase in events reported.
V. CAUSAL CLASSIFICATION MODEL

A root cause classification model has been developed for the field of trans-
fusion medicine as part of a medical event reporting system for transfusion
medicine (31). This classification approach was based on the Eindhoven Class-
ification Model (21,24,36). It has three major categories of causes, which are
grouped as (a) technical (equipment, software, and forms), (b) organizational
(policies procedures, and protocols), and (c) human causes (knowledge-based,
rule-based, and skill-based). The classification of human failures is consis-
tent with the theoretical framework of Rasmussen (27,28), and the latent techni-
cal and organizational factors are consistent with the framework of Reason
(12,25). Table 2 is the Eindhoven Classification Model, medical version for
transfusion medicine.
To illustrate how this classification model works, we have selected a
transfusion event that was reported, investigated, and diagrammed as to what
happened and for which a root cause analysis was performed.
102
Table 2 Eindhoven Classification Model for Medical Domain
Category Code Definition
1. Latent Errors Errors that result from underlying system failures

A. Technical Refers to physical items such as equipment, physical installations, software, materials,
labels, and forms
External TEX Technical failures beyond the control and repsonsibility of the investigating organization
Design TD Failures due to poor design of equipment, software, labels or forms
Construction TC Correct design was not followed accurately during construction
Materials TM Material defects not classified under TD or TC
B. Organizational
External OEX Failures at an organizational level beyond the control and responsibility of the investigating
organization
Transfer of knowledge OK Failures resulting from inadequate measures taken to ensure that situational or domain
specific knowledge or information is transferred to all new or inexperienced staff
Protocols/Procedures OP Failures related to the quality and availability of the protocols with the department (too
complicted, inaccurate, unrealistic, absent, or poorly presented)
Management Priorities OM Internal management decisions in which safety is relegated to an inferior position when
faced with conflicting demands or objectives; a conflict between production needs and
safety (e.g., decisions made about staffing levels)
Culture OC Failures resulting from collective approach and its attendant modes of behavior to risks in
the investigating organization
2. Active Errors Errors or failures that result from human behavior
A. Human
External HEX Human failures originating beyond the control and responsibility of the investigation
organization
Kaplan and Battles
Knowledge-based
Knowledge-based errors HKK The inability of an individual to apply existing knowledge to a novel situation
Rule-based
Qualifications HRQ Incorrect fit between an individuals qualification, training, or education and a particular task
Coordination HRC A lack of task coordination within a health-care team in an organization
Verification HRV Failures in the correct and complete assessment of a situation including relevant conditions
of the patient and materials to be used before starting the intervention
Intervention HRI Failures that result from faulty task planning (selecting the wrong protocol) and/or
execution (selecting the correct protocol but carrying it out incorrectly)
Monitoring HRM Failures during monitoring of process or patient status during or postintervention
Skill-based
Slips HSS Failures in performance of fine motor skills
Tripping HST Failures in whole body movements
3. Other
Managing Error for System Improvement
A. Patient-related factor PRF Failures related to patient characteristics or conditions beyond the control of staff and
influence treatment
B. Unclassifiable X Failures that cannot be classified in any other category
103
104
Kaplan and Battles
Figure 5 Photograph of a unit of red cells with an out-of-sequence transfer label.

A medical technologist on the second shift in a blood bank was releasing

units from quarantine to inventory when she noticed an out-of-sequence number
on the back of a unit of red blood cells (see Fig. 5). The unit was isolated until
the labels were corrected. It was determined that no incorrect labels had been used
either in testing or in component production, therefore no harm was done. Clearly
this event was a near miss.
Figure 6 is a causal tree diagramming this event. When classifying this
event, the first question to ask would be: Were there any technical failures? The
separation markings between different number sequence labels were not promi-
nent, providing unclear guidance as to where to tear the roll of labels in order to
separate adjacent blocks of numbers. In addition, the markings provided little
feedback when the tear had been done incorrectly (allowing little chance for
recovery from the error). Clearly, there was a design failure in the label itself,
classified as a TD. Next we looked for any organizational failures. There might
have been a failure in the procedure for checking the accuracy or consistency of
the label on the unit prior to placing it into quarantine. These procedures should
be reviewed. If the procedure is not clearly written, another contributing root
cause would be classified as OP. However, even a very clear and explicit
procedure might not be effective if the detectability of the error (feedback) is poor.
The phlebotomist made a slip by tearing separate labels at the wrong place. This
action would be classified as an HSS skill-based behavioral error. However,
without redesigning the label, this event would likely recur.
VI. SUMMARY
The safety of transfusion can be improved if one can identify errors that are an
indication of a systems weak points before they result in an adverse outcome to
a donor or a patient. Doing this requires a focus not only on adverse events, but
on the capture and recording of near-miss events as well. In order to capture
near-miss events, it is necessary for everyone in an organization to identify and
report those conditions and actions that have the potential to adversely affect
patient safety. It is also necessary to avoid assigning blame when an error is
identified, but rather to find the root causes of the error. Without an adequate
understanding of the causes of error, there is little likelihood the error can be
corrected and prevented in the future. It is essential to look for and to eliminate
those things that set up humans for failure. A useful goal is to create a safety
culture where everyone seeks to identify and report conditions that may compro-
mise transfusion safety.
106
Kaplan and Battles
Figure 6 A causal tree outlining a labeling error.

REFERENCES
1. Linden JV, Kaplan HS. Transfusion errors: causes and effects. Transfus Med Rev
1994; 8(3):169183.
2. Guidelines for Good Manufacturing Practices. U.S. Food and Drug Administration,
1993.
3. Monger P. Packaging security in the pharmaceutical industry: an MCA inspectors
viewpoint. Pharmaceut Technol 1993; 4:81.
4. Nagel D. Human error in aviation operations. In: Wiener EL, ed. Human Factors in
Aviation. San Diego, CA: Academic Press, 1988:
5. Shulman IA, Lohr K, Derdiarian AK, Picukaric JM. Monitoring transfusion practices a
strategy for improving transfusion practices. J Bone Joint Surg 1992; 74-A:652658.
6. Honig CL, Bove JR. Transfusion-associated fatalities: review of Bureau of Biologics
reports 19761978. Transfusion 1980; 20:653661.
7. Myhre B. Fatalities from blood transfusion. JAMA 1980; 244:13331335.
8. Camp FR, Monaghan WP. Fatal blood transfusion reactions: an analysis. Am J
Forensic Med Pathol 1981; 2:143150.
9. Edinger S. A closer look at fatal transfusion reactions. Med Lab Obs 1985; 4:4145.
fusion 1990; 30:583590.
11. Mummert TB, Tourault MA. Review of transfusion related fatalities: many prevent-
able. Hosp Technol Scanner 1993; 4:13.
12. Reason J. Human Error. New York: Cambridge University Press, 1990.
13. Linden JV, Paul B, Dressler KP. A report of 104 transfusion errors in New York State.
Transfusion 1992; 32:601660.
14. Taswell HF, Smith AM, Sweatt MA, Pfaff KJ. Quality control in the blood bank: a
new approach. Am J Clin Pathol 1974; 62:491495.
15. Classen DC, Pestotnik SL, Evans RS, Burke JP. Computerized surveillance of adverse
drug events in hospitalized patients. JAMA 1991; 266:28172851.
16. Classen DC, Pestotnik MS, Evans RS, Lloyd JF, Burke JP. Adverse drug events in
hospitalized patients. JAMA 1997; 277:301306.
17. Reynard WD, Billings CE, Cheaney ES, Hardy R. The Development of the NASA
Aviation Safety Reporting System. Moffett Field, CA: National Aeronautics and
Space Administration Science and Technology Branch, 1986.
18. Billings, CE. Personal communication
19. Murphy WG, McClelland DBL. Deceptively low morbidity from failure to practice
safe blood transfusion: an analysis of serious blood transfusion errors. Vox Sang
1989; 57:5962.
20. Gambino R, Mallon P. Near misses: an untapped data base to find root causes. Lab
Rep 1991; 13:4144.
21. Van der Schaaf TW. Near miss reporting in the chemical process industry. Ph.D.
thesis, Eindhoven University of Technology, 1992.
22. Heinrich HW. Industrial Accident Prevention. New York: McGraw-Hill, 1931.
23. Zapt D, Reason JT. Introduction to error handling. Appl Psychol 1994; 43:427432.
24. Kaplan HS, Battles JB, Van der Schaaf TW, Shea CE, Mercer SQ. Identification and
classification of the causes of events in transfusion medicine. Transfusion 1998;

38:10711081.
25. Reason JT. Forward. In: Bogner MS, ed. Human Error in Medicine. Hillsdale, NJ:
Lawrence Erlbaum Associates; 1994:viixv.
26. Reason JT. Managing the Risks of Organizational Accidents. Aldershot: Ashgate,
1997.
27. Rasmussen J. The definition of human error and a taxonomy for technical system
design. In: Rasmussen J, Duncan K, Lepant J, eds. New Technology and Human
Error. London: John Wiley & Sons LTD, 1987:2330.
28. Rasmussen J. Information processing and human-machine interaction: an approach
to cognitive engineering. New York: Elsevier, 1986.
29. Perper JA. Life-threatening and fatal therapeutic misadventures. In: Bogner MS, ed.
Human Error in Medicine. Hillsdale, NJ: Lawrence Erlbaum Associates, 1994:2752.
30. Curtain LL. When negligence becomes homicide. Nursing Manage 1997; 2:78.
31. Paget MA. The Unity of Mistakes: A Phenomenological Interpretation of Medical
Work. Philadelphia: Temple University Press, 1988.
32. Berwick DM. Continuous improvement as an ideal in health care. N Engl J Med
1989; 32:5356.
33. Lucas DA. Organizational aspects of near miss reporting. In: van der Schaaf TW,
Lucas DA, Hale AR, eds. Near Miss Reporting as a Safety Tool. Oxford: Butterworth-
Heinemann, 1991:
34. Lucas DA. Understanding the human factors in disasters. Interdisciplinary Sci Rev
1992; 17:185190.
35. Leape LL, Bates DE, Cullen DJ, et al. System analysis of adverse drug event: ADE
prevention study group. JAMA 1995; 274:3543.
36. Battles JB, Kaplan HS, Van der Schaaf TW, Shea CE. Attributes of medical event
reporting systems. Arch Pathol Lab Med 1998; 122:132139.
5
Graft-Versus-Host Disease
Iain J. Webb and Kenneth C. Anderson

Dana-Farber Cancer Institute and Harvard Medical School, Boston,
Massachusetts
I. INTRODUCTION
The clinical course of allogeneic bone marrow and peripheral blood stem cell
transplant patients is frequently complicated by graft-versus-host disease (GVHD).
This condition results when lymphocytes in the donor hematopoietic stem cell
(HSC) component recognize the HLA antigens of the recipient as foreign,
generating a characteristic immune response (1,2). Fever, diarrhea, liver function
test (LFT) abnormalities, and a characteristic cutaneous rash are the major clinical
manifestations of this condition. However, GVHD may also result from the
infusion of viable T lymphocytes within cellular blood components. This condi-
tion, which is accompanied by marrow aplasia and pancytopenia, was first
described in 1966 (3). In the following decades, subsequent reports established
transfusion-associated graft-versus-host disease (TA-GVHD) as a distinct disease
entity. TA-GVHD is resistant to most immunosuppressive therapeutic modalities.
Consequently, treatment is rarely successful, making the recognition of patient
groups warranting preventive measures essential.
Recently, new therapeutic strategies utilizing the potential graft-versus-
leukemia (GVL) effect of allogeneic lymphocytes have been developed, empha-
sizing the need to further understand the pathogenesis of TA-GVHD. Adoptive
immunotherapy, the infusion of allogeneic lymphocytes in patients who have
relapsed following allogeneic hematopoietic progenitor cell (HPC) transplanta-
tion, has been shown to produce both remissions and a clinical syndrome similar
to TA-GVHD (4). Whether the GVL effect is separable from GVHD is the object
of ongoing laboratory and clinical studies.
109
110 Webb and Anderson
II. CLINICAL PRESENTATION AND COMPLICATIONS

Symptoms and signs of TA-GVHD typically appear 810 days posttransfusion,
with death typically occurring 34 weeks posttransfusion (5). As seen in cases of
GVHD following HPC transplantation, a characteristic cutaneous eruption ap-
pears, associated with watery diarrhea, LFT abnormalities, and fever. Watery
diarrhea may be profuse, and LFT elevations may be marked and accompanied
by extensive hepatocellular damage. Nonspecific manifestations include an-
orexia, nausea, and vomiting. The maculopapular exanthem typically develops
centrally, then progresses to involve the extremities. In severe cases, generalized
erythroderma and bullae may appear.
The development of marrow aplasia distinguishes TA-GVHD from GVHD
occurring following allogeneic HPC transplantation. Thrombocytopenia and leu-
kopenia are late features. Complications of pancytopenia such as hemorrhage and
infection ensue and lead to patient death. The time course is rapid, with death
occurring 13 weeks following development of clinical symptoms.
III. DIFFERENTIAL DIAGNOSIS

There are no pathognomonic features to differentiate TA-GVHD from a variety
of viral illnesses and drug reactions (6). Patients who are receiving transfusions
typically suffer from comorbid conditions, which may obscure the clinical
features of TA-GVHD, particularly should the clinician have a low index of
suspicion. Characteristic pathological changes in the skin, liver, and bone marrow
may aid diagnosis, but only the documentation of donor-derived lymphocytes in
the recipient circulation and/or tissues will confirm it. Techniques used to detect
donor-derived lymphocytes in the patient will be discussed below. Pathological
examination of a skin biopsy may reveal degeneration of the epidermal basal cell
layer with vascuolization; dermal-epithelial layer separation and bullae forma-
tion; mononuclear cell migration into, and infiltration of, the upper dermis;
hyperkeratosis; and degenerative dyskeratosis (7,8). Liver biopsies reveal degen-
eration and eosinophilic necrosis of the small bile ducts with intense periportal
inflammation and mononuclear (lymphocytic) infiltration. Bone marrow aspirates
reveal lymphocytic infiltration, pancytopenia, and possibly fibrosis.
IV. INCIDENCE
The epidemiological data on TA-GVHD are derived from analysis of reports of
single or very small groups of patients. A prospective study on the development
of TA-GVHD has never been undertaken and would be very difficult to perform.
Estimates of the incidence and identification of patient groups at risk are therefore
Graft-Versus-Host Disease 111
subject to the limitations of retrospective data. Cases of TA-GVHD are most

certainly underreported, with at least two factors contributing to the underreport-
ing: lack of recognition and the absence of definitive diagnostic studies in many
instances. Thus, while over 200 cases of presumed TA-GVHD have been reported
or referenced in the Japanese- and English-language literature, definitive diagnos-
tic tests have been performed in only a handful of cases. Lists of published case
reports and small series have been compiled in several articles (9,10). Rare
survivors have been documented, but the overall reported mortality is approxi-
mately 90% (5,11,12).
V. PATIENT GROUPS AT RISK

TA-GVHD has been reported in patients with hematological or solid malignancies
or congenital immunodeficiency states, as well as in infants and adults with
apparently intact immune systems, but not in patients afflicted with the acquired
immune deficiency syndrome (AIDS). In addition, the disease has also only rarely
been reported in recipients of organ transplants or in patients on immunosuppres-
sive medications.
Initially, cases of TA-GVHD were reported in patients with severe com-
bined immunodeficiency or Wiskott-Aldrich syndromes (Table 1), in newborns
with erythroblastosis fetalis (Table 2), and in patients with Hodgkins disease
or non-Hodgkins lymphoma, acute myelocytic or lymphoblastic leukemia, or
chronic lymphocytic leukemia (Table 3) (13). Although its true incidence remains
unknown, TA-GVHD is estimated to occur in 0.11.0% of patients with hemato-
logical malignancies or lymphoproliferative diseases, and patients carrying the
above diagnoses are felt to be at risk for TA-GVHD. Certain subgroups of patients
with hematological malignancies, including those suffering from chronic lym-
phocytic leukemia who are treated with fludarabine, may be at higher risk
because of the prolonged effects of this purine analog on cell-mediated immune
function (1416).
Table 1 TA-GVHD in Patients with

Congenital Immunodeficiency Syndromes:
Clinical Setting
Severe combined immunodeficiency syndrome
Thymic hypoplasia
Wiskott-Aldrich syndrome
Leniers disease
5-Nucleotidase deficiency
Nonspecified immunodeficiency
Table 2 TA-GVHD in Patients with

Immunodeficiency States: Clinical Setting
Hemolytic disease of the fetus or newborn
Premature newborns
Neonatal alloimmune thrombocytopenia
Neonatal immunosuppressive medication
Cytotoxic drugs and irradiation may be related to the development of

TA-GVHD (17). TA-GVHD was originally recognized in patients with solid
tumors receiving intensive therapy for neuroblastoma (18,19). In one series, 4 of
34 patients with solid tumors (lung and germ cell cancer) who were treated with
high doses of chemotherapy and autologous marrow infusions and subsequently
received transfusions of nonirradiated blood cells developed TA-GVHD (20). In
addition, case reports documenting TA-GVHD in patients with cervical, renal,
esophageal, lung, bladder, or prostate carcinoma who did not receive aggressive
chemotherapy indicated that a broader spectrum of patients with solid tumors may
be at risk.
TA-GVHD has also been documented in premature infants who received
unirradiated blood products in the setting of hyaline-membrane disease, suspected
sepsis, and respiratory distress syndrome, but also in infants without these
Table 3 Posttransfusion GVHD in Patients with

Malignancies: Clinical Setting
Hematological
Hodgkins disease
Non-Hodgkins lymphoma
Acute myelocytic leukemia
Acute lymphocytic leukemia
Chronic lymphocytic leukemia
Aplastic anemia
Solid tumors
Neuroblastoma
Lung carcinoma
Glioblastoma
Rhabdomyosarcoma
Cervical carcinoma
Esophageal carcinoma
Renal adenocarcinoma
Autologous hematopoietic progenitor cell transplantation
complications (10,21). Such patients did not have congenital immunodeficiency

syndromes or erythroblastosis fetalis.
Finally, TA-GVHD has been reported in apparently immunocompetent
adults (Table 4). Clinical settings in which TA-GVHD has been reported in
immunologically normal hosts include pregnancy; cardiac, vascular, and abdom-
inal surgeries; alpha thalassemia; rheumatoid arthritis; trauma; and short-course
glucocorticoid therapy (22,23). The syndrome was initially described in 1986
following blood transfusion to a patient in Japan who had surgery for an aortic
aneurysm and was not recognized to be immunodeficient (24). A survey of
340 Japanese hospitals documented postoperative erythroderma, identical to
TA-GVHD, in 96 of the 63,257 patients who underwent cardiac surgery, with a
mortality rate of 90% (25).
More recently, fatal GVHD has been reported in a number of HLA-
heterozygous transfusion recipients who received a transfusion with a one-way
HLA match (26,27). These recipients shared a haplotype with related or unre-
lated HLA-homozygous donors. Directed donations from immediate family mem-
bers increase the likelihood of TA-GVHD because such donors share HLA
antigens with recipients; homozygosity for HLA types is likely to be present
among not only first-degree relatives but also all related recipient-donor pairs
(28). The frequency of one-way HLA matches has been calculated in different
countries (29). The reported risk varies from 1:16,835 in France to 1:874 in Japan,
where there is less diversity in HLA antigen expression and, consequently,
TA-GVHD cases have been more frequently reported (9,10,29).
The transfusion of platelets from donors sharing at least two antigens at the
HLA-A and B loci (HLA-matched platelets) has been demonstrated to result in
satisfactory platelet increments in alloimmunized patients refractory to standard
platelet therapy (30). However, the provision of platelets from donors sharing
HLA antigens may also predispose to TA-GVHD (31).
Table 4 TA-GVHD in
Immunocompetent Patients:
Clinical Setting
Pregnancy
Cholecystectomy
Cardiac surgery
Vascular surgery
Gastrointestinal surgery
Abdominal surgery
Alpha thalassemia
Liver transplantation
Pancreosplenic transplantation
The risk factors predisposing to TA-GVHD are only partially defined.

TA-GVHD does not always occur in immunodeficient patients receiving unirra-
diated blood components, while it may affect individuals with apparently normal
immune function, particularly in the setting of a one-way HLA match. Blood
transfusion itself may be immunosuppressive (32). Hence, an argument can be
made that almost every patient who requires transfusion of a blood component is
potentially immunocompromised in some way that could facilitate the develop-
ment of TA-GVHD. The low incidence of TA-GVHD in presumably immuno-
competent patients may result from underrecognition of the syndrome, but likely
also reflects effective defense mechanisms in individuals with truly intact immune
function.
VI. PATHOGENESIS
The immune status of the host and the extent of HLA mismatch between donor
and recipient together determine the extent of any host-versus-graft response.
The ability of the transfusion recipient to mount an immune response against
donor T lymphocytes is fundamental to the pathogenesis of TA-GVHD. Thus,
Billingham (33) has proposed three requirements for the development of GVHD:
(a) differences in histocompatibility (HLA) antigens between the donor and the
recipient, (b) the presence of immunocompetent cells in the graft, and (c) inability
of the host to reject these immunocompetent cells. Usually the larger number of
immune cells in the immunocompetent host will eliminate the donor-derived
T cells via a host-versus-graft reaction. However, if an immunocompetent HLA-
heterozygous individual is transfused with even a small number of functional
T lymphocytes derived from a donor who is homozygous for one of the recip-
ients HLA haplotypes, the recipients immune system does not recognize the
major histocompatibility antigens on the donor cells as being foreign and is
therefore incapable of eliminating them. Some of the donor-derived T cells will
recognize those host HLA antigens that are encoded by the unshared haplotype
as being foreign, undergo clonal expansion, and establish TA-GVHD (26).
The TA-GVHD expressed under these circumstances may be expected to occur
regardless of the hosts immune status since the failure to eliminate donor-derived
T lymphocytes is based on the genetics of the HLA system rather than on
variables that contribute to immune competence per se.
Fast et al. (34) have provided important insights into the pathogenesis of
TA-GVHD, implicating recipient CD4+, CD8+, and natural killer cells in control-
ling TA-GVHD. In a mouse model, varying numbers of parental lymphoid cells
were injected into unirradiated F1 hybrid recipients, providing donor lymphocytes
homozygous for an HLA haplotype present in the recipient. The effect of selective
depletion of recipient CD4+, CD8+, and natural killer cells on the regulation of
TA-GVHD was assessed. Depletion of CD4+ cells increased the number of donor
cells necessary to induce TA-GVHD, while depletion of recipient CD8+ cells or
natural killer cells decreased the number of door cells required to produce
TA-GVHD. Thus, CD4+ cells may be involved in the pathogenesis of TA-GVHD,
and CD8+ and natural killer cells may be protective. According to this model,
patients at high risk for TA-GVHD would be those with impaired CD8 and natural
killer cell function, especially if they receive blood products with a one-way
HLA match.
No cases of TA-GVHD have been reported in individuals with AIDS,
further emphasizing the role of CD4+ and CD8+ lymphocytes in the pathogenesis
of TA-GVHD. It is possible that TA-GVHD may account for some of the
nonspecific signs and symptoms presently attributed to infections, drug reactions,
and other coexistent medical conditions in AIDS patients. Alternatively, it may
be that some qualitative aspect of the human immunodeficiency virus (HIV)
related immune deficit may alter the predisposition of AIDS patients to develop
TA-GVHD. The murine study reported by Fast et al. (34) suggests that the latter
explanation may be valid, as CD4+ lymphocyte function declines early while
CD8+ function is preserved until late in the course of HIV disease. In the mouse
model, both of these features would be expected to decrease the likelihood of
developing TA-GVHD. In contrast, it has been proposed that reactive recipient
CD8+ T cells are required for the development of GVHD (35). Activated
HIV-1infected CD4+ T cells express an HLA class 2derived peptide mimicked
by the carboxy-terminus of the HIV-1 envelope. It is hypothesized that the
activation of CD8+ T cells against the HIV-infected CD4+ T cells may preclude
the development of GVHD. Consequently, clarification of the role of both CD4+
and CD8+ cells in AIDS will likely provide further insight into the pathogenesis
of GVHD and vice versa.
In recipients of allogeneic HPC transplants, the spectrum between GVHD
and HLA alloimmunization and the extent of mononuclear cell microchimerism
depend on the dose of allogeneic cells transfused, the immune competence of the
recipient, as well as the extent of HLA similarity between donor and recipient
(36). For example, in an animal model of postbone marrow transplant GVHD,
reducing the number of cells transplanted resulted in a stepwise increase in
rejection rates of donor cells; conversely, reduction in the size of the marrow
innoculum could be compensated by increasing host immunosuppression (37).
Since transfused cellular blood products are rarely HLA-antigen tested or
matched, the first of Billinghams three requirements (33)that there be dif-
ferences in HLA between the donor and recipientis almost always present
in the setting of blood component transfusion. TA-GVHD is mediated by the
viable T lymphocytes that inevitably contaminate nonirradiated transfused cellu-
lar blood components. The naturally or iatrogenically immunosuppressed recipi-
ent has only a limited capacity to generate an effective host-versus-graft reaction,
and greater HLA disparity increases the probability that donor lymphocytes will
attack host tissues.
VII. DIAGNOSIS
The differential diagnosis for TA-GVHD is broad: myriad factors such as infec-
tions and drug reactions can result in the development of fevers, skin rashes, and
LFT abnormalities. Histological findings in the skin and gastrointestinal tract may
suggest the diagnosis of TA-GVHD, but are not pathognomonic. The only
definitive approach to the diagnosis of TA-GVHD is the identification of donor-
derived lymphocytes in the circulation or tissues of the affected host (Table 5).
This requires either careful HLA typing or some other technique that reliably
distinguishes between host and donor cells (3841). Blood samples adequate for
HLA typing are frequently not available, since circulating host blood cells are
rapidly eliminated. Polymerase chain reactionbased methods for HLA typing
may be a useful substitute (4043). Otherwise, the patients HLA type may be
deduced from those of surviving first-degree relatives (26). Cytogenetics have
been employed in the event that donor and recipient have been of different gender
or when the disease has followed transfusion of granulocytes donated by individ-
uals with Philadelphia chromosome-positive chronic myeloid leukemia (44).
Increasingly sophisticated techniques for confirming the diagnosis of TA-GVHD
have included the detection of polymorphisms for restriction fragment lengths
and human microsatellite markers (27,41). Even so, this syndrome is still fre-
quently diagnosed only at autopsy.
VIII. THERAPY
Treatment of TA-GVHD is only rarely effective. Attempted immunosuppressive
therapies have included glucocorticoids, antithymocyte globulin, cyclosporine,
cyclophosphamide, and anti-T-cell monoclonal antibodies (8,23,25,4547). Al-
Table 5 Methods Used to Diagnose or Document TA-GVHD

Conventional HLA typing
Deduction of patients HLA type from those of family members
DNA-based HLA typing using the polymerase chain reaction (PCR)
Cytogenetics
Restriction fragment length polymorphism
Polymorphism of human microsatellite markers
Identification of donor T cells by above techniques in skin biopsies
though some of these agents have been successful in the treatment of post-HPC
transplant GVHD, they have been ineffective for TA-GVHD. Rare responses to
some of the commonly used agents have been reported (7,48), contributing to
anecdotal experiences from which it is difficult to extract guidelines for clinical
practice. A therapeutic trial of glucocorticoids or other immunosuppressive agents
is often attempted but is frequently ineffective in this serious and increasingly
frequent disease. There appears to be no advantage to early diagnosis and
treatment, as patients diagnosed early in the course of the disease fare no better
than those diagnosed later. Thus, prevention is critically important.
IX. PREVENTION: GAMMA IRRADIATION

Irradiation of blood components is indicated in patient groups at high risk for the
development of TA-GVHD (Table 6). The relatively low frequency of TA-GVHD
in immunocompetent patients receiving blood from unrelated donors has thus far
precluded the extension of gamma-irradiation to all transfused cellular blood
components. Issues relating to cost, the logistics of irradiation in emergency and
small clinic settings, and the exceedingly low risk of TA-GVHD in most cases
Table 6 Patient Groups at Risk for Developing TA-GVHD

Clearly supported
Patients with selected immunodeficiencies:
Congenital immunodeficiencies
Hodgkins disease
Chronic lymphocytic leukemia treated with fludarabine
Newborns with erythroblastosis fetalis
Intrauterine transfusions
Recipients of hematopoietic progenitor cell transplants
Recipients of blood products donated by relatives
Recipients of HLA-selected (matched) platelets or platelets known to be
homozygous
Probably at risk
Patients with:
Other hematological malignancies
Solid tumors treated with cytotoxic agents
Recipient-donor pairs from genetically homogeneous populations
Premature and, possibly, term neonates
No defined risk
Patients with:
Acquired immunodeficiency syndrome (AIDS)
Immunosuppressive medications
have also been raised (49,50). However, irradiation is likely indicated within
genetically homogeneous populations, since transfusion of unirradiated compo-
nents between unrelated donor-recipient pairs may be expected to result in
TA-GVHD when a one-way HLA match occurs by chance.
The American Association of Blood Banks (AABB) currently recommends
the irradiation of blood components at 2.5 Gy to the center of the component,
with no area receiving less than 1.5 Gy (51). This irradiation dose is in excess of
the levels found to abrogate mixed lymphocyte culture (MLC) reactivity, as this
dose is inadequate to prevent TA-GVHD (52,53). Instead, the recommended
2.5 Gy dose is based on limiting dilution assays (LDAs) using visual assessment
of T-cell proliferation, which demonstrate a greater than 5 log reduction in viable
T cells. However, levels of host cytotoxic T-lymphocyte precursors (pCTLs) and
interleukin-2secreting helper T-lymphocyte precursors (pHTLs), which are not
measured in LDAs, may be more predictive of GVHD development following
allogeneic BMT (5458).
Significant variability in irradiation practice exists between centers. For
example, blood component irradiation practice was examined in a survey of 2250
blood centers, hospital blood banks, and transfusion services that are institutional
members of the AABB (59). Only 12.3% of the institutions had on-site facilities
for the irradiation of blood components. Of 9,397,516 components transfused in
1989, 952,516 (10.1%) were irradiated, and 44 cases of TA-GVHD were identi-
fied. There was marked variability in blood component irradiation practice, even
among groups in whom the risk of TA-GVHD was well defined. For example,
12, 19, and 32% of institutions did not provide irradiated components to recipients
of allogeneic BMT, patients receiving autologous BMT, and those with congenital
immunodeficiencies, respectively. Irradiated blood components were provided by
51.4, 34, 32, and 20% of institutions to patients with leukemias, Hodgkins
disease, non-Hodgkins lymphoma, and solid tumors, respectively, and 24.5%
provided irradiated blood products to patients with AIDS.
X. NEW DIRECTIONS
A. Leukoreduction to Prevent TA-GVHD
In theory, leukoreduction of cellular blood components could be used to decrease
TA-GVHD. However, since neither the number nor quality of T cells that mediate
GVHD are presently defined, targets for leukoreduction to prevent TA-GVHD
cannot be established. Further, TA-GVHD has been reported in both immu-
nocompetent and immunodeficient recipients of transfusions leukoreduced by
filtration, suggesting that standard leukoreduction procedures cannot completely
prevent TA-GVHD (60,61). Newer technologies appear capable of decreasing the
number of contaminating leukocytes (62); whether the levels of leukoreduction
achieved will be adequate to prevent TA-GVHD remains to be established.

Leukoreduction could theoretically deplete sufficient numbers of T-helper lym-
phocyte precursors (pHTL) and cytotoxic T-lymphocyte precursors (pCTL) to
avoid TA-GVHD. However, the effect of neither gamma irradiation nor leuko-
reduction on these cell populations has yet been determined. Until data demon-
strate adequate highly efficient removal of pHTL and pCTL by leukoreduction,
gamma irradiation of all cellular blood components should be utilized to prevent
TA-GVHD in patient groups judged to be at risk.
B. Donor Lymphocyte Infusions and TA-GVHD

Patients with leukemia relapsing following allogeneic bone marrow transplanta-
tion may respond to infusions of lymphocytes from the original bone marrow
donor, without receiving chemotherapy or other treatment. For example, the
European Group for Blood and Marrow Transplantation (EGBMT) documented
complete remissions in 73% of 84 patients with relapsed chronic myelogenous
leukemia (CML) (4). Eighty-seven percent of patients remained in remission at 3
years. GVHD of Grade II or greater developed in 41% of patients, and myelo-
suppression occurred in 34% of patients. The development of these features
associated with TA-GVHD was associated with the attainment of remission,
suggesting a potential relationship between TA-GVHD and the GVL effect of the
lymphocyte infusions.
The donor lymphocyte populations responsible for GVL, TA-GVHD, and
post-HPC transplantation GVHD have not been determined. CD4+ cells have
been implicated in GVL (63,64), while it has been observed that the depletion of
either CD6+ or CD8+ T lymphocytes from donor bone marrow can prevent
GVHD at the time of allogeneic transplantation (65,66). Further experiments to
identify the role of both donor and host lymphocyte populations in TA-GVHD
and GVL are necessary. These studies will be important to determine how
best to prevent TA-GVHD and optimize the GVL effect of lymphocytes while
minimizing GVHD.
REFERENCES
1. Ferrara JL, Deeg HJ. Graft versus host disease. N Engl J Med 1991; 324:667764.
2. Antin JH, Ferrara JLM. Cytokine dysregulation and acute graft-versus-host disease.
Blood 1992; 80:29642968.
3. Hathaway WE, Brangle RW, Nelson TL, Roeckel IE. Aplastic anemia and alympho-
cytosis in an infant with hypogammaglobulinemia: Graft-versus-host reaction? J Ped
1966; 68:713722.
4. Kolb HJ, Schattenberg A, Goldman JM, Hertenstein B, Jacobsen N, Arcese W,
Ljungman P, Ferrant A, Verdonck L, Niederwieser D, van Rhee F, Mittermueller J,
de Witte T, Holler E, Ansari H. Graft-versus-leukemia effect of donor lymphocyte

transfusions in marrow grafted patients. European Group for Blood and Marrow
Transplantation Working Party Chronic Leukemia. Blood 1995; 86:20412050.
5. Anderson KC. Clinical indications for blood component irradiation. In: Baldwin ML,
Jefferies LC, eds. Irradiation of Blood Components. Bethesda, MD: American
Association of Blood Banks, 1992:3149.
6. Shivdasani RA, Anderson KC. Transfusion-associated graft-versus-host disease:
scratching the surface. Transfusion 1993; 33;696697.
7. Prince M, Szer J, van der Weyden MB, Pedersen JS, Holdsworth RF, Whyte G.
Transfusion associated graft-versus-host disease after cardiac surgery: response to
antithymocyte-globulin and corticosteroid therapy. Aust NZ J Med 1991; 21:4346.
8. OConnor NTJ, Mackintosh P. Transfusion associated graft versus host disease in an
immunocompetent patient. J Clin Pathol 1992; 45:621622.
9. Ohto H, Anderson KC. Survey of transfusion-associated graft-versus-host disease in
immunocompetent recipients. Transfusion Med Rev 1996; 10:3143.
10. Ohto H, Anderson KC. Posttransfusion graft-versus-host disease in Japanese new-
borns. Transfusion 1996; 36:117123.
11. Andersen CB, Ladefoged SD, Taaning E. Transfusion-associated graft-versus-graft
and potential graft-versus-host disease in a renal allotransplanted patient. Hum Path
1992; 23:831834.
12. Mori S, Matsushita H, Ozaki K, Ishida A, Tokuhira M, Nakajima H, Kizaki M,
Sugiura H, Kikuchi A, Handa M, et al. Spontaneous resolution of transfusion-
associated graft-versus-host disease. Transfusion 1995; 35:431435.
13. von Fliedner V, Higby DJ, Kim U. Graft-versus-host reaction following blood product
transfusion. Am J Med 1982; 72:951961.
14. Briz M, Cabrera R, Sanjuan I, Fores R, Diez JL, Herrero M, Regidor C, Algora M,
Fernandez MN. Diagnosis of transfusion-associated graft-versus-host disease by
polymerase chain reaction in fludarabine-treated B-chronic lymphocytic leukaemia.
Br J Haematol 1995; 91:409411.
15. Briones J, Pereira A, Alcorta I. Transfusion-associated graft-versus-host disease
(TA-GVHD) in fludarabine-treated patients: is it time to irradiate blood component?
Br J Haematol 1996; 93:739741.
16. Williamson LM, Wimperis JZ, Wood ME, Woodcock B. Fludarabine treatment and
transfusion-associated graft-versus-host disease. Lancet 1996; 348:472473.
17. Kessinger A, Armitage J, Klassen L, Landmark J, Hayes J, Larsen A, Purtilo D. Graft
versus host disease following transfusion of normal blood products to patients with
malignancies. J Surg Oncol 1987; 36:206209.
18. Woods WG, Lubin BH. Fatal graft-versus-host disease following a blood transfusion
in a child with neuroblastoma. Paediatrics 1981; 67:217221.
19. Kennedy J, Ricketts R. Fatal graft versus host disease in a child with neuroblastoma
following a blood transfusion. J Ped Surg 1986; 21:11081109.
20. Postmus PE, Mulder NH, Elema JD. Graft versus host disease after transfusions of
non-irradiated blood cells in patients having received autologous bone marrow: a
report of 4 cases following ablative chemotherapy for solid tumors. Eur J Cancer Clin
Oncol 1988; 24:889894.
21. Berger RS, Dixon SL. Fulminant transfusion-associated graft-versus-host disease in

a premature infant. J Am Acad Dermatol 1989; 9:205207.
22. Sheehan T, McLaren KM, Brettle R, Parker AC. Transfusion-induced graft versus
host disease in pregnancy. Clin Lab Haematol 1987; 9:205207.
23. Otsuka S, Kunieda K, Kitamura F, Misawa K, Sasaoka I, Hirose M, Kasuya S, Saji
S, Noma A. The critical role of blood from HLA-homozygous donors in fatal
transfusion-associated graft-versus-host disease in immunocompetent patients. Trans-
fusion 1991; 31:260264.
24. Hathaway WE, Fulginit VA, Pierce CW, Githens JH, Pearlman DS, Muschenheim F,
Kempe CH. Graft-vs-host reaction following a single blood transfusion. J Am Med
Assoc 1967; 201:10151020.
25. Juji T, Takahashi K, Shibata Y, Ide H, Sakakibira T, Ino T, Mori S. Posttransfusion
graft-versus-host disease in immunocompetent patients after cardiac surgery. N Engl
J Med 1989; 321:56.
26. Shivdasani RA, Haluska FG, Dock NL, Dover JS, Kineke EJ, Anderson KC.
Graft-versus-host disease associated with transfusion of blood from unrelated HLA-
homozygous donors. N Engl J Med 1993; 328:766770.
27. Petz LD, Calhoun L, Yam P, Cecka M, Schiller G, Faitlowicz AR, Herron R, Sayah
D, Wallace RB, Belidegrun A. Transfusion-associated graft-versus-host disease in
immunocompetent patients: report of a fatal case associated with transfusion of blood
from a second-degree relative, and a survey of predisposing factors. Transfusion
1993; 33:742750.
28. Kantre MH. Transfusion-associated graft-versus-host disease: Do transfusions from
second-degree relatives pose a greater risk than those from first-degree relatives?
Transfusion 1992; 32:323327.
29. Ohto H, Yasuda H, Noguchi M, Abe R. Risk of transfusion-associated graft-versus-
host disease as a result of directed donations from relatives. Transfusion 1992;
32:691693.
30. Yankee RA, Grumet FC, Rogentine GN. Platelet transfusion therapy: The selection
of compatible platelet donors for refractory patients by lymphocyte HLA typing.
N Engl J Med 1969; 281:12081212.
31. Benson K, Marks AR, Marshall MJ, Goldstein JD. Fatal graft-versus-host disease
associated with transfusions of HLA-matched, HLA-homozygous platelets from
unrelated donors. Transfusion 1994; 34:432437.
32. Perkins HA. Transfusion-induced immunologic unresponsiveness. Transfusion Med
Rev 1988; 2:196203.
33. Billingham RE. The biology of graft-versus-host reactions. Harvey Lectures 1966-67;
62:2178.
34. Fast LD, Valeri CR, Crowley JP. Immune responses to major histocompatibility
complex homozygous lymphoid cells in murine F1 hybrid recipients: implications for
transfusion-associated graft-versus-host disease. Blood 1995; 86:30903096.
35. Habeshaw JA, Dalgleish AG, Hounsell EF. Absence of GVH in AIDS. J AIDS 1994;
7:12871289.
36. Dzik WH. Mononuclear cell microchimerism and the immunomodulatory effect of
37. Uharek L, Gassmann W, Glass B, Steinmann J, Loeffler H, Mueller-Ruchholtz W.
Influence of cell dose and graft-versus-host disease on rejection rates after allogeneic
transplantation. Blood 1992; 79:16121620.
38. Kunstmann E, Bocker T, Roewer L, Sauer H, Mempel W, Epplen JT. Diagnosis of
transfusion-associated graft-versus-host disease by genetic fingerprinting and poly-
merase chain reaction. Transfusion 1992; 32:766770.
39. Suzuki K, Akiyama H, Takamoto S, Maruyama Y, Sakamaki H, Akagi K, Maeda Y,
Takenaka M, Onozawa Y. Transfusion-associated graft-versus-host disease in a
presumably immunocompetent patient after transfusion of stored packed red cells.
Transfusion 1992; 32:358360.
40. Hayakawa S, Chishima F, Sakata H, Fuji K, Ohtani K, Kurashina K, Hayakawa J,
Suzuki K, Nakabayashi H, Esumi M, Nemoto N, Sakurai I, Satoh K. A rapid
molecular diagnosis of posttransfusion graft-versus-host disease by polymerase chain
reaction. Transfusion 1993; 33:413417.
41. Wang L, Juji T, Tokunaga K, Takahashi K, Kuwata S, Uchida S, Tadokoro K, Takai
K. Brief Report: Polymorphic microsatellite markers for the diagnosis of graft-
versus-host disease. N Engl J Med 1994; 330:398401.
42. Saito M, Takamatsu H, Nakao S, Shiobara S, Matsuda T, Tajika E, Nakamura T, Kaji
T, Yoshida T. Transfusion-associated graft-versus-host disease after surgery for
bladder cancer. Blood 1993; 82:326330.
43. Uchida S, Wang L, Yahagi Y, Tokunaga K, Tadokoro K, Juji T. Utility of fingernail
DNA for evaluation of chimerism after bone marrow transplantation and for diagnos-
tic testing for transfusion-associated graft-versus-host disease [letter]. Blood 1996;
87:40154016.
44. Matsushita H, Shibata Y, Fuse K, Kimura M, Iinuma K. Sex chromatin analysis of
lymphocytes invading host organs in transfusion-associated graft-versus-host disease.
Virchows Archiv B Cell Pathol 1988; 55:237239.
45. Arsura EL, Bertelle A, Minkowitz S, Cunningham J, Jr., Grob D. Transfusion-
associated-graft-versus-host disease in a presumed immunocompetent patient. Arch
Int Med 1988; 148:19411944.
46. Otsuka S, Kunieda K, Hirose H, Takeuchi H, Mizutani M, Nagaya G, Sato G, Kasuya
S, Matsutomo K, Noma A, Saji S. Fatal erythroderma (suspected graft-versus-host
disease) after cholecystectomy. Transfusion 1989; 29:544548.
47. Vogelsang GB. Transfusion-associated graft-versus-host disease in nonimmunocom-
promised hosts. Transfusion 1990; 30:101103.
48. Cohen D, Weinstein H, Mihm M, Yankee R. Nonfatal graft-versus-host disease
occurring after transfusion with leukocytes and platelets obtained from normal
donors. Blood 1979; 53:10531059.
49. Lind SE. Has the case for irradiating blood products been made? Am J Med 1985;
78:543544.
50. Perkins H. Should all blood from related donors be irradiated? Transfusion 1992;
32:302303.
51. Preparation of blood components. In: Klein H, ed. Standards for Blood Banks
and Transfusion Services. Bethesda, MD: American Association of Blood Banks,
1996:14.
52. Sproul AM, Chalmers EA, Mills KI, Burnett AK, Simpson E. Third party mediated
graft rejection despite irradiation of blood products. Br J Haematol 1992; 80:251

252.
53. Lowenthal RM, Challis DR, Griffiths AE, Chappell RA, Goulder PJ. Transfusion-
associated graft-versus-host disease: report of an occurrence following the adminis-
tration of irradiated blood. Transfusion 1993; 33:524529.
54. Kaminski E, Hows J, Man S, Brookes P, Mackinnon S, Hughes T, Avakian O,
Goldman JM, Batchelor JR. Prediction of graft versus host disease by frequency
analysis of cytotoxic T cells after unrelated donor bone marrow transplantation.
Transplantation 1989; 48:608613.
55. van Els CA, Bakker A, Zwinderman AH, Zwaan FE, van Rood JJ, Goulmy E. Effector
mechanisms in graft-versus-host disease in response to minor histocompatibility
antigens. II. Evidence of a possible involvement of proliferative T cells. Transplan-
tation 1990; 50:6771.
56. Irschick EU, Hladik F, Niederwieser D, Nussbaumer W, Holler E, Kaminski E, Huber
C. Studies on the mechanism of tolerance or graft-versus-host disease in allogeneic
bone marrow recipients at the level of cytotoxic T-cell precursor frequencies. Blood
1992; 79:16221628.
57. Theobald M, Nierle T, Bunjes D, Arnold R, Heimpel H. Host-specific interleukin-2-
secreting donor T-cell precursors as predictors of acute graft-versus-host disease in
bone marrow transplantation between HLA-identical siblings. N Engl J Med 1992;
327:16131617.
58. Schwarer AP, Jiang YZ, Brookes PA, Barrett AJ, Batchelor JR, Goldman JM, Lechler
RI. Frequency of anti-recipient alloreactive helper T-cell precursors in donor blood
and graft-versus-host disease after HLA-identical sibling bone-marrow transplanta-
tion. Lancet 1993; 341:203205.
59. Anderson KC, Goodnough LT, Sayers M, Pisciotto PT, Kurtz SR, Lane TA, Anderson
CS, Silberstein LE. Variation in blood component irradiation practice: implica-
tions for prevention of transfusion-associated graft-versus-host disease. Blood 1991;
77:20962102.
60. Akahoshi M, Takanashi M, Masuda M, Yamashita H, Hidano A, Hasegawa K,
Kasajima T, Shimizu M, Motoji T, Oshimi K, et al. A case of transfusion-associated
graft-versus-host disease not prevented by white cell-reduction filters. Transfusion
1992; 32:169172.
61. Hayashi H, Nishiuchi T, Tamura H, Takeda K. Transfusion-associated graft-versus-
host disease caused by leukocyte-filtered stored blood. Anesthesiology 1993;
79:14191421.
62. Webb IJ, Schott DM, Cook J, Anderesen JW, Barrett BB, Anderson KC. Cobe Spectra
LRS for preparation of leukoreduced single donor apheresis platelets without labora-
tory filtration. Blood 1996; 88:335a (Abstr 1327).
63. Faber LM, van Luxemburg-Heijs SAP, Veenhof WFJ, Willemze R, Falkenburg JHF.
Generation of CD4+ cytotoxic T-lymphocyte clones from a patient with severe
graft-versus-host disease after allogeneic bone marrow transplantation: implications
for graft-versus-leukemia reactivity. Blood 1995; 86:28212828.
64. Giralt S, Hester J, Huh Y, Hirsch-Ginsberg C, Rondon G, Seonh D, Lee M, Gajewski
J, Van Besien K, Khouri I, Mehra R, Przepiorka D, Korbling M, Talpaz M, Kantarjian
H, Fischer H, Deisseroth A, Champlin R. CD8-depleted donor lymphocyte infusion
as treatment for relapsed chronic myelogenous leukemia after allogeneic bone

marrow transplantation. Blood 1995; 86:43374343.
65. Champlin R, Ho W, Gajewski J, Feig S, Burnison M, Holley G, Greenberg P, Lee K,
Schmid I, Giorgi J, Yam P, Petz L, Winston D, Warner N, Reichert T. Selective
depletion of CD8+ T lymphocytes for prevention of graft-versus-host disease after
allogeneic bone marrow transplantation. Blood 1990; 76:418423.
66. Soiffer RJ, Murray C, Mauch P, Anderson KC, Freedman AS, Rabinowe SN,
Takvorian T, Robertson MJ, Spector N, Gonin R, Miller KB, Rudders RA, Freeman
A, Blake K, Coral F, Nadler LM, Ritz J. Prevention of graft-versus-host disease by
selective depletion of CD6-positive T lymphocytes from donor bone marrow. J Clin
Oncol 1992; 10:11911200.
6
Transfusion-Related
Acute Lung Injury
Mark A. Popovsky
American Red Cross Blood ServicesNew England Region;
Harvard Medical School; and Beth Israel Hospital, Boston,
Massachusetts
I. INTRODUCTION
Although more than three centuries have passed since the recognition that
transfusion can be associated with deadly complications, it is only in the last
1015 years that pulmonary injury has been generally appreciated as a possible
transfusion outcome. In fact, before the mid-1980s the only well-recognized
manifestations of pulmonary injury from transfusion were anaphylactic reactions
and circulatory overload. Unfortunately, the respiratory system can be compro-
mised by another, immunologically driven type of reaction. Such reactions
were originally designated by a variety of descriptive terms, including non-
cardiogenic pulmonary edema (1), allergic pulmonary edema (2), hypersensitivity
reaction (3), and leukoagglutinin transfusion reaction (4). Only in the last decade
has there been a broader appreciation that this type of lung injury has an
immunological basis.
II. CLINICAL PRESENTATION

Transfusion-related acute lung injury (TRALI) is a life-threatening complication
that, when it presents fulminantly, is indistinguishable from the adult respiratory
distress syndrome (ARDS) secondary to other etiologies (e.g., sepsis, aspiration,
125
126 Popovsky
Table 1 Signs and Symptoms of TRALI

Acute respiratory distress
Acute pulmonary edema
Hypotension
Hypoxemia
Fever
Setting: transfusion of plasma-containing blood products within 12 hours
toxic inhalation) (5,6). As with ARDS, TRALI is a syndrome characterized by

acute respiratory distress.
The symptoms of TRALI include severe bilateral pulmonary edema, severe
hypoxemia (arterial oxygen tensions of 3050 torr are frequently observed)
(57), and cyanosis. The respiratory distress may first be manifested as dysp-
nea or cyanosis (peripheral or central). Although the edema may first be con-
fined to the lower lung fields, over several hours it usually involves the entire
lung. Roentgenograms classically demonstrate white-out by interstitial and
alveolar infiltrates (3,5,6), but in the first few hours a patchy pattern may be
observed. For patients who are in a decubitus condition in the operating room,
the edema may initially manifest itself in only the dependent areas of the lung
(Table 1).
Other frequent manifestations include fever (12C elevation) and mild
to moderate hypotension, although infrequently, hypertension may be ob-
served. When hypotension occurs, it is usually unresponsive to intravenous fluid
administration.
All of these symptoms arise in the setting of recent transfusion of plasma-
containing blood components, always within 16 hours and usually within 12
hours. In contrast to circulatory overload, patients with TRALI have normal
central venous pressure and normal or low pulmonary wedge pressures.
While this discussion addresses the severe, classic presentation, many
patients may present with milder forms of respiratory distress that still represent
this syndrome.
III. COMPLICATIONS
TRALI differs from ARDS in several important ways. Unlike ARDS with its
attendant morbidity and mortality (death rate of approximately 4050%), approx-
imately 80% of patients with TRALI improve both clinically and physiologically
within 4896 hours of the original insult, provided there is prompt and vigorous
respiratory support (58) While in many ARDS patients the lung injury is
irreversible, in TRALI the pulmonary lesion is typically transient. The pO2 levels
Transfusion-Related Acute Lung Injury 127
return to their pretransfusion levels. Roentgenograms demonstrate rapid clearing

of the edema fluid. In one large study of this syndrome 100% of 36 patients
required oxygen support, while 72% required short term mechanical ventilation
(8). There was a subset of patients, however, with a more prolonged course. In
about 20% of cases, pulmonary infiltrates persisted for at least 7 days, but even
these patients show no evidence of permanent sequelae (8). Although data are
limited, it appears that 58% of patients die from complications related to the
pulmonary insult (5,6,8). As more cases are reported, there is a growing appreci-
ation of the risk of death. In a recent report two children with malignant
osteopetrosis died of TRALI following transfusion of single donor platelets (9).
For obvious reasons, TRALI is an important clinical diagnosis.
IV. DIFFERENTIAL DIAGNOSIS

A. Circulatory Overload
Respiratory distress, cyanosis, and tachypnea are prominent features of circula-
tory overload (7,10). Tachycardia and hypertension are usually present. Symp-
toms begin within several hours of transfusion of any type of blood component.
The setting is often rapid infusion in either very young or old patients.
B. Bacterial Contamination
Fever, hypotension, and vascular collapse are prominent features of bacterial
contamination (11). Respiratory distress is infrequently observed. Patients may
present with disseminated intravascular coagulation. The onset of symptoms is
within 12 hours of transfusion of cellular or plasma-containing blood compo-
nents. Although platelet concentrates or apheresis platelets are the most frequently
implicated components (12), red cell concentrates may be involved.
C. Anaphylactic Transfusion Reactions

Respiratory distress, cyanosis related to laryngeal edema, and bronchospasm, not
pulmonary edema, are the dominant symptoms of this complication (13,14).
Erythema and urticarial eruption are prominent and typically involve conflunaries
of the trunk, face, and neck. Hypertension is usually severe and frequently occurs
within seconds to minutes after the initiation of transfusion of a plasma protein
containing blood component or derivative (as little as a few milliliters). Fever is
not a manifestation of anaphylactic reactions.
128 Popovsky
V. INCIDENCE
The incidence rate of TRALI is unknown. In one study from the mid-1980s, 1 in
5000 plasma-containing transfusions were associated with this reaction (6,8). This
study took place during a period in which blood banks in the United States were
converting to red cell concentrates containing significantly less plasma (reduced
from an average of 100 to approximately 4060 mL); therefore, one might assume
that the frequency has decreased. On the other hand, the number of reports in the
literature has increased dramatically. From 1951 when the syndrome was first
described until 1985, there were fewer than 40 case reports in the English
language literature (1531). Since then descriptions of at least 94 recipient
reactions have appeared in the literature (8,9,3248,52,53,6769), and the author
is aware of an additional 65 unpublished cases.
In a study by Clarke et al. (41), 46 (0.34%) of 14,602 transfusion were
associated with severe respiratory reactions to random donor platelets over a
2-year period in a single general hospital. Rare reactions were also seen in
recipients of red cells. The reactions are most frequently seen in patients with
hematological malignancies. The average age of the platelets at transfusion was
4.5 days, significantly greater than that of the controls.
Other investigators have failed to find an at-risk population, although
recurring cases have been rarely described (39). The male:female ratio among
reactors is approximately 1:1 Cases have been described in both very young and
elderly transfusion recipients (from one month to 87 years) (8,39). Most patients
had no history of transfusion reactions, and there are no common denominators,
disease associations, or underlying conditions that necessitated the transfusions.
A report from New Zealand found an incidence of 0.001% of ARDS
reactions from 1981 to 1987 from a total of 440,000 transfused blood components
(49). The nearly 20-fold difference in reported incidents in these studies may
reflect the level of awareness at a medical center, rather than changes in
incidence. At the Mayo Clinic, for instance, transfusions are often administered
under the supervision of blood bank staff who are trained to identify transfusion
reactions.
There is reason to believe that TRALI may be significantly underdiagnosed.
In a series of 40 patients with pulmonary edema in the operative setting,
Cooperman and Price (50) found that 50% of cases were attributed to circulatory
overload or an unknown cause. It is conceivable that some of these cases
represented TRALI. Culliford and colleagues (25) described a type of non-
cardiogenic pulmonary edema in three patients following cardiopulmonary by-
pass that is consistent with this diagnosis, but tests that might have supported it
were not performed.
In an analysis of transfusion-associated fatalities reported to the U.S. Food
and Drug Administration from 1976 to 1985, acute pulmonary injury was
implicated in 12.1% (31) of 256 cases (51). In this study, respiratory death from
acute-onset pulmonary edema was the third most common cause of death from
transfusion. This type of complication was more frequent than bacterial contam-
ination or anaphylaxis. The relative importance of TRALI may increase as other
important complications, specifically transfusion-transmitted hepatitis, recede as
a consequence of improved blood donor screening and testing.
Finally, in a recent study of hypotensive reactions associated with platelet-
containing products, 3 of 24 reactions were consistent with, but unrecognized
as TRALI (52). This report underscores two points: (a) that TRALI remains
underdiagnosed and (b) that it may be confused with other complications of
transfusion.
VI. IMPLICATED BLOOD COMPONENTS

As stated previously, TRALI is associated with transfusion of blood components
containing plasma. These components include whole blood, red blood cells
(prepared in citrate-phosphate-dextrose or citrate-phosphate-dextrose-adenine 1)
as well as protein-poor anticoagulant-preservative solutions (such as Additive
Solution-1 or Additive Solution-3), granulocytes collected by apheresis, platelet
concentrates, and platelets collected by apheresis and cryoprecipitate (16,15
28). In most instances the implicated blood component contains more than 60 mL
of plasma, but it is apparent that in some instances smaller quantities (e.g.,
cryoprecipitate contains only 1015 mL of plasma and platelet concentrate, 4050
mL of plasma) are sufficient to initiate the pulmonary events described above.
It is noteworthy that commercially available plasma derivatives, such as albumin,
plasma protein fraction, intravenous immune globulin, and gamma globulin, that
manufactured from large pools of plasma donors using Cohn fractionation
procedures have not been associated with any case reports.
VII. MECHANISM
Although the precise mechanism of TRALI is unknown, there are sufficient clues
to assume that it is an immune-mediated event. Unlike most immunologically
triggered transfusion reactions, in TRALI pathological antibodies are typically of
donor, rather than recipient, origin (Table 2). Numerous reports have docu-
mented the presence of human leukocyte antigen (HLA)specific antibodies or
leukoagglutinins in the plasma of the donors of implicated blood components
(3,4,8,18,26,2935,3739,42).
Popovsky and Moore (8) found such antibodies in 89% of 36 cases in one
series. In about half of the cases studied, the HLA-A or HLA-B antibodies of the
implicated donor corresponded with one or more HLA epitopes of the recipient
130 Popovsky
Table 2 Immunological
Findings in TRALI
Donor plasma
HLA-specific antibodies
Granulocyte-specific antibodies
Leukoagglutinins
Recipient plasma
HLA-specific antibodies
Granulocyte-specific antibodies
Leukoagglutinins
(8,26). Goeken and colleagues (29) as well as others have confirmed these
findings (31). In other cases, neutrophil-specific antibodies (anti-NA2, anti-5b,
and anti-NB2) have been identified in the serum of implicated units (27,32,37).
These antibodies are usually found in the blood of multiparous donors. On the
other hand, in 5% of reported cases, similar specificities of antibodies are
found in the pretransfusion serum of the transfusion recipient (5,7,8). Finally,
in 515% of cases, no antibody has been identified in either the patient or
the donor.
The likely explanation of antibodies in the donor, rather than antibodies in
the recipient, as the causative agent is that the substrate with which the
leukocyte antibodies can reactnamely, the recipients entire circulating and
marginated pool of leukocyteis far larger than the quantity of donor leukocytes
present in a single transfused component. It appears that TRALI begins with
passive transfer of antibody from donor plasma to the recipient (5,7), which sets
off a chain of reactions.
Underscoring the primacy of the antibody-mediated theory are a handful of
case reports of TRALI associated with interdonor incompatibility. In a recent
report, a 67-year-old male with acute myelogenous leukemia received a pool of
four platelet concentrates and developed the signs and symptoms of classic
TRALI. In the ensuing workup, one of the platelet donors was found to have
anti-A2 and anti-A28 in the plasma. The recipient tissue type was negative for
both A2 and A28, and the plasma did not contain HLA granulocyte antibodies.
One of the other donors of the platelets used in the pool was found to be A28
positive (53).
What is the relationship of these antibodies to this reaction? Much of the
current understanding is gleaned from studies of ARDS. Although the mecha-
nisms involved in the development of ARDS are complex, there is considerable
evidence to support a major role for complement activation in neutrophil influx
into the lung, causing damage to the pulmonary microvasculature. When comple-
ment is activated, C5a promotes neutrophil aggregation, margination, and seque-

stration in the microvasculature of the lung (54,55), which is at least partially
related to increased granulocyte adhesion (56). Experimental studies in rabbits,
as well as observations in ARDS patients, suggest that when complement-
activated neutrophils release their proteases, oxygen radicals, and acidic lipids,
the underlying pulmonary vascular endothelium is damaged, with subsequent
extravasation of protein-laden fluid into the adjacent interstitium and alveoli
(57,58). Larsen et al. (59) demonstrated that C5 fragments consistently produced
lung inflammation characterized by neutrophil accumulation and edema. In all
likelihood, these pathological changes account for the radiographic and clinical
findings seen in TRALI. Brittingham (20) noted that when 50 mL of blood known
to contain leukoagglutinins was transfused to a healthy person, the result was a
severe pulmonary reaction characterized by hypoxemia, pulmonary edema, hypo-
tension, and fever. This study suggested that passive transfer of leukocyte
antibodies may play an important, if not decisive, role in triggering the action (7).
As lymphocytotoxic (i.e., HLA) antibodies readily fix complement, passive
transfusion of these antibodies probably accounts for complement activation of
the sequence of events described previously.
Seeger et al. (60) have described a powerful model for understanding the
relation between donor antibodies and the development of TRALI. Using an ex
vivo rabbit lung model, these investigators found that acute lung injury charac-
terized by severe lung edema resulted from the infusion of an admixture of
complement, anti-5b, and 5b-positive human neutrophils. These changes were
seen 36 hours after infusion, which parallels the clinical presentation in humans.
Other investigators have documented a similar pathological timeline (58). When
complement, anti-5b, or 5b-positive granulocyte antigen was deleted from the
experiments, no pathological changes occurred. While these data suggest that
the correspondence of antibody specificity for a recipient epitope is important in
the pathogenesis of the respiratory decompensation observed in TRALI, cases are
left unexplained in which the HLA or neutrophil-specific antibody does not share
epitopes of the recipient. Despite the fact that approximately 12% of blood
donors have HLA-specific antibodies, TRALI is an infrequent result of transfu-
sion. Although there is a report of two episodes of TRALI involving the same
recipient and the same antibody-positive donor (39), current understanding does
not allow for a characteristic profile of at-risk blood recipient.
In all likelihood, other factors such as the character of the antibody, the
nature and distribution of the related antigen, the extent of complement activation,
and the immune status of the recipient are important variables that determine the
final clinical response (47,61). Silliman et al. (45) identified a cohort of TRALI
patients in whom no HLA or leukocyte antibodies were found. Rather, they
describe the presence of a neutrophil priming agenta lipid in the blood
components given to the patients who developed TRALI. They postulated that
132 Popovsky
this lipid developing during routine storage of blood components and that it
primes polymorphonuclear oxidase. These investigators (48) found that at the
time blood components (e.g., whole blood, red blood cells, platelet concentrate)
become outdated, they contain a priming agent that enhance polymorphonuclear
NADPH-oxidase activity by 2.1 to 2.8 fold. As this study represents the first
description of an alternative, nonantibody-mediated model of TRALI, other
laboratories will need to confirm these findings.
In a retrospective study from the same laboratory reporting the work
described above, 10 consecutive patients suspected of TRALI were evaluated for
PMN priming activity (62). These investigators used as controls patients having
febrile or urticarial reactions. They found that postreaction sera from TRALI
patients demonstrated a significant increase in PMN priming activity (2.1-fold)
compared with the patients prereaction sera as well as that of the control group.
In all cases the priming activity from serum samples was inhibited by pretreat-
ment of the isolated PMNs with 400 M WEB 2170, an inhibitor of the
platelet-activating factor receptor. Silliman and colleagues found that PMN-
generating lipids have a significant effect on lung function in an isolated, perfused
rat lung model (63). When lipopolysaccharide formed at day 42 of red blood cell
storage was infused, significant changes in pulmonary artery pressure and lung
weight were observed.
One other mechanism that merits discussion involves cytokines. Numerous
laboratories have shown that cytokines play a central role in the modulation of
inflammatory immune responses. Several reports implicate cytokines, including
tumor necrosis factor (TNF) and interleukin-8 in the pathogenesis of IgG-
mediated hemolytic transfusion reactions (HTRS) (64,65). Of interest is the
observation that patients experiencing HTRs may have hypoxemic or hypercapnic
respiratory failure or both (66). It may be relevant that during HTRs the
concentrations of TNF increase and TNF has been implicated in the development
of septic ARDS (66). It has been suggested that the release of significant
quantities of TNF from degranulating neutrophils may contribute to the injury of
pulmonary capillary endothelium seen in ARDS. How these observations pre-
cisely relate to TRALI remains a matter of conjecture.
VIII. DIAGNOSIS
Because there is no diagnostic test or pathognomonic sign, TRALI remains a
diagnosis of exclusion. One must rule out other causes of respiratory distress and
pulmonary edema in the transfusion setting: myocardial infarction, circulatory
overload, and bacterial infection. Normal central venous and pulmonary wedge
pressures are consistent with TRALI. The demonstration of lymphocytotoxic,
HLA-or granulocyte-specific antibodies in donor or recipient serum is strongly
suggestive of the diagnosis (7). The presence of a positive reverse lymphocyte

crossmatch between donor serum and the patients lymphocytes provides impor-
tant, supportive results, as does correspondence of antibody and antigen.
IX. TREATMENT
While the earlier reports of TRALI described a fulminant picture of respiratory
distress, it is clear that not all cases are associated with life-threatening compli-
cations. Respiratory support should be as intensive as dictated by the clinical
picture. In almost all cases oxygen supplementation is necessary, and if the
hypoxemia is severe, intubation and mechanical ventilation are important inter-
ventions (6,8). Once a diagnosis is seriously entertained, therapeutic measures
should be started promptly. Pressor agents may be useful in case of sustained
hypertension. Corticosteroids are probably of marginal value, and diuretics have
no role because the underlying pathology involves microvascular injury, rather
than fluid overload (36).
X. MANAGEMENT OF FUTURE TRANSFUSIONS

In the majority of cases in which a donor or antibody has been implicated,
no special measures are necessary to manage future transfusions of plasma-
containing components. However, if antibody in the recipient has been identified,
it has been suggested that filters that significantly reduce leukocyte content
be used for transfusion of cellular components. However, given the low re-
ported incidence of TRALI, no data are available to either support or refute
this approach.
XI. PREVENTION
Strategies aimed at reducing the incidence of reactions are complicated by the
absence of a profile of those recipients at greatest risk and by the lack of a
diagnostic test. Any effort will suffer for lack of sensitivity or specificity.
Therefore, measures must focus on limiting exposure. As a consequence, there
is no consensus on the subject, and few blood collectors have taken specific
steps. However, some investigators have made recommendations that include the
following (5):
1. Donors who have been implicated in TRALI should be permanently
deferred or subsequent donations limited to the production of frozen-
deglycerolized or washed red blood cells. Such steps will prevent the
use of plasma from blood components prepared from these donors.
134 Popovsky
2. Multiparous donors (two or three pregnancies as a suggested threshold

number) should be prospectively identified and their blood either
screened for HLA or granulocyte antibodies or diverted for uses other
than whole blood, fresh frozen plasma, or single donor apheresis
platelets.
If the second group of recommendations were followed, these measures would be
expected to decrease the frequency of TRALI by limiting any components
containing large amounts of plasma from donors most likely to have produced
these antibodies. However, there are significant weaknesses to this approach.
First, donor histories are not necessarily accurate, and women having had
alloimmunizing exposures to fetal leukocytes through either abortion or ectopic
pregnancy may neglect to report this part of their medical history. Second, this
fails to address the donor who has been immunized by transfusion, unless the
donors blood was tested for HLA or granulocyte-specific antibodies. Third, red
blood cells, random donor platelet concentrates, and cryoprecipitate contain less
than 50 mL of plasma, but such components have been associated with TRALI.
To make this approach more effective, one would need to defer all blood
component production from multiparous donors (estimated to be 530% of donor
bases), but this would result in a significant loss of many safe blood donors.
Finally, tests for HLA and granulocyte antibodies are time consuming and are
not routinely available in many blood centers or hospital blood banks. Clearly,
more sensitive and specific measures are needed for a more effective strategy to
prevent TRALI.
XII. CONCLUSION
TRALI is an important, life-threatening transfusion complication, the incidence
rate of which is unknown but is most likely underrecognized. There may be
several mechanisms that lead to a common pathway of microvascular insult and
alveolar injury. In most cases, this syndrome is reversible, particularly with rapid
diagnosis and treatment.
REFERENCES
1. Carilli AD, Ramanamurty MV, Chang YS, et al. Noncardiogenic pulmonary edema
following blood transfusion. Chest 1978; 74:310312.
2. Kernoff PBA, Durrant IJ, Rizza CR, Wright FW. Severe allergic pulmonary oedema
after plasma transfusion. Br J Haematol 1972; 23:777781.
3. Wolf CFW, Canale VC. Fatal pulmonary hypersensitivity reaction to HLA-incompati-
ble blood transfusion: report of a case and review of the literature. Transfusion 1976;
16:135140.
4. Ward HN. Pulmonary infiltrates associated with leukoagglutinin transfusion reac-

tions. Ann Intern Med 1970; 73:689694.
5. Popovsky MA, Chaplin HC Jr, Moore SB. Transfusion-related acute lung injury: a
neglected, serious complication of hemotherapy. Transfusion 1992; 32:589592.
6. Popovsky MA. The role of leukocyte depletion in prevention of transfusion-related
acute lung injury. In: Sweeney J, Heaton A, eds. Clinical Benefits of Leukodepleted
Blood Products. Austin, TX: R.G. Landes, 1995.
7. Popovsky MA. Transfusion-related acute lung injury. In: Popovsky M, ed. Transfu-
sion Reactions. Bethesda, MD: American Association of Blood Banks, 1996.
8. Popovsky MA, Moore SB. Diagnostic and pathogenetic considerations in transfusion
related acute lung injury. Transfusion 1985; 25:573577.
9. Jeter EK, Madysatha P, Rodriguiz RM, Key LL. Transfusion-associated acute lung
injury induced by granulocytebound cytophilic IgG (abstr). Transfusion 1994;
34(suppl):65S.
10. Stack G, Judge JV, Snyder EL. Febrile and non-immune transfusion reactions. In:
Rossi EC, Simon TL, Moss GS, Gould SA, eds. Principles of Transfusion Medicine.
2d ed. Baltimore, MD: Williams & Wilkins, 1996.
11. Goldman M, Blajchman MA. Blood product-associated bacterial sepsis. Transfus
Med Rev 1991; 5:73-83.
transfusion medicine. In: Nance SJ, ed. Blood Safety: Current Challenges. Bethesda,
MD: American Association of Blood Banks, 1992:213228.
13. Vengelen-Tyler V, ed. Technical Manual. 12 ed. Bethesda, MD: American Association
of Blood Banks, 1996:549.
14. Pineda AA, Taswell HF. Transfusion reactions associated with anti-IgA antibodies:
report of four cases and review of the literature. Transfusion 1975; 15:1015.
15. Barnard RD. Indiscriminate transfusion: a critique of case reports illustrating hyper-
sensitivity reactions. NY State J Med 1951; 51:23992401.
16. Ward HN, Lipscomb TS, Cawley LP. Pulmonary hypersensitivity reaction after blood
transfusion. Arch Intern Med 1968; 122:362366.
17. Byrne JP, Dixon JA. Pulmonary edema following blood transfusion reaction. Arch
Surg 1971; 102:9194.
18. Andrews AT, Zmijewski CM, Bowman HS, Reihart JK. Transfusion reaction with
pulmonary infiltration associated with HLA-specific leukocyte antibodies. Am J Clin
Pathol 1976; 66:483487.
19. Philipps E, Fleischner FG. Pulmonary edema in the course of a blood transfusion
without overloading the circulation. Dis Chest 1966; 50:619623.
20. Brittingham TE. Immunologic studies on leukocytes. Vox Sang 1957; 2:242248.
21. Silver H. Non-cardiac pulmonary edema following blood transfusion. Hartford Hosp
Bill 1973; 28:336338.
22. Thompson JS, Severson CD, Parmely MJ, et al. Pulmonary hypersensitivity
reactions induced by transfusion of non-HLA leukoagglutinins. N Engl J Med 1971;
284:11201125.
23. Felbo M, Jensen KG. Death in childbirth following transfusion of leukocyte-
incompatible blood. Acta Haematol (Basel) 1962; 27:113119.
136 Popovsky
24. Dubois M, Lotze MT, Diamond WJ, et al. Pulmonary shunting during leuko-
agglutinin-induced non-cardiac pulmonary edema. JAMA 1980; 244:21862189.
25. Culliford AT, Thomas S, Spencer FC. Fulminating noncardiogenic pulmonary edema.
A newly recognized hazard during cardiac operations. J Thorac Cardiovasc Surg
1980; 80:868875.
26. Popovsky MA, Abel MD, Moore SM. Transfusion-related acute lung injury associ-
ated with passive transfer of antileukocyte antibodies. Am Rev Respir Dis 1983;
128:185189.
27. Yomtovian R, Kline W, Press C, et al. Severe pulmonary hypersensitivity associated
with passive transfusion of a neutrophil-specific antibody. Lancet 1984; 1:244246.
28. Reese EP Jr, McCullough JJ, Craddock PR. An adverse pulmonary reaction to
cryoprecipitate in a hemophiliac. Transfusion 1975; 15:583588.
29. Goeken NE, Schulak JA, Nghiem DD, et al. Transfusion in donor-specific blood
transfusion patients resulting from transfused maternal antibody. Transplantation
1984; 38:306307.
30. Dhainaut JF, Brossard Y, Dimercurio JP, et al. Oedeme pulmonaire lesionnel gravis-
sisme au cours du 20 echange plasmatique pour incompatibilite Rhesus. In: Hemo-
perfusion, Echanges Plasmatique en Reanimation. Paris: Expansion Scientifique
Francaise, 1981:291298.
31. Campbell DA, Swartz RD, Waskeritz JA, Turcotte JG. Leukoagglutination with
interstitial pulmonary edema (letter). Transplantation 1982; 34:300301.
32. VanBuren NL, Stroncek DF, Clay MA, et al. Transfusion-related acute lung in-
jury caused by an NB2 granulocyte-specific antibody in a patient with thrombotic
thrombocytopenic purpura. Transfusion 1990; 30:4245.
33. Eastlund T, McGrath PC, Britten A, Propp R. Fatal pulmonary transfusion reaction
to plasma containing donor HLA antibody. Vox Sang 1989; 56:6366.
34. Eastlund DT, McGrath PC, Burkart P. Platelet transfusion reaction associated with
interdonor HLA incompatibility. Vox Sang 1988; 55:157160.
35. OConnor JC, Strauss RG, Goeken NE, Knox LB. A near-fatal reaction during
granulocyte transfusion of a neonate. Transfusion 1988; 28:173176.
36. Levy GJ, Shabot MM, Hart ME, et al. Transfusion-associated noncardiogenic pulmo-
nary edema: Report of a case and a warning regarding treatment. Transfusion 1986;
26:278281.
37. Nordhagen R, Conradi M, Dromtorp SM. Pulmonary reaction associated with
transfusion of plasma containing anti-5b. Vox Sang 1986; 51:102107.
38. DeWolf AM, Van Den Berg BW, Hoffman HJ, Zundert AA. Pulmonary dysfunction
during one-lung ventilation caused by HLA-specific antibodies against leukocytes.
Anesth Analg 1987; 66:463467.
39. Leaman B, Anderson D, Walker I, Heddle N. Two episodes of non-cardiogenic
pulmonary edema in a patient caused by the same donors blood (abstr). Transfusion
1994; 34(suppl):26S.
40. Win N, Kaye T, Jones L, Knowles S. Transfusion-related acute lung injury (TRALI)
(abstr). Transfus Med 1995; 5(suppl 1):33.
41. Clarke G, Podlosky L, Petrie L, Boshkov L. Severe respiratory reactions to random
donor platelets: an incidence and nested case-control study (abstr). Blood 1994;
84(suppl 1):465a.
42. Lin M, Chen CC, Hong CC, et al. Transfusion-associated respiratory distress in
Taiwan. Vox Sang 1994; 67:372376.
43. Reissman P, Manny N, Shapira SC, et al. Transfusion-related adult respiratory distress
syndrome. Isr J Med Sci 1993; 29:303307.
44. Popovsky MA, Chaplin HC, Moore SB. Transfusion-related acute lung injury (letter).
Transfusion 1993; 33:444445.
45. Silliman C, Pitman J, Thurman G, Ambruso DR. Neutrophil (PMN) priming agents
develop in patients with transfusion-related acute lung injury (abstr). Blood 1992;
80:(suppl 1):261a.
46. Bux J, Hoch J, Bindl L, et al. Transfusionassoziierte acute Lungeninsuffizienz. Dtsch
Med Wochenschr 1994; 119:1924.
47. Gans ROB, Duurkens VAM, van Zundert AA, Hoorntje SJ. Transfusion-related acute
lung injury. Intensive Care Med 1988; 14:654657.
48. Silliman CC, Thurman GW, Ambruso DR. Stored blood components contain agents
that prime the neutrophil NADPH oxidase through the platelet-activating-factor
receptor. Vox Sang 1992; 53:133136.
49. Henderson RA. Acute transfusion reactions. NZ Med J 1990; 103:509511.
50. Cooperman LH, Price HL. Pulmonary edema in the operative and postoperative
period: a review of 40 cases. Ann Surg 1970; 172:833891.
fusion 1990; 14:654670.
52. Hume HA, Popovsky MA, Benson K, Glasman AB, et al. Hypotensive reactions: a
previously uncharacterized complication of platelet reaction? Transfusion. In press.
53. Virchis AE, Contreras M, Novarrete C, Jan-Mohamed R, et al. Transfusion-related
acute lung injury (TRALI) due to inter-donor incompatibility (abstr). Blood. In press.
54. Jacob JS, Craddock PR, Hammerschmidt DE, Moldow CF. Complement-induced
granulocyte aggregation: an unsuspected mechanism of disease. N Engl J Med 1980;
302:789794.
55. Hammerschmidt DE, Jacob HS. Adverse pulmonary reactions to transfusion. Adv
Intern Med 1982; 27:511530.
56. Craddock PR, Hammerschmidt DE, Moldow CF, et al. Granulocyte aggregation as a
manifestation of membrane interactions with complement: possible role in leukocyte
margination microvascular occlusion, and endothelial damage. Semin Hematol 1979;
16:140147.
57. Shasby DM, Van Benthuysem KM, Tate RM, et al. Granulocytes mediate acute
edematous lung injury in rabbits and in isolated rabbits lungs perfused with phorbol
myristate acetate: role of oxygen radicals. Am Rev Respir Dis 1982; 125:443-447.
58. Sznajder JI, Fraiman A, Hal JB, et al. Increased hydrogen peroxide in the expired
breath of patients with acute hypoxemic respiratory failure. Chest 1989; 96:606612.
59. Larsen GL, McCarthy K, Webster RO, et al. A differential effect of C5a and C5a des
Arg in the induction of pulmonary inflammation. Am J Pathol 1980; 100:179192.
60. Seeger W, Schneider U, Kreusler B, et al. Reproduction of transfusion-related acute
lung injury in an ex vivo lung model. Blood 1990; 76:1438444.
61. Westaby S. Mechanisms of membrane damage and surfactant depletion in acute lung
injury. Intensive Care Med 1986; 12:25.
62. Silliman CC, Paterson A, Dickey WO, Stroncek D, Popovsky MA, Caldwell SA,
138 Popovsky
Ambruso DR. The association of biologically active lipids with the development of
transfusion-related acute lung injury: a retrospective study. Transfusion. In press.
63. Silliman C, Allard J, Voelkel N, Ambruso D. Plasma and lipids from stored packed
red cells (PRBCs) cause acute lung injury in an animal model (abstr). Blood 1996.
64. Davenport RD, Polak TJ, Kunkel SL. White cell-associated procoagulant induced by
ABO incompatibility. Transfusion 1994; 34:943949.
65. Davenport RD, Streiter RM, Sandiford TJ, Kunkel SL. Interleukin-8 production in
red blood cell incompatibility. Blood 1990; 76:24392442.
66. Capon SM, Goldfinger D. Acute hemolytic transfusion reaction, a paradigm of the
systemic inflammatory response: new insights into pathophysiology and treatment.
Transfusion 1995; 35:513520.
67. Boshkov L, Silliman C, Clarke G, Dickey W, Ambruso D. Transfusion-related acute
lung injury (TRALI) following platelet transfusion (abstr). Blood 1995; 86:1403.
68. Lindgren L, Ylihankala A, Holme L, Koskimies S, et al. Transfusion-related acute
lung injury (TRALI) after fresh-frozen plasma in a patient with a coagulopathy. Acta
Anaesthes Scand 1996; 40:641644.
69. Bux J, Becker F, Seeger W, Kilpatrick D, et al. Transfusion-related acute lung injury
due to HLA-A2-specific antibodies in recipient and NB1-specific antibodies in donor
blood. Br J Haem 1996; 93:707713.
7
Clinical Effects of the
Immunomodulation Associated
with Allogeneic Blood Transfusions
Jos O. Bordin
Escola Paulista de Medicina, So Paulo, Brazil
Morris A. Blajchman
McMaster University and Hamilton Centre Canadian Red Cross
Society, Hamilton, Ontario, Canada
I. INTRODUCTION
Considerable data have accumulated over the past two decades suggesting that
allogeneic blood transfusions (ABTs) may be associated with immunomodulation
in recipients (1,2). Depending on patient category, this immunomodulatory effect
can be clinically detrimental or beneficial (Table 1). Thus, it has been suggested
that the ABT-associated immunomodulation might adversely affect the overall
clinical outcome in patients undergoing curative surgery for a variety of malig-
nant tumors by downregulating the recipients immune system permitting unreg-
ulated tumor cell growth (35). In addition, many observational studies have
implicated this ABT-associated immunomodulatory effect with an increased
prevalence of postoperative bacterial infection episodes after abdominal, open-
heart, and orthopedic surgery (69). These adverse clinical effects, however, have
yet to be proven to be associated with ABT. In contrast, there is clear evidence
that ABT-associated immunomodulation can be beneficial for selected categories
of patients, such as: increasing allograft survival in renal allograft recipients (10);
decreasing the recurrence rate in women with recurrent spontaneous abortion
(11); and possibly reducing the relapse rate in patients with Crohns disease (12).
139
140 Bordin and Blajchman
Table 1 Potential Clinical Effects of Immunomodulation Associated with Allogeneic

Blood Transfusions
1. Adverse
Increase in cancer recurrence rate
Increase in prevalence of postoperative bacterial infection
2. Beneficial
Improvement of renal allograft survival
Reduction in prevalence of spontaneous recurrent abortion
Decrease in relapse rate of Crohns disease
Adoptive immunotherapy in chronic myelogenous leukemia reducing the relapse
rate after bone marrow transplantation
This chapter summarizes the currently available evidence relating to the clinical
effects of ABT-associated immunomodulation.
II. ALLOGENEIC BLOOD TRANSFUSIONS

AND CANCER RECURRENCE
The possible association between perioperative ABT and increased cancer recur-
rence was first suggested by Gantt in 1981 (13). Since then, over 100 retrospec-
tive and prospective studies evaluating the effect of perioperative ABT on cancer
recurrence and/or overall outcome in patients with a malignancy undergoing
cancer surgery have been published (1416). However, because the available data
are mostly from retrospective or prospective observational studies, they are not
regarded as being conclusive, and the question as to whether or not ABTs
influence tumor growth is still unresolved.
Evidence for a possible adverse effect of ABT in patients with a malignancy
has been reported in approximately 60% of observational studies evaluating
patients with a large variety of malignancies, including breast, lung, kidney,
prostate, stomach, cervix, vulva, head and neck, larynx, soft tissue, bone, and liver
metastases; the remaining 40% showing no deleterious effect (3,14).
Since 1993, three prospective, randomized, controlled clinical trials (RCTs)
have been published that investigated the association of perioperative ABT with
colorectal cancer recurrence (1719) (Table 2). An RCT conducted in Rotterdam,
The Netherlands, randomized 475 colorectal cancer patients to receive either
allogeneic or autologous blood products perioperatively (17). This study reported
no difference in either the cancer-specific survival rate at 4 years (68% vs. 64%,
p = 0.60) or the cancer recurrence rate at 4 years (62% vs. 56%, p = 0.50).
However, the data showed that the relative risk (RR) of cancer recurrence
was increased in transfused patients who had received either allogeneic (RR =
Table 2 Results of Prospective, Randomized Clinical Trials Analyzing the Effect of Perioperative Blood Transfusions on Clinical
Outcome (recurrence rate and/or overall prognosis) in Patients with Malignant Tumors
Aa Bb Cc
Tumor site Colorectal Colorectal Colorectal

Number of patients 475 120 697
Number of subjects per arm (1) 133 (1) 48 (1) 337
(2) 112 (2) 52 (2) 360
Type of blood transfusion per arm (1) Autologous BC-PRBCs (1) Autologous BC-PRBCs (1) LR-Allogeneic PRBCs
(2) Allogeneic BC-PRBCs (2) Allogeneic BC-PRBCs (2) Allogeneic BC-PRBCs
Rate of tumor recurrence per arm (1) 62% (1) 16.7% Transfusedd = 30%
(2) 56% (2) 28.9% Not transfused = 26%
p = 0.50 p = 0.11 p = 0.22
Blood Transfusion and Immunomodulation
Relative risk of tumor recurrence compared Arm (1) 1.8; p = 0.04 Arm (1) 0.96; p = 0.96 Arm (1) 0.91; p = 0.53
with untransfused patients Arm (2) 2.1; p = 0.01 Arm (2) 7.01; p = 0.006
Cancer-specific survival rate (1) 68% (1) vs (2) Transfusedd = 69%
(2) 64% Not transfused = 81%
p = 0.60 p = 0.20 p < 0.001
PRBCs = Packed red blood cells; BC-PRBCs = buffy coatreduced PRBCs; LR-Allogeneic PRBCs = leukocyte-reduced (48 hours old) allogeneic PRBCs.
aFrom Ref. 17.
bFrom Ref. 18.
cFrom Ref. 19.
dPatients transfused with either LR-PRBCs or BC-PRBCs.
141
2.1, p = 0.01) or autologous (RR = 1.8, p = 0.04) blood, compared to subjects that
had not been transfused (17). Since this study actually compared outcomes among
patients receiving autologous blood versus recipients of buffy coatreduced
allogeneic blood, it is possible that the degree of leukocyte removal resulting from
buffy coat reduction may have diminished the ABT-associated immunomod-
ulatory effect that might have been observed had unmodified allogeneic blood
been used. Buffy coatreduced allogeneic blood represents the removal of
approximately 80% of the leukocytes present in a unit of whole blood. Interest-
ingly, this level of leukocyte reduction has been shown to prevent donor-specific
T-cell responsiveness associated with nonbuffy coatreduced blood (20).
Thus it is possible that the widespread use in Western Europe of buffy-coat
reduced allogeneic cellular blood components may significantly lessen the
immunomodulatory risk of cellular blood component transfusions in Western
Europe compared to that occurring in allogeneically transfused patients in North
America (20).
The second RCT, from Munich, Germany, randomized 120 patients with
potentially curative colorectal carcinoma to receive either leukocyte-reduced
allogeneic blood or predeposited autologous blood. After a median follow-up of
22 months, tumor recurrence was detected in 28.9% of the patients transfused
with allogeneic blood compared with 16.7% of the subjects transfused with
autologous blood (p = 0.11). The authors also reported that the need for allo-
geneic blood was an independent predictor of cancer recurrence (RR = 7.01; 95%
Cl = 1.7727.75; p = 0.006) (18).
The third prospective RCT was also from The Netherlands and compared
recipients of leukocyte-reduced allogeneic packed red blood cells (PRBCs) with
buffy coatreduced allogeneic PRBC transfusions in patients also undergoing
curative surgery for colorectal cancer. This study did not detect a difference in
patient survival between the two transfusion arms. As in the Rotterdam study,
transfused patients who received either allogeneic or autologous blood had a
significantly lower 3-year survival than untransfused study subjects (69% vs.
81%, p < 0.001) (19). Interestingly, recurrence rates appeared not to be affected
by the need for transfusion (30% vs. 26% p = 0.22).
In 1993, the results of the available observational studies of patients with
colorectal carcinoma were subjected to meta-analysis by two groups of investi-
gators independently (4,5). The literature-based meta-analysis from the first team
of investigators reported that the RRs of cancer recurrence, cancer-associated
death, and death from any cause in subjects transfused with allogeneic blood were
1.80, 1.76, and 1.63, respectively (4). These authors concluded that their analysis
supported the hypothesis that perioperative ABT was associated with an increased
risk of cancer recurrence and death from this disease (4). The second team of
meta-analysts concluded that perioperative ABT increased the RR of cancer
recurrence by 37% (95% Cl = 2056%) (5).
Blood Transfusion and Immunomodulation 143
A third meta-analysis, performed to try to show a quantitative synthesis of

the published observational studies by the random-effects method before any
adjustment for all known confounders, concluded that the risk of ABT-related
adverse outcome ranged from 6% in breast cancer to 262% in head and neck
carcinoma. In this analysis, the ABT-associated adverse outcome was found to be
statistically significant for all cancer sites, except breast (15).
The results of the three RCTs mentioned above (1719) were subjected
recently to yet another meta-analysis (16). For the latter meta-analysis, subjects
receiving buffy coatreduced allogeneic RBC transfusions were allocated to the
treatment group, while the control subjects were those assigned to receive either
autologous or leukocyte-reduced allogeneic PRBCs. The results of this analysis
gave an RR very close to unity, not supporting the hypothesis that ABT was
associated with a deleterious effect. However, it is important to point out that this
analysis could not rule out an ABT effect smaller than a 33% increase in risk of
cancer recurrence (16).
The association of transfusion history and cancer risk has also been
evaluated in approximately 37,000 cancer-free women aged between 55 and 69
years. The results of this study indicate that the RR is 2.20 (95% Cl = 1.353.58)
for non-Hodgkins lymphoma and 2.53 (95% Cl = 1.34 4.78) for renal carci-
noma (21). Thus, despite considerable advances in our knowledge over the past
decade about risk factors associated with malignancy, the issue of the association
between ABT and cancer recurrence is still unclear (22).
III. ALLOGENEIC BLOOD TRANSFUSIONS

AND BACTERIAL INFECTION
The relationship between ABT-associated immunomodulation and postoperative
bacterial infectious complications has been reviewed recently by several authors
(1,9,16,23). Again, most of the available clinical data are from uncontrolled
studies, and unequivocal evidence for the existence of an adverse ABT effect
relating allogeneic transfusions as an independent prognostic factor for post-
operative septic complications has not yet been established.
Over the past 5 years, six prospective RCTs have evaluated the impact of
blood transfusions on the prevalence of postoperative bacterial infections in blood
transfusion recipients (7,8,17,2426) (Table 3). In 1992, an RCT from Denmark
reported on 197 patients who underwent curative colorectal surgery (7). Study
subjects were assigned randomly to receive either leukocyte-reduced allogeneic
whole blood or nonleukocyte-reduced allogeneic whole blood (7). The re-
sults showed that patients transfused with nonleukocyte-reduced allogeneic
whole blood had a significantly higher prevalence of postoperative infections
Table 3 Results of Prospective, Randomized Clinical Studies Analyzing the Effect of Perioperative Allogeneic, Leukocyte-Reduced
Allogeneic and Autologous Transfusions on the Prevalence of Postoperative Bacterial Infections in Recipients
144
Aa Bb Cc Dd Ee Ff
Type of surgery Colorectal Colorectal Colorectal Cardiac Colorectal Colorectal

Number of patients 197 475 120 914 589 697
Number of patients (1) 48 (1) 133 (1) 53 (1) 285 (1) 118 (1) 215
per study arm (2) 56 (2) 112 (2) 37 (2) 287 (2) 142 (2) 231
(3) 294 (3) 155 (3) 251
Type of blood (1) LR-WB (1) Autologous (1) Autologous (1) LR-BC-PRBCs (1) LR-BC-PRBCs (1) LR-Allogeneic
component (stored) BC-PRBCs PRBCs (fresh) (stored) PRBCs
transfusion per (2) Whole (2) Allogeneic (2) Allogeneic (2) LR-BC-PRBCs (2) BC-PRBCs (2) Allogeneic
study arm Blood BC-PRBCs BC-PRBCs (stored) (3) Not transfused BC-PRBCs
(3) BC-PRBCs (3) Not transfused
Prevalence rate (1) 2% (1) 25% (1) 12% (1) 16.7% (1) 0% (1) 42%
of infection per (2) 23% (2) 27% (2) 27% (2) 17.7% (2) 18.3% (2) 36%
study arm (3) 22.7% (3) 1% (3) 24%
Statistical (1) vs. (2); (1) vs. (2); NS (1) vs. (2); (1) vs. (2); NS (1) vs. (2), (1) vs. (2),
significance p < 0.01 p < 0.05 (1) and (2) vs. (3); p < 0.0001 p < 0.75
p < 0.05 (1) vs. (3); (1) and (2) vs. (3);
p < 0.0001 p < 0.01
PRBCs = packed red blood cells; LR-WB (stored) = leukocyte-reduced whole blood filtered after storage); Autologous BC-PRBCs = autologous buffy
coatreduced PRBCs; Allogeneic BC-PRBCs = allogeneic buffy coat-reduced PRBCs; NS = not significant.
aFrom Ref. 7.
bFrom Ref. 17.
cFrom Ref. 8.
dFrom Ref. 24.
eFrom Ref. 25.
Bordin and Blajchman
fFrom Ref. 26.

than patients who received 99.98% leukocyte-reduced allogeneic blood (23% vs.
2%; p < 0.01).
In another RCT, which had been designed to measure cancer-related
prognosis, no difference was detected in the prevalence of postoperative infection
in study patients who were transfused with allogeneic buffy coatreduced PRBCs
compared to those who received autologous whole blood (17). In this study,
however, transfused subjects who received either allogeneic or autologous red
blood cell transfusions had a higher prevalence of postoperative infection than
those who were not transfused (17).
In the RCT from Munich, in which study subjects were assigned to receive
either autologous PRBCs or buffy coatreduced PRBCs, a significantly higher
postoperative infection rate in patients transfused with buffy coatreduced
PRBCs was seen compared to that in subjects transfused with autologous PRBCs
(27% vs. 12% p = 0.036) (8). Using multivariate regression analysis to adjust for
other risk factors, the RR of postoperative infections in the allogeneic group
versus that in the autologous transfusion group was 2.84 (p = 0.047; 95% Cl =
1.027.98). The observed number of noninfectious complications was similar in
the two groups of patients. In contrast to that seen in the study of Busch et al.
(17), autologous blood recipients did not have a greater prevalence of postoper-
ative infectious complications than that seen in those who were not transfused (8).
A three-arm RCT including 914 cardiac surgery patients has been re-
ported from Leiden, The Netherlands (24). In this study, patients were assigned
randomly to receive allogeneic leukocyte-reduced buffy coatreduced PRBCs
filtered within 24 hours of collection, allogeneic leukocyte-reduced buffy coat
reduced PRBCs filtered after storage, or nonleukocyte-reduced buffy coatreduced
PRBCs. Patients who received allogeneic leukocyte-reduced buffy coatreduced
PRBCs showed a lower prevalence of both postoperative infection (22.7% vs.
17.3%; p < 0.05) and mortality rate than those transfused with nonleukocyte-
reduced buffy coatreduced PRBCs. This study, however, detected no difference
in mortality or in the prevalence of postoperative infection (16.7% vs. 17.7%)
between the two leukocyte-reduced groups of study subjects (24).
Another RCT randomly assigned 589 colorectal carcinoma patients to
receive either allogeneic poststorage leukocyte-reduced buffy coatreduced
PRBCs or allogeneic buffy coatreduced PRBCs (25). The results indicated that
patients transfused with allogeneic poststorage leukocyte-reduced buffy coat
reduced PRBCs had a significantly lower prevalance of postoperative bacterial
infections than subjects who received allogeneic buffy coatreduced PRBCs (0%
vs. 18.3%, p < 0.001) (25).
Recently, yet another large multicenter RCT randomly assigned 697
patients undergoing colorectal surgery to receive either allogeneic leukocyte-
reduced buffy coatreduced PRBCs or allogeneic buffy coatreduced PRBCs.
Leukocyte reduction was performed within 48 hours of blood collection. No
difference was detected in the prevalence of postoperative bacterial infection

in the study subjects in the two transfusion arms. There was, however, a higher
number of bacterial infectious episodes in patients who received a transfusion
(allogeneic or autologous) compared with those who did not (39% vs. 24%, p >
0.01) (26).
The results of three of these RCTs (8,17,19) were subjected recently to a
meta-analysis (16). The results did not support the hypothesis that a deleterious
ABT-associated effect occurred; however, the author could not rule out an ABT
effect of a <33% increase in infectious risk (16).
It has been suggested that the risk of postoperative infection might increase
with the number of allogeneic blood units transfused. In a study of patients who
underwent surgery for a penetrating colonic injury, the calculated risk of a
bacterial infection was 7.5% for untransfused patients and 25, 37, and 57%,
respectively, for patients transfused with 15, 6 9, or 10 allogeneic blood units
(27). Similarly, a retrospective analysis of patients undergoing surgery for gastric
carcinoma reported that those patients who developed postoperative infections
had received a higher number of allogeneic blood products compared to those
who did not (28). Using multiple logistic and receiver operating characteristic
(ROC) curve analysis, a recent prospective study in 267 patients with colorectal
cancer reported that the incidence of infection was significantly higher in patients
transfused with allogeneic blood (28.8%) than in either patients transfused with
autologous blood (8.0%) or those who had not been transfused (6.3%) (p =
0.001). These results also showed a significant trend associated with an increasing
number of allogeneic blood units transfused and risk of infection (p < 0.0002) (6).
In a recently reported RCT, colorectal cancer patients transfused perioperatively
with more than three units of PRBCs had a higher corrected RR for postopera-
tive infections than patients transfused with one to three units (3.6 vs. 1.6; p <
0.05) (26).
Based on the available observational studies, the postoperative bacterial
infection rate in allogeneically transfused patients varied from 20 to 30% com-
pared to 510% for either untransfused or autologous blood recipients (1).
The definition of the term infection in such patients, however, is crucially
important. Limiting the definition of infection complication to positive cultures
underestimates prevalence, while extending the definition to include fever prob-
ably overestimates prevalence.
In conclusion, the six available European RCTs have contributed im-
portantly to our understanding about a possible association between perioperative
blood transfusion and bacterial complications; however, they give contradictory
results. In an effort to resolve controversies about causal relationships, some
investigators have recommended the use of meta-analysis. Two types of meta-
analysis can be done: one is a meta-analysis of the literature (MAL) and the other
is meta-analysis of individual patient data (MAP). In this instance, a MAL cannot
be done because of heterogeneity of the data from the six RCTs (29). It may,
however, be possible to do a MAP. Such an international collaborative individual
patient-based meta-analysis might allow for a definitive conclusion to be reached
as to whether ABTs increase susceptibility to postoperative infection. Such a
study has been proposed recently (23).
IV. OTHER CLINICAL MANIFESTATIONS OF ALLOGENEIC

BLOOD TRANSFUSIONASSOCIATED
IMMUNOMODULATION
A. Blood Transfusion and Allograft Transplantation
Since the report from Opelz et al. over 25 years ago, it is widely accepted that
ABTs improve renal allograft survival (30,31). In spite of the fact that human
leukocyte antigen (HLA) matching is one of the most important predictors for the
continuing long-term function of cadaveric renal allografts, patients receiving
ABTs have a significantly better allograft survival rate than those not transfused,
regardless of the number of HLA-A, HLA-B, and HLA-DR locus mismatches
(10,32). Even with HLA-identical sibling renal allografts, about 30% of al-
logeneically transfused recipients experience graft rejection compared to 75% of
untransfused recipients (33).
Because of the availability of improved treatment regimens for rejection
episodes and the potent action of the currently available immunosuppressive
drugs, the ABT effect on renal allograft survival has declined over the last decade
(10,34,35). Nevertheless, a collaborative multicenter study reporting on the
outcome of more than 58,000 cadaveric renal transplantation since the advent of
the use of cyclosporine reports that subjects receiving ABTs are still more likely
to have successful renal allografts than those who did not (10). This study
reported that the one-year renal allograft survival rate of subjects receiving
pretransplant ABTs was 35% higher than that observed in untransfused patients
(10). Similar results have been obtained also in patients receiving living-related-
donor renal transplants (36).
For renal allograft recipients, the following still need to be defined: the
optimal number of allogeneic blood units necessary to achieve the optimal
ABT-associated effect; the allogeneic blood component required to produce such
an effect; the optimal volume of blood required with each transfusion; the optimal
timing of the ABT to produce the effect; the coexisting hazards of the ABT-
associated immunomodulatory effect; and whether ABTs are even still needed
(32). Nonetheless, patients receiving larger numbers of ABTs have been shown
to have better one-year allograft survival rates than subjects transfused with few.
Interestingly, subjects transfused with more than 10 units show poorer overall
graft survival rates, suggesting that multitransfused patients are more likely to
develop cytotoxic antibodies and thus are at greater risk for earlier and more
severe allograft rejection (36).
Patients given leukocyte-reduced blood components such as washed RBC
concentrates or frozen-deglycerolized RBCs have been shown to have poorer
one-year cadaveric graft survival than recipients of whole blood or unmodified
RBC concentrates, indicating that the allogeneic leukocytes present in the donor
blood participate in the production of the ABT-associated beneficial effect (37).
Further studies are required to understand how ABTs induce their clinical effect
in renal transplantation; however, ABT is still considered an efficacious interven-
tion that can be useful in the management of some patients scheduled for kidney
transplantation. This treatment modality continues to be used in donor-directed
renal allograft transplantation.
B. Blood Transfusion and Recurrent

Spontaneous Abortions
The maintenance of a pregnancy depends on the immunological equilibrium
between the fetus and the maternal immune response to the fetus. When the two
spouses share HLA antigens, this immunological balance may be modified and
maternal blocking antibodies may not be formed, predisposing the woman to
recurrent abortions. Based on this theory, the use of allogeneic leukocyte transfu-
sions has been proposed as a form of immunotherapy to treat women with
recurrent spontaneous abortions (38). Overall, the results of prospective non-
randomized studies of women with recurrent spontaneous abortions indicate a
success rate of about 75% with the use of either paternal or third-party leukocytes
compared to a success rate of only 50% receiving autologous or no leuko-
cytes (1,38).
Because the therapeutic efficacy of allogeneic leukocyte infusions in
women with recurrent spontaneous abortion became increasingly controversial
and also because of the lack of a large RCT of sufficient sample size to indicate
the most appropriate treatment method for such patients, the American Society
for Reproductive Immunology (ASRI) initiated the conduct of a worldwide
collaborative individual patientbased meta-analysis to evaluate all the available
patient data relating to the efficacy of allogeneic leukocyte immunotherapy for
women with recurrent spontaneous abortion (12). This meta-analysis of individ-
ual patient data was done by two separate and independent analytic teams.
Both agreed that leukocyte immunotherapy represented an effective treatment
for patients with recurrent spontaneous abortion. Even though such an effect
appears to be small, with only 810% of affected women likely to achieve one
additional live birth, allogeneic leukocyte immunotherapy is now widely ac-
cepted as an efficacious form of treatment for women with recurrent spontaneous
abortion (11,38).
The use of allogeneic leukocyte immunotherapy has been associated with

various transfusion risks, including alloimmunization to leukocyte antigens,
neonatal graft-versus-host-disease, and congenital anomalies. Nonetheless, such
side effects are rare, and the number of affected infants (3%) in the treated group
was similar to the number of affected newborns seen in the control group (11,38).
C. Blood Transfusion and Inflammatory Bowel Disease

Following surgery in patients with Crohns disease, the recurrence rate of a bowel
obstruction or perforation has been estimated at about 50% at 10 years. Because
immunological mechanisms might be involved in the pathogenesis of this disease,
several clinical studies have analyzed whether the postoperative recurrence rate
is affected by the immunomodulatory effects associated with ABTs administered
perioperatively. The results from the available studies indicate that the recurrence
rate in the two groups is similar (37.5% in the ABT-transfused group vs. 40.5%
in the untransfused group) (1). However, all the data pooled in this evaluation had
been generated by retrospective observational studies that evaluated patients
subjected to different surgical treatments for different follow-up periods.
ABTs have also been proposed as a major risk factor in the development of
postoperative bacterial infections in Crohns disease patients in one study (39).
Such an association did not reach statistical significance in a similar study (40).
Multiple ABTs have been also associated with a significantly lower peripheral
total lymphocyte and T-cell counts following surgery in Crohns disease patients
(12). Many factors might affect the recurrence rates in patients with Crohns
disease as well as the prevalence of bacterial infections following the surgical
treatment; therefore, large well-designed RCTs are needed to ascertain the impact
of ABT in the clinical activity of patients with Crohns disease.
V. TRANSFUSION AND IMMUNE FUNCTION

The immunogenicity of soluble, particulate, or cellular antigens associated with
the major histocompatibility complex (MHC) molecules (in humans, the HLA
antigens) present in transfused blood products depends on the ability of antigen-
presenting cells (APCs) to present them to recipient T cells. In addition, co-
stimulatory signals are required to make possible the generation and amplification
of antigen-specific T-cell responses and effector function (42,41). The CD28
receptor is the major costimulatory signal for T cells, and CD28/B7 interaction
represents a critical pathway for transplantation tolerance and immune reactivity
(42,43). Currently, T-helper cells are categorized into two major subsets: Th1 and
Th2. CD28 costimulation enables the development of Th2 cytokineproducing
cells and in the absence of CD28 costimulation, T cells will not be primed to
produce Th2 cytokines. This could result in the failure to produce the Th1 subset
(43). The Th1 subset when activated produces interleukin-2 (IL-2), interferon-
(IFN-), tumor necrosis factor- (TNF-), and lymphotoxin (LT), but not IL-4 or
IL-5. On the other hand, the Th2 sub-set of helper cells produces IL-4, IL-5, IL-6,
and IL-10, but not IL-2, IFN-, TNF-, or LT. Th1 cells stimulate those cells
involved in the cellular immune response, while Th2 cells are associated with the
humoral immune response. The B7-1 protein delivers a costimulatory signal
through CD-28 and CTLA-4 T cell receptors, which regulate IL-2 secretion, while
the B7-2 protein provides a critical early costimulatory signal that results in T-cell
clonal proliferation (42). Impairment of one of the costimulatory pathway signals
thus could result in T-cell anergy (41,44).
ABTs have been shown to be associated with various alterations in in vitro
measurements of immunological responsiveness in recipients. The most com-
monly observed functional modification is a lowering in the helper: suppressor
(CD4:CD8) lymphocyte ratio. Such an abnormality has also been shown to occur
in patients with hemophilia A after treatment with factor VIII concentrates (45).
Recently, two prospective clinical trials have reported that the use of very
high-purity factor VIII concentrates retards the decline in CD4 counts over time
compared with intermediate purity factor VIII concentrates (4648). Moreover,
CD4 cell counts were shown to remain relatively stable in HIV-infected hemo-
philiacs treated exclusively with recombinant factor VIII for 3.5 years (49).
However, the use of high-purity factor VIII products was shown to neither retard
the development of AIDS nor decrease the risk of death in HIV-infected patients
with hemophilia (50). In addition to quantitative alterations of Tcell subsets,
functional abnormalities in lymphocytes have also been described in patients with
hemophilia. These include decreased proliferative responses to mitogens, de-
creased natural killer (NK) cell activity, diminished cell-mediated immunity,
hypergammaglobulinemia due to polyclonal B cell activation, T-cell activation,
and inadequate monocyte function (45,51). It has been suggested that some of
these functional immunological alterations could be prevented with the use of
very high-purity factor VIII concentrates (52).
Other functional immunological alterations associated with ABT include
suppression of lymphocyte blastogenesis, decreased antigen presentation, and
reduction in delayed-type hypersensitivity (53).
VI. MECHANISMS OF ALLOGENEIC

TRANSFUSIONASSOCIATED IMMUNOMODULATION
Although the mechanisms of the ABT-associated immunomodulatory effect re-
main to be elucidated, it is generally accepted that this biological phenomenon is
mediated by the allogeneic leukocytes present in transfused cellular blood prod-
ucts (1). The Class ll HLA antigens present in transfused allogeneic leukocytes
appear to elicit an immune response by recipients T cells, but HLA compati-
bility between blood donor and recipient may result in the persistence of circu-
lating donor mononuclear cells in the recipient (54,55). In this context, it has
been demonstrated, using the polymerase chain reaction technique, that trans-
fused male leukocytes can persist for 16 days in female patients receiving
multiple ABTs (56). This prolonged survival of small numbers of transfused
allogeneic leukocytes within the recipients circulation, a phenomenon known
as mononuclear microchimerism, can downregulate the immune response of
the recipient, inducing tolerance for donor alloantigens and predisposing the
recipient to the development of a chronic transfusion-associated graft-versus-host
disease (54,55).
In vitro investigations have suggested that leukocytes lose their immuno-
genicity during blood storage. Consequently, it has been hypothesized that
the transfusion of allogeneic blood, stored for prolonged periods of time, could
result in transfusion-associated immunomodulation due to recipient T-cell an-
ergy (44).
The hypothesis that the immunomodulatory effect of ABT is related to the
presence of allogeneic leukocytes has also been supported by research involving
experimental animals. Data from such studies indicate that ABTs accelerate tumor
growth and enhance metastatic nodule formation in both inbred and outbred
experimental animals (57,58). In such studies, it has been shown that allogeneic-
ally transfused experimental animals (mice and rabbits) inoculated with synge-
neic tumor cells develop significantly higher numbers of pulmonary nodules than
animals given syngeneic blood (57,58). These studies show that animals receiving
allogeneic buffy-coat leukocytes develop significantly higher numbers of pulmo-
nary nodules than animals given either plasma or prestorage leukocyte-reduced
whole blood (58). Conceivably, the prestorage removal of allogeneic leukocytes
prevented the accumulation of the soluble biological mediators synthesized
and released by the donor allogeneic leukocytes during storage. It is possible that
such biologically active substances are involved in the ABT-associated im-
munomodulation induced. Relevantly, allogeneic leukocytes have been recog-
nized as the blood component responsible for the increased susceptibility to
gut-derived infection in a murine model (59).
Further clues about the mechanism of ABT-associated immunomodulation
have been provided by experimental data showing that the ABT-associated tumor
growthpromoting effect can be adoptively transferred. In these experiments,
naive animals that had received spleen cells from allogeneically transfused
animals (inbred and outbred) developed a greater number of pulmonary meta-
static nodules than was observed in animals that had received spleen cells from
animals, transfused with syngeneic blood. This ABT-associated tumor growth
promoting effect could not be adoptively transferred using spleen cells derived
from animals that had been transfused with prestorage leukocyte-reduced al-
logeneic blood (57). Additionally, it has been observed that the ABT-associated
immunomodulatory effect occurred following the infusion of ABT-conditioned
splenic T cells but was not seen after the infusion of ABT-conditioned splenic B
cells, even though the presence of the B cells enhanced the extent of the tumor
growthpromoting effect of the ABT-conditioned splenic T cells.
The ABT-associated tumor growthpromoting effect also was observed in
adoptive-transfer experiments in which ABT-conditioned spleen cells were in-
stilled intraperitoneally in diffusion chambers that allowed only the release of
soluble substances. The results of these latter experiments suggest that the
ABT-associated tumor growthpromoting effect might be mediated by biologi-
cally active substances released by conditioned T cells. The nature of these
substances still needs to be defined.
The mechanism of the ABT-associated immunomodulatory effect thus is
still unresolved. It is probably due to a combination of mechanisms involving
active immunosuppression, host anergy and clonal deletion. With regard to clonal
deletion, the infusion of foreign MHC antigens could result in the deletion of
recipients T-cell clones, whose T-cell receptor is directed against foreign MHC
antigens. The development of an active suppressor cell network following ABT
thus may result in shift towards a Th2-type immune response. Preliminary clinical
data have recently suggested that ABTs elicit a Th2-type response predominantly
(60). Relevantly, a significant increase in soluble IL-2R and IL-6 concentration
has been detected in colorectal carcinoma patients receiving allogeneic whole
blood transfusions but not in patients transfused with leukocyte-reduced al-
logeneic blood (61).
Several in vitro studies have shown that low molecular weight com-
ponents found in factor VIII products inhibit the proliferative response of
mononuclear cells to phytohemaglgutinin (52). In these studies, high-purity
factor VIII concentrates have been shown to reduce the induced expression of
T-cell activation molecules such as the IL-2 receptor (CD25), the transferrin
receptor (CD71), CD38, CD11a/CD18, and HLA-DR (52). This inhibitory action
of factor VIII concentrates may, at least in part, be due to contamination with
transforming growth factor- (TGF-) (62). TGF- may also be involved in
mechanisms by which veto cells downregulate the immune responsiveness of
the host (63).
In contrast to active suppression, anergy refers to nonreactivity by the host.
As outlined earlier, the absence of costimulatory signals may result in T-cell
unresponsiveness. Of interest, it has been shown that IL-10, a Th2-type cytokine,
prevents in vitro expression of the CD80 family of costimulatory molecules, on
macrophages (64).
VII. SUMMARY
The potential clinical importance of the immunomodulatory effects associated
with ABT is one of the major current concerns in transfusion medicine. Many
questions relating to this phenomenon remain to be elucidated. Data from both
retrospective and prospective studies, including three RCTs, addressing the issue
of immunomodulatory effects of perioperative ABT on cancer prognosis have
yielded contradictory results. The conclusions from six RCTs examining the
prevalence of postoperative septic complications following ABTs also have been
contradictory. In contrast, there is clear-cut evidence that ABT-associated im-
munomodulation might be beneficial for selected groups of patients. Accordingly,
transfusions of allogeneic cellular blood components have been shown to improve
renal allograft survival as well as reducing the recurrence rate in women with
recurrent spontaneous abortions.
The immunomodulatory effect of ABT is generally accepted as being
mediated by the allogeneic leukocytes and/or their products present in transfused
cellular allogeneic blood products. The presence of the allogeneic leukocytes in
these products has thus been associated with diverse adverse biological activities,
including febrile transfusion reactions, graft-versus-host disease, alloimmuniza-
tion to leukocyte antigens causing platelet refractoriness, and immunomodulation.
Data from animal-based research suggest that the ABT-associated im-
munomodulatory effect is immunologically mediated and that this effect is due
to the presence of allogeneic leukocytes in the transfused blood products. Accord-
ingly, the experimental animal data show that prestorage leukocyte reduction
might prevent the ABT-associated growth enhancement of animal tumors. A sim-
ilar protective effect of leukocyte reduction has also been observed in some
human studies of ABT-associated postoperative bacterial infections, but clear-cut
evidence for the clinical benefit of leukocyte reduction is not yet available.
Addendum: Subsequent to completion of this chapter, many publications have

appeared relevant to the topic of immunomodulation and allogeneic blood transfu-
sions. These are summarized in an editorial in Transfusion published in 1999 (65).
In this editorial, the issue of transfusion-associated immunomodulation and the
case for universal white cell reduction is discussed in some detail (65). In addition,
E. C. Vamvakas and M. A. Blajchman have co-edited a book entitled The Im-
munomodulatory Effects of Blood Transfusion, recently published by the AABB
Press (1999). This volume contains 13 chapters, in which most of the published data
relating to the issue of the immunomodulatory effect of allogenic blood transfusions
are reviewed in some detail (66). Thus, for the latest information relating to the topic
of this chapter, the reader is referred to these two recent publications.
REFERENCES
1. Bordin JO, Heddle NM, Blajchman MA. Biologic effects of white blood cells present
in transfused cellular blood products. Blood 1994; 84:17031721.
2. Meryman HT. Transfusion induced alloimmunization and immunosuppression and
the effect of white blood cell depletion. Transfus Med Rev 1989; 3:180193.
3. Francis DMA. Relationship between blood transfusion and tumour behaviour. Br J
Surg 1991; 78:14201428.
4. Chung M, Steinmetz OK, Gordon PH. Perioperative blood transfusion and outcome
after resection for colorectal carcinoma. Br J Surg 1993; 80:427432.
5. Vamvakas EC, Moore SB. Perioperative blood transfusion and colorectal cancer
recurrence: a qualitative statistical overview and meta-analysis. Transfusion 1993;
33:754765.
6. Vignali A, Braga M, Gianotti L, Radaelli G, Gentilini O, Russo A, Di Carlo V. A
single unit of transfused allogeneic blood increases postoperative infections. Vox
Sang 1996; 71:170175.
7. Jensen LS, Andersen AJ, Christiansen PM, et al. Postoperative infection and natural
killer cell function following blood transfusion in patients undergoing elective
colorectal surgery. Br J Surg 1992; 79:513516.
8. Heiss MM, Mempel W, Jauch KW, et al. Beneficial effect of autologous blood
transfusion on infectious complications after colorectal cancer surgery. Lancet 1993;
342:13281333.
9. Vamvakas EC, Moore SB. Blood transfusion and postoperative septic complications.
Transfusion 1994; 34:714727.
10. Opelz G. The role of HLA matching and blood transfusions in the cyclosporine era.
Transplant Proc 1989; 21:609612.
11. Coulam CB, Clark DA, Collins J, et al. Worldwide collaborative observational study
and meta-analysis on allogeneic white blood cell immunotherapy for recurrent
spontaneous abortion. Am J Reprod Immunol 1994; 32:5572.
12. Tartter PI, Heimann TM, Aufses, Jr, AH. Blood transfusion, skin test reactivity, and
lymphocytes in inflammatory bowel disease. Am J Surg 1986; 151:358361.
13. Gantt CL. Red blood cells for cancer patients. Lancet 1981; 2:363.
14. Bordin JO, Blajchman MA. Immunosuppressive effects of allogeneic blood transfu-
sions: Implications for the patient with a malignancy. Hematol Oncol Clin North Am
1995; 9:205218.
15. Vamvakas EC. Perioperative blood transfusion and cancer recurrence: meta-analysis
for explanation. Transfusion 1995; 35:760768.
16. Vamvakas EC. Transfusion-associated cancer recurrence and postoperative infection:
meta-analysis of randomized, controlled clinical trials. Transfusion 1996; 36:175186.
17. Busch ORC, Hop WCJ, van Papendrecht MAWH, Marquet RL, Jeekel J. Blood
transfusions and prognosis in colorectal cancer. N Engl J Med 1993; 328:1372 1376.
18. Heiss MM, Jauch KW, Delanoff C, Mempel W, Schildberg FW. Blood trans-
fusion modulated tumor recurrencea randomized study of autologous versus
allogeneic blood transfusion in colorectal cancer surgery. J Clin Oncol 1994;
12:18591867.
19. Houbiers JGA, Brand A, van de Watering LMG, et al. Randomised controlled trial
comparing transfusion of leucocyte-reduced or buffy-coat-reduced blood in surgery
for colorectal cancer. Lancet 1994; 344:573578.
20. van Prooijen HC, Aarts-Riemens MI, van Oostendorf WR, Hene RJ, Gmelig-Meyling
FHJ, de Weger RA. Prevention of donor-specific T-cell unresponsiveness after
buffy-coat-reduced blood transfusion. Br J Haematol 1995; 91:219223.
21. Cerhan JR, Wallace RB, Folsom AR, Potter JD, Munger RG, Prineas RJ. Transfusion
history and cancer risk in older women. Ann Inter Med 1993; 119:815.
22. 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; 70:187194.
23. Blajchman MA. Allogeneic blood transfusions, immunomodulation, and postopera-
tive bacterial infection: Do we have the answers yet? Transfusion 1997; 37:121125.
24. van de Watering LM, Houbiers JG, Hermans J, et al. White blood cell depletion
reduces postoperative mortality in patients undergoing cardiac surgery (abstract
#SY3DI-06). Proceedings of the 24th Congress of the International Society of Blood
Transfusion, Mukuhari, Japan, 1996, p. 37.
25. Jensen LS, Kissmeyer-Nielsen P, Wolff B, Qvist N. Randomised comparison of white
blood cell-reduced versus buffy-coat-poor transfusion and complications after col-
orectal surgery. Lancet 1996; 348:841845.
26. Houbiers JG, van de Velde CJ, van de Watering LM, Hermans J, Schreuder S, Bijnen
AB, Pahplatz P, Schattenkerk ME, Wobbes T, de Vries JE, Klementschitsch P, van de
Maas, AH, Brand A. Transfusion of red blood cells is associated with increased
incidence of bacterial infection after colorectal surgery: a prospective study. Transfu-
sion 1997; 37:126134.
27. Dawes LG, Aprahamian C, Condon RE, Malongi MA. The risk of infection after
colon injury. Surgery 1986; 100:796803.
28. Pinto V, Baldonedo R, Nicolas C, Barez A, Perez A, Aza J. Relationship of transfusion
and infectious complications after gastric carcinoma operations. Transfusion 1991;
31:114118.
29. Vamvakas EC, Blajchman MA. A proposal for an individual patient data-based
meta-analysis of randomized controlled trials of allogeneic transfusion and postoper-
ative bacterial infection. Transfus Med Rev 1997; 11:180194.
30. Opelz G, Sengar DPS, Mickey MR, Terasaki PI. Effect of blood transfusions on
subsequent kidney transplants. Transplant Proc 1973; 5:253259.
31. Opelz G, Terasaki PI. Improvement of kidney-graft survival with increased numbers
of blood transfusions. N Engl J Med 1978; 299:799803.
32. Blajchman MA, Singal DP. The role of red blood cell antigens, histocompatibility
antigens, and blood transfusions on renal allograft survival. Transfus Med Rev 1989;
3:171179.
33. Norman DJ, Wetzsteon P, Barry JM, Fischer S. Blood transfusions are beneficial in
HLAidentical sibling kidney transplants. Transplant Proc 1985; 17:23472348.
34. van Twuyver E, Mooijaart RJD, ten Berge IJM, et al. Pretransplantation blood
transfusion revisited. N Engl J Med 1991; 325:12101213.
35. Ross WB, Yap PL: Blood transfusion and organ transplantation. Blood Rev 1990;
4:252258.
36. Sells RA, Scott MH, Prieto M, Bone JM, Evans GM, Millar R, Hillis AN. Early
rejection following donor-specific transfusion prior to HLA-mismatched living re-
lated renal transplantation. Transplant Proc 1989; 21:11731174.
37. Horimi T, Terasaki PI, Chia D, Sasaki B. Factors influencing the paradoxical effect
of transfusions on kidney transplants. Transplantation 1983; 35:320323.
38. Unander AM. The role of immunization treatment in preventing recurrent abortion.
Transfus Med Rev 1992; 6:116.
39. Tartter PI, Driefuss RM, Malon AM, Heimann TM, Aufses AH. Relationship of
postoperative septic complications and blood transfusions in patients with Crohns
disease. Am J Surg 1988; 155:4348.
40. Peters WR, Fry RD, Fleshman JW, Kodner IJ. Multiple blood transfusions reduce the
recurrence rate of Crohns disease. Dis Col Rect 1989; 32:749753.
41. Mincheff MS, Meryman HT. Costimulatory signals necessary for induction of T cell
proliferation. Transplantation 1990; 49:768772.
42. Guinan EV, Gribben JG, Boussiotis VA, Freeman GJ, Nadler LM. Pivotal role of the
B7:CD28 pathway in tranplantation tolerance and tumor immunity. Blood 1994;
84:32613282.
43. Webb LMC, Feldmann M. Critical role of CD28/B7 coestimulation in the develop-
ment of human Th2 cytokineproducing cells. Blood 1995; 86:34793486.
44. Mincheff MS, Meryman HT, Kapoor V, Alsop P, Wotzel M. Blood transfusion and
immunomodulation: a possible mechanism. Vox Sang 1993; 65:1824.
45. Watson HG, Ludlam CA. Immunological abnormalities in haemophiliacs. Blood Rev
1992; 6:2633.
46. de Biasi R, Rocino A, Miraglia E, Mastrullo L, Quirino AA. The impact of a very
high purity factor VIII concentrate on the immune system of human immuno-
deficiency virus-infected hemophiliacs: a randomized, prospective, two-year compar-
ison with an intermediate purity concentrate. Blood 1991; 78:19191922.
47. Seremetis SV, Aledort LM, Bergman GE, et al. Three-year randomised study of
high-purity or intermediate-purity factor VIII concentrates in symptom-free HIV
seropositive haemophiliacs: effects on immune status. Lancet 1993; 342:700 703.
48. Hilgartner MW, Buckley JD, Operskalski EA, Pike MC, Mosley JW. Purity of factor
VIII concentrates and serial CD4 counts. Lancet 1992; 341:13731374.
49. Mannucci PM, Brettler DB, Aledort LM, et al. Immune status of human im-
munodeficiency virus seropositive and seronegative hemophiliacs infused for 3.5
years with recombinant factor VIII. Blood 1994; 83:19581962.
50. Goedert JJ, Cohen AR, Kessler CM, et al. Risks of immunodeficiency, AIDS, and
death related to purity of factor VIII concentrate. Lancet 1994; 344:791792.
51. Blumberg N, Heal JM. Effects of transfusion on immune function: cancer recurrence
and infection. Arch Pathol Lab Med 1994; 118:371379.
52. Vermont-Desroches C, Rigal D, Blourde C, Bernaud J. Immunosuppressive property
of a very high purity antihaemophilic preparation: a low molecular weight component
inhibits an early step of PHA induce cell activation. Br J Haematol 1992; 80:370377.
53. Blajchman MA, Bordin JO. Mechanisms of transfusion-associated immunosuppres-
sion. Curr Opin Hematol 1994; 1:457461.
54. Dzik WH. Mononuclear microchimerism and the immunomodulatory effect of

55. Nusbacher J. Blood transfusion is mononuclear cell transplantation. Transfusion
1994; 34:10021006.
56. Adams PT, Davenport RD, Reardon DA, Roth MS. Detection of circulating donor
white blood cells in patients receiving multiple transfusions. Blood 1992; 80:551
555.
57. Blajchman MA, Bardossy L, Carmen R, Sastry A, Singal DP. Allogeneic blood
transfusion-induced enhancement of tumor growth: two animal models showing
amelioration by WBC-depletion and passive transfer using spleen cells: Blood 1993;
81:18801882.
58. Bordin JO, Bardossy L, Blajchman MA. Growth enhancement of established tumors
by allogeneic blood transfusion in experimental animals and its amelioration by
WBC-depletion: the importance of the timing of the leukodepletion. Blood 1994;
84:344348.
59. Gianotti L, Pyles T, Alexander JW, Fukushima R, Babcock GF. Identification of the
blood component responsible for increased susceptibility to gut-derived infection.
Transfusion 1993; 33:458465.
60. Kirkley SA, Cowles J, Pellegrini VD, Harris CM, Boyd AD, Blumberg N. Cytokine
secretion after allogeneic or autologous blood transfusion. Lancet 1995; 345:527.
61. Jensen LS, Hokland M, Nielsen HJ. A randomized controlled study of the effect of
bedside white blood cell depletion on the immunosuppressive effect of whole blood
transfusion in patients undergoing elective colorectal surgery. Br J Surg 1996;
83:973977.
62. Wadhwa M, Dilger P, Tubbs J, Mire-Sluis A, Barrowcliffe T, Thorpe R. Identification
of transforming growth factor-b as a contaminant in factor VIII concentrates: a
possible link with immunosuppressive effects in hemophiliacs. Blood 1994; 84:2021
2030.
63. Verbanac KM, Carver FM, Haisch CE, Thomas JM. A role for transforming growth
factor-beta in the veto mechanism in transplant tolerance. Transplantation 1994;
57:893900.
64. Ding L, Linsley PS, Huang LY, Germain RN, Shevach EM. IL-10 inhibits macro-
phage costimulatory activity by selectively inhibiting the up-regulation of B7 expres-
sion. J Immunol 1993; 151:12241234.
65. Blajchman MA. Transfusion-associated immunomodulation and universal white cell
reduction: Are we putting the cart before the horse? Transfusion 1999; 39:665670.
66. Vamvakas EC, Blajchman MA. Immunomodulatory Effects of Blood Transfusion.
Bethesda, MD: AABB Press; 1999.
8
Overview of Infectious Disease
Roger Y. Dodd
Holland Laboratory, American Red Cross, Rockville, Maryland
I. INTRODUCTION
Infection has always been an adverse outcome of blood transfusion. A great deal
of effort has been directed towards reducing the risk of this outcome, and these
efforts have been extraordinarily successful, so that it is true that, in objective
terms, the blood supply is safer than it has ever been (13). Nevertheless, the issue
of blood safety continues to attract a great deal of attention from the media,
politicians, and professionals. As a natural consequence of this attention, there
continues to be interest in further reduction of the frequency of transfusion-
transmitted infection. However, an important component of any program de-
signed to improve safety is an assessment of the current level of risk, along with
the benefits attributable to the proposed intervention. This should form a basis for
logical decision making and appropriate use of resources. In addition, it is
important that responsible and accurate information about risk be freely available
to those who prescribe or receive blood components. At the same time, it must
be recognized that public sentiment will not always support decisions made
purely upon the basis of costs and benefits. It is clear, for example, that there
is little tolerance for failure to take actions that may prevent transmission of
HIV, whereas there is much less concern about the greater risk of transfusion-
associated bacterial sepsis.
Transfusion-transmitted disease occurs because our measures to assure
blood safety are not perfect. At interview, donors may be unaware of their own
risk of infectivity or deny it. Laboratory errors may occur, and even if they do
not, current serological tests cannot identify all infectious individuals, since virus
159
160 Dodd
may circulate prior to the development of detectable levels of antibodies or of

viral antigens.
This chapter will address the means that can be used to estimate the residual
risk of infection from blood components and will provide estimates that are
current at the time of writing. It will also introduce some upcoming measures that
are anticipated to provide further reductions in risk.
II. APPROACHES FOR ASSESSING INFECTION RISK

A number of different approaches have been used to measure or estimate the
residual risk of transfusion-transmitted infection or disease. Each has advantages
and disadvantages and includes sources of error. Consequently, at this time,
assessment of transfusion risk is an inexact process. Ideally, the outcomes of a
variety of measures should be considered in order to arrive at an appropriate
synthesis. A problem with almost all approaches actually stems from the very
success of those measures designed to assure blood safety. That is, the frequency
of adverse outcomes is extremely low, resulting in estimates with large confidence
intervals or a need for extremely large studies or both.
A. Surveillance
Surveillance is an organized approach to collecting and reviewing data about
posttransfusion disease or infection. It is an important tool but cannot necessarily
be used to generate useful estimates of risk. Two aspects of surveillance are
applicable. In the first, the blood provider or some other agency actively solicits
reports of posttransfusion disease from establishments that have used the blood.
This procedure requires a high degree of attention from the medical community
and effective record keeping across institutions. Even so, a reporting system will
only capture cases that are recognized, usually as a result of overt sickness,
although some cases come to light as a result of serological investigation. Thus,
even when successful in identifying disease, reporting is an insensitive measure
of the frequency of posttransfusion infection. Also, it suffers from the disadvan-
tage that there is not necessarily a causal linkage between transfusion and disease;
this is particularly true when, as is the case for many infections, incidence rates
in the community exceed those among blood recipients.
Formal programs of surveillance for posttransfusion disease have been
established at the national level in at least two countries. The hemovigilance
system in France involves a formal mechanism of reporting and assessing adverse
outcomes, as does the English SHOT (Serious Hazards of Transfusion) program.
These programs have been productive but suffer from the fact that there is really
no control over the initial recognition and reporting of posttransfusion disease.
Overview of Infectious Disease 161
In the United States, adverse effects of transfusion (with special focus on deaths)
must be reported to the Food and Drug Administration (FDA).
The second aspect of surveillance relates to follow-up investigation of
locally or nationally reportable disease. For example, in the United States,
viral hepatitis and acquired immunodeficiency syndrome (AIDS) are reportable.
In some states human immunodeficiency virus (HIV) infection is also reportable.
Further studies are performed to assess risk factors associated with these reported
diseases; for AIDS, all cases are investigated, whereas for hepatitis, only a
proportion are evaluated. It is reasonable to suppose that the majority of AIDS
cases are reported, but more intensive local surveillance clearly shows that
hepatitis is underreported (4,5). Again, surveillance mechanisms will only iden-
tify disease, and even in the case of HIV, where every infection will likely result
in disease, the length of the incubation period precludes contemporaneous esti-
mates of infection rates. Nevertheless, surveillance studies do demonstrate tem-
poral variation. They have effectively demonstrated a continuing reduction in the
incidence of hepatitis that is epidemiologically linked to transfusion (68) and
have revealed that, by December 1998, only 39 of the total of 8760 transfusion-
associated AIDS cases in the United States could be linked to transfusions given
after the implementation of anti-HIV testing (9). It should be noted that great care
is necessary in attributing infection to transfusion, and many cases of apparent
transfusion AIDS were not confirmed upon careful follow-up (10). In some cases,
surveillance data may be the only available way to estimate risk, as is the case
for transfusion-associated malaria in the United States.
B. Prospective Studies
Conceptually, the most satisfactory way to define the risk of transfusion-associated
infection is to perform a prospective study on a population of blood recipients,
thus providing a direct measure of the frequency of infection. This necessitates
the identification of a suitable recipient population, collection of pretransfusion
samples, and posttransfusion follow up with sampling over a sufficient time
period for evolution of markers of infection. The study population should be
selected to avoid confounding factors and with the expectation that the recipients
will survive for a reasonable amount of time after transfusion. A good example
of such a study is the series of investigations performed by Nelson and colleagues
on cardiac surgery patients in Baltimore and Houston (1114). Much greater
value is gained by retaining samples of all donations along with sufficient
information to access the donors again in the future. In the context of posttransfu-
sion follow-up studies, the classic examples are the hepatitis studies performed
in the late 1970s, particularly in the United States (1518). Such studies have
proven to be a rich mine of information over many years, although it should be
emphasized that great caution must be taken in extrapolating all of the data to
162 Dodd
current times. However, Alters study (17) continues to the present day. Even
further value may be obtained from prospective studies that include a controlled
evaluation of an intervention. For example, the posttransfusion study reported by
Blajchman and colleagues assessed the impact of surrogate markers [alanine
aminotransferase (ALT) and hepatitis B core antibody (anti-HBc)] in reducing the
incidence of posttransfusion hepatitis (19). Similarly, studies by Bowden and
colleagues compared the frequency of cytomegalovirus (CMV) infection and
disease in bone marrow transplant patients receiving CMV-seronegative or leuko-
reduced blood components (20).
The major advantage of a properly conducted prospective study is that each
posttransfusion infection may be unequivocally identified and that proper atten-
tion to follow-up testing of the donors can clearly establish that the infection was
indeed acquired by transfusion of a given donors blood. Ideally, such investiga-
tion should include serological or genomic characterization of the specific strain
of the isolates from donor and recipient. The size of the recipient population is
known, along with the number of products transfused, so infection rates can be
defined, along with confidence intervals.
The major disadvantages of prospective studies are their complexity and
cost. In fact, as will be shown below, the frequency of residual infections is now
such that enormous studies have to be performed even to identify a single
infection. Nelsons studies, for example, included up to 9294 transfused, and 2238
untransfused patients, involving transfusion of 120,312 components (13). Even if
one or two infections are identified, the confidence intervals for the resulting risk
estimates are large. More useful data on the risk of hepatitis C were, however,
obtained from this study, with the earliest report showing transmission of HCV
from 0.45% of donations before testing was initiated and 0.19 and 0.03% risks
after the respective initiation of surrogate testing and version 1.0 anti-HCV tests
(12). The resources needed to mount comprehensive prospective studies are no
longer available, and in any case it could be argued that the expenditure would
not be justified. Another difficulty is the extent to which a study can be regarded
as representative. As pointed out above, the posttransfusion hepatitis studies of
the 1970s, although still widely quoted, cannot be regarded as representative of
todays situation, as they involved very different donor populations (some of which
were paid) and less comprehensive procedures for donor screening and collection
of medical histories. Nelsons study was performed in two metropolitan areas that
could not be presumed to be representative of the country as a whole. Finally, the
tests involved may not be adequate to provide the best measure; the ALT tests
used in the original post transfusion hepatitis (PTH) studies were neither specific
nor sensitive, and reevaluation of Nelsons study on HCV seroconversion, using
a newer test procedure, essentially doubled the per-unit incidence estimate (12,14).
Another direct measure of the risk of infectivity for HIV was developed by
Busch et al. (21) and Vyas et al. (22) who pooled samples of blood donations and
used viral culture and PCR technology to look directly for evidence of HIV in
blood, tested and issued as seronegative. Ultimately, this study yielded one isolate
from an investigation representing the equivalent of 160,000 donations (21,22).
Clearly, this represented an enormous effort, which was potentially subject to
criticism about the sensitivity or specificity of the methods involved. In addition,
this heroic study generated but a single data point, with its resultant wide
confidence interval.
Finally, in cases where a diagnostic test is available, but is not used for
donor screening, it is possible to perform a seroprevalence study among blood
donors. If there is some knowledge about the frequency of transmission of the
agent from seropositive donors, then it is possible to make a reasonable estimate
of the risk of recipient exposure and infection in the absence of the test. Once a
test is implemented, however, the problem becomes the measurement of efficacy
of the test and other methods must be used. At this time, donor seroprevalence
studies of this type are being performed in order to assess the risk of exposure to
Trypanosoma cruzi, the agent of Chagas disease, in the United States (23,23a).
C. Other Means of Estimating Risk

A variety of techniques may be used to estimate residual risk: all require some
assumptions, which are a possible source of error. Perhaps the simplest approach
is to measure, or estimate, the true sensitivity of the serological screening test(s)
used and to multiply this by the prevalence of the marker in the donor population.
If necessary, the risk can then be adjusted to reflect the anticipated infectivity of
the false-negative samples. In this way Alter has recently estimated that the
residual risk of HBV infection in the United States is about 1:250,000 (24,25).
She based this calculation on HBsAg and anti-HBc seroprevalence rates of 0.03
and 1.0% respectively, along with sensitivities of 99.9 and 99% for the two tests.
She assumed that all HBsAg false negatives would be infectious but estimated
that only 4% of anti-HBc false negatives would be infectious. The critical
assumption for this estimate is the value for test sensitivity, which may be affected
by a number of factors, particularly if window period donations are a major source
of posttransfusion infection. In some cases, sensitivity can be estimated from
other, prospective studies.
In the case of posttransfusion HIV infection, it is generally accepted that
tests are highly sensitive and that laboratory errors are rare. Consequently,
collection of blood during the window period is probably the most important
source of such infections (2629). Thus, the risk of donor infectivity translates to
the probability of collecting a donation during the window period. Assuming that
there is no association between exposure to HIV and desire to give blood, then
the risk is simply stated as: incidence of infection probability of being in
window period. The probability of being in the window period is length of
164 Dodd
window period time between donations. As discussed below, it is possible to

estimate the window period with reasonable accuracy, but the other assumptions
require some caution. It is known that some HIV-positive donors do acknowledge
that they donated blood in order to obtain a test result (30,31), and it seems likely
that such individuals might be more likely to seek testing shortly after risk
behavior. Also, it is not clear that first-time donors are equivalent to habitual
donors. Some give only once, and at the initiation of testing the prevalence rates
for HIV antibodies were higher among first time donors, even though none of the
donors had been previously tested (32). Some estimates of risk incorporate an
adjustment for this difference (28). Ward and colleagues were the first to use this
approach (26), and subsequently Cumming and colleagues developed a more
exhaustive estimate, refining many of Wards original estimates (32). More
recently, Busch and colleagues have used insensitive (detuned) assays for
anti-HIV to identify and quantitate the frequency of seropositive donations
collected within 6 months following seroconversion. In this way, it is possible to
estimate the incidence of HIV infection among first-time donors, which proved
to be 2.4-fold greater than that for repeat donors (32a).
Given that HIV incidence rates and donation frequency can be observed,
the most sensitive component of this type of estimate is the length of the window
period. The concept of using lookback information for this purpose was intro-
duced by Kleinman and Secord (33) and was subsequently developed further by
Petersen and colleagues (27) and formed the basis for current estimates of risk
(28, 29). Briefly, a population of repeat blood donors who had seroconverted was
identified. The disposition of the last donation before seroconversion was deter-
mined and the HIV status of the recipient(s) of that donation was established.
It was found that there was an inverse relationship between the risk of infection
and the time between the seropositive and previous, seronegative, donation.
This relationship was fitted to a mathematical model, which clearly indicated that
the infectious window period was on the order of 45 days. Further, the observed
data fit the model so well that it was clear that there was relatively little variation
in the length of the window period, and additionally this suggested that the
observed infections were not dominated by laboratory errors resulting in false-
negative findings. Had this been the case, a strong time relationship would not be
expected. It should be noted that these studies support the concept that HIV
infection does not result in extraordinarily lengthy preseroconversion periods.
This lookback approach provides a good measure of the infectious window
period, which probably differs from the overall period between exposure to HIV
and the appearance of detectable serological markers.
Petersens original study relates to data collected up to the end of 1990 and
thus defines a window period that reflects the sensitivity of the tests in use at that
time. Anti-HIV tests have increased in sensitivity, as demonstrated by their ability
to identify seroconversion earlier; the test in most common use in the United
States detects seroconversion some 23 days earlier than those in use in 1990 (34).
These data have been used to arrive at an estimate of risk of 1:340,000 per
component in the United States as of the beginning of 1995. This estimate is
reduced to 1:420,000 if an adjustment is made for the impact of other tests (28).
Schreiber et al. estimated the decrease in the window period attributable to the
increased sensitivity of a variety of tests, including those that detect the HIV
genome (29).
The use of window period estimates along with the incidence of new
infection is also applicable to other transfusion-transmitted agents, as exemplified
by the analysis published by the Retrovirus Epidemiology in Donors Study
(REDS) group (29). Lookback studies on other agents have not yet been per-
formed or if performed, have not been analyzed in a way that provides an estimate
of the window period.
D. Infection and Disease

For many agents, infection does not necessarily result in disease. Consequently
estimates of the risk of infection may not provide useful information about the
long-term outcomes of posttransfusion infection or the eventual resource usage.
Therefore, it is important to understand the natural history of posttransfusion
disease in order to establish a complete assessment of risk. In general, this has
not proven simple, and some prospective estimates do not appear to have been
supported by subsequent findings. Although it is clear that essentially all individ-
uals who are infected with HIV will develop AIDS, the outcome of transfusion-
transmitted infection with human T-cell lymphotrophic virus (HTLV), hepatitis B
virus (HBV), or hepatitis C virus (HCV) is less clear. Early prospective studies
on blood recipients with non-A, non-B hepatitis (NANB) suggested that up to
60% had histological evidence of serious or severe liver disease (35,36). How-
ever, long-term follow-up studies have suggested that, despite these pathological
findings, symptomatic disease is uncommon and there does not appear to be
excess mortality after 1820 years of observation (37). It now appears likely,
however, that up to 20% of chronically infected patients may eventually develop
symptomatic and potentially severe liver disease (38).
III. QUANTITATIVE RISK ESTIMATES

A. HIV
Table 1 outlines key estimates for the residual risk of HIV infection from the fully
screened and tested voluntary blood supply in the United States. The methodol-
ogy is noted along (where appropriate) with the number of observations that
generated the estimate. At first glance, the risk estimates appear to be widely
166 Dodd
Table 1 Published Risk of HIV Infection from Fully Screened and Tested Blood in
the United States
Risk/Unit Observations Method Publication year Ref.
1:40,000 NA Estimate 1988 26

1:68,000 3 Lookback 1988 33
1:153,000 NA Estimate 1989 32
1:61,000 1 Viral culture 1991 21
1:60,000 2 Recipient seroconversion 1992 13
1:160,000 1 Viral culture and PCR 1994 22
1:210,000 36 Lookback, window 1994 27
1:420,000 NA Estimate 1995 28
1:493,000 NA Estimate 1996 29
1:676,000a NA Estimate 1996 29
NA = Not applicableestimate based upon window period and population figures.

aEstimate adjusted to include the effect of testing for HIV p24 antigen.
disparate. However, all of the estimates (prior to the implementation of HIV p24
antigen screening) fall within one order of magnitude. Further, there is a clear
trend over time towards lower risk estimates, commensurate with improve-
ments in test sensitivity and reductions in the frequency of test-positive dona-
tions. What is in fact remarkable is the similarity of the estimates, particularly
when those made during the same time interval are compared. In addition, recent
studies show, as might be expected, that there is considerable regional variation
in the residual risk in the United States, implying that national estimates will
vary from those developed for limited regions. One recent national estimate
of 1:420,000 (28) includes an adjustment for components infectious but sero-
negative for HIV, but that would nevertheless be withheld from transfusion
because of other test markers. This estimate is based upon a window period of 25
days, derived from lookback studies, subsequently corrected for the improved
sensitivity of current tests, and an incidence rate of 3.7 infections per 100,000
donations from repeat donors, representing 78% of donations. For first-time
donors, an estimate of 6.6 incident HIV infections per 100,000 was made.
Schreiber et al. (29) came up with a remarkably similar estimate of 1:493,000,
based upon careful studies in five large blood centers. Additionally, they esti-
mated the impact of additional testing, including that for HIV p24 antigen,
implying a current risk of 1:676,000 (29). As pointed out above, the risk of
disease and ultimately death from HIV infection is 100%. Consequently, every
potential infection may also be regarded as a potential death, albeit many years
after the original transfusion.
B. Hepatitis Viruses
One estimate of the risk of posttransfusion HBV infection (1:250,000 per unit) in
the United States is outlined above; it should be noted that this contrasts with a
reporting rate of one acute, clinical case per 30,000 recipients. In a preliminary
report from his study of cardiac surgery patients in Baltimore and Houston,
Nelson et al. suggest that the recipient seroconversion rate for anti-HBc implies
a per unit risk of about 1:2,500 after implementation of surrogate testing (ALT
and anti-HBc), with a further reduction to 1:25,000 after anti-HCV was im-
plemented (39). However, these findings have not been confirmed by additional
testing (particularly among the donors). Also, within the same study, there were
five seroconversions among 2,334 nontransfused patients. This rate did not differ
significantly from the overall seroconversion rate of 39 among the 9,449 trans-
fused patients. There are no contemporary prospective studies that define the
frequency of posttransfusion HBV infection. Surveillance studies now suggest
that only 1% of reported cases of hepatitis B have prior transfusion as a risk
factor (8).
Risk estimates based upon seroconversion rates and a window period of 59
days suggest that one unit in every 63,000 may be infectious for HBV (29). This
estimate is, however, dependent upon a window period that does not necessarily
reflect the true period of infectivity and upon adjustments that reflect the transient
nature of HBs antigenemia. Thus, this estimate should be regarded with some
cautionit seems likely that the risk is lower. Reasonable estimates of the
outcomes of HBV infection are that there will be clinically apparent acute disease
in 35% of cases, of which less than 0.5% will be fulminant, but of these, 60%
may result in death. Chronic disease appears to be very rare for adults, with a
long-term outcome in fewer than 5% of those infected.
The studies of Nelson et al. have been most instructive in the early
development of risk estimates for HCV infection. In their evaluation of transfused
cardiac surgery patients, they observed an incidence of one HCV infection per
3300 transfused components. This study represented donations tested by a first-
generation (c100-3) anti-HCV test and for anti-HBc and ALT elevations (12).
Subsequently, the study was extended and recipients (but not donations) were
tested using a so-called second-generation, multiantigen procedure. On this basis,
the residual risk was one infection per 1672 components (14) Kleinman and
others made an estimate of the potential improvement attributable to second-
generation testing. They based their estimate upon the increased frequency of
detection of HCV antibodies among donors and suggested that the residual risk
was between 1:2000 and 1:6000 (40). In Blajchman et als study, the frequency
of hepatitis C infection after the introduction of the first-generation test was 1.6
and 2.7 cases per thousand recipients of surrogate tested and nontested compo-
nents, respectively (19). A recent paper (41) points out that, as with HBV, care
168 Dodd
must be taken in attributing iatrogenic HCV infection to transfusion. Careful

molecular studies of isolates from 17 apparent posttransfusion HCV cases among
recipients of blood screened by second-generation tests showed that the source
was not transfusion, but was likely patient-to-patient transmission.
Schreiber et al. (29) estimated the current risk of residual posttransfusion
HCV infection as one in 103,000 units if all risk is attributable to window period
infections. In this case it is clear that the incidence of new infections is low, but
the window period of 82 days has been estimated from prospective studies of
blood recipients and may not necessarily reflect the true period of infectivity (29).
Another concern about this estimate is that of immunosilent HCV infection. Alter
has pointed out that there are individuals who are HCV antibody negative but
positive for HCV RNA by PCR (5). Accordingly, she has developed a different
estimate based upon her estimate of the sensitivity of so-called version 2 tests,
along with the seroprevalence rate for anti-HCV among donors. She concluded
that the test sensitivity was 90% and the residual risk was about 1:4000, based
upon a prevalence rate of 0.24% (25). However, newer tests are more sensitive,
and this estimate is no longer appropriate. More interestingly, as data from
genome amplification testing of blood donations are emerging, it is becoming
apparent that most, if not all donors identified as HCV RNA positive but antibody
negative are actually in the early stages of seroconversion. In the United States,
there have been no major prospective posttransfusion hepatitis studies that
generate meaningful risk data since the introduction of multiantigen anti-HCV
tests. Those that have been presented or discussed have not identified any cases
among several hundred patients evaluated.
At the time of writing, blood agencies in Europe and in the United States
are implementing nucleic acidbased testing of donor samples for HIV and/or
HCV RNA. In general, such testing is being achieved using amplification tech-
nology (primarily the reverse transcriptase-polymerase chain reaction [RT-PCR],
and transcription-mediated amplification [TMA]) on small to moderate-sized
(16128) pools of samples. These techniques should provide further information
about the incidence of infection and window period risk. Early experience has
suggested that the frequency of positive findings is either compatible with, or
lower than, predictions made from incidence and window period estimates.
One of the greatest concerns about posttransfusion HCV infection is the
frequency and severity of long-term outcomes. Although there is little disagree-
ment about the frequency of chronic infection, the frequency of long-term
elevations of serum transaminases, or even the frequency of significant patholog-
ical change on biopsy, these figures do not appear to be congruent with the
frequency or severity of clinically apparent disease. Overall, a synthesis of current
data suggests that among those infected with HCV, almost all will develop
chronic infection, as shown by circulating HCV RNA. Perhaps 25% of infected
individuals will have some acute, symptomatic disease, albeit mild. Fewer than
0.25% suffer fatal acute disease. Some 70% of infected individuals will have
chronic manifestations of infection. However, of all those infected, it appears that
up to 15% may eventually be diagnosed with liver failure and perhaps 3.5%
overall will die of liver disease after an incubation period of 20 years (38). There
does not seem to be any significant excess mortality during 1820 years of
follow-up of individuals originally diagnosed with posttransfusion NANB (37).
Follow-up studies do not provide reliable data beyond this period, although it
must be noted that in Japan some individuals have been shown to progress to liver
cancer (42).
There continues to be discussion about additional forms of viral hepatitis
that may be transmitted by transfusion. Although enteric viruses (such as HAV)
may be transmitted very infrequently as a result of collection of blood during an
asymptomatic preacute phase (43,44), there is also some evidence that post-
transfusion hepatitis (or at least, serum transaminase elevation) occurs in the
absence of any markers to known viruses (5,45,46). Most studies suggest that
these events are self-limited and that maximum ALT levels tend to be low.
Of great interest is Blajchman et al.s observation that this form of posttransfusion
hepatitis appears to occur among recipients of autologous transfusions as fre-
quently as it does among allogeneic blood recipients (19). Thus, great caution
should be used in attributing all (or perhaps any) such cases to an infectious agent.
An HCV-like virus, variously termed hepatitis G virus (47) or GB virus
type C (48), has been identified by molecular cloning and has been shown to be
associated with a minority (about 15%) of cases of residual posttransfusion
hepatitis. The agent is clearly transmissible by transfusion, but appears to have
little, if any, clinical relevance (4951).
Similarly, a DNA virus identified among three individuals with hepatitis
and termed TT virus (after one of the patients) has been shown to occur with
relatively high frequency and to be transmissible by transfusion. Again, however,
it has not proven possible to associate this virus with any evidence of liver
disease, and it appears to offer little, if any, risk to blood recipients (51a,51b).
C. Other Agents
The risk of HTLV infection was originally defined in Nelsons studies. In his
cohort, one infection was noted among the recipients of almost 70,000 tested
components (13). However, anti-HTLV-I tests are improving, particularly in their
ability to detect infection with HTLV-II, and as of 1997 two tests have been
licensed with a specific claim for detection of HTLV-II. The REDS group
estimated that window period infections would present a residual risk of about
1:641,000 (29). It is generally accepted that individuals who are seropositive for
HTLV-I have only about a 4% lifetime chance of developing either HTLV-
associated myelopathy or adult T-cell leukemia lymphoma (ATL). Although the
170 Dodd
former disease has been found in association with transfusion, this does not
appear to be the case for ATL. Published window period risk estimates for key
agents are outlined in Table 2.
Malaria is infrequently transmitted by transfusion in the United States: in a
1991 review, Nahlen et al. showed that the average reporting rate for clinically
apparent malaria is approximately one per 34 million units transfused (52);
surveillance data show that these rates have not changed appreciably through the
surveillance report of 1997 (53). The risk may be higher in countries with
populations having closer ties and more frequent travel to malarious areas.
An additional concern is the possibility of reestablishment of malaria in countries
from which it had previously been eliminated.
Infection with CMV is a relatively frequent outcome of transfusion, with
some studies indicating that 12% of blood units may be capable of transmitting
CMV. In almost all cases, however, such infection has little impact upon im-
munologically competent recipients but often has profound or fatal consequences
for immunocompromised patients. The risk of such outcomes is not well defined,
and increasing concentration on the provision of CMV seronegative or leuko-
reduced products has largely eliminated the concern for those patient populations
most at risk (20,54).
Another disease that has attracted some recent attention is Creutzfeldt-
Jakob disease (CJD). This is a prion disease with an extended incubation period.
It occurs with an incidence rate of about one case per million population annually,
and concern arises when a person with a history of blood donation is diagnosed
with this disease. The agent is clearly transmissible, although there is no current
evidence to show that it can be transmitted by transfusion. Indeed, case-control
Table 2 Estimated Risk of Infection by Selected Transfusion-Transmissible Agents,

United States, 1998, and Potential Impact of Genome Amplification Testing (GAT)
Infection risk per unit
Agent Window period (days) Serological tests Serological tests + GAT
HBV 59 1:63,000 1:110,000

HCV 82 1:103,000 1:368,000
HIV 16 1:676,000 1:990,000
HTLV 51 1:641,000 NA
Malaria NA 1:3,000,000 NA
T. cruzi NA <1:48,000a NA
NA = Not applicable.
aSee text.
Source: Adapted from Refs. 29, 52.

and lookback studies have failed to show any evidence of such transmission
(5557). Animal studies suggest that the transmissible agent, when obtained from
brain tissue, may be infectious by parenteral inoculation. Conversely, intracranial
inoculation of blood from infected animals can also transmit disease. Recognition
that CJD had been transmitted by certain lots of human-derived growth hormone
has generated concern, and donors who received this material are now perma-
nently deferred (58), even though blood-to-blood transmission does not appear to
have been noted. This concern has been extended to recipients of dura mater and
to those with a family history of CJD. Despite the absence of any evidence of
transmission of CJD by transfusion, stringent requirements are now in place for
donor deferral and for recall of components of manufactured products collected
from donors subsequently found to be affected by, or judged to be at risk of, CJD.
These requirements are based solely upon the theoretical possibility of transmis-
sion. Additional concern has arisen, at least in the United Kingdom, as a result of
the apparent association between bovine spongiform encephalopathy (BSE) and
new variant CJD (nv CJD) in humans. Because the characteristics of the BSE
agent are not well known, specific measures have been proposed to reduce the
possible risk of transmission.
In the United Kingdom and some other countries, uniform leukoreduction
of components has been introduced, as it is thought that the agent of nvCJD is
more likely to be associated with leukocytes. In the United Kingdom, domes-
tically obtained plasma is no longer used for the manufacture of plasma deriva-
tives. This has led in the United States and Canada to consideration of procedures
to defer donors who have spent enough time in the British Isles to be considered
at some increased risk for nvCJD infection.
Finally, routine testing for markers of syphilis infection continues in most
countries. There appears to be essentially no risk of syphilis transmission by
transfusion at this time, and consequently no estimate is provided.
D. Bacterial Contamination
Since the adoption of closed plastic bag collection systems, bacterial contamina-
tion has not generally been regarded as a serious problem in transfusion medicine.
It is indeed true that bacterial contamination of red cell concentrates and fresh
frozen plasma is infrequent (5962). The major concern with red cells has been
sepsis or endotoxin shock resulting from transfusion of products with accumu-
lated high counts of Yersinia enterocolitica. In United States, such events occur
with a frequency of about one case per 3 million units transfused (63). A similar
rate of contamination with pseudomonads has been reported (64). These events
usually involve red cells that stored for 20 days or more. Bacterial outgrowth is
limited to organisms that can multiply at the 4C storage temperature of red cells.
It seems likely that the source of Y. enterocolitica is enteric infection in the donor,
172 Dodd
with a concomitant low-level bacteremia. Interestingly, a number of cases of

Yersinia sepsis have been noted among autologous blood recipients, providing an
exception to the general rule that you cannot be infected by your own blood (65).
Because autologous units tend to be stored for prolonged periods, the risk may
actually be greater than for allogeneic units.
Platelet concentrates are stored at 2022C and rapid outgrowth of a variety
of bacterial species can occur. There are many estimates of the frequency of
contamination of platelet units, with very wide variation, depending upon the
actual measure used. For example, careful sterility cultures suggest that one to
two platelet concentrates in every thousand may contain bacteria (59). Although
much higher rates have been reported, such studies may not have been adequately
corrected for false-positive cultures. Quantitative evaluation suggests that many
such contaminants may only be present at very low levels, even at platelet outdate
(66). In contrast, prospective studies looking for sepsis among platelet recipients
have indicated that as many as one platelet concentrate in every 12,000 may
contain pathogenic levels of contaminants (67). This appeared to hold for random
donor platelets and for apheresis products; the per patient risk is, of course,
greater for the former, given that five to six packs constitute a therapeutic dose.
It is not yet clear whether these rates are truly representative, since passive and
active surveillance suggests that posttransfusion sepsis is much less frequent
when a large number of institutions are included. Again, as a result of the
dynamics of bacterial growth, the risk of sepsis does increase with the time the
concentrate has been stored.
There are a number of sources of bacteria that may contaminate platelets.
It is likely that a variety of enteric bacteria that contaminate platelet units are
actually derived from the donors circulation; it is also known that bacteria can
enter the bloodstream as a result of dental or other manipulation of the teeth.
Others may be accidental environmental or skin contaminants. It should be noted
that there is no easy solution to this problem, since it is essentially impossible to
render skin aseptic at more than the most superficial level. Recently, an outbreak
of Serratia marcescens sepsis was reported; it apparently resulted from accidental
external contamination of the blood containers during manufacturing (68). One
comprehensive review notes that the overall fatality rate for transfusion-associ-
ated sepsis is about 25% (59). Unlike most other outcomes described in this
chapter, death occurs shortly after transfusion: it is possible that it has gained a
lower degree of public concern because it is seen as but one of the many mishaps
that can occur during hospitalization.
E. Laboratory Errors
As indicated above, laboratory errors are a possible source of risk to blood
recipients, inasmuch as improper performance of a test could lead to a false-
negative finding, but there is no clear measure of the frequency of such errors.
Systematic errors that might result in the failure of a complete test run are unlikely
to occur because of the need to obtain satisfactory results with test calibrators and
(where used) external controls, along with routine quality assurance measures.
Random errors such as sample transposition or omission may occur, although
modern, automated systems in place in blood centers should preclude such
problems. In order to affect a recipient, such an error would have to coincide with
the presence of an infectious sample. On the other hand, Young reported that 18,
27, and 55 HBsAg testing errors were reported to FDA in 1985, 1986, and 1988,
respectively (69). If every one of these reflected failure to identify an EIA-
positive product (which is by no means clear), then the error rate could have been
as high as 1855 per 18,000, assuming 18 million collections of whole blood and
plasma per year and an EIA-reactive rate of 0.1%. This would translate to an
overall error rate of 0.10.3%. In order to assess the risk of transmission of
disease, this error rate would have to be multiplied by the prevalence of true-
positive results. Thus, an error rate of 0.1% might lead to a risk of infection of
about one in 14 million for HIV (with a prevalence of six per hundred thousand);
the figure for HCV, with a prevalence of 0.16% would, however, be about one in
600,000. It is reasonable to assume that increased attention to transfusion safety
and quality assurance in testing over the past years will have resulted in low rates
of laboratory errors. Consequently, laboratory testing errors probably have no
significant contribution to the risk of infection from transfusion. Other widely
publicized errors certainly do occur but most often reflect failure to withhold a
safe, test-negative product because of a prior false-positive test result on the donor
of that unit (70).
IV. EMERGING INFECTIONS

One of the lessons of the AIDS epidemic is that there is always potential for new
diseases that may impact the safety of transfusion. While it is clearly not possible
to predict the outbreak of an entirely new disease, it does imply a need to be alert
to the implications of newly described diseases. More important, not all emerging
infections are novel; in most cases an old disease appears in a new environment.
At least three examples are relevant. First, immunocompromised patients have
become a population at singular risk of adverse outcomes from otherwise benign
infections. Thus, as immunocompromising or immunoablative therapies increase,
so does the risk of disease, even if the risk of infection does not change. In this
context, consider the impact of CMV or B19 infection on bone marrow transplant
recipients (20,71). Second, population movements are now more common than
they have ever been, and infections move with their hosts. The risk of transfu-
sion-transmitted Chagas disease is now apparent in the United States and
174 Dodd
elsewhere solely as a result of emigration from countries in which human

infection with T. cruzi is endemic (72). Finally, changes in ecology may have a
profound and unexpected impact. Changes in the relationship between humans
and the countryside have resulted in major increases in the population and range
of hosts and vectors of the agents of Lyme disease, babesiosis, ehrlichiosis, and
other tickborne infections in the United States (73,74).
Of necessity, it is difficult to provide risk estimates for emerging diseases.
However, if the potential threat is significant, it may be appropriate to mount
studies to define such a threat as a means to assist in rational decision making.
Such a process was used for HTLV in the United States (75), and at the time of
writing studies to define the risk of T. cruzi infection are under way (23).
Estimates of the number of potentially infected individuals have been published:
there may be at least 100,000 T. cruziinfected legal immigrants in the United
States (72). Even if only 1% were to donate, this would generate a tangible donor
prevalence rate of about 1:12,000a figure that has, at least in part, been
supported by seroprevalence studies reported to date, in which one in 70009000
donations are seropositive in some parts of the United State (23). Even so, there
has been no evidence of infection from lookback studies on these seropositive
donors. To date, a worst-case estimate of infectivity of 1:48,000 could be derived
from the seroprevalence rates of around 1:8,000 in Miami and Los Angeles and
an infectivity of one out of six, which would represent the upper 95% confidence
interval of zero infections among 18 lookbacks. However, other information
suggests that national seroprevalence rates would be less than 1:33,000 and that
infectivity would also be lower. Some estimates for the frequency of infection
with Babesia microti have also been made for areas of known endemicity
indeed, one study reported one transmission among 600 units in Connecticut.
It appears that this agent is becoming more widespread and additional species of
piroplasms are being recognized in unexpected localities (74,76,77).
V. SURVIVAL AFTER TRANSFUSION

One other component of the overall estimation of risk of infectious disease relates
to the likelihood that the blood recipient will actually survive long enough to be
affected by the disease resulting from the infection. While this has no effect upon
the definition of the risk of infection for any patient, it does affect the estimation
of the societal costs of posttransfusion infection. For many years it has been
recognized that individuals who are transfused are quite likely to die of their
underlying disease. Additionally, the majority of transfusions are given to those
aged 65 years or more. Early studies in which attempts were made to identify
recipients of blood from donors who were potentially infected with HIV showed
that any given component had a 50% chance of having been transfused to an
individual who died within one year of the transfusion (26). This does not
necessarily mean that 50% of patients die, since those who did die may have
received more blood components. As discussed above, the overall mortality
among patients originally entered into posttransfusion hepatitis studies was
around 50% after 18 years of follow-up. However, a recent publication from
Vamvakas and Taswell suggests that posttransfusion mortality was not this high
in a community hospital system. However, posttransfusion mortality did vary
significantly according to age at transfusion (78). In summary, then, economic
estimates that define the overall costs of transfusion-associated disease should
recognize that up to 50% of the chronic disease outcomes will not contribute,
since the patients will have died before disease can be manifested.
VI. POOLED PLASMA PRODUCTS

The issue of risk from pooled plasma products is complex and will merely be
outlined here. Therapeutic products are derived from plasma by industrial-scale
fractionation procedures requiring the pooling of large numbers (perhaps tens of
thousands) of individual units. As a consequence, even a modest risk of residual
infectivity per unit may translate to a high risk of contamination of the pool itself,
albeit with significant dilution. Some fractionated products, such as albumin,
which incorporate a pasteurization step, have always been regarded as safe.
In general, immune globulins have also been regarded as safe, although the
mechanism underlying such safety has been unclear. Recently, some intravenous
immunoglobulin preparations have been shown to transmit HCV, perhaps as a
result of gentler fractionation procedures, in association with the absence of
anti-HCV, resulting from the implementation of sensitive screening tests (79).
Viral inactivation procedures for immunoglobulin preparations are now being
implemented. Of much more concern, labile clotting factor concentrates have, in
the past, offered an almost uniform risk of infection for HBV, HCV, and HIV.
Fortunately, procedures for viral inactivation of labile plasma products were
under development around the time that transfusion-associated AIDS was recog-
nized. Unfortunately, these procedures came too late to prevent HIV infection
among a majority of users of these products. Progressive improvements in these
inactivation procedures involving the use of heat and of solvent-detergent meth-
ods have essentially eliminated the risk of infection with HIV, HCV, and HBV,
the last of which has also been affected by the use of HBV vaccine.
One concern about the solvent-detergent approach to viral inactivation is
that it is ineffective against nonenveloped viruses. Although transfusion-associated
infection with nonenveloped viruses is unusual, this concern was illustrated by
outbreaks of HAV infection among recipients of solvent detergent-treated anti-
hemophilic factor concentrates in Europe (80). The source of the viral contami-
176 Dodd
nation remains unclear, but there have been no further reports of any outbreaks.
The B19 parvovirus is also nonenveloped and has been readily transmitted by
concentrates that had been subject to early, and less stringent, heat-treatment
regimens (81). It is unclear whether these transmissions led to any clinically
apparent disease. Currently, in some countries, fresh frozen plasma is also being
subjected to inactivation, either by methylene blue photoinactivation or solvent-
detergent treatment (8284). A possible disadvantage of the solvent-detergent
procedure is that it must be performed on a pooled product, consequently, risk
from enveloped viruses will be reduced or eliminated but at the cost of some
increase in the risk of exposure to HAV or B19. A solvent-detergenttreated
frozen plasma product has also been licensed and is in use in the United States.
Phase 4 (postlicensure) clinical studies showed that this product could indeed
transmit B19 parvovirus, as some lots were contaminated with a relatively high
titer of the virus. Preventive measures, including testing for viral DNA in final
products, have been implemented.
VII. OPTIONS FOR RISK REDUCTION

As with any other aspect of life, blood transfusion is not without risk of adverse
effects. A proportion of these adverse effects are due to infectious agents. The risk
of incapacitating disease or death from such infection is very low compared to
other everyday or medical risks. For example, one study showed that adverse
events occurred in 3.7% of all hospitalizations; of these, 6.5% led to permanent
disability and 13.6% to death. In other words, the risk of dying in hospital is about
one in 200: about half of these deaths were attributable to negligent acts and were
thus preventable (85,86). McCullough reviewed some other medical risks: a
1:50,000 annual risk of death from oral contraceptives, for example (2). Never-
theless, there is little tolerance for even rare transfusion mishaps, and there is thus
pressure to continue to try to reduce the risk to some as-yet-to-be-defined
tolerable level. Perhaps one perceptual problem is that the risk is not diffuse, as
is the risk of reacting to a drug; rather, it is focused in a single blood unit. Thus,
it is perhaps natural to believe that the offensive unit should be identifiable.
As pointed out above, none of the multiple layers of blood safety is
impregnable and perfection will not be achievable, but there may be some room
for improvement. Some donors with positive test results for anti-HIV acknowl-
edge, on further questioning, that they were aware of risk factors or behaviors
(30,31), so behavioral research may define ways to improve the screening
process. Indeed, as tests improve and the window periods narrow, it will be more
and more important to focus on recent risk exposures rather than those in the
distant past. At the same time, a greater proportion of seropositive donors are
unable to recognize their risk, so there will be clear limits to the improvement
that can be gained, particularly for infections such as HCV where 40% or more
of the cases may be unaccompanied by identifiable risk (5,87).
Serological testing is unlikely to improve greatly, although it might be
possible to test for additional markers, as was the case for the HIV p24 antigen.
There is widespread anticipation that genome-based testing will offer significant
benefits. Indeed, the benefits of such testing have been projected (see Table 2).
It seems unlikely that any procedure for genome amplification testing of individ-
ual donations will be available in the immediate future but testing of pooled
samples has been implemented on a broad scale in Europe and the United States.
Although this application was implemented to reduce viral load in plasma for
further manufacture, it is clear that testing in small pools of 1624 will have
sufficient sensitivity to achieve significant improvement in the detection of HCV
window period donations and has the potential to detect at least some of the
remaining HIV window period donations. It will probably not have significant
effect upon HBV safety, however.
There may be some prospect for viral inactivation of cellular products, but
such procedures will almost certainly require the addition of chemicals or drugs
to the components; these, of course, will have their own risks, such as genotoxic-
ity. These risks may not outweigh the benefits to be gained. Finally, it will be very
important to consider emerging diseases in order to determine whether interven-
tions are necessary or appropriate. Ultimately, it must to be recognized that the
marginal gains that might be achieved by additional measures may not justify
the resources required to implement them.
REFERENCES
1. Dodd RY. The risk of transfusion-transmitted infection. N Engl J Med 1992;
327:419421.
2. McCullough J. The nations changing blood supply system. JAMA 1993; 269:2239
2245.
3. AuBuchon JP, Birkmeyer JD, Busch MP. Safety of the blood supply in the United
States: opportunities and controversies. Ann Intern Med 1997; 127:904909.
4. Alter MJ, Mares A, Hadler SC, Maynard JE. The effect of underreporting on the
apparent incidence and epidemiology of acute viral hepatitis. Am J Epidemiol 1987;
125:133139.
5. Alter MJ, Margolis HS, Krawczynski K, Judson FN, Mares A, Alexander WJ, Hu PY,
Miller JK, Gerber MA, Sampliner RE, Meeks EL, Beach MJ. The natural history of
community-acquired hepatitis C in the United States. N Engl J Med 1992; 327:1899
1905.
6. Alter MJ, Hadler SC, Margolis HS, Alexander WJ, Hu PY, Judson FN, Mares A,
Miller JK, Moyer LA. The changing epidemiology of hepatitis B in the United States.
Need for alternative vaccination strategies. JAMA 1990; 263:12181222.
178 Dodd
7. CDC. Hepatitis Surveillance Report No. 55. Atlanta: Centers for Disease Control and
Prevention, 1994: 136.
8. CDC. Hepatitis Surveillance Report No. 56. Atlanta: Centers for Disease Control and
Prevention, 1995: 133.
9. CDC. HIV/AIDS surveillance Report Vol. 10, No 2. Atlanta: Centers for Disease
Control and Prevention, 1999: 143.
10. Selik RM, Ward JW, Buehler JW. Trends in transfusion-associated acquired immune
deficiency syndrome in the United States, 1982 through 1991. Transfusion 1993;
33:890893.
11. Cohen ND, Munoz A, Reitz BA, Ness PK, Frazier OH, Yawn DH, Lee H, Blattner
W, Donahue JG, Nelson KE, Polk BF. Transmission of retroviruses by transfusion
of screened blood in patients undergoing cardiac surgery. N Engl J Med 1989;
320:11721176.
12. Donahue JG, Muoz A, Ness PM, Brown DE, Jr, Yawn DH, McAllister HA, Jr, Reitz
BA, Nelson KE. The declining risk of post-transfusion hepatitis C virus infection.
N Engl J Med 1992; 327:369373.
13. Nelson KE, Donahue JG, Muoz A, Cohen ND, Ness PM, Teague A, Stambolis VA,
Yawn DH, Callicott B, McAllister H, Reitz BA, Lee H, Farzadegan H, Hollingsworth
CG. Transmission of retroviruses from seronegative donors by transfusion during
cardiac surgery. A multicenter study of HIV1 and HTLV-I/II infections. Ann Intern
Med 1992; 117:554559.
14. Nelson KE, Donahue JG, Stambolis V. Post-transfusion hepatitis C virus infection.
Reply. N Engl J Med 1992; 327:16011602.
15. Aach RD, Szmuness W, Mosley JW, Hollinger FB, Kahn RA, Stevens CE, Edwards
VM, Werch J. Serum alanine aminotransferase of donors in relation to the risk of
non-A, non-B hepatitis in recipients. The Transfusion-Transmitted Viruses Study.
N Engl J Med 1981; 304:989994.
16. Koziol DE, Holland PV, Alling DW, Melpolder JC, Solomon RE, Purcell RH, Hudson
LM, Shoup FJ, Krakauer H, Alter HJ. Antibody to hepatitis B core antigen as a
paradoxical marker for non-A, non-B hepatitis agents in donated blood. Ann Intern
Med 1986; 104:488495.
17. Alter HJ, Purcell RH, Holland PV, Alling DW, Koziol DE. Donor transaminase and
recipient hepatitis. Impact on blood transfusion services. JAMA 1981; 246:630634.
18. Stevens CE, Aach RD, Hollinger FB, Mosley JW, Szmuness W, Kahn R, Werch J,
Edwards V. Hepatitis B virus antibody in blood donors and the occurrence of non-A,
non-B hepatitis in transfusion recipients. An analysis of the Transmission-Transmitted
Viruses Study. Ann Intern Med 1984; 101:733738.
19. Blajchman MA, Bull SB, Feinman SV. Post-transfusion hepatitis: Impact of non-A,
non-B hepatitis surrogate tests. Lancet 1995; 345:2125.
20. Bowden RA, Slichter SJ, Sayers M, Weisdorf D, Cays M, Schoch G, Banaji M, Haake
R, Welk K, Fisher L, McCullough J, Miller W. A comparison of filtered leukocyte-
reduced and cytomegalovirus (CMV) seronegative blood products for the preven-
tion of transfusion-associated CMV infection after marrow transplant. Blood 1995;
86:35983603.
21. Busch MP, Eble BE, Khayam-Bashi H, Heilbron D, Murphy EL, Kwok S, Sninsky
J, Perkins HA, Vyas GN. Evaluation of screened blood donations for human
immunodeficiency virus type 1 infection by culture and DNA amplification of pooled

cells. N Engl J Med 1991; 325:15.
22. Vyas GN, Rawal BD, Babu G, Busch MP. Diminishing risk of HIV infection from
transfusion of seronegative blood (abstr). Transfusion 1994; 34:63S
23. Leiby DA, Read EJ, Lenes BA, Yund AJ, Stumpf RJ, Kirchhoff LV, Dodd RY.
Seroepidemiology of Trypanosoma cruzi, etiologic agent of Chagas disease, in US
blood donors. J Infect Dis 1997; 176:10471052.
23a. Leiby DA, Fucci MH, Stumpf RJ. Trypanosoma cruzi in a low-to moderate-risk blood
donor population: seroprevalence and possible congenital transmission. Transfusion
1999; 39:310315.
24. CDC. Public Health Service inter-agency guidelines for screening donors of blood,
plasma, organs, tissues, and semen for evidence of hepatitis B and hepatitis C.
MMWR 1991; 40 (RR-4):117.
25. Alter MJ. Residual Risk of transfusion-associated Hepatitis. Bethesda, MD: National
Institutes of Health, 1995: 2327.
26. Ward JW, Holmberg SD, Allen JR, Cohn DL, Critchley SE, Kleinman SH, Lenes BA,
Ravenholt O, Davis JR, Quinn MG, et al. Transmission of human immunodeficiency
virus (HIV) by blood transfusions screened as negative for HIV antibody. N Engl J
Med 1988; 318:473478.
27. Petersen LR, Satten GA, Dodd R, Busch M, Kleinman S, Grindon A, Lenes B, HIV
Seroconversion Study Group. Duration of time from onset of human immunodefi-
ciency virus type 1 infectiousness to development of detectable antibody. Transfusion
1994; 34:283289.
28. Lackritz EM, Satten GA, Aberle-Grasse J, Dodd RY, Raimondi VP, Janssen RS,
Lewis WF, Notari EP, Petersen LR. Estimated risk of transmission of the human
immunodeficiency virus by screened blood in the United States. N Engl J Med 1995;
333:17211725.
29. Schreiber GB, Busch MP, Kleinman SH, Korelitz J. The risk of transfusion-transmitted
viral infections. N Engl J Med 1996; 334:16851690.
30. Doll LS, Petersen LR, White CR, Ward JW, HIV Blood Donor Study Group. Human
immunodeficiency virus type 1-infected blood donors: behavioral characteristics and
reasons for donation. Transfusion 1991; 31:704709.
31. Williams AE, Thomson RA, Schreiber GB, Watanabe K, Bethel J, Lo A, Kleinman
SH, Hollingsworth CG, Nemo GJ. Estimates of infectious disease risk factors in US
blood donors. JAMA 1997; 277:967972.
32. Cumming PD, Wallace EL, Schorr JB, Dodd RY. Exposure of patients to human
immunodeficiency virus through the transfusion of blood components that test
antibody-negative. N Engl J Med 1989; 321:941946.
32a. Janssen RS, Satten GA, Stramer SL, Rawal BD, OBrien TR, Weiblen BJ, Hecht FM,
Jack N, Cleghorn FR, Kahn JO, et al, New testing strategy to detect early HIV-1
infection for use in incidence estimates and for clinical and prevention purposes.
JAMA 1998; 280:4248.
33. Kleinman S, Secord K. Risk of human immunodeficiency virus (HIV) transmission
by anti-HIV negative blood. Estimates using the lookback methodology. Transfusion
1988; 28:499501.
34. Busch MP, Lee LLL, Satten GA, Henrard DR, Farzadegan H, Nelson KE, Read S,
180 Dodd
Dodd RY, Petersen LR. Time course of detection of viral and serologic markers
preceding human immunodeficiency virus type 1 seroconversion: Implications for
screening of blood and tissue donors. Transfusion 1995; 35:9197.
35. Alter HJ. Youll wonder where the yellow went: a 15-year retrospective of post-
transfusion hepatitis. In: Moore SB, ed. Arlington, VA: AABB, 1987:5386.
36. Aledort LM. Consequences of chronic hepatitis C: a review article for the hematol-
ogist. Am J Hematol 1993; 44:2937.
37. Seeff LB, Buskell-Bales Z, Wright EC, Durako SJ, Alter HJ, Iber FL, Hollinger FB,
Gitnick G, Knodell RG, Perrillo RP, Stevens CE, Hollingsworth CG, NHLBI Study
Group. Long-term mortality after transfusion-associated non-A, non-B hepatitis.
N Engl J Med 1992; 327:19061911.
38. Koretz RL, Abbey H, Coleman E, Gitnick G. Non-A, non-B post-transfusion hepati-
tis: Looking back in the second decade. Ann Intern Med 1993; 119:110115.
39. Nelson KE, Ahmed F, Stambolis V, Ness PM, Yawn D, McAllister H. Incident
hepatitis C virus (HCV) and hepatitis B virus (HBV) infections in transfused
cardiac surgery patients: infection rates during different methods of donor screening.
Bethesda, MD: NIH Consensus Development Conference on Infectious Disease
Testing for Blood Transfusions. National Institutes of Health, 1995: 6769.
40. Kleinman S, Alter H, Busch M, Holland P, Tegtmeier G, Nelles M, Lee S, Page E,
Wilber J, Polito A. Increased detection of hepatitis C virus (HCV)-infected blood donors
by a multiple-antigen HCV enzyme immunoassay. Transfusion 1992; 32:805813.
41. Allander T, Gruber A, Naghavi M, Beyene A, Soderstrom T, Bjorkholm M, Grillner
L, Persson MAA. Frequent patient-to-patient transmission of hepatitis C virus in a
haematology ward. Lancet 1995; 345:603607.
42. Kiyosawa K, Sodeyama T, Tanaka E, Gibo Y, Yoshizawa K, Nakano Y, Furuta S,
Akahane Y, Nishioka K, Purcell RH, Alter HJ. Interrelationship of blood transfusion,
non-A, non-B hepatitis and hepatocellular carcinoma: analysis by detection of
antibody to hepatitis C virus. Hepatology 1990; 12:671675.
43. Hollinger FB, Khan NC, Oefinger PE, Yawn DH, Schmulen AC, Dreesman GR,
Melnick JL. Posttransfusion hepatitis type A. JAMA 1983; 250:23132317.
44. Noble RC, Kane MA, Reeves SA, Roeckel I. Posttransfusion hepatitis A in a neonatal
intensive care unit. JAMA 1984; 252:27112715.
45. Alter HJ, Purcell RH, Shih JW, Melpolder JC, Houghton M, Choo Q-L, Kuo G.
Detection of antibody to hepatitis C virus in prospectively followed transfusion
recipients with acute and chronic non-A, non-B hepatitis. N Engl J Med 1989;
321:14941500.
46. Alter HJ. New kit on the block: evaluation of second-generation assays for detection
of antibody to the hepatitis C virus. Hepatology 1992; 15:350353.
47. Linnen J, Wages J, Jr., Zhang-Keck ZY, Fry KE, Krawczynski KZ, Alter H, Koonin
E, Gallagher M, Alter M, Hadziyannis S, Karayiannis P, Fung K, Nakatsuji Y, Shih
JWK, Young L, Piatak M, Jr, Hoover C, Fernandez J, Chen S, Zou JC, Morris T,
Hyams KC, Ismay S, Lifson JD. Molecular cloning and disease association of
hepatitis G virus: a transfusion-transmissible agent. Science 1996; 271:505508.
48. Simons JN, Leary TP, Dawson GJ, Pilot-Matias TJ, Muerhoff AS, Schlauder GG,
Desai SM, Mushahwar IK. Isolation of novel virus-like sequences associated with
human hepatitis. Nature Med 1995; 1:564569.
49. Miyakawa Y, Mayumi M. Hepatitis G virustrue hepatitis virus or an accidental

tourist? N Engl J Med 1997; 336:795796.
50. Alter HJ, Nakatsuji Y, Melpolder J, Wages J, Wesley R, Shih JWK, Kim JP.
The incidence of transfusion-associated hepatitis G virus infection and its relation to
liver disease. N Engl J Med 1997; 336:747754.
51. Alter MJ, Gallagher M, Morris TT, Moyer LA, Meeks EL, Krawczynski K, Kim JP,
Margolis HS. Acute non-A-E hepatitis in the United States and the role of hepatitis
G virus infection. N Engl J Med. 1997; 336:741746.
51a. Nishizawa T, Okamoto H, Konishi K, Yoshizawa H, Miyakawa Y, Mayumi M.
A novel DNA virus (TTV) associated with elevated trasaminase levels in post-
transfusion hepatitis of unknown etiology. Biochem Biophys Res Commun 1997;
241:9297
51b. Simmonds P, Davidson F, Lycett C, Prescott LE, MacDonald DM, Ellender J, Yap
PL, Ludlam CA, Haydon GH, Gillon J, et al. Detection of a novel DNA virus (TTV)
in blood donors and blood products. Lancet 1998; 352:191195.
selection criteria for United States travellers to malarious areas. Transfusion 1991;
31:798804.
53. CDC. Malaria surveillanceUnited States, 1994. MMWR 1997; 46 (suppl):118.
54. Preiksaitis JK. Indications for the use of cytomegalovirus-seronegative blood prod-
ucts. Transf Med Rev 1991; 5:117.
55. Esmonde TFG, Will RG, Slattery JM, Knight R, Harries-Jones R, De Silva R,
Matthews WB. Creutzfeldt-Jakob disease and blood transfusion. Lancet 1993;
341:205207.
56. Klein R, Dumble LJ. Transmission of Creutzfeldt-Jakob disease by blood transfusion.
Lancet 1993; 341:768768.
57. Dodd RY, Sullivan MT. Creutzfeldt-Jakob disease and transfusion safety: tilting at
icebergs? Transfusion 1998; 38:221223.
58. Holland PV. Why a new standard to prevent Creutzfeldt-Jakob disease? Transfusion
1988; 28:293294.
59. Goldman M, Blajchman MA. Blood product-associated bacterial sepsis. Transfus
Med Rev 1991; 5:7383.
60. Wagner SJ, Friedman LI, Dodd RY. Transfusion-associated bacterial sepsis. Clin
Microbiol Rev 1994; 7:290302.
61. Sazama K. Bacteria in blood for transfusion: review. Arch Pathol Lab Med 1994;
118:350365.
62. Klein HG, Dodd RY, Ness PM, Fratantoni JA, Nemo GJ. Current status of microbial
contamination of blood components: summary of a conference. Transfusion 1997;
37:95101.
63. Tipple MA, Bland LA, Murphy JJ, Arduino MJ, Panlilio AL, Farmer JJ, III, Tourault
MA, Macpherson CR, Menitove JE, Grindon AJ, Johnson PS, Strauss RG, Bufill JA,
Ritch PS, Archer JR, Tablan OC, Jarvis WR. Sepsis associated with transfusion of
red cells contaminated with Yersinia enterocolitica. Transfusion 1990; 30:207213.
64. Lifson AR, Stanley M, Pane J, OMalley PM, Wilber JC, Stanley A, Jeffery B,
Rutherford GW, Sohmer PR. Detection of human immunodeficiency virus DNA
182 Dodd
using the polymerase chain reaction in a well-characterized group of homosexual and

bisexual men. J Infect Dis 1990; 161:436439.
65. Richards C, Kolins J, Trindade CD. Autologous transfusion-transmitted Yersinia
enterocolitica. JAMA 1992; 268:15411542.
66. Leiby DA, Kerr KL, Campos JM, Dodd RY. A retrospective analysis of microbial
contaminants in outdated random-donor platelets from multiple sites. Transfusion
1997; 37:259263.
67. Morrow JF, Braine HG, Kickler TS, Ness PM, Dick JD, Fuller AK. Septic reactions
to platelet transfusions. A persistent problem. JAMA 1991; 266:555558.
68. Gong J, Hgman CF, Hambraeus A, Johansson CS, Eriksson L. Transfusion-associated
Serratia marcescens infection: studies of the mechanism of action. Transfusion 1993;
33:802808.
69. Young FE. Efficacy of new tests and the safety of the blood supply. Transfusion 1990;
30:45.
70. Dodd RY, Houlihan K, Lamberson HV. The specificity of enzyme immunoassays for
antibodies to human immunodeficiency virus: impact on record and donor manage-
ment [letter]. Transfusion 1993; 33:693.
71. Cohen BJ, Beard S, Knowles WA, Ellis JS, Joske D, Goldman JM, Hewitt P, Ward
KN. Chronic anemia due to parvovirus B19 infection in a bone marrow transplant
patient after platelet transfusion. Transfusion 1997; 37:947952.
72. Schmuis GA. Trypanosoma cruzi, the etiologic agent of Chagas disease: status in
the blood supply in endemic and nonendemic countries. Transfusion 1991; 31:547
557.
73. Gerber MA, Shapiro ED, Krause PJ, Cable RG, Badon SJ, Ryan RW. The risk of
acquiring Lyme disease or babesiosis from a blood transfusion. J Infect Dis 1994;
170:231234.
74. Herwaldt BL, Kjemtrup AM, Conrad PA, Barnes RC, Wilson M, McCarthy MG,
Sayers MH, Eberhard ML. Transfusion-transmitted babesiosis in Washington state:
first reported case caused by a WA1-type parasite. J Infect Dis 1997; 175:12591262.
75. Williams AE, Fang CT, Slamon DJ, Poiesz BJ, Sandler SG, Darr WF, Shulman G,
McGowan El, Douglas DK, Bowman RJ, Peetoom F, Kleinman SH, Lenes B, Dodd
RY. Seroprevalence and epidemiological correlates of HTLV-1 infection in U.S. blood
donors. Science 1988; 240:643646.
76. Persing DH, Herwaldt BL, Glaser C, Lane RS, Thomford JW, Mathiesen D, Krause
PJ, Phillip DF, Conrad PA. Infection with a Babesia-like organism in northern
California. N Engl J Med 1995; 332:298303.
77. Herwaldt BL, Persing DH, Prcigout EA, Goff WL, Mathiesen DA, Taylor PW,
Eberhard ML, Gorenflot AF. A fatal case of babesiosis in Missouri: identification of
another piroplasm that infects humans. Ann Intern Med 1996; 124:643650.
78. Vamvakas EC, Taswell HF. Mortality after blood transfusion. Transf Med Rev 1994;
8:267280.
79. CDC. Outbreak of hepatitis C associated with intravenous immunoglobulin adminis-
trationUnited States, October 1993;June 1994. MMWR 1994; 43:505509.
80. Mannucci PM, Gdovin S, Gringeri A, Colombo M, Mele A, Schinaia N, Ciavarella
N, Emerson SU, Purcell RH, Italian Collaborative Group. Transmission of hepatitis
A to patients with hemophilia by Factor VIII concentrates treated with organic solvent
and detergent to inactivate viruses. Ann Intern Med 1994; 120:17.
81. Azzi A, Ciappi S, Zakvrzewska K, Morfini M, Mariani G, Mannucci PM. Human
parvovirus B19 infection in hemophiliacs first infused with two high-purity, virally
attenuated Factor VIII concentrates. Am J Hematol 1992; 39:228230.
82. Lambrecht B, Mohr H, Knver-Hopf J, Schmitt H. Photoinactivation of viruses in
human fresh plasma by phenothiazine dyes in combination with visible light. Vox
Sang 1991; 60:207213.
83. Wieding JU, Hellstern P, Khler M. Inactivation of viruses in fresh-frozen plasma.
Ann Hematol 1993; 67:259266.
84. Horowitz B, Bonomo R, Prince AM, Chin SN, Brotman B, Shulman RW. Solvent/
detergent-treated plasma: a virus-inactivated substitute for fresh frozen plasma. Blood
1992; 79:826831.
85. Leape LL, Brennan TA, Laird N, Lawthers AG, Localio AR, Barnes BA, Hebert L,
Newhouse JP, Weiler PC, Hiatt H. The nature of adverse events in hospitalized
patients. Results of the Harvard Medical Practice Study II. N Engl J Med 1991;
324:377384.
86. Brennan TA, Leape LL, Laird NM, Hebert L, Localio AR, Lawthers AG, Newhouse
JP, Weiler PC, Hiatt HH. Incidence of adverse events and negligence in hospitalized
patients. Results of the Harvard Medical Practice Study I [see comments]. N Engl J
Med 1991; 324:370376.
87. Alter MJ, Hadler SC, Judson FN, Mares A, Alexander WJ, Hu PY, Miller JK, Moyer
LA, Fields HA, Bradley DW, Margolis HS. Risk factors for acute non-A, non-B
hepatitis in the United States and association with hepatitis C virus infection. JAMA
1990; 264:22312235.
9
Blood Donor Screening and
Supplemental Testing
Principles, Procedures, and Consequences
Jay E. Valinsky
I. INTRODUCTION
Blood transfusion has become increasingly safe worldwide since the introduction
of screening tests for transfusion-transmitted infectious diseases (TTD) (1) and
an extensive series of other measures to minimize the risk to recipients of blood,
blood components, and derivatives. Two events triggered dramatic improvements
in blood safety in the United States in the 1970s: the conversion to all-volunteer,
unpaid donors of blood for transfusion (2) and the widespread introduction of
tests for hepatitis B surface antigen that could be applied at the level of mass
screening (3). Cases of posttransfusion hepatitis dropped from about 25% of
recipients to about 5% of recipients.
The repertoire of blood screening tests for TTD expanded following the
introduction of tests for HIV-1 in the mid-1980s, screening tests for HIV, HTLV,
HCV, and HIV-1 p24 antigen, and tests of surrogate markers for liver dam-
age/disease (ALT and anti-HBc). Incremental improvements in blood safety
accompanied the initial implementation of each of these tests, and further
improvements were seen with increasingly more sensitive generations of tests.
Nucleic acid amplification testing (NAT) for HCV and HIV, and potentially other
transfusion-transmitted infectious agents, promises to reduce the already low
prevalence of these markers in the donor population and, by extension, the
exposure of transfusion recipients even further.
It should be emphasized that the ongoing improvements in blood safety did
not result solely from the introduction of new screening tests, but also by (a) more
carefully defining donor eligibility; (b) directly questioning donors about risk
185
186 Valinsky
behavior; (c) the establishment of registries of deferred donors; (d) the quarantine
of products until analysis of test results was complete and process controls
permitted labeling and release to inventory; (e) the introduction and enforcement
of current good manufacturing practices (cGMP); and (f) extensive monitoring
and investigation of adverse incidents, errors, and accidents (4).
All blood and plasma collected in the United States is screened for a variety
of infectious disease markers as well as ABO group and Rh type (up to 14
screening tests). A list of current screening tests is provided in Table 1. The U.S.
Food and Drug Administration (FDA) requires many of these screening tests for
the qualification of products and donors. Others, like assays for alanine amino-
transferase (ALT) and anti-hepatitis B core antibody, are still in use as surrogate
tests for nonA-E hepatitis or as product qualification tests for plasma destined
for further manufacture to blood derivatives. Still others (e.g., cytomegalovirus,
sickle cell trait) are used at the discretion of the blood centers as a means of
identifying products that may have specific clinical or therapeutic applications.
Table 1 Current Blood Screening and Supplemental Tests
Screening tests Supplemental tests
HIV-1/HIV-2 EIAa HIV-1/HIV-2 Algorithmb (HIV-1 Western blot,b

or Immunofluorescence Assayb (I FA, HIV-2
EIA,b HIV-2 Western blot)
HIV-1 p24 Antigen EIAa
HTLV-I/II EIAa HIV-1 p24 antigen neutralizationb
Hepatitis B surface Ag EIAa HTLV-I/II Western blot, alternative EIAb or PCR
Hepatitis C EIA (HCV 2.0 or 3.0)a HCV-RIBA slot immunoblot assayb
Anti-hepatitis B core EIAa HBsAg neutralizationb
Alanine aminotransferase (ALT)c Syphilis (FTA-ABS)b
Syphilis (PK-TP or RPR)a
ABO/Rha
NAT-HCV/HIVd
Cytomegalovirus (CMV)c,e
Sickle cellc,e
Antibody screenc,e
Extended antigen typingc,e
HLA matchc,e
aDenotes required test.
bFDA-approved confirmatory test.
cTest not required.
dCurrently being performed under Investigational New Drug applications (IND).
eTest routinely performed by most blood centers on a portion of inventory to identify products for
which there are specific clinical or therapeutic indications.

Blood Donor Screening and Testing 187
The focus of this chapter will be on the tests currently employed in the
United States for screening of volunteer blood doors for TTD and on the
disposition of donors and products as determined by test results. Other chapters
in this volume will directly address the issues of blood safety. General information
about the test algorithms, test kit qualification, controls, sensitivity and specificity,
and quality control will be followed by more specific information about the perfor-
mance and outcomes of each of the individual tests. The discussions in this chapter
are not intended to compare tests from different vendors nor to endorse the use of
specific tests. The tables that follow describe donor and product disposition reflect
requirements described in various FDA guidances and memoranda.
II. GENERAL TECHNIQUES FOR BLOOD SCREENING

A. Test Devices, Test Kits, and Test Performance
The screening of donated blood for markers of TTD is typically performed on
samples of serum or plasma using serological tests for diagnostic antigens,
antibodies, or enzymes. Blood screening tests are generally enzyme-linked im-
munosorbent assays (EIA/ELISA), agglutination tests (e.g., ABO/Rh screening,
syphilis), or chemistry tests (e.g., ALT). The serological tests employed in donor
screening are qualitative or semi-quantitative and are generally used to detect the
presence or absence of a particular analyte.
While donor screening tests may be performed using a sequence of manual
steps on a variety of devices, mass screening of blood donors in the contemporary
setting is routinely conducted on automated or semi-automated devices. Auto-
mated testing platforms have similar properties, namely: (a) positive identifica-
tion of specimens and reagents by bar-code scanning; (b) robotic aliquoting of
specimens, controls, and reagents; (c) temperature-controlled incubators that are
integral parts of the system; (d) spectrophotometric detection systems; and
(e) computerized systems to calculate test results, manage the testing process,
provide batch records, and reinforce current good manufacturing practices
(cGMP). Current testing platforms are designed to provide relatively rapid
throughput of samples and turnaround of test results, typically in 58 hours.
Rapid tests (5) for infectious disease markers have also been developed.
These are typically self-contained devices that use flow-through or other tech-
niques to trap analytes and reagents on a membrane and produce a color reaction
that indicates whether the result is positive or negative. These tests may prove
valuable in regions of the world where automated technologies are prohibited by
cost or local conditions (6), but they may be limited in sensitivity and specificity
and, in the context of mass screening, may be difficult to quality control.
In most cases, blood screening tests are provided in the form of test kits
containing all of the reagents necessary for test performance, including solid
188 Valinsky
phases for performing EIA/ELISA, buffers, test reagents, and controls. Each kit
component is identified with a part number and has a defined expiration date. The
components are linked to a manufacturing master lot number with an independ-
ent expiration date. Components from different kit lots should not be inter-
changed, since test performance is based in part on the set of components linked
to a given master lot. Sample age, storage, and transport conditions are specified
in the manufacturers instruction circulars to ensure optimal test performance.
The EIA/ELISA tests used for detection of TTD markers use antigen or
antibody capture techniques (7). The antigen or antibody capture reagents (e.g.,
purified antibodies, monoclonal antibodies, purified antigens, cell or viral lysates,
or combinations thereof) are bound to a solid phase, typically a plastic support.
Different vendors have deployed a variety of test matrices, including 96-well
plastic microtiter/microwell plates, plastic beads, and microparticles, but the basic
principles are the same.
The EIA/ELISA tests are performed in sequential steps. Samples of plasma
or serum are incubated with matrix-bound capture reagents for a sufficient period,
at an appropriate temperature, to allow optimal binding of the analyte present in
the test sample to occur. Test controls (see below) are incubated simultaneously.
The matrix is then washed to remove unbound serum or plasma. The wash step
is followed by the addition of a conjugate reagent in one or several steps. These
reagents are typically enzyme-linked (e.g., horseradish peroxidase, alkaline phos-
phatase) anti-human Ig antibodies, enzyme-conjugated antigens, or biotin-avidin-
enzyme conjugates. During this incubation, the conjugate binds to the analyte
bound in the first step to the solid phase by the capture reagent. The matrix is
washed again to remove unbound conjugate. In the last step, or detection step,
reporter molecules or substrates that react with the bound enzyme conjugates are
added. Detection occurs as the specifically bound conjugates convert substrates
to chromogenic, fluorogenic, or chemiluminescent products or react with other
reporter molecules that can be detected spectrophotometrically. The signal (e.g.,
absorbance, relative fluorescence) produced in the test is proportional to the
amount of analyte initially present in the test sample. In qualitative tests, the
signal is evaluated relative to a threshold or cut-off value computed from
control values (see below).
B. Positive and Negative Kit Controls, Calculation of

Assay Cut-Off Values, and the Use of External Controls
Control samples are included in, and are an integral part of, assay kits
(manufacturers kit controls). Like other components of the kits, the controls
are linked to the master kit lot number. Both positive and negative controls are
typically analyzed in replicate as part of the test procedures. The results of the
replicate tests are averaged. The averages can then be used to assess kit perfor-
mance by comparing the actual results to ranges described in the manufacturers

package inserts. If the controls do not meet the performance characteristics
defined in the package insert, the test is invalid.
Positive kit controls are typically prepared from known antibody- or
antigen-positive specimens and are designed to react strongly in the test. The
positive controls are generally not used as kit calibrators but, rather, are controls
that show that the test procedure is working and that the reagents used in the test
perform within the limits described in the package insert.
Negative controls are usually prepared from normal human serum or
plasma. In most current assays, negative control values are used as kit calibrators.
The negative control is typically used to calculate the test cut-off (CO) or
threshold value. The cut-off calculation is test specific. Calculation of the cut-off
usually entails computing the mean of the negative control values and then adding
a constant factor to the mean, or alternatively, multiplying the mean by a constant
factor. Thus, the cut-off for a specific test is variable and related to the perfor-
mance of the test in a given test run. The constants are determined by the
manufacturer and specified in the package insert. The cut-off calculation method
is set during development of the test to optimize the differences between reactive
and nonreactive samples, thereby balancing sensitivity and specificity.
If the signal (e.g., absorbance) (S) produced by a given test sample is equal
to or greater than the calculated cut-off value, the specimen is considered to be
reactive in the test. This may also be expressed as the signal-to-cut-off ratio
(S/CO). If the S/CO ratio is 1.0, the specimen is considered reactive. The use of
the S/CO ratio rather than the absolute absorbance value for reporting results is
valuable, since it normalizes the test results, thereby permitting inter- and
intra-assay comparisons.
In addition to the manufacturers kit controls, external or run controls
should also be analyzed with each test run. By definition, external controls are
not part of the test kit or linked to the master lot. Positive external controls are
generally prepared from weakly reactive specimens or from strongly reactive
specimens diluted to react near the cut-off. Since this control is intended to mimic
a weakly reactive test sample, even small changes in test conditions or perfor-
mance (e.g., temperature fluctuations, sampling errors, time variances) might
result in this control giving a false-negative result. Thus, if the external control
fails, it may be an indicator that actual samples tested in that run, with similar
near-cut-off reactivity, might also have given false-negative findings in the test.
Negative external controls are useful to monitor the performance of the kit
negative control. These reagents are prepared from negative human serum or
plasma. Unlike the negative kit controls, they cannot be used to calculate the kit
cut-off. Failure of the external control may also invalidate the test run. If an
external control fails, the test run is suspect, whether or not the manufacturers
kit controls are in range.
190 Valinsky
The use of external controls has been somewhat controversial (8) because:
1. The performance characteristics of some of the external control
reagents in current use are not well defined. There is presently no
requirement for FDA clearance of these reagents, nor is there an abso-
lute requirement that they be manufactured under cGMP conditions.
2. External control values are sometimes used to define test performance
characteristics that exceed the manufacturers specifications. For exam-
ple, quantitative criteria for external control performance (e.g., controls
must perform within certain ranges of S/CO to be valid) have been
applied to semi-quantitative or qualitative tests.
3. There remain questions about how frequently external controls should
be run during the day (e.g., with each set of manufacturers controls,
with a batch of specimens, twice daily).
The Health Care Financing Administration (HCFA) has recently published rules
for the use of positive and negative external controls for in vitro diagnostic tests
(9). In commenting on these rules, FDA notes that the Clinical Laboratory
Improvement Amendments rules for use of these controls should be followed
but cautions that the use of external controls goes beyond the claims of the
manufacturers package inserts, that the manufacturers guidelines must be fol-
lowed, and that external controls must not be used as substitutes for the calibrators
included in the test kits (10).
C. Quality Control for Screening Tests

Quality control (QC) measures must be incorporated into the daily operation of
laboratories performing blood donor screening tests. QC provides a quantitative
or semi-quantitative approach to evaluation of test performance and verifies that
the tests are operating within the limits described in the manufacturers package
inserts. QC as defined herein addresses neither the issue of accuracy of the tests
nor the correct reporting of test results. These issues should be addressed in
periodic audits as part of an overall quality improvement plan.
The performance of the manufacturers positive and negative kit controls is
the first measure of adequacy of the test. If these controls perform within the
ranges described by the vendor and all test parameters are met, the test run is
considered valid, and the results for all samples tested on that run are considered
valid. If the manufacturers controls are out of range, the test is invalid and should
be repeated. This is obviously the case when the calibrator (e.g., negative control)
fails and the cut-off value cannot be calculated as well as when the positive
control fails.
Failure of the external control(s) when manufacturers kit controls are in
range also renders the test run suspect. The results for reactive (positive) speci-
mens must be retained and the samples subjected to the rest of the test algorithm.
The nonreactive (negative) samples should be retested (10). The procedure used
to assess the validity of the external control should be designed cautiously, so as
not to exceed the specifications of the test set by the manufacturer. Several
attempts have been made to establish performance rules for external controls
[e.g., Westgard rules (11) or other standard deviation rules]. However, these
statistical process control methods are not strictly applicable to the semi-
quantitative or qualitative tests used in donor screening. If the limits are too
stringent (e.g., values for the external control must be within 2 standard deviations
of mean), the frequency of failed runs due to random error alone is unreasonably
high (nearly 5%). If the limits are too relaxed (e.g., 3 standard deviations from
the mean), it is likely that a considerable percentage of run failures may go
undetected. Thus, it is more appropriate to apply a pass/fail system to qualify
external control performance for a particular run. Nonetheless, statistical process
control methods (e.g., control charts) should be applied rigorously to track shifts
or trends in laboratory performance using both the external and manufacturers
control values.
Lot acceptance criteria should also be established in each laboratory.
Upon receipt, a new master lot of test reagents should be quarantined pending
qualification for use. Lot acceptance may involve testing the new lot of reagents
against a panel of specimens that challenges the performance of the new lot in
the specific laboratory environment. The current lot should be tested in parallel.
This is an important step, since considerable variation in sensitivity is often
observed among master lots, even though they all meet manufacturing and
FDA specifications. The external controls should be included in the lot accep-
tance panel. This is most important for the positive external control, since it is
weakly reactive and may give false-negative results if the sensitivity of the new
lot is lower than the current lot. For purposes of tracking and trending the
performance of the external controls, it is also recommended that they be
requalified with each new master lot of reagents and that statistical parameters
(e.g., mean, standard deviation) be recalculated. This may be done conveniently,
for example by testing 20 replicates of the controls as part of lot validation.
As part of a quality improvement plan and to complement these quality control
measures, laboratories should actively participate in certified proficiency testing
programs managed by external organizations (e.g., College of American Pathol-
ogists, state health departments).
D. Test Sensitivity, Specificity, Window Period, Efficiency,

and Predictive Values
Two commonly used indices of assay performance are test sensitivity and
specificity. Sensitivity often has two interpretations. In comparing two tests, for
192 Valinsky
example, one test may be considered more sensitive if it is able to detect the
presence of an analyte in a sample when the other does not. On the other
hand, one test may be considered more sensitive if it can detect an analyte at
higher dilution. While the less sensitive of the two tests may detect the pres-
ence of low levels of analyte, it may still not detect all infected individ-
uals, thereby leading to false-negative results. Thus, while both definitions
apply, sensitivity, in the context of donor screening, is the ability of the assay
to identify true positives (TP) (e.g., infected individuals) in the test population,
while minimizing the number of false negatives (FN). The sensitivity of most
commercially available test kits currently exceeds 98%. Because the emphasis
is on the detection of as many positive donors as possible, most existing EIA
tests for donor screening produce a significant number of false-positive results
(see below).
As test sensitivity increases, the ability to identify individuals at earlier
stages of infection also improves. This reduces the time between exposure/infection
and detection [i.e., the window period (12)]. It is important to note that the length
of the window period is more a function of the sensitivity of the test than it is of
the disease. This has been shown clearly in the cases of HIV (13) and HCV
(14,15), in which the window period has decreased markedly with improvements
in test sensitivity in progressive generations of test kits.
The counterpoint of test sensitivity is specificity. Specificity can be defined
as the ability of a test to identify noninfected individuals in a population (true
negatives, TN), thereby avoiding false positives (FP). The specificity of most
commercially available EIA test kits exceeds 99%. It should be remembered that
sensitivity and specificity are essentially reciprocal values.
A convenient means of comparing tests, which takes both sensitivity and
specificity into account, is test efficiency. This is the ability of an assay to
correctly identify positives and negatives in the test population.
The predictive value (positive or negative) of a test is a semi-quantitative
assessment of the value of the test that takes the actual population prevalence into
account. Thus, for a given specificity and sensitivity, the value of tests may differ
depending on the populations tested, e.g., low-risk blood donors versus high-risk
populations in a sexually transmitted disease clinic. In the case of blood donor
testing, the negative predictive value of a test may be the more valuable
parameter, since it estimates the ability of the test to detect false negatives.
Equations for the calculation of the parameters defined in this section are
found in Table 2.
The ability of these serological assays for infectious disease markers
to detect antibodies or antigens may vary regionally, reflecting the genotypic
distribution of the infectious agent (16,17). In some instances it has been
Table 2 Calculation of Sensitivity, Specificity, Efficiency, and Predictive Values
Parameter Calculation
Sensitivity (TP/TP + FN) 100%

Specificity (TN/TN + FP) 100%
Efficiency (TP/TP + FN) 100%
Positive predictive value (PPV) (TP + TN/TP + FP + TN + FN) 100%
Negative predictive value (NPV) TN/(TN + FN) 100%
TP = True positives; TN = true negatives; FP = false positives; FN = false negatives.
shown that mutations in the infectious agent may go undetected in some tests
(18,19).
E. Screening Test Algorithms

The same basic algorithm applies to all serological screening tests. Serum or
plasma samples are subjected to an initial screening test. If the initial result is
nonreactive (NR) or negative (NEG), the donated blood unit is considered
acceptable for the individual test. If the initial screening result is reactive (R) or
positive (POS), the sample is considered initially reactive (IR) and the test is
repeated in duplicate. When repeat tests are performed, the final interpretation is
derived from the analysis of the three test results. If two or three results are
reactive, the sample is considered repeatedly reactive (RR). If only one of the
three tests is reactive, the specimen is considered initially reactive only and
classified as nonreactive (NR).
Exceptions to this general algorithm include: (1) ALT testing, in which a
single determination is made and the activity, expressed in international units
(IU), is reported as normal or elevated; (b) NAT for HCV and HIV, in which
specimens are tested in pools and positive pools are resolved to the individual
samples (see below) and; (c) syphilis testing on some automated platforms in
which an indeterminate (IND) result is possible. IND results are treated opera-
tionally as positive results in this case.
Three consequences arise from a repeatedly reactive screening test result
on the current donation. First, the labeling of the implicated unit of blood, and
therefore its placement into inventory for transfusion, is interdicted. Second, the
donor is placed in a deferral registry that may affect future donations. Third,
supplemental testing is ordered where appropriate and available to confirm the
screening test result. The outcomes for donors and products based on screening
test results are summarized in Table 3.
194
Table 3 Product and Sample DispositionImplications of Screening Test Results
Screening result
Result Suitable for Interdict Sample to Product Product
Analyte Initial Repeat interpretation transfusion shipment confirmation label disposition
HBc NR None NR Yes No N/A ABO/Rh Inventory

R 1 or 2 Ra RR No Yes N/A Biohazard Discard
RBC/PLT;
plasma to
manufacture
HBsAg
HIV-1/2 NR None NR Yes No No ABO/Rh Inventory
HCV
HTLV-I/II R 1 or 2 R RR No Yes No Biohazard Discard
HIV-1 p24 Ag
Syphilis NR or NEG None NR/NEG Yes No No ABO/Rh Inventory
R/POS/INDb R/POS/IND R/POS/IND No Yes Yes Biohazard Discard
ALTc N None N Yes No N/A ABO/Rh Inventory
E or S None E or S No Yes N/A Biohazard Discard
aEach IR sample is tested in duplicate. If either one or both of the repeat tests is reactive, the interpretation is repeatedly reactive (RR).
bIndeterminate syphilis results are treated as positive.
cResults for ALT testing are linked to International Units of enzyme activity.
Valinsky
III. SUPPLEMENTAL AND CONFIRMATORY TESTING

Additional, more specific supplemental or confirmatory tests are used to verify a
reactive result in the screening tests (Table 1). As noted above, screening tests are
not 100% specific. The balance between sensitivity and specificity in the design
of screening tests is critical, since even small changes in sensitivity can have large
consequences for the absolute number of donors deferred and units lost as a
consequence of false-positive results. This is illustrated in Table 4. In the low
prevalence U.S. blood donor population, for example, the ratio of confirmed (i.e.,
true) positives to total number of repeatedly reactive samples detected in the
screening tests ranged from about 0.8 (HCV) to essentially zero (HIV-1 p24
antigen). The sensitivity of supplemental tests must be at least as good as the
screening tests to which they are linked. Generally, supplemental tests use a
different methodology than do the associated screening test (e.g., HIV-1 Western
blot instead of HIV-1/2 EIA).
The results of supplemental testing have no implications for labeling,
shipment of blood products, or placement of the donor in the deferral registry for
the current donation. Rather, these results are utilized for donor notification,
counseling, reentry, in-date product retrieval, and lookback. However, the donor
deferral status may be modified based on subsequent test results. Donor and
product dispositions, dictated by the screening and confirmatory test results, are
summarized in Tables 5 and 6.
A. Supplemental Testing Methods

Supplemental tests in current use include Western blots (WB), microscope-based
immunofluorescence assays (IFA), immunoblots, EIA/ELISA neutralization tech-
niques, and molecular methods, such as PCR.
Table 4 Prevalence of Serological Markers Among Blood Donors
EIA RR Confirmed Ratio

Marker (%) (%) Conf/RR
Hepatitis B surface antigen 0.055 0.025 0.45

Hepatitis C antibody 0.195 0.156 0.80
HIV-1/2 antibody 0.080 0.012 0.15
HIV-1 p24 antigen 0.030 <1/107 0
HTLV-I/II antibody 0.11 0.006 0.55
Anti-hepatitis B core antibody 0.665 N/A N/A
ALT 0.200 N/A N/A
Syphilis 0.240 0.200 0.83
Source: Summary of New York Blood Center screening and confirmatory test results
19981999. N = 700,000; see Ref. 28 For HIV-1 p24 antigen data.
Table 5 Donor DispositionImplications of Test Results
Screening test result Confirmatory result Surveillance Deferral type Reentry available
196
Syphilis RR or FTA Pos No Permanent Yes

Indeterminate FTA-NEG No None N/A
HBsAg RR Positive No Permanent No
Not neutralized Yes 1st occurrencetemporary 8 wk 1st occurrenceYes
2nd occurrencepermanent 2nd occurrenceNo
HBc RR N/A Yes 1st hit 2 hits permanent No
HCV RR Negative Yes Permanent 1st occurrenceYes: 2ndNo
Indeterminate No Permanent No
Positive No Permanent No
ALT Elevated N/A Temporary 1 yr for 1st 1 yr, 2nd occurrence moderate elevation Yesautomatic after 1 yr
moderate elevation or 1st occurrence super elevated
HIV-1/2 RR Negative/HIV-2 NR Yes Permanent 1st occurrenceYes; 2ndNo
Negative/HIV-2 RR No Permanent No
Indeterminate No Permanent No
HIV p24 Indeterminate Yes 1st occurrence temporary 8 wk 2nd occurrenceNo
Antigen RR 2nd occurrence permanent
HTLV-1/2 RR Negative Yes Permanent if 2nd occurrence negative No
HTLV-I/II or indeterminate
Indeterminate Yes Permanent if 2nd occurrence negative No
or indeterminate
Positive (WB or EIAs) No Permanent No
Valinsky
The combination of a HIV-1/2 RR and a HIV-1 p24 antigen RR on any single or combination of donations will prevent reentry.
The combination of a HBsAg RR and HBc RR on any single or combination of donations will result in permanent deferral.
Table 6 Hospital In-Date Product Retrieval and Lookback
Screening test result Confirmatory result Lookback In-date product retrieval
Syphilis RR or indeterminate FTA Positive or negative N/A N/A

HBsAg RR Not neutralized N/A 5 years
Positive 6 months
HBc RR N/A N/A 5 years
HCV RR Negative N/A Earliest available records
Indeterminate HCV RIBA 3.0 N/A
Positive Earliest available records
Blood Donor Screening and Testing
ALT Reactive N/A N/A N/A

HIV-1/2 RR Negative/HIV2 NR N/A 5 years
Negative/HIV 2 RR 5 years
Indeterminate/HIV 2 NR N/A
Indeterminate/HIV 2 RR 5 years
Positive 5 years
HIV p24 Antigen RR Indeterminate N/A 3 months
Positive 3 months
HTLV-1/2 RR WB negative N/A 5 yrcellular products only
WB indeterminate N/A
WB positive 5 yrcellular products only
Positive in 2 EIAs 5 yrcellular products only
197
198 Valinsky
1. Western Blots
The Western blot is one of the most widely used tests for the confirmation of the
presence of antibodies to retroviruses. A Western blot assay is typically prepared
from partially purified viral proteins derived from viral lysates, recombinant
proteins, peptides, or combinations thereof. The viral proteins are separated by
SDS-polyacrylamide gel electrophoresis according to apparent molecular weight.
Following electrophoretic separation, the proteins are arrayed in the gel as
concentrated bands. The proteins are then transferred electrophoretically from
the polyacrylamide to strips of nitrocellulose paper, where the proteins/peptides
are stabilized and stored until use. The nitrocellulose strips, therefore, contain an
image of the proteins as originally displayed in the polyacrylamide gel.
To perform the test, samples of serum or plasma are incubated with the
nitrocellulose test strips and processed in a manner similar to an EIA. Antibodies
in the test sample bind to specific proteins or antigens arrayed on the nitro-
cellulose strip. Binding is revealed through a series of reactions with enzyme-
antibody conjugates and substrates. Colored products produced by the enzymatic
reaction precipitate and bind to the nitrocellulose strips. The identity of the
antibody binding sites is established both by the position of the stained protein
band on the strip (i.e., apparent molecular weight) and by the reactivity estimated
by the staining intensity. The specificity of the test arises from this specific
binding of antibody to recognizable virus-associated antigens, and the sensitivity
of the test arises, in part, from the concentration of the viral antigens in
electrophoretic bands. The test kits include positive and negative control samples
that give predictable patterns of antibody binding and intensity that can be used
to calibrate the position and reactivity of the test samples.
Typically results are reported as positive (POS) if there is a pattern of
antibody binding to diagnostic bands specific for the virus in question or negative
(NEG) if there are no specific binding patterns or no antibody binding at all.
Results may also be indeterminate (IND) if there are binding patterns that do not
meet the criteria for positive. Failure of the control reagents to produce the
expected pattern of binding and/or intensity invalidates the entire run. A discussion
of Western blot tests for particular viral markers can be found in subsequent sections.
2. Immunofluorescence Assays
An alternative to the Western blot for the detection of viral antibodies is the
Immunofluorescence assay (IFA). Unlike the Western blot, IFA does not identify
specific antibodies. A positive result simply indicates whether antibodies to the
virus in question were present in the test sample. Immunofluorescence assays
have been used as confirmatory tests for HIV, HTLV, and syphilis.
In this test, which is basically an indirect immunofluorescence assay, cells
infected with a particular virus are deposited and fixed in wells etched or
circumscribed on a glass microscope slide. The fixed cells are then incubated with
test serum or plasma. After a series of washes to remove unbound test sample,
the cells are incubated with a fluorescently tagged anti-human IgG. Viral antigens
present intracellularly are revealed by fluorescence microscopy. The specificity
of the test is assessed by comparing the immunofluorescence intensity and
patterns in the infected cells with those observed in noninfected cells treated with
the same test sample. Staining intensity and patterns produced by the test samples
can also be compared with those produced by the positive and negative controls.
The fluorescence intensity is usually scored on an arbitrary scale of 0 to 4+.
Additional quality controls include evaluation of patterns of fluorescence (e.g.,
membrane vs. ctyoplasmic, hazy vs. well-demarcated fluorescence).
3. Immunoblot (Slot Immunoblot)

The immunoblot assays are similar to Western blots in that the test is performed
on a strip of nitrocellulose paper and follows the general principles for incubation
and deposition of a colored enzymatic product at the site of binding of a specific
antibody. The difference is in the preparation of the nitrocellulose strip. In this
case, the antigens are purified recombinant, synthetic, or viral peptides and are
applied, rather than electrophoresed, to the nitrocellulose strip. One advantage of
this methodology is that the concentration of antigen applied to any given location
on the strip can be optimized for sensitivity and specificity. The readout of the
strip is essentially cleaner, since there are no extraneous, nonspecific, or nonviral
proteins present. In addition, this format may result in a reduction of indetermi-
nate results relative to the Western blot (2022). In the current immunoblot assays
(e.g., RIBA tests, see below), two concentrations of human immunoglobulin are
also transferred to the strip. These are methodological controls that show whether
or not the anti-human IgG-developing reagents are functional. The test sample
must react positively with these internal strip controls in order for the strip to be
valid. In addition, positive and negative controls must react appropriately for the
entire test run to be valid. A detailed discussion of the application of this format
to confirmatory testing for HCV follows in a later section.
4. EIA Neutralization Tests

Neutralization tests are generally used to confirm the presence of viral antigens
in samples repeated reactive in EIA tests.
The principle of the neutralization test is competition for binding of antigen
in the test sample between antibody free in solution and antibody bound to the
solid phase. The test is usually performed in two parts. One aliquot of the test
specimen is preincubated with a known amount of antibody to the antigen in
question (neutralizing antibody). A second aliquot is incubated with a nonreac-
tive antibody (control antibody). The EIA test is performed following these
200 Valinsky
incubations. If the antigen is present in the neutralized sample, some or all will
bind to the antibody in solution and form an antigen-antibody complex. Antigen
in the complex will be unavailable for binding to the antibody on the solid phase.
The signal in the EIA reaction will be reduced relative to the sample incubated
with the control antibody. The percentage of inhibition or neutralization can then
be determined by comparing the neutralized and nonneutralized values. The
sample is considered confirmed positive if assay-specific threshold values are
achieved (see below). It has recently been shown that neutralization assays may
be prone to artifacts that produce false-positive (23,24) and/or indeterminate
results (25).
B. Quality Control for Supplemental and

Confirmatory Tests
The procedures described above for quality control of screening tests apply
equally to confirmatory or supplemental tests. Positive and negative kit controls
must meet criteria defined in the manufacturers package inserts. Failure of the
controls invalidates the test run. Any positive results stand, and specimens must
be reported as confirmed positive. Negative or indeterminate results should be
repeated. External controls should also be included as in screening tests with the
same consequences in the event of control failures. Lot acceptance panels should
also be used. Lot validation for EIA-based confirmatory tests is managed in the
same way as for screening tests. In the case of blot-type assays, or IFA, lot
validation involves preparation of panels for lot-to-lot comparisons that contain
manufacturers kit controls and external controls only.
IV. NUCLEIC ACID AMPLIFICATION TESTING

In 1999, nucleic acid amplification testing (NAT) ushered in a new era for
screening of donated blood. Unlike the serological blood screening tests currently
in use, NAT is a molecular technology used to detect nucleic acids rather than
antibodies or antigens. Using NAT methods [e.g., PCR or transcription-mediated
amplification (TMA)], nucleic acid sequences can be amplified in vitro up to
107-fold. By combining the process of nucleic acid amplification with the
appropriate detection systems (e.g., liquid hybridization, biotinylated probes,
chemiluminescence), high-sensitivity, high-specificity tests for viral infections
have been created that are amenable to mass screening. Because amplification of
the target sequences can be effected in vitro, NAT ostensibly increases the
sensitivity of tests for viral markers, in contrast to current serological tests for
antibodies or antigens. The latter are limited by the degree to which antibody
production had been amplified in vivo and the amount of viremia in the blood.
The principal objective of the introduction of NAT was to implement assays

that would effectively reduce the window period for certain TTD. Recent data
suggest that this is feasible for HCV and HIV (26) (i.e., reduction from 7080
days to 1030 days for HCV and from 22 days to 11 days for HIV). It is also
inferred that the implementation of NAT will reduce the residual risk for HCV
from 1.100,000 to 1/500,000-1/1 million and HIV from 1/677,000 to 1/1 million
(27). Therefore, the focus has been on the implementation of tests for these two
viruses. Thorough reviews of the implications of NAT with regard to im-
provements in blood safety and the ramifications for blood availability have been
presented (27,28). NAT for HIV and HCV were the initial tests implemented for
donor screening.
In this section, general procedures for NAT will be discussed. The applica-
tion to blood donor screening will be addressed in the section on specific test
performance. Specific guidelines for the design of these tests have been provided
by FDA in a recent guidance (28).
Two of several available NAT technologies have recently been employed
for blood screening on a massive scale. Both depend on the extraction of
RNA from samples of plasma. The first is polymerase chain reaction (PCR) (29).
In the application to blood screening, viral RNA is first isolated from EDTA-
plasma samples. This is accomplished by concentration of viral particles by
high-speed centrifugation of the plasma and lysis of virus particles with a
chaotropic agent followed by alcohol precipitation. The assay then proceeds
through several steps: (a) reverse transcription of the RNA to cDNA; (b) ampli-
fication of the target cDNA by thermocycling in the presence of target spe-
cific complementary primers, DNA polymerase, deoxynucleoside triphosphates
(dNTP), and Mg2+; (c) hybridization of the amplified products (amplicons) to
oligonucleotide probes specific to the targets; and (d) detection of the probe-
bound amplified products. Positive and negative control samples and an internal
method control are analyzed simultaneously. In addition, an external control
with a known number of copies of RNA is analyzed to check the sensitivity of
the method.
An alternative method, transcription-mediated amplification (TMA) is also
currently being used for blood screening. In this assay format, viral genomic RNA
is extracted from plasma samples. Target-specific oligonucleotides are hybridized
to viral sequences, and the hybrids are captured on a magnetic microparticles.
Reverse transcriptase is used to generate cDNA corresponding to the target
sequences, and then RNA polymerase is used to generate multiple RNA
amplicons from this DNA template. The reaction is carried out isothermally at
41.5C. In the next steps, the amplicons are hybridized to acridinium ester-labeled
probes that are detected by chemiluminescence (30,31). Controls similar to those
used in the PCR assays are employed as well.
202 Valinsky
IV. SPECIFIC TESTING ALGORITHMS

The following sections contain detailed descriptions of test algorithms applied to
specific TTD markers. Tables 3, 5, and 6 can be used to aid in tracking the
disposition of both donors and products following positive or reactive test results.
These tables reflect the intent of a variety of FDA memoranda and guidances.
The administration of these policies and procedures for donor deferral, confiden-
tial unit exclusion, in-date product retrieval, and lookback may vary in the details
from center to center.
A. HIV Testing (Antibodies to HIV-1 and HIV-2 and

HIV-1 p24 Antigen)
Both FDA (32) and AABB standards (33) require that all units of blood intended
for use in transfusion be screened for the presence of antibodies to HIV-1 and
HIV-2. Additionally, AABB standards require, and FDA recommends, testing for
the presence of HIV-1 p24 antigen. The EIA antibody test is typically performed
as a so-called combined test in which HIV-1 and HIV-2 antibodies are detected
simultaneously in an antibody capture assay. This combined test was introduced
in the United States in 1992. The test in most widespread use employs recombi-
nant HIV-1 env and gag and HIV-2 env proteins bound to the solid phase. Binding
of anti-HIV antibodies is detected by incubation of the immobilized antigen
anti-body complex with HIV-1 env and gag and HIV-2 env proteins conjugated
to horseradish peroxidase. Color formation following incubation with substrates
is proportional to the amount of HIV-specific antibody bound.
Confirmation of HIV-1/HIV-2 screening test results is accomplished using
a combination of Western blot or IFA (34) and ELISA assays. Because the initial
screening test was a combination test for HIV-1 and HIV-2, the confirmatory
algorithm uses an approach that may lead to the discrimination between these
two analytes.
The most commonly used algorithm employs an FDA-approved HIV-1
Western blot performed on donor samples repeatedly reactive in the HIV-1/HIV-2
combined EIA. The interpretations in the manufacturers package inserts conform
to recommendations made by the Centers for Disease Control (35). Valid test
results for the HIV-1 Western Blot are:
POSITIVEThe sample contains antibodies specific for HIV-1. The
sample is positive for HIV-1 antibodies if any two of the following bands
are present: p24, gp41, and/or gp120/160. Several studies (36,37) indi-
cate that false-positive results are not uncommon. Therefore, counseling
of donors with such patterns requires careful interpretation of the results
as well as an evaluation of risk.
NEGATIVEFor HIV-1, NO bands of any kind may be visible on the test

strips. If this is the case, the interpretation is that the sample does not
contain detectable antibodies to HIV-1 but may contain antibodies to
HIV-2. Supplemental testing using an FDA-approved test for HIV-2 is
recommended.
INDETERMINATEIn this case, any combination of bands not consis-
tent with the positive pattern results in an interpretation of indeterminate.
No conclusive determination about the presence of antibodies to HIV-1
or HIV-2 antigens can be made on this sample. Supplemental testing
using an FDA-approved test for HIV-2 is recommended.
Table 7 summarizes these band patterns and HIV-1 Western blot interpretative
standards currently in use.
Alternatively, an FDA-approved HIV-1 IFA can be used in place of the
Western blot in the initial step of the confirmatory algorithm. The valid results
for this test are (34,38):
Table 7 Interpretation of HIV-1 Western Blots
Gene product
env gaga pola Nonviral bandsa Interpretation
gp41
gp120 p24 p31 (e.g. p70)
gp160
None None None None Negative
Any p24 Any Any or none
Any p24 Any Any or none Positive
Any 2 None Any or none Any or none
Any Other Pattern Any or none Indeterminate
None None None Any Indeterminate
aAdditional reactivities for gag (e.g., p17, p26, and p55) and pol (e.g., p56 and p66) gene products
may also be present in band patterns associated with a fully reactive specimen. However, these are
not diagnostic bands defined in the CDC criteria nor in the manufacturers package inserts. Nonviral
bands (NVB) commonly appear in the p70 region of the Western blot strip but may also appear in
other regions. NVBs do not appear to be associated with HIV-specific antigens. Nonetheless, they are
recorded during the reading of the strips. Their appearance can affect the interpretation of the strip in
only one case, namely that instance in which NVBs appear on a strip devoid of HIV diagnostic bands.
Under current guidelines, this strip must be interpreted as Indeterminate.
204 Valinsky
POSITIVESpecific cytoplasmic fluorescence staining is observed in the

HIV-1infected cells and that there is a significant difference in the
intensity of fluorescent staining and the pattern of staining between
the infected and noninfected cells.
NEGATIVEBoth the uninfected and infected cells treated with test
sample have an appearance similar to the uninfected and infected cells
treated with the negative control serum and the uninfected cells treated
with positive control serum.
INDETERMINATEAn interpretation of indeterminate pertains when:
(a) there is fluorescent staining in both infected and uninfected cells;
(b) one cannot distinguish differences in staining intensity between in-
fected and uninfected cells; or (c) duplicates are discordant. The inter-
pretation of indeterminate cannot lead to any conclusions about the
presence or absence of antibodies to HIV-1.
Regardless of whether WB or IFA was used as the primary supplemental

test, HIV-2 assays are performed on all samples with indeterminate or negative
results. This step is introduced to address the possibility that an HIV-2positive
sample escaped detection in the HIV-1 confirmatory tests. In the case of the
Western blot, there is extensive crossreactivity of antibodies with HIV-1 (p24)
and HIV-2 (p26) gag proteins. It should be noted that in the United States only
three HIV-2positive blood donors have been identified since 1992 (39). An
FDA-licensed EIA for antibodies to HIV-2 is used in this portion of the algorithm.
The valid results are:
REPEATEDLY REACTIVEThe presence of antibodies to HIV-2 anti-

gens is suspected. Additional supplemental tests (e.g., HIV-2 Western
blot) are recommended.
NONREACTIVEAntibodies to HIV-2 are not detectable in the sample.
Samples that are reactive in the HIV-2 EIA may be subjected to an HIV-2
Western blot. Currently these blots are available for research use only. Valid test
results for this assay are:
POSITIVEAntibodies to p26 (gag) and gp34 (env transmembrane) or

gp105 (gp34 trimer) are present in the sample. It is likely, therefore, that
the specimen is positive for HIV-2 antibodies.
NEGATIVEThe sample does not contain detectable antibodies to HIV-2.
INDETERMINATEAny pattern of reactivity that does not produce a
positive result. The results are inconclusive regarding the presence of
HIV-2 antibodies.
The algorithm for final interpretation of the HIV-1/2 testing, based on the
combined results of the Western blot or IFA and EIA tests, is presented in Table 8.
B. HIV-1 p24 Antigen

Screening for HIV-1 p24 antigen, required by AABB (33) and recommended by
FDA (40), was implemented in the United States in March 1996 (41). The purpose
of this test was to reduce the window period through the earlier detection of
donors who were antigenemic but who had not yet produced anti-HIV antibodies.
The test is an antigen capture assay in an EIA format. For the tests in common
use, monoclonal antibodies to HIV-1 p24 antigen are coated onto a solid phase.
Virus particles present in the test sample are lysed in a specific buffer. Binding
of the HIV-1 p24 antigen released by this process is revealed in several steps
following binding of antibody-enzyme conjugates and substrate. This test is a
qualitative test for the presence of HIV-1 p24 antigen. The reactivity of test
samples from specimens that also contain antibodies to HIV-1 p24 antigen may
be significantly reduced as a result of the formation of antigen-antibody com-
plexes. The qualitative tests approved for donor screening do not incorporate an
antigen-antibody dissociation step.
Table 8 Final Interpretation of HIV-1/2 Test Results
HIV-1/2 HIV-1 WB
EIA or IFA HIV-2 EIA HIV-2 WB Final interpretation
NR N/A N/A N/A Negative

RR POS N/A N/A HIV positive
RR IND RR POS HIV-2 positivea
IND RR IND Indeterminate
IND RR NEG Indeterminate
IND NR N/A Indeterminate
RR NEG RR POS HIV-2 positive
NEG RR IND Indeterminate
NEG RR NEG Negative on supplemental testsb
NEG NR N/A Negative on supplemental testsc
aIn some cases itis not possible to differentiate between HIV-1 and HIV-2 using this algorithm because
of extensive antibody cross-reactivity.
bDonor is not eligible for reentry.
cDonor may be eligible for reentry.
206 Valinsky
Samples repeatedly reactive in the screening test are confirmed using a

neutralization test. In this case, one set of test specimens is preincubated with
purified human anti-HIV globulins (neutralizing antibody), and a second set is
incubated with control, nonneutralizing antibody. If the reactivity of the neutral-
ized sample is at least 40% of the control, nonneutralized sample, the test sample
is considered confirmed positive for HIV-1 p24 antigen. Valid test results for this
test are:
POSITIVEthe sample treated with the nonneutralizing antibody is
reactive in the test, and, in the case of the sample treated with neutraliz-
ing antibody, the signal is reduced by greater than or equal to 40%.
This result implies that in the absence of anti-HIV antibodies, or other
signs of HIV infection, that the donor may be in a preseroconversion
phase.
INDETERMINATEAn indeterminate result may be obtained in two
ways: (a) specimens are reactive in the screening test but neutralization
is less than 40%, or (b) the neutralization test is invalid (e.g., the result
of the sample treated with the nonneutralizing antibody is nonreactive).
In both of these cases, the donor status is unclear. Retesting of the
original or fresh specimens is recommended as is follow-up after 8
weeks, to determine whether seroconversion has occurred.
It should be noted that extensive follow-up of donors with positive HIV-1 p24
antigen neutralization results by RT-PCR and by monitoring over time revealed
that the vast majority are false positive. The donors are routinely PCR negative
and do not show evidence of HIV-1 seroconversion. All observations to date
indicate that the prevalence of true positives (HIV-1 p24 antigen-positive, anti-
body-negative) is <1/107 donations (41).
C. Human T-Lymphotropic Virus Types I and II

Screening of U.S. blood donors for human T-lymphotrophic virus I (HTLV-I) was
implemented in 1988. The original screening tests had package insert intended
use claims for HTLV-I only. Nonetheless, HTLV-II antibodies were also detected
in these tests as a consequence of cross-reactivity with HTLV-I antigens. The
sensitivity of these tests for HTLV-II was limited. (It is likely that <50% of
HTLV-II infections were detected.) In 1997, FDA recommended the implementa-
tion of donor screening tests that specifically detected HTLV-II (42). The tests in
current use are combined tests for HTLV-I and II that contain viral env and gag
antigens specific for HTLV-I and II bound to the solid phase. The sensitivity of
these test for HTLV-II is markedly enhanced (43,44).
Until recently, most supplemental testing for HTLV-I/II was performed

using investigational or research use only assays, typically Western blots. Early
on, the difficulty with these Western blots was their relative insensitivity for
HTLV env antibodies and, in some cases, the inability to clearly distinguish
between HTLV-I/II. Radioimmunoprecipitation assays (RIPA) (45) or IFA tests
(46) were also added to the repertoire of HTLV supplemental tests to overcome
some of the sensitivity issues. The difficulty in performing tests like RIPA and
IFA, quality control issues, and the lack of commercial success of some assays
has limited the alternatives. There are currently no FDA-approved supplemental
tests for HTLV. Since the implementation of the requirement for HTLV-II
screening tests, there are no tests available for routine use that distinguish
between HTLV-I and HTLV-II. This raises the complexity of donor notification
and counseling.
One investigational/research use only Western blot is still in use, how-
ever. This Western blot strips contain HTLV-I viral lysate and a recombinant
HTLV-I env protein (rp21e). The inclusion of this recombinant protein enhances
the sensitivity for envelope antibodies. In some cases it is possible to discriminate
between HTLV-I and HTLV-II based on band intensities of rp21e, p19, and p24
in these blots (43). For this test, the presence of nonviral bands on a strip that
does not elicit diagnostic bands does not affect an interpretation of negative.
Common nonviral bands are in the molecular weight ranges of 70, 5155, and 43
kDa. These bands are likely to be related to HLA antigens.
The valid test results for this Western blot are:
POSITIVEThe sample contains, at a minimum, antibodies to p24 (gag)

and gp46 or p21env (env). Some specimens may contain p19 (gag) and
p21env only and no p24. These may be early seroconversion cases that
should be verified by follow-up testing or by PCR.
NEGATIVENo viral bands detected.
INDETERMINATEViral-specific bands are present but do not meet the
criteria for positive. Up to 70% of repeatedly reactive specimens may
elicit indeterminate test results.
Recent data suggest that many donors described as positive for HTLV
using these algorithms may be false positive (47,48). Because of this, and
because of the unavailability of licensed confirmatory tests, an algorithm was
proposed in which a second FDA-licensed EIA test is used to confirm the
presence of HTLV antibodies in a given specimen (49). This test algorithm is
based on results that suggest that many specimens repeatedly reactive in one EIA
may not be reactive in a second licensed test. Thus, only those specimens that
are reactive in both tests are considered positive for HTLV. Although it is
208 Valinsky
not an obligatory part of the algorithm, additional testing using the investiga-
tional or research use only Western blot is suggested for those samples positive
in the two EIAs. This algorithm does not allow discrimination of HTLV-I and
HTLV-II. Additional testing using research use only PCR tests may be valuable
in this regard.
D. Hepatitis B Surface Antigen

The hepatitis B virus (HBV) tests employed in donor screening are not com-
prehensive and cannot necessarily be used to diagnose the status of infection (50).
Screening for HBV consists of tests for hepatitis B surface antigen and antibodies
to hepatitis B core antigen (see below). The donor screening algorithms do not
include tests for antibodies to hepatitis B surface antigen (anti-HBsAg) or for
hepatitis B e-antigen (HBeAg or anti-HBe). The current HBSAg assays are
sensitive to 0.10.2 ng/mL, corresponding to about 107108 virus particles.
Consequently, detection of both early phase acute infections as well as chronic
carriers is possible with the conventional serological assays. The use of HBV
NAT or more sensitive serological tests is currently the subject of intense
discussion (27,28). Current antigen capture EIA/ELISA for HBsAg can be
performed either in qualitative or quantitative modes, but the qualitative tests are
used in blood screening. Monoclonal anti-HBsAg antibodies are typically used as
the capture reagents. Binding of antigens is detected using antibody enzyme
conjugates and a chromogenic substrate.
Confirmation of the presence of hepatitis B surface antigen in serum or
plasma is performed using a neutralization test. In this instance, a specimen is
confirmed positive if the reduction of a specific absorbance is at least 50% upon
addition of neutralizing antibody and the nonneutralized control generates a
signal greater than or equal to the cutoff of the assay. The valid test results are:
POSITIVEThe test is reactive for the sample incubated with the control,
nonneutralizing antibody (S/CO 1) and neutralization is greater than or
equal to 50%. High concentrations of antigen in a sample may lead to a
prozone effect. Samples that produce absorbance values greater than
the cut-off, but which are weakly neutralized (<50%), should be diluted
and retested until a valid result is obtained.
NEGATIVE or NOT NEUTRALIZABLEThe test is nonreactive for
the sample incubated with the control, nonneutralizing antibody (S/CO
< 1). The neutralization can assume any value.
It should be noted that there are a significant number of samples weakly
reactive in the EIA that are weakly neutralized in the confirmatory test but none-
theless are confirmed positive. These anti-HBc nonreactive and PCR-negative

samples are likely to be false positives (50). Despite this finding, donors must be
indefinitely deferred.
E. Antibodies to Hepatitis C Virus

Screening tests for hepatitis C virus (HCV) were introduced as single antigen EIA
tests in 1990. Multiantigen tests were introduced in 1992 (HCV 2.0) and in 1996
(HCV 3.0). The difference in sensitivity between the HCV 3.0 and HCV 2.0 test
is small but significant (51); nonetheless both tests are licensed by FDA and
remain in common use. The HCV EIA tests are antibody capture assays based on
recombinant proteins bound to the solid phase. These proteins represent various
regions of the HCV genome. Peptides representing the core (c22-3), NS3 and NS4
(c200 or c33c/c100-3), and, in some cases, NS5 regions of the HCV genome are
included in the test. Antibody binding is detected using enzyme-conjugated
antibodies and the appropriate substrates.
HCV EIA reactivity is confirmed with a slot immunoblot assay (recombi-
nant immunoblot assay, RIBA). In this assay, HCV recombinant and synthetic
peptides representing several regions of the HCV genome are deposited on
nitrocellulose strips. In the current version of this test (RIBA 3.0), the peptides
included are c100 (p)/5-1-1(p) (NS-4), c33c (NS3), c22 (p) (core), and NS5. In
addition to these HCV-specific peptides, two internal IgG controls and a control
for the presence of antibodies to superoxide dismutase (SOD) are included on the
strip. The latter is included because the recombinant peptides used in this test and
in the EIA are expressed as fusion proteins with SOD. The SOD control is used
to eliminate the possibility that antibody reactivity might be due to anti-SOD
antibodies and not to HCV-specific antibodies. Detection of anti-HCV antibodies
is revealed using an anti-human IgG-enzyme conjugate and a chromogenic
enzyme substrate system (52). Bands are scored for presence of antibody and for
binding intensity (04+). Failure of the high and/or low concentration IgG
controls or the kit positive or negative controls invalidates the test run. Valid test
results for the HCV RIBA 3.0 assay are:
POSITIVEAt least 2 HCV bands having 1+ or greater reactivity are
present.
NEGATIVENo HCV bands having 1+ or greater reactivity are present
or the SOD band having 1+ or greater reactivity is present alone.
INDETERMINATEA single HCV band having a reactivity of 1+ or
greater is present or the SOD band with at least 1+ reactivity is present
in conjunction with any combination of HCV bands of 1+ reactivity or
greater.
210 Valinsky
F. Antibodies to Hepatitis B Core Antigen

Screening for antibodies to hepatitis core antigen (anti-HBc) was implemented in
the mid-1980s as a surrogate marker for non-A, non-B hepatitis and, somewhat
more controversially, as a surrogate marker for HIV. The value of anti-HBc
screening was reduced following the introduction of specific and sensitive tests
for HCV (53). It nonetheless remains a required test. In some quarters, it is still
considered valuable in reducing the risk of HBV infections both in the transplant
setting and in some instances in which HBsAg tests are negative, but the risk of
HBV transmission nonetheless exists (54).
The tests currently in use for donor screening detect both IgM and IgG and
thus have limited diagnostic value in detecting current infections. Approximately
1% of donors tested with the current anti-HBc EIAs are repeatedly reactive. Most
of these donors are negative for HBsAg, and most never develop HBV infections.
In all likelihood these are false-positives. There are currently no confirmatory
tests in routine use for anti-HBc. Thus a significant number of these donors are
deferred, and in-date products must be retrieved (Tables 5,6), perhaps unneces-
sarily. It should be noted that the deferral status of donors is affected by the
combination of HBsAg and anti-HBc test results.
G. Alanine Aminotransferase
Alanine aminotransferase (ALT) an enzyme involved in amino acid metabolism,
is found in highest concentrations in the liver and kidney. The enzyme catalyzes
the conversion of alanine and -aminoglutaric acid to pyruvic acid and glu-
tamic acid.
The appearance of the enzyme in the serum is often taken as an indicator
of tissue damage. ALT testing of blood donors was initiated as a surrogate marker
for non-A, non-B hepatitis in the mid- to late 1980s. Its value has been reduced
significantly following the introduction of specific tests for HCV (55). Neither
AABB nor FDA requires donor screening for ALT. However, ALT screening is a
requirement for the shipment of recovered plasma for further manufacture in the
countries of the European Union (56).
A variety of tests have been used to measure the ALT activity in serum.
These generally take advantage of coupled enzyme reactions utilizing the prod-
ucts of the ALT reaction in subsequent enzymatic reactions, which, in turn, lead
to the formation of chromogenic products that can be detected spectrophotomet-
rically. Both kinetic and endpoint tests have been used. The cut-off for acceptable
ALT activity is defined in terms of either studies on the distribution of ALT
activities in the test population or by using the normal values defined by the
manufacturers package insert (57). The valid results for the ALT test are:
NORMALALT activities, expressed in international units (IU), <60120 IU.

ELEVATEDALT activities are 60120 IU.
SUPERELEVATEDALT activities are 120 IU.
H. Syphilis Testing
Serological tests for syphilis (STS) were the first infectious disease marker tests
applied to blood for transfusion. Syphilis testing was essentially the first test for
transfusion-transmitted diseases. Currently the STS employed for screening blood
donors include manual reagin tests [e.g., rapid plasma reagin tests (RPR)] or
microhemagglutination (MHA) tests antibodies to Treponema pallidum. These
tests appear in large measure to identify biological false positivesindividuals
who retain antibodies for many years following exposure and a much smaller
subset of individuals who may have active or untreated syphilis (58). Transmis-
sion of syphilis by blood transfusion is an extremely rare occurrence. STS is still
required for donor screening.
The RPR test detects the presence of reagin in the serum or plasma of
infected individuals using a cardiolipin antigen bound to carbon particles.
When the reagin is present, an agglutination reaction occurs, which can be scored
visually. Test results are either reactive or nonreactive. Reactive results should be
interpreted cautiously, should be confirmed in a quantitative test and should be
reviewed in the context of health history. Confirmation using a quantitative test
is typically not part of the screening algorithm, but rather a treponemal test,
FTA-ABS (see below), is used.
Alternative tests, either EIA or MHA-TP tests, have been automated for
mass screening and are more appropriately deployed in the context of donor
screening than the reagin tests. The MHA-TP type tests, which are in wide use,
analyze serum or plasma samples for the presence of antibodies to T. pallidum.
The presence of antibody is detected in an agglutination reaction measured by
light scattering or other photometric methods. The results are typically reported
as reactive or nonreactive (59). In some cases, results may be reported as
indeterminate. These are technically treated as reactive, and samples are reflexed
to the confirmatory algorithm.
Reactive findings in the screening tests are confirmed using the fluores-
cent treponemal antibody absorption test (FTA-ABS). In this assay, test serum
is treated with a sorbent derived from a Reiter strain culture of T. pallidum
to absorb nonspecific antibodies. This absorbed sample is then incubated
with T. pallidum fixed to microscope slides. If the sample contains antibody,
it will bind to the fixed cells. Specific antibody binding is detected by fluo-
rescence microscopy following incubation with FITC-conjugated anti-human
212 Valinsky
antibody. Test results are compare to positive and negative control sera. Valid
results are reactive or nonreactive based on staining intensity. Because the
false-positive rate for blood donor samples is significant (6062), additional
testing, using RPR or other assays, may be employed for counseling pur-
poses. The composite of interpretations of syphilis results can be found in
Table 9.
I. Nucleic Acid Amplification Testing

U.S. blood centers embarked on nucleic acid amplification testing (NAT) for
HCV and HIV in the spring of 1999. Testing is currently being performed under
Investigational New Drug (IND) applications. All allogeneic whole blood, di-
rected, and apheresis donors are being tested. The TMA-based multiplex assays
and a PCR-based assay are being evaluated in the INDs. The first objective of the
IND is to assess the operational aspects of NAT, evaluate test performance in a
mass screening setting, and assess the impact of NAT on blood availability.
During this phase of the study (Phase 1), there was no requirement to release
blood for transfusion based on NAT results. In Phase 2 of the study, it is expected
that there will be a requirement for the release of blood based on NAT results.
Most importantly, data will be collected throughout to assess the efficacy of NAT
in reducing the window period.
NAT is currently being conducted in pools of 1624 samples (primary
pool). Testing is being conducted in this manner since, at this writing, the
technology for single unit testing is not available. Sample pools are constructed
by robotic transfer of sample aliquots to tubes in which the extraction of RNA is
carried out. Amplification and detection of the target sequences then follow
Table 9 Interpretation of Syphilis Confirmatory Test Results
FTA-ABS
Screening test result result RPR result Interpretation
Nonreactive N/A N/A Negative

Repeatedly reactive Nonreactive Nonreactive Negative, false-positive screen
Repeatedly reactive Reactive Reactive Positive syphilis serology,
suggests current infection
Repeatedly reactive Nonreactive Reactive Negative, false-positive screen
Repeatedly reactive Reactive Nonreactive Positive syphilis serology,
suggests past infection
standard protocols in manual or semi-automated systems. FDA has defined

requirements for the sensitivity of the tests in sample pool (100 copies/mL in
the pool, at a 95% detection rate) as well as in individual samples (5000
copies/mL, at a 95% detection rate) (28). These levels conform to those proposed
by the European Union (63).
The NAT algorithms are dependent upon the testing format used. In each
case, however, positive primary pools must be resolved to identify the implicated
samples and donors. For the PCR-based assay in current use, primary pools of 24
samples are currently being tested. After the RNA extraction step, amplification
and detection are carried out independently for HCV RNA and HIV RNA. If the
pool is negative for HCV or HIV, the units represented in that pool are considered
negative. If the pool is positive for HCV or HIV or both, four mini-pools
(secondary pools) containing six samples each are created and tested using the
same methodology. Units represented in negative secondary pools are reported as
negative. If a positive secondary pool is detected, the six samples contained
therein are tested individually in the third stage of pool resolution. Negative
individual samples are exonerated, while donors associated with positive samples
are deferred, units are discarded, and the donors invited to participate in follow-up
studies that test for seroconversion. This algorithm applies equally to HCV and
HIV tests.
The TMA-based assays are multiplex tests in which HCV and HIV RNA
are assessed simultaneously. In addition, the current test configuration permits a
more direct resolution of positive primary pools. Primary pools of 16 samples
each are tested. Units represented in a negative pool are considered negative for
both HCV and HIV RNA. Samples in a positive pools are tested individually.
Units associated with negative results on the individual samples are exonerated
for both HCV and HIV RNA. Positive samples are considered NAT positive and
are reflexed to discriminatory TMA assays tests that determine whether the
reactivity was due to HCV, HIV, or both. Donor deferral, unit discard and
follow-up studies are also performed in the INDs employing the TMA-based test.
Both PCR and TMA tests can give rise to inconclusive results. This is the case,
for example, when primary or secondary pools are positive, but the result cannot
be confirmed in tests on the individual samples. This is also true in the case of
positive multiplexed TMA tests in which the discriminatory tests do not reveal a
definitive result.
Preliminary data suggest that the sensitivity and specificity of the PCR and
TMA tests are comparable, but that there may be some differences in test
performance characteristics (FDA Workshop on Implementation of NAT, Decem-
ber 14, 1999). The algorithms for disposition of products and donors based on
NAT results are presented in Table 10.
214 Valinsky
Table 10 Nucleic Acid Amplification Testing for HCV and HIV RNA
1. Possible Results
Suitable for
NAT HCV or HIV Result Interpretation transfusion
PCR Assaya
NEG Negative Yes
POS Positive No
Pos Pool/Indiv. Negb Inconclusive No
Pos Pool/QNSc Inconclusive No
TMA Assayd
NEG Negative Yes
POS Positive No
HIV POS Positive for HIV-1 RNA No
HCV POS Positive for HCV RNA No
NAT POS Positiveno discrimination No
between HIV and HCV
aSpecimens are tested in primary pools of 24 plasma samples. If the primary pool is
negative, all 24 samples are reported negative. If the primary pool is positive, specimens
are retested in 4 pools of 6 samples each. If a positive pool(s) is (are) identified, specimens
are retested as individual samples. The result reported is the final result. Hold implies that
test results during the pool resolution were inconclusive. QNS implies that the pool
resolution could not be completed.
bThis result reflects a positive test on the primary pool, but negative results in testing of
individual samples in the pool.

cThis result occurs when the pool testing algorithm cannot be completed because of
technical or other reasons.

dSpecimens are tested in primary pools of 16 plasma samples. If the primary pool is
negative, all 16 samples are reported negative. If the primary pool is positive, all 16
specimens are retested individually. If one or more specimens are positive, they are
subjected to discriminatory TMA assays to distinguish between HIV and HCV. All negative
samples are released. The result reported is the final result following discriminatory testing.
In those cases where the discriminatory assay fails, the result is reported as positive.
2. Disposition of Blood Products According to Test Results
Product disposition
Result Interdict
interpretation shipment Product label RBC/PLT Plasma
Negative No For transfusion Inventory Inventory

Positive Yes Biohazard Discard Studya
Inconclusive Yes Biohazard Discard Studya
QNS Yes Biohazard Discard Discard
aSpecimens in this category are submitted for further analysis as part of the IND studies.
3. Management of Donor Deferral Entries
Donor Recipient
NAT result for Deferral type notification lookback
HCV
Negative None None No
Positive 1 yr unless seroconverts to In person Yes
HCV antibody POS
Pos pool/Indiv neg Surveillance In person No
Pos pool/QNS 1 yr or until retest negative In person No
HIV
Negative None None No
Positive 6 mo unless seroconverts In person Yes
to HIV antibody POS
Pos pool/Indiv. neg Surveillance In person No
Pos. pool/QNS 6 mo or until retest negative In person No
Note: In those cases where the TMA discriminatory assay is inconclusive, the donors are managed as
if the result were positive.
REFERENCES
1. Busch M, Chamberland M, Epstein J, et al. Vox Sang 1999; 77:6776.
2. Eastlund T. Monetary blood donation incentives and the risk of transfusion transmit-
ted diseases. Transfusion 1998; 38:874882.
3. Alter HJ, Holland PV, Purcell RH, et al. Post-transfusion hepatitis after exclusion of
commercial hepatitis B antigen positive donors. Ann Int Med 1972; 77:691699.
4. U.S. Food and Drug Administration. Guidance for Industry. Quality assurance in
blood establishments. Rockville, MD: CBER Office of Communication, Training and
Manufacturers Assistance, 1995.
5. Giles RE, Perry KR, Parry JV. Simple rapid test devices for anti-HIV screening: Do
they come up to the mark? J Med Virol 1999; 59:104109.
6. Mvere D, Constantine M, et al. Rapid and simple hepatitis assays: encouraging results
from a blood donor population in Zimbabwe. Bull WHO 1996; 74:1924.
7. Constantine NT, Callahan JD, Watts DM. Retroviral Testing: Essentials for Quality
Control and Laboratory Diagnosis. Boca Raton, FL: CRC Press, 1992.
8a. Linden JV, Wethers J, Dressler KP. Controversies in Y medicine: use of external
controls in transmissible disease testing: pro. Transfusion 1994; 34:550551.
8b. Epstein JS. Controversies in transfusion medicine: use of external controls in
transmissible disease testing: con. Transfusion 1994; 34:552553.
9. Code of Federal Regulations, 42CFR Part 493. Washington, DC: U.S. Government
Printing Office, 2000.
10. U.S. Food and Drug Administration. Draft Guidance for Industry: revised recommen-
dations for the invalidation of test results when using licensed and 510(k) cleared
216 Valinsky
blood borne pathogen assays to test donors. Rockville, MD: CBER Office of
Communication, Training and Manufacturers Assistance, 1999.
11. Green GA 4th, Carey RN, Westgard JO, et al. Quality control for qualitative assays:
quantitative QC procedure designed to assure analytical quality required for an
ELISA of hepatitis B surface antigen. Clin Chem 1997; 43:16181621.
12. Schreiber GB, Busch MP, Kleinman SH, Korelitz JJ. The risk of transfusion-
transmitted viral infections. The Retrovirus Epidemiology Donor Study. N Engl J
Med 1996; 334:16851690.
13. Busch MP, Lee LL, Satten GA, et al. Time course of detection of viral and serologic
markers preceding human immunodeficiency virus type 1 seroconversion: impli-
cations for screening of blood and tissue donors, Transfusion 1995; 35:9197.
14. Kleinman S, Busch MP, Korelitz JJ, Schreiber GB. The incidence/window period
model and its use to assess the risk of transfusion-transmitted human immunodefi-
ciency virus and hepatitis C virus infection. Transfus Med Rev 1997; 11:155172.
15. Vrielink H, Reesink HW, van den Burg PJ, et al. Performance of three generations of
anti-hepatitis C virus enzyme-linked immunosorbent assays in donors and patients.
Transfusion 1997; 37:845849.
16. Lee JH, Roth WK, Zeuzem S. Evaluation and comparison of different hepatitis C
virus genotyping and serotyping assays. J Hepatol 1997; 26:10011009.
17. Cheingsong-Popov R, Osmanov S, Pau CP, et al. Serotyping of HIV type 1 infections:
definition, relationship to viral genetic subtypes, and assay evaluation. UNAIDS
Network for HIV-1 Isolation and Characterization. AIDS Res Hum Retroviruses
1998; 14:311318.
18. Schable C, Zekeng L, Pau CP, et al. Sensitivity of United States HIV antibody tests
19. Jongerius JM, Webster M, Cuypers HT, et al. New hepatitis B virus mutant form in
a blood donor that is undetectable in several hepatitis B surface antigen screening
assays. Transfusion 1998, 38:5659.
20. Zaaijer HL, van Rixel T, van Exel-Oehlers P, et al. New anti-human immunodefici-
ency virus immunoblot assays resolve nonspecific western blot results. Transfusion
1997; 37:193198.
21. Kleinman SH, Busch MP, Stramer SL, Watanabe K. Use of HIV-1/HIV-2 recombinant
immunoblot assay (RIBA) to resolve infection status of donors with possible false
positive HIV Western blot (WB) (abstract). Transfusion 1997; 37(suppl):47S.
22. Vrielink H, Zaajier HL, van der Poel CL, et al. New strip immunoblot confirmation
of HTLV-I/II infection. Vox Sang 1996; 70:114116.
23. Kleinman S, Busch M, Rawal B, Glynn S. Evaluation of HBsAg neutralization test
positive (Neut POS) donors with negative (NEG) anti-HBc (abstract). Transfusion
1998; 38(suppl):92S.
24. Stramer SL, Aberle-Grasse J, Bordsky JP, et al. US blood donor screening with p24
antigen (Ag): one year experience (abstr). Transfusion 1997; 37 (suppl):1S.
25. Strauss D, Valinsky JE, Kessler D, et al. Specificity of HIV-1 p24 antigen neutraliza-
tion (abstract). Transfusion 1996; 36(suppl):38S.
26. Murthy KK, Henrard DR, Eichberg JW, et al. Redefining the HIV-infectious window
period in the chimpanzee model: evidence to suggest that viral nucleic acid testing
can prevent blood-borne transmission. Transfusion 1999; 39:688693.
27. Busch MP, Kleinman SH, Jackson B, et al. Committee report. Nucleic acid amplifi-
cation testing of blood donors for transfusion-transmitted infectious diseases: report
of the Interorganizational Task Force on Nucleic Acid Amplification Testing of Blood
Donors. Transfusion 2000; 40:143145.
28. U.S. Food and Drug Administration. Guidance for Industry: in the manufacture and
clinical evaluation of in vitro tests to detect nucleic acid sequences of human
immunodeficiency viruses types 1 and 2. Rockville, MD: CBER Office of Commu-
nication, Training and Manufacturers Assistance, 1999.
29. Tilston P, Morris DJ, Klapper PE, Corbitt G. Commercial assay for hepatitis C virus
RNA. Lancet 1994; 344:201202.
30. MacDonough SH, Giachetti C, Yang Y, et al. High throughput assay for the simulta-
neous or separate detection of human immunodeficiency virus-1 (HIV-1) and hepa-
titis type C virus (HCV). Infusionther Transfusionmed 1998; 25:1649.
31. Myers TW, Gelfand DH. Reverse transcription and DNA amplification by a Thermus
termophilus DNA polymerase. Biochemistry 1994; 30:76617666.
32. Code of Federal Regulations, Title 21 CFR Part 610. 610.45. Washington, DC: U.S.
Government Printing Office.
33. Menitove J, ed. Standards for Blood Banks and Transfusion Services. 19th ed.
34. Food and Drug Administration. Memorandum: use of Fluorognost HIV-1 immuno-
fluorescent assay (IFA). Rockville, MD: CBER Office of Communication, Training
and Manufacturers Assistance, 1992.
35. Centers for Disease Control. Interpretive criteria used to report Western blot results
for HIV-1 antibody testing, MMWR 1991; 40:692695.
36. Kleinman S, et al. False-positive HIV-1 test results in a low risk screening setting
of voluntary blood donation. Retrovirus Epidemiol Donor Study, JAMA 1998;
280:10801085.
37. Sayre KR, Dodd RY, Tegtmeier G, et al. False-positive human immunodeficiency
virus type 1 western blot tests in non-infected blood donors. Transfusion 1996;
36:4552.
38. Human Immunodeficiency Virus Type 1 (HIV-1) Fluorognost HIV-1 IFA.
Manufacturers package insert. Vienna, Austria: Waldheim Pharmazeutika,
Ges.m.b.H, 1992.
39. Centers for Disease Control and Prevention. Update: HIV-2 infections among blood
and plasma donorsUnited States. JAMA 1995; 274:10071008.
40. U.S. Food and Drug Administration. Memorandum: recommendations for donor
screening with a licensed test for HIV-1 antigen. Rockville, MD: CBER Office of
Communication, Training, and Manufacturers Assistance, 1995.
41. Stramer S, Aberle-Grasse J, Bordsky J, et al. United States blood donor screening
with p24 antigen: (one year experience) (abstract). Transfusion 1997; 37(suppl):1S.
42. Food and Drug Administration. Guidance for industry: donor screening for anti-
bodies to HTLV-II. Rockville, MD: CBER Office of Communication, Training and
Manufacturers Assistance, 1997.
43. Vrielink J, Reesink HW, Zaaijer JL, et al. Sensitivity and specificity of four assays to
detect human T-lymphotropic virus type I or type I/II antibodies. Transfusion 1996;
36:344346.
218 Valinsky
44. Liu H, Sha M, Stramer S, et al. Sensitivity and specificity of human T-lymphotropic
virus (HTLV) types I and II polymerase chain reaction and several serologic assays
in screening a population with a high prevalence of HTLV-II. Transfusion 1999;
39:11851193.
45. Anderson DW, Epstein JS, Lee TH, et al. Serological confirmation of human
T-lymphotrophic virus Type I infections in healthy blood and plasma donors. Blood
1989; 74:25852591.
46. Gallo D, Penning LM, Hanson CV. Detection and differentiation of antibodies to
human T-cell lymphotropic virus types I and II by the immunofluorescence method.
J Clin Micro 1991; 29:23452347.
47. Busch MP, Kleinman SH, Stramer SL. False positive HTLV-I/II supplemental test
results in blood donors (abstr). Transfusion 1999; 39:105S.
48. Stramer SL, Layug L, Trenbeath J, et al. Use of a second EIA in an HTLV-1/HTLV-2
algorithm (abstr). Transfusion 1998; 38(suppl):81S.
49. American Association of Blood Banks. Association bulletin #99-9. Dual enzyme
immuno assay (EIA) approach for deferral and notification of anti-HTLV-I/II EIA
reactive donors. Bethesda, MD: American Association of Blood Banks, 1999.
50. Strauss D, Valinsky JE, Kessler D, Bianco C. Decreased specificity of HBsAg
confirmation as a result of decreased specificity of the screening test (abstract).
51. Busch MP, Tobler LH, Stramer S, et al. Yield of HCV EIA 3.0 vs EIA 2.0 in screening
of US blood donors (abstr). Transfusion 1997; 37(suppl):111S.
52. Skidmore S. Recombinant immunoblot assay for hepatitis C antibody. Lancet 1990;
335:1346.
53. Schifman RB, Rivers SL, Sampliner RE, Kramer JE. Significance of isolated hepatitis
B core antibody in blood donors. Arch Int Med 1993; 153:22612266.
54. Lok A. Hepatitis B infection: pathogenesis and management. J Hepatol 2000;
32(suppl):8997.
55. Cable R, Badon S, Pray C, Popovsky MA. Limited utility of alanine aminotransfer-
ase screening of hepatitis C antibody screened blood donors. Transfusion 1997;
37:206210.
56. Committee for Proprietary Medicinal Products. Plasma derived medicinal products:
position paper on ALT testing, CPMP/BWP/385/99, 1999.
57. American Association of Blood Banks. Association bulletin #95-7. FDA notification
requirement: ALT procedure change, change to the circular of information. Bethesda,
MD: American Association of Blood Banks, 1995.
58. Orton S, Liu H, Cable R, et al. Prevalence of circulating T. pallicum in STS+/F
TA-Abs+ blood donors (abstr). Transfusion 1999; 39(suppl):2S.
59. Larsen SA, Hunter EF, Kraus SJ. A manual of tests for syphilis. Washington, DC:
American Public Health Association, 1990.
60. Win N, Islam SI, Peterkin MA, Walker ID. Positive direct antiglobulin test due
to antiphospholipid antibodies in normal healthy blood donors. Vox Sang 1997;
72:182184.
61. Fosdick M, Winslow D, Larsen S. Clinical significance of MHATP as a confirmatory
test for the Olympus PK-TP system (abstr). Transfusion 1994; 34(suppl):49S.
62. Strauss D, Del Valle C, Valinsky JE, et al. Donor screening for syphilis using the
Olympus PK-TP assayimplications for donor notification and re-entry. Transfusion
1993; 33(suppl):38S.
63. Committee for Proprietary Medicinal Products. The introduction of nucleic acid
amplification technology (NAT) for detection of hepatitis C virus RNA in plasma
pools, addendum CPMP/BWP/390/97 to CP/BWP/269/95, 1998.
10
Bacterial Contamination
Laura L. Spinelli and Mark E. Brecher

University of North Carolina, Chapel Hill, North Carolina
Transfusion-transmitted bacterial infection remains a persistent complication of

transfusion. Currently the aggregate risk of contracting a viral infection [hepatitis
B virus (HBV), hepatitis C virus (HCV), human immunodeficiency virus types 1
and 2 (HIV-1/2), of human T-cell lymphotrophic virus types I and II (HTLV-I/II)]
is estimated at 1 in 34,000 (1), while the incidence of platelet bacterial contami-
nation is approximately 1 in 1000 and is thought to cause severe morbidity or
death in as many as 150 people per year in the United States (2). While much
attention has been focused by the public and the media on transfusion-transmitted
disease (particularly HIV), increasing awareness and improved testing and direct
questioning of donors have significantly decreased the transmission of hepatitis,
retroviruses, and other viruses. Bacterial contamination of blood products, how-
ever, has been largely overlooked and is now thought to be the major cause of
mortality from transfusion-transmitted disease. The risk of receiving a bacterially
contaminated platelet may be 50- to 250-fold higher than the risk of transfusion-
related infection per unit associated with HIV-1, HBV, and HTLV-I/II (2).
I. TRANSFUSION-TRANSMITTED BACTERIAL
INFECTION OF PLATELETS
Sepsis due to transfusion of bacterially contaminated platelets is the most com-
mon transfusion-transmitted disease, principally because of the fact that platelets
are stored for up to 5 days at 2024C (3). In the years 19861991 there was a
total of 182 transfusion-related deaths reported to the U.S. Food and Drug
Administration, (FDA) 29 of which were due to bacterially contaminated blood
products. Some 21 (72%) of these deaths were associated with the transfusion of
221
222 Spinelli and Brecher
bacterially contaminated platelets (2). Five fatalities were reported in both 1990
and 1991. Five to six million random platelet concentrates and single donor
apheresis are transfused per year in the United States (46). This would mean that
there is at least a one in a million chance of death due to sepsis per unit transfused.
However, it is widely suspected that platelet bacterial sepsis is frequently unrec-
ognized and thus underreported.
The contamination rate for platelets is approximately 1 in 1000 per individ-
ual unit (7,8). Single unit contamination is similar for both platelet concentrates
made from whole blood and single-donor apheresis concentrates, but the ultimate
risk of sepsis is likely to be 610 times greater with pooled random units, as there
is a 6- to 10-fold increased donor exposure (7,9). It has been estimated that as
many as 150 people per year in the United States suffer severe morbidity and or
mortality as a consequence of a platelet transfusion (2).
In a recent study of symptomatic bacteremia following platelet transfusion
in 161 bone marrow transplant recipients in Hong Kong, it was found that 1 in
2000 units of platelet concentrates were bacterially contaminated. This translated
to 1 in 350 pooled platelets being contaminated. Of those patients who were
febrile (elevation of temperature of 1C) following platelet transfusion, 1 in 4
(27%) were found to have received a bacterially contaminated unit. Of those
found to have a 2C rise in temperature following a platelet transfusion, 50%
were found to have received a bacterially contaminated unit (10). In this multiply
transfused patient population, the chance of receiving a bacterially contaminated
platelet was 1 in 16. Of the 10 patients who are known to have received a
bacterially contaminated unit, 4 suffered from septic shock.
It is felt that the incidence of sepsis due to red cell transfusions has
decreased over the last 30 years (11), most likely due to the introduction of
disposable collection sets and bags in the 1960s. Yet, the frequency of bacterial
sepsis due to platelet transfusion has increased, largely due to increasing use of
this component and prolonged storage time at 20-24C.
A. Organisms and Source

Skin commensals such as Staphylococcus epidermidis and Bacillus cereus are the
organisms most often implicated in platelet bacterial contamination (12). These
organisms typically do not grow at 06C but survive and multiply readily at
2024C, the storage temperature of platelets. In descending order, the organisms
most commonly implicated in fatalities (Table 1) are Staphylococcus aureus,
Klebsiella pneumoniae, Serratia marcescens, and S. epidermidis (13). Other
isolated organisms include Salmonella, Escherichia coli, Pseudomonas aer-
uginosa, and B. cereus (1319). Fatalities due to platelet contamination tend to
be equally divided between gram-positive and gram-negative organisms (13).
Potential sources of contaminant organisms include contamination of the collec-
Bacterial Contamination 223
Table 1 Reported Fatalities Due to Platelet Sepsis in the United States,

19891991
Year No. Associated bacteria Gram stain result
1987 4 Klebsiella oxytoca Negative

Serratia marcescens Negative
Klebsiella pneumoniae Negative
Staphylococcus epidermidis Positive
1988 2 Alpha-hemolytic streptococci Positive
Serratia marcescens Negative
1989 3 Salmonella cholera Livingston Negative
Staphylococcus warneri Positive
Enterobacter aerogens Negative
1990 5 Staphylococcus aureus (4) Positive
Streptococcus mitis Positive
1991 5 Pseudomonas aeruginosa Negative
Staphylococcus aureus (2) Positive
Escherichia coli Negative
Klebsiella pneumoniae Negative
tion bag, tubing, or anticoagulant and donor-related factors such as transient

bacteremia. However, contamination of platelets is thought to occur principally
during phlebotomy because of inadequate sterilization and or skin core removal
by the collection needle. Despite excellent technique, one cannot assure a sterile
venipuncture, because organisms harbored in sebaceous glands and hair follicles
cannot be completely disinfected. Scarring or dimpling of the venipuncture site
due to prior donations has also been recognized as a risk factor for aseptic
venipuncture, as these areas frequently contain recessed pits that are difficult
to adequately sterilize (20,21). One percent iodine solution is the only topi-
cal cleanser that has been shown to be 100% effective in providing donor
skin sterility. Donors who are allergic to iodine are often cleansed with some
type of acetone alcohol solution, which has been shown to effect only a 50%
reduction in skin microbials (22). It would seem prudent not to prepare platelets
from such donors.
B. Clinical Presentation
The clinical sequelae of the transfusion of bacterially contaminated platelets
may range from asymptomatic to mild fever (which may be indistinguishable
from a nonhemolytic transfusion reaction) to acute sepsis, hypotension, and death.
The presentation is much more varied and often less severe than that of pa-
tients infected by transfusion of bacterially contaminated red cells (9). In fact,

it is felt that sepsis due to transfusion of contaminated platelets is vastly un-
derrecognized and underreported. Patients in need of platelet transfusion are
often leukopenic, and fever can be readily attributed to other infectious causes
(36). Much of the underreporting may occur because the organisms most fre-
quently found in platelet contamination are skin commensals that are the same
organisms implicated in catheter sepsis. Therefore, sepsis, which in these pa-
tients is often attributed to catheters, may very well have its origin in platelets
transfused several hours before. The overall mortality rate of platelet-associated
sepsis reported in the literature is 26% (12). Any patient who develops fever
within 6 hours of platelet infusion should be started on empiric broad-spectrum
antibiotics (10,23).
II. TRANSFUSION-TRANSMITTED BACTERIAL

INFECTION OF RED CELLS
Of the 29 fatalities due to transfusion of bacterially contaminated blood products
reported to the FDA from 1986 to 1991, 8 (28%) were associated with transfusion
of red cells (2). Considerable attention has been paid to red cellassociated sepsis
because of the unusual organisms and the high mortality associated with these
episodes. Yet, data from the Centers for Disease Control and Prevention (CDC)
for 19871994 suggest a contamination rate of less than one per million red cell
units (22 cases per 28 million units of red cells) (2).
A. Organisms and Source

The most commonly implicated organism in bacterial contamination of red cells
is Yersinia enterocolitica, usually serotype 0:3 (2426). Seven of the eight
fatalities due to sepsis secondary to transfusion of contaminated red cells reported
to the FDA in the years 19861991 were caused by Y. enterocolitica (2). This
organism grows readily in the presence of dextrose and iron at 4C, the temper-
ature at which red cells are stored. Contamination is attributed to transient
bacteremia in asymptomatic infected donors. On retrospective analysis, up to
50% of donors implicated in contamination of products recalled gastrointestinal
symptoms in the weeks prior to and surrounding donation (23,2730). White cells
of infected persons contain viable organisms that multiply during storage. These
cells lose membrane integrity and disintegrate, releasing bacteria into the stored
units (31,32). Consequently, contamination is directly related to storage time, with
a significantly higher incidence as units age more than 25 days (24). In fact, all
seven cases of Yersinia contamination reported to the CDC in the years between
1986 and 1991 were older than 25 days (2). However, cases of Yersinia red cell
sepsis have been reported following as little as 714 days of storage (29,33).
Serratia marcescens has been implicated in the contamination of plastic
blood containers manufactured in a plant in Belgium, which affected several
Danish and Swedish transfusion centers. The outbreak was thought to involve
the manufacturing process and resulted in the closing of a blood bag manufac-
turing plant (34,35). Serratia liquefaciens has been recently recognized as both a
red cell and a platelet contaminant in the United States (L. Bland, personal
communication). Other gram-negative organisms that have been repeatedly asso-
ciated with red cell contamination are Pseudomonas and Enterobacter spp.
(12,24,36,37).
B. Infection of Autologous Red Cell Units

Although autologous blood is generally considered a safer blood product, to
date there have been at least five cases of bacterial contamination of autologous
red cell units, four due to Y. enterocolitica and one due to S. liquefaciens (3842;
S. Cookson, personal communication). Fortunately, all recipients survived. Upon
retrospective questioning, all patients infected by Yersinia recalled gastrointesti-
nal symptoms in days prior to donation. In the case of Serratia contamination,
the patients infected toe ulcer was presumed to be the source.
C. Clinical Presentation
Symptoms associated with transfusion of contaminated red cells are more severe
and rapid in onset than those caused by an infected platelet transfusion. Patients
frequently develop high fever (temperatures as high as 109F have been observed)
and chills during or immediately following transfusion (42,43). From 1987 to
February 1996, 20 recipients of Yersinia-infected red cells in 14 states were
reported to the CDC. Twelve of the 20 recipients died in 37 days or less following
transfusion. The median time to death was 25 hours! Of the 7 who developed
disseminated intravascular coagulation (DIC), 6 died. Signs, symptoms, and
complications are summarized in Table 2 (44).
In an anesthetized patient, hypotension, oozing, oliguria/anuria, and fever
should alert the anesthesiologist to the problem. Affected recipients may also
experience nausea, vomiting, and diarrhea. In severe instances, DIC, vascular
collapse, and death can rapidly ensue. If a septic transfusion reaction is suspected,
the transfusion should be discontinued immediately and not restarted. The mini-
mum number of organisms necessary to cause clinical symptoms is not known,
but concentrations of 108 CFU/mL or greater have been associated with severe
reactions and death (11).
Table 2 Morbidity and Mortality Associated with Yersinia

enterocoliticaContaminated Red Blood Cells
Signs, symptoms, and mortality Number Percentage
Chills 16 80
Fever 14 70
Hypotension 13 65
DIC 7 35
Death 12 60
Source: Ref. 60.
III. TRANSFUSION-TRANSMITTED BACTERIAL

INFECTION OF PLASMA AND CRYOPRECIPITATE
Cell-free products such as plasma and cryoprecipitate are stored in the frozen state
and thus are rarely associated with contamination. However, in some cases
Pseudomonas cepacia and P. aeruginosa have been cultured from cryoprecipitate
and plasma thawed in contaminated water baths (45,46).
IV. TRANSFUSION-TRANSMITTED BACTERIAL

INFECTION OF PLASMA PROTEIN CONCENTRATES
Human serum albumin is a good culture medium and preserves viability. The
heating step (60C for 10 hours) in the manufacturing of albumin is performed to
inactivate certain viruses, not to assure bacterial sterility (47). Achieving bacterial
sterility would require autoclaving; albumin would denature at these extreme
temperatures. On occasion, specific lots of albumin product have been found to
be contaminated with bacteria, typically Pseudomonas species. These lots have
produced transient bacteremias, hypotension, and febrile reactions in recipients
(48). Most recently, two patients in two different hospitals developed Entero-
bacter cloacae septicemia after receiving albumin (47,49,50). This resulted in a
worldwide recall of 5, 20, and 25% albumin, Monoclate-P (antihemophilic
factor), and plasma protein fraction. It is suspected that cracks in the seal may
have been responsible for the contamination.
A. Detection
There is no universally accepted test, method, or device used in the detection of
bacterially contaminated blood product. Ideally, the test must be sensitive (pref-
erably to 104105 CFU/ml), specific, inexpensive, simple to use, capable of
detecting the commonly implicated organisms, adaptable for new bacterial threats,
and usable in a hospital environment. Approaches currently under investigation
are summarized below.
1. Bacterial Staining
Stain of a smear immediately prior to transfusion is sensitive to approximately
106 CFU/mL for gram-positive and 108 CFU/mL for gram-negative organisms
(51,52). Gram stain is most often employed, although Wrights stain and acridine
orange have also been recommended (53,54). This test lacks the necessary
sensitivity to detect all clinically significant bacterial contamination and has been
associated with both high false-positive and false-negative rates (53).
Acridine orange stains have a sensitivity of 104 CFU/mL, but require a
fluorescent microscope for visualization (55,56). While there is an increased rate
of detection, the cost and time required have precluded any further interest in this
approach.
2. Bacterial Culture
Because only 0.11 mL of product is typically cultured on an agar plate, the
sensitivity of this method is limited. This may be overcome in part by the use of
automated blood culture systems (57,58). This procedure generally requires
greater than 24 hours for the culture to grow and often is fraught with false
positives resulting from inoculation/culture technique.
3. Endotoxin Assays
The Limulus amebocyte lysate (LAL) assay is the most sensitive, detecting 104
mg of endotoxin per 0.1 mL of plasma, or 104 CFU/mL (59). However, this test
is not widely available and is of limited value; gram-positive bacteria are not
detected by this method because they do not contain endotoxin in their cell walls.
4. Bacterial Ribosomal Assays

A technique that is sensitive to 105 CFU/mL is the use of rRNA (ribosomal RNA)
chemiluminescence linked probes. This technique utilizes a single-stranded
(DNA) probe complementary to a highly conserved bacterial rRNA found in all
bacterial species. Requiring only 23 hours to perform, this method has detected
1001000 CFU/mL of S. aureus and the majority of B. cereus, P. aeruginosa,
S. aureus, and S. epidermidis discontaminants exceeding 104 CFU/mL (60,61).
The probe assay is depicted in Figure 1. The bacteria are enzymatically lysed,
which causes bacterial rRNA to be released. A labeled DNA probe that combines
with the complementary rRNA to form a stable DNA:RNA hybrid is then added.
Any unhybridized probe is selectively hydrolyzed by the addition of base.
228
Figure 1 The rRNA chemiluminescence linked probe assay is schematically represented in four stages. (1)
RNA preparation: bacteria are enzymatically lysed and the rRNA is released. (2) Hybridization: a labeled DNA
probe combines with the complementary rRNA to form a stable DNA:RNA hybrid. (3) Differential hydrolysis:
the unhybridized probe is selectively hydrolyzed by base. (4) Detection: a chemiluminescence reaction is
Spinelli and Brecher
initiated in the presence of base and hydrogen peroxide. The amount of luminesence is quantitated in relative
light units (RLU). (From Ref. 23.)
A chemiluminescent reaction is then initiated in the presence of base and hydro-

gen peroxide, and the resulting amount of luminescence is quantitated in relative
light units (RLU). Recently the use of probe technology to bacterial rRNA has
been applied to detect bacterial contamination of food. DuPont has developed two
machines (the Riboprinter Microbial Characterization System and the Bax sys-
tem), which, in a matter of hours, can scan a sample for dangerous bacteria. In the
future, this technology might be applied not only to detection of food contamina-
tion, but also to detection of blood product contamination and prevention of
resultant catastrophic events (62).
5. Polymerase Chain Reaction

The polymerase chain reaction (PCR) has been evaluated in the detection of
Yersinia. Feng et al. were able to detect 103 organisms in 100 L (5 103
CFU/mL) of blood (63). Unfortunately, this technique, as described, is time
consuming, identifies only Yersinia, and is subject to problems of contamination,
thereby compromising specificity.
6. Altered Biochemistry
Bacterial metabolism during storage alters certain biochemical properties, namely
glucose, pCO2, pO2, and pH (6467). As bacteria proliferate, they consume
oxygen and glucose, causing a decrease in these analytes, and CO2 is transiently
increased (Fig. 2). Changes in glucose can be rapidly and inexpensively detected
by the use of glucometry (Fig. 3) (65). Two groups have studied the utilization of
dipstick reagents to detect decreased glucose and pH in contaminated platelet
concentrates (64,65). In platelets contaminated with 107 CFU/mL, the overall
sensitivity and specificity were 94% and 98%, respectively (65). In some cases,
S. aureus and K. pneumoniae were detected in the 103105 CFU/mL range (65).
Change in pH has been detected using a CO2-sensitive label placed on the outside
surface of platelet containers. During aerobic growth of bacteria, CO2 diffuses
through the wall of the container and the color in the label changes. Unfortunately,
platelets themselves release CO2 in relation to their number and metabolism,
limiting the utility of this method (68). An alternative technique is the use of
CO2-sensitive labels on gas-permeable plastic bags with culture medium, trans-
ferring only part of the platelet component into it. Unfortunately, this procedure
is also quite insensitive, requiring large numbers of bacteria for detection (69).
7. Visual Inspection
An important observation made by Kim et al. was that grossly contaminated
addition solution (AS) red cell units were darker in appearance than their attached
segmented tubing. As a result of bacterial growth and metabolism, red cells are
Figure 2 Growth of Yersinia enterocolitica in AS red cells. (a) Growth curves of two
inoculated units indicated by dashed lines. (b) Oxygen tension plotted over storage time
for two sterile (culture negative) and the two inoculated (culture positive) AS3 red cell
units. (c) CO2 tension plotted over storage time for two sterile and the two inoculated AS3
red cell units. (d) Free supernatant hemoglobin over storage time for the same two sterile
and two inoculated AS red cell units. (Figures 2a, b, and d modified from Ref. 23.)
Figure 3 (a) Glucose analysis of sterile platelets stored in PL-732 bags (mean, mean
-2SD, and mean -3SD) versus platelets stored red in PL-732 bags with CPD and inoculated
with S. aureus (day 0). (b) Glucose analysis of sterile platelets stored in CLX bags (mean,
mean -2SD, and mean-3SD) versus platelets stored in CLX bags with CP2D (double
dextrose) and inoculated with S. marcescens (day 0).
lysed and oxygen consumed, causing a decrease in the oxygen saturation of

hemoglobin and a darkening of the bag compared to the attached segmented
tubing (Fig. 2) (70,71). In contaminated red cells, as a rule, the bacteria can be
isolated only from the bags, and the segmented tubing attached to the bag is
invariably sterile. While the same concentration of bacteria would be expected to
be present in the bag and segmental tubing, most of the bacteria are actually
killed. Since only a very few survive, the greater volume found in the bag
compared with the volume in the segmented tubing favors growth in the bag.
We have observed this color change as early as day 24 of storage with Y. en-
terocolitica and day 20 of storage with S. liquefaciens (unpublished observations).
Careful inspection of units, comparing the color of the bag to that of the segments,
is a quick and easy screening tool.
The use of visual inspection is more difficult with platelets. One proposed
way to detect contamination is the observation of decreased swirling or
streaming in platelet bags (37,54,72). This method is based on the observation
that viable, intact platelets are discoid in shape and produce a swirling phenom-
enon when gently squeezed and visualized against a light source. Platelets do not
swirl if the pH is low or they have been exposed to cold (7380). Platelet
concentrates contaminated with bacteria frequently have a decreased pH as the
bacteria respire and produce CO2, thus causing a decreased swirling effect
when visualized. However, several investigators remain skeptical as to the clinical
utility of this method. Significantly, a recent study found that more than one third
of swirling-negative platelets had a normal pH and as many as 18% of
5-day-old platelet concentrates did not swirl and thus would be unnecessarily
destroyed (81). Aggregated platelet concentrates can also be indicative of bacte-
rial contamination (particularly with gram-negative organisms) and should not
be infused.
B. Prevention
Many methods have been proposed to aid in the prevention of bacterial contam-
ination of blood products. While many show great promise, there has not been
wide acceptance and institution of any one technique.
1. Donor Questioning
Most cases of red cell contamination are thought to be attributed to transient
asymptomatic bacteremia in donors. Questions are asked prior to donation to
tease out prospective bacteremic donors (24,82). Histories of recent dental
procedures, gastrointestinal or genitourinary manipulation, and breast feeding
may all be associated with bacteremia and would cause a potential donor to be
deferred. Some have suggested asking donors about a recent history of diarrhea
or abdominal pain to help identify persons infected with Y. enterocolitica (82).

Unfortunately, this does not appear to be a specific method of prevention, as this
would defer approximately 9.7% of potential donors, most of whom are not
infected (83). In addition, this method lacks sensitivity, as only 13 of 20 donors
associated with Y. enterocoliticacontaminated red cells recalled a history of
gastrointestinal symptoms (44). Of patients with gastrointestinal symptoms who
have a stool culture performed, only 2% grow Yersinia (8486).
2. Skin Core Contamination

Most cases of platelet contamination arise as the result of inadequate skin
sterilization prior to donation. In addition, skin fragments are drawn up into the
collection bag during the initial phase of donation and provide a source of
infection (20,21). A variety of simple methods to prevent skin core contamination
are under investigation. Possibilities include the use of a trochar instead of a
large-bore needle for collection, placement of a filter screen at the base of the
collection bag, and disposal of the first few milliliters of blood (87,88). A study
performed on 22,000 blood donations by the Red Cross in the Netherlands in
which the first 10 mL of donor blood were diverted from the primary bag showed
that 16 of the first 5 mL aliquots were bacterially contaminated, while only 2 of
the second 5 mL aliquots were culture positive. Similar reports have been
documented in vitro studies of collection needles (22,23,88,89,102).
3. Prestorage Culture
Prestorage culture as a preventive measure against the transfusion of bacteri-
ally contaminated products is problematic. It has been found that 0.20.5%
of units are culture positive (57). As a result of artifactual contamination dur-
ing culturing, the rate of positive cultures may be higher than the amount
of bacterial contamination actually present in stored blood products. A high
number of false positives would likely result in the unnecessary disposal of
noncontaminated units. Additionally, as numerous studies in our and other
laboratories have demonstrated, even with known bacterially inoculated units,
units are frequently culture negative for several days before later cultures be-
come positive (24,90,91). This is likely due to the antibacterial effects of plasma
and granulocytes. Alternatively, breakdown of granulocytes or skin plugs may
lead to a late release of bacteria; thus, many cases may not be detected by
prestorage culture.
4. Bacteriocidal Techniques
The use of antibiotics to assure a sterile product, while effective, is not felt to be
an acceptable solution. Fears of the selection and development of antibiotic-
resistant strains of bacteria and of merely trading one rare event (drug anaphy-
laxis) for another (bacterial sepsis) preclude such use (37,51). A promising
approach is the use of light or ionizing radiation to produce not only non-
immunogenic, but also sterile, blood products (9295). A photochemical decon-
tamination system for platelet concentrates using long-wavelength ultraviolet
radiation and 8-methoxypsoralen has been shown to inactivate 2530 logs/h of
E. coli or S. aureus (96). The techniques that utilize light are restricted to
nonopaque products such as plasma and possibly platelets. With regard to plate-
lets, there has been concern as to the effects this technology might have on
metabolic and functional properties as well as concerns of mutagenicity, logistics,
and cost (97). Because of such concerns and technical hurdles, such methodology
has, to date, been limited to the research laboratory.
5. Storage Time
In 1991, the Blood Product Advisory Committee (BPAC) of FDA reviewed
all cases of Yersinia sepsis from red cell units reported to either FDA or the
CDC during the late 1980s. At that time all reported cases of red cellassociated
yersinia sepsis in the United States had occurred in units older than 25 days.
As a result of the time necessary for the bacteria to attain a lethal concentration,
the BPAC proposed reducing the storage time of red cells from 42 to 25 days (98).
This recommendation was subsequently rejected for the following reasons:
1. A questionnaire distributed at the time revealed that 20% of red cell
units in stock at over 1500 blood banks and transfusion services were
more than 28 days old (98). Discarding such units would have severely
compromised the nations blood supply.
2. A shorter outdate would then require recruitment of new donors; it
was estimated that the addition of a quarter of a million donations
per year would be required to replace the losses due to outdates
(98,99). This would involve additional risk since first-time donors are
known to be at a greater risk of carrying disease because their blood
has not been repeatedly tested, as has the blood of veteran donors
(98,99).
3. Units less than 25 days old can also cause sepsis. While decreasing the
storage time may lessen the problem, it would not eliminate it entirely
(29,33,59,100).
4. Older units are less likely to transmit viruses such as HIV (101).
Increased platelet storage time has also been associated with increased
probability of contamination. At present, the allowable storage time for platelets
is 5 days, although in 1983 the shelf life of platelets was increased to 7 days
(Fig. 4). This increase was associated with an increase in the number of sep-
Figure 4 Platelet shelf life timeline.
tic events in older platelets, and the shelf life was again reduced to 5 days
in 1986 (102). Unfortunately, merely decreasing the shelf life did not elimi-
nate the problem of bacterial contamination. We and other laboratories have
shown that even a moderate inoculate (1050 CFU/mL) of certain bacteria, such
as B. cereus and P. aeruginosa, can have a minimal lag phase with a doubling time
of 12 hours (60). This can lead to a bacterial load of 108 CFU/mL in just 12
days. One approach to minimize platelet associated sepsis taken by the Oncology
Center of Johns Hopkins was to limit the transfusion of platelets to 4 days of
storage (103). It is unlikely that there will be further national reductions in storage
time of platelets because this would cause severe shortages of this component.
Other concerns within the blood banking community are also affecting
platelet availability. Because of the increased complexity of viral screening of
blood products, an increasing awareness of the potential for errors, and a need to
minimize cost and maximize efficiency of operations, many blood collection
organizations such as the American Red Cross are attempting to centralize or
regionalize disease marker testing. While commendable in concept, this central-
ization of testing can, because of the transport time of samples, result in the
decreased availability of one-day-old platelets. In 1982, the mean age of distrib-
uted platelets was 1.6 days, in 1983 (after extension of the dating period to 5 days)
it was 2.0 days, and in 1992 (after addition of increased laboratory testing) it was
2.5 days. In 1983, only 5% of issued platelets were older than 3 days. In 1992,
just 10% were older than 3 days. But with the introduction of centralized testing
by the Red Cross, the mean age of issued platelets increased to 2.7 days, with
20% older than 3 days (89). Not only can this decrease the available shelf life of
an already precariously limited supply of platelets, it can also decrease the
availability of fresh platelets, the most hemostatically effective and the least likely
to be bacterially contaminated.
6. Storage Temperature and Separation of Components

After inoculation of whole blood with Y. enterocolitica, there is an initial rapid
decline in the number of viable organisms, followed by a resumption of growth
after a lag phase of approximately 5 days (55,56,104106). When whole blood is
placed immediately at 4C, growth occurs more rapidly than if held at 10C for
24 hours prior to 4C storage (107). Bacterial growth of most pathogens is
inhibited by cold temperatures, but host defense mechanisms present in fresh
whole blood are also inhibited by storage at 4C. Yersinia with plasmid-encoded
complement resistance can become complement sensitive when blood is transiently
stored at 20C (105). Because phagocytosis and complement activation are impaired
by storage at 4C, it may be advantageous to allow red blood cells to remain in contact
with plasma for several hours prior to separation and storage of components.
Platelets are stored for a maximum of 5 days at 2024C because of concern
regarding the potential for bacterial contamination and progressive decline in
platelet function (storage lesion) if platelets are stored for longer periods of time
at these temperatures (108,109). While storage of platelets at 4C results in a
significantly lower rate of bacterial contamination, it also causes a temperature-
induced activation of platelets and a rapid decline in functional ability and in vitro
viability. In a recent study, cold storage of platelets was performed, mimicking
the endogenous inhibition of platelet activation through the addition of specific
second-messenger stimulators. It was reported that the platelets stored at 4C for
9 days displayed no loss in cell number, and that the study platelets recovered
partial functional ability and viability compared with control platelets stored at
22C for 5 days (110). If a practical method for storing platelets at 4C is
perfected, it would have the potential to reduce the risk of bacterial contamination
of this component.
7. Leukoreduction
Because of concern regarding the accumulation of leukocyte and platelet break-
down products in stored blood, as well as the immunogenicity of white cell
fragments, there has been considerable emphasis placed on the potential benefits
of prestorage versus poststorage leukoreduction (111,112). However, this posed a
potential dilemma, as it was possible that prestorage filtration would cause an
increase in bacterial contamination of blood products and septic complications
because of removal of the phagocytic leukocytes that would normally remove low
levels of bacteria present in these components. Fortunately, in the case of
Y. enterocolitica contamination of red cells, prestorage leukoreduction actually is
associated with a decrease in bacterial growth in inoculated red blood cells
(Table 3) (56,104,106,113,114). The mechanism by which leukoreduction re-
moves bacteria is multifactorial. Bacteria that have been phagocytized (Fig. 5)
but not killed are removed with the white blood cells. Alternatively, organisms
may be adsorbed to leukocytes or activated complement, to then be bound
indirectly to the filter, or the bacteria may directly adhere to the filter fibers
(116,117). Studies on leukoreduction of platelets have not shown such promising
results. Unlike the decrease in Y. enterocolitica growth in prestorage leukoreduced
Figure 5 Transmission electron micrograph showing (A) penetration and engulfment

and (B) internalization of Yersinia enterocolitica in leukocytes from a buffy coat prepara-
tion of a normal blood donor. The innoculation dose was 109 CFU added to 100 mL of a
buffy coat preparation. The unit was stored for 2 hours at room temperature before
sampling. The microorganisms are indicated by arrowheads. (From Ref. 115.)
red cells, neither a positive nor a negative effect on bacterial growth has been
demonstrated in prestorage leukoreduction of platelets (60,118,119).
8. Frozen Components
Contamination of blood components has occurred during the thawing process
(45,46). Use of a plastic overwrap that prevents outlet ports from direct exposure
to water from the warming bath is an easy and effective method to prevent
contamination (120). Disinfectants and frequent changing of water baths have not
been shown to prevent growth of bacterial species (11,37).
9. Immunoassay
The Red Cross is currently developing an immunochromatographic technique
with Binax, Portland, ME (22,89). This method is intended to be applied im-
mediately prior to issuance of the blood component. The goals of this effort are
to detect the 810 bacteria that account for 95% of bacterially infected platelets,
to detect >105 CFU/mL, to be performed in less than 20 minutes, to be simple
and require no special equipment, and to have a distinct yes or no answer. As yet,
it has not entered clinical trials.
Recently a monoclonal antibody (MAb900) has been described that detects
a prokaryotic elongation factor (EF-Tu). EF-Tu is one of the most abundant
proteins in prokaryotes and is not present in eukaryotic cells. As such, it is hoped
that the use of anti EF-Tu may make a very rapid (46 h) and effective screening
tool possible (121).
Table 3 Yersinia enterocolitica Growth and Prestorage Leukocyte Reduction

by Filtration
Inoculating
concentration Filtered Control
(CFU/mL) growth/total (%) growth/total (%) Filter type Ref.
100 0/10 (0) 10/10 (100) Sepacell PL-5N 31

65 1/5 (20) 4/4 (100) Leukotrap 58
65 1/5 (20) 4/4 (100) Leukotrap RC (Pall RC300) 58
10/150 3/8 (37) 8/8 (100) Pall BPF4 113
0.3132 3/24 (12) 16/24 (67) Sepacell R-500 106
2030,000 6/30 (20) 22/30 (73) Cellselect, NPBI 108
1.5 2/6 (33) 6/6 (100) Leukotrap RC (Pall RC300) 72
Total 16/88 (18) 70/86 (81)
V. TRANSFUSION-TRANSMITTED SYPHILIS
Treponema pallidum, the agent that causes syphilis, is a thinwalled, motile,
spirochete. This organism cannot be visualized with Gram stain, nor does it grow
on bacteriological media or cell culture, yet it is considered to be a bacterium,
and infection is treated with penicillin. Although it is a bacterium, it is often
treated as a distinct entity, different from other transfusion-transmitted bacterial
infection and is thus addressed in this separate section of the text. Donors infected
with T. pallidum may be asymptomatic with negative serology during periods of
spirochetemia (122). We are only aware of three cases of transfusiontransmitted
syphilis in the past 27 years (123125). While the organism is killed by storage
at 4C, it may live for 15 days at these cold temperatures (126,127). Subse-
quently, a rare posttranfusion infection may be associated with transfusion of a
very fresh unit of red blood cells from a donor who was in the seronegative phase
at the time of donation. There is greater concern with platelet transfusion, since
this component is stored at 2024C, temperatures suitable for growth of this
organism. The cardiolipin test is used in screening donated blood, but this test lacks
sensitivity for detecting an acute infection. Nevertheless, there is an extremely
low rate of transfusiontransmitted syphilis infection for a number of reasons:
1. Many infected donors are screened with donor questioning.
2. Although an insensitive test in the acute post infections setting, the
cardiolipin assay does pick up a number of infected donors.
3. Refrigerator storage results in the death of spirochetes.
4. Most patients receiving platelets are on antibiotics at the time of
transfusion, which would be bactericidal for any transmitted viable
organisms.
5. Because there is a high correlation between infection with T. pallidum

and viruses such as HIV (128131) and hepatitis B and C (132,133),
donors in the seronegative phase of syphilis may be excluded as a result
of a positive test for one of these viruses.
VI. INVESTIGATION, THERAPY, AND REPORTING

When transfusion-related sepsis is suspected, one should obtain the residual
blood in the bag for culture and stain. One should also culture the patients
blood and compare its isolates with the blood bag isolates. Antibiotic sensitivi-
ties may be helpful in confirming that an organism isolated from a blood bag
is the same organism isolated from a patients blood. We recommend that
antibiotic therapy be considered if the patient is not already receiving broad-
spectrum antibiotics.
It may be critical to locate other blood components from the same donor
since other liquid-stored components may also be bacterially contaminated. In
such cases, liquid components from the same donation should be identified,
quarantined, tested for contamination, and destroyed. If blood components from
the same donation have already been administered to other patients, the clinicians
caring for those patients should be notified.
Confirmed cases should be reported to:
Investigations and Prevention Branch
Hospital Infections Program, A-07
Centers for Disease Control and Prevention
(404) 639-1550
FAX (404) 639-3770
Fatalities must be reported to:
Office of Compliance
Center for Biologics Evaluation and Research
Food and Drug Administration
(301) 594-1191
VII. CONCLUSION
Bacterial contamination of blood components remains a problem today. While the
bacterial contamination of red cells, fresh frozen plasma, and cryoprecipitate
occur rarely, it is estimated that approximately 1 in 1000 platelet units are
contaminated as a consequence of storage at 2024C. Thus, the risk of receiving
a bacterially contaminated platelet transfusion exceeds the combined risk of
receiving an HIV-, hepatitis B, or hepatitis Ccontaminated blood component.

Unfortunately, recognition of bacterial sepsis is problematic because patients
often do not become symptomatic until several hours following a contaminated
platelet transfusion and the spectrum of organisms overlaps those seen in catheter
sepsis. It is therefore suspected that cases of platelet-related bacterial sepsis go
undiagnosed and are often falsely attributed to catheter sepsis.
Ironically, those patients who are the most ill equipped to handle a bolus of
intravenous bacteria are the ones most often exposed (Table 4). Recipients of
platelets tend to be the sickest patients, often having received chemotherapy with
resultant temporary marrow aplasia. It is in these patients who are immuno-
suppressed that sepsis from platelet contamination most often occurs.
Unlike testing for viruses, which only needs to be performed once at the
time of donation, testing of blood components for bacterial contamination due
to bacterial growth during storage must be performed close to the time of
transfusion. Methods are now being described for the rapid and sensitive testing
of blood components for bacterial contamination. Such methods may have the
added benefit of prolonging the shelf life of platelets, which is currently time
limited because of the increasing risk of bacterial contamination over extended
storage times.
Every year documented cases of fatalities from bacterially contaminated
blood components (principally platelets) are reported. It may be more cost-
effective to implement platelet bacterial testing than the currently implemented
tests for syphilis, anti HIV2, HIV p24 antigen, anti HTLVI/II, and antibody to
hepatitis B core antigen (anti-HBC) routinely performed on all donations in the
United States. Recent studies demonstrating high concentrations of cytokines
(TNF-, IL1, IL-8, and IL-1) following transfusion-associated sepsis offer a
glimmer of hope for possible future prospects for therapy with anticytokines
or cytokine antagonists (134). Efforts to perfect a method of refrigerated storage
of platelets also remain a possibility (110). However, our duty as blood bankers
Table 4 Factors Affecting Outcome of Transfusion of

Bacterially Contaminated Blood Products
Virulence of the organism
Immune status and general condition of the recipient
Concentration and bolus dose of bacteria transfused
Timely recognition and therapeutic intervention
Intensity of patient monitoringi.e., inpatient vs. outpatient
Medicines the patient is receivingi.e., antibiotics
and clinicians is to assure the safest blood product possible for both today
and tomorrow.
There remains some controversy regarding the actual risk of bacterial
contamination of blood, and no perfect screening or prevention methodology
currently exists. The perfect should not be the enemy of the good, and im-
plementation of partial solutions that have little risk of causing harm should be
encouraged (135). Rapid bacterial screening of platelets will likely be available
in the near future.
REFERENCES
1. Schreiber GB, Busch MP, Kleinman SH, Korelitz. The risk of tranfusion-transmitted
viral infections. N Engl J Med 1996; 334:16851690.
2. Association Bulletin 96-6. Bacterial Contamination of Blood Components. AABB
Faxnet, No. 294, August 1996. American Association of Blood Banks.
3. Dodd RY. Adverse consequences of blood transfusionquantitative risk estimates.
In: Blood Supply Risks, Perceptions, and Prospects for the Future. Bethesda, MD:
American Association of Blood Banks, 1994:124.
4. Devine P, Linden JV, Hoffstadter LK, et al. Blood donor-, apheresis-, and transfu-
sion-related activities: results of the 1991 American Association of Blood Banks
Institutional Membership Questionnaire. Transfusion 1993; 33:770782.
5. Devine P, Postoway N, Hoffstadter L, et al. Blood donation and transfusion
practices: the 1990 American Association of Blood Banks Institutional Membership
Questionnaire. Transfusion 1992; 32:683687.
6. Surgenor DM, Wallace EL, Hao SHS, et al. Collection and transfusion of blood in
the United States, 1982-1988. N Engl J Med 1990; 322:16461651.
7. Barrett BB, Andersen JW, Anderson KC. Strategies for the avoidance of bacterial
contamination of blood components. Transfusion 1993; 33:228233.
8. Halpin TJ, Kilker S, Epstein J, Tourault M. Bacterial contamination of platelet
poolsOhio, 1991. MMWR 1992; 41:36)37.
9. Morro JF, Braine HG, Kickler TS, Ness PM, Dick JD, Fuller AK. Septic reactions
to platelet transfusions. A persistent problem. JAMA 1991; 266:555558.
10. Chiu EKW, Yuien KY, Lie AKW, Liang R, Lau YL, Lee ACW, Kwong YL, Wong
S, Ng MH, Chan TK. A prosective study of symptomatic bacteremia following
platelet transfusion and its management. Transfusion 1994; 34:950954.
11. Wallas CH. Transfusion of blood components contaminated with bacteria. Transfu-
sion transmitted infections. In: Smith DM, Dodd RY, eds. Transfusion Transmitted
Infections. Chicago: ASCP Press, 1991:195206.
12. Goldman M, Blajchman MA. Blood product-associated bacterial sepsis. Transfusion
Med Rev 1991; 5(1):7383.
13. BPAC discusses invalidation of test results, platelet usage. Blood Bank Week 1992;
9:13.
14. Rhame FS, Root RK, MacLowry JD, Dadismon TA, Bennett JV. Salmonella
septicemia from platelet transfusions. Study of an outbreak traced to a hemato-

genous carrier of Salmonella cholerae-suis. Ann Intern Med 1973; 78:633641.
15. Heal JM, Jones ME, Forey J, Chaudhry A, Stricof RL. Fatal Salmonella septicemia
after platelet transfusion. Transfusion 1987; 27:25.
16. Arnow PM, Weiss LM, Weil D, Rosen NR. Escherichia coli sepsis from contami-
nated platelet transfusion. Arch Intern Med 1986; 146:321324.
17. Yomtovian R, Lazarus HM, Goodnough LT, Hirschler NV, Morrissey AM, Jacobs
MR. A prospective microbiologic surveillance program to detect and prevent the
transfusion of bacterially contaminated platelets. Transfusion 1993; 33:902909.
18. Buchholz DH, Young VM, Friedman NR, Reilly JA, Mardiney MR. Detection and
quantitation of bacteria in platelet products stored at ambient temperature. Transfu-
sion 1973; 13(5):268275.
19. Halpin TJ, Kilker S, Epstein J, Tourault M. Bacterial contamination of platelet
poolsOhio, 1991. MMWR 1992; 41(3):3637.
20. Gibson T, Norris W. Skin fragments removed by injection needles. Lancet 1958;
2:983985.
21. Lilly HA, Lowbury EJL, Wilkins MD. Limits to progressive reduction to resident
skin bacteria by disinfecion. J Clin Pathol 1979; 32:382385.
22. NIH workshop examines bacterial contamination of blood components. CCBC
Newslett 1995; 13:16.
23. Krishnan LS, Brecher ME. Transfusion transmitted bacterial infection. Hematol/
Oncol Clin North Am 1995; 9:167185.
24. Tipple MA, Bland LA, Murphy MF, et al. Sepsis associated with transfusion of red
cells contaminated with Yersinia enterocolitica. Transfusion 1990; 30:207213.
25. Update: Yersinia enterocolitica bacteremia and endotoxin shock associated with red
blood cell transfusionUnited States, 1987-1988. MMWR 1991; 37:577578.
blood cell transfusionsUnited States, 1991. MMWR 1991; 40:176178.
27. Bjune G, Rund T, Eng J. Bacterial shock due to transfusion with Yersinia en-
terocolitica infected blood. Scand J Infect Diss 1984; 16:411412.
28. Wright DC, Selss IF, Vinton KF, et al. Fatal Yersinia enterocolitica sepsis after blood
transfusion. Arch Pathol Lab Med 1985; 109:10401042.
29. Jacobs J, Jamaer D, Vandeven J, et al. Yersinia enterocolitica in donor blood: a case
report and review. J Clin Microbiol 1989; 27:11191121.
blood cell transfusionUnited States, 1987-1988. MMWR 1988; 37:577578.
31. Hogman CF, Gong J, Hambraeus A, Johansson CS, Eriksson L. The role of white
cells in the transmission of Yersinia enterocolitica in blood products. Transfusion
1992; 32:654657.
32. Pollack C, Straley SC, Klempner MS. Probing the phagolysosomal environment of
human macrophages with a Ca2+-responsive operon fusion in Yersinia pestis.
Nature 1986; 22:834683.
33. Jensenius M, Hoel T, Heier HE. Yersinia enterocolitica septicemia after blood
transfusion. Tidsskr Nor Laegeforen 1995; 115:940942.
34. Hogman CF, Fritz H, Sandberg L. Post-transfusion Serratia marcescens septicemia
(editorial). Transfusion 1993; 33:189191.
35. Heltberg O, Show F, Gerner-Smidt P, Kolmos HJ, Dybkjer E, Gutschik E, Jerne D,

Jepsen OB, Weischer M, Frederiksen W, Sorensen H. Nosocomial epidemic of
Serratia marcescens septicemia ascribed to contaminated blood transfusion bags.
Transfusion 1993; 33:221227.
36. Sazama K. Bacteria in blood for transfusion: a review. Arch Pathol Lab Med 1994;
18:350365.
37. Wagner SJ, Friedman LI, Dodd RY. Transfusion associated sepsis. Clin Microbiol
Rev 1994; 7:290302.
38. Richards C, Kolins J, Trindale CD. Autologous transfusion-transmitted Yersinia
enterocolitica (letter). JAMA 1992; 268:15411542.
39. Sire JM, Michelet C, Mesnard R, et al. Septic shock due to Yersinia enterocolitica
after autologous transfusion (letter). Clin Infect Dis 1993; 17:954955.
40. Duncan KL, Ransley J, Elterman M. Transfusion-transmitted Serratia liquefaciens
from an autologous blood unit (letter). Transfusion 1994; 34:738739.
41. Haditsch M, Binder L, Gabriel C, et al. Yersinia enterocolitica septicemia in
autologous blood transfusion. Transfusion 1994; 34:907909.
42. Braude AI, Williams D, Siemienski J, Murphy R. Shock-like state due to transfusion
of blood contaminated with gram-negative bacilli. Arch Intern Med 1953; 92:7584.
43. Sazama K. Reports of 355 transfusion-associated deaths: 1976-1985. Transfusion
1990; 30:583590.
44. Cookson ST, Arduino MJ, Aguero SM, Jarvis WR, and the Yersinia Study Group.
Yersinia enterocolitica-contaminated red blood cells (RBCs)an emerging threat to
blood safety. 1996; ICAAC.ABS:2352.
45. Rhame FS, McCullough JJ, Cameron S, et al. Pseudomonas cepacia infections
caused by thawing cryoprecipitate in a contaminated water bath (abstr). Transfusion
1979; 19:653.
46. Casewell MW, Slater NGP, Cooper JE. Operating theatre water-baths as a cause of
Pseudomonas septicemia. J Hosp Infect 1981; 2:237240.
47. Albumin recall. Lancet 1996; 348:1026.
48. Sgouris JT, Rene A, eds. Proceedings of the Workshop on Albumin, 1975. Bethesda,
MD: National Heart and Lung Institute, National Institutues of Health, Department
of Health, Education and Welfare, 1976:284.
49. Recalls issued for Centeon products. AABB Regulatory Update, October 1996; 3:6.
50. Centeon issues world wide recall of albumin and PPF. Americas Blood Centers
Newslett 1996; 39:12.
51. Food and Drug Administration: Blood Products Advisory Committee, Center for
Biologist Evaluation and Research, 36th Meeting, Rockville, MD, May 29, 1992,
Publication No. 92-026824, pp. 440445.
52. Walter CW, Kundsin RB, Button LN. New technique for detection of bacterial
contamination in a blood bank using plastic equipment. N Engl Med 1957; 257:364
369.
53. Barrett BB, Andersen JW, Anderson KC. Strategies for the avoidance of bacterial
contamination of blood components. Transfusion 1993; 33:228233.
54. Yomtovian R, Lazarus HM, Goodnough LT, et al. A prospective microbiologic
surveillance program to detect and prevent the transmission of bacterially contami-
nated platelets. Transfusion 1993; 33:902909.
55. Arduino MJ, Bland LA, Tipple MA, et al. Growth and endotoxin production of
Yersinia enterocolitica and Enterobacter agglomerans in packed erythrocytes. J Clin
Microbiol 1989; 27:14831485.
56. Kim DM, Brecher ME, Bland LA, et al. Prestorage removal of Yersinia entercolitica
from red cells with white cell-reduction filters. Transfusion 1992; 32:658662.
transfusion medicine. In: Nance SJ, ed. Blood safety: Current Challenges. Bethesda,
MD: American Association of Blood Banks, 1992:213228.
58. Myhre BA, Nakusako YY, Schott R. A rapid method for determining the sterility of
frozen-reconstituted blood. Transfusion 1977; 17:195202.
59. Arduino MJ, Bland LA, Tipple MA, et al. Growth and endotoxin production of
Yersinia enterocolitica and Enterobacter agglomerans in packed erythrocytes. J Clin
Microbiol 1989; 27:14831485.
60. Brecher NE, Hogan JJ, Boothe G, et al. The use of a chemiluminescence-linked
universal bacterial ribosomal RNA gene probe and blood gas analysis for the rapid
detection of bacterial contamination in white cell reduced and nonreduced platelets.
Transfusion 1993; 33:450457.
61. Brecher ME, Hogan JJ, Boothe G, et al. Platelet bacterial contamination and the use
of a chemiluminescence-linked universal bacterial ribosomal RNA gene probe.
Transfusion 1994; 34:750755.
62. Sullivan A. Machines speed testing of food for bacteria, Wall Street 1996; (July 15):
63. Feng P, Keasler SP, Hill WE. Direct identification of Yersinia enterocolitica in blood
by polymerase chain reaction amplification. Transfusion 1992; 32:850854.
64. Wagner SJ, Robinette D. Evaluation of swirling, pH, and glucose tests for the
detection of bacterial contamination of platelet concentrates. Transfusion 1996;
36:989993.
65. Burstain JM, Brecher ME, Workman K, Foster M, Faber GH, Mair D. Rapid
identification of bacterially contaminated platelets using reagent strips. Transfusion
1997;
66. Myhre BA, Demainew SH, Yoshimori RN, Nelson E, Carmen RA. pH changes
caused by bacterial growth in contaminated platelet concentrates. Ann Clin Lab Sci
1985; 15:509514.
67. Arpi M, Bremmelgaard A, Abel Y, Olsson K, Hansen L. A novel screening method
for the detection of microbial contamination of platelet concentrates (letter). Vox
Sang 1993; 65:335336.
68. Arpi M, Bremmelgaard A, Abel Y, Olsson K, Hansen L. A novel screening method
for the detection of microbial contamination of platelet concentrates. An experimen-
tal pilot study. Vox Sang 1993; 65:335336.
69. Hogman CF, Gong J. Studies of one invasive and two noninvasive methods for
detection of bacterial contamination of platelet concentrates. Vox Sang 1994;
67:351355.
70. Kim DM, Estes TJ, Brecher ME et al. WBC filtration, blood gas analysis and plasma
hemoglobin in Yersinia enterocolitica contaminated red cells. Transfusion 1992;
32:41S.
71. Kim DM, Brecher ME, Bland LA, Estes TJ, Carmen RA, Nelson EJ. Visual
identification of bacterially contaminated red cells. Transfusion 1992; 32:221225.
72. Myre BA, Walker LJ, White ML. Bacteriocidal properties of platelet concentrates.
Transfusion 1974; 14:116123.
73. Murphy S, Kahn RA, Holme S, et al. Improved storage of platelets for transfusion
in a new container. Blood 1982; 60:194200.
74. Holme S, Murphy S. Quantitative measurements of platelet shape by light transmis-
sion studies: application to storage of platelets for transfusion. J Lab Clin Med 1978;
92:5364.
75. Holme S, Murphy S. Platelet storage at 22C for transfusion: interrelationship of
platelet density and size, medium pH, and viability after in vivo infusion. J Lab Clin
Med 1983; 101:161174.
76. Lindberg JE, Slichter SJ, Murphy S, et al. In vitro function and in vivo viability of
stored platelet concentrates. Effect of a secondary plasticizer component of PVC
storage bags. Transfusion 1983; 23:294299.
77. Murphy S, Sayar SN, Gardner FH. Storage of platelet concentrates at 22C. Blood
1970; 35:549557.
78. White JG, Krivit W. Ultrastructural basis for shape change induced in platelets by
chilling. Blood 1967; 30:625635.
79. Murphy S, Gardner FH. Platelet preservation. Effect of storage temperature on
maintenance of platelet viabilitydeleterious effect of refrigerated storage. N Engl
J Med 1969; 280:10941098.
80. Holme S, Vaidja K, Murphy S. Platelet storage at 22C: effect of type of agitation
on morphology, viability, and function in vitro. Blood 1978; 52:425435.
81. Bertolini F, Murphy S. A multicenter inspection of the swirling phenomenon in
platelet concentrates prepared in routine practice. Transfusion 1996; 36:128132.
82. Grossman BJ, Kollins P, Lau PM, et al. Screening blood donors for gastrointestinal
illness: a strategy to eliminate carriers of Yersinia entercolitica. Transfusion 1991;
31:500501.
83. Katz L, MacPherson JL, Zuck TF. Yersinia and blood donation (letter). Transfusion
1992; 32:191.
84. Hoogkamp-Korstanje JA, de Koning J, Samsom JP. Incidence of human infection
with Yersinia enterocolitica serotypes 03, 08 and 09 and the use of indirect
immunofluorescence in diagnosis. J Infect Dis 1986; 153:138141.
85. Marks MI, Pai CH, Lafleur L, Lackman L, Hammerberg O. Yersinia enterocolitica
gastroenteritis: a prospective study of clinical, bacteriologic, and epidemiologic
features. J Pediatr 1980; 96:2631.
86. Mingrone MG, Fantasia M, Figura N, Guglielmetti P. Characteristics of Yersinia
enterocolitica isolated form children with diarrhea in Italy. J Clin Microbiol 1987;
25:13011304.
87. Olthuis H, Putlaert C, Verhagen C, Valk L. Method for removal of contaminating
bacteria during venapuncture (abstract). Int Soc Blood Transf Regional Congress
1995; 77.
88. Figueroa PI, Yoshimori R, Nelson E, Macabasco G, Carmen R. Distribution of
bacteria in fluid passing through an inoculated collection needle (abstr). Transfusion
1995; 35(suppl):11S.
89. Bacterial detection device. Blood Bank Week 1995; (Sept. 29):2.
90. Stenhouse MAE, Milner LV. Yersinia enterocolitica: a hazard in blood transfusion.
Transfusion 1982; 22:396398.
91. Van Noyen R, Vandepitte J, Wauters G, Selderslaghs R. Yersinia enterocolitica: its
isolation by cold enrichment from patients and healthy subjects. J Clin Pathol 1981;
34:10521056.
92. Lin L, Wiesehahn GP, Morel PA, et al. The use of 8-methoxypsoralen and long-
wavelength ultraviolet radiation for decontamination of platelet concentrates. Blood
1989; 74:515525.
93. Matthews JL, Newman JT, Sogandares-Bernal F, et al. Photodynamic therapy of
viral contaminants with potential for blood banking applications. Transfusion 1988;
28:8185.
94. Pamphilon DH, Corbin SA, Saunders J, Tandy NP. Applications of ultraviolet light
in the preparation of platelet concentrates. Transfusion 1989; 29:379383.
95. Prince AM, Stephan W, Brotman B. B-propiolactone/ultraviolet irradiation; a review
of its effectiveness for inactivation of viruses in blood derivatives. Rev Infect Dis
1983; 5:92107.
96. Lin L, Wiesehahn GP, Morel PA, Corash L. Use of 8-methoxypsoralen and
long-wavelength ultraviolet radiation for decontamination of platelet concentrates.
Blood 1989; 74:517525.
97. Johnson RB, Napychank PA, Murphy S, Snyder EL. In vitro changes in platelet
function and metabolism following increasing doses of ultraviolet-B irradiation.
Transfusion 1993; 33:249255.
98. FDA committee endorses education and research to combat rare bacterial reaction:
rejects operational changes for now. Council Commun Blood Centers (CCBC)
Newslett 1991; (May 10):14.
99. FDA Blood Products Advisory Committee supports educational efforts on Yersinia
enterocolitica. Blood Bank Week 1991; 20:13.
100. Kostreski F. Dont change storage rules, blood bankers warn. CAP Today 1991; 5,
45.
101. Donegan E, Lenes BA, Tomasulo PA, et al. Transmission of HIV-1 by component
type and duration of shelf storage before transfusion (letter). Transfusion 1991;
30:851852.
102. Braine HG, Kickler TS, Charache P, et al. Bacterial sepsis secondary to platelet
transfusion: an adverse effect of extended storage at room temperature. Transfusion
1986; 26:391393.
103. Food and Drug Administration: Blood Products Advisory Committee, Center for
Biologist Evaluation and Research, 36th Meeting, Rockville, MD, May 29, 1992,
Publication No. 92-026824, p. 433.
104. Buchholz DH, AuBuchon JP, Snyder EL, et al. Removal of Yersinia enterocolitica
from AS-1 Red cells. Transfusion 1992; 32:667672.
105. Gibb AP, Martin Km, Davidson GA, et al. Modeling the growth of Yersinia
enterocolitica in donated blood. Transfusion 1004; 34:304310.
106. Pietersz RNI, Reesink HW, Pauw W, et al. Prevention of Yersinia enterocolitica
growth in red blood cell concentrates. Lancet 1992; 340:755756.
107. Pietersz RNI, Reesink HW, Dekker MA, et al. Elimination of Yersinia enterocolitica
by a 20h hold of whole blood and removal of leukocytes by filtration (abstr).

108. Lazarus HM, Herzig RH, Warm SE, Fishman DJ. Transfusion experience with
platelet concentrates stored for 24 to 72 hours at 22C: importance of storage time.
109. Punsalang A, Heal JM, Murphy PJ. Growth of gram-positive and gram-negative
bacteria in platelet concentrates. Transfusion 1989; 29:596599.
110. Connor J, Currie LM, Allan H, Livesey SA. Recovery of in vitro functional activity
of platelet concentrates stored at 4C and treated with second-messenger effectors.
Transfusion 1996; 36:691698.
111. Blajchman MA. The effect of leukodepletion on allogenic donor platelet survival
and refractoriness in an animal model. Semin Hematol 1991; 28(suppl):1417.
112. Brecher ME, Pineda AA, Torloni AS, et al. Prestorage leukocyte depletion: effect
on leukocyte and platelet metabolities, erythrocyte lysis, metabolism and in vivo
survival. Semin Hematol 1991; 28(suppl):39
113. Wenz B, Burns ER, Freundlich LF. Prevention of growth of Yersinia enterocolitica
in blood by polyester fiber filtration. Transfusion 1992; 32:663666.
114. Kim DM, Estes TJ, Brecher ME, et al. WBC filtration, blood gas analysis and
plasma hemoglobin in Yersinia enterocolitica contaminated red cells. Transfusion
1992; 32:41S.
115. Hogman CF, Engstrand L. Factors affecting growth of Yersinia enterocolitica in
cellular blood products. Transfus Med Rev 1996; 10:259275.
116. Goldman M, Delage G. The role of leukodepletion in the control of transfusion-
transmitted disease. Transfus Med Rev 1995; 9:919.
117. AuBuchon JP, Pickard C. White cell reduction and bacterial proliferation (letter).
Transfusion 1993; 33:533534.
118. Sherburne B, McCullough A, Dzik WH, et al. Bacterial proliferation in platelet con-
centrates is unaffected by pre-storage leukocyte depletion. Blood 1991; 78(suppl):350a.
119. Wenz B, Ciavarella D, Freundlich L. Effect of prestorage white cell reduction on
bacterial growth in platelet concentrates. Transfusion 1993; 33:520523.
120. Smith DA, Monaghan WP, Orcult RM et al. A presealed overwrap method of
protecting frozen blood components during water immersion thawing. Transfusion
1981; 21:447449.
121. Weber S, Lottspeich F, Kohl J. An epitope of elongation factor Tu is widely
distributed within the bacterial and archaeal domains. J Bacteriol 1995; 177:1119.
122. Seidl S. Syphilis screening in the 1990s. Transfusion 1990; 30:773774.
123. Soendjojo A, Boedisantoso M, Ilias MI, et al. Syphilis demblee due to a blood
transfusion. Br J Veneral Dis 1982; 58:149150.
124. Risseeuw-Appel IM, Kothe FC. Transfusion syphilis: a case report. Sex Transm Dis
1983; 10:200201.
125. Chambers RW, Foley HT, Schmidt PJ. Transfusion of syphilis by fresh blood
components. Transfusion 1969; 9:3234.
126. Van der Sluis JJ, Onvlee PC, Kothe FCHA, et al. Transfusion syphilis, survival of
Treponema pallidum in donor blood I. Report of an orientating study. Vox Sang
1984; 47:197204.
127. Van der Sluis JJ, ten Kate FJW, Vuzevski VD, et al. Transfusion syphilis, survival
of Treponema pallidum in donor blood II. Dose dependence of experimentally

determined survival times. Vox Sang 1985; 49:390399.
128. Quinn TC, Cannon RO, Glasser D, et al. The association of syphilis with risk of
human immunodeficiency virus infection in patients attending sexually transmitted
dissease clinics. Arch Intern Med 1990; 150:12971302.
129. Nelson KE, Vlahov D, Cohn S, et al. Sexually transmitted diseases in a population
of intravenous drug users: association with seropositivity to the human immuno-
deficiency virus (HIV). J Infect Dis 1991; 164:457463.
130. Potterat JJ. Does syphilis facilitate sexual acquisition of HIV? (letter). JAMA 1987;
258:473.
131. Otten MW Jr, Zaidi AA, Peterman TA, et al. High rate of HIV seroconversion among
patients attending urban sexually transmitted disease clinics. AIDS 1994; 8:549
553.
132. Rosenblum L, Darrow W, Witte J, et al. Sexual practices in the transmission of
hepatitis B virus and prevalence of hepatitis delta virus infection in female prosti-
tutes in the United States. JAMA 1992; 267:24772481.
133. Thomas DL, Cannon RO, Shapiro CN, et al. Hepatitis C, hepatitis B, and human
immunodeficiency virus infections among non-intravenous drug-using patients
attending clinics for sexually transmitted diseases. J Infect Dis 1994; 169:990995.
134. McAllister SK, Bland LA, Arduino MJ, et al. Patient cytokine response in transfu-
sion-associated sepsis. Infect Immun 1994; 62:21262128.
135. Leveton LB, Sox HC, Stoto MA. HIV and blood supply: an analysis of crisis
decisionmaking, executive summary. Transfusion 1996; 36:919927.
11
HIV and Blood Transfusion
Celso Bianco and Maria Rios

I. INTRODUCTION
Human immunodeficiency virus (HIV) transmission by transfusion of blood and

blood products has been the biggest driver of change in the history of transfusion
medicine. Most of the medical world in the 1970s and the early 1980s was
celebrating the control of infectious diseases. The hemophilia community was
celebrating the enjoyment of a normal life provided by the use of plasma-derived
clotting factor concentrates. In this environment of excitement and trust in the
power of medical science and pharmaceuticals, the initial reactions to the news
that the acquired immunodeficiency syndrome (AIDS) could possibly be trans-
mitted by transfusion were denial and disbelief. The first report suggesting that
clotting factor concentrates might transmit HIV was published in July 1982 (1).
It described three hemophilia patients who developed immunosuppression and
opportunistic infections. By December 1992, four more cases had been identified
among patients with hemophilia A (2). The first case of suspected transmission
of AIDS by blood transfusion was also reported at that timea 20-month-old
child who developed AIDS after receiving multiple transfusions, including a
transfusion of platelets derived from blood donated by a male who subsequently
was found to have AIDS (3). The findings about AIDS and blood transfusions
prior to the identification of the etiological agent were reviewed intensively (4,5),
and both the medical community and the patient community had no choice but to
accept the growing evidence that AIDS was transmitted by a bloodborne infec-
tious agent (6).
Assays for blood donor screening for HIV were licensed by the U. S. Food
and Drug Administration (FDA) in March 1985 and led to a remarkable reduction
251
252 Bianco and Rios
in the transmission of HIV by transfusion. The development of viral inactivation

technologies applicable to plasma derivatives stemmed the spread of AIDS
through the large pools required for the manufacture of clotting factor concen-
trates. Unfortunately, the development was not smooth and six hemophilia
patients previously seronegative for HIV seroconverted between September 1986
and September 1987. None had risk factors for HIV infection other than hemo-
philia. The implicated lots had been produced from plasma that had been screened
for HIV and then heated at 60C for 30 hours in the lyophilized state. Subse-
quently, concentrates produced by this process were removed from distribution
(7). A similar event was identified by the Transfusion Safety Study (8).
More recently the Institute of Medicine of the National Academy of
Sciences reviewed the history of the first few years of AIDS and transfusion in a
landmark report. This report concluded that the AIDS epidemic was mismanaged
by the blood-collecting agencies, professional organizations, hemophilia organi-
zations, and the federal government. The report was vehemently criticized by the
blood banking community for judging history in hindsight and for acceptance
of allegations and opinions as facts without critical examination and without
placement in the context of contemporary knowledge (9).
The tragedy of the first few years led to the development of successful
programs for the prevention of transmission of AIDS by transfusion that included
donor education, adequate medical history, physical examination, and mecha-
nisms that allowed exclusion of individuals at risk in a confidential manner (10).
The current screening tests for HIV antibodies are highly sensitive, and the
Western blot confirmatory test is highly specific. Finally, additional screening
tests capable of detecting antibodies to HIV variants, HIV-1 p24 antigen, and HIV
RNA in the plasma of infected individuals have made blood transfusion extremely
safe. In addition, all plasma derivatives are virally inactivated by chemical and
physical methods and therefore do not transmit lipid-enveloped viruses such as
HIV, hepatitis B virus (HBC), hepatitis C virus (HCV), and human T-cell
lymphotropic virus (HTLV). For instance, the solvent detergent process was
licensed in 1985 and became extremely successful (11).
II. THE VIRUS

The first retroviruses associated with human disease were HTLV-I and HTLV-II,
discovered in 1980 and 1982, respectively (these viruses are the subject of another
chapter in this book). Human immunodeficiency virus type 1 (HIV-1, formerly
called LAV and HTLV-III) was first recognized in 1983 and identified as a
lentivirus, the most complex group of retroviruses. In 1985 another virus closely
related to HIV-1 was characterized and named HIV-2. Soon thereafter the simian
HIV and Blood Transfusion 253
immunodeficiency virus (SIV) was isolated from a captive macaque monkey.

The biology and structure of HIV-1 and HIV-2 have been extensively reviewed
(12,13). The genome of retroviruses is contained in two strands of RNA that must
be transcribed into complementary DNA (cDNA) and integrate into the host viral
genome for completion of its life cycle. All have a characteristic genome
arrangement: they have a gag (group antigen) gene, a pol (polymerase) gene, and
an env (envelope) gene flanked by two long terminal repeats (LTR). In addition,
they have genes that regulate proviral gene expression both temporally and
quantitatively. HIV contains more than six genes not found in other retroviruses.
The tat gene encodes for the TAT protein. TAT binds to transactivating transcrip-
tion elements and induces the synthesis of proviral RNA transcripts that encode
for structural proteins of the virus. The rev gene codes for the REV protein, which
increases expression of structural proteins. HIV proviral gene expression utilizes
mRNA splicing to generate several viral proteins. In the absence of REV, these
mRNA are only spliced into those that encode for regulatory proteins. The nef
gene encodes for the NEF protein. NEF-negative strains replicate more rapidly
than NEF-positive strains. Two other genes, vif and vpu, are required for virion
assembly and maturation. Finally, a second transactivator gene supplementing tat
is vpr. Its product, the VPR protein, accelerates the rate of production of viral
proteins. Reduction of TAT levels downregulates overall viral gene expression.
The balance between TAT and REV is the major event determining the viral
latency/expression switch and subsequent disease development. HIV-2 has all the
regulatory genes present in HIV-1 but differs from it in genomic organization.
HIV-2 infection predominates in West Africa, and there is evidence that it is
less virulent.
The most important HIV proteins are briefly described in Table 1. Figure 1
shows the HIV-1 genes and corresponding proteins.
III. GENETIC VARIATION

Genetic variation is the hallmark of the HIV infection. Diversity leads to the
existence of major groups (Group M, or main, and Group O, outlier). The Group
M viruses can be divided into at least eight distinct clades (A through H). Clades
differ among themselves in 2030% of their amino acid sequences of the gp120
protein. Isolates within clades (and within the same individuals) vary by 520%
and are called quasispecies. Diversity can represent a problem for diagnostic tests.
This issue is further discussed in the testing section of this chapter.
Diversity enables the virus to escape surveillance by the immune system.
The virus establishes a persistent infection without being cleared by the immune
system and induces immunodeficiency that allows its survival. Most of the
254 Bianco and Rios
Table 1 Major Genes and Proteins of HIV-1 and HIV-2
Virus type
Gene HIV-1 HIV-2
gag
Precursor p55 p56
Core p24 p26
Matrix p17 p16
Nucleocapsid p9
Nucleocapsid p7
pol
Reverse transcriptase p66, p51 p68, p53
Endonuclease p31 p34
env
Precursor gp160 gp140
Surface gp120 gp105 (125)
Trans-membrane gp41 gp36 (41)
Regulatory
Vif p23
Vpr p15
Tat p14
Rev p19
Vpu p16
Nef p27
diversity in HIV is generated during the reverse transcription of RNA into cDNA.
The HIV reverse transcriptase (RT) lacks 3 exonuclease proofreading activity
allowing the occurrence of errors during the transcription process. The overall
mutation rate is very high, in the order of 3.4 105 per base per replication cycle.
The degree of genetic variation observed in HIV infection is phenomenalup to
20% within an infected individual (14,15). This high degree of diversity is
particularly seen in patients receiving antiprotease therapy, reaching more than
7% by week 60 (16,17). Diversity is observed even within the same tissue. HIV-1
genomes infecting different regions of the brain of one study subject with HIV
encephalitis (HIVE) had a mosaic structure, being assembled from different
combinations of evolutionarily distinct lineages in p17 (gag), pol, individual
hypervariable regions of gp 120 (V1/V2, V3, V4, and V5), and gp41/nef (18).
Antigenic stimulation influences the dynamics of HIV replication, including the
relative expression of different HIV variants (19). Naturally occurring recombi-
nant HIV strains have been found in infected patients in regions of the world
where multiple genotypic variants coexist (20).
HIV Genome
1000 vpu 8000
gag vif tat
nef
5LTR pol vpr env 3LTR
rev
gag precursor
p15
VPR env precursor
p55 p23
HIV and Blood Transfusion
MA p15 VIF p16 gp160

p17 VPU
CA
p24
NC NC SU NEF
p7 P9 gp120 gp41
Enzyme precursors p19 REV TAT

p66 p51 p31
RT
p66
p51 p31
Figure 1 Graphic representation of the HIV genome and respective gene products.
5LTR- gag-pol-vpr-vpu-env-tat-rev-nef-LRT3 HIV-1 gene productsthe gag (group anti-
gen) region encodes for the structural proteins: p17 (matrix), p24 (capsid), and p15 (p7, p9
nucleocapsid). The pol region encodes for the nonstructural proteins protease (physically
part of the pol open reading frame), p31 (integrase), and p66/p51 (heterodimer reverse
transcriptase). The env (envelope) region encodes for the envelope proteins gp120 (exposed
to the external surface of the virus) and for gp41 (envelope trans-membrane). The genes vpr,
255
tat, rev, and nef encode for p15, p14, p19, and p27, respectively, these are regulatory proteins
involved in viral replication. The genes vif and vpu encode for p23 and p16, respectively.
256 Bianco and Rios
IV. HIV-1 (HIV M AND HIV O) AND HIV-2

The strains of HIV-1 that have caused the worldwide pandemic of AIDS have
been designated as group M viruses. Another group of HIV-1 viruses character-
ized by extensive genetic divergence from group M strains have been identified
recently and classified as group O viruses. Group O viruses are rare but have been
reported in patients from West and Central Africa (Cameroon, Gabon, Niger,
Nigeria, Senegal, and Togo), nationals of these countries living in Europe, and
one French national (21). The first case of Group O infection in the United States
was identified in April 1996 in a woman who had come to the United States from
Africa (22). Enzyme immunoassay (EIA) kits commercially available in the
United States do not consistently detect antibodies elicited by infection with
HIV-1 group O strains. For this reason, until FDA-licensed assays are able detect
Group O variants, individuals who have emigrated from West and Central Africa
are deferred from donating blood and tissue. Tests capable of detecting HIV
Group O variants have been licensed and in use in European countries for several
years (23).
HIV is found worldwide, with distribution ranging from 0.01 to 5.6% of the
population. The prevalence may be even higher in some African countries.
The prevalence of HIV-1 infection among blood donors in the United States is
approximately 1:20,000. The prevalence of HIV infection among the general
population has been estimated at 1:250. Currently, 33 U.S. states require reporting
of HIV infection in addition to AIDS. It is expected that a more precise measure
of incidence with increased reporting will allow better monitoring of the effect of
the epidemic because of the therapies that have reduced AIDS incidence.
The major modes of transmission are either homosexual or heterosexual
intercourse, parenteral transmission through needle sharing among injecting drug
users (IDU), and through transfusion of contaminated blood or blood products.
HIV is also transmitted from mother to infant, either intrapartum, perinatally, or
via breast-feeding. By December 1998, 679,739 cases of AIDS had been reported
to CDC since the beginning of the epidemic. Of these, 48% occurred among men
who had sex with men and 26% among injection drug users. Men who had sex
with men and also injected drugs constituted 10% of the reported cases. Ten
percent of the cases occurred in heterosexual contacts of individuals at risk (sex
with an injection drug user, a bisexual male, a transfusion recipient, or a person
with hemophilia) (24).
HIV is present in seminal fluid, cervical mucus, and vaginal fluid. There is
a stronger association of efficacy of HIV transmission with receptive anal
intercourse, but the virus can also be transmitted during vaginal intercourse.
The probability of a woman becoming infected by an HIV-positive male partner
during vaginal intercourse is estimated at 0.2% per encounter, and probability is
even lower for infection from a woman to a man in vaginal intercourse.
Maternal transmission of HIV accounts for over 90% of HIV infections in

infants and children. The efficiency of transmission from mother to child ranges
from 13 to 42% in the absence of intervention (HIV treatment). Between 50 and
70% of the mother-to-child transmissions occur late in pregnancy or during birth,
when maternal blood may enter the fetal circulation, or by neonatal mucosal
exposure to the virus during labor or delivery. Breast-feeding increases the risk
of HIV transmission by 14% over the risk of HIV infection during pregnancy or
delivery (13).
V. HIV AND DISEASE

Because CD4+ T cells and macrophages are the only cells found to be infected
consistently in vivo, these are considered critical for viral replication. However,
HIV has been found in B cells, natural killer (NK) cells, CD8+ T cells, peripheral
and follicular dendritic cells, eosinophils, precursors of CD4+ bone marrow cells,
Langerhans cells, megakariocytes, astrocytes, oligodendrocytes, renal epithelial
cells, cervical cells, gastrointestinal epithelial cells, and cells from organs such as
liver, lungs, salivary glands, eyes, prostate, testes, and adrenal glands (13).
The CD4 molecule is the primary receptor for HIV, and essentially any cells that
express this molecule, even in a transient stage, may be susceptible to infection.
The viral gp 120 binding CD4 is the first step in viral infection. That CD4 binding
is critical, but not sufficient, for viral infection is clearly demonstrated by the
fact that human brain and skin cells expressing the CD4 molecule bind to
HIV envelope but their membranes fail to fuse. This observation indicates a
requirement for other components (co-receptors) for proper fusion and conse-
quent cell infection.
The two most important co-receptors are the chemokine receptors CXCR4
(receptor for SDF-1) and CCR5 (receptor for MIP-1, MIP-, and RANTES).
The chemokines that bind to these receptors can block HIV infection, and
natural resistance to infection is observed in individuals who have defect in
the gene encoding for the CCR5 (25,26). Mutant CCR5 co-receptors can con-
fer resistance to HIV-1 infection. p120 binds to CD4 (T4) on T lymphocytes
and other cells and leads to infection. Co-receptors on T cells (fusin) and
macrophages (CCR5) also play a role in the binding and penetration of the virus.
A polymorphism in the CCR5 gene (with a 32 bp deletion) reduces the ability of
HIV to enter cells, slows decline in CD4 cells, and leads to a reduced (>100)
viral load. About 50% of long-term survivors of HIV are heterozygous for CCR5
deletion (27). HIV resistance has also been observed in individuals with a
combination of two separate CCR5 mutations (28). The distribution of the
homozygous deleted CCR5 genotype among 566 persons with hemophilia and
97 transfusion recipients known to have been exposed to HIV indicated that the
258 Bianco and Rios
lack of CCR5 expression protected persons from infection. Only individuals who
are homozygous for CCR5 are protected, and there is no difference in the rate of
progression to AIDS between infected heterozygous and homozygous wild-type
subjects (29).
VI. CELL TROPISM AND CONSEQUENCE OF INFECTION

The tropism of the virus for monocyte/macrophage (M-tropic) or CD4+ T cell
(T-tropic) is determined by the co-receptor used for entry. M-tropic virus can only
use CCR5, and T-tropic virus uses CXCR4. There are isolates capable of using
both CCR5 and CXCR4. In the course of infection of CD4+ T cells, virions are
released almost exclusively to the plasma surrounding the cells and there is
extensive cell death. In the course of infection of monocytes and macrophages,
the virions are often released into intracytoplasmic vacuoles and there is less cell
death. The tropism for macrophages or CD4+ T cells seems to be determined by
changes in env genes. Other CD4+ cells like microglia and follicular dendritic
cells can also be infected and function as anatomical reservoirs, protecting the
virus from antiretroviral drugs.
VII. PATHOGENY
The evidence demonstrating that HIV-1 and HIV-2 cause AIDS is overwhelming.
This evidence has been contested by a small number of prominent scientists (30),
but this opinion has been totally rejected by others on the basis of availability of
sound scientific data.
Primary HIV infection is followed by a retroviral syndrome that develops
in 4090% of cases (31). Usual symptoms are flu-like and include fever,
fatigue, pharyngitis, weight loss, myalgias, headache, and nausea. About 50% of
the patients present lymphadenopathy and night sweats. These symptoms disap-
pear after a few weeks. During this period, HIV-1 RNA concentrations vary
widely (104106 molecules/mL). Because of the difficulty in identifying the date
of exposure to HIV-1 infection in persons other than transfusion recipients,
studies of the incubation periods for AIDS have been limited. One study based
on a cohort of 84 homosexual and bisexual men showed that the maximum
likelihood estimate for the proportion of infected homosexual men developing
AIDS was 0.99 (90% CI, 0.381). Furthermore, the maximum likelihood estimate
for the mean incubation period for AIDS in homosexual men was 7.8 years (90%
CI, 4.215.0 years), which was close to the estimate of 8.2 years for adults
developing transfusion-associated AIDS (32). European data on transfusion-
associated (TA) AIDS cases reported in 1989 estimated the median incubation
period in adults from 6.5 to 11 years (33).
HIV infection results in progressive elimination of helper T lymphocytes

with consequent immunodeficiency. Most individuals remain asymptomatic for
several years prior to development of AIDS. The definition of AIDS is based on
the presence of low CD4+ counts (CD4 < 200/L or CD4+ T cells < 14% of total
lymphocytes) or opportunistic infections. The normal concentration of CD4+ T
cells ranges between 800 and 1400/L. AIDS-defining illnesses include tuberculo-
sis, atypical mycobacteria, fungal infection, cytomegalovirus retinitis, (CMV)
Pneumocystis carinii pneumonia, oral and esophageal candidiasis, Herpes sim-
plex infection, cryptosporidia, isospora, cryptococcal meningitis, toxoplasmosis,
Kaposis sarcoma, and non-Hodgkin lymphomas. A number of neurological
diseases also occur. Among them are encephalitis, aseptic meningitis, progressive
multifocal leukoencephalopathy (PML), central nervous system (CNS) lym-
phoma, and AIDS dementia. Thrombocytopenia is very common in late HIV and
responds poorly to usual therapy. Parvovirus B19 may be an important infection
in AIDS patients. However, aggressive antiretroviral treatment may effectively
diminish transfusion requirements among HIV-infected individuals with pure red
blood cell (RBS) aplasia resulting from parvovirus B19 infection (34). The course
of HIV infection and progression to AIDS has been extensively reviewed (13).
Until the recent past the prognosis of AIDS was bleak, and death usually occurred
within 812 years of diagnosis. Recent therapeutic advances have changed this
picture entirely, with substantial reduction in mortality. HIV mortality has de-
creased from 29.4 to 8.8 per 100 person-years from 1995 to 1997 (35).
The hypothesis that blood transfusions could directly accelerate progression
to AIDS through activation of HIV-1 expression and/or transfusion-related im-
munosuppression was proposed several years ago. In essence, allogeneic leuko-
cytes would present an antigenic challenge to HIV-infected mononuclear cells,
which would then proliferate and increase viral production. In vitro incubation of
HIV-infected mononuclear cells with allogeneic blood components (mainly leu-
kocytes) results in increased viral replication. However, preliminary data suggest
that leukocyte reduction prior to RBC transfusion does not alter viral replication
compared with standard packed RBC transfusions (33)
VIII. HIV TRANSMISSION BY TRANSFUSION OF BLOOD

AND BLOOD PRODUCTS
Most recipients of HIV-infected blood become seropositive (37). HIV transmis-
sion by transfusion was also recognized through testing of repositories in clinical
laboratories. For instance, 18 of 211 patients with leukemia who had been
multiply transfused before the availability of screening for HIV antibody were
found to be seropositive for HIV (38). Life-table analysis of 116 seropositive
recipients suggested that AIDS develops in 49% (95% CI, 3662%) within 7
260 Bianco and Rios
years after infection (39). By December 1998 there were 8,760 cases of AIDS
among recipients of blood, blood components, or tissue and 5,145 among patients
with hemophilia or other coagulation disorders among the 688,200 cases of AIDS
reported to CDC (24). This number is somewhat smaller than that predicted by
modeling studies suggesting a total of 15,000 eventual cases of AIDS attributable
to infection by blood transfusion prior to July 1985 (40,41). The incidence of
transfusion-associated HIV-1 infection in San Francisco was estimated to have
risen rapidly from the first occurrence in 1978 to a peak in late 1982 of
approximately 1.1% per transfused unit. The decrease after 1982 coincided with
the implementation of high-risk donor deferral measures. It is estimated that,
overall, approximately 2,135 transfusion recipients were infected with HIV-1 in
the San Francisco region alone (37). By December 1998, 39 cases of AIDS had
been reported among recipients of blood that had screened negative for HIV
antibody (transfused after introduction of screening tests in 1985). The number
of transfusion-associated AIDS cases will certainly be affected by the newer HIV
therapies. Accurate data about transmission will not be available until reporting
of HIV infection (instead of AIDS) is required for the entire United States.
One of the most important sources of information about HIV transmission
by transfusion was a repository of approximately 200,000 sera from blood donors
established in late 1984 and early 1985 by the Transfusion Safety Study (TSS),
which was sponsored by the National Heart, Lung, and Blood Institute (NHLBI).
Collections were made in the four metropolitan areas with the highest prevalence
of AIDS prior to the availability of screening tests for antibodies to HIV in March
1985. Retrospective testing of this repository showed an overall anti-HIV-1
prevalence of 16 cases per 10,000 donations (42). The vast majority of individuals
who received transfusion of blood from seropositive donors became positive for
antibodies to HIV, and many developed AIDS. This was documented by extensive
lookback studies. Essentially, a seropositive donation triggered a search for
records of all prior donations made by that donor and notification of recipients
and testing whenever possible. One of the earliest studies identified seropositive
donors among the units transfused to each of 19 patients who developed AIDS
and were seropositive for HIV (43). HIV-1 was isolated from the blood of blood
donors and HIV-positive blood recipients for months after the transfusion event
clear documentation that HIV viremia was persistent (44). In one rather informa-
tive case, blood components harvested from a single, asymptomatic, seropositive
donor were transfused into a group of cancer patients. Of 10 living recipients,
9 had antibodies to the virus. Cultures for HIV were positive in 7 of the 9
seropositive recipients. Six seropositive recipients had developed immunological
and clinical sequelae of HIV infection (45). A similar case of transmission
occurred in a bone marrow transplant recipient (46).
In the study of a cohort of pediatric patients in a large private metropolitan
hospital, of the 775 children identified as having received transfusions during the
project period, 644 (83%) were located, and 443 (69%) were evaluated for HIV-1
infection. Among those evaluated, 33 (7%) had antibody to HIV-1 (47).
IX. LOOKBACK FOR HIV

These studies were so effective at identifying patients exposed to HIV prior to the
availability of blood donor screening tests that lookback for HIV became part
of the routine of blood collecting facilities and was ultimately mandated by FDA.
Some authors found that an expanded lookback could help identify recipients
at risk. They cross-referenced blood donor lists with case reports of AIDS made
to the Department of Health and identified several seropositive recipients of blood
donated by these individuals (48). This approach was not widely utilized because
of ethical questions and concerns about confidentiality.
One of the most tragic aspects of transfusion-associated HIV was the
inadvertent transmission of the virus to spouses and sexual partners. One CDC
study verified that 7 of 32 (21.9%) female partners of male recipients were
themselves infected with HIV-1, as compared with none of 14 male partners of
female recipients (p = 0.08). Transmission was not associated with frequency
of unprotected vaginal intercourse (49). In another study, blood recipients who
transmitted HIV-1 to their sexual partners had higher mean viral RNA levels than
did nontransmitting recipients (4.3 vs. 3.6 log10 copies/mL; p = 0.05), suggesting
that viral load contributed to heterosexual infectivity (50). The most important
factor associated with transmission of HIV in a review of 132 recipients who
tested positive for antibody to HIV-1 as part of the Transfusion Safety Study was
viremia (51). Review of data also suggested that older RBC units were less likely
to transmit HIV than fresher units, probably because of the viability of lympho-
cytes (52).
Another sad aspect of the AIDS epidemic took place in less developed
countries in Latin America, Africa, and Asia. For instance, plasmapheresis banks
reusing needles played a major role in the dissemination of HIV infection in
Mexico, where paid donors provided a third of the blood used in 1986 (53). By
1996, 9 of the 12 Central and South American countries studied screened all
donors for HIV (54). However, donor screening is spotty at best in many areas of
the Middle East, India, and Africa. Although the contribution of transfusion-
transmitted infection to the HIV epidemic has not been accurately assessed, an
estimated 510% of HIV infections in developing countries are due to blood
transfusion. In a study conducted one year after implementation of HIV blood
screening in the largest hospital in the capital city of the Democratic Republic of
the Congo, an estimated 25% of pediatric HIV infections and 40% of infections
among children over one year of age were due to transfusion (55).
262 Bianco and Rios
X. IDIOPATHIC CD4+ T-CELL LYMPHOCYTOPENIA,

OR AIDS WITHOUT HIV
In the course of the International Conference on AIDS in 1992, there were several
case reports of patients with profound immunosuppression without serological or
molecular evidence of infection by HIV. The general media immediately pub-
lished these reports and raised substantial concerns about a potential new epi-
demic of immunosuppression caused by an unknown infectious agent (56,57).
The case definition of idiopathic CD4+ T-cell lymphocytopenia (ICL) was broad:
a CD4+ cell count lower than 400/L in an individual who was negative for HIV
serology. Follow-up studies demonstrated that transmissible agents did not cause
this condition (58,59). A CDC task force reviewed more than 230,000 cases of
AIDS and 47 patients with ICL. It confirmed that the disorder was rare and
represented various clinical and immunological states not associated with a
transmissible agent (60). The best explanation for the ICL phenomenon came
from a study of the distribution of CD4+ cells in normal individuals. Individuals
with a low CD4+ count are at the tail end of the normal distribution. Essen-
tially, ICL appeared to be an artifact of the normal statistical distribution of CD4+
values (61).
XI. DONOR SELECTION

In the absence of screening assays prior to 1985, a large amount of effort was
developed in donor screening procedures and potential surrogate tests. In early
1983, the New York Blood Center (NYBC) introduced confidential unit exclusion
(CUE). The procedure was designed to allow members of groups at increased risk
of AIDS to confidentially designate their donations for laboratory studies and not
for transfusion. AIDS-related questions in medical history led to a 2% increase in
donor rejections; 97% of donors said their blood could be used for transfusions;
1.4% said their blood could be used for laboratory studies only; and 1.6% did
not respond (62). CUE has been less effective recently in the detection of
HIV-positive individuals than in the early days of donor screening for HIV.
A study performed in Canada showed that only 1.7% of individuals positive on
HIV Western blot used CUE (63). In the last few years none of the HIV-positive
donors to NYBC have used CUE (C. Bianco, personal communication).
About 50% of the donors who tested positive for HIV in 1985 used CUE.
This high degree of effectiveness may have been unique to NYBC (64). Overall,
only 35% of units donated by window-period donors were not transfused
because of the CUE option. Although donors who confidentially exclude their
blood from transfusion were 21 times more likely to have HIV antibody, the rarity
of window-period donors and the infrequency of confidential exclusion by
window-period donors cause the CUE option to have minimal impact on transfu-
sion safety. Only a miniscule number of HIV-positive donors identified by NYBC

in the last 10 years opted for CUE (C. Bianco, personal communication).
A major effort has been dedicated to understanding why HIV-positive
individuals donate blood despite the educational materials provided to donors and
questions asked during medical history. Of 388 seropositive men who donated
between 1988 and 1989, 56% had had sex with men, 10% had used drugs
intravenously, 8% had had sex with intravenous drug users, and 27% had no
identified risk. Of 124 seropositive women, 58% had had sex with men at risk
for HIV (81% of whom used drugs intravenously), 5% had used drugs intrave-
nously, and 41% had no identified risk. Racial and ethnic minorities made up 68%
of seropositive donors (black, 38%; Hispanic, 30%) (65). More recent data
from the study show an increase in individuals with no identified risk and in
minorities. A study sponsored by FDA in the early 1990s suggested that direct
questions about risk behavior would be effective in the identification of individ-
uals with risk behavior (66). Direct oral questions were reasonably well accepted
by potential blood donors (67). On April 23, 1992, FDA recommended adoption
of the direct questioning approach to medical history by all regulated blood
establishments (68).
XII. TESTING
The history of introduction of HIV screening tests for blood donor screening has
been extensively reviewed (69).
A. Surrogate Testing
Prior to the availability of a specific screening test for HIV, there were many
attempts to identify surrogate assays that could assist in donor screening. For in-
stance, some scientists at CDC proposed the use of the antibody to the hepatitis
B core antigen (HBcAb) and the detection of urinary neopterin for identification
of donors at risk (70). Others proposed the use of levels of CD4+ cells for the
screening of blood donors (71). However, a study of the correlation of HBcAb
and CD4+ cell counts with use of CUE (a correlate of risk behavior) suggested
that the contribution of these assays in the absence of a specific screening assay
for HIV would be relatively small (72). In addition, determination of T-helper/
T-suppressor cell ratios, HBcAb, and immune complexes among 18 sets of blood
donors who donated blood to recipients who subsequently developed AIDS could
not distinguish suspected transmitters from controls. None of the assays was as
sensitive and specific as the later developed tests for antibody to HIV (73).
264 Bianco and Rios
B. Specific Tests
The first specific tests for antibodies to HIV became available in March 1985.
Within several weeks, the entire blood supply was being screened using the
first-generation assays. These assays had a number of recognized problems, such
as lack of specificity and known cross-reactivity with antibodies to HLA antigens
that were present on the cell line used to grow the virus. There was so much
concern about lack of specificity that the FDA Memorandum to Blood Establish-
ments published in February 1995 determined that a sample was reactive only
when, after an initial reactive screen, it was repeated in duplicate and one of the
duplicates was reactive. There were no confirmatory assays available, and there
was widespread concern about notifying donors of reactive test results in the
absence of adequate sensitivity and specificity data. Most blood-collecting facil-
ities delayed notification of reactive blood donors until the first quarter of 1987,
when the Western blot confirmatory assay was licensed (74). Despite these
caveats, the early HIV screening assays contributed immensely to the safety of
the blood supply. The development of blood donor screening tests for HIV has
been extensively reviewed in the past (75).
C. Testing for HIV-2, HIV-1 Group O

FDA recommended that screening of blood donors for HIV-2 be implemented by
June 1, 1992, a few months after licensing a combined test for antibodies to HIV-1
and HIV-2 (76). Between June 1992 and June 1995, there were two HIV-2
positive donors identified among an estimated 74 million blood and plasma
donations that had been screened for HIV-2 (77). The next version of screening
assays for HIV-1 must also be able to detect HIV-1 Group O variants, as discussed
previously (2123).
D. Confirmation and Indeterminates

The licensure of the Western blot assay for antibodies to HIV in 1987 introduced
the then new concept of indeterminate test results. Fearful that test results that
were not clearly negative could be associated with risk of HIV transmission to
recipients, FDA determined that donors with reactive results on screening tests
that were not negative (a blank strip) on Western blot could not be accepted as
blood donors. Several subsequent studies documented the lack of specificity of
these EIA assays and have clearly shown that individuals with indeterminate test
results do not constitute a risk to the safety of the blood supply. For instance,
donors with indeterminate supplemental test results for HIV identified between
1990 and 1993 were tested by EIA, Western blot, and peptide assays for
antibodies to HIV-2 and HIV-1 subtype O. Peripheral blood mononuclear cells

and/or plasma from the follow-up samples were also tested for HIV-1 DNA and/or
RNA by polymerase chain reaction. The study concluded that contemporary
blood donors classified as indeterminate in supplemental HIV testing are infre-
quently infected with HIV and recommended that donors whose follow-up
samples test negative in anti-HIV-1/2 EIAs and negative or persistently indeter-
minate in Western blots be considered eligible for reinstatement (78). Another
study addressed donors who were positive for HIV-1 by Western blot but were
possibly falsely positive because they lacked reactivity to p31. Many of these
individuals were tested by PCR and shown not to be infected with HIV-1.
The false-positive rate of Western blotpositive donors was 4.8% of the EIA
reactive donors and 0.0004% (1 in 251,000) of all donors (95% CI, 1 in 173,000
to 1 in 379,000 donors) (79).
E. Addressing the Window of Seroconversion for HIV

As mentioned above, the introduction of screening tests for antibodies to HIV led
to a substantial reduction of transmission by blood and blood products. However,
a small number of cases continued to occur, mostly because of the window of
seroconversion, that is, the period between the first presence of infectious virus
in the circulation of a recently infected individual and the ability of the screening
assay to detect the HIV infection. Reducing or closing the window has been
the ultimate goal of every assay improvement and new assay technology im-
plemented since 1985, as shown in Table 2. The first well-documented cases of
transmission to 13 blood recipients who had received blood from 7 donors who
screened negative for antibody to HIV at the time of donation were published in
1988. All 7 donors seroconverted (80).
Table 2 Evolution of Blood Donor Screening Assays for HIV
HIV per million

Date Test transfusions Window (days)
Pre-1985 N.A. 2400

19851987 HIV-1 (1.0) 7.7 56
19871990 HIV-1 (1.0+) 5.7 42
19911992 EIA-2 (2.0) 4.5 33
19921996 HIV-1/2 (3.0) 1.8 22
1996 P24 Ag 1.3 16
(?) DNA PCR (1.3) (16)
1999 RNA PCR 0.9 11
266 Bianco and Rios
Estimates of risk associated with the window of seroconversion for HIV in

the 1980s were derived from studies of donor-recipient pairs. In a study of 4,163
adults undergoing cardiac surgery who received 36,282 transfusions of blood
components, the investigators found one case of HIV-1 transmission by transfu-
sion of screened blood components or 0.003% per unit (81). This large multicen-
ter study among cardiac surgery patients was updated in 1992 and included
11,532 patients who received 120,312 units in three hospitals. Two new HIV
infections were detected, for a rate of 0.0017% with an upper limit of the 95% Cl
of 0.0053% (82). The study of 179 seropositive donors and 39 recipients of blood
also addressed the issue of seroconversion from these donors. The window period
averaged 45 days, with few, if any, donors remaining infectious and seronegative
for longer than 6 months (83).
Transfusion risk estimated based on the window period of assays in use for
donor screening was 1:153,000 per unit transfused in 1989 (84). More recently,
two major studies generated more accurate and up-to-date estimates of window-
period transmission. A collaboration between the Centers for Disease Control and
Prevention (CDC) and the American Red Cross estimated that in 19921993 one
donation in every 360,000 (95% CI, 210,0001,140000) was made during the
window period. In addition, it estimated that 1 in 2,600,000 donations was HIV
seropositive but was not identified as such because of an error in the labora-
tory. The study also estimated that 1542% of window-period donations were
discarded because they were seropositive on laboratory tests other than the HIV-
antibody test. The final estimated risk for a transfusion recipient was 1:450,000
660,000 donations of screened blood (85). The other study was based on
incidence data collected through the Retroviral Epidemiology Donor Study
supported by the National Heart, Lung and Blood Institute (86). It estimated the
risk of donations made by donors whose units passed all screening tests during
an infectious window as 1:493,000 (95% CI, 202,0002,778,000).
Improvement in the sensitivity of anti-HIV assays has resulted in significant
shortening of the preseroconversion window period. A study comparing the
sensitivity of early screening assays with current generation assays showed days
of shortening of the HIV window period as follows: contemporary anti-HIV-1/2
EIAs, 20.3 days (95% Cl, 8.032.5); p24 antigen and DNA PCR, 26.4 days (95%
Cl, 12.638.7); and RNA PCR, 31.0 days (95% Cl, 16.745.3) (87).
Testing donor specimens with both a sensitive HIV-1 EIA (3A11 assay) and
a less sensitive modification of the same EIA (3A11-LS assay) allows for the
differentiation between individuals with early HIV-1 infection from those infected
for longer periods. This is a result of the increase in the affinity of antibodies that
accompanies development of the immune response in the infected individual.
This approach allowed estimation of incident infections among first-time donors
as 7.18 per 100,000 per year (95% CI, 4.5111.20/100,000) versus repeat blood
donors with 2.95/100,000 per year (95% CI, 1.146.53/100,000). Thus, the
incidence of HIV infection among first-time donors is 2.4 times higher than that
of repeat donors (88). Interestingly, concern over the theoretical possibility of
disease transmission via blood from donors who develop Creutzfeldt-Jakob
disease has led to proposals for the deferral of donors who are 50 years of age or
older. This would require recruitment of a higher number of first-time donors to
replace the deferred individuals. The estimated increase in the risk of infected
units would be 12% for HIV, 21% for HCV, and 22% for HBsAg (89). The po-
tential effects of the introduction of research assays for HIV RNA will be
discussed later in this chapter.
F. Blood Donor Screening for HIV-1 p24 Antigen

The desire to shorten the window of seroconversion for HIV led to the develop-
ment of a test for HIV-1 p24 antigen that could identify viral particles in
circulation, very much like the very effective HBsAg assays. Unfortunately, large
clinical trials showed that the contribution of the HIV-1 p24 antigen assays would
be miniscule. For instance, a large trial carried out in 1989 among over 500,000
volunteer blood donors in the United States failed to identify a single individual
positive for HIV-1 p24 antigen and negative on the screening test for antibodies
to HIV-1. The finding suggested that the available HIV-1 p24 antigen test would
not add substantially to the safety of the U.S. blood supply (90).
However, concerns about the risk of transmission of HIV by donations
made in the window of seroconversion continued to rise with the infrequent, but
periodic, identification of cases of transmission of HIV by transfusion (91,92).
The incidence rate of HIV in certain populations was also extremely high. In one
study of 2300 emergency room patients, 180 (7.8%) were Western blot (WB)
positive for HIV-1 antibodies. Of 2120 antibody-negative or WB-indeterminate
patients, none of whom were identified on clinical grounds as having primary
HIV-1 infection, 6 (0.28%) were HIV-1 p24 antigenpositive and had serologies
consistent with primary HIV-1 infection. Of these 6, 3 were seronegative even
with third-generation antibody ELISA assays. Thus 3 of 183 individuals showing
up at an emergency room in a large metropolitan area were positive for HIV p24
antigen and negative for antibodies for HIV-1 (93).
The issue of HIV-1 p24 antigen screening was addressed by the FDA Blood
Products Advisory Committee (BPAC). After extensive discussion, the committee
decided not to recommend its introduction. The decision was based on two major
issues: lack of cost benefit, considering the miniscule number of individuals that
would be detected, and fear of the magnet effect, that is, fear that individuals
at risk may be tempted to donate blood in order to obtain confidential test results
(94,95). Following the meeting, FDA received a letter from a U.S. congress-
man questioning BPACs decision. The FDA commissioner reconstituted BPAC.
On August 8, 1995, FDA issued a Memorandum to Registered Establishments
268 Bianco and Rios
recommending adoption of the HIV-1 p24 antigen screening assay as soon as

licensed (96). The first test was licensed on March 14, 1996, and implemented
overnight by the blood banking community. By March 1999, with almost 40
million units screened, there had been five cases of individuals positive on the
HIV-1 p24 antigen test and negative on the antibody test, for a rate of 1:8,000,000.
This rate is much smaller than that predicted at the time of test introduction
(about seven cases per year). The most accepted explanation is that recently
infected individuals have a viral syndrome, do not feel well, have a fever, and do
not donate.
G. Blood Donor Screening by Nucleic Acid

Amplification Technologies
Obviously, HIV-1 p24 antigen screening could not close the window of
seroconversion for HIV. In an attempt to further address this issue, FDA convened
a meeting in September 1994 to discuss the feasibility of closing the window
through nucleic acid amplification technologies (NAT) (97). That meeting con-
cluded that NAT was promising, but that the technology was not practical for
mass donor screening. In October 1996 NHLBI awarded contracts to manufac-
turers for the development of tests and equipment that would allow mass
screening of blood donors HCV and HIV RNA using cGMP-compliant automated
methodology.
The availability of commercial assays based on (NAT) encouraged the
development of approaches for screening blood donors even before automated
equipment became available. One of these approaches was to use pools of donor
samples. Pools would reduce the number of specimens that needed to be tested
because the vast majority of blood donors are negative for HIV (and for HCV).
Pools had already been used successfully to determine risk of HIV transmission
in a study carried out in San Francisco. In that study, 1530 pools of leukocytes
from 50 donors each were tested for HIV by culture, by PCR, or by both.
One pool was positive for both, and the calculated risk of HIV transmission
for the donor population (November 1987December 1989) was estimated at
1:61,171 (98).
In 1999, U.S. blood centers implemented NAT screening of blood donors
under Investigational New Drug (IND) applications submitted to FDA as clinical
research studies in collaboration with the manufacturers. The study objectives are
to determine the specificity and reproducibility of the NAT test systems on pooled
samples in high volume blood donor screening laboratories, to assess the impact
of NAT testing on blood component availability, and to evaluate the use of NAT
tests to improve the safety of the blood supply. Two test systems are being used:
the Roche Molecular Systems COBAS AMPLISCREEN for HCV and HIV and
the GenProbe Pooled Plasma HIV-1/HCV Amplified Assay. The Roche COBAS
AMPLISCREEN system is based on PCR amplification of target cDNA using

virus-specific complementary primers, that is, hybridization of the amplified
products to oligonucleotide probes specific to the target. The entire test pro-
cess takes approximately 6 hours. The GenProbe HIV-1/HCV test, which de-
tects HIV-1 and/or HCV RNA in a single tube multiplex format, is based on
transcription-mediated amplification (TMA), which utilizes two enzymes to
produce RNA amplicons via DNA intermediates. These assays are being carried
out in pools of plasma samples that include 1624 donors. If the pool is positive
in NAT, it is broken down into smaller pools and retested until the positive sample
is identified. Although data from preclinical studies are encouraging, in that
several donors positive on NAT and negative for antibodies to HCV have been
identified, specificity, sensitivity, reproducibility, and run validity data for these
test systems are unknown in the donor screening laboratory setting. To date, with
over 2 million units screened, only one donor positive for HIV-1 on NAT and
negative for antibodies to HIV-1 or p24 antigen has been confirmed to have
been identified.
Similar systems have been implemented by manufacturers of plasma deriv-
atives and by several countries in Europe, and recent experience suggests that
NAT testing reduces the window of seroconversion for HCV. The data for HIV
are not as encouraging (99101).
Experimental data obtained in chimpanzees indicate that NAT may actually
completely close the window of seroconversion for HIV. Apparently, the initial
replication of HIV occurs in cells of the immune system that are not circulating.
When viral particles appear in the circulation, NAT becomes positive. There was
no demonstrable infectivity in either plasma or peripheral blood mononuclear
cells obtained before molecular markers were detectable. This suggested that the
infectious window is considerably shorter than the total window as measured
from exposure and that NAT might not only shorten the seronegative window, but
also totally prevent transfusion-transmitted HIV infection (102).
One issue highlighted by the introduction of NAT under a research protocol
was the appropriateness of using postexposure prophylaxis (PEP) against HIV in
case a donated unit was transfused and HIV-1 NAT test results became available
only after the transfusion. CDC has issued extremely useful guidelines for PEP
that are applicable to recipients of potentially infected blood products and to
health care workers inadvertently exposed to HIV-positive materials (103).
XIII. DONOR NOTIFICATION AND COUNSELING

A fundamental piece of the prevention of disease transmission by transfusion is
donor notification and counseling. During this process, donors with positive
results for infectious disease markers are notified of test results and are edu-
270 Bianco and Rios
cated about the risk of transmitting infections to recipients of their donations

(104). Unfortunately, the donor counseling process for the majority of the
other donors, that is, those with repeatedly reactive results for HIV on EIA or on
HIV-1 p24 antigen with negative or indeterminate supplemental test results, is not
very effective.
Donor notification has been trivialized by the bureaucratic approach of
FDA guidances: (a) all donors with repeatedly reactive results are deferred from
donating blood even if negative on subsequent donations; (b) donors with
indeterminate test results, that is, with any bands in Western blot even if clearly
nonviral bands, are never acceptable as blood donors again; (c) donors who had
repeatedly reactive test results on more than one donation in the past, even in the
first generation of nonspecific EIAs, are not acceptable; and (d) there are no
negative results for HIV-1 p24 antigen supplemental tests. Results are either
positive or indeterminate. Thus, regulations have created a cadre of individuals
in limbo. These donors cannot be cleared even after years of negative screening
test results and negative NAT results. Donors are extremely frustrated by these
rules. Blood centers are very frustrated by these rules. They do not contribute to
the safety of the blood supply.
XIV. COST/BENEFIT CONSIDERATIONS

There have been many attempts to identify strategies that facilitate the im-
plementation of blood donor screening assays for HIV in third world countries.
Most of these approaches have been rejected outright, despite evidence that they
would have contributed substantially to the safety of the blood supply in these
countries. One study showed a reduction of 70% in costs when pools of 10
samples were used for initial EIA screen (105). Another study reused wells on
EIA plates that had tested negative on previous rounds and concluded that there
was no significant reduction in sensitivity of the test samples or controls when
run in parallel with new plates (106). Obviously, these procedures are unaccept-
able in the United States or Europe. They have rarely been used in third world
countries because of concerns about tests that are less sensitive than those used
in industrialized countries.
A review of the addition of newer blood donor screening tests for HIV since
1992 (HIV-2, HIV-1 p24 antigen, requirement for HIV-1 type O sensitivity,
research NAT) show that their contribution to prevention of HIV transmission by
transfusion of blood and blood products is miniscule, if any. This impression is
confirmed by studies designed to estimate the cost-effectiveness of expanding the
HIV testing protocol for donated blood by adding HIV-1 p24 antigen detection or
RNA PCR (at costs of $5/unit and $8/unit, respectively). The addition of p24
antigen testing would prevent eight more cases at a net additional cost of $60
million annually ($2.3 million/quality-adjusted life-year); RNA PCR testing

would prevent 16 more cases at a net additional cost of $96 million annually
($2.0 million/quality-adjusted life-year). The authors concluded that the cost-
effectiveness of these additions was far below that of most medical interventions.
On the other hand, HIV antibody testing prevents 1568 cases of transfusion-
acquired HIV infection each year at a modest cost of $3,600 per quality-adjusted
year of life saved (107). [This study of test efficacy projected based on anticipated
window-period reductions (6 days for p24 antigen, 11 days for RNA PCR) and
donor seroconversion rates derived from the Retrovirus Epidemiology Donor
Study on the basis of current estimates of HIV prevalence rates in blood donors
(1/10,000) and 16 million annual transfusions in the United States.]
XV. CONCLUSION
The control of HIV transmission by transfusion of blood and blood products is
one of the biggest success stories of science and medicine. We certainly will
benefit from the study of all the efforts applied by many people in many different
areas to address this major issue. The major issue that still remains unresolved
for the transfusion medicine community is how to effectively transmit infor-
mation about the safety of the blood supply instead of the risks associated with
blood transfusions.
REFERENCES
1. Centers for Disease Control and Prevention. Epidemiologic notes and reports:
Pneumocystis carinii pneumonia among persons with hemophilia A. MMWR 1982;
31:365367,
2. Centers for Disease Control and Prevention. Update on acquired immunodeficiency
syndrome (AIDS) among patients with hemophilia A. MMWR 1982; 31:644646,
3. Centers for Disease Control and Prevention. Epidemiologic notes and reports:
possible transfusion-associated acquired immune deficiency syndrome (AIDS)
California. MMWR 1982; 31:652654,
4. Jett JR, Kuritsky JN, Katzmann JA, Homburger HA. Acquired immunodeficiency
syndrome associated with blood-product transfusions. Ann Int Med 1983; 99:621
624,
5. Curran JW, Lawrence DW, Jaffe, H, Kaplan JE, et al. Acquired immuno-deficiency
syndrome (AIDS) associated with transfusions. N Engl J Med 1984; 310:6975,
6. Bove, JR. Transfusion associated AIDSa cause for concern. N Engl J Med 1984;
310:115116,
7. Remis RS, OShaughnessy MV, Tsoukas C, Growe GH, Schechter MT, Palmer RW,
Lawrence DN. HIV transmission to patients with hemophilia by heat-treated,
donor-screened factor concentrate. CMAJ 1990; 142:12471254.
272 Bianco and Rios
8. Dietrich SL, Mosley JW, Lusher JM, Hilgartner MW, Operskalski EA, Habel L,
Aledort LM, Gjerset GF, Koerper MA, Lewis BH, et al. Transmission of human
immunodeficiency virus type 1 by dry-heated clotting factor concentrates. Transfu-
sion Safety Study Group. Vox Sang 1990; 59:129135.
9. Zuck TF, Eyster ME. Blood safety decisions, 1982 to 1986: perceptions and
misconceptions. Transfusion 1996; 36:928931.
10. Bianco C. Protecting the blood supply. Transfus Med Rev 1989; 3 (suppl 1):912,
11. Horowitz B, Prince AM, Hamman J, Watklevicz C. Viral safety of solvent/detergent-
treated blood products. Blood Coagul Fibrinolysis 1994; (suppl 3)S2128.
12. Martin MA. The molecular and biological properties of the human immunodefici-
ency virus. In: Stamatoyannopoloulos G, Neinhuis AW, Majerus PW, Varmus H,
eds. The Molecular Basis of Blood Diseases. 2d ed. Philadelphia: W.B. Saunders
Company, 1994:863908.
13. Fauci AS, Desrosiers RC. Pathogenesis of HIV and SIV. In: Coffin JM, Hughes SH,
Varmus HE, eds. Retroviruses. Plainview, NY: Cold Spring Harbor Laboratory
Press, 1997: 587635.
14. Wain-Hobson S. Running the gamut of retroviral variation. Trends Microbiol 1996;
4:135141.
15. Vartanian JP, Sala M, Henry M, Wain-Hobson S, Meyerhans A. Manganese cations
increase the mutation rate of human immunodeficiency virus type 1 ex vivo. J Gen
Virol 1999; 80:19831986.
16. Berkhout B. HIV-1 evolution under pressure of protease inhibitors: climbing the
stairs of viral fitness. J Biomed Sci 1999; 6:298305.
17. Brown AJ, Korber BT, Condra JH. Associations between amino acids in the
evolution of HIV type 1 protease sequences under indinavir therapy. AIDS Res Hum
Retroviruses 1999; 15:247255.
18. Morris A, Marsden M, Halcrow K, Hughes ES, Brettle RP, Bell JE, Simmonds P.
Mosaic structure of the human immunodeficiency virus type 1 genome infecting
lymphoid cells and the brain: evidence for frequent in vivo recombination events in
the evolution of regional populations. J Virol 1999; 73:87208731.
19. Ostrowski MA, Krakauer DC, Li Y, Justement SJ, Learn G, Ehler LA, Stanley SK,
Nowak M, Fauci AS. Effect of immune activation on the dynamics of human
immunodeficiency virus replication and on the distribution of viral quasispecies.
J Virol 1998; 72:77727784.
20. Burke DS. Recombination in HIV: an important viral evolutionary strategy. Emerg
Infect Dis 1997; 3:253259.
21. De Leys R, Vanderborght B, Vanden Haesevelde M, et al. Isolation and partial
characterization of an unusual human immunodeficiency retrovirus from two per-
sons of West-Central African origin. J Virol 1990; 64:12071216.
22. Centers for Disease Control and Prevention. Identification of HIV-1 Group O
InfectionLos Angeles County, California, 1996. MMWR 1996; 45:561565.
23. van Binsbergen J, Keur W, v d Graaf M, Siebelink A, Jacobs A, de Rijk D, Toonen
J, Zekeng L, Afane Ze E, Gurtler LG. Reactivity of a new HIV-1 group O third
generation A-HIV-1/-2 assay with an unusual HIV-1 seroconversion panel and
HIV-1 group O/group M subtyped samples. J Virol Methods 1997; 69:2937.
24. Centers for Disease Control and Prevention. HIV AIDS Surveillance Report 1998;
10:14.
25. Cao Y, Qin L, Zhang L, Safrit J, Ho DD. Virologic and immunologic characteriza-
tion of long-term survivors of human immunodeficiency virus type 1 infection. N
Engl J Med 1995; 332:201208.
26. Pantaleo G, Menzo S, Vaccarezza M, Graziosi C, Cohen OJ, Demarest JF,
Montefiori D, Orenstein JM, Fox C, Schrager LK, et al. Studies in subjects with
long-term nonprogressive human immunodeficiency virus infection. N Engl J Med
1995; 332:209216.
27. de Roda Husman AM, Koot M, Cornelissen M, Keet IP, Brouwer M, Broersen SM,
Bakker M, Roos MT, Prins M, de Wolf F, Coutinho RA, Miedema F, Goudsmit J,
Schuitemaker H. Association between CCR5 genotype and the clinical course of
HIV-1 infection. Ann Intern Med 1997; 127:882.
28. Quillent C, Oberlin E, Braun J, et al. HIV-1-resistance phenotype conferred by
combination of two separate inherited mutations of CCR5 gene. Lancet 1998;
351:1418.
29. Wilkinson DA, Operskalski EA, Busch MP, Mosley JW, Koup RA. A 32-bp deletion
within the CCR5 locus protects against transmission of parenterally acquired human
immunodeficiency virus but does not affect progression to AIDS-defining illness.
J Infect Dis 1998; 178:11631166.
30. Duesberg P, Rasnick D. The AIDS dilemma: drug diseases blamed on a passenger
virus. Genetica 1998; 104:85132.
31. Quinn TC. Acute primary HIV infection. JAMA 1997; 278:5862.
32. Lui KJ, Darrow WW, Rutherford GW 3d. A model-based estimate of the mean
incubation period for AIDS in homosexual men. Science 1988; 240:13331335.
33. Kalbfleisch JD, Lawless JF. Estimating the incubation time distribution and ex-
pected number of cases of transfusion-associated acquired immune deficiency
syndrome. Transfusion 1989; 29:672676.
34. Mylonakis E, Dickinson BP, Mileno MD, Flanigan T, Schiffman FJ, Mega A, Rich
JD. Persistent parvovirus B19 related anemia of seven years duration in an
HIV-infected patient: complete remission associated with highly active antiretroviral
therapy. Am J Hematol 1999; 60:164166.
35. Palella FJ Jr, Delaney KM, Moorman AC, et al. Declining morbidity and mortality
among patients with advanced human immunodeficiency virus infection. HIV
Outpatient Study Investigators. N Engl J Med 1998; 26 (338):853860.
36. Groopman JE. Impact of transfusion on viral load in human immunodeficiency virus
infection. Semin Hematol 1997; 34 (suppl 2):2733.
37. Busch MP, Young MJ, Samson SM, Mosley JW, Ward JW, Perkins HA. Risk of
human immunodeficiency virus (HIV) transmission by blood transfusions before the
implementation of HIV-1 antibody screening. The Transfusion Safety Study Group.
38. Minamoto GY, Scheinberg DA, Dietz K, Gold JW, Chein N, Gee T, Reich LM,
Hoffer J, Mayer K, Armstrong D, et al. Human immunodeficiency virus infection
in patients with leukemia. Blood 1988; 71:11471149.
39. Ward JW, Bush TJ, Perkins HA, Lieb LE, Allen JR, Goldfinger D, Samson SM,
Pepkowitz SH, Fernando LP, Holland PV, et al W, Bush TJ, Perkins HA, Lieb LE,
274 Bianco and Rios
Allen JR, Goldfinger D, Samson SM, Pepkowitz SH, Fernando LP, Holland PV,
et al. The natural history of transfusion-associated infection with human immuno-
deficiency virus. Factors influencing the rate of progression to disease. N Engl J
Med 1989; 321:947952.
40. Kalbfleisch JD, Lawless JF. Estimating the incubation time distribution and ex-
pected number of cases of transfusion-associated acquired immune deficiency
syndrome. Transfusion 1989; 29:672676.
41. Selik RM, Ward JW, Buehler JW. Trends in transfusion-associated acquired immune
deficiency syndrome in the United States, 1982 through 1991. Transfusion 1993;
33:890893.
42. Kleinman SH, Niland JC, Azen SP, Operskalski EA, Barbosa LH, Chernoff Al,
Edwards VM, Lenes BA, Marshall GJ, Nemo GJ, et al. Prevalence of antibodies to
human immunodeficiency virus type 1 among blood donors prior to screening. The
Transfusion Safety Study/NHLBI Donor Repository. Transfusion 1989; 29:572
580.
43. Jaffe HW, Sarngadharan MG, DeVico AL, Bruch L, Getchell JP, Kalayanaraman
VS, Haverkos HW, Stoneburner RL, Gallo RC, Curran JW. Infection with HTLV-
III/LAV and transfusion-associated acquired immunodeficiency syndrome. Sero-
logic evidence of an association. JAMA 1985; 254:770773.
44. Feorino PM, Jaffe HW, Palmer E, Peterman TA, Francis DP, Kalyanaraman VS,
Weinstein RA, Stoneburner RL, Alexander WJ, Raevsky C, et al. Transfusion-
associated acquired immunodeficiency syndrome. Evidence for persistent infection
in blood donors. N Engl J Med 1985; 312:12931296.
45. Anderson KC, Gorgone BC, Marlink RG, Ferriani R, Essex ME, Benz PM,
Groopman JE. Transfusion-acquired human immunodeficiency virus infection among
immunocompromised persons. Ann Intern Med 1986; 105:519527.
46. Antin JH, Smith BR, Ewenstein BM, Arceci RJ, Lipton JM, Page PL, Rappeport
JM. HTLV-III infection after bone marrow trasplantation. Blood 1986; 67:160163.
47. Lieb LE, Mundy TM, Goldfinger D, Pepkowitz SH, Brunell PA, Caldwell MB,
Ward JW. Unrecognized human immunodeficiency virus type 1 infection in a cohort
of transfused neonates: a retrospective investigation. Pediatrics 1995; 95:717721.
48. Samson S, Busch M, Ward J, Garner J, Salk S, Fernando L, Holland P, Rutherford
G, Benjamin R, Perkins H. Identification of HIV-infected transfusion recipients: the
utility of cross-referencing previous donor records with AIDS case reports. Trans-
fusion 1990; 30:214218.
49. OBrien TR, Busch MP, Donegan E, Ward JW, Wong L, Samson SM, Perkins HA,
Altman R, Stoneburner RL, Holmberg SD J. Heterosexual transmission of human
immunodeficiency virus type 1 from transfusion recipients to their sex partners.
J AIDS 1994; 7:705710.
50. Operskalski EA, Stram DO, Busch MP, Huang W, Harris M, Dietrich SL, Schiff ER,
Donegan E, Mosley JW. Role of viral load in heterosexual transmission of human
immunodeficiency virus type 1 by blood transfusion recipients. Transfusion Safety
Study Group. Am J Epidemiol 1997; 146:655661.
51. Busch MP, Operskalski EA, Mosley JW, Lee TH, Henrard D, Herman S, Sachs DH,
Harris M, Huang W, Stram DO. Factors influencing human immunodeficiency virus
type 1 transmission by blood transfusion. Transfusion Safety Study Group. J Infect

Dis 1996; 174:2633.
52. Donegan E, Lenes BA, Tomasulo PA, Mosley JW. Transmission of HIV-1 by
component type and duration of shelf storage before transfusion. Transfusion 1990;
30:851852.
53. Volkow P, Perez-Padilla R, del-Rio C, Mohar A. The role of commercial plasma-
pheresis banks on the AIDS epidemic in Mexico. Rev Invest Clin 1998; 50:221226.
54. Schmunis GA, Zicker F, Pinheiro F, Brandling-Bennett D. Risk for transfusion-
transmitted infections diseases in Central and South America. Emerg Infect Dis
1998; 4:511.
55. Lackritz EM. Prevention of HIV transmission by blood transfusion in the developing
world: achievements and continuing challenges. AIDS 1998; 12 (suppl A):S8186.
56. Duncan RA, von Reyn CF, Alliegro GM, Toossi Z, Sugar AM, Levitz SM. Idiopathic
CD4+ T-lymphocytopeniafour patients with opportunistic infections and no evi-
dence of HIV infection. N Engl J Med 1993; 328:393398.
57. Laurence J. T-cell subsets in health, infectious disease, and idiopathic CD4+
T-lymphocytopenia. Ann Intern Med 1993; 119:5562.
58. Ho DD, Cao Y, Zhu T, Farthing C, Wang N, Gu G, Schooley RT, Daar ES. Idiopathic
CD4+ T-lymphocytopeniaan analysis of five patients with unexplained opportu-
nistic infections. N Engl J Med 1993; 328:386392.
59. Spira TJ, Jones BM, Nicholson JK, Lal RB, Rowe T, Mawle AC, Lauter CB,
Shulman JA, Monson RA. Idiopathic CD4+ T-lymphocytopeniaimmunodeficiency
without evidence of HIV infection. N Engl J Med 1993; 328:380385.
60. Smith DK, Neal JJ, Holmberg SD. Unexplained opportunistic infections and CD4+
T-lymphocytopenia without HIV infection. An investigation of cases in the United
States. The Centers for Disease Control Idiopathic CD4+ T-lymphocytopenia Task
Force. N Engl J Med 1993; 328:373379.
61. Busch MP, Valinsky JE, Paglieroni T, Prince HE, Crutcher GJ, Gjerset GF, Oper-
skalski EA, Charlebois E, Bianco C, Holland PV, et al. Screening of blood donors
for idiopathic CD4 + T-lymphocytopenia. Transfusion 1994; 34:192197.
62. 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. Evidence of volunteer blood donor cooperation. Transfusion 1985;
25:39.
63. Chiavetta JA, Nusbacher J, Wall A. Donor self-exclusion patterns and human
immunodeficiency virus antibody test results over a twelve-month period. Transfu-
sion 1989; 29:8183.
64. Petersen LR, Lackritz E, Lewis WF, Smith DS, Herrera G, Raimondi V, Aberle-
Grasse J, Dodd RY. The effectiveness of the confidential unit exclusion option.
Transfusion 1994; 34:865869.
65. Petersen LR, Doll LS and The HIV Blood Donor Study Group. Human im-
munodeficiency virus type 1-infected blood donors: epidemiologic, laboratory, and
donation characteristics. Transfusion 1991; 31:698703.
66. Mayo DJ, Rose AM, Matchett SE, Hoppe PA, Solomon JM, McCurdy KK.
Screening potential blood donors at risk for human immunodeficiency virus.
Transfusion 1991; 31:466474.
276 Bianco and Rios
67. Gimble JG, Friedman LI. Effects of oral donor questioning about high-risk behav-
iors for human immunodeficiency virus infection. Transfusion 1992; 32(5):446449.
68. Food and Drug Administration. Memorandum to all registered blood establishments.
Revised recommendations for the prevention of HIV transmission by blood and
blood products. (http//:www.fda.gov/cber/publications).
69. Busch MP. Transfusion associated AIDS. In: Rossi EC, Simon TL, Moss GS, Gould
SA, eds. Principles of Transfusion Medicine Baltimore: Williams & Wilkins, 1996:
699708.
70. Fuchs D, Hausen A, Reibnegger G, Reissigl H, Schonitzer D, Spira T, Wachter H.
Urinary neopterin in the diagnosis of acquired immune deficiency syndrome. Eur J
Clin Microbiol 1984; 3:7071.
71. Galel SA, Lifson JD, Engleman EG. Prevention of AIDS transmission through
screening of the blood supply. Annu Rev Immunol 1995; 13:201227.
72. Pindyck J, Waldman A, Oleszko W, Zang E, Bianco C. Prevalence of viral
antibodies and leukocyte abnormalities among blood donors considering themselves
at risk of exposure to AIDS. Ann NY Acad Sci 1984; 437:472484.
73. McDougal JS, Jaffe HW, Cabridilla CD, Sarngadharan MG, Nicholson JK,
Kalyanaraman VS, Schable CA, Kilbourne B, Evatt BL, Gallo RC, et al. Screening
tests for blood donors presumed to have transmitted the acquired immunodeficiency
syndrome. Blood 1985; 65:772775.
74. The Consortium for Retrovirus Serology Standardization. Serological diagnosis of
human immunodeficiency virus infection by Western blot testing. JAMA 1988;
260:675679.
75. Groopman JE, Chen FW, Hope JA, Andrews JM, Swift RL, Benton CV, Sullivan
JL, Volberding PA, Sites DP, Landesman S, et al. Serological characterization of
HTLV-III infection in AIDS and related disorders. J Infect Dis 1986; 153(4):736742.
76. Centers for Disease Control and Prevention. Testing for antibodies to human
immunodeficiency virus type 2 in the United States. MMWR 1992; 41 (RR-no.12).
77. Centers for Disease Control and Prevention. Update: HIV-2 infection among blood
donors and plasma donorsUnited States, June 1992-June 1995. MMWR 1995;
44:603606.
78. Busch MP, Kleinman SH, Williams AE, Smith JW, Ownby HE, Laycock ME, Lee
LL, Pau CP, Schreiber GB. Frequency of human immunodeficiency virus (HIV)
infection among contemporary anti-HIV-1 and anti-HIV-1/2 supplemental test-
indeterminate blood donors. The Retrovirus Epidemiology Donor Study. Transfu-
sion 1996; 36:3744.
79. Kleinman S, Busch MP, Hall L, Thomson R, Glynn S, Gallahan D, Ownby HE,
Williams AE. False-positive HIV-1 test results in a low-risk screening setting of
voluntary blood donation. Retrovirus Epidemiology Donor Study. JAMA 1998;
280:10801085.
80. Ward JW, Holmberg SD, Allen JR, Cohn DL, Critchley SE, Kleinman SH, Lenes
BA, Ravenholt O, Davis JR, Quinn MG, et al. Transmission of human im-
munodeficiency virus (HIV) by blood transfusions screened as negative for HIV
antibody. N Engl J Med 1988; 318:473478.
81. Cohen ND, Munoz A, Reitz BA, Ness PK, Frazier OH, Yawn DH, Lee H, Blattner
W, Donahue JG, Nelson KE, et al. Transmission of retroviruses by transfusion of
screened blood in patients undergoing cardiac surgery. N Engl J Med 1989;

320:11721176.
82. Nelson KE, Donahue JG, Munoz A, Cohen ND, Ness PM, Teague A, Stambolis VA,
Yawn DH, Callicott B, McAllister H, et al. Transmission of retroviruses from
seronegative donors by transfusion during cardiac surgery. A multicenter study of
HIV-1 and HTLV-I/II infections. Ann Intern Med 1992; 117:554559.
83. Petersen LR, Satten GA, Dodd R, Busch M, Kleinman S, Grindon A, Lenes B.
Duration of time from onset of human immunodeficiency virus type 1 infectiousness
to development of detectable antibody. Transfusion 1994; 34:283289.
84. Cumming PD, Wallace EL, Schorr JB, Dodd RY. Exposure of patients to human
immunodeficiency virus through the transfusion of blood components that test
antibody-negative. N Engl J Med 1989; 321(14):941946.
Lewis WF, Notari EP 4th, Petersen LR. Estimated risk of transmission of the human
333:17211725.
transmitted viral infections. The Retrovirus Epidemiology Donor Study. N Engl J
Med 1996; 334:16851690.
87. Busch MP, Lee LL, Satten GA, Henrard DR, Farzadegan H, Nelson KE, Read S,
preceding human immunodeficiency virus type 1 seroconversion: implications for
88. Janssen RS, Satten GA, Stramer SL, Rawal BD, OBrien TR, Weiblen BJ, Hecht
FM, Jack N, Cleghorn FR, Kahn JO, Chesney MA, Busch MP. New testing strategy
to detect early HIV-1 infection for use in incidence estimates and for clinical and
prevention purposes. JAMA 1998; 280:4248.
89. Busch MP, Glynn SA, Schreiber GB. Potential increased risk of virus transmission
due to exclusion of older donors because of concern over Creutzfeldt-Jakob disease.
Transfusion 1997; 37(10):9961002.
90. Alter HJ, Epstein JS, Swenson SG, VanRaden MJ, Ward JW, Kaslow RA, Menitove
JE, Klein HG, Sandler SG, Sayers MH, et al. Prevalence of human immuno-
deficiency virus type 1 p24 antigen in U.S. blood donorsan assessment of the
efficacy of testing in donor screening. The HIV-Antigen Study Group. N Engl J Med
1990; 323:13121317.
91. Roberts CR, Longfield JN, Platte RC, Zielmanski KP, Wages J, Fowler A. Transfu-
sion-associated human immunodeficiency virus type 1 from screened antibody-
negative blood donors. Arch Pathol Lab Med 1994; 118:11881192.
92. Centers for Disease Control and Prevention. Persistent lack of detectable HIV-1
antibody in a person with HIV infectionUtah, 1995. MMWR 1996; 45:181185.
93. Clark SJ, Kelen GD, Henrard DR, Daar ES, Craig S, Shaw GM, Quinn TC.
Unsuspected primary human immunodeficiency virus type 1 infection in sero-
negative emergency department patients. J Infect Dis 1994; 170:194197.
94. Busch MP, Alter HJ. Will human immunodeficiency virus p24 antigen screening
increase the safety of the blood supply and, if so, at what cost? Transfusion 1995;
35:536539.
278 Bianco and Rios
95. Korelitz JJ, Busch MP, Williams AE. Antigen testing for human immunodeficiency
virus (HIV) and the magnet effect: Will the benefit of a new HIV test be offset by
the numbers of higher risk, test-seeking donors attracted to blood centers? Retro-
virus Epidemiology Donor Study. Transfusion 1996; 36:203208.
96. Centers for Disease Control and Prevention. U.S. Public Health Service guidelines
for testing and counseling blood and plasma donors for human immunodeficiency
virus type 1 antigen. MMWR 1996; 45 (No. RR-2):19.
97. Hewlett IK, Epstein JS. Food and Drug Administration conference on the feasibility
of genetic technology to close the HIV window in donor screening. Transfusion
1997; 37:346351.
98. Busch MP, Eble BE, Khayam-Bashi H, Heilbron D, Murphy EL, Kwok S, Sninsky
J, Perkins HA, Vyas GN. Evaluation of screened blood donations for human
immunodeficiency virus type 1 infection by culture and DNA amplification of
pooled cells. N Engl J Med 1991; 325:15.
99. Yerly S, Pedrocchi M, Perrin L. The use of polymerase chain reaction in plasma
pools for the concomitant detection of hepatitis C virus and HIV type 1 RNA.
Transfusion 1998; 38:908914.
100. Cardoso MS, Koerner K, Kubanek B. Mini-pool screening by nucleic acid testing
for hepatitis B virus, hepatitis C virus, and HIV: preliminary results. Transfusion
1998; 38:905907.
101. Schottstedt V, Tuma W, Bunger G, Lefevre H. PCR for HBV, HCV and HIV-1
experiences and first results from a routine screening programme in a large blood
transfusion service. Biologicals 1998; 26:101104.
102. Murthy KK, Henrard DR, Eichberg JW, Cobb KE, Busch MP, Allain JP, Alter HJ.
Redefining the HIV-infectious window period in the chimpanzee model: evidence
to suggest that viral nucleic acid testing can prevent blood-borne transmission.
Transfusion 1999; 39:688693.
103. Centers for Disease Control and Prevention. Public Health Service Guidelines for
the Management of Health-Care Worker Exposures to HIV and Recommendations
for Postexposure Prophylaxis. MMWR 1998; 47(RR-7):128.
104. Bianco C, Kessler D. Donor notification and counseling. Management of blood
donors with positive test results. Vox Sang 1994; 67(suppl 3):255259.
105. Emmanuel JC, Bassett MT, Smith HJ, Jacobs JA. Pooling of sera for human
immunodeficiency virus (HIV) testing: an economical method for use in developing
countries. J Clin Pathol 1988; 41:582585.
106. Smillie JM, Ala FA. Reducing the cost of anti-HIV screening. J Virol Methods 1988;
19:181184.
107. AuBuchon JP, Birkmeyer JD, Busch MP. Department of Pathology. Cost-effective-
ness of expanded human immunodeficiency virus-testing protocols for donated
12
Hepatitis Viruses and
Blood Transfusion
Alfred M. Prince
Lindsley F. Kimball Research Institute, New York Blood Center, New
York, New York
This review will summarize our expanding understanding of the transmission of

hepatitis viruses by blood transfusion. In addition to the well-known hepatitis B
virus (HBV) and hepatitis C virus (HCV), the recently identified hepatitis G virus,
not yet definitively identified as a hepatitis-causing agent, and the most recently
identified TT virus (TTV) agent will be reviewed.
I. HEPATITIS B VIRUS
The classic studies of Krugman at the Willowbrook School were the first to identify
two distinct hepatitis agents, originally termed MS-1 and MS-2, now known as
hepatitis A and hepatitis B virus (1). This made it possible to identify an antigen
discovered in liver tissue in 1967 by immunofluorescence as a hepatitis Bspecific
antigen, (HBsAg) first called the S.H. antigen (2). Blumberg identified an antigen
in the serum of an Australian aborigine that was originally interpreted as a genetic
marker conferring susceptibility to a variety of diseases: leukemia, leprosy, and all
forms of hepatitis (3). However, eventually it became clear that the Australia
antigen was the surface coat of HBV and thus was hepatitis B specific (4).
Numerous studies revealed a statistical asociation between HBsAg and
hepatocellular carcinoma (e.g., Ref. 5). That these reflected an etiological relation-
ship was shown when prospective follow-up of HBsAg carriers and matched
controls revealed that the relative risk of development of hepatocellular carcinoma
was about 224:1 (6), establishing HBV as the leading oncogenic virus in the world.
279
280 Prince
A. Modes of Transmission of HBV

HBV is readily transmitted by sexual (7) and parenteral routes, such as unsteril-
ized syringes and needles. Primate experiments using HBV-susceptible gibbons
(8) and chimpanzees (9) have revealed that semen and saliva can transmit HBV
infection when administered subcutaneously and intravaginally (8).
In addition to the above modes of transmission, horizontal transmission
(i.e., that occurring without apparent parenteral, sexual, or perinatal exposure,) is
common, especially in the developing world (10). In Africa most children are
infected with HBV by the age of 5, and transmission from infected mothers is
rare (11). Horizontal transmission between uninfected and infected infants is the
likely mechanism. In institutions for the mentally retarded the virus is readily
transmitted between carriers, patients, and staff (12). The virus is also readi-
ly transmitted from carrier mothers to their infants, especially in Asia. Both
transplacemental and intravaginal routes have been postulated. Thus, numerous
modes of transmission exist other than blood transfusion.
B. Infectivity of Blood Containing HBsAg

Following the studies indicating an association between HBsAg (formerly called
Australia antigen, SH antigen, hepatitis-associated antigen), numerous studies
were carried out to assess the infectivity of HBsAg-containing blood. Holland
et al. reported that 50% of transfused patients receiving at least 1 HBsAg unit
developed icteric posttransfusion hepatitis, as compared to only 7% of control
patients receiving no HBsAg (13). These findings led to rapid institution of
screening for HBsAg in the developed world. Numerous assays have been used
to screen blood donors for HBsAg. The first of these was the Ouchterlony, or agar
gel diffusion, assay. The New York Blood Center was the first to screen all blood
for this antigen using this relatively insensitive assay. Subsequently a series of
more sensitive assays was introduced: counterelectrophoresis, passive hemagglu-
tination inhibition, radioimmunoassay, and enzyme immunoassays. The first
radioimmunoassay was stated to be 10 times as sensitive as the agar gel diffusion
test; however, it soon became clear that much of this increase was due to
nonspecificity of the radioimmune assay (14).
C. Use of Anti-HBc for the Detection of

HBsAg-Negative, HBV-Infective Blood
In a report that remains provocative and controversial, Hoofnagle et al. found four
cases of posttransfusion hepatitis B in which no HBsAg- or anti-HBscontaining
blood had been transfused, although each had received a unit with antibody to the
hepatitis core (HBcAg) (15). This study raised the possibility that anti-HBc alone
Hepatitis Viruses and Blood Transfusion 281
could be a marker of infectivity. This hypothesis was supported recently by Chung

et al., who found that exclusion of donors with isolated anti-HBc had the highest
sensitivity and specificity (66.7 and 96.1%, respectively) for prevention of
posttransfusion B (16). This hypothesis is also in accord with a more recent
finding. Rehermann et al. reported that HBV DNA is detectable up to decades
after a primary, apparently self-limited, HBV infection (17). It would not be
surprising that HBV DNA, which is largely if not entirely in the HBV virion (the
Dane particle), would be a marker of infectivity.
Although anti-core testing may detect donors who are HBsAg negative and
infectious, this involves considerable loss to the blood supply. It has been
estimated that the prevalence of anti-HBc in the U.S. blood supply averages 2.6%,
ranging from 0.55 to 6.38% in different geographic regions (18). The loss would
be approximately 300,000 units per year in the United States. Nevertheless, this
test is now in use, primarily as a surrogate maker for HIV and for prevention of
transmission of HBV (19). The radioimmunoassay (Corab, Abbott Labs, North
Chicago, IL) is at least twice as specific as the more widely used enzyme
immunoassay method (Corzyme, Abbott Labs, North Chicago, IL) and thus
would reduce blood wastage (20).
D. The Estimated Risk of Transmission of

HBV Infection in the United States
Schreiber et al. have reported a very sophisticated approach for estimating the
risk of transmission of bloodborne agents in 1996, when second- and third-
generation screening assays were in place (21). Using data from 586,507 blood
donors who donated more than once between 1991 and 1993, they estimated the
incidence of each of the infections screened for in subjects negative when initially
tested. These were then adjusted for the duration of the window period, i.e., the
period while the infection is inclubating but serological markers are not yet
positive. Using this approach they estimated that the risk of donating blood during
the HBV incubation period was 1 in 63,000 (31,000147,000). It was estimated
that addition of polymerase chain reaction (PCR) to routine screening would
reduce the window period by 25% and that this would reduce the overall risk by
42% or 81 cases per 12 million units transfused. These estimates may be low,
especially if HBV DNA persisting for decades after acute infections, as demon-
strated by the results of Rehermann et al. (17), is infectious.
E. Impact of PCR Screening on the Transmission of HBV

As indicated above, Schreiber et al. (21) estimated that PCR screening would
reduce the transmission of HBV by 42%. Sankary et al. tested 375 units of plasma
with alanine aminotransferase (ALT) elevations and negative HBsAg tests and
282 Prince
estimated that these would be derived from 47,500 blood donations having both
normal and abnormal ALT levels (22). Only one unit tested positive by PCR, and
even in this case contamination could not be excluded since seroconversion did
not occur and PCR testing was negative on follow-up. If the PCR result is correct,
one positive in 47,500 donations would correspond to 252 PCR positives in the
12 million units of blood transfused in the United States annually. This would be
a higher estimate than the 81 cases preventable by PCR testing given by Schreiber
et al. (21).
II. HEPATITIS C VIRUS

By 1974 a prospective posttransfusion follow-up study had provided definitive
evidence that most transfusion-related hepatitis was not related to HBV. The au-
thors somewhat prematurely coined the term hepatitis C virus for the agent of
these infections (23). The non-B cases did not resemble hepatitis A epidemiolog-
ically and were subsequently shown not to be hepatitis A by immune electron
microscopy (24). The generally accepted terminology for these cases became
non-A, non-B (NANB) hepatitis, recognizing that more than one agent could be
involved. Investigators at Chiron undertook to clone the genome from an NANB
agent (25). After 5 years of discouragement they succeeded in cloning the first
NANB-specific clone (5-1-1) using expression cloning and a human source of
antibody. Because the sequence of the clone was unique, calling the cloned virus
HCV was justified. These workers rapidly cloned and sequenced most of the viral
genome, revealing it to have close similarity in genome organization, and in
hydrophilicity, to the flaviviruses. Ligation of four different cloned sequences and
expression of these provided the first generation (C100-3) of anti-HCV diagnostic
tests. Addition of capsid (C22), NS3, and NS5 proteins provided the second- and
third-generation assays generally used today for diagnosis and for blood screen-
ing. A prospective posttransfusion follow-up study estimated that first-generation
assays reduced the incidence of HCV infection by 86%, with a further reduction
due to the use of second-generation assays of 57% (26). These improved assays
dramatically reduced the incidence of posttransfusion transmission of HCV (27),
but the usefulness of the NS5 component has been questioned, because its
inclusion did not yield additional positives confirmable by PCR and by peptide
assays (28). PCR assays are useful for confirmation of the results of the serolog-
ical assays, although only 93% of serologically confirmed assays are PCR
positive (29). Presumably the PCR-negative cases were self-limited. A large
posttransfusion study in which 50 probable and 11 possible cases of non-B were
observed was retrospectively analyzed by testing of antibody to C100-3, capsid
polypeptides, and RT-PCR (23,30). All of the probable cases and five of the
possible cases were shown to be due to HCV infection. This study suggested that
all bloodborne non-A, non-B infections in the population studied were likely due
to HCV.
Application of anti-HCV tests rapidly confirmed the importance of these
assays: 80% of cases of chronic posttransfusion hepatitis from different parts of
the world were anti-C100-3 positive (25). Self-limited cases tended to show
transient or no antibody by this assay (31,32). HCV infections are characterized
by a high rate (6080%) of chronic infections. Chronic liver disease in HCV
infections may be manifested as chronic active hepatitis, cirrhosis, and hepato-
cellular carcinoma (33). HCV infections differ in that respect from HBV infec-
tions, where only about 5% of adult infections become chronic.
A strip radio immunoblot assay (RIBA) was manufactured by Chiron
to provide some measure of confirmatory specificity for the ELISA assays.
The concordance between third-generation enzyme immunoassays and the third-
generation immunoblot assay was 96% (34). The third-generation assay is con-
siderably more sensitive than the second-generation assays. Fifty-seven sera with
indeterminate results with the second-generation assay were retested with the
third-generation assay. Thirty-three (57.9%) showed at least one additional band
with the third-generation assay and thus were classified as positive (35).
A. Should ALT Testing Be Used for the Prevention of

HCV Posttransfusion Hepatitis?
It has been estimated that ALT testing would detect about three window-phase
donations per million units of blood transfused after institution of second-
generation screening assays. This contrasts with ALT elevations of approximately
1800 units per million units prior to anti-HCV testing. Only 8 of 10,000 units with
ALT elevations and negative anti-HCV tests are estimated to be infected with
HCV. The cost of ALT testing per year of life saved was calculated to be
$7.9 million. The authors concluded that ALT screening of volunteer blood
donors should be stopped (36). A policy that ALT could be discontinued was
adopted in the United States (19), but European fractionators are still required
to use plasma with normal ALT levels, thus U.S. blood centers still test for
ALT in order to comply with European requirements for plasma shipped to
European fractionators.
B. Sources of HCV Infection Other than Transfusion

In most cases transfused individuals who develop HCV infection will have been
infected by transfused blood; however, this is not always the case. Probably the
most common source of HCV infection worldwide is intravenous drug abuse (37).
If a transfused patient is a drug user or the spouse of a drug user, this could give
rise to HCV infection in a transfusion recipient. Surprisingly, drug use in donors
284 Prince
who had denied drug use during predonation questioning was found to be a major
factor in HCV transmission (38). Intrafamilial transmission of HCV among
spouses has been documented in some (39,40) but not all studies (41). Sex
partners whose spouses were anti-HCV positive were 3.7 times as likely to be
anti-HCV positive as partners of seronegative spouses (42). HCV RNA has been
detected, albeit in very low concentration, in both saliva and urine (43) and rarely
in breast milk (44). Vertical transmission from infected mothers appears to be
extremely rare, except when the mother is coinfected with HIV (45,46). Surgical
intervention, use of nondisposable needles or syringes, and dental therapy have
been implicated in the transmission of HCV (47,48). All of the above possible
routes of transmission need to be evaluated in the analysis of posttransfusion
cases, especially in cases where blood has been screened with second- and third-
generation assays. It should be remembered that over 90% of HCV infections
have been acquired outside the transfusion setting and current testing has reduced
transfusion-transmitted HCV to an extremely low level (33).
C. Impact of PCR Screening on Transmission of HCV

Cases negative by ELISA and RIBA assays, but positive by PCR, were identified
(34). Such cases were reported to be identifiable among blood donors (49), but a
subsequent study including the sera tested in the first study failed to confirm this
finding, suggesting PCR contamination in the first study (50).
Window period analysis has suggested that the risk of HCV transmission
of blood tested by second-generation assays is 1:103000 (28,000288,000) and
that this would be reduced by 72% following introduction of PCR testing (21).
This would seem to be a very minimal estimate.
PCR testing for HCV was introduced in the United States in 1999 under
Investigational New Drug (IND) protocols. Samples from 16128 units are
pooled before testing. Positive pools are then resolved to find the individual
positive sample.
D. Immunity in HCV Infections

Both in humans and in rechallenged chimpanzees, there is evidence of absent or
weak immunity after repeated exposure to even the same isolate of HCV (51).
Wyatt et al. analyzed this phenomenon in the extreme situation of homologous
challenge of chronically infected chimpanzees (52). Two factors were identified
as playing a role in the reinfection of these animals. In some cases immunity
seemed to be limited to major quasispecies in the inoculum, thus reinfection was
accompanied by emergence of minor quasispecies present in the inoculum.
In other cases fitness of certain quasispecies emerging after challenge was
postulated. The lack of immunity in HCV infections must be taken into consid-
eration when postulating reactivation of chronic HCV infection because many

such cases probably represent dual infections.
E. Course of Hepatitis C
Hepatitis C viremia, as detected by the presence of HCV RNA, occurs within 13
weeks after exposure. About 90% of patients will have detectable anti-HCV 3
months after infection. Chronic hepatitis C is a common clinical syndrome in the
United States and is now the most common reason for liver transplantation. At
least 70% of patients have persistent or intermittent ALT elevations on long-term
follow-up, with liver biopsies showing chronic inflammatory changes. The virus
persists because it continuously evolves into differing variants (quasispecies) as
neutralizing antibodies are developed by the host. About 80% of patients with
chronic HCV will eventually develop cirrhosis and 15% will develop hepato-
cellular carcinoma after 20 years. It is not understood why 15% of those infected
with HCV spontaneously recover or why patients with chronic disease may have
markedly different clinical courses.
Interferon- has been used in the treatment of HCV for several years.
Unfortunately, the treatment results in a sustained response in only 1020% of
patients. Almost all patients who receive interferon experience an uncomfortable
flu-like syndrome early in the course of treatment. More recently, interferon-
and the antiviral agent ribavirin have been given as combined therapy. Sustained
response rates have approximately doubled (to 4050%) using this combination.
There is some controversy about which patients should be treated. However, most
clinicians agree that patients with a persistently elevated ALT, positive HCV
RNA, and a liver biopsy showing portal or bridging fibrosis should be treated.
III. NON-ABC HEPATITIS

In recent years there has been considerable interest in attempting to characterize
a putative non-A, non-B, non-C, non-D, non-E virus causing posttransfusion
hepatitis. Some have termed this elusive entity hepatitis F. In a very careful study,
Japanese workers tested the hypothesis that this entity reflected silent HBV
infection. PCR for HBV identified this agent in 18 of 20 cases of acute and 17 of
20 chronic cases of putative hepatitis F. Sequencing revealed T-to-C mutations in
DR2 and an eight nucleotide deletion of the 3 terminus of the X genecoding
region. These mutations were thought to lead to suppression of replication
and expression of the HBV DNA (53). Clearly, this was not the elusive hepatitis
G or F.
Alter and Bradley (54) summarized the evidence for the existence of a
non-ABC form of hepatitis: Bradleys chimpanzee studies revealed a putative
286 Prince
chloroform-resistant agent that did not induce tubular ultrastructural changes,

such as those produced by HCV. This observation was, however, never confirmed.
Furthermore, HCV sometimes resists the effects of chloroform, presumably
because the chloroform does not always penetrate to all parts of the infectious
material (Prince and Brotman, unpublished data). The chloroform-resistant agent
appeared to have a diameter of 27 nm, which is very similar to that of the
core of HCV. The 27 nm agent was not shown to be a human hepatitis agent
by development of antibodies in inoculated chimpanzees. Putative clinical evi-
dence for the existence of an additional non-ABC agent was the observation
of multiple episodes of hepatitis in patients. This cannot be taken as evidence
for a non-ABC agent in the light of the weak or absent immunity to HCV,
described above.
Further putative evidence for the existence of non-ABC agent(s) was drawn
from posttransfusion follow-up studies. In three such studies 40, 12, and 11% of
NANB cases were classified as non-ABC by serlogical exclusion (reviewed in
Ref. 54). These observations may reflect lenient criteria for the diagnosis of
hepatitis because when the incidence of non-ABC hepatitis was compared in
patients receiving allogeneic and autologous transfusions, the incidence was
found to be almost identical (0.55 vs. 0.51%). Thus, a high proportion, or even
all, of the non-ABC cases may represent cases with low levels of transaminase
elevation related to surgery and hospitalization, rather than to viral infection.
An additional factor is that most of these cases were diagnosed by the relatively
insensitive first-generation assays and PCR was not done.
Additional evidence supporting the conclusion that posttransfusion studies
did not identify non-ABC cases is the fact that, in a retrospective posttransfusion
follow-up study in which very strict criteria were used to define probable viral
hepatitis, each of 50 such cases could be identified as HCV based on serology
and PCR (23,30).
In an ongoing study, Alter and Bradley (54) review cases of chronic
hepatitis that serologically appear to be due to a non-ABC agent. These cases are
less severe than HCV cases clinically, except in reports from Greece, and have a
lower rate of development of long-term chronicity. These findings could reflect
infection with a non-ABC agent or, alternatively, could reflect very mild HCV
cases, which are known to have a tendency to be HCV seronegative, or only
transiently seropositive, and which tend to be self-limited and thus unlikely to
progress to chronicity (31).
IV. HEPATITIS G VIRUS

A viral agent of putative human origin, the GB agent, was passaged into tamarins
and marmosets, producing severe short incubation hepatitis by the sixth passage.
Investigators from Abbott Laboratories used representational difference analysis

and this material to isolate clones from two flavi-like viruses, GBV-A and
GBV-B (55). Both of these agents turned out to be of tamarin origin. Shortly
therafter these investigators isolated a novel human flavivirus GB virus C (56).
A similar, if not identical, virus termed hepatitis G virus (HGV) was inde-
pendantly isolated by another group of investigators (57). These viruses (re-
viewed in Ref. 58) are positive single-stranded RNA viruses containing about
9400 nucleotides, have the genomic organization of flaviviruses, and are distantly
related to HCV. GB virus C and HGV share 96% of their deduced amino acid
sequences and may thus be considered different isolates of the same virus.
One to two percent of voluntary blood donors have HGV RNA (59), and the
infection is clearly transmissible by transfusion. HGV frequently exists as a mixed
infection with HBV, HCV, or HIV but does not appear to intensify disease caused
by the other viruses.
It is not yet clear whether HGV itself causes hepatitis. In Alters study there
were 35 HGV infections among 357 transfusion recipients. Only three had
hepatitis with HGV as the sole viral marker, and in these the hepatitis was mild
(59). Long-term follow-up of HGV infections has revealed that 75100% became
chronically infected (60) but that chronic hepatitis did not develop in any patient
infected with HGV alone (61). The findings of this study did not support HGV
as an etiological agent of non-ABC hepatitis. However, it has been reported that
HGV sequences were detected by PCR in three of six cases of fulminant hepatitis
(62). This observation must be pursued, especially since most cases of fulminant
hepatitis are serologically non-ABC.
It is of interest that both HCV and HGV associate with lipoproteins and are
therefore largely precipitated by antibodies to B lipoprotein but not by anti-IgG
(63). This property provides a defense against antibody-mediated neutralization
and may contribute to the high rate of chronic infection by these viruses.
V. THE TTV AGENT

A surprising and possibly important finding was recently reported from the
laboratory of M. Mayumi in Japan. Using representational difference analysis,
these workers isolated a viral clone (N22) of 500 nucleotides from a patient with
posttransfussion non-ABC hepatitis (64). Primers made to the sequence of the
clone permitted detection of the clone in serum and characterization of the density
of the associated virions, designated the TT virus (1.26 g/cm2). The nucleic acid
of this particle was sensitive to DNase I and was equally amplified with and
without reverse transcription, thus indicating the genome to contain DNA.
The TTV DNA sequence was not detectable by PCR in human leukocytes or
human placenta. Thus the TTV sequence was not of host origin. This conclusion
288 Prince
was also drawn from the fact that no homology to the TTV sequence was found
among 1,731,752 DNA sequences and 154,072 protein sequences in the National
Institute of Genetics (Mishima, Japan) database.
TTV DNA was detected in three of five patients with non-A to G post-
transfusion hepatitis. This virus may belong to the parvovirus group. Its role in
posttransfusion hepatitis deserves further study. The data so far presented suggest
that TTV is an extremely common infection, and that it is not associated with
acute or chronic hepatitis.
VI. APPROACHES TO REDUCTION IN THE

COST OF ANTIVIRAL SCREENING
The cost of antiviral screening in the developed world is high and renders these
tests beyond the reach of many developing world countries. One approach that
could be considered is the simpification of the screening tests themselves and
their manufacture in selected centers in the developing world. Dipstick assays
lend themselves to this kind of application (65). The Program for Appropriate
Health Care (PATH) in Seattle, Washington, has instituted development of
dipstick diagnostics for use in blood screening and transfers the manufacturing
responsibility to sites in the developing world. PATH maintains responsibility for
quality control.
It is clearly desirable that alternate and less expensive assays approach the
sensitivity and reliability of present second- and third-generation assays, if
possible. However, if the choice is no testing versus a slight reduction in
sensitivity, this should be made at the level of the developing country. In all cases
panels of test sera should be provided by international organizations to assess test
performance.
One means of cost reduction is to use two sequential ELISA assays and to
limit the use of confirmatory assays, or PCR, to those samples positive in both
screening assays (66). An evaluation of this strategy found no samples with
discordant screening results to be positive in RIBA or PCR tests.
VII. RESIDUAL RISK OF POSTTRANSFUSION

HEPATITIS IN THE DEVELOPED WORLD
There has been an extraordinary decline in the risk of posttransfusion hepatitis
during the past 20 years. This has been due to largely better donor selection
including avoidance of paid commercial donors and avoidance of high-risk
donors through thorough questioning and self-exclusion policies. Further reduc-
tions have been due to specific serological tests for HCV and the progressive
improvement of these tests and those for HBsAg.
To what extent will PCR tests, shortly to be introduced, prevent all residual
posttransfusion hepatitis? This important question will doubtless be clarified
in the next few years. It is not unlikely that posttransfusion hepatitis will soon
be eliminated. Reviews such as the present will then, at best be of only histori-
cal interest.
REFERENCES
1. Krugman S, Giles JP, Hammond J. Infectious hepatitis. Evidence for two distinc-
tive clinical, epidemiological and immunological types of infection. JAMA 1967;
200:365373.
2. Prince AM, Fiji H, Gershon RK. Immunohistochemical studies on the etiology of
anicteric hepatitis in Korea. Am J Hyg 1964; 79:365381.
3. Blumberg BS, Gerstley BJS, Hungerford DA, London WT Sutnick AI. A serum
antigen (Australia antigen) in Downs syndrome, leukemia and hepatitis. Ann Int Med
1967; 66:924931.
4. Prince AM. An antigen detected in the blood during the incubation period of serum
hepatitis Proc Natl Acad Sci USA 1968; 60:814821.
5. Prince AM, Szmuness W, Michon J, Demaille J, Diebolt G, Linhard J, Quenum C,
Sankale M. A case/control study of the association between primary liver cancer and
hepatitis B infection in Senegal. Int J Cancer 1975; 16:376383.
6. Beasley RP, Lin CC, Hwang L-Y, Chien C-S. Risk of hepatocellular carcinoma in
hepatitis B infection: a prospective study in Taiwan. In: Szmuness W, Alter HJ,
Maynard JE, eds. Viral Hepatitis. 1981 International Symposium. Franklin Institute
Press, 1982:261272.
7. Szmuness W, Much MI, Prince AM, Hoofnagle JH, Cherubin CE, Harley EJ, Block
GH. On the role of sexual behavior in the spread of hepatitis B infection. Ann Int
Med 1975; 83:489495.
8. Scott RM, Snitbhan R, Bancroft WH, Alter HJ, Tingpalapong M. Experimental
transmission of hepatitis B virus by semen and saliva. J Infec Dis 1980; 142:6771.
9. Alter HJ, Purcell RH, Gerin JL, London WT, Kaplan PM, McAuliffe WJ, Wagner J,
Holland PV. Transmission of hepatitis B to chimpanzees by hepatitis B surface
antigen-positive saliva and semen. Infect Immun 1977; 16:928933.
10. Davis LG, Weber DJ, Lemon SM. Horizontal transmission of hepatitis B virus.
Lancet 1989; i:889893.
11. Prince AM, White T, Pollock N, Riddle N, Brotman B, Richardson L. The epidemi-
ology of hepatitis B infections in Liberian infants. Infect Immun 1981; 32:675680.
12. Breuer B, Friedman SM, Millner ES, Kane MA, Synder RH, Maynard JE. Transmis-
sion of hepatitis B virus to classroom contacts of mentally retarded carriers. JAMA
1985; 254:31903195.
13. Holland PV, Alter HJ, Purcell RH, Walsh JH, Morrow AG. The infectivity of blood
containing the hepatitis associated antigen. In: Prier JE, Freidman H, eds. Australia
Antigen. Baltimore: University Park Press, 1973:191203.
290 Prince
14. Prince AM, Brotman B, Jass D, Ikram H. Specificity of the direct radioimmunoassay
for the detection of hepatitis B antigen. Lancet 1973; i:13461350.
15. Hoofnagle JH, Seeff LB, Bales ZB, Zimmerman HJ, Veterans Administration Hepa-
titis Cooperative Study Group. Type B hepatitis after transfusion with blood contain-
ing antibody to hepatitis B core antigen. N Engl J Med 1978; 298:13791383.
16. Chung HT, Lee JSK, Lok ASF. Prevention of posttransfusion hepatitis B and C by
screening for antibody to hepatitis C virus and antibody to HBcAg. Hepatology 1993;
18:10451049.
17. Rehermann B, Ferrari C, Pasqunelli C, Chisari FV. The hepatitis B virus persists for
decades after recovery from acute viral hepatitis despite active maintenance of a
cytotoxic T lymphocyte response. Nature Med 1996; 2:11041108.
18. Kline WE, Bowmann RJ, McCurdy KKE, OMalley JP, Sandler SG. Hepatitis B core
antibody (anti-HBc) in blood donors in the United States: implications for surrogate
testing programs. Transfusion 1987; 27:99102.
19. Infectious disease testing for blood transfusions. Consensus development conference
statement. National Institutes of Health, June 911, 1995.
20. Hanson MR, Polesky HF. Evaluation of routine anti-HBc screening of volunteer
blood donors: a questionable surrogate test for non-A, non-B hepatitis. Transfusion
1987; 27:107108.
transmitted viral infections. N Engl J Med 1996; 334:16851690.
22. Sankary TM, Yang G, Romeo JM, Ulrich PP, Busch MP, Rawal BD, Vyas GN. Rare
detection of hepatitis B and hepatitis C virus genomes by polymerase chain reaction
in seronegative donors with elevated alanine aminotransferase. Transfusion 1994;
34:656660.
23. Prince AM, Brotman B, Grady GF, Kuhns WJ, Hazzi C, Levin RW, Millian SJ. Long
incubation posttransfusion hepatitis without serological evidence of exposure to
hepatitis B virus. Lancet 1974; ii:241246.
24. Feinstone SM, Kapikian AZ, Purcell RH, Alter HJ, Holland PV. Transfusion associ-
ated hepatitis not due to viral hepatitis type A or B. N Engl J Med 1975; 292:767770.
25. Choo Q-L, Weiner AJ, Overby LR, Kuo G, Houghton M, Bradley DW. Hepatitis C
virus: the major causative agent of viral non-A, non-B hepatitis. Br Med Bull 1990;
46:423441.
26. Gonzalez A, Esteban JI, Madoz P, Viladomiu L, Genesca J, Muniz E, Enriquez J,
Torras X, Hernndez JM, Quer J, Vidal X, Alter HJ, Shih JW, Esteban R, Guardia J.
Efficacy of screening donors for antibodies to the hepatitis C virus to prevent
transfusion-associated hepatitis: final report of a prospective trial. Hepatology 1995;
22:439445.
27. Donahue JG, Munoz A, Ness PM, Brown DE Jr, Yawn DH, McAllister HA Jr, Reitz
BA, Nelson KE. The declining risk of post-transfusion hepatitis C virus infection.
N Engl J Med 1992; 327:369373.
28. Uyttendaele S, Claeys H, Mertens H, Verhaert H, Vermylen C. Evaluation of
third-generation screening and confirmatory assays for HCV antibodies. Vox Sang
1994; 66:122129.
29. Silva AE, Hosein B, Boyle RW, Fang CT, Shindo M Waggoner JG, Hoofnagle JH,
DiBisceglie AM. Diagnosis of chronic hepatitis C: comparison of immunoassays and

the polymerase chain reaction. Am J Gastroenterol 1994; 89:493496.
30. Prince AM, Brotman B, Inchausp, Pascual D, Nasoff M, Hosein B, Wang CY.
Patterns and prevalence of hepatitis C virus infection in posttransfusion non-A, non-B
hepatitis. J Infect Dis 1993; 167:12961301.
31. Prince AM, Brotman B, Huima T, Kraduledat P, Houghton M, Kuo G, Choo Q-L,
Polito A, diNello R, Nelles MJ. Distinction between chronic and self-limited forms
of hepatitis C infection. In: Shikata T, Purcell RH, Uchida Y, eds. Viral Hepatitis C,
D, and E. Amsterdam: Elsevier Science Publishers, 1991:716.
32. Yuki N, Hayashi N, Kasahara A, Hagiware H, Ohkawa K, Fusamoto H, Kamada T.
Hepatitis C virus replication and antibody responses toward specific hepatitis C virus
proteins. Hepatology 1994; 19:13601365.
33. Alter MJ. Review of serologic testing for hepatitis C virus infection and risk of
posttransfusion hepatitis C. Arch Path Lab Med 1994; 118:342345.
34. Craxi A, Valenza M, Fabiano C, Magrin S, Fiorentino G, Diquattro O, Pagliaro L.
Third-generation hepatitis C virus tests in asymptomatic anti-HCV-positive blood
donors. J Hepatol 1994; 21:730734.
35. Buffet C, Charnaux N, Laurent-Puig P, Chopineau S, Quichon JP, Briantais MJ,
Dussaix E. Enhanced detection of antibodies to hepatitis C virus by use of a
third-generation recombinant immunoblot assay. J Med Virol 1994; 43:259261.
36. Busch MP, Korelitz JJ, Kleinman SH, Lee SR, AuBuchon JP, Schreiber GB, Retro-
virus Epidemiology Donor Study. Declining value of alanine aminotransferase in
screening of blood donors to prevent posttransfusion hepatitis B and virus infection.
Transfusion 1995; 35:903910.
37. Verbaan H, Andersson K, Eriksson S. Intravenous drug abusethe major route of
hepatitis C virus transmission among alcohol-dependent individuals. Scand J Gas-
troenterol 1993; 28:714718.
38. Conry-Cantilena C, VanRaden M, Gibble J, Melpolder J, Shakil AO, Viladomiu L,
Cheung L, DiBisceglie A, Hoofnagle J, Shih JW, Kaslow R, Ness P, Alter HJ. Routes
of infection, viremia, and liver disease in blood donors found to have hepatitis C virus
infection. N Engl J Med 1996; 334:16911696.
39. Oshita M, Hayashi N, Kasahara A, Yuki N, Takehara T, Hagiwara H, Hayakawa Y,
Yasumatsuya Y, Kishida Y, Fusamoto H, Kamada T. Prevalence of hepatitis C virus
in family members of patients with hepatitis C. J Med Virol 1993; 41:251255.
40. Lissen E, Alter HJ, Abad MA, Torres Y, Perez-Romero M, Leal M, Pineda JA,
Torronteras R, Snchez-Quijano A. Hepatitis C virus infection among sexually
promiscuous groups and the heterosexual partners of hepatitis C virus infected index
cases. Eur J Clin Microbiol Infect Dis 1993; 12:827831.
41. Bresters D, Mauser-Bunschoten EP, Reesink HW, Roosendall G, van der Poel CL,
Chamuleau RAFM, Jansen PLM, Weegink CJ, Cuypers HTM, Lelie PN, van den
Berg HM. Sexual transmission of hepatitis C virus. Lancet 1993; 342:210211.
42. Thomas DL, Zenilman JM, Alter Shih JW, Galai N, Carella AV, Quinn TC. Sexual
transmission of hepatitis C virus among patients attending sexually transmitted
diseases clinics in Baltimorean analysis of 309 sex partnerships. J Infect Dis 1995;
171:768775.
43. Numato N, Ohori H, Hayakawa Y, Saitoh Y, Tsunoda A, Kanno A. Demonstration of
292 Prince
hepatitis C virus genome in saliva and urine of patients with type C hepatitis. J Med
Virol 1993; 41:120128.
44. Zimmermann R, Perucchini D, Fauchere J-C, Joller-Jemelka H, Geyer M, Huch R,
Huch A. Hepatitis C virus in breast milk. Lancet 1995; 345:928.
45. Roudot-Thoraval F, Pawlotsky J-M, Thiers V, Deforges L, Girollet P-P, Guillot F,
Huraux C, Aumont P, Brechot C, Dhumeaux D. Lack of mother-to-infant transmis-
sion of hepatitis C virus in human immunodeficiency virus-seronegative women: a
prospective study with hepatitis C virus RNA testing. Hepatology 1993; 17:772777.
46. Zanetti AR, Tanzi E, Paccagnini S, Principi N, Pizzocolo G, Caccamo ML, DAmico
E, Cambi G, Vecchi L, Lombardy Study Group on Vertical HCV Transmission.
Mother-to-infant transmission of hepatitis C virus. Lancet 1995; 345:289291.
47. Mele A, Sagliocca L, Manzillo G, Converti F, Amoroso P, Stazi MA, Ferrigno L,
Rapicetta M, Franco E, Adamo B, Palumbo F, Sbreglia C, Pan A, Pasquini P, SEIVA
Collaborating Group. Risk factors for acute non-A, non-B hepatitis and their relation-
ship to antibodies for hepatitis C virus: a case-control study. Am J Public Health 1994;
84:16401643.
48. Chen T-Z, Wu J-C, Yen F-S, Sheng W-Y, Hwang S-J, Huo T-I, Lee S-D. Injection
with nondisposable needles as an important route for transmission of acute commu-
nity-acquired hepatitis C virus infection in Taiwan. J Med Virol 1995; 46:247251.
49. Sugitani M, Inchauspe G, Shindo M, Prince AM. Sensitivity of serological assays for
detection of HCV viremic blood donors. Lancet 1992; 339:10181019.
50. Prince AM, Scheffel JW, Moore B. A search for hepatitis C virus polymerase chain
reaction positive but seronegative subjects among blood donors with elevated ami-
notransferase. Transfusion 1997; 37:211214.
51. Prince AM, Brotman B, Huima T, Pascual D, Jaffery M, Inchauspe G. Immunity in
hepatitis C infection. J Infect Dis 1992; 165:438443.
52. Wyatt CA, Andrus L, Brotman B, Huang F, Lee D-H, Prince AM. Immunity in
chimpanzees chronically infected with hepatitis C virus: role of minor quasispecies
in reinfection. J Virol 1998; In Press.
53. Uchida T, Shimojima M, Gotoh K, Shikata T, Tanaka E, Kiyosawa K. Silent
hepatitis B virus mutants are responsible for non-A, non-B, non-C, non-D, non-E
hepatitis. Microbiol Immunol 1994; 38:281285.
54. Alter HJ, Bradley DW. Non-A, non-B hepatitis unrelated to the hepatitis C virus
(non-ABC) Semin Liver Dis 1995; 15:110120.
55. Simons JN, Pilot-Matias TJ, Leary TP, Dawson GJ, Desai SM, Schlauder GG,
Muerhoff AS, Erker JC, Buijk SL, Chalmers ML, Van Sant CL, Mushahwa IK.
Identification of two flavivirus-like genomes in the GB hepatitis agent. Proc Natl
Acad Sci USA 1995; 92:34013405.
56. Simons JN, Leary TP, Dawson GJ, Pilot-Matias TJ, Muerhoff AS, Schlauder GG,
Desai SM. Isolation of novel virus-like sequences associated with human hepatitis.
Nat Med 1995; 1:564569.
57. Linner J, Wages J, Jr, Zhang-Keck ZY, Fry KE. Molecular cloning and disease
association of hepatitis G virus: a transfusion transmissible agent. Science 1996;
271:505508.
58. Miyakawa Y, Mayumi M. Hepatitis G virusa true hepatitis virus or an accidental
tourist? N Engl J Med 1997; 336:795796.
59. Alter HJ, Nakatsuji Y, Melpolder J, Wages J, Wesley R, Shih JW-K, Kim JP.
The incidence of transfusion-associated hepatitis G virus infection and its relation to
liver disease. N Engl J Med 1997; 336:747754.
60. Lefrere J-J, Loiseau P, Maury J, Lasserre J, Mariotti M, Ravera N, Lerable J, Lefevre
G, Morand-Joubert L, Girot R. Natural history of GBV-C/hepatitis G virus infection
through the follow-up of GBV-C/hepatitis G virus-infected blood donors and recipi-
ents studied by RNA polymerase chain reaction and anti-E2 serology. Blood 1997;
90:37763780.
61. Alter MJ, Gallagher M, Morris TT, Moyer LA, Meeks EL, Krawczynski K, Kim JP,
Margolis HS, Sentinel Counties Viral Hepatitis Study Team. Acute non-A-E hepatitis
in the United States and the role of hepatitis G virus infection. N Engl J Med 1997;
336:741746.
62. Yoshiba M, Okamoto H, Mishiro S. Detection of the GBV-C hepatitis virus genome
in serum from patients with fulminant hepatitis of unknown aetiology. Lancet 1995;
346:11311132.
63. Sato K, Tanaka T, Okamoto H, Miyakawa Y, Mayumi M. Association of circulating
hepatitis G virus with lipoproteins for a lack of binding with antibodies. Biochem
Biophys Res Commun 1996; 229:719725.
64. Nishizawa T, Okamoto H, Konishi K, Yoshizawa H, Miyakawa Y, Mayumi M.
A novel DNA virus (TTV) associated with elevated transaminase levels in post-
transfusion hepatitis of unknown etiology. Biochem Biophy Res Comm 1997;
241:9297.
65. Mulyanto DR, Suwignya S, Tsauri S, Itoh K, Mizui M, Tsuda F, Okamoto H,
Yoshizawa H, Mishiro. An easy dipstick assay for anti-core antibodies to screen
donors for hepatitis C virus viremia. Vox Sang 1996; 70:229231.
66. Allain JP, Kitchen A, Aloysius S, Reeves I, Petrik J, Barbara JAJ, Williamson LM.
Safety and efficacy of hepatitis C virus antibody screening of blood donors with two
sequential screening assays. Transfusion 1996; 36:401405.
13
Human T-Cell Leukemia Virus
Maria Rios and Celso Bianco

I. THE VIRUS: ISOLATION, CHARACTERIZATION,

AND BIOLOGICAL PROPERTIES
The human T-cell leukemia viruses type I (HTLV-I) and type II (HTLV-II) are
human retroviruses in the oncornavirus family. This family includes the human
immunodeficiency virus (HIV), the bovine leukemia virus (BLV), and the simian
T-cell leukemia virus (STLV) (1). HTLV-I was first isolated from a patient with
cutaneous T-cell lymphoma (2,3). The virus was quickly associated with human
diseases, such as adult T-cell leukemia, leading to the addition of the retrovirus
family to the category of human pathogens (27). HTLV-II was isolated from a
patient with a T-cell variant hairy cell leukemia (8).
Retroviruses are enveloped viruses that need to be transcribed into DNA
and integrated into the host cell genome in order to replicate. The viral particles
carry the enzyme reverse transcriptase (RT), which transcribes the viral RNA into
complementary DNA (cDNA). The cDNA is subsequently converted into double-
stranded DNA (dsDNA) and integrated into the host cell genome as proviral DNA
(9). The genomic material of retroviruses is composed of three major regions: gag
(group antigen), which encodes for structural proteins; pol, which encodes for
nonstructural proteins, such as polymerase and proteases; and env, which encodes
for envelope proteins. HTLV-I and HTLV-II also have a set of genes named
tax/rex that encode for proteins that regulate viral replication. The genomic
sequence of the virus is flanked by two long terminal repeats (LTR) with elements
that regulate transcription and join the host cell genome after integration (Fig. 1).
The tax protein increases the rate of transcription initiation by acting on the
promoter located in the 5LTR. The tax protein also transactivates heterologous
295
296
HTLV Genome
1000 8000
5LTR gag pol env tax 3LTR
p19 p24 p15 gp46 p21

MA CA NC SU TM
Protease p40/p37 tax
p27/p26 rex
Polymerase
Figure 1 Graphic representation of the HTLV genome and respective gene products.
5LTR-gag-pol-env-tax/rex-LRT3HTLV gene productsthe gag (group antigen) region
encodes for the following structural proteins: p19 (matrix), p24 (capsid), and p15 (nu-
cleocapsid). The pol region encodes for the nonstructural proteins protease and reverse
transcriptase. The env (envelope) region encodes for the envelope proteins gp46 (exposed to
the external surface of the virus) and for p21 (envelope transmembrane). The tax/rex region
encodes for p40 (tax) a regulatory, nonstructural protein that regulates activation of viral
transcription by interacting with LTR, and for the p27 (rex) nonstructural protein that
Rios and Bianco
regulates viral replication at postranscriptional level. (From Refs. 11,12.)

Human T-Cell Leukemia Virus 297
promoters such as the IL-2 and IL-2R promoter, the GM-CSF promoter, the c-fos
and c-cis promoter. The rex protein controls HTLV gene expression at the post
transcriptional level (10,11). Retroviruses infect a wide range of cellular types in
vitro. Fortunately, their ability to transform these cells is limited.
Unlike other retroviruses, HTLV has highly conserved genomic sequences
(12). The rate of base substitutions is estimated at approximately 1% per 1000
years (13,14). The diversity of the entire collection of HTLV sequences identified
to date is lower than that observed within a single patient in late stage of HIV
infection (1417).
II. EPIDEMIOLOGY
HTLV-I is found worldwide. However, it clusters preferentially around the
equatorial belt (Table 1). The prevalence of HTLV-I/II infection in the United
States is approximately 0.0004%, based on blood donor screening. About half of
the positive individuals are infected with HTLV-II (37).
The epidemiology of HTLV-II is not as well known as that of HTLV-I.
Infection is highly prevalent among intravenous drug users (IVDUs) in the United
States and in Europe (4346). Infection with HTLV-II is also frequent among
Native American populations from North, Central, and South America (4556),
and the prevalence can be as high as 2030% (54,57).
The major routes of transmission for HTLV-I and HTLV-II are sexual
contact and from mother to child through breast-feeding, leading to clear familial
clustering of HTLV-I carriers (56,58). HTLV-I infection has also been reported in
children who have not been breast-fed, indicating that other perinatal routes of
transmission have to be considered (60). Vertical transmission (mother-to-child)
occurs with a frequency of 1030% (29,58,61). HTLV-I/II and HIV share the
same routes of infection. However, HTLV-I/II appears to be less infectious than
HIV. This difference has been attributed to viral load in the course of infection
and to the fact that, in the host, HTLV-I always remain cell-associated while
HIV-1 is both cell-associated and cell-free (6264). The rate of infection among
females is higher than that among males (24,29). Studies involving steady
heterosexual couples documented significant concordance of HTLV-I serological
status (65). The efficacy of sexual transmission has been estimated at 53% (44).
Serologically discordant couples were predominantly female positive and male
negative, indicating higher efficiency of transmission from males to females than
from females to males (59,64,65). HTLV-I has been isolated from semen (66).
Transmission from females to males may be associated with other factors, such
as sexual intercourse during menses and bleeding during intercourse (44).
The modes of transmission of HTLV-II are less well documented than that
of HTLV-I. Sexual transmission of HTLV-II has been difficult to study because
298 Rios and Bianco
Table 1 Prevalence of HTLV-I Infection in Various Regions of the World
Region Countries Prevalence (%) Ref.
Africa Cameroon, Gabon, Ivory Coast, 110 1821

Kenya, Tanzania, Zaire
Caribbean islands Barbados, Guadeloupe, Haiti, 17 2125
Jamaica, Martinique, Tobago
Europe France, Italy, United Kingdom, Spain 0.1 19,2628
Japanese islands Kyushu, Okinawa 130 29
South and Central Argentina, Brazil, Bolivia, 0.310 3036
America Colombia, Honduras, Paraguay,
Peru, Venezuela, Uruguay
North America United States, Canada <0.1 37
Pacific islands Australia, Malaysia, Taiwan, 1.714 3842
Vietnam
of the frequent coincidence of intravenous drug abuse. For instance, studies of

female prostitutes have shown that intravenous drug abuse is the major risk factor
for seropositivity (4346). Studies of Guaymi Indians from Panama and Native
Americans from New Mexico show concordance of seropositivity among married
couples, similar to what has been observed with HTLV-I (47). Vertical transmis-
sion of HTLV-II through breast-feeding has also been reported (67,68). Parenteral
transmission through contaminated needles is observed among IVDUs (4446)
and in countries without disposable needles where sterilization of multiple use
needles is inadequate.
Transmission by blood and blood products will be discussed later in
this chapter.
III. ASSOCIATION OF HTLV-I INFECTION WITH DISEASE

The first two diseases identified as associated with HTLV-I were the hematolog-
ical malignancy known as adult T-cell leukemia (ATL) and the progressive
neurodegenerative disease known as tropical spastic paraparesis (TSP) in the
Caribbean and Africa (69,70), and as HTLV-Iassociated myelopathy (HAM) in
Japan. Other diseases reported to be associated with HTLV-I infection are
infectious dermatitis (71), uveitis (72), opportunistic infections of the lungs,
chronic renal insufficiency, and lymphadenopathy (73,74). It has been suggested
that malignant diseases are caused by monoclonal proliferation of cells containing
integrated provirus and that nonmalignant diseases result from polyclonal prolif-
eration of HTLV-Iinfected cells or are a consequence of the hosts immunologi-

cal reaction to the virus (7577).
A. Adult T-Cell Leukemia

Adult T-cell leukemia was first described in Japan in 1974 and recognized as a
specific disease in 1977 (78,79). The virus was first isolated in the United States
from leukocytes from an ATL patient in 1980 (1) and called HTLV. The virus was
also isolated in Japan in 1981 from a cell line derived from an ATL patient (MT-I)
and called adult T-cell leukemia virus (ATLV) (80). Documentation that the U.S.
and Japanese isolates were identical led to the agreement that the newly identified
retroviruses should be called HTLV-I (81,82).
ATL is a mature T-cell lymphoma characterized by monoclonal integration
of the virus into the leukemic cell genome (81,83). The leukemic phase of ATL
is characterized by circulating CD4+/CD25+ T cells (79). ATL develops after a
long incubation period. In Japan, the age of onset of ATL ranges from 24 to 85
years, with an average of 58 years, but in the Caribbean and South America onset
occurs at an average of 40 years (81,8385). A proportion of HTLV-Iinfected
individuals develop ATL after an extraordinary long period of incubation, 2030
years (86,87). The lifetime risk for individuals infected before the age of 20 is
5% (84). The male-to-female ratio is 1.4:1, opposite to that HTLV-I of asymp-
tomatic carriers. The major clinical findings at onset are enlargement of peripheral
lymph nodes, hepatomegaly, splenomegaly, and skin lesions. Hypercalcemia is
present in 32% of the patients. Immunosuppression accompanied by opportunistic
infections is also common. The white blood cell count varies from normal up to
5 105/L. Leukemic cells have cleaved or multilobular nuclei known as flower
cells (79). There are four clinical forms of ATL according to their clinical course:
acute, chronic, lymphomatous, and smoldering (82). The prognosis of ATL is very
poor, with a mean survival time in Japan of 6.2, 10.2, and 24.3 months,
respectively, for acute, lymphomatous, and chronic ATL. ATL does not respond
well to chemotherapy. Recent reports suggest that treatment with zidovudine and
interferon-alpha may be effective (86,88,89).
B. Tropical Spastic Paraparesis/

HTLV-IAssociated Myelopathy
Tropical spastic paraparesis/HTLV-Iassociated myelopathy (TSP/HAM) was
first associated with HTLV-I in 1985 in Martinique (69) and in 1986 in Africa
(70) and Japan (87). In Japan, this disease is strongly associated with history of
blood transfusions (90). The lifetime risk of development of TSP/HAM among
HTLV-Iinfected individuals is less than 2% (91). The main neurological features
of TSP/HAM are spasticity or hyperreflexia of the legs, disturbances of bladder
300 Rios and Bianco
function, weakness of leg muscles, and sensory disturbances. Spinal cord atrophy,
signs of funicular demyelinization, axonal loss, and gliosis are often found in the
lower thoracic spinal cord. Sometimes the white matter presents lesions in
magnetic resonance imaging.
Analysis of the cerebrospinal fluid shows pleocytosis, high titer of IgG,
and oligoclonality (92). Neuropathologically it appears as a typical immune-
inflammatory disease with infiltration of the parenchyma by CD4+ and CD8+ T
cells. In later stages, the number of inflammatory cells diminishes and they are
almost exclusively CD8+ T-cells. TSP/HAM patients have a very intense im-
munological response to HTLV-I antigens. The relationship between the immune
response and damage to the central nervous system is unclear. There is indication
that HTLV-I proviral load may be a factor, because TSP/HAM patients have 50
times more HTLV-I proviral DNA in peripheral blood lymphocytes than asymp-
tomatic carrier (93). Patients with TSP/HAM show some improvement after
treatment with oral corticosteroids.
C. Infective Dermatitis
Infective dermatitis was described in 1966 in Jamaican children. The association
with HTLV-I infection was made in 1990 by detection of viral genome in skin
biopsies (71). Infective dermatitis is characterized by crusted scabies, corneal
opacities, chronic bronchiectasis, parasitic worm infestation, and progression to
more severe HTLV-Iassociated diseases such as ATL and TSP/HAM (94,95).
D. HTLV-IAssociated Uveitis
HTLV-Iassociated uveitis is characterized by moderate to severe cellular infil-
tration of the vitreous body (vitreous opacity), mild iritis, and moderate retinal
vasculitis (72). Virological and molecular biological findings suggest that cyto-
kines produced by HTLV-Iinfected T cells in the eye play a central role in the
pathogenesis of the disease. The proviral load in blood lymphocytes from patients
with uveitis is higher than that of asymptomatic carriers (72). The inflammation
responds to topical and systemic treatment with corticosteroids and reoccurs in
about 60% of the cases after therapy is discontinued. Some patients present
typical Sjgrens syndrome (9698).
E. Other Diseases
Other diseases observed in HTLV-Iinfected individuals are opportunistic infec-
tions of the lungs, chronic renal insufficiency, lymphadenopathy, arthropathy, and
autoimmune diseases (99101). HTLV-Iassociated infiltrating pneumonitis has
also been reported in some individuals in Japan (100102).
IV. DISEASE ASSOCIATED WITH HTLV-II

HTLV-II has not been clearly associated with any disease. The original isolate of
HTLV-II came from a patient with hairy cell leukemia. Large surveys of patients
with hairy cell leukemia have identified only sporadic examples of HTLV-II
seropositivity (103105). Subsequently, several cases of large granular lympho-
cytic leukemia (LGLL), a T-cell malignancy with natural killer cell phenotype,
have been reported (106). However, several surveys of LGLL did not show clear
excess prevalence of HTLV-I antibodies or sequences of the HTLV-II genome in
these patients (107111).
New Mexico has a large Native American population with high prevalence
of HTLV-II infection. The incidence of lymphoproliferative disorders among
these individuals is not higher than that of the normal population. However, this
lack of association cannot be directly documented because the HTLV-II status of
the registered cases has not been determined (112). One case of HTLV-II
associated mycosis fungoides positive by PCR and negative for antibodies
has been reported (113). A syndrome of severe skin disease, eosinophilia,
and dermatopathic lymphadenopathy has been reported among intravenous
drug abusers co-infected with HTLV-II and HIV-1 (114). A growing number of
TSP/HAM cases associated with HTLV-II are being reported (115121). Some
of the case reports had features reminiscent of the ataxic form of TSP/HAM
reported from Jamaica. Cases have had oligoclonal bands in the CSF reminiscent
of what has been observed in HTLV-Ipositive cases.
An increased incidence of infectious diseases was observed during prospec-
tive follow-up of human T-lymphotropic virus type II and Iinfected blood
donors. Compared with seronegative controls, HTLV-II infection was associated
with an increased incidence of bronchitis, bladder and/or kidney infections, and
oral herpes. HTLV-I infection was associated with increased incidence of bladder
and/or kidney infection (122). Other possible conditions identified in medical
surveys of the Guaymi Indians, drug abuser and transfusion cohorts that are being
evaluated in association with HTLV-II are an adult polyarthritis, eczema of the
skin, and asthma.
V. TRANSMISSION OF HTLV-I AND HTLV-II BY

BLOOD AND BLOOD PRODUCTS
Retrospective studies performed in Japan have clearly documented the transmis-
sion of HTLV-I by transfusion of cellular components of blood, but not by plasma
or by plasma derivatives. The estimated frequency of infection by transfusion of
a seropositive cellular component was 63% (123). This study led to the im-
plementation of screening of United States blood donors for antibodies to
HTLV-I/II in December 1988 (124). The previously observed rate of transmission
302 Rios and Bianco
was confirmed by a large donor-recipient study performed in Jamaica prior to the

availability of screening tests. That study showed that the median time to
seroconversion among infected recipients was 51 days and that infectivity was
inversely proportional to the age of the transfused product, suggesting that the
loss of infectivity was associated with loss of viability of leukocytes (125).
Transmission of HTLV-I/II by transfusion seems to be rare and has been estimated
at 1:641,000 based on large studies of seroconversion among blood donors at four
major U.S. blood centers (126).
Transmission of HTLV-II by transfusion of cellular products has also been
documented (127,128). Retrospective studies based on lookback programs lead
to the identification of recipients infected by HTLV-II through the transfusion of
blood that had tested negative on an HTLV-I screening assay, showing the lower
sensitivity of HTLV-I based screening assays to detect antibodies to HTLV-II
(47,128,129). More sensitive assays containing specific HTLV-II epitopes have
since been licensed and are in use for blood donor screening (130).
While transmission of HTLV-I/II infection by transfusion is well docu-
mented in the literature, transmission of disease has been infrequent. There is one
report suggesting that two long-term survivors of hematological malignancies,
one with Hodgkins and one with acute promyelocytic leukemia, developed ATL
6 months and 11 years after blood transfusions, respectively (131). There are
several reported cases of TSP/HAM attributed to transfusion in the literature.
The incubation time varied from 6 to 24 months, with an average of 15 months
(132138).
VI. LABORATORY ASSAYS FOR HTLV-I/II

Blood donor screening for HTLV-I and HTLV-II was recommended by the U.S.
Food and Drug Administration (FDA) in December 1988, as screening assays for
antibodies to these viruses became licensed (124). All antibody-screening assays
licensed in the United States contain, in addition to HTLV-I antigens, HTLV-II
specific proteins (130) and are based on enzyme immunoassay (EIA) principles
(microtiter plates, beads, or microparticles containing antigens that are incubated
with serum or plasma of the prospective blood donor). Japan uses particle
agglutination assays based on the same principles and similar reagents.
Every year, 10,00015,000 U.S. blood donors are found to be repeatedly
reactive on commercial EIA screening assays for HTLV-I/II and require perfor-
mance of additional, more specific tests for confirmation of screening test results.
They include Western blots (WB), radioimmunoprecipitation assay (RIPA), re-
combinant immunoblot (RIBA), and immunofluorescence. Unfortunately, the
sensitivity and specificity of these assays is poorly defined. At present, there is
no gold standard. None of these supplemental assays has been licensed by

FDA. In addition, because of miniscule demand, it is unlikely that they will ever
be licensed.
The Western blot is the most used supplemental assay for HTLV-I/II.
A sample is positive when it reacts with at least two gene products (gag, p24, and
env, p21 or gp46). In the most commonly available Western blot (Cambridge
Biotech), gag encoded proteins (p19 and p24) are well represented, while the
envelope protein gp46 is not. Therefore other supplemental assays such as RIPA
are used to increase sensitivity for env products. RIPA is performed with a lysate
of MT-2 cells and is positive when antibodies to gp61/68 are detected.
Supplemental assays generate a large number of indeterminate test results
and provide poor type discrimination between HTLV-I and HTLV-II due to high
homology between the two viral types. (See Fig. 2.) FDA approved in June 1999
the use of a second licensed test for antibodies to HTLV-I/II (manufactured by a
different supplier) on samples that test repeatedly reactive on a screening test for
Provirus 65% (nt)
LTR 31% (nt)
gag, Matrix, p19 56% (aa)
gag, capsid, p24 83% (aa)
gag, nuclocapsid, p15 67% (aa)
Protease, p14 24% (aa)
Polymerase 85 kD 56% (aa)
env Surface, gp 46 61% (aa)
env, transmembrane, p21 84% (aa)
Non-translated 28% (nt)
tax, p40/p37 78% (aa)
rex, p27/p26 61% (aa)
0 10 20 30 40 50 60 70 80 90
% homology
Figure 2 Degree of homology between HTLV-I and HTLV-II at the nucleotide (nt) and
amino acid levels.
304 Rios and Bianco
antibodies to HTLV-I/II. It appears that more than 60% of these samples test
nonreactive in the second assay. Donors with discordant EIA reactivity remain
eligible to donate as long as the screening assay for the next donation tests
nonreactive. The dual EIA strategy facilitates donor counseling by removing
false-positive test results and avoiding indeterminate Western blot test results on
a substantial number of donors.
Amplification of specific viral sequences through the polymerase chain
reaction has been the most productive approach for confirmation of EIA results
and for the discrimination HTLV type (4347, 128,129). Unfortunately, nucleic
acid amplification technologies for HTLV-I/II are not readily available to most
blood-collecting facilities and clinical laboratories.
VII. DONOR NOTIFICATION AND COUNSELING

The lack of accuracy of confirmatory results, together with the unclear con-
sequences of viral infection, have made donor notification, counseling, and
follow-up of donors with repeatedly reactive results of EIA for antibodies to
HTLV-I and HTLV-II a difficult task. Counseling of donors for HTLV-I/II
has been the subject of an extensive review by the public health authorities and
the blood banking community. The outcome was published in 1993 (139).
Counseling should stress the difference between HTLV-I/II and HIV. Counseling
should also be tailored to viral type because of differences in disease association
(or lack thereof) between the two viral types. In addition, the low frequency of
disease development over the lifetime of infected individuals (25%) should
be emphasized.
Donors with positive test results are told not to donate blood or tissues in
the future and to use precautions for the prevention of transmission of the virus
by blood or by sexual contact. HTLV-I/IIpositive mothers are discouraged
from breast-feeding when practicable to prevent mother-to-child transmission.
However, this approach should be avoided in countries where breast-feeding is
critical to child survival, e.g., in areas where diarrheal disease leads to high
mortality. Use of condoms is recommended for couples in which one of the
partners is HTLV-I/II positive. In case these couples desire to have babies,
unprotected sex should be limited to the period of fertility. Such decisions require
careful discussions between physician and patient, and no clear guidelines have
been defined.
Counseling of donors with indeterminate test results is always difficult
because of the undefined messages that must be conveyed. These donors are told
that they are probably not infected. However, they are strictly barred from
donating blood.
REFERENCES
1. Poiesz BJ, Sherman M, Saksena N, et al. The biology and epidemiology of the
human T-cell lymphoma/leukemia viruses. In: Neu HC, Levy JA, Weiss RA, eds.
Frontiers of infectious diseases: Focus on HIV London: Churchill Livingston, 1993:
189205.
2. Poiesz BJ, Ruscetti FW, Gazdar AF, Bunn PA, Gallo RC. Detection and isolation of
type C retrovirus particles from fresh and cultured lymphocytes of a patient with
cutaneous T-cell lymphoma. Proc Natl Acad Sci USA 1980; 77:74157419.
3. Yoshida M, Miyoshi I, Hinuma Y. Isolation and characterization of retrovirus from
cell lines of human adult T-cell leukemia and its implication in the disease. Proc
Natl Acad Sci 1984; 79:20312035.
4. Popovic M, Reitz MS, Jr, Sarngadharan MG. The virus of Japanese adult T-cell
leukemia is a member of the human T-cell leukemia/lymphoma virus Group. Nature
1982; 300:6366.
5. Watanabe T, Seiki M, Yoshida M. ATLV (Japanese isolate) and HTLV (US isolate)
are the same strain of retroviruses. Virology 1984; 133:238241.
6. Blattner WA, Kalyanaram VS, Robert-Guroff M, Lister A, Galton DAG, Sarin PS,
Crawford MH, Catovsky D, Greaves M, Gallo R. The human type-C retrovirus,
HTLV, in blacks from the Caribbean region, and relationship to adult T-cell
leukemia/lymphoma. Int J Cancer 1982; 30:257264.
7. Doski H, Denic S, Patel N, Krishnamurthy M, Levine P, Clark JW. Adult T-cell
Leukemia/Lymphoma in Brooklyn. JAMA 1988; 259:22552257.
8. Kalyanaraman VS, Sarngadharan MG, Robert-Guroff M. A new subtype of human
T-cell leukemia virus (HTLV-II) associated with a T-cell variant of hairy cell
leukemia. Science 1982; 218:571573.
9. Weiss R, Teich N, Varmus H, Coffin J, eds. RNA Tumor Viruses. Cold Spring
Harbor, NY: Cold Spring Harbor Laboratory, 1985.
10. Inoue J, Seiki M, Yoshida M. The second pX product p27xIII of HTLV-I is required
for gag gene expression. FEBS 1986; 209:187190.
11. Inoue J, Yoshida M, Seiki M. Transcriptional (p40x) and post-transcriptional
(p27xIII) regulators are required for the expression and replication of human T-cell
leukemia virus type I genes. Proc Natl Sci USA 1987; 84:36533657.
12. Holland J J, ed. Genetic Diversity of RNA Viruses. Berlin: Spring-Verlag, 1992.
13. Sherman MP, Saksena NK, Dube DK, Yanangihara R, Poiesz BJ. Evolutionary
insights in the origin of human T-cell lymphoma/leukemia virus type I (HTLV-I)
derived from sequence analyses of a new HTLV-I variant from Papua New Guinea.
J Virol 1992; 66:25562563.
14. Gessain A, Gallo RC, Franchini G. Low degree of human T-cell leukemia virus type
I genetic drift in vivo as means of monitoring viral transmission and movement of
ancient human populations. J Virol 1992; 66:22882295.
15. Sherman MP, Saksena NK, Dube DK, Yanangihara R, Poiesz BJ. Evolutionary
insights in the origin of human T-cell lymphoma/leukemia virus type I (HTLV-I)
derived from sequence analyses of a new HTLV-I variant from Papua New Guinea.
J Virol 1992; 66:25562563.
306 Rios and Bianco
16. Myers G, Korber B, Wain-Hobson S, Smith RF, Pavlakis GN. Human retroviruses
and AIDS. 1993. Los Alamos, NM:Los Alamos National Laboratory, 1993.
17. Pedrosa Martins L, Chenciner N, Wain-Hobson S. 1992. Complex intra-patient
sequence variation in the V1 and V2 hypervariable regions of the HIV-1 gp120
envelope sequences. Virology 64:837845.
18. Verdier M, Denis F, Sangare A, et al. Prevalence of antibody to human T cell
leukemia virus type I (HTLV-I) in populations of Ivory Coast, West Africa. J Infect
Dis 1989; 160:363370.
19. Touz E, Gessain A, Lyon-Caen O, Gout O. Tropical spastic paraparesis/HTLV-I-
associated myelopathy in Europe and Africa: clinical and epidemiological aspects.
JAIDS Hum Retrov 1996; 13(S1):S3845.
20. Wiktor SZ, Piot P, Mann JM, et al. Human T-cell lymphotropic virus type I (HTLV-I)
among female prostitutes in Kinshasa, Zaire. J Infect Dis 1990; 161:10731077.
21. Blattner WA, Saxinger C, Riedel D. A study of HTLV-I and its associated risk factors
in Trinidad and Tobago. J AIDS 1990; 90:11021108.
22. Blattner WA, Blaney DW, Robert-Guroff M, Sarngadharan MG, Kalyanaraman VS,
Sarin PS, Jaffe ES, Gallo, RC. Epidemiology of human T-cell leukemia/lymphoma
virus. J Infect Dis 1983 147:406416.
23. Riedel D, Evans AS, Saxinger C, Blattner WA. A historical study of human
T-lymphotropic virus type I transmission in Barbados. J Infect Dis 1989; 159:603
609.
24. Murphy EL, Figueroa JP, Gibbs WN, et al. Human T-lymphotropic virus type I
(HTLV-I) seroprevalence in Jamaica: I. Demographic determinants. Am J Epidemiol
1991; 133(11):11141124.
25. Maloney EM, Murphy EL, Figueroa JP, Gibbs WN, Cranston B, Hanchard B,
Holding-Cobham M, Malley K, Blattner WA. Human T-lymphotropic virus type-I
(HTLV-I) seroprevalence in Jamaica II. Geographic and ecological determinants.
Am J Epidemiol 1991; 133:11251134.
26. Taylor GP. The epidemiology of HTLV-I in Europe. JAIDS Hum Retrov 1996;
13(S1):S814.
27. Hale A, Leung T, Sivasubramaniam S, Kenny J, Sutherland S. Prevalence of
antibodies to HTLV in antenatal clinic attendees in south east London. J Med Virol
1997; 52:326329.
28. Voevodin A, Gessain A. Common origin of human T-lymphotropic virus type-I from
Iran, Kuwait, Israel, and La Reunion Island. J Med Virol 1997; 52:7782.
29. Yamaguchi K. Human T-lymphotropic virus type I in Japan. Lancet 1994; 343:213
216.
30. Gabbai AA, Bordin JO, Vieira-Filho JBP, Kuroda A, Oliveira ASB, Cruz MV,
Ribeiro AAF, Delaney SR, Henrard DR, Rosario J, Roman GC. Selectivity of human
T-lymphotropic virus type-I (HTLV-1) and HTLV-2 infection among different
populations in Brasil. Am J Trop Med Hyg 1993; 49:664671.
31. de Rivera IL, Amador L, Mourra S, Li S, Rasheed S. Geographical clustering of
human T-cell lymphotropic virus type 1 infection in Honduras. J Clin Microbiol
1995; 33:29993003.
32. Maloney EM, Ramirez H, Levine A, Blattner WA. A survey of the human T-cell
lymphotropic virus type I (HTLV-I) in south-western Colombia. Int J Cancer 1989;

44:419423.
33. Gotuzzo E, Sanchez J, Escamilla J, Carrillo C, Phillips IA, Moreyra L, Stamm W,
Ashley R, Roggen EL, Kreiss J, Piot P, Holmes KK. Human T cell lymphotropic
virus type I infection among female sex workers in Peru. J Infect Dis 1994;
169:5459.
34. Pino N, Peralta LM, Pampuro S, Pimental E, Libonatti O. HTLV-I/II seroprevalence
and co-infection with other pathogens in blood donors in Buenos Aires. J AIDS
1994; 7:206207.
35. Vasquez P, Sanchez G, Volante C, Ramirez E, Soto G, Lee H. Human T-lympho-
tropic virus type I: new risk for Chilean population. Blood 1991; 78:850.
36. Merino F, Robert-Guroff M, Clark J, Bundo-Barcho M, Blattner WA, Gallo RC.
Natural antibodies to human T-cell leukemia/lymphoma virus in healthy Venezuelan
population. Int J Cancer 1984; 34:501506.
37. Khabbaz RF, Heneine W, Grindon A, Hartley TM, Shulman G, Kaplan J. Indeter-
minate HTLV serologic results in U.S. blood donors: are they due to HTLV-I or
HTLV-II? J AIDS 1992; 5:400-404.
38. Bastian I, Gardner J, Webb D, Gardner I. Isolation of a human T-lymphotropic virus
type-I strain from Australian aboriginals. J Virol 1993; 67:843851.
39. Doherty RR. HTLV-I in Australia and Oceania: long-term resident or recent immigr-
ant? Med J Aust 1996: 164:8486.
40. Yang Y-C, Hsu T-Y, Liiu M-Y et al. Molecular subtyping of T-lymphotropic virus
type I (HTLV-I) by a nested polymerase chain reaction-restriction fragment poly-
morphism analysis of the envelope gene: two distinct lineages of HTLV-I in Taiwan.
J Med Virol 1997; 51:2531.
41. Yanagihara R, Jenkins CL, Alexander SS, et al. Human T lymphotropic virus type
I infection in Papua New Guinea: high prevalence among the Hagahai confirmed
by Western blot analysis. J Infect Dis 1990; 162:649654.
42. Yamashita M, Achiron A, Miura T et al. HTLV-I from Iranian Mashhadi Jews in
Israel is phylogenetically related to that of Japan, India, and South America rather
than that of Africa and Melanesia. Virus Genes 1995; 10:8590.
43. Khabbaz RF, Hartel D, Lairmore M, Horsburgh CR, Schoenbaum EE, Roberts B,
Hartley TM, Frieldland G. Human T lymphotropic virus type II (HTLV-II) infection
in a cohort of New York intravenous drug users: an old infection? J Infect Dis 1990;
163:252.
44. Khabbaz RF, Onorato IM, Cannon RO, et al. 1992 Seroprevalence of HTLV-I and
HTLV-II among intravenous drug users and persons in clinics for sexually transmit-
ted diseases. N Engl J Med 1992; 326:375380.
45. The HTLV European Research Network. Seroepidemiology of the human T-cell
leukaemia/lymphoma viruses in Europe. J Acquir Immune Defic Syndr Hum
Retrovirol 1996; 13(1):6877.
46. Lee H, Sawanson P, Shorty VS, Zack JA, Rosenbalt JD, Chen IS-Y. High rate of
HTLV-II infection in seropositive drug abusers in New Orleans. Science 1989;
244:471.
47. Hjelle B, Wilson C, Cyrus S, Bradshaw P, Schammel C, Wiltbank T, Alexander S.
308 Rios and Bianco
Human T-cell leukemia virus type II infection frequently goes undetected in

contemporary US blood donors. Blood 1993; 81:16411644.
48. Laimore MD, Jacobson S, Garcia F, et al. Isolation of human T-cell lymphotropic
virus type 2 from Guaymi Indians in Panama. Proc Natl Acad Sci USA 1990;
87:88408844,
49. Lal RB, Povoa M, Lal AA. Seroprevalence of HTLV-II in Paragaminos, state of
Para, Brazil. Lett J AIDS 1992; 5:634636.
50. Fujiyama C, Fujiyoshi T, Mivra T, Yashiki S, Matsumoto D, Zaninovic V, Blanco
O, Harrington Jr W, Byrnes JJ, Hayami M, Sonoda S. A new endemic focus of T-cell
lymphotropic virus type among Ornico natives in Colombia. J Infect Dis 1992;
168:10751077.
51. Hjelle B, Cyrus S, Swenson SG. Evidence for sexual transmission of human T
lymphotropic virus type II. Ann Intern Med 1992; 116:9091.
52. Biglone M, Gessain A, Quiruelas S, Fay O, Taborda MA, Fernandez E, Lupo S,
Panzita A, de Th G. Endemic HTLV-II infection among Tobas and Matacos
Amerindians from North of Argentina. J AIDS 1993; 6:631633.
53. Lavine PH, Jacobson S, Elliot R, Cavallero A, Colclough G, Stephenson C, Knigge
RM, Drumond J, lishimura M, Taylor ME, Wiktor S, Shaw GM. HTLV-II in Florida
Indians. AIDS Res Hum Retrov 1993; 9:123128.
54. Maloney EM, Biggar RJ, Neel JV, et al. Endemic human T-cell lymphotropic virus
type II infection among isolated Brazilian Amerindians. J Infect Dis 1992; 166:100
107.
55. Lairmore MD, Jacobson S, Gracia F, et al. Isolation of human T-cell lymphotropic
virus type 2 from Guaymi Indians in Panama. Proc Natl Acad Sci USA 1990;
87:88408844.
56. Ijichi S, Zaninovic V, Leon FE et al. Identification of human T-cell leukemia virus
type IIb infection in the Wayu, an aboriginal population of Colombia. Jpn J Cancer
Res 1993; 84:12151218.
57. Ishak R, Harrington WJ, Azevedo VN, et al. Identification of human T cell
lymphotropic virus type IIa infection in the Kayapo, an indigenous population of
Brazil. AIDS Res Hum Retrovir 1995; 11:813821.
58. Kajiyama W, Kashiwagi S, Ikematsu H, Hayashi J, Nomura H, Okochi K. Intra-
familial transmission of adult T cell leukemia virus. J Infect Dis 1986; 154:851857.
59. Ichimaru M, Ikeda S, Kinoshita K, Hino S, Tsuji Y. Mother-to-child transmission of
HTLV-1. Cancer Detect Prev 1991; 15:177181.
60. Fujino T, Fujiyoshi T, Yashiki S, Sonoda S, Otsuka H, Nagata Y. HTLV-I transmis-
sion from mother to fetus via placenta (letter). Lancet 1992; 340:1157.
61. Tajima K, Cartier L. Epidemiological features of HTLV-I and adult T-cell leukemia.
Intervirology 1995; 38:238246.
62. Bartolomeu C, Saxinger WC, Clark JW, et al. Transmission of HTLV-I and HIV
among homosexual men in Trinidad. JAMA 1987; 257:26042608.
63. Figureoa JP, Ward E, Morris J, et al. Incidence of HIV and HTLV-I infection among
sexually transmitted disease clinic attendees in Jamaica. J AIDS Hum Retrovir 1997;
15:232237.
64. Bulterys M, Landesman S, Burns DN, Rubinstein A, Goedert JJ. Sexual behavior
and injection drug use during pregnancy and vertical transmission of HIV-1. J AIDS
Hum Retrovir 1997; 15:7682.
65. Kaplan JK, Khabbaz RF, Murphy EL, Hermansen S, Roberts C, Lal R, Heneine V,
Wright D, Matijas L, Thomson R, Rudolph D, Switzer WM, Kleinman S, Busch M,
Schreiber GB, and the Retrovirus Epidemiology Donor Study Group. Male-to-
female transmission of human T-cell lymphotropic virus type I and II: association
with viral load. J AIDS 1996; 12:193201.
66. Iwahara Y, Takehara N, Kataoka R, Sawada T, Ohtsuki Y, Nakachi H, Maehama T,
Okayama T, Miyoshi I. Transmission of HTLV-I to rabbits via semen and breast
milk from serositive healthy persons. Int J Cancer 1990; 45:980983.
67. Lal RB, Gongora-Biachi RA, Pardi D, Switzer WM, Goldman I, Lal AA. Evidence
for mother-to-child transmission of human T lymphotropic virus type II. J Infect Dis
1993; 586591.
68. Heneine V, Woods T, Green D, Fukuda K, Giusti R, Castillo L, Armien B, Garcia
F, and Kaplan J E. Detection of HTLV-II in breast milk of HTLV-II infected mothers
(letter). Lancet 1992; 340:11571158.
69. Gessain A, Vernant JC, Maurs L. Antibodies to human T-lymphotropic virus type I
in patients with tropical spastic paraparesis. Lancet 1985; ii:407409.
70. Gessain A, Francis H, Sonan T. HTLV-I and tropical spastic paraparesis in Africa.
Lancet 1986; ii:698.
71. La Grenade L. HTLV-I-associated infective dermatitis: past, present, and future.
J AIDS Hum Retrov 1996; 13(S1):S4649.
72. Mochizuki M, Ono A, Ikeda E, Hikita, Watanabe T, Yamaguchi K, Sagawa K, Ito
K. HTLV-I uveitis. J AIDS Hum Retrov 1996; 13(S1):S5056.
73. Osato M, Yamaguchi K, Yamasaki H, Suzushima H, Asou N, Sakata K, Kawakita
M, Takatsuki K. A human T-cell lymphotropic virus type-I carrier with chronic renal
failure, aplastic anaemia, myelopathy, uveitis, Sjogren syndrome and panniculitis.
Int Med 1996; 35(9):742745.
74. Yamaguchi K, Takatsuki K. Adult T cell leukemia-lymphoma. Baillieres Clin
Hematol 1993; 6:899915.
75. Franchini, G. Molecular mechanisms of human T-cell leukemia/lymphotropic virus
type I infection. Blood 1995; 86:36193639.
76. Blattner WA, Pombo-de-Oliveira, MS. HTLV-I and HTLV-II. In: Merigan C, Bartlett
JG, Bolognesi D, eds. Textbook of AIDS Medicine, 1998: 10031029.
77. Manns A, Hisada M, La Grenade L Human T-lymphotropic virus type I infection.
Lancet 1999; 353:19511958.
78. Yodoi J, Takatsuki K, Masuda T. N Engl J Med 1974; 290:572574.
79. Uchiyama T, Yodoi J, Sagawa, K., Takatsuki K, Uchino, H. Adult T-cell leukemia:
clinical and hematological features of 16 cases. Blood 1977; 50:481492.
80. Hinuma Y, Nagata K, Hanaoka M, et al. Adult T-cell leukemia: antigen in adult
T-cell leukemia cell line and detection of antibodies to the antigen in human sera.
Proc Natl Acad Sci USA 1981; 78:64766480.
81. Yodoi J, Uchiyama T. Diseases associated with HTLV-I: virus, IL-2 receptor
deregulation and redox regulation. Immunol Today 1992; 13:405410.
82. Shimoyama M, members of Lymphoma Study Group. The fourth nationwide study
310 Rios and Bianco
of T-cell leukemia/lymphoma (ATL) in Japan: estimates of risk of ATL and its

geographical and clinical features. Br J Hematol 1991; 79:428439.
83. Yoshida M, Seiki M, Yamaguchi K, Takatsuki K. Monoclonal integration of human
T-cell leukemia provirus in all primary tumors of adult T-cell leukemia suggests
causative role of human T-cell leukemia virus in the disease. Proc Natl Acad Sci
USA 1984; 81:25342537.
84. Cleghorn FR, Manns A, Falk R, et al. Effect of human T-cell leukemia virus
type I infection on non-Hodgkins lymphoma incidence. J Natl Cancer Inst 1995;
87:10091014.
85. Pombo de Oliveira MS, Matutes E, Schulz T, et al. T-cell malignancies in Brazil.
Clinico-pathological and molecular studies of HTLV-I-positive and -negative cases.
Int J Cancer 1995; 60:823827.
86. Tajima K, Kuroishi T. Estimation of rate of incidence of ATL among ATLV
(HTLV-I) carriers in Kyushu, Japan. Jpn J Clin Oncol 1985; 15:423430.
87. Takatsuki K, Matsuoka M, Yamaguchi K. Adult T-cell leukemia/lymphoma in Japan.
J AIDS Hum Retrovir 1996; 13 (suppl 1):s1519.
88. Gill PS, Harrington W, Kaplan M, et al. Treatment of T-cell leukemia-lymphoma
with a combination of interferon alpha and zidovudine. N Engl J Med 1995;
332:17441748.
89. Hermine O, Bouscary D, Gessain A, et al. Brief report: treatment for adult T-cell
leukemia-lymphoma with zidovudine and interferon alpha. N Engl J Med 1995;
332:17491751.
90. Osame M, Izumo S, Igata A. Blood transfusion with HTLV-I associated myelopathy.
Lancet 1986; ii:104105.
91. Maloney EM, Cleghorn FR, Morgan OS, et al. Incidence of HTLV-I associated
myelopathy/Tropical spastic paraparesis (HAM/TSP) in Jamaica and Trinidad.
J AIDS Human Retrovir 1998; 17:167170.
92. Link H, Cruz M, Gessain A, Gout O, de The G, Kam-Hansen S. Chronic progressive
myelopathy associated with HTLV-I; oligoclonal lgG and anti-HTLV-I IgG antibod-
ies in the cerebrospinal fluid and serum. Neurology 1989; 39:15661572.
93. Kira J, Koyanagi Y, Yamada T, et al. Increased HTLV-I proviral DNA in HTLV-I-
associated myelopathy: a quantitative polymerase chain reaction study. Ann Neurol
1991; 29:194201.
94. Tsukasaki K, Yamada Y, Ikeda S, Tomonaga M. Infective dermatitis among patients
with ATL in Japan. Int J Cancer 1994; 57:293.
95. Hanchard B, LaGrenade L, Canberry C, et al. Childhood infective dermatitis
evolving into adult T-cell leukaemia after 17 years [letter]. Lancet 1991; 338:1593
1594.
96. Ohba N, Matsumoto M, Sameshima Y, et al. Ocular manifestations in patients with
human T-lymphotropic virus type I. Jpn J Ophthalmol 1989; 33:112.
97. Mochizuki M, Tajima K, Watanabe T, Yamaguchi K. Human T lymphotropic virus
type-I uveitis. Br J Ophthalmol 1994; 78:149154.
98. Sagawa K, Mochizuki M, Masuoka K, et al. Immunopathological mechanisms of
human T lymphotropic virus type-I (HTLV-I) uveitis. Detection of HTLV-I infected
T cells in the eye and their constitutive cytokine production. J Clin Invest 1995;
95:852858.
99. Ohshima K, Kikuchi M, Masuda Y, et al. HTLV-I associated lymphoadenopathy.

Cancer 1992; 69:239248.
100. Nishioka K, Maruyama I, Sato K, Kitajima I, Nakajima Y, Osame M. Chronic
inflamatory arthropathy associated with HTLV-I. Lancet 1989; 1:441.
101. Nakada K, Yamaguchi K, Furugen S, et al. Monoclonal integration of HTLV-I
proviral DNA in patients with strongyloidiasis. Int J Cancer 1987; 40:145148.
102. Mita S, Sugimoto M, Nakamura M, et al. Increased human T lymphotropic virus
type-I (HTLV-I) proviral DNA in peripheral blood mononuclear cells and broncho-
alveolar lavage cells from Japanese patients with HTLV-I-associated myelopathy.
Am J Trop Med Hyg 1993; 48:170177.
103. Rosenblatt JD, Gasson JC, Glaspy J, et al. Relationship between human T-cell
leukemia virus-II and atypical hairy cell leukemia: serologic study of hairy cell
leukemia patients. Leukemia 1987; 1:397401.
104. Hjelle B, Mills R, Swenson S, Mertz G, Key C, Allen S. Incidence of hairy cell
leukemia, mycosis fungoides, and chronic lymphocytic leukemia in a first known
HTLV-II-endemic population. J Infect Dis 1991; 163:435440.
105. Kalyanaraman VS, Sarngadharan MG, Robert-Guroff M, et al. A new subtype of
human T-cell leukemia virus (HTLV-II) associated with a T-cell variant of hairy cell
leukemia. Science 1982; 218:571573.
106. Hjelle B, Appenzeller O, Mills R, et al. Chronic neurodegenerative disease associ-
ated with HTLV-II infection. Lancet 1992; 339:645646.
107. Pombo de Oliveira MS, Matutes E, Schulz T, et al. T-cell malignancies in Brazil.
clinico-pathological and molecular studies of HTLV-I positive and negative cases.
Int J Cancer 1995; 60:823827.
108. Pawson R, Schulz TF, Matutes E, Catovsky D. The human T-cell lymphotropic
viruses types I/II are not involved in T-prolymphocytic leukemia and large granular
lymphocytic leukemia. Leukemia 1997; 11:13051311.
109. Jacobson S, Lehky T, Nishimura M, Robinson S, McFarlin D, Dhib-Jalbut S.
Isolation of HTLV-II from a patient with chronic, progressive neurological disease
clinically indistinguishable from HTLV-I-associated myelopathy/tropical spastic
paraparesis. Am Neur Assoc 1993; 14:392396.
110. Loughran TP, Sherman MP, Ruscetti FW, et al. Prototypical HTLV-I/II infection is
rare in patients with large granular lymphocyte leukemia. Leukemia Res 1994;
18:423429.
111. Loughran TP, Coyle T, Sherman MP, et al. Detection of human T-cell leukemia/
lymphoma virus, type II, in patient with large granular lymphocyte leukemia. Blood
1992; 80:11161119.
112. Hjelle B, Mills R, Swenson S, Mertz G, Key C, Allen S. Incidence of hairy cell
leukemia, mycosis fungoides, and chronic lymphocytic leukemia in a first known
HTLV-II-endemic population. J Infect Dis 1991; 163:435440.
113. Zucker-Franklin D, Hooper WC, Evatt BL. Human lymphotropic retroviruses
associated with mycosis fungoides: evidence that human T-cell lymphotropic virus
type II (HTLV-II) as well as HTLV-I may play a role in the disease. Blood 1992;
80:15371545.
114. Kaplan MH, Hall WW, Susin M, et al. Syndrome of severe skin disease, eosinophilia
312 Rios and Bianco
and dermatopathic lymphadenopathy in patients with HTLV-II complicating human

immunodeficiency virus infection. Am J Med 1991; 91:300309.
115. Martin MP, Biggar RJ, Hamlin-Green G, Staal S, Mann D. AIDS Res Hum Retrovir
1993; 9:715719.
116. Semmes OJ, Majone F, Cantemir C, Turchetto L, Hjelle B, Jeang K-T. HTLV-I and
HTLV-II tax: differences in induction of micronuclei in cells and transcriptional
activation of viral LTRs. Virology 1996; 217:313379.
117. Fouchard N, Flaguel B, Bagot M, Avril MF, Hermine O, Sigaux F, Merle-Beral H,
Troussard X, Delfraissy JF, de The G. Lack of evidence of HTLV-I/II infection in T
CD8 malignant or reactive lymphoproliferative disorders in France: a serological
and/or molecular study of 169 cases. Leukemia 1995; 9:20872092.
118. Woods TC, Graber JM, Hershow RC, Khabbaz RF, Kaplan JE, Heineine W.
Investigation of proviral load in individuals infected with human T-lymphotropic
virus type II. AIDS Res Hum Retrovir 1995; 11:12351239.
119. Lehky TJ, Flerlage N, Katz D, Houff S, Hall WH, Ishii K, Monken C, Dhib-Jalbut
S, McFarland HF, Jacobson S. Human T-cell lymphotropic virus type II-associated
myelopathy: clinical and immunologic profiles. Ann Neurol 1996; 40:714723.
120. Doeneief G, Marlink R, Bell K, Marder K, Renjifo B, Sern Y, Mayeux R.
Neurological consequences of HTLV-II infection in injection-drug users. Neurology
1996; 46: 15561560.
121. Yokoi K, Kawai H, Akaike M, Mine H, Saito S. Presence of human T-lymphotropic
virus type II-related genes in DNA of peripheral leukocytes from patients with
autoimmune thyroid diseases. J Med Virol 1995; 45:392398.
122. Murphy EL, Glynn SA, Fridey J, Smith JW, Sacher RA, Nass CC, Ownby HE,
Wright DJ, Nemo GJ. Increased incidence of infectious diseases during prospective
follow-up of human T-lymphotropic virus type II- and I-infected blood donors.
Retrovirus Epidemiology Donor Study. Arch Intern Med 1999; 159(13):14851491.
123. Okochi K, Sato H, Hinuma Y. A retrospective study on transmission of adult T-cell
leukemia virus by blood transfusion: seroconversion in recipients. Vox Sang 1984;
46:245253.
124. Public Health Service Working Group. Licensure of screening tests for antibodies
to human T-lymphotropic virus type I. MMWR 1988; 37:736747.
125. Manns A, Wilks RJ, Murphy EL, et al. A prospective study of transmission by
transfusion of HTLV-I and risk factors associated with seroconversion. Int J Cancer
1992; 51:886891.
126. Schreiber GB, Busch MP, Kleinman SH, Korelitz JJ, the risk of transfusion-
127. Sullivan MT, Williams AE, Fang CT, Grandinett T, Poiesz BJ, Ehrlich GD.
Transmission of human T-lymphotropic virus types I and II by blood transfusion.
Arch Intern Med 1991; 151:20432048.
128. Rios M, Khabbaz RF, Kaplan JE, Hall WW, Kessler D, Bianco C. Transmission of
human T-cell leukemia virus type II by transfusion of human T-cell lymphotropic
virus type I screened blood products. J Infect Dis 1994; 170:206210.
129. Nelson KE, Donahue JG, Munoz A, et al. Transmission of retroviruses from
seronegative donors by transfusion during cardiac surgery. A multicenter study of
HIV-1 and HTLV-I/II infections. Ann Intern Med 1992; 117:554559.
130. Food and Drug Administration (CBER). Guidance for Industry: Donor Screening
for Antibodies to HTLV-II. August 1997
131. Chen YC, Wang CH, Su IJ, Hu CY, Chou MJ, Lee TH, Lin DT, Chung TY, Liu CH,
Yang CS. Infection of human T-cell leukemia virus type I and development of
human T-cell leukemia lymphoma in patients with hematologic neoplasms: a
possible linkage to blood transfusion. Blood 1989; 74(1):388394.
132. Bartholomew C, Jack N, Edwards J, Charles W, Corbin D, Cleghorn FR, Blattner
WA. HTLV-I serostatus of mothers of patients with adult T-cell leukemia and
HTLV-I-associated myelopathy/tropical spastic paraparesis. J Hum Virol 1998;
1(4):302305.
133. Domingues RB, Muniz MR, Jorge ML, Mayo MS, Saez-Alquezar A, Chamone DF,
Scaff M, Marchiori PE. Human T cell lymphotropic virus type-1-associated myelo-
pathy/tropical spastic paraparesis in Sao Paulo, Brazil: association with blood
transfusion. Am J Trop Med Hyg 1997; 57(1):5659.
134. Kuroda Y, Takashima H, Yukitake M, Sakemi T. Development of HTLV-I-associated
myelopathy after blood transfusion in a patient with aplastic anemia and a recipient
of a renal transplant. J Neurol Sci 1992; 109(2):196199.
135. Saxton EH, Lee H, Swanson P, Chen IS, Ruland C, Chin E, Aboulafia D, Delamarter
R, Rosenblatt JD. Detection of human T-cell leukemia/lymphoma virus type I in a
transfusion recipient with chronic myelopathy. Neurology 1989; 89(6):841844.
136. Kaplan JE, Litchfield B, Rouault C, Lairmore MD, Luo CC, Williams L, Brew BJ,
Price RW, Janssen R, Stoneburner R, et al. HTLV-I-associated myelopathy associ-
ated with blood transfusion in the United States: epidemiologic and molecular
evidence linking donor and recipient. Neurology 1991; 41(2 Pt 1):192197.
137. Kurosawa M, Machii T, Kitani T, Tokumine Y, Kawa K, Maekawa I, Kawamura T,
Miyake T, Kanda M. HTLV-I associated myelopathy (HAM) after blood transfusion
in a patient with CD2+ hairy cell leukemia. Am J Clin Pathol 1991; 95(1):7276.
138. Delamarter RB, Carr J, Saxton EH. HTLV-I viral-associated myelopathy after blood
transfusion in a multiple trauma patient. Clin Orthop 1990; (260):191194.
139. Khabbaz RF, Fukuda K, Kaplan JE, Bianco C, Blattner W, Busch M, Dodd R,
Epstein J, Gilcher R, Jackson C, Katz L, Kleinman S, Murphy EL, Nemo G, Poiesz
BJ, Rios M, Sloand E, Sullivan M, William AE. Guidelines for counseling persons
infected with human T-lymphotropic virus type I (HTLV-I) and type II (HTLV-II)
Ann Int Med 1993; 118:448454.
14
Transfusion-Acquired
Cytomegalovirus Infection
Approaching Resolution
Gary E. Tegtmeier
Community Blood Center of Greater Kansas City, Kansas City, Missouri
I. INTRODUCTION
Cytomegalovirus (CMV) is one of the human herpesviruses; all are highly

successful intracellular parasites of humans. With the exception of varicella virus,
they cause asymptomatic or mild primary infections, establish latent infections in
their hosts, can periodically reactivate and infect other susceptible hosts, and
persist lifelong. As long as the hosts immune system is uncompromised, the
balance between virus and host is maintained. However, immunosuppression
induced by disease or therapy can alter the balance between virus and host,
resulting in systemic virus infection and concomitant disease.
CMVs disease-producing potential was first recognized in infants who
were infected in utero (1). Although posttransfusion CMV infections were first
described in 1966 (2), it was not until the 1980s that severe CMV disease was
unequivocally linked to the transfusion of CMV-seropositive blood in premature
infants (3).
The high frequency and serious consequences of CMV infection in im-
munocompromised patients was documented in solid organ and bone marrow
transplant (BMT) patients in the mid- to late 1970s (1). The dominant role of the
donor organ versus blood product in CMV transmission was quickly established
in solid organ transplantation (4). It was more gradually recognized that trans-
fused blood might be the source of CMV in BMT patients (5), particularly
seronegative patients receiving marrow from CMV-negative donors (6). The ef-
ficacy of CMV-seronegative blood products in reducing the risk of transfusion-
acquired (TA) CMV infection in such patients was soon established (68).
315
316 Tegtmeier
The use of leukoreduced blood products to lower the risk of TA CMV infection
was established subsequently, first in neonates (9), then in BMT patients (10).
Progress in controlling TA CMV infections has been achieved, but an
intractable, low level of risk persists in BMT patients (11). Effective surveillance
for early CMV infections using sensitive laboratory tests such as the pp65
antigenemia test and the availability of effective antiviral agents have reduced
morbidity and mortality in this patient population. Nevertheless, prevention
remains the ultimate goal, one that will certainly be realized in the context of
ongoing research on CMV latency, molecular methods for CMV detection, and
progress in viral inactivation technology.
The aim of this chapter is to review the following areas: (a) recent progress
in understanding the sites of CMV latency and the natural history of CMV
infection as they relate to TA CMV infections; (b) the epidemiology of TA CMV
in recipients; (c) the epidemiology of CMV in donors; and (d) the interventions
currently available to prevent TA CMV infections. The focus will be on the
transmission of CMV by blood components to the seronegative recipient.
II. CMV LATENCY AND THE NATURAL HISTORY OF

CMV INFECTION: RECENT FINDINGS
CMV is a highly evolved infectious agent. By virtue of persistent shedding after
primary infection and a capacity to induce latent infection with periodic reactiva-
tion, CMV maximizes its potential to spread. CMV is transmitted both vertically
from mother to her unborn or newborn child and horizontally by exposure to body
fluids, sexual contact, transplantation, and blood transfusion.
CMV can cause disease in immunocompromised patients, an outcome that
varies in severity with the degree of the patients immunosuppression, whether
disease-induced, iatrogenic, or due to prematurity. Disease manifestations in
immunocompromised patients include pneumonitis, retinitis, and gastritis.
CMV has the largest genome of any known animal virus. It codes for
220230 open reading frames, 62 of which have been shown to be unnecessary
for virus growth in cell culture. It is, therefore, likely that the proteins encoded
by the latter are involved in modulating CMV infection in vivo, including the
establishment of latent infection (12).
Our understanding of the site(s) and mechanism(s) of CMV latency and
how the virus is reactivated is incomplete. However, recent experimental data
have established the monocytemacrophage lineage as the likely site of CMV
latency in the bone marrow and peripheral blood. CMV DNA was found in five
of six seropositive subjects, predominantly in the nonT-cell population, and
specifically in adherent cells and CD14+ cells. CMV DNA was also detected in
three of nine seronegative subjects (13).
Transfusion-Acquired Cytomegalovirus 317
CMV has been successfully reactivated in vitro by means of allogeneic

stimulation of progenitor or peripheral blood mononuclear cells (14,15). Virus
was recovered after long-term culture from macrophages expressing dendritic cell
markers (14).
In a recent study (16), CMV genomes were detected in 0.0040.01%
mononuclear cells from granulocyte colony-stimulating factormobilized pe-
ripheral blood from 7 of 10 seropositive donors at copy numbers of 213 genomes
per infected cell. Genome-positive cells were found in one of two seronegative
donors. The application of the in situ detection and quantitation methods used in
this study will permit a comprehensive analysis of the distribution of latent viral
genomes and transcripts in various cell populations. Such work will be crucial to
understanding the transmission and the pathogenesis of CMV infections in blood
and organ transplant recipients.
Evidence that actively infected donors might be the source of infectivity in
some blood products dates back to a report from Rumania describing the isolation
of CMV from buffy coats of 2 of 35 healthy blood donors (17). Numerous
studies followed that attempted to isolate CMV from blood of over 1400 donors
without success (1). Retrospective (18,19) and prospective (19) testing of blood
donors for IgM anti-CMV also support the notion that actively infected donors
may transmit CMV.
A study of recently infected pregnant women demonstrated white blood
cellassociated CMV DNA by PCR in 100% of samples during the first month
of infection, which fell to 90% during the second month and to 0% by 6 months
(20). Viral load fell even more rapidly: at month 1, 60% of positive samples at
viral loads of >10 genome equivalents (GE) per 105 white blood cells (WBCs)
compared to 3.3% of the positive samples at month 2.
Similar findings were reported from a study of seroconverting adolescents
where 7585% of WBC samples were positive by PCR within 16 weeks of
infection, declining to 025% 48 weeks after infection (21). CMV DNA in plasma
was also monitored and was detected at a lower frequency, 2540%, at 816
weeks. However, some WBC and plasma samples were positive as late as
60 weeks after infection.
Further evidence that actively or recently infected donors may be responsi-
ble for some TA CMV infections comes from a study of 168 seroconverting blood
donors, whose plasma samples were tested for CMV DNA by PCR both pre-
and postseroconversion (22). Three (1.6%) were found positive, one in a pre-
seroconversion sample and two in postseroconversion samples. Viral load ranged
from 400 to 1000 copies/mL. No attempt was made to demonstrate the possible
infectivity of the detected DNA in vitro.
Whether CMV DNA can be detected in remotely infected seropositive
donors or in seronegative donors is a subject of controversy. Publications have
appeared reporting the detection of CMV DNA at high frequency in monocytes
318 Tegtmeier
or WBC from both seropositive and seronegative donors (13,2325). By contrast,

other reports have either failed or only occasionally identified CMV DNA by PCR
from seropositive or seronegative donors (26,27). Preliminary results from a large
multilaboratory study (28) indicate that the disparate findings among different
labs appear not to be due to lack of assay sensitivity; rather, the reports of positive
results among seronegative donors are more likely due to amplicon contamination
or lack of assay specificity.
III. EPIDEMIOLOGY OF TA CMV

That CMV can be transmitted by leukocytes contaminating blood products has
been suspected for a long time (29,30). The association of transmission with
seropositive donors was firmly established in the 1980s (3). However, it is still
uncertain whether all seropositive donors can transmit infection, or whether
infectivity is restricted to a subset of donors who either have experienced a recent
primary infection or a reactivated infection. Circumstantial evidence favors the
former hypothesis in that high infection frequencies were seen in exchange-
transfused infants (31), recipients of fresh whole blood (1), and patients receiving
granulocyte transfusions (29,30).
Other donor-related variables in CMV transmission include the type of
blood product transfused and the age of the blood product. Whereas cellular-
containing blood products such as whole blood, red blood cells, platelets, and
granulocytes have been associated with CMV transmission, no evidence has been
published implicating fresh-frozen plasma in TA CMV infections (32). The very
high rates of TA CMV infection seen in the early prospective studies of cardiac
surgery patients also suggested that fresh blood may be more infectious than
stored blood, although this has never been proven in controlled studies.
Studies of immunocompetent, CMV-seronegative cardiac surgery patients
carried out in the 1960s and 1970s showed high rates of TA infection, ranging
from 8.3 to 66.7% (1). These studies were carried out primarily in patients who
had received large volumes of fresh heparinized whole blood. These early
findings contrast markedly with more recently published data in seronegative
immunocompetent recipients showing TA infection rates in the range of 0.91.2%
(1,33,34).
The transmission of CMV by transfused blood is influenced by a multiplic-
ity of factors. Clearly, infectious virus, free or cell associated, and/or latently
infected cells must be present in donor blood. In theory, one virion or latently
infected cell should suffice, but at present the infectious dose of cell-free virus or
latently infected cells is unknown. If latently infected cells are the vehicle for
transmission, they must survive long enough in the recipient for viral activation
and release of virions to occur.
If recently infected donors are the source of infection, they may have
low-level viremia and/or larger numbers of latently infected cells than seroposi-
tive donors who were infected at a more remote time. Because leukocytes are not
distributed evenly across different blood components, both the kind and number
of components a patient receives affects the level of risk. Recipient factors that
play a role in transmission are the degree of HLA matching between donor and
recipient, iatrogenic or disease-induced immunosuppression, and the cytokine
environment in the recipient.
Patient populations known to be at risk for severe TA CMV morbidity and
mortality include infants, especially premature infants, pregnant women, solid
organ transplant patients, BMT patients, HIV-infected patients, and patients
with malignancy.
Rates of infection in transfused CMV-seronegative premature infants were
shown to be quite high in early studies, ranging from 24.0 to 31.8% (3,35). Later
investigations failed to corroborate the high rates originally seen; infection rates
varied from 0 to 8.7% (19,3638). No clear explanations for these differences in
rates could be discerned; these studies were carried out in different geographic
regions. The observed differences in rates were not readily correlated with donor
anti-CMV prevalence, the amount or age of the blood transfused, or infant
birth weight.
Seronegative pregnant women undergoing primary CMV infection transmit
the infection with high efficiency to the fetus, often with devastating conse-
quences (39). Although the risk of TA CMV in this patient group has not been
rigorously assessed, a recent study failed to show CMV conversions in 162 CMV-
seronegative pregnant women, 8 of whom were transfused prior to delivery (38).
The risk of TA CMV infections in CMV-seronegative allogeneic BMT
patients receiving marrow from CMV seronegative donors and routine blood
product support has been assessed in three studies. The infection rates ranged
from 31.8 to 50.0% (6,8,10) not dissimilar to the 21.4% rate seen in CMV-
seronegative autologous BMT patients receiving routine blood products after
transplantation (10).
Highly variable levels of risk have been reported in CMV-seronegative
solid organ transplant recipients who received organs from CMV-seronegative
donors, as summarized in Table 1 (4056). An extraordinary number of variables
influencing the potential TA risk combine to account for the heterogeneous results
seen in this setting. The variables include the incidence of CMV infections in the
donor population, the CMV antibody test employed, the age of the transfused
blood, and the methods of blood component preparation.
At the recipient level, the potential of TA CMV infection and disease in
organ transplant patients is influenced by the following factors: the immuno-
suppressive regimen employed; the immunogenicity of the transplanted organs;
320 Tegtmeier
Table 1 Posttransfusion CMV Infections in

Seronegative Recipients Receiving Seronegative Organs
Range of PT CMV
Type of transplant infections (%) Ref.
Kidney 020 4044

Heart 026 45,46
Heart/Lung 4.533 4649
Liver 7.1100 5056
posttransfusion sepsis; and graft leukocytes that establish microchimeras and

thereby induce tolerance to transfused leukocytes.
TA CMV infections in seronegative oncology patients have been monitored
in several studies and range in incidence from 0 to 22% (5759). Symptomatic
CMV disease, however, is rare outside of the bone marrow transplant setting.
The American Association of Blood Banks (AABB) has recommended the use of
either CMV-seronegative or leukoreduced blood products for CMV-seronegative
patients who receive chemotherapy intended to produce severe neutropenia (60).
The rationale is to prevent potential candidates for BMT from acquiring an
infection that could result in serious morbidity after transplantation.
The risk of TA CMV infections in HIV-infected patients who are CMV
seronegative is not known. Given the devastating clinical consequences CMV can
effect in HIV-infected patients and the possibility that it may hasten the progres-
sion of AIDS, the use of CMV-seronegative or leukoreduced blood products for
CMV-seronegative patients infected with HIV has been recommended (60).
Although transmission of second strains to CMV seropositive, HIV-infected
patients is possible, provision of CMV-safe products has not been recom-
mended. With the recent public presentation of the data from the Viral Activation
Transfusion Study (VATS), which showed no benefit to CMV-infected patients
with HIV-1 infection from leukocyte reduction of red blood cells, this concern
may be allayed.
IV. PREVENTION OF TA CMV INFECTION BY

DONOR ANTIBODY SCREENING
A 1971 study of exchange-transfused infants in Germany was the first to suggest
that CMV-seropositive donors might transmit the infection and that seronegative
donors might be less likely to transmit (31). Eight of 15 seronegative infants
(53%) receiving CMV-seropositive donor blood became infected, while none of
20 seronegative infants receiving seronegative blood became infected. This was
the first published data suggesting that CMV-seronegative donor blood posed a
reduced risk of CMV transmission.
A large prospective study conducted in 1976 further suggested that CMV-
seronegative blood products posed a lesser risk of CMV transmission than those
from seropositive donors (61). Three of 86 seronegative recipients (3%) receiving
seronegative blood became infected compared to 13 of 54 seronegative recipients
(24%) receiving seropositive blood.
The seminal study demonstrating the efficacy of CMV-seronegative blood
products in preventing neonatal CMV infections appeared in 1981 (3). None of
90 seronegative infants given CMV-seronegative blood products became infected.
In contrast, 10 of 74 (13.5%) seronegative infants given unscreened blood
products (some of which were seropositive) became infected, 5 with serious or
fatal infections. All serious or fatal infections occurred in infants who weighed
less than 1500 g.
Seropositive infants in both the 1971 and 1981 studies showed evidence of
posttransfusion CMV infection at varying frequencies whether they received
CMV-seropositive or seronegative blood. Although it is not known whether the
infections were transfusion acquired or maternally derived, none were symptomatic.
Adding further credibility to the 1981 study from Stanford was a 1983
publication from the Medical College of Virginia, which reported on 178 transfused
neonates (35). Eight (4%) infants became infected, all of whom weighed less than
1050 g. Three of these infants died of CMV-related syndromes, and three others had
symptomatic CMV infections. Infected infants received more blood products than
uninfected infants, a greater percentage of which were CMV antibody positive, and
infected infants were more likely to be born to seronegative mothers.
The combined weight of these studies led to the implementation of donor
screening for antibodies to CMV (anti-CMV) to provide an inventory of CMV-
seronegative blood products with a reduced risk of CMV transmission to sero-
negative low birth weight infants. This became an AABB standard in the
mid-1980s.
It is unclear whether the use of seronegative blood products has been
restricted to seronegative at-risk infants despite a subsequent study from the
Stanford group showing that the use of CMV-seronegative blood products might
place low birth weight infants born to seropositive mothers at risk for developing
CMV disease (62). This was thought to be a consequence of iatrogenic blood loss
with replacement by CMV-seronegative blood, catabolic loss of remaining ma-
ternally acquired antibody, and CMV infection acquired from maternal secretions
during birth or from breast milk.
The impetus for providing CMV-seronegative blood products to CMV-
seronegative BMT recipients with CMV-seronegative marrow donors was gen-
erated by studies from Seattle (6), Glasgow (7), and Minneapolis (8). The
Minnesota and Seattle studies demonstrated that those recipients had greatly
322 Tegtmeier
reduced rates of primary CMV infections compared to recipients who were given
unscreened blood products. The results of these investigations are summarized
in Table 2.
Prevention of TA CMV infection in CMV-seronegative solid organ trans-
plant patients receiving CMV-seronegative organs by the use of CMV-safe
blood products has been recommended (60). Due to the limited transfusion
requirements of renal transplant patients, provision of CMV-seronegative blood
products is feasible. However, the more intensive use of blood in heart, heart-
lung, and liver transplants may outstrip the available CMV-screened inventory.
Fortunately, filtered products can be supplied under these circumstances.
A. Antibody Screening Tests

Early studies documenting the efficacy of anti-CMV negative blood in preventing
TA CMV infections included complement fixation and indirect hemagglutination
(3,31,61). The assays used in these studies were in-house, unstandardized assays
of undefined sensitivity and specificity. An indirect hemagglutination test based
on the one used in the 1981 Stanford study became the first commercially
available assay in 1983.
Other commercially available assays soon appeared including solid-phase
fluorescence, enzyme immunoassay (EIA), and passive latex agglutination. Com-
parative evaluation of these assays followed, which reported test sensitivities and
specificities ranging from 89 to 100% (6366). However, these evaluations were
carried out in the 1980s, and no reliable supplemental test was available to
corroborate the specificity of reactive screening test results, i.e., no gold standard
existed then or now.
Table 2 Posttransplant CMV Infections in CMV

Seronegative BMT with CMV Seronegative Marrow Donors
According to CMV Serological Status of Blood Products
No. infected/No. transfused (%)
CMV-seronegative CMV unscreened

Study blood products blood products
Miller (8) 2/45 (4) 14/44 (32)

MacKinnon (7) 0/22a (0) N.D.
Bowden (6) 1/32b (3) 8/25 (32)
aSix recipients given marrow from CMV seropositive donors.
bp < 0.007.
Currently available commercial assays for anti-CMV are listed in Table 3.

Both the Hemagen hemagglutination test and Olympus particle agglutination test
are automated assays that can be run on the Olympus PK7200. The Immucor solid
phase red cell adherence assay is semi-automated, as is the enzyme immunoassay
from Abbott. The latex agglutination test from Becton Dickinson is manual.
No comparative evaluations of these contemporary anti-CMV assays have been
published.
Despite the probable high quality of currently available commercial tests,
breakthrough infections continue to occur with low frequency in high-risk pa-
tients. The possible explanations for the failure of antibody screening are several.
The donor could be in the window period, i.e., the period of time when a newly
infected donor is infectious, but antibody negative. Strain variation may also
account for false-negative antibody tests in donors, but given the broad reactivity
of strain AD169, which serves as the basis for most commercial tests, this
explanation is not likely. Finally, donor antibody levels may fall to undetectable
levels. One publication has shown evidence for this phenomenon (67). However,
given that CMV is thought to reactivate periodically, it is doubtful that most
infected individuals would lose detectable antibody.
Although accurate data on the percentage of the U.S. blood supply being
screened for anti-CMV are unavailable, it is thought that 2025% of allogeneic
donations are screened each year to provide anti-CMV negative blood products
for high-risk patients. The extent to which FDA guidance to provide leukoreduced
blood products to all patients in the United States by 2002 will affect the use of
anti-CMV negative blood is not clear. However, a majority of the panelists at a
recent Canadian consensus conference (68) recommended the continued use of
CMV-seronegative blood components to high-risk patients despite the fact that
all cellular containing blood products in Canada are now leukoreduced.
Table 3 CMV Antibody Tests for Blood Donor Screening
Manufacturer Test name Type of test
Olympus (Fujirebio) PKTM CMV-PA Particle agglutination

Abbott CMV Total Ab EIA Enzyme immunoassay
Becton Dickinson CMV Scan Latex agglutination
Immucor Capture-CMV Solid phase red cell adherence
Hemagen PK CMV Hemagglutination
324 Tegtmeier
B. Epidemiology of CMV in Blood Donors

The wide variation in anti-CMV prevalence in blood donors around the world
was highlighted by a 1973 study (69), which reported prevalences ranging from
40% in Western Europe to 100% in third world countries. Although a systematic
study of anti-CMV prevalence in U.S. blood donors has not been published, the
previous report showed prevalences of 45% and 79% for donors from Albany and
Houston, respectively. A 1985 telephone survey of seven U.S. blood centers
yielded anti-CMV prevalence rates ranging from 30 to 70% (70).
In the United States CMV antibody prevalence varies according to age and
gender. Younger donors have lower rates than older donors, and at any given age,
female donors show higher rates than males. These relationships are illustrated in
Fig. 1. Donors 1725 years of age had a 30% prevalence, while donors over 65
years of age were 80% positive. A steady, age-related increase in prevalence can
be seen between these extremes. With the exception of the over 65 group, females
have higher prevalence rates than males.
Ethnic variations in anti-CMV prevalence seen in Kansas City donors are
summarized in Table 4. Caucasians showed the lowest prevalence (46%), while
African-American and Asian donors had higher prevalence of 64% and 76%,
respectively. Thus, the intersecting variables of geography, age, gender, and
ethnicity will determine the relative ease or difficulty that a given blood center
will have in providing an adequate inventory of CMV-seronegative blood prod-
ucts for high-risk patients.
100
Male Female Total
90
80
%Anti-CMV Positive
70
60
50
40
30
20
10
0
1725 2635 3645 4655 5664 >65 Total
Age Ranges (Years)
Figure 1 Prevalence of anti-CMV by gender and age in Kansas City blood donors
from 1992 to 1995. Data from 20,000 donors screened for the first time by enzyme
immunoassay.
Table 4 Prevalence of Anti-CMV in Kansas

City Blood Donors by Ethnicity
Number (%)
Ethnicity Tested Positivea
Caucasian 50,243 23,275 (46)

African-American 1,291 826 (64)
Asian 59 45 (76)
aBy enzyme immunoassay.
V. LEUKOREDUCTION
A 1977 publication from Duke was the first to suggest leukoreduction as a means
of reducing the risk of TA CMV infections (71). Leukocyte-poor units prepared
by inverted centrifugation of whole blood were given to eight cardiac surgery
patients, one of whom became infected compared to four of six controls who were
given unmanipulated whole blood.
The reduced risk of CMV transmission of frozen deglycerolized red cells
was demonstrated in two studies. One from Boston followed renal dialysis
patients (72). None of 21 seronegative recipients became infected after receiving
frozen blood compared to 3 seronegative recipients who became infected after
receiving conventional blood. One hundred and six seronegative neonates in
Houston were given frozen, deglycerolized red cells, none of which showed
evidence of posttransfusion CMV infection (9).
Later studies employed saline washing to reduce the leukocyte count of red
cell units. Despite this intervention, one published in 1986 reported 6 of 54
neonates (11%) with posttransfusion CMV infection (73). A second reported
seroconversion of one infant of 76 transfused (1.5%) with at least one saline-
washed CMV-seropositive unit (74). No contemporary controls were available
for comparison.
An Australian study was the first to demonstrate the effectiveness of
leukoreduction by filtration (75). Nine of 42 infants (21%) receiving unfiltered
seropositive blood became infected compared to 0 of 30 infants given filtered
blood.
None of 48 premature neonates given filtered blood in three Connecticut
intensive care nurseries became infected with CMV (76). No control population
was followed by these investigators. It should be noted that only the studies from
Duke and Australia included controls.
Six additional studies (10,58,7780) using either filtration and/or centrifu-
gation to leukocyte-reduce red blood cells and platelets given to patients with
326 Tegtmeier
hematological malignancies or to BMT patients have further documented the

efficacy of leukoreduction as a means of preventing TA CMV infections. The
results are summarized in Table 5. Two of these studies were randomized (10,58),
one was a cohort study (77), two represented case series (78,79), and one
compared leukoreduction to prior experience using CMV-seronegative blood
products (80).
The collective weight of the evidence from these investigations support the
conclusion that leukoreduced blood components carry a much reduced risk of
CMV transmission. In 1997, the Ad Hoc Committee on Prevention of CMV
Transmission of the AABB concluded that the leukocyte reduction level currently
accepted for the prevention of alloimmunization to HLA molecules reduces
CMV transmission to a level at least equivalent with that observed with CMV-
seronegative components (60).
Identical standards have been set by AABB and FDA to qualify blood
components as leukoreduced. Red blood cells and apheresis platelets must contain
<5 106 WBCs per unit. Platelets derived from whole blood must contain <8.3
105 WBC per unit or <5 106 WBC per pool of six.
Quality control of leukoreduced products is essential. It is clear that more
consistent results are obtained when blood components are filtered in the labora-
tory rather than at the bedside. Moreover, bedside filters have yielded results at
odds with those generated by manufacturers (81), and, until recently, counting the
number of residual WBCs has been difficult. Filter failures have been reported in
Table 5 Prevention of TA CMV Infections in BMT Patients and

Patients with Hematological Malignancies by Leukoreduction
No. with CMV

infection (%)/
No. transfused
Author (Ref.) Year Control LRa
Murphy (58) 1988 2/9 (22) 0/11

DeGraan-Hentzen (77) 1989 10/86 (12) 0/59
DeWitte (78) 1991 None 0/28
Bowden (10) 1991 7/30 (23) 0/35
Van Prooijen (79) 1994 None 0/60
Pamphilon (80) 1999 0/114b 0/62
aLR = Leukoreduced.
bExperience in patients given CMV-seronegative blood.
donors with hemoglobin AS (82,83), and adverse recipient reactions have sur-
faced recently, e.g., red-eye syndrome (84).
VI. ARE LEUKOREDUCTION AND ANTIBODY

SCREENING EQUALLY EFFECTIVE?
Only one prospective, randomized, controlled trial has been carried out to date
that has sought to answer this question (11). Bedside filters from one manufac-
turer were used to provide leukoreduced blood components to patients random-
ized to one arm of this study, while CMV-seronegative blood components were
supplied to the other. Recipients were CMV-seronegative BMT patients who
received marrow from CMV-seronegative donors.
In the primary analysis, defined as infections occurring after day 21
posttransplant, there was no significant difference between the probability of
CMV infection or disease in patients receiving CMV-seronegative blood (0.8%;
95% CI 0.128%) compared to patients receiving leukoreduced blood compo-
nents (1.2%; 95% CI 0.35.0%). No patients in the former group and three
patients in the latter group developed CMV disease.
A secondary analysis was performed that included all infections diagnosed
between day 0 and day 100. Similar rates of CMV infection were seen in the
CMV-screened group (1.6%) and the leukoreduced group (2.4%). However, no
CMV disease was observed in patients receiving CMV-seronegative blood com-
ponents compared to 2.4% in patients given leukoreduced components. All six
infected patients in the filtered arm died, whereas the four infected patients in the
CMV-screened arm survived without evidence of disease.
The results of the secondary analysis were surprising. The 100% mortality
of infected patients in the filtered arm has been ascribed to a statistical anomaly.
The authors of this paper also pointed out that four of five patients who developed
CMV infection before day 21 had discrepant or equivocal CMV antibody tests at
randomization and suggested that these patients may have been infected prior to
the receipt of blood components in the study. Some believe this study over-
estimates the rate of CMV infection and disease in patients who received
leukoreduced components.
At the conclusion of this trial, CMV-seronegative blood products were
again given to seronegative patients with CMV-seronegative donors in Seattle. Of
117 patients followed, two became infected and one developed CMV-related
disease. Based on this experience, it was concluded that CMV infection occurs
in 23% of patients regardless of the interventions employed to prevent it and
that CMV disease occurs in 12%. Since 1994, both CMV-seronegative and
leukoreduced blood products have been used to support transplant patients in
Seattle (85).
328 Tegtmeier
Since 1997 apheresis platelets that qualify as leukoreduced have been part
of the CMV-safe product mix used to support BMT patients in Seattle. Of 387
patients transplanted during this interval, 4% have experienced primary CMV
infection. It is unclear whether pheresis platelets carry a higher risk of CMV
transmission compared to filtered products, but the question begs to be answered.
Thus, there remains an intractable low level of risk for primary CMV infection
in CMV-seronegative transplant patients who are supported with CMV-safe
blood products (85).
VII. WHAT IS THE SOURCE OF RESIDUAL RISK?

The limitations of CMV antibody screening and leukoreduction techniques have
been outlined previously. A plausible, but yet unproved, hypothesis is that the
residual risk posed by CMV-safe blood products is a consequence of window
period donations. Data presented earlier have documented high levels of cell-
associated CMV DNA and lower levels of CMV DNA in plasma of recently
infected immunocompetent individuals who experienced asymptomatic CMV
infections. If plasma DNA is infectious, and this remains to be proven, then
individuals early in the course of an asymptomatic infection would be seronega-
tive and potentially infectious. Such donors would not be identified by antibody
testing, and although infected cells would be retained by a filter, CMV DNA in
plasma is not likely to be retained. Molecular testing of donor blood may soon
provide the means of finally resolving the problem of TA CMV infections.
REFERENCES
1. Ho M, Cytomegalovirus: Biology and Infection. 2nd ed. New York: Plenum, 1991.
2. Kaariainen L, Klemola E, Paloheimo J. Rise of cytomegalovirus antibodies in an
infectious mononucleosis-like syndrome after transfusion. Br Med J 1966; 1:12701272.
3. Yeager AS, Grumet FC, Hafleigh EB, Arvin AR, Bradley JS, Prober CG. Prevention
of transfusion-acquired cytomegalovirus infections in newborn infants. J Pediatr
1981; 98:281287.
4. Ho M, Suwansirikul S, Dowling JN, Youngblood LA, Armstrong JA. The trans-
planted kidney as a source of cytomegalovirus infection. N Engl J Med 1975;
293:11091112.
5. Meyers JD, Flournoy N, Thomas ED. Risk factors for cytomegalovirus infection after
human marrow transplantation. J Infect Dis 1986; 153:478488.
6. Bowden RA, Sayers M, Flournoy N, Newton B, Banaji M, Thomas ED, Meyers JD.
Cytomegalovirus immune globulin and seronegative blood products to prevent
primary cytomegalovirus infection after marrow transplantation. N Engl J Med 1986;
314:10061010.
7. Mackinnon S, Burnett AK, Crawford RJ, Cameron S, Leask BG, Sommerville RG.
Seronegative blood products prevent primary cytomegalovirus infection after bone

marrow transplantation. J Clin Pathol 1988; 41:948950.
8. Miller WJ, McCullough J, Balfour HH Jr, Haake RJ, Ramsay NK, Goldman A,
Bowman R, Kersey J. Prevention of cytomegalovirus-infection following bone
marrow transplantation: a randomized trial of blood product screening. Bone Marrow
Transplant 1991; 7:227234.
9. Brady MT, Milam JD, Anderson DC, Hawkins EP, Speer ME, Seavy D, Bijou H,
Yow MD. Use of deglycerolized red blood cells to prevent posttransfusion infection
with cytomegalovirus in neonates. J Infect Dis 1984; 150:334339.
10. Bowden RA, Slichter SJ, Sayers MH, Mori M, Cays MJ, Meyers JD. Use of
leukocyte-depleted platelets and cytomegalovirus-seronegative red blood cells for
prevention of primary cytomegalovirus infection after marrow transplant. Blood
1991; 79:246250.
reduced and cytomegalovirus (CMV) seronegative blood products for the prevention
of transfusion-associated CMV infection after marrow transplant. Blood 1995;
86:35983603.
12. Mocarski Jr. ES. Cytomegaloviruses and their replication. In: Fields BN, Knipe DM,
Howley PM, eds. Virology. Philadelphia: Lippincott, Raven, 1996:24472492.
13. Taylor-Wiedeman J, Sissons JGP, Sinclair JH. Monocytes are a major site of
persistence of human cytomegalovirus in peripheral blood mononuclear cells. J Gen
Virol 1991; 72:20592064.
14. Sderberg-Nauclr C, Fisk KN, Nelson JA. Reactivation of latent human cytomega-
lovirus by allogeneic stimulation of blood cells from healthy donors. Cell 1977;
91:119126.
15. Hahn G, Jores R, Mocarski ES. Cytomegalovirus remains latent in a common
precursor of dendritic and myeloid cells. Proc Natl Acad Sci 1998; 95:39373942.
16. Slobedman B, Mocarski ES. Quantitative analysis of latent human cytomegalovirus.
J Virol 1999; 78:48064812.
17. Diosi P, Moldovan E, Tomescu N. Latent cytomegalovirus infection in blood donors.
Br Med J 1969; 4:660662.
18. Beneke JS, Tegtmeier GE, Alter JH, Luetkemeyer RB, Solomon R, Bayer WL.
Relation of titers of antibodies to CMV in blood donors to the transmission of
cytomegalovirus infection. J Infect Dis 1984; 150:883888.
19. Lamberson Jr. HV, McMillan JA, Weiner LB, Williams ML, Clark DA, McMahon
CA, Lentz EB, Higgins AP, Dock NL. Prevention of transfusion-associated cytomeg-
alovirus (CMV) infection in neonates by screening blood donors for IgM to CMV. J
Infect Dis 1988; 157:820823.
20. Revello MG, Zavattoni M, Sarasini A, Percivalle E, Simoncini L, Gerna G. Human
cytomegalovirus in blood of immunocompetent persons during primary infection:
prognostic implications for pregnancy. J Infect Dis 1998; 177:11701175.
21. Zhanghellini F, Boppana SB, Emery VC, Griffiths PD, Pass RF. Asymptomatic
primary cytomegalovirus infection: virologic and immunologic features. J Infect Dis
1999; 180:702707.
22. Tegtmeier GE, Henderson SE, Blosser JK, Drew WL, Miner R. CMV DNA in plasma
330 Tegtmeier
of seroconverting and anti-CMV seroprevalent blood donors (abstract). Transfusion

1999; 39:116S.
23. Stanier P, Taylor DL, Kitchen AD, Wales N, Tryhorn Y, Tyms AS. Persistence of
cytomegalovirus in mononuclear cells in peripheral blood from blood donors. Br Med
J 1989; 299:897898.
24. Bevan IS, Daw RA, Day PJ, Ala FA, Walker MR. Polymerase chain reaction for
detection of human cytomegalovirus infection in a blood donor population. Br J
Haematol 1991; 78:9499.
25. Larsson S, Soderberg-Nauclr C, Wang FZ, Moller E. Cytomegalovirus DNA can be
detected in peripheral blood mononuclear cells from all seropositive and most
seronegative healthy blood donors over time. Transfusion 1998; 38:271278.
26. Bitsch A, Kirchner H, Dupke R, Bein G. Failure to detect human cytomegalovirus
DNA in peripheral blood leukocytes of healthy blood donors by the polymerase chain
reaction. Transfusion 1992; 32:612617.
27. Nolte FS, Emmens RK, Thurmond C, Mitchell PS, Pascuzzi C, Devine SM, Saral R,
Wingard JR. Early detection of human cytomegalovirus viremia in bone marrow
transplant recipients by DNA amplification. J Clin Microbiol 1995; 33:12631266.
28. Roback JD, Hillyer CD, Drew WL, Luka J, Mocarski ES, Soderberg-Naucler C,
Woxenius S, Busch MP. Comparison of 7 PCR assays for detection of cytomegalo-
virus (CMV) DNA in PBMCS of blood donors: results of a blinded multicenter study
(abstr). Blood 1999; 94:246a.
29. Winston DJ, Ho WG, Howell CL, Miller MJ, Mickey R, Martin WJ, Lin CH, Gale
RP. Cytomegalovirus infections associated with leukocyte transfusions. Ann Intern
Med 1980; 93:671675.
30. Hersman J, Meyers JD, Thomas ED, Buckner CD, Clift R. The effect of granulocyte
transfusions on the incidence of cytomegalovirus infection after allogeneic marrow
transplantation. Ann Intern Med 1982; 96:149152.
31. Luthardt T, Siebert H, Lsel I, Quevedo M, Todt R. Cytomegalovirus-infektionen bei
Kindern mit Blutaustausch-transfusion im Neugeborenenalter. Klin Wochenschr
1971; 49:8186.
32. Bowden R, Sayers M. The risk of transmitting cytomegalovirus infection by fresh
frozen plasma. Transfusion 1990; 30:762763.
33. Wilhelm JA, Matter L, Schopfer K. The risk of transmitting cytomegalovirus to
patients receiving blood transfusion. J Infect Dis 1986; 154:169171.
34. Preiksaitis JK, Brown L, McKenzie M. The risk of cytomegalovirus infection in
seronegative transfusion recipients not receiving exogenous immunosuppression. J
Infect Dis 1988; 157:523529.
35. Adler SP, Chandrika T, Lawrence L, Baggett J. Cytomegalovirus infections in
neonates acquired by blood transfusions. Pediatr Infect Dis 1983; 2:114118.
36. Smith D Jr, Wright P, Estes W, Krueger L, Wallas C, Moldovan R, Tanley P.
Posttransfusion cytomegalovirus infection in neonates weighing less than 1250 grams
(abstr). Transfusion 1983; 23:420.
37. Tegtmeier GE. The use of cytomegalovirus-screened blood in neonates. Transfusion
1988; 28:201203.
38. Preiksaitis JK, Brown L, McKenzie M. Transfusion-acquired cytomegalovirus infec-
tion in neonates. A prospective study. Transfusion 1988; 28:205209.
39. Stagno S. Cytomegalovirus. In: Remington J, Klein G, eds. Infectious Diseases of the
Fetus and Newborn. Philadelphia: WB Saunders, 1995:312353.
40. Marker SC, Howard RJ, Simmons RL, Kalis JM, Connelly DP, Najarian JS, Balfour
Jr. HM. Cytomegalovirus infection: A quantitative prospective study of three hundred
consecutive renal transplants. Surgery 1981; 89:660671.
41. Harris KR, Saeed AA, Digard NJ, Whiteford K, Geoghegan TA, Lee HA, Slapak M.
Cytomegalovirus titers in kidney transplant donor and recipient: influence of cyclo-
sporine A. Transplant Proc 1984; 16:3133.
42. Rubin RH, Tolkoff-Rubin NE, Oliver R, Rota TR, Hamilton J, Betts RF, Pass RF,
Hillis W, Szmuness W, Farrel ML, Hirsch MS. Multi-center seroepidemiologic study
of the impact of cytomegalovirus infection on renal transplantation. Transplantation
1985; 40:243249.
43. Boyce NW, Hayes K, Gee D, Holdsworth SR, Thomson NM, Scott D, Atkins RC.
Cytomegalovirus infection complicating renal transplantation and its relationship to
acute transplant glomerulopathy. Transplantation 1988; 45:706709.
44. Chou S, Norman DJ. The influence of donor factors other than serologic status on
transmission of cytomegalovirus to transplant recipients. Transplantation 1988;
46:8993.
45. Preiksaitis JK, Rosno S, Grumet C, Merigan TC. Infections due to herpesviruses in
cardiac transplant recipients: role of the donor heart and immunosuppressive therapy.
J Infect Dis 1983; 147:974981.
46. Wreghitt TG, Hakim M, Gray JJ, Kucia S, Wallwork J, English TA. Cytomegalovirus
infections in heart and heart and lung transplant recipients. J Clin Pathol 1988;
41:660667.
47. Dummer JS, Montero CG, Griffith BP, Hardesty RL, Paradis IL, Ho M. Infections in
heart-lung transplant recipients. Transplantation 1986; 41:725729.
48. Wreghitt T. Cytomegalovirus infections in heart and heart-lung transplant recipients.
J Anti-Microb Chem 1989(suppl E); 23:4960.
49. Smyth RL, Scott JP, Borysiewicz LK, Sharples LD, Steart S, Wreghitt T, Gray JJ,
Higenbottam TW, Wallwork J. Cytomegalovirus infection in heart-lung transplant
recipients: risk factors, clinical associations and response to treatment. J Infect Dis
1991; 164:10451050.
50. Singh N, Dummer JS, Kusne S, Breinig MK, Armstrong JA, Makowka L, Starzl TE,
Ho M. Infections with cytomegalovirus and other herpesviruses in 121 liver trans-
plant recipients: transmission by donated organ and the effect of OKT3 antibodies. J
Infect Dis 1988; 158:124131.
51. Barkholt LM, Ericzon B-G, Ehrnst A, Forsgren M, Anderson JP. Cytomegalovirus
infections in liver transplant patients: incidence and outcome. Transplant Proc 1990;
22:235237.
52. Gorensek MJ, Carey WD, Vogt D, Goormastic M. A multivariate analysis of risk
factors for cytomegalovirus infection in liver transplant recipients. Gastroenterology
1990; 98:13261332.
53. Pillay D, Charman H, Burroughs AK, Smith M, Rolles K, Griffiths PD. Surveillance
for CMV infections in orthotopic liver transplant recipients. Transplantation 1992;
53:12611265.
54. Sutherland S, Bracken P, Wreghitt TG, OGrady J, Calne R, Williams R. Donated
332 Tegtmeier
organ as a source of cytomegalovirus in orthotopic liver transplantation. J Med Virol

1992; 37:170173.
55. Manez R, Kusne S, Martin M, Linden P, Irish W, Torre-Cisneros J, Kramer D, Ho M,
Starzl TE. The impact of blood transfusion on the occurrence of pneumonitis in
primary cytomegalovirus infection after liver transplantation. Transfusion 1993;
33:594597.
56. Falagas ME, Snydman DR, Ruthazer R, Griffith J, Werner BG. Primary cytomegalo-
virus infection in liver transplant recipients: comparison of infections transmitted via
donor organs and via transfusions. Clin Infec Dis 1996; 23:292297.
57. Cox F, Hughes WT. Cytomegaloviremia in children with acute leukemia. J Pediatr
1975; 87:190194.
58. Murphy MF, Grint PC, Hardiman AE, Lister TA, Waters AH. Use of leukocyte-poor
blood components to prevent primary cytomegalovirus (CMV) infection in patients
with acute leukemia (letter). Br J Haematol 1988; 70:253254.
59. Preiksaitis JK, Desai S, Vaudry W, Roberts S, Akabutu J, Grundy P, Wilson B,
Boshkov L, Hannon J, Joffres M. Transfusion- and community-acquired cytomega-
lovirus infection in children with malignant disease: a prospective study. Transfusion
1997; 37:941946.
60. American Association of Blood Banks, Association Bulletin #97-2, Leukocyte reduc-
tion for the prevention of TA cytomegalovirus. Bethesda, MD: American Association
of Blood Banks, 1997.
61. Monif GRG, Daicoff GI, Flory LF. Blood as a potential vehicle for the cyto-
megaloviruses. Am J Obstet Gynecol 1976; 126:445448.
62. Yeager AS, Palumbo PE, Malachowski N, Ariagno RL, Stevenson DK. Sequelae of
maternally derived cytomegalovirus infections in premature infants. J Pediatr 1983;
102:918922.
63. Phipps PH, Gregoire L, Rossier E, Perry E. Comparison of five methods of cytomeg-
alovirus antibody screening of blood donors. J Clin Microbiol 1983; 18:12961300.
64. Beckwith DG, Halstead DC, Alpaugh K, Schweder A, Blount-Fronefield DA, Toth K.
Comparison of a latex agglutination test with five other methods for determining the
presence of antibody against cytomegalovirus. J Clin Microbiol 1985; 21:328331.
65. Adler SP, McVoy M, Biro VG. Detection of cytomegalovirus antibody with latex
agglutination. J Clin Microbiol 1985; 22:6870.
66. Taswell HF, Reisner RK, Rabe DE, Shelley CD, Smith TF. Comparison of three
methods for detecting antibody to cytomegalovirus. Transfusion 1986; 26:285289.
67. Waner JL, Weller TH, Kevy SV. Patterns of cytomegalovirus complement fixing
antibody activity: A longitudinal study of blood donors. J Infect Dis 1973; 127:538543.
68. Canadian Blood Services, Canadian Blood and Marrow Transplant Group, Consensus
Conference: Prevention of Post-Transfusion CMV in the Era of Universal Leuko-
reduction, Toronto, Canada, January 78, 2000.
69. Krech U. Complement-fixing antibodies against cytomegalovirus in different parts of
the world. Bull WHO 1973; 49:103106.
70. Tegtmeier GE. Transfusion-transmitted cytomegalovirus infections: significance and
control. Vox Sang 1986; 51(suppl1):2230.
71. Lang DJ, Ebert PA, Rodgers BM, Boggess HP, Rixse RS. Reduction of postperfusion
cytomegalovirus infections following the use of leukocyte-depleted blood. Transfu-

sion 1977; 17:391395.
72. Tolkoff-Rubin NE, Rubin RH, Keller EE, Baker GP, Stewart JA, Hirsch MS.
Cytomegalovirus infection in dialysis patients and personnel. Ann Intern Med 1978;
89:625628.
73. Demmler GJ, Brady MT, Bijou H, Speer ME, Milam JD, Hawkins EP, Anderson DC,
Six H, Yow MD. Posttransfusion cytomegalovirus infection of neonates: role of
saline-washed red blood cells. J Pediatr 1986; 108:762765.
74. Luban NL, Williams AE, MacDonald MG, Mikesell GT, Williams KM, Sacher RA.
Low incidence of cytomegalovirus infection in neonates transfused with washed red
blood cells. Am J Dis Child 1987; 141:416419.
75. Gilbert GL, Hayes K, Hudson IL, James J. Prevention of transfusion-acquired
cytomegalovirus infection in infants by blood filtration to remove leukocytes. Lancet
1989; 2:12281231.
76. Eisenfeld L, Silver H, McLaughlin J, Klevjer-Anderson P, Mayo D, Anderson J,
Herson V, Krause P, Savidakis J, Lazar A. Prevention of transfusion-associated
cytomegalovirus infection in neonatal patients by the removal of white cells from
77. de Graan-Hentzen YC, Gratama JW, Mudde GC, Verdonck LF, Houbiers JG, Brand
A, Sebens FW, van Loon AM, The TH, Willemze R. Prevention of primary cytomeg-
alovirus infection in patients with hematologic malignancies by intensive white cell
depletion of blood products. Transfusion 1989; 29:757760.
78. De Witte T, Schattenberg A, Van Dijk BA, Galama J, Olthuis H, Van der Meer JW,
Kunst VA. Prevention of primary cytomegalovirus infection after allogeneic bone
marrow transplantation by using leukocyte-poor random blood products from cyto-
megalovirus-unscreened blood bank donors. Transplantation 1990; 50:964968.
79. van Prooijen HC, Visser JJ, van Oostendorp WR, de Gast GC, Verdonck LF.
Prevention of primary transfusion-associated cytomegalovirus infection in bone
marrow transplant recipients by the removal of white cells from blood components
with high-affinity filters. Br J Haematol 1994; 87:144147.
80. Pamphilon DH, Rider JR, Barbara JA, Williamson LM. Prevention of transfusion-
transmitted cytomegalovirus infection. Transf Med 1999; 9:115123.
81. Ledent E, Berlin G. Inadequate white cell reduction from red cell concentrates by
filtration. Transfusion 1994; 34:765768.
82. Bodensteiner D. White cell reduction in blood donors with sickle trait (letter).
83. Ould Amar AK, Cesaire R, Krob-Bauchet B, Robert P, Maier H, Bucher B. Altered
filterability of fresh sickle trait donor blood (letter). Vox Sang 1997; 73:5556.
84. Centers for Disease Control and Prevention. Adverse ocular reactions following
transfusion-United States, 19971998. MMWR 1998; 47:4950.
85. Nichols G. Is leukoreduction a suitable alternative to CMV screening? Are there
alternate methods? Canadian Blood Services, Canadian Blood and Marrow Trans-
plant Group Consensus Conference, Prevention of Post-Transfusion CMV in the Era
of Universal Leukoreduction, Toronto, Canada, January 78, 2000.
15
Creutzfeldt-Jakob Disease,
Other Human Transmissible
Spongiform Encephalopathies,
and Transfusion of Blood and
Blood Products
Celso Bianco
I. INTRODUCTION
This chapter addresses human transmissible spongiform encephalopathies (TSEs)

and the theoretical possibility of transmission of these diseases by the transfusion
of blood and blood products. TSEs are rapidly progressing, fatal diseases,
characterized by mental deterioration, cerebellar dysfunctions, involuntary move-
ments, and psychiatric alterations. The brains of these patients show spongiform
degeneration and deposition of an amyloid protein called prion protein (PrPres).
This protein is the protease-resistant form of a cellular protein called PrPc, which
is enclosed by a gene designated PRNP. PrPres is pathognomonic for this group
of illnesses. The diagnosis of human TSE can be strongly suspected on clinical
grounds (e.g., by a characteristic electroencephalogram in a number of cases), but
it can be only confirmed by histological or biochemical investigations of brain
material after biopsy or autopsy. Over 85% of the cases of human TSEs are
sporadic, appearing in individuals without family history. About 10% of the cases
are familial. More than 80 cases of iatrogenic transmission of human TSEs have
been reported. A new form of Creutzfeldt-Jakob disease (CJD) named new variant
CJD (nvCJD) was described in 1996, and there is evidence indicating that it is
335
336 Bianco
related to bovine spongiform encephalopathy (BSE). These issues have been

reviewed (13).
II. NATURE OF THE TRANSMISSIBLE AGENT

The causative agent of TSEs is controversial. Initially, the etiology of TSE was
attributed to a slow virus infection. However, transmission studies performed in
the scrapie model indicated that the infective material was devoid of measurable
amounts of nucleic acid. The evidence for an infectious protein, the prion, was
carefully reviewed (4). Essentially, the prion is a conformationally altered form
of a normal cellular membrane protein called PrPc. Abnormal PrP, PrPsc, or PrPres
can induce conformational changes in normal PrP and is consequently infec-
tious. These conformational changes may require cofactors. PrPsc and PrPres are
highly resistant to proteases and form rod-shaped multimers that precipitate as
amyloid-causing spongiform alterations in the brain (5). Some authors still favor
a viral etiology for CJD or believe that the cofactor for the alterations of PrP is a
virus (6). Currently, the majority of investigators in the area accept the prion
etiology for TSEs.
III. TSEs
The major human TSEs are kuru, sporadic CJD, iatrogenic CJD, nvCJD, familial
CJD, Gerstmann-Strassler-Scheinker syndrome (GSS), and fatal familial insom-
nia (FFI). Animal TSEs have been known for many years and are useful models
for the human disease. Among them are scrapie in sheep and goats, transmissible
mink encephalopathy (TME), chronic wasting disease of elks, and bovine spongi-
form encephalopathy (BSE), also known as mad cow disease. TSEs appear to
be transmitted by the ingestion of infected tissues from diseased animals. This has
been documented for both TME and BSE (2,3).
A. Kuru
Kuru was the first TSE ever recognized. It occurred among the Fore peoples
inhabiting the Central Highlands of Papua-New Guinea. It was apparently spread
by cannibalism performed as a rite of respect and mourning for the dead. Since
the cessation of cannibalism in the 1950s, the disease has nearly disappeared.
The incubation time exceeded 40 years. The disease was transmitted experimen-
tally to different species of Old and New World monkeys and apes in classic
experiments by Asher, Gibbs, and Gajdusek. Brain contained the highest amount
of infectivity, while serum and blood did not transmit the infection (7). Occasion-
ally, spleen and lymph nodes transmitted disease after intracerebral inoculations,
Human Transmissible Spongiform Encephalopathies 337
but not by administration through a nasogastric tube (four attempts with chim-
panzees) (7). Clinically, Kuru is characterized by a preponderance of cerebel-
lar symptoms; mental deterioration appears later in the course of the disease.
In histological sections, relatively large deposits of PrP (so-called kuru plaques)
can be observed.
B. Sporadic CJD
The sporadic form of CJD is the most common type of human TSE. CJD was first
described in 192021. Most affected individuals die within one year of onset of
symptoms (reviewed in Refs. 4,5). In many cases, a characteristic electroenceph-
alogram with periodic triphasic waves can be observed. The human disease can
be transmitted to nonhuman primates (7) and less regularly to nonprimates (cats,
guinea pigs, mice, and hamsters) (8).
The incidence of CJD has been extensively analyzed by Schonberger et al.
from the CDC. There were 3642 deaths between 1979 and 1994, or 228 deaths/
year, for a rate of about one case per million inhabitants per year in the United
States. The average annual age-adjusted death rate during the study period was
0.95 deaths per million persons, ranging from 0.78 in 1980 to 1.11 in 1987 (9).
The incidence was zero among 5- to 19-year-olds, and reached 3.65.8/1,000,000
among individuals over 60 years of age. The incidence has remained constant
over the years and was the same in all U.S. states. There are no reports of
transmission of CJD by sexual or casual contact. The overall mortality rate in
Europe is 0.69 at one year (10). Interestingly, the phenotypic expression of the
disease may be linked to polymorphism at codon 129 of PrP. In contrast to the
normal population, 69% of CJD patients were homozygous for methionine at
codon 129 (vs. 42% of controls). Homozygosity on codon 129 is considered to
be a genetic predisposing factor for CJD (11).
Clinically, sporadic CJD it is characterized by an early onset of mental
dysfunctions (e.g., memory loss or behavioral abnormalities). However, in a
substantial proportion of cases (15% in a large study of confirmed CJD cases),
cerebellar symptoms are the first signs of the disease (12). The most extensive
experience of experimental transmission of TSEs was reported by Brown et al.
(12) The most effective route of transmission is intracerebral inoculation, and the
highest infectivity is found in the brain of infected individuals, followed by spinal
cord, cerebrospinal fluid, and eye.
C. Iatrogenic CJD
The possibility of iatrogenic transmission of TSEs was first raised in 1974, with
the association of a case of CJD with a corneal transplant performed 18 months
earlier (13). A second case was described more recently with onset of disease 30
338 Bianco
years after a corneal transplant (14). However, these cases are not well docu-
mented, and transmission of CJD by corneal transplants is unlikely. Silver
electrodes used for stereotactic electroencephalograms during neurosurgery trans-
mitted CJD to two patients, although the electrodes had been cleaned and
sterilized by 70% ethanol and formaldehyde vapor between uses (15). No other
transmissions by either of these routes have been reported.
Starting in 1985, CJD was identified among 7 of 6284 recipients of human
pituitary-derived growth hormone prepared from human cadaveric pituitary
glands (16). Concerned about the theoretical possibility of transmission by
transfusion, the U.S. Food and Drug Administration (FDA) recommended as a
precautionary measure that individuals who received human pituitary-derived
growth hormone be deferred from donating blood (November 25, 1987). In
Europe, development of CJD among recipients of human pituitary-derived growth
hormone was also a serious problem. By November 1994, 11 cases had been
reported in the United States, 14 in the United Kingdom, and 32 in France
(L. Schonberger, personal communication). Large batch sizes contributed to the
spread. Duration of the disease was significantly longer than that observed in
sporadic CJD. Estimated mean incubation time was 8.9 years in France and more
than 10 years in the United Kingdom and the United States. Clinical symptoms
consisted mainly of cerebellar ataxia. Homozygosity for codon 129 of the PRNP
gene, the main genetic determinant for susceptibility to CJD, was significantly
associated with disease. The mean incubation time was significantly longer in
heterozygous individuals than in homozygous individuals (11 vs. 8.95 years).
There have been four transmissions traceable to gonadotropin use in Australia.
Gonadotropin was also prepared from cadaveric human pituitary glands but has
had a far more limited use than human growth hormone.
Iatrogenic transmission of CJD has also been documented among recipients
of human dura mater transplants (17). Incubation times were less than 2 years for
some patients. Most, but not all, of the dura mater transmissions have been
associated with use of a single product manufactured in large batches by a single
manufacturer. To date approximately 60 cases have been traced to this source.
Recently, Japan identified 43 additional cases associated with transplants of dura
mater (18). Most of the transmissions occurred in codon 129 homozygous
individuals.
D. New Variant CJD

A new variant of CJD was first described in 1996 (19,20). By July 1999 there had
been 39 cases identified in the United Kingdom, one in France, and one in the
Republic of Ireland. The major characteristic of nvCJD is the age of onset: all
patients have been under the age of 50, while sporadic CJD occurs mainly in the
elderly, with a peak of incidence at an age of 70 (9). The histological picture is
characterized by relatively large and abundant deposits of PrP (kuru-like plaques)

surrounded by prominent vacuoles (florid plaques) in the neutrophil. In contrast
to sporadic CJD, PrP deposits have been found in peripheral lymphoid organs
(tonsils, spleen) (21). This disease seems to be associated with BSE for several
reasons: the geographical distribution, the clinical presentation, the molecular
similarity of the nvCJD- and BSE-associated PrPs (22), and the biological
properties of the infectious agents (23).
E. Familial TSEs
Three types of CJD, namely familial CJD, GSS (24), and FFI (25), are hereditary
diseases. They are associated with mutations in the normal cellular gene that
codes for the prion protein. GSS evolves very slowly, over a period of years.
Patients with FFI are unable to sleep. Transmission of these TSEs to animals has
been occasionally successful (12).
F. Animal TSEs
Scrapie in sheep has been known for more than 200 years. In 1961, scrapie was
successfully transmitted to laboratory mice (26). Gajdusek and Gibbs were able
to transmit the disease to New and Old World monkeys (reviewed in Refs. 27,28).
In 1975, Manuelidis described the transmission of CJD to guinea pigs by
intracerebral injection of brain tissue (29). These observations were extended to
hamsters, mice, and rats (29,30). The exact route of natural transmission of
scrapie has not been elucidated. Animal experiments are complicated by the
species barrier, i.e., an animal species is less sensitive to a TSE agent isolated
from another species than the originally infected species. In addition, intracereb-
ral injections accept only a limited sample volume (3050 L), raising the
possibility that infectivity is frequently underestimated. Mad cow disease, or
bovine spongiform encephalopathy has been attributed to the practice of feeding
cattle with sheep offal.
IV. RELATIONSHIP BETWEEN nvCJD AND

MAD COW DISEASE
A cluster of nvCJD cases with a unique neuropathological picture among young
patients was reported in 1996 (31). The authors of the report suggested that these
patients were infected by a new strain of prions related to mad cow disease.
The molecular characteristics of the variant strains have been determined (32).
One case of variant CJD has been reported in France (33), and another appears
to have been identified in the Republic of Ireland (commentary in Lancet, June
340 Bianco
26, 1999). The public concern was so intense that it led the European Community
to ban importation of British beef until measures designed to control the spread
of the infection were instituted. The ban was lifted as of August 1,1999. Little is
known about the relative efficiency of various routes of infection by the BSE or
the nvCJD agent. However, the spread of the disease in cattle and its transmission
to other species, presumably through food, suggests that the oral route may be
efficient. In the United States, the Centers for Disease Control and Prevention
(CDC) established surveillance systems for CJD and CJD variants (34). No cases
of BSE have so far been identified. In addition, an analysis of CJD between 1979
and 1994 found no evidence of the variants form of CJD (35). Unfortunately,
because of the very low incidence of CJD and the long incubation period, there
will be a long period before more definitive answers become available.
V. CJD AND BLOOD TRANSFUSION

On October 18, 1994, the American Red Cross (ARC) reported to FDA that a
64-year-old blood donor who had donated more than 90 times over more than 30
years had died with a clinical diagnosis of CJD. Plasma from his donations was
often pooled for further manufacture of plasma derivatives. On October 27, 1994,
ARC voluntarily recalled components of the last four donations made by the
donor. On November 17, 1994, Baxter and ARC initiated voluntary market
withdrawal of implicated lots of IVIG, Factor VIII AHF, albumin, and plas-
ma protein fraction. Soon, Miles withdrew alphaI-proteinase inhibitor lots and
Sandoz withdrew IVIG that contained donations made by this donor. In Novem-
ber 1994 hemophilia treaters in New York initiated notification of patients who
had received implicated lots of Factor VIII AHF that they had received prod-
ucts that contained plasma from a donor who later developed CJD. This set of
events triggered worldwide concerns about transmissibility of CJD by blood and
blood products.
VI. EXPERIMENTAL ATTEMPTS TO TRANSMIT

TSEs BY BLOOD
The initial concerns about transmissibility of CJD by transfusion were raised by
Manuelidis (36), who inoculated buffy coat cells from peripheral blood of two
CJD patients into the brains of rodents. After 200500 days, these animals showed
spongiform degeneration of the brain, while control experiments never resulted
in CJD. In 1993, these investigators inoculated buffy coat cells from normal
volunteers with no family history of dementia into the brains of hamsters. After
a long period of observation, 26 of 30 buffy coats (86.7%) induced CJD-like
changes in the animals brains. The investigators concluded that the CJD agent
endemically infects humans but only infrequently produces dementia (37).

This interpretation is highly controversial, because these are the logical controls
for the early experiments performed with CJD buffy coats and raise serious
questions about assay specificity. Transmission of classical CJD to animals by
intracerebral inoculation was attempted in other studies (38,39). Data presented
in these papers are conflicting. For instance, one author finds CJD infectivity in
blood clots injected intracerebrally into mice in one out of three patients (38).
Another author injected plasma of a pregnant woman with CJD intracerebrally
into mice (39). The neat plasma was noninfectious, but after threefold concentra-
tion it was highly infectious. In addition, the patients leukocytes were negative,
while cord blood was infective. Another author described transmission of iatro-
genic CJD by buffy coat injected into hamsters but provides no details of dose,
route, or development of alterations in the animal (40). The group of Manuelidis
also reported transmission of CJD by buffy coat from a patient with an unclear
neurological disease (41), from patients with Alzheimers disease (42), and
from healthy individuals (37). Those results could not be repeated (43,44).
Rohwer attributed the experimental results to the occurrence of a late-onset
wasting disease associated with Clostridium difficile in the hamsters (45). In
addition, two studies with goats as donor and indicator animals could not
demonstrate any infectivity in blood (46,47). Studies in minks inoculated with the
TME agent did not reveal any infectivity in serum or any other component of
blood (48,49). Purified and concentrated lymphocytes from peripheral blood did
not transmit the disease while spleen and mesenteric lymph nodes were clearly
infectious. No infectivity could be found in serum or in blood clots of scrapie-
infected goats and sheep (5052). In addition, no infectivity has been detected
in buffy coat of cattle infected orally with the BSE agent up to 18 months
postinoculation (53,54).
A number of long-term, well-controlled experiments have failed to dem-
onstrate transmission of disease by blood from kuru and CJD patients into
monkeys and apes (7,12). In addition, Gajdusek transfused blood (>300 mL) from
three different CJD patients into three chimpanzees. None of these animals
developed TSE after an observation period of more than 20 years (7,12). It should
be noted that transmission of CJD by routes other than the intracerebral route
has rarely been successful, and the reproducibility of the phenomenon has often
been questioned.
VII. LACK OF CORRELATION BETWEEN TSEs AND

TRANSFUSION OF BLOOD AND BLOOD PRODUCTS
Several investigators have examined the theoretical possibility of CJD transmis-
sions by transfusion in humans. A study of the transfusion histories of 202 definite
342 Bianco
and probable cases of CJD performed in England and Wales between 198084
and 199092 showed that 21 of the patients had received blood transfusions and
29 had donated blood (55). The frequency of blood transfusions or donations did
not differ between CJD cases and matched controls, leading the investigators to
conclude that the evidence did not suggest that transfusion was a major risk factor
for development of CJD (55).
No cases of CJD among persons with hemophilia had been reported in the
medical literature until October 1994. The Medline database contained 1485
references on CJD and 6385 references on hemophilia between January 1976 and
October 1994. None of these references linked CJD and hemophilia. Unfortu-
nately, this type of search cannot be repeated because of the high number of
articles addressing the theoretical possibility of transmission of CJD by blood
transfusion. An extensive review of mortality data between 1979 and 1994
performed by CDC did not identify a single CJD death in individuals with a
clotting disorder or hemoglobinopathy (35).
A recent multicenter European case control study found no significant risk
of CJD associated with surgery and blood transfusions (56). In a follow-up of one
CJD patient who was a frequent blood donor, none of the blood recipients
developed CJD. Eighteen blood recipients have died of nonneurological disor-
ders; nine were alive at the time of the investigation. The time periods for
follow-up ranged from 1 to 22 years after the blood transfusions.
The distribution of infectivity in plasma derivatives in experimental TSE
models was studied by spiking normal plasma with trypsinized cells from a
scrapie-infected hamster. The plasma was fractionated using the classical Cohn
method and fractions injected intracerebrally into animals. The study showed a
potential but minimal risk of acquiring CJD from the administration of plasma
protein concentrates (57).
VIII. LOOKBACK STUDIES

Studies of recipients who received blood and blood products from donors who
later developed CJD have not yet led to a documented case of CJD. A collabora-
tive lookback study performed by the American Red Cross, the New York Blood
Center, and the CDC tracked 178 recipients of blood units derived from donors
who developed CJD. Nine of these recipients lived between 13 and 24 years after
transfusion, and 41 lived more than 5 years. No CJD was revealed in any of these
recipients (58). However, only 2 patients had been followed for more than 20
years after the transfusion.
CJD has not been identified among patients who received large amounts
of blood or blood products (35,59). CDC has examined tissues from 30 patients
with severe hemophilia who died with CNS symptoms since 1983 and found
no evidence of CJD. These patients had received clotting factor concentrates

for 1523 years. In addition, none of 101 hemophilia A patients, 76 of whom
lived 1117 years after receiving more than 100 units of cryoprecipitate, devel-
oped CJD (60). Another study carried out by the Veterans Administration in-
volved review of records of 8614 inpatient episodes of care and 543 death
certificates of veterans who received plasma products from a donor who later
developed CJD. It did not yield any cases of potential transmission after 7 years
of follow-up (61). One case of potential transmission to a liver transplant
recipient who also received transfusions of albumin has recently been reported.
One of the albumin donors died 3 years later from a dementia clinically charac-
terized as CJD (62). Obviously, the liver transplant recipient was exposed to a
variety of drugs and biologics, making it difficult to determine the exact source
of disease.
IX. MEASURE TO PREVENT THE THEORETICAL

TRANSMISSION OF THE TSE AGENT BY
TRANSFUSION OF BLOOD AND BLOOD PRODUCTS
A. Market Withdrawals and Recalls
On December 15, 1994, the issue of CJD and transfusion was reviewed by the
FDA Blood Products Advisory Committee. After extensive discussion, the Com-
mittee recommended that in-date cellular products of blood from donors who later
develop CJD should be withdrawn from distribution. In cases in which these
products had been transfused, the Committee recommended that physicians and
recipients be notified. In the case of plasma pooled for further manufacture, the
Committee recommended against recall of manufactured products because of
the lack of evidence for transmission. The hemophilia community, extremely
concerned about the theoretical potential for transmission of CJD, was quite
dissatisfied with this recommendation, leading FDA to convene a new advisory
committee to review the possibility of transmission of CJD by plasma derivatives.
The special advisory committee met on June 22, 1995, and recommended that all
plasma products containing donations from individuals who later died of CJD be
withdrawn from the market. This recommendation was based on precautionary
principles, despite the lack of evidence for transmissibility of CJD by these
products. FDA issued a memorandum to blood establishments on August 8, 1995,
recommending quarantine of these products. FDA also indicated that release of
these products might need to occur because of shortages and that the released
products should bear a warning disclosing risks and benefits. However, this
clause has not been invoked because of lack of acceptability of such products by
the patients.
344 Bianco
B. Deferral of Older Donors

The American Red Cross announced on December 5, 1996, that it would only use
plasma from donors under the age of 60 as a source of plasma for the manufacture
of plasma derivatives. This step was taken to lessen major market withdrawals
of plasma derivatives manufactured from volunteer donors because of rare reports
of older donors developing CJD. This policy was also adopted by some collec-
tors of source plasma who deferred donors over the age of 50. Interestingly, Busch
et al. have shown, based on data from the Retroviral Epidemiology Donor Study,
that a blanket removal of donors age 50 and over increases the risk of human
immunodeficiency virus (HIV) transmission by 12% hepatitis C virus (HCV) by
21%, and hepatitis B virus (HBV) by 22% (63).
C. Deferral of Donors with Family History or

Potential Exposure to Infectious Materials
Since 1987, individuals who had received human growth hormone of human
pituitary origin have been deferred from donating blood according to an FDA
recommendation. An FDA memorandum issued on August 8, 1995, advised that
individuals who had a family history of CJD or had received dura mater
transplants be deferred from donating blood or plasma. The deferral policies were
extended with the issuance of another memorandum on December 11, 1996,
entitled Revised Precautionary Measures to Reduce the Possible Risk of Trans-
mission of Creutzfeldt-Jakob Disease (CJD) by Blood and Blood Products.
It defined the categories of iatrogenic, familial, and possibly familial risks.
It added dura mater transplants to the deferrals and defined three questions to be
asked during the donor history. It also elaborated on actions to be taken when
donors were identified as at increased risk of developing CJD or were subse-
quently diagnosed with CJD. Blood donors with one blood relative with history
of CJD were to be deferred and the collected unit discarded. If there were two or
more blood relatives, the family was considered a family at risk for CJD, and
in-date products from prior donations were to be placed in quarantine. In order to
address concerns about large amounts of blood products placed in quarantine,
FDA indicated that donors could be subjected to genetic testing for familial forms
of PrP. If the tests were negative (i.e., no mutations associated with increased
susceptibility were identified), products could be released from quarantine. If the
tests were not performed or the donors tested positive, recipients of products from
these donors were to be notified and counseled. The same applied for donors who
subsequently died of CJD.
The American Association of Blood Banks issued Bulletin #96-4 containing
guidance for recipient notification. This bulletin was published in the AABB
News Briefs issue of June 1996. The guidance suggested that appropriate com-
mittees within an institution (Internal Review Board, Ethics Committee, etc.)

review the issues and define criteria for recipient notification.
On September 8, 1998, FDA changed its policy regarding classic CJD.
Deferral of donors with family history of CJD (two or more family members) was
retained. However, withdrawal of products was restricted to products made with
plasma from donors who later developed nvCJD, not classic CJD. The latter was
retained as a precautionary measure because of the limited follow-up of individ-
uals exposed to BSE. So far, there have been no cases of nvCJD identified in the
United States. The European Community has had a similar policy: given the lack
of specific information on nvCJD, as a precautionary measure it would be prudent
to withdraw batches of plasma-derived medicinal products from the market if a
donor to a plasma pool is subsequently strongly suspected, by the reference
center, of having nvCJD. In addition, FDA and the U.S. Department of Agricul-
ture have established rigid criteria for the importation of products that could be
used for manufacture of drugs or for human consumption. These include fetal
bovine serum, gelatin, and prepared foods.
D. Deferral of Donors Who Have Spent

6 Months or More in the U.K.
In June 1999, the TSE Advisory Committee of FDA recommended that persons
who had been in the United Kingdom for an aggregate amount of 6 months or
more between 1980 and 1996 should be deferred from donating blood. This
represents 2% of all blood donors in the United States and 3.1% of all donors in
New York (A. Williams et al. personal communication). The Committee reflected
concerns about nvCJD and about the fact that the United Kingdom had been
discarding the plasma donated by their own donors and was using plasma from
paid United States blood donors to use in their plasma fractionation plants. The
Committee was not swayed by arguments expressed by the transfusion medicine
community indicating that the policy would lead to major shortages. In addition,
the Committee rejected indications that replacement of these regular blood donors
with first-time donors with much higher prevalence of HIV, HCV, and HBV
would increase the risk for diseases known to be transmitted by transfusion.
The Committee preferred to adhere to the precautionary principles adopted by
their British counterparts. This deferral policy will be reviewed every 6 months.
A similar policy is currently under consideration in Canada.
E. Leukoreduction
The argument that leukoreduction, a process that removes the vast majority of
leukocytes from blood and blood components, may decrease CJD infectivity of
blood is derived from two different types of observation. First Klein et al. (64)
346 Bianco
demonstrated that mice lacking mature B lymphocytes do not develop clinical

symptoms of scrapie when inoculated with infectious material outside the brain
(i.e., intravenously and intraperitoneally). Second, PrP Sc has been detected in
tonsils and appendices of patients with nvCJD (21). In addition, PrPc is expressed
at different stages of leukocyte differentiation from CD34+ stem cells to mature
lymphocytes and monocytes but not granulocytes (65). Because infectivity could
be present in circulating leukocytes, leukoreduction would be a practical way to
reduce the risk of nvCJD. Several countries have implemented leukoreduction
or have committed to implementation of leukoreduction procedures based on
these theoretical arguments. It should be noted that the arguments are based
on observations made in experimental models that may bear no relevance to
nvCJD transmission.
As of July 1999, France, Portugal, Ireland, Luxembourg, and Canada had
implemented universal leukoreduction. The United Kingdom had decided that
leukoreduction for all blood for transfusion should be extended as soon as
practically possible, and the FDA Blood Products Advisory Committee recom-
mended implementation of universal leukoreduction, regardless of its role in the
prevention of CJD transmission. The Committee indicated that that universal,
prestorage leukoreduction would benefit recipients of blood components.
While leukoreduction has a number of well-documented benefits, it is
difficult to assess the role it would play in the prevention of transmission of
nvCJD. For instance, the degree of leukoreduction required to prevent transmis-
sion is unknown (particularly because transmission has not yet been observed).
In addition, the subclass of white blood cells that may carry the nvCJD agent has
not been determined. Prestorage leukoreduction has not yet been applied univer-
sally. Thus infrequent allergic reactions, rare hypotensive reactions (FDA letter
dated may 4, 1999, http:/www.fda.gov/cdrh/safety/hypoblrf.htm), and a red eye
syndrome observed among recipients of one filter brand raise some concerns (66).
The arguments that have been presented for implementing this practice in
the context of nvCJD include the suggestion that white blood cells may be
involved in the transport of the CJD and nvCJD agents. It is not yet clear what
proportion of the bloodborne infectivity is distributed into white blood cells and
which subtype classes of white blood cells (T, B, or dendritic cells) carry
infectivity for CJD or nvCJD. Furthermore, there is no evidence to date suggest-
ing that CJD/nvCJD is spread by blood transfusion. Consequently, implementa-
tion of universal leukoreduction to prevent CJD/nvCJD transmission is not based
on epidemiological or experimental data.
F. Screening Tests for CJD/nvCJD

There are no assays currently available for the screening of blood donations for
CJD or nvCJD. Thus, in the absence of cases of transmission of CJD and nvCJD
by blood, it is impossible to assess the influence of any of the current preventive

measures on the risk of transmission of CJD and nvCJD. However, various assays
based on monoclonal antibodies to PrP have been developed. Discrimination
between normal PrP and PrPsc or PRPres is based on resistance to digestion by
proteinase K (67,68). These assays are useful in the identification of cows with
BSE and in the diagnosis of sick patients (71,72).
G. Inactivation/Removal of the Agent

from Blood and Plasma
The infectivity of various cellular blood components and plasma derivatives has
been studied through spiking human blood with the scrapie agent and by
intracerebral, intravenous, and intraperitoneal inoculation of a mouse-adapted
strain of human CJD using hamsters and mice as assay animals. The data of such
studies suggest that when CJD is present (at low concentration) in blood of
infected animals, partitioning into various cellular and plasma compartments
occurs. In cellular blood components, CJD is present in leukocytes and platelets.
In plasma CJD is recovered from cryoprecipitate and fractions I, II, and III of
the Cohn fractionation method, but not in fractions IV and V and albumin (57,69).
These experiments have shown that fractionation reduces the infectious load
of plasma.
H. Reduction of the Size of Pools Used

for Manufacture of Derivatives
The pool size of plasma for fractionation varies from 1,000 to 10,000 units when
source plasma collected by aphersis is used; for recovered plasma the pool size
is also quite variable (30,00060,000 donor units from whole blood donations).
The question has been raised whether reduction of the pool size may decrease the
risk of CJD transmission. The probability that pooled donor plasma will contain
a donation from an individual with a disease has been analyzed for a range of
disorders and different pool sizes (70). Using this analysis, the probability that a
CJD patient has contributed to a pool of 10,000 donors is 0.8%; if the pool size
increases to 100,000 donors, the probability increases to 7.6%. However, it is
unlikely that a person with symptoms of CJD will donate blood. Using mathe-
matical modeling, Brown et al. (57) concluded that the chance of contracting CJD
from a pooled blood product to which a patient with CJD has contributed would
be extremely small, no matter what the size of the donor pool. Limitation of the
pool size is not likely to reduce such a risk, because the infectivity does not
saturate the pool.
Other diagnostic tools (e.g., detection of brain proteins in the cerebrospinal
fluid or in the peripheral blood) are under development (68,69). A specific
348 Bianco
immune response, often the basis for a diagnostic test, has not been observed.
Tests indicative of the disease before the onset of clinical symptoms do not
yet exist.
X. OVERALL ASSESSMENT
The assessment of the potential risk of TSE transmission by transfusion has been
a very difficult task. The reality of the AIDS tragedy hit the transfusion medicine
community after years of dismissive statements that minimized risks. In addition,
the hemophilia community, devastated by the transmission of HIV and HCV,
has exerted substantial political pressure, demanding safety and compensation.
Now both the scientific and the blood banking community are afraid of repeating
the same mistakes. The phenomenon could be called the fear of another AIDS
mistake. Medical experts do not want to risk a statement such as there is
sufficient evidence to indicate that the transmission of TSEs by transfusion of
blood and blood products is unlikely out of fear of being proven wrong.
In September 1998 FDA suspended a recall of plasma products that had been
manufactured from pools containing a unit donated by an individual who later
developed classical CJD. They retained recalls for nvCJD and the discard of
components from these donors. These derivative recalls had caused substantial
shortages of plasma derivatives all over the world. However, the fear of another
AIDS mistake prevented the regulators from accepting the overwhelming evi-
dence that individual components also do not transmit classical CJD. Another
interesting observation is the adoption of leukoreduction by several European
countries and Canada despite the lack of epidemiological or experimental evi-
dence that it will contribute to the prevention of CJD transmission. The action
was based on the experiments of Klein et al. (64) suggesting that B cells played
a role in dissemination scrapie in a mouse model quite distant from nvCJD.
However, evidence that blood and plasma have very low infectivity, derived from
the same model, has been summarily ignored.
Despite the evidence against the transmission of classical CJD by transfu-
sion, many of the restrictions, geographic deferrals, recommendations for leuko-
reduction, etc. continue to be applied because the period of follow-up for nvCJD
has been relatively short. The peak of BSE in England occurred in 1990, the first
case of nvCJD was identified in 1996, and as of May 31, 1999, there were 42
definite and probable cases of nvCJD.
The most solid recent assessment has been made by Paul Brown, from the
National Institutes of Health at a meeting in February 1999, Although experi-
mental studies indicate that blood donations from individuals with CJD might be
capable of transmitting disease, the available epidemiological evidence indicates
that bloodborne infection does not occur. He attributed the disparity to (a) very
low to absent levels of blood infectivity in patients with CJD; (b) dilution
of infectivity in large donor pools; (c) loss of infectivity (13 logs) during
plasma fractionation; and (d) the comparative inefficiency of transmission of the
infectious agent by parenteral routes. Items b and c are only applicable to
plasma derivatives.
ACKNOWLEDGMENTS
The author thanks Maria Rios for assistance in the preparation of this manuscript.
The author also wants to credit the report to the Scientific Committee on
Medicinal Products and Medical Devices of the European Community led by Dr.
J. Lwer, dated October 21, 1998, for its insights into the subject. The report is
posted at the following internet site: http://europa.eu.int/comm/dg24/health/sc/
scmp/outcome_en.html.
REFERENCES
1. Haywood A M. Transmissible spongiform encephalopathies. N Engl J Med 1997;
337:18211828.
2. Johnson RT, Gibbs CJ Jr. Creutzfeldt-Jakob disease and related transmissible spongi-
form encephalopathies. N Engl J Med 1998; 339:19942004.
3. Knight R. The relationship between new variant Creutzfeldt-Jakob disease and bovine
spongiform encephalopathy. Vox Sang 1999; 79:203208.
4. DeArmond SJ, Prusiner SB. Etiology and pathogenesis of prion diseases. AM J
Pathol 1995; 146:785.
5. Gajdusek DC. Nucleation of amyloidogenesis in infectious and non-infectious amy-
loidoses of the brain. Ann NY Acad Sci 1994; 724:173.
6. Manuelidis L. The dimensions of Creutzfeldt-Jakob disease. Transfusion 1994;
34:915.
7. Asher DM, Gibbs CJ Jr, Gajdusek DC. Pathogenesis of subacute spongiform enceph-
alopathies. Ann Clin Lab Sci 1976; 6:84103.
8. Manuelidis EE, Kim JH, Mericangas JR, Manuelidis L. Transmission to animals of
Creutzfeldt-Jakob disease from human blood. Lancet 1985; ii:896.
9. Holman RC, Khan AS Belay ED, Schonberger LB. Creutzfeldt-Jakob disease in the
United States, 1979-1994: using national mortality data to assess the possible
occurrence of variant cases. Emerg Infect Dis 1996; 2:333337.
10. Will RG, Alperovitch A, Poser S, Pocchiari M, Hofman A, Mitrova E, de silva R,
D Alessandro M, Delasnerie-Laupretre N, Zerr I, van Duijn C. Descriptive epidemi-
ology of Creutzfeldt-Jakob disease in six European countries, 1993-1995. Ann Neurol
1998; 43:763767.
11. Parchi P, Castellani R, Capellari S, Ghetti B, Young K, Chen SG, et al. Molecular
basis of phenotypic variability in sporadic Creutzfeldt-Jakob disease. Ann Neurol
1998; 39:767778.
350 Bianco
12. Brown P, Gibbs CJ Jr, Rodgers-Johnson P, Asher DM, Sulima MP, Bacote A,
Goldfarb LG, Gajdusek DC. Human spongiform encephalopathy: the National Insti-
tutes of Health series of 300 cases of experimetally transmitted disease. Ann Neurol
1994; 35:513529.
13. Duffy P, Wolf J, Collins G, DeVoe A G, Steeten B, Cowen D. Possible person-to-
person transmission of Creutzfeldt-Jakob disease. N Engl J Med 1974; 290:692693.
14. Heckmann JG, Lang CJG, Petruch F, Druschky A, Erb C, Brown P, Neundrfer B.
Transmission of Creutzfeldt-Jakob disease via corneal transplant. J Neurol Neurosurg
Psychiatry 1997; 63:388390.
15. Bernoulli C, Siegfried J, Baumgartner G, Regli F, Rabinowicz T, Gajdusek DC, Gibbs
CJ Jr. Danger of accidental person-to-person transmission of Creutzfeldt-Jakob
disease by surgery. Lancet 1977; i:478479.
16. Fradkin JE, Schonberger LB, Mills JL, et al. Creutzfeldt-Jakob disease in pituitary
growth hormone recipients in the United States. JAMA 1991; 265:880.
17. Martinez-Lage JF, Poza M, Sola J, Tortosa JG, Brown P, Cervenakova L, Esteban JA,
Mendonza A. Accidental transmission of Cretzfeldt-Jakob disease by dural cadaveric
grafts. J Neurol Neurosurg Psychiatry 1994; 57:10911094.
18. Centers for Disease Control. Cretzfeldt-Jakob disease associated with cadaveric dura
mater graftsJapan, January 1979-May. MMWR 1996; 46:10661096.
19. Will RG, Ironside JW, Zeidler M, Cousens SN, Estibeiro K, Alperovitch A, Poser S,
Pocchiari M, Hofman A, Smith PG, A new variant of Cretzfeldt-Jakob disease in the
UK. Lancet 1996; 347:921925.
20. Will RG, Knight RSG, Zeidler M, Ironside JW, Cousens SN, et al. Reporting of
suspect new variant Creutzfeldt-Jakob disease. Lancet 1997; 349:847.
21. Hill AF, Butterworth RJ, Joiner D, Rossor MN, Thomas DJ, Frosh A, Tolley N, Bell
JE, Spencer M, King A, AI-Sarraj S, Ironside JW, Iantos PL, Collinge J. Investigation
of variant Cretzfeldt-Jakob disease and other human prion diseases with tonsil biopsy
samples. Lancet 1999; 353:183189.
22. Collinge J, Sidle KCL, Meads J, Ironside J, Hill AF. Molecular analysis of prion strain
variation and the etiology of new variant CJD. Nature 1999; 383:685690.
23. Bruce ME, Will RG, Ironside JW, McConnell I, Drummond D, Suttie A, McCardle
L, Chree A, Hope J, Birkett C, Cousens S, Fraser H, Bostock CJ. Transmissions to
mice indicate that new variant CJD is caused by the BSE agent. Nature 1997;
389:498501.
24. Gerstmann J, Strussler E, Scheinker I. ber eine eigenartige hereditr-familire
Erkrankung des Zentralnervensystems. Zugleich ein Beitrag zur Frage des vor-
zeitigen lokalen Alterns. Z Ges Neurol Psych 1936; 154:736762.
25. Lugaresi E, Medori R, Montagna P, Baruzzi A, Cortelli P, Lugaresi A, et al. Fatal
familial insomnia and dysautonomia with selective degeneration of thalamic nuclei.
N Engl J Med 1986; 315:9971003.
26. Chandler R L. Encephalopathy in mice produced by inoculation with scrapie brain
material. Lancet 1961; i:107108.
27. Gibbs CJ Jr, Gajdusek DC, Morris JA. Viral characteristics of the scrapie agent in
mice. In: Gajdusek DC, Gibbs, CJ, Alper M, eds Slow, Latent and Temperate Virus
Infections. National Institute of Neurologic Diseases and Blindness, 1965:195202.
28. Gibbs CJ Jr, Gajdusek DC. Isolation and characterization of the subacute spongiform
virus encephalopathies of man:kuru and Creutzfeldt-Jakob disease. J Clin Pathol

1972; 25(suppl. 6):8496.
29. Manuelidis EE, Angelo JN, Gorgacz EJ, Manuelidis L. Transmission of Creutzfeldt-
Jakob disease to Syrian hamster. Lancet 1977; i:479479.
30. Tateishi J, Ohta M, Koga M, Sato Y, Kuroiwa Y. Transmission of chronic spogiform
encephalopathy with kuru plaques and leukomalacia to small rodents. Ann Neurol
1979; 5:581584.
31. Will RG, Ironside JW, Zeidler M, Cousens SN, Estibeiro K, Alperovitch A, et al.
A new variant of Creutzfeldt-Jakob disease in the UK. Lancet 1996; 347:921925.
32. Collinge J, Sidie KCL, Meads J, Ironside J, Hill AF. Molecular analysis of prion strain
variation and the etiology of new variant CJD. Nature 1996; 383:685690.
33. Chazot, G, Broussoille, E, Lapras C, Blattler C, Aguzzi A, Koop N. New variant of
Creutzfeldt-Jakob disease in a 26-year old French man (letter). Lancet 1996; 347:1181.
34. Reingold A, Rothrock G, Starr M, et al. Surveillance for Creutzfeldt-Jakob disease
United States. MMWR 1996; 45:665668.
35. Holman RC, Kahn AS, Belay ED, Schonberger LB. Creutzfeldt-Jakob disease in the
United States, 1979-1994: using national mortality data to assess the possible
occurence of variant cases. Emerg Infect 1996; 2:333337.
36. Manuelidis EE. Transmission of Creutzfeldt-Jakob disease from man to guinea pig.
Science 1975; 190:571572.
37. Manuelidis L, Manuelidis EE. A transmissible Creutzfeldt-Jakob disease-like agent
is prevalent in the human population. Proc Natl Acad Sci 1993; 90:77247728.
38. Tateishi J. Transmission of Creutzfeldt-Jakob disease from human blood and urine
into mice. Lancet 1985; ii:10741074.
39. Tamai Y, Kojima H, Kirajima R, Taguchi F, Ohtani Y, Kawaguchi T, Miura S, Sato
M, Ishihara Y. Demonstration of the transmissible agent in tissue from a pregnant
woman with Creutzfeldt-Jakob disease. N Engl J Med 1992; 327:649649.
40. Deslys JP, Lasmzas C, Dormont D. Selection of specific strains in iatrogenic
Creutzfeldt-Jakob disease (letter). Lancet 1994; 343:848849.
41. Manuelidis EE, Manuelidis L, Pincus JH, Collins WF. Transmission, from man to
hamster, of Creutzfeldt-Jakob disease with clinical recovery. Lancet 1978; ii:4042.
42. Manuelidis EE, de Figueiredo JM, Kim JH, Fritch WW, Manuelidis L. Transmission
studies from blood of Alzheimer disease patient and healthy relatives. Proc Natl Acad
Sci USA 1988; 85:48984901.
43. Godec MS, Asher DM, Master CL, Rubi JU, Payne JA, Rubi-Villa DJ, Wagner EE,
Rapoport SI, Schapiro MB. Evidence against the transmissibility of Alzheimers
disease. Neurology 1992; 41:13201320.
44. Godec MS, Asher DM, Kozachuk WE, Masters CL, Rubi JU, Payne JA, Rubi-Villa
DJ, Wagner EE, Rapoport SI, Schapiro MB. Blood buffy coat from Alzheimers
disease patients and their relatives does not transmit spongiform encephalopathy to
hamsters. Neurology 1994; 44:11111115.
45. Rohwer RG. Alzheimers disease transmission: possible artifact due to intercurrent
illness. Neurology 1992; 42:287288.
46. Pattison IH, Millson GC. Distribution of scrapie agent in the tissues of experimentally
inoculated goats. J Comp Pathol Ther 1962; 72:233244.
352 Bianco
47. Pattison IH, Millson GC, Smith K. An examination of the action of whole blood,
blood cells or serum on the goat scrapie agent. Res Vet Sci 1964; 5:116121.
48. Marsh RF, Burger D, Hanson RP. Transmissible mink encephalopathy: behavior of
the disease agent in mink. Am J Vet Res 1969; 30:16371642.
49. Marsh RF, Miller JM, Hanson RP. Transmissible mink encephalopathy: studies on
the peripheral lymphocyte. Infect Immun 1973; 7:352355.
50. Hadlow WJ, Eklund CM, Kennedy RC, Jackson TA, Whitford HW, Boyle CC.
Course of experimental scrapie virus infection in the goat. J Infect Dis 1974;
129:559567.
51. Hadlow WJ, Kennedy RC, Race RE, Eklund CM. Virologic and neurohistologic
findings in dairy goats affected with natural scrapie. Vet Pathol 1980; 17:187199.
52. Hadlow WJ, Race RE, Kennedy RC. Temporal distribution of transmissible mink
encephalopathy virus in mink inoculated subcutaneously. J Virol 1987; 61:3235
3240.
53. Wells GAH, Dawson M, Hawkins SAC, Austin AR, Green RB, Dexter I, Horigan
MW, Simmons MM. Preliminary observations on the pathogenesis of experimental
bovine spongiform encephalopathy. In: Gibbs CJ Jr, ed. Bovine Spongiform Enceph-
alopathy: The BSE Dilemma. New York: Springer-Verlag, 1996:2844.
54. Wells GAH, Hawkins SAC, Green RB, Austin AR, Dexter I, Spencer YI, Chaplin
MJ, Stack MJ, Dawson M. Preliminary observations on the pathogenesis of ex-
perimental bovine spongiform encephalopathy (BSE): an update. Vet Rec 1998;
142:103106.
55. Esmonde TFG, Will RG, Slattery JM, Knight R, Harries-Jones R, de Silva R,
341:205207.
56. van Duijn CM, Delasnerie-Lauprtre N, Masullo C, Zerr I, de Silva R, Wientjens
DPWM, Brandel J-P, Weber T, Bonavita V, Zeidler M, Alprovitch A, Poser S,
Granieri E, Heye N, Hensen S, Mller N. Creutzfeldt-Jakob disease and blood
transfusion (letter). Lancet 1994; 343:298299.
57. Brown P, Rohwer RG, Dunstan BC, Macauley C, Gajdusek DC, Drohan WN.
The distribution of infectivity in blood components and plasma derivatives in ex-
perimental models of transmissible spongiform encephalopathy. Transfusion 1998;
38:810816.
58. Sullivan MT, Schonberger LB, Kessler D, et al. Creutzfeldt-Jakob disease (CJD)
investigational lookback study. Transfusion 1997; 37(suppl):2S.
59. Operaskalski EA, Mosley JW. Pooled plasma derivatives and Creutzfeldt-Jakob
disease (letter). Lancet 1995; 346:12241224.
60. Evatt BL. Prions and hemophilia: assessment of risk. Haemophilia 1998; 4:628633.
61. Rahman A, Bullman TA, Kang HK, Muniz A, Kizer KW, Roselle GA, Booss J.
A search for Creutzfeldt-Jakob disease in veterans treated with potentially contami-
nated blood products. Infect Control Hosp Epidemiol. In press.
62. Crange A, Gray F, Cesaro P, Adle-Biassette H, Duvoux C, Cherqui D, Bell J, Parchi
P, Gambetti P, Degos J-D. Creutzfeldt-Jakob disease after liver transplantation. Ann
Neurol 1995; 38:269272.
63. Busch MP, Glynn SA, Schreiber GB. Potential increased risk of virus transmission
due to exclusion of older donors because of concern over Creutzfeldt-Jakob disease.

Transfusion 1997; 37:9961002.
64. Klein MA, Frigg R, Flechsig E, Raeber AJ, Kalinke U, Bluethmann H, Bootz F, Suter
M, Zinkernagel RM, Aguzzi A. A crucial role for B cells in neuroinvasive scrapie.
Nature 1997; 890:687690.
65. Dodelet VC, Cashman NR. Prion protein expression in human leukocyte differentia-
tion. Blood 1998; 91:15561561.
66. Centers for Disease Control. Adverse ocular reactions following transfusions
United States, 1997-1998, MMWR 1998; 47:4951.
67. Korth C, Stierli B, Streit P, Moser M, Schaller O, Fischer R, Schulz-Schaeffer W,
Kretzschmar H, Raeber A, Braun U, Ehrensperger F, Hornemann S, Glockshuber R,
Riek Safar J, Wille H, ltri V, Groth D, Serban H, Torchia M, Cohen FE, Prusiner SB.
Eight prion strains have PrPSc molecules with different conformations. Nat Med
1998; 4:11571165.
68. Otto M, Wiltfang J, Schuetz E, Zerr I, Otto A, Pfahlberg A, Gefeller O, Uhr M, Giese
A, Weber T, Kretzschmar HA, Poser S. Diagnosis of Creutzfeldt-Jakob disease by
measurement of S100 protein in serum: prospective case-control study. BMJ 1998;
316:577582.
69. Dormont D. Evaluation du risque de transmission des agents transmissibles non
conventionnels par lalbumine plasmatique humaine. (Evaluation of transmission of
unconventional agents by human albumin.) Ann Franc Anesth Reanim 1996; 15:560
568.
70. Lynch TJ, Weinstein MJ, Tankersley DL, Fratantoni JC, Finlayson JS. Considerations
of pool size in the manufacture of plasma derivatives. Transfusion 1996; 35:513529.
71. Hsich G, Kenney K, Gibbs CJ, Lee KH, Harrington MG. The 14-3-3 brain protein in
cerebrospinal fluid as a marker for transmissible spongiform encephalopathies.
N Engl J Med 1996; 335:924930.
72. Billeter M, Wuthrich K, Oesch B. Prion (PrPSc)-specific epitope defined by a
monoclonal antibody. Nature 1997; 390:7477.
16
The Protozoan Parasites
Malaria and Chagas Disease
Silvano Wendel
Hospital Sirio Libanes, So Paulo, Brazil
I. INTRODUCTION
Protozoan infections affect millions of people in the world, mainly in tropical,

developing countries. Malaria and American trypanosomiasis (Chagas disease)
have been linked to blood transfusion transmission for decades. Both parasites
have a complex life cycle involving a definitive host (insect or mammal), and
humans are usually intermediary hosts, quite often as a result of invasion of wild
habitats, where these diseases are in equilibrium as zoonoses. Transfusion-transmitted
parasites are usually restricted to tropical areas, and immunocompromised recip-
ients are the most affected; a few cases of Chagas disease are reported in indus-
trialized, developed countries, whereas transfusion-transmitted malaria (TTM)
has been described in hundreds of recipients worldwide. Nevertheless, the on-
going migration processes that have taken place during recent decades have
opened new frontiers for these somewhat geographically restricted diseases.
II. MALARIA
A. Historical Aspects and Epidemiology
The first case of TTM was reported in 1911 by Wolsey (1), and more than 3000
cases have been described so far (2), although these numbers may represent less
than 50% of the actual cases (3). Although malaria has been eradicated in almost
all European countries, the United States, Australia, and Japan, it is still present
in 102 countries (46), where more than 120 million people are infected annually,
355
356 Wendel
with 1 million deaths and nearly 300 million carrying the parasite. Countries in
tropical Africa are responsible for 80% of all clinical cases and more than 90%
of all parasite carriers in the world. Excluding the African continent, 90% of cases
reported to WHO are from 19 different countries, with some 75% of them
concentrated in 9 countries: India, Brazil, Afghanistan, Sri Lanka, Thailand,
Indonesia, Vietnam, Cambodia, and China (decreasing order of frequency).
Malaria is concentrated in certain regions within a single country. For example,
though 85% of the Brazilian geographic territory is exposed to malaria, less than
15% of the Brazilian population lives in the affected area, with nearly 99% of all
cases detected in the Amazon region. In addition, nearly 40% of the world
population, or 1.8 billion people (1992), still remain exposed to varying degrees
of risk of malarial infection (4,5).
Transmission by blood components is either a consequence of the sanitary
conditions of a country or region, when complete eradication has not yet been
achieved (e.g., tropical Africa, India, Sri Lanka, Brazilian Amazon basin, etc.), or
a result of imported cases into developed, nonendemic places by immigrants or
travelers from endemic areasmainly described in France, the United Kingdom,
the United States, and Spain (311).
The annual incidence of TTM ranges from 0.25 cases/million units in the
United States (12) to more than 50 cases/million units in endemic regions (79).
Even in nonendemic countries where complete eradication has been accom-
plished, the ever-increasing migration flow played a great role in the last three
decades, as observed in France in the 1980s when more than 100 cases were
described (10). Table 1 shows the prevalence of infected donors from some
endemic and nonendemic countries.
B. The Agent and Its Life Cycle

Human malaria is transmitted by four different species: Plasmodium falciparum,
Plasmodium vivax, Plasmodium ovale, and Plasmodium malariae, all of which
require two different hosts to complete their life cycle: a mosquito (Anopheles),
where the sexual stage ensues, and humans, where the asexual stage is observed.
When infected female mosquitoes bite humans, they release, through their
salivary glands, small asexual sporozoitesmotile, spindle-shaped, 1015 m
long organismsinto the bloodstream. These circulate for a very short time
(<60 min), until hepatocytes invasion (exoerythrocytic stage), followed by a
schizogony step (a repeated nucleus division), which lasts from 5 to 31 days,
ending with the rupture of mature schizonts and release of merozoites (from 2,000
for P. malariae to 40,000 for P. falciparum) into the bloodstream. The merozoites
in turn invade circulating red blood cells (erythrocytic stage). Another schizogony
step occurs in red cells, with generation of schizonts (ring forms) and subsequent
release of additional merozoites (424 depending on Plasmodium species), infect-
Protozoan Parasites 357
Table 1 Prevalence of Malaria-Infected Donors in Some Countries
%
Infected
Author (Ref.) Country donors Species Method
Chikwem et al. (42) Nigeria 4.0 P.f. Thick and thin blood films
Ibhanesebor et al. (24) Nigeria 40 P.f. Thick and thin blood films
Ferreira et al. (19) Brazila 32 P.f. IgG antibodies (IFAT)
24 P.v.
37 P.f.+P.v.
Kiesslich et al. (20) Brazila 15.7 P.f.+P.v. IFAT
0.8 Thick films
1.6 Acridine orange (QBC)
Hong et al. (41) Vietnam 0.2 P.f. PCR (repetitive genomic DNA)
Chiodini et al. (27) United 1.5b P.f. ELISA (antibody)
Kingdom 0.45c
Tabor (11) France 10b IFAT
P.f. = Plasmodium falciparum; P.v. = Plasmodium vivax; IFAT = indirect immunofluorescence assay;
PCR = polymerase chain reaction.
aDonors only from the Brazilian Amazon, a highly endemic area. Other parts of the country are not
considered.
bDonors from endemic, tropical countries.
cDonors from nonendemic countries, never exposed to malaria. This group most likely represent
false-positive results, as none were confirmed by IFAT.
ing other red cells and continuing this asexual stage. This schizogonic process is
regular, lasting 48 hours for P. malariae, P. vivax, and P. ovale (Tertian fever) and
72 hours for P. falciparum (Quartan fever).
The merozoites released from liver cells cannot subsequently reinvade other
hepatocytes and perpetuate the process. However, some P. vivax and P. ovale
sporozoites give rise to hypnozoites (a very latent exoerythrocytic form), which
stay dormant inside the hepatic parenchyma for approximately up to 5 years
before starting the exoerythrocytic schizogonic stage, inducing late relapses if
proper treatment is not applied.
In parallel, after several divisions, some erythrocytic parasites undergo
another cycle step and differentiate into sexual gametocytes. They are released in
the bloodstream and, instead of invading red cells, are ingested by another
mosquito during a blood meal, beginning a sexual reproduction (fertilization) in
the insect stomach. This step leads to the development of oocysts under the basal
membrane, which undergo a sporogony (a reduction division), releasing thou-
sands of free sporozoites in the insects hemocele that migrate towards the
mosquito salivary gland, ending their life cycle (6,11,13,14).
358 Wendel
In addition to classical, vectorial transmission, malaria can also be trans-

mitted by blood transfusion, organ transplants, sharing of intravenous drugs or
needles, accidental laboratory exposure, or congenitally. The life cycle of Plas-
modium and the different patterns of transmission are shown in Figures 1 and 2,
respectively.
C. Transmission by Blood Components

Infected blood donors may harbor parasites for many years, especially those
partially immune from endemic regions where such immunity leads to an absence
of symptoms, or after chemoprophylaxis. Different periods of infection are seen
for each species; it is widely known that P. falciparum can be cleared within one
year, although longer periods (813 years) have been described. P. vivax and
P. ovale usually do not persist for more than 34 years, but infections after 68
years or even more have also been described. However, longer periods of
longevity ranging from 10 to 50 years were seen for P. malariae, as reported in
the former Soviet Union and the United States (79). Because of the shorter
infectivity periods for P. falciparum, P. ovale, and P. vivax, it seems unlikely that
a donor will remain infective after 5 years of infection.
TTM is transmitted by asexual forms present in the red blood cells, usually
from asymptomatic donors who carry a low level of parasitemia and, rarely, in
the period between subclinical parasitemia and the onset of symptoms (usually
associated with previous chemoprophylaxis, a suppressive rather than therapeutic
regimen). Gametocytes are not infective, and free merozoites are present in the
bloodstream for a very short period and do not maintain viability in the stored
units. The chances of infection by circulating sporozoites right after the insect
biting, though theoretically possible within a 60-minute period (15), has not been
associated with transfusion cases. Transmission from symptomatic donors is very
unlikely and, when present, usually occurs in endemic regions.
The minimum number of parasites that leads to infection is not known; for
P. vivax, a total of 10 parasites/mL was successful in inducing experimental
transmission (7). In addition, an infection rate of 12 parasites/mL will result in
nearly half a million parasites in a single whole blood unit, a number far above
the minimum necessary to induce an active infection in the recipient (3).
Although transmission occurs mainly by the use of red cell components,
other blood components (untreated liquid or frozen plasma, platelets, and granu-
locyte concentrates) cannot be regarded as cell-free and, therefore, devoid of
viable infective parasites (16). Infectivity remains even in deglycerolized post-
thaw units stored at 70C. Red cell preservatives containing adenine seem to
enhance the plasmodia viability. No infection has been associated with freeze-
dried plasma or industrialized derivatives (albumin, immunoglobulin, etc.), even
if derived from potentially infected donors; these donations can be regarded as
Figure 1 The life cycle of malaria. Although recently infected donors bearing sporozo-
ites in the bloodstream are able to induce infection in experimentally infected volunteers,
this is a very unusual situation. TTM has been observed basically from donors in the
erythrocytic stage through the transfusion of merozoite-infected red cells.
safe if used exclusively for fractionation. Though viability at 46C storage is less
than 5 days for P. malariae and up to 10 days for P. falciparum, holding units for
710 days in order to achieve a relative protection is not justified.
A survey covering the 19731980 period showed distribution of 38% for
P. malariae, 42% for P. vivax, 42% for P. ovale, and 20% for P. falciparum (7).
Another survey conducted in the United States (12) covering the 19721988
period, when 45 TTM cases were reported, showed that a distribution of 38% for
P. malariae, 29% for P. falciparum, 24% for P. vivax, and 9% for P. ovale.
360 Wendel
Figure 2 Malaria transmission: The natural, vectorial transmission through Anopheles

mosquitoes is shown in the left. Malaria can be transmitted through blood transfusion,
organ transplants, sharing of intravenous drugs or needles, accidental laboratory exposure,
or congenitally.
Conversely, the distribution among different endemic regions also shows a wide
variation; in Africa, P. falciparum is the most common agent (4,5), while P. vivax
is the most common in India (17) and Brazil (1820).
D. Clinical Symptoms
The incubation period depends on several factors: the number of viable transfused
parasites, the species and the strain, the host immunity, and the previous use of
malarial chemoprophylaxis. On average, TTM has a longer incubation period than
that observed in natural transmission, with a 16-day mean period (829 days) for
P. falciparum, a 19.6day mean period (830 days) for P. vivax and P. ovale, and
a 57.2-day mean period (6106 days) for P. malariae (2,7,8).
The first symptoms are usually nonspecific, mainly fever that only reaches
its peculiar periodicity in 2 weeks. Since most physicians in nonendemic areas
are unaware of this possibility, a period is observed between the onset of the
symptoms and diagnosis ranging from 12 to 43 days, but periods of 160 days up
to one year have been reported. Even in endemic countries, a long delay may be
observed (18). Symptoms are more severe in splenectomized, organ-transplanted,
or immunodeficient recipients. In developed countries, TTM is usually associated
with previously splenectomized recipients, cardiopulmonary bypass surgery pa-
tients, and organ transplant recipients (2123), whereas in endemic countries,
particularly in Africa, neonates who undergo exchange transfusions are one of the
most severely affected patient groups (24).
The rate of lethal cases is highly related to late diagnosis (especially in
nonimmune recipients) and plays an important role when infection is due to
P. falciparum (with cerebral, renal, and pulmonary lesions). This was particularly
important in the United States after the Vietnam war when a 24-fold higher
mortality was seen in patients diagnosed in civilian hospitals compared to those
in military hospitals (11). Fatalities are still high, with some reports of up to 20%
of cases, much higher than the imported cases (<1%), where the index of
suspicion is higher. Thus, one must always bear in mind this possibility whenever
fever ensues after a transfusion episode. Since no exoerythrocytic stage is
observed in TTM, no relapse is found (irrespective of the causative species) after
appropriate treatment.
E. Preventive Measures
A dramatic decrease in TTM should be expected after malaria eradication in the
United States. American troops returning from Korea, Vietnam, and, lately,
Somalia brought back a remarkable number of infected personnel responsible for
TTM, despite preventive measures having been taken. On the other hand,
migratory movements, particularly from Africa and India into Europe and from
Southeast Asia and South America into North America (3), were also responsible
for TTM. Therefore, specific measures have been taken in order to prevent it.
The first strategy is taking an accurate history including place of birth,
previous residence locations, immigration status, and a complete account of
traveling during the previous 3 years. The use of specific questionnaires aims at
excluding three types of infected donors:
1. Nonimmune travelers who acquired malaria abroad, preventing re-
lapses after specific treatment (P. vivax and P. ovale) or asymptomatic
parasitemia following inadequate chemoprophylaxis
2. Military personnel returning from endemic areas
3. Foreign visitors or immigrants from endemic regions who are usually
immune and asymptomatic
Although rules in different countries vary (16,25), blood donors are gener-
ally not accepted until 6 months after chemoprophylaxis has been finished in the
absence of symptoms. As a general policy, those with alleged past infection or
362 Wendel
who had taken chemoprophylaxis in the presence of symptoms are not accepted
during a 3-year period. This measure is highly effective for P. falciparum,
P. vivax, and P. ovale but not for P. malariae, for which longer periods of
infectivity are recorded. The questions apply only to products where viable red
cells are present, but may be disregarded for donations intended exclusively as
sources for plasma for fractionation. Since the world malarial zone is so vast, an
alphabetical list of countries reporting malaria transmission and a geographical
malarial map (provided by WHO) (4,5) should be present in every collection
facility. Unfortunately, avoiding infected donors only by using questionnaires
brings two additional problems (26):
1. Low specificityA study in the United States (12), where the incidence
of TTM is 0.25 cases/million units, claimed that reducing the interval
to 6 months after travel, irrespective of prophylaxis, would allow some
70,000315,000 (mean of 44,000) additional blood units to be collected
each year, with an additional calculated risk of 0.03 cases/million units
in the annual incidence. The authors stated that the 6-month deferral
was accurate for 88% of cases and the 3-year deferral period used for
preventing P. vivax and P. ovale (which accounted for only 33% of
cases) would have prevented only one case in 17 years.
2. Low sensitivityAfter a careful examination of all cases of TTM in the
United States, it was stated that at least 50% of them could be avoided
if the aforementioned questionnaire was correctly filled out (12);
additionally, the interviewer must rely upon the accuracy of the donors
history, which may be in error. This same study reported that a positive
history could be ascertained in only 2030% of the implicated donors,
since many of them were infected in childhood or do not remember
the episode.
In the light of these problems, some nonendemic countries have introduced
specific serological screening (see below), which is applicable only to selected
donors at increased risk, such as immigrants from or individuals born in endemic
areas or those who had malaria at least 3 years before donation (10,16,27). With
this strategy, about 10% of donors in France (10) who were tested by an indirect
immunofluorescence assay (IFAT) were confirmed as infective; in the United
Kingdom a recent work detecting antibodies by an antiglobulin ELISA (27) with
P. falciparum antigens reported a 98.5% saving of nonplasma components from
donors from tropical areas, which would lead to approximately 40,000 units
recovered annually. Other, more radical positions, such as a permanent exclusion
of all persons born or having lived in endemic areas, have been proposed, but they
should be carefully analyzed, especially in locations with blood shortage (28).
In countries where malaria is endemic, the exclusion for 3 years of donors
previously infected may be inappropriate. The use of some screening tests, such
as detection of malarial antibodies, may also be inadequate, since most cases

denote only a previous infection and not the presence of circulating parasites.
In such cases testing for malarial antigens with monoclonal antibodies may be
helpful (29).
The use of inactivating agents might be a suitable alternative in the near
future. Currently, three agents are under investigation:
1. Merocyanine 540 (MC 540)This photosensitizing dye has been used
in preclinical trials for purging leukemia and lymphoma cells (30). In
the presence of light irradiation, it has been demonstrated to protect
experimental animals challenged with P. yoeliiinfected cells with no
deleterious effect on red cells. Its inactivating action is probably
dependent on binding to the parasitophorous vacuole membrane of
intracellular parasites. However, its absorption spectrum overlaps that
of hemoglobin, requiring that the treatment be performed only at low
hematocrits (31), which would carry some inconvenience for a routine
red cell malaria inactivation in the blood bank setting.
2. Pc4This silicon phthalocyanin (32), a psoralenic photosensitizing
dye, is used for the photodynamic treatment of tumor cells and steril-
ization of other infectious agents present in blood components (33).
A considerable inactivation of P. falciparum in infected red blood cells
was observed (3 log10), in either low (35%) or high (60%) hematocrit
levels, after a 40-minute light exposure. In addition, a strong activity
was obtained even in the absence of light, rendering this agent a
suitable one for malaria inactivation. Although its mechanism is still
unknown, it seems very likely that it plays a role through oxidative
stress and generation of singlet oxygen, as with crystal violet in the
photoinactivation of Trypanosoma cruzi (see below). Unfortunately, no
further results are available concerning the effect when large volumes
(e.g., a full RBC unit) are concerned, including the effect of plasma
present in the treated components or their posttransfusional survival.
3. Crystal violet (gentian violet)This agent has been used for Trypano-
soma cruzi inactivation for over 40 years and has also shown some
effect in inactivating P. bergheiinfected red blood cells (34). However,
this issue still deserves further investigation.
1. Screening Based on Detection of Antibodies

Detection of antibodies is achieved mainly by indirect immunofluorescence assay
(IFAT), indirect hemagglutination assay (IHA), or ELISA using P. falciparum,
nonhuman parasites (showing some cross-reactivity with human antibodies), or
purified or recombinant antigens. These tests are useful to interpret high-titered
or negative sera denoting, respectively, the presence or absence of infection.
364 Wendel
However, low-titered sera must, in most cases, be supplemented by a thorough

interview with the donor (3). Though IgG antibodies usually correlate with the
number of past malarial attacks (19), the presence of antibodies does not neces-
sarily mean the presence of circulating parasites, especially among donors from
endemic regions. Antibodies detected by IFAT are present in >95% of infected
donors, usually arising 714 days after the onset of infection by P. falciparum
(with longer periods for other species), persisting for several years in those who
are subject to continuous exposure (immune donors from endemic countries),
especially when infection occurs by P. malariae, where a long latency period and
high titers are found. After proper treatment, these antibodies tend to persist for
612 months. Thus, this method is highly recommended for screening donors in
nonendemic countries; a substantial array of experience has been gained in France
for over 15 years (10). On the other hand, it seems to be unsuitable in highly
endemic regions, where a high number of antibody-positive donors are to be
found; in such cases, the adoption of antigen detection is recommended (see
below). Although in the window phase the sensitivity of IFAT or ELISA ranges
from 50 to 80% (35), it is unlikely that donors in the acute infection period will
feel well enough to donate blood, especially in nonendemic countries. A proposed
action of combined IFAT and ELISA tests for malarial antibodies screening (27)
in blood donors is shown in Table 2. Because of the wide window phase period,
detection of antibodies is not recommended as diagnosis of early, acute cases.
2. Screening Based on Detection of Antigens

The most used method of antigen detection is the thick smear test. However,
it lacks sensitivity, with a threshold of 10100 parasites/L (the lower limit
achieved only in the hands of highly qualified technicians), which is by far more
than the minimum necessary for the transmission. It is also highly subjective,
leads to lapses of concentration by the technical staff because it is a very boring
procedure, and cannot be used as a mass screening method, particularly where
blood services are not fully established or developed, as in regions where TTM
is an important problem. One study in the United States showed prevention by
thick smear of only 10% of potential TTM and positivity in 31% (8/26) of donors
implicated in TTM (12).
The detection of parasites within erythrocytes has been achieved by IFAT
in endemic countries (13,29). Sensitivity reaches 530 parasites/L, but lower
levels of parasitemia may reach up to 2.5 106 parasites in one unit of blood (7).
PfHRP-2, a water soluble, histidine-rich antigen present in immature
P. falciparum gametocytes (36), and the production of monoclonal antibodies
(mAb) against it have been the basis of two distinct methods for circulating
antigen detection. The first is an ELISA test, which achieved a 98% sensitivity
in field studies in Thailand (38). The second is a dipstick test, based on mAb
Table 2 Proposed Action as the Result of a Positive Malaria ELISA Screening Test
for Donors Donating in Nonendemic Countries
Test results Action
ELISA screen-reactive Repeat ELISA + IFAT; if one is positive, reject donation

or borderline
ELISA positive Exclude from further donation
IFAT negative No medical follow-up
ELISA positive Reject donation
IFAT positive Clarify history; last possible exposure to malaria
(i.e., last visit to malarial area)?
Medical history >2 years Exclude from further donation
No medical follow-up
Medical history Review in local infectious diseases unit or advise donor:
6 months to 2 years positive antibodies test
Will need investigation if febrile in the next year or in-
form physician of antibody result and suggest options
as above
For long-term follow-up all donors should be retested by IFAT. If not exposed to malaria, antibodies
will usually disappear within 3 years, and donors may be reintegrated to the donor panel, if not rejected
by otherwise cause.
Source: Adapted from Ref. 27.
fixed onto a cellulose strip, which allows detection of circulating P. falciparum

antigen present in whole blood. This method has an overall sensitivity of
96.5100% when parasitemia is 60/L (with a lower level in cases with smaller
parasitemia) (39) and is particularly useful for screening in urban areas, al-
though comprehensive studies among blood donors are still pending. Unfortu-
nately, both tests dependent on the PfHRP-2 antigen are species specific and
should be regarded with caution in countries where P. falciparum is not the most
common agent.
Recently, the detection of parasites in the bloodstream using the automatic
acridine orange dye test (QBC system) has gained experience in the Amazon
region (20,40), with an overall sensitivity of 73%.
A screening procedure using the polymerase chain reaction (PCR) was
published using Vietnamese blood donors, with a sensitivity for P. falciparum of
14 parasites/50 L, and for P. vivax of 40130 parasites/50 L, or roughly 100
times that achieved by blood films. Although this test has been applied in a
developing country, its high cost, particularly in regions with limited resources
where malaria is of great concern, still renders it unsuitable for routine use.
366 Wendel
F. Diagnosis and Treatment

Diagnosis is supported by clinical findings and by the presence of circulating
parasites. Thick smears should be done every 8 hours for 23 days by an
experienced technician, followed by serological tests (which are usually positive
when symptoms are present). The current treatment of malaria is quite complex.
Chloroquine and several other antimalarial agents are effective against the
erythrocytic forms of P. vivax and P. ovale, while primaquine is used for hepatic
forms (which is not necessary for TTM, as only the erythrocytic stage occurs).
Quinine is effective against most P. falciparum strains, although an increase in
resistant forms has been described. Treatment must be rapidly instituted, since
outcome correlates with time to treatment. In case of doubt, it is advisable to treat
the case as a drug-resistant P. falciparum strain until a final diagnosis can be
achieved. The subject has been reviewed elsewhere (13,14).
G. Conclusion
A great deal of concern still exists related to the increase in incidence of TTM
for several reasons (3): increased travel by asymptomatic individuals from en-
demic areas to nonendemic regions (e.g., immune donors to a group of nonim-
mune recipients); the atypical incubation period (6 to >60 days); the lack of
awareness of attending physicians in most nonendemic countries, who may
additionally be unable to reach an accurate diagnosis when inadequate malarial
chemoprophylaxis is performed (which may change TTM natural evolution); the
ever-increasing appearance of drug-resistant strains, particularly of P. falciparum,
which causes the most severe cases; the resurgence of malaria in previously
eradicated areas; and, finally, the settlement of individuals in previously un-
habitated and virgin areas, mainly in South America, with a dramatic increase of
vector-transmitted malaria.
III. AMERICAN TRYPANOSOMIASIS (CHAGAS DISEASE)

A. Historical Aspects and Epidemiology
American trypanosomiasis, or Chagas disease, whose agent is the protozoa
Trypanosoma cruzi, occurs only in the Americas. Chagas disease was initially
described in 1909 in the hinterlands of Brazil, by Carlos Chagas, who demon-
strated by a series of several papers the nature of the protozoa, its morphology
in the bloodstream, its life cycle in the digestive system of invertebrates
(triatomines), cultivation in agar blood, and transmission to vertebrates (4349).
Transmission by blood transfusion was first suggested by Mazza in Argen-

tina in 1936 (50). Others later supported his original concept in Brazil, Uruguay,
and Argentina.
The first donors found to be infected were described in 1949 in Belo
Horizonte (Brazil) (51) and confirmed by others in So Paulo in 1951 (52).
The first two cases of transfusion-transmitted Chagas disease were published in
1952 (53) in Brazil, and during the same period the value of chemoprophylaxis
with crystal violet (gentian violet) was studied (54). Though regarded as a strictly
Latin American problem, Chagas disease transmission through blood compo-
nents became recognized in North America in the late 1980s (5559). Several
authors demonstrated the existence of infected donors in this region, although
lookback studies have not proved the transmission of T. cruzi to any recipient
from these infected donors (6062).
Some 90 million people are at risk in endemic areas and 1824 million are
possibly infected (63,64) in 18 Latin American countries (it has not yet been
described in Cuba or the Dominican Republic). It is estimated that 23 million
people manifest any chronic feature (cardiac or gastrointestinal), with nearly
45,000 annual deaths (65). As a result, Chagas disease is the main cause of early
retirement and years lost from incapacity; when Chagas disease is measured by
disability-adjusted life-years (DALYs), it shows the highest DALY in all Latin
America, whereas the remaining infectious diseases (malaria, schistosomiasis,
leishmaniasis, leprosy, filariasis, and onchocercosis) with public importance
represent altogether less than 25% of infectious diseaserelated economic burden.
On a global scale, Chagas disease represents the third tropical disease in DALYs,
right after malaria and schistosomiasis (66).
B. The Agent
Trypanosoma cruzi is a long and slender protozoa, belonging to the order
Kinetoplastida, family Trypanosomatidae, with a single nucleus, a flagellum, and
a kinetoplast, a DNA particle present in the mitochondria. There are three stages
in its evolutive life cycle.
1. AmastigotesThese are round, intracellular forms, with 1.54.0 m in

diameter, found as clusters in infected cells. No flagellum is seen in this
form. They are found in vertebrates in macrophages, muscle fibers, testis,
ovaries, thyroid and adrenal glands, and in the central nervous system.
2. EpimastigotesThese forms have a juxtanuclear kinetoplast and a
flagellum. They are rarely found in the bloodstream of vertebrates and
are found mainly in the foregut of the insect vector.
368 Wendel
A
Figure 3 (A and B) Trypomastigote forms present in the bloodstream from a patient with
acute transfusion-transmitted Chagas disease. This evolutive form is seen only in the acute
phase, where high parasitemia is usually present. In the chronic phase, this form is seldom
observed, usually recovered only by enrichment methods.
3. TrypomastigotesThere are usually C or U shaped, 1220 m in

diameter, with the flagellum emerging from a postnuclear kinetoplast,
allowing great motility (Fig. 3). In vertebrates they are found in
the bloodstream, lymph, and cerebrospinal fluid, especially in the
acute phase.
There are more than 100 different strains, each with a particular preference
for human, domestic, or sylvatic reservoirs (67). In the bloodstream, a poly-
morphism is also observed where two different polar forms are present.
One is mainly macrophagetropic (slender forms, represented by the Y strain)
with preferential parasitism for spleen, liver, and bone marrow cells, with high
susceptibility to complement lysis in experimental infected animals, inducing
high parasitemia and animal mortality. The nonmacrophage form (a broad one,
represented by the CL strain) induces a preferential tropism for muscle cells
(cardiac and striated), with an almost negligible parasitemia, low acute mortality
in infected animals, high resistance to complement-mediated lysis, and a higher
susceptibility to chronic persistence. In addition to the morphological differences
among different strains, polymorphism can also be observed according to dif-

ferent isoenzymes patterns (zymodemes) (68,69), kDNA cleavage by RFLP
(schyzodemes) (70), or molecular sequencing of several parasite genes (71).
C. The Vector
The invertebrates responsible for transmission of T. cruzi are hematophagous bugs
belonging to the family Reduviidae and subfamily Triatominae (Fig. 4), with over
110 different species listed worldwide. Only 40 are adapted to human habitats,
and they are a significant vector for T. cruzi only in the New World, where they
can be detected from latitude 42N (northern California, Utah, Maryland) to
latitude 46S (Patagonia). They are known in English as the kissing bug or
cone-nose bug, in Spanish as vinchuca, and in Portuguese as barbeiro. The most
important reduviids are shown in the Table 3.
D. Mechanism of Transmission and

Life Cycle (Vertebrate-Invertebrate)
In nature, invertebrate triatomines feed basically on monkeys, marmosets, sloths,
armadillos, and skunks and live mainly in holes in bark of trees. For peridomestic
or intradomiciliary transmission (where cats, dogs, rodents, and humans are the
370
Figure 4 Chagas disease patterns of transmission. Sylvatic reservoirs (monkeys, marmosets, sloths,
armadillos, skunks) perpetuate the zonosis in the wild, while cats, dogs, rodents, and humans are the
main domestic reservoirs. In this setting some species have adapted to some very rudimentary huts,
where they are prone to feed on humans. If the blood meal derived from an already infected individual,
the infective cycle will be closed; additionally, infected persons may transmit Chagas disease through
blood transfusion, organ transplants, congenitally, breast feeding, or by accidental laboratory exposure,
Wendel
leading to new acute cases. Further details about transmission and life cycle can be found in the text.
Table 3 The Most Important Reduviid Species Linked to Chagas Disease

Transmissiona
Species Country
Triatoma infestans Argentina, Bolivia, Brazil, Chile, Uruguay, southern Peru

Rodnius prolixus Colombia, Mexico, Venezuela, Central America
Triatoma dimidiata Ecuador, Mexico, Central America
Triatoma sordida Argentina, Bolivia, Brazil, Paraguay
Triatoma brasiliensis Brazil
Panstrongylus megistus Brazil
Triatoma berberi Mexico
Rodnius pallescens Panama
aTriatoma infestans, Rodnius prolixus, and Triatoma dimidiata are considered as the primary vectors
for Chagas disease transmission, whereas the remaining five species are secondary ones. T. infestans
is the only one that has mainly an intradomiciliar behavior, being the target of a multicountry effort
in South America for its complete eradication (the Southern Cone Ministerial Initiative) (72).
main reservoirs), some species are well adapted to very rudimentary human
dwellings made with wattle or bamboo, having unplastered walls with palm-
thatched roofs. These huts or shacks are usually found in isolated poor areas of
Latin America, grouped into small clusters or scattered around tiny villages.
After having a blood meal, the infected bug defecates in the sucking area,
leaving infective metacyclic trypomastigotes in feces, entering the host through
the feeding puncture or through mucosal membranes or by local scratching.
In humans, invasion into the macrophages occurs either via specific cellular
receptor or by endocytosis, and after a 20- to 30-hour period, a binary division
begins in the cytoplasm, generating 50500 intracellular amastigotes every 12
hours, with later differentiation to epimastigotes and trypomastigotes, which lyse
the infected cells, leading to release into the bloodstream. This, in turn, enables
the invasion of other cells, leading to host death or, more frequently, to the
development of an immune process that controls parasitemia, although with no
parasite eradication. Thus, parasitemia is present only in the acute process, while
in the chronic stage of Chagas disease, subpatent parasitemia is observed,
evidenced only by parasite-enrichment methods. There are no spontaneous cures
or normal relapses, except when a host immunosuppressive status occurs, such as
after organ transplants, oncological treatments, or in AIDS (73).
Chagas disease is also transmitted congenitally and by breast-feeding,
accidental laboratory contamination, organ transplants (74,75), or blood transfu-
sion (Fig. 4) (67).
372 Wendel
E. Prevalence Among Blood Donors

Since the Americas comprise more than 20 different countries, a wide array of
infected donors is observed in different regions, ranging from as low as 0.01%
in the United States to 60% in certain Bolivian cities (63,67,7680), as shown in
Table 4 and Figure 5. In So Paulo, Brazil, a study of over 105,000 blood
donations yielded a repeatedly reactive rate of 1.76%, with a final prevalence of
1.03% (95% CI, 0.971.09%) using an experimental confirmatory test (Western
blot) (81). In California, where 40% of donors are of Latin origin, the estimated
prevalence ranged from 0.1 to 1.1% (84). A multicenter study of Hispanic donors
(as evidenced by the surname given by the donors) demonstrated a prevalence
rate of 0.166% (85). Another U.S. study with donors who responded positively to
a broad risk concerning Chagas disease (n = 3978) gave a repeatedly reactive
(RR) rate (by ELISA) in 15 (0.38%) donors, 8 (0.201%) confirmed by radio-
immunoprecipitation assay (RIPA) (86). As a negative control, donors who
during questioning denied risk for Chagas disease (n = 3224) were tested. Four
(0.124%) were RR (ELISA); one was (0.031%) confirmed positive on RIPA.
Overall 19 (0.264%) of 7,202 donors were RR, with 9 (0.125%) confirmed by
RIPA. A study by Winkler et al. (87) of 13,309 donors gave 16 repeatedly reactive
samples by ELISA (0.12%), 9 confirmed by RIPA (final rate of 0.07%). Another
U.S. (62) study in a low-risk population (Texas and Oklahoma), where 100,089
donors were tested by ELISA found an overall repeat reactivity of 0.15%. Two
percent were confirmed by RIPA, giving a final confirmed positivity of 0.003%,
or 1 in 33,000 donors. However, all of these infected donors came from a single
area in Texas; the final rate in this particular region was 1 of 7700 donors. These
data clearly indicate that T. cruziinfected donors are no longer restricted to Latin
American countries.
Continuous surveys display a progressive decrease in the prevalence of
T. cruzi among Latin American blood donors. Three main reasons for this are
(a) implementation of efficient sanitary programs, (b) urbanization of the popu-
lation and (c) replacement of paid donors with volunteer, altruistic donors.
Although better transfusion services are in place in Latin America today, many
places are devoid of such efficient programs, still rendering blood transfusion as
the second most important route for Chagas disease transmission.
F. Transmission by Blood Components

With the exception of lyophilized plasma (89) and blood derivatives subjected to
sterilization procedures (e.g., albumin, gamma globulin and clotting factor con-
centrates), all blood products are infective. Trypanosoma cruzi remains viable at
4C for at least 18 days (90) or up to 250 days when kept at room temperature
(91) (92,93). The viability of the parasite is somewhat lower in frozen compo-
Table 4 Prevalence of Trypanosoma cruzi Antibody Among Blood Donors from

Several Countries
Positive results by Positive results by

Country Samples Year screening tests (%) supplemental tests (%)
Argentina 194,752 1993 6.7

498,380 1994 5,6
Bolivia 1,298 1990 25.0
Brazil 105,506 1992 1.97 1.03a
835,764 1993 0.44
1,099,601 1994 0.70
Chile 163,979 1992 1.33
Colombia 1,716 1994 1.5
Costa Rica 2,574 1991 1.01
Ecuador 44,172 1994 0,11
El Salvador 20,438 1994 1,47
Guatemala 34,070 1994 1,4
Honduras 27,885 1994 1.24b
Mexico 3,419 199192 1.28
Paraguay 30,252 1994 5.3
Peru 1,481 1994 2.9
Uruguay 57,205 1994 0.8
Venezuela 961,933 198492 1.20c
584,795 1993 1.14
United Statesd 988 1991 0.1 to 1.1e
7,835 1992 0.17f
3,978 1994 0.38 0.201 (HR)g
3,224 1994 0.12 0.03 (LR)g
13,309 1995 0.12h 0.07
23,978 1997 0.30 0.14 (HR)i
25,487 1997 0.13 0.004 (LR)i
100,089 1997 0.15 0.003 (LR)j
Data are mainly derived from screening tests, except when specified in results by supplemental tests.
aWestern blot.
bFrom Ref. 82.
cFrom Ref. 83
dRIPA data from supplemental results in United States: HR = high-risk donors; LR = low-risk donors.
eFrom Ref. 84.
fFrom Ref. 85.
gFrom Ref. 86.
hFrom Ref. 87.
iFrom Ref. 88.
jFrom Ref. 63.
Source: Modified from Refs. 63, 67, 79, 80.

374 Wendel
Figure 5 Estimated prevalence of Chagas disease among blood donors from 16 different
countries from the American continent, based on data from Table 4, and considering each
country as a whole, although marked regional differences are observed within each country
(Adapted from Refs. 63, 67, 80, 104.) Countries depicted in white have unknown data.
nents (up to 24 hours), although several patients with hemophilia treated only
with cryoprecipitate have been infected (94,95). In many places, blood transfu-
sion is recognized as the second most important way of transmission of Chagas
disease; on the other hand, it can be recognized as the main important route in
industrialized countries (e.g., Canada, the United States, and Spain) (5558,96).
The true number of reported cases is grossly underestimated, since no more
than 300 cases have been recently published in the literature (63,67), clearly
representing the tip of the iceberg. One reason for this is the lack of knowledge
or awareness of the disease (especially observed in industrialized, nonedemic
areas). In addition, there has been a decrease in observed cases in most developed
Latin American urban services, particularly in Argentina, Brazil, Chile, and

Uruguay (97).
Although the number of reported cases in countries of the northern hemi-
sphere is quite low, the recent and intense emigration from these countries is of
some concern. Currently, it is estimated that there are at least 7,000,000 legal
Latin American immigrants in the United States, 250,000 in Europe, 150,000 in
Japan, and 80,000 in Australia (Fig. 6) (63,67,77,98). Although they still do not
represent a great percentage of the donor pool, the observed prevalence in some
selected donor population in the United States (several times higher than for HIV,
HTLV I/II, or HBsAg) clearly confirms what was predicted a few years ago (78).
Furthermore, it is expected that 50,000370,000 immigrants to North America are
infected by T. cruzi (99101), nearly 75,000 of whom might bear some cardiac
manifestation (102). Chagas disease is slowly but gradually changing its natural
geographical limits, putting other countries at risk unexpected only a decade
ago (103).
G. Risk Factors and Probability of Infection

The possibility of infection by blood components depends on several factors, such
as the amount of transfused blood, the parasite strain, presence of parasitemia at
the time of donation, and the recipient immune status. Additionally, the probabil-
ity of infection is highly dependent on whether screening tests are performed in
blood banks (67,78,104).
1. Unscreened Blood Components

In 1972, Cerisola et al. in Argentina (94) reported that the risk of transfusion-
transmitted Chagas disease could be estimated based on Newtons binomial
distribution:
P = 1 (1f)n
where f = the prevalence of infected donors in the population and n = the number
of transfused units. However, no attention was paid to the probability of a patient
becoming infected when receiving one unit of infected blood; this is actually not
100%, but rather an average of 1225% although higher rates (47.6%) were
observed in Bolivia, a hyperendemic region (107). Thus, it seems rather prudent
to add to the original Cerisola formula the k factor for infectivity. It is not known
why the infectivity is not very high, but the low parasitemia rate (1 parasite per
20 mL of blood) and the simultaneous presence of inhibitory antibodies in the
plasma may be responsible. On the other hand, the survival rate (SR) of recipients
1 or 2 years after the transfusion event, averages 4050% (108). Therefore, the
376
Figure 6 The main pattern of human migration from Latin America to North America, Europe, Australia, and Japan, with the
corresponding expected legal immigrants in each region, although the actual number is expected to be more than this. (Based on Refs.
Wendel
63,67,77.)
final estimated risk formula for transfusion-transmitted Chagas disease when

unscreened units are used should be calculated as:
P = 1 (1f)n k SR
In addition, one can speculate that the risk is also proportional to the ratio of
components produced (CP) from a single whole blood unit related to the number
of individual patients transfused with such units (PtTx), i.e., one whole blood unit
is transformed into x components that are transfused into y recipients.
Naturally, this additional index (CP/PtTx) should be included in the formula
above; however, it seems to be quite difficult to really estimate it, especially in
Latin America.
2. Screened Blood Components

Currently, there are several screening tests for T. cruzi antibodies. However, a
Brazilian report showed that transmission is still possible despite serological
screening. A rate of 0.79% (12 of 1503) false-negative results has been reported,
due mainly to the use of a single screening test (IHA, the least sensitive method
available). The chances decreased when two different screening procedures are
used, a strategy promoted by WHO (110,111), the Pan-American Health Organi-
zation (PAHO) (112) and the Brazilian Ministry of Health (113). With the use of
two screening methods, Takei reported a sensitivity of at least 99.7% (114).
3. Partially Screened Components

While evaluating data from 12 different Latin American countries (Bolivia, Chile,
Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Honduras, Nicaragua,
Paraguay, Peru, and Venezuela) with some 1,200,000 donations in 199394,
Schmunis et al. (104) reported that only Honduras had a full screening program
for T. cruzi. Among the other countries, by calculating the number of donations,
the coverage of the blood supply, the actual prevalence of infected donors, the
sensitivity of the diagnostic kits used for screening, and the fractionation index
(i.e., the number of final components produced from the whole blood units
donated), these authors reached that the probability of getting a T. cruzi transfu-
sion-transmitted infection, P(I), ranged from as low as 2.1:10,000 units in
Nicaragua to 219:10,000 units in Bolivia.
A different approach to calculate the residual risk based on screened
components would be one linked to the window period, as already calculated for
some viral diseases (113,114,115). This method would have limited utility in
calculating the residual transfusion risk of Chagas disease, since the likelihood of
donors being in the window period is quite remote as infection is usually acquired
during childhood. In addition, such studies involve hundreds of thousands of
378 Wendel
donors, with millions of units testeda number quite difficult to be fully

evaluated in Latin America.
Based on the previous formula and a test sensitivity of 99.7% in a location
with a prevalence rate of 1%, one can predict that the risk of an acute transfu-
sion-transmitted Chagas disease (24 months after transfusion) is approximately
1:200,000 units, not accounting for the survival rate. Although these estimates are
theoretical and no clear-cut prospective epidemiological studies have confirmed
these data, limited studies have reinforced the estimated risk (116).
Data presented at the Brazilian Ministry of Health in 1999, using the
assumptions previously reported (104) and a national prevalence of 0.83%,
showed that this risk would now be 1:109,000 units (117) in Brazil.
H. Clinical Symptoms in Recipients

The clinical findings observed in recipients of infected units are almost the same
as those observed when infection occurs by insect transmission, except that the
chagoma of inoculation, a typical swelling of skin, face, or eyelids (representing
the entry site of parasites), is not observed. The incubation period varies from 20
to 40 days (range 8120 days). Fever is by far the most common and sometimes
the only manifestation. Lymphadenopathy and hepatosplenomegaly may also be
present, and the association of these latter symptoms must always be suspected
as Chagas disease in recipients transfused from untested Latin American donors.
Cardiac arrhythmia, with disturbances of atrioventricular conduction, ECG
alterations, or reduction of ejection fraction may occur, leading to pericardial
effusion and/or cardiac arrest. Death, although uncommon, may be seen in the
most severe cases, usually in immunocompromised recipients.
The central nervous system can also be affected, with somnolence, fatigue,
and tremors as the most common symptoms. Myoclonus, seizures, meningitis, or
meningoencephalitis, although very rare, is seen in the most severe cases, usually
in immunocompromised patients.
The gastrointestinal system is usually spared in acute Chagas disease.
A key feature of transfusion-transmitted Chagas disease is that approximately
20% of infected recipients are completely asymptomatic, raising no suspicion of
diagnosis.
After the acute phase, a spontaneous recovery will ensue after 68 weeks,
but may extend up to 4 months. Thereafter, the disease follows its natural course
to an indeterminate phase (persisting for years or decades), which represents the
majority of cases. The chronic phase, with cardiac, gastrointestinal, or neuro-
logical symptoms, is observed after several years and usually is not linked to an
acute phase.
When the heart is affected, the myocardium becomes thin, with right
and left chamber enlargement. Disturbances of atrioventricular conduction and
Adams-Stokes syndrome (due to right bundle branch block) may occur. Apical
aneurysm of the left ventricle with attached thrombi to the endocardium is often
found, leading to peripheral embolism.
The gastrointestinal system is compromised in 810% of patients, with
denervation of autonomic parasympathetic ganglia (Auerbachs plexus) with
esophageal or colonic hypotonia. Enlargement of internal organs (esophagus or
colon) are known as megasyndromes (e.g., megacolon or megaesophagus).
I. Clinical Symptoms in Donors

One of the key issues concerning the clinical findings among blood donors is that
they are, in the vast majority, asymptomatic. Nevertheless, they can be subdivided
into three main groups.
1. Serological Chagas diseaseOnly antibodies are found, without any
evidence of symptoms, comprising 6080% of all infected donors.
2. Latent chronic phaseDonors show some visceral abnormalities as
evidenced by different diagnostic tests (x-ray, CT scan, ECG) but are
still asymptomatic, requiring a close monitoring and follow-up, since
clinical symptoms may develop in the future.
3. Symptomatic donorsConsidered the tip of the iceberg; many times
the symptoms are not persistent or specific.
As there can be an overlapping between the two latter groups, the true
prevalence of each one is difficult to determine, although about 50% of patients
with any heart or gastrointestinal abnormalities show clinical symptoms. Accord-
ing to Gontijo (118), a study of 291 chagasic donors revealed that 54% were in
the indeterminate phase, 38% had cardiac symptoms, 18% had digestive symp-
toms, and 10% had both cardiac and digestive symptoms. Among them, 79.5%
with chronic cardiopathy and 75% with esophagopathy were in the initial stage
of clinical evolution; severe heart disease was found in only 3.1% of cases.
J. Preventive Measures
Prevention of Chagas disease can be accomplished by three different strate-
gies (119).
1. Anamnesis and Questionnaires to the Donors

Paid blood donors are no longer accepted in Brazil, although they still represent
almost 40% of donors in some Latin American regions (most notably in Bolivia).
An increased rate of infected donors is associated with a higher age, first-time
donation and longer duration of living in endemic areas (120,121). In endemic
380 Wendel
and nonendemic regions, donors who lived in infested dwellings or acknowledge

having been bitten by the bug must be deferred from donation. Specific question-
naires are in use in California to identify high-risk donors; in 70 of 3492 eligible
donors who were disqualified due to their answers, 45 samples were tested for
T. cruzi antibodies, and two had a positive result (122); similar findings were
reported in a subsequent North American study (86).
2. Serological Tests
Several method are available for screening, some of which are licensed in the
United States. Complement fixation was gradually replaced by IHA, IFA, or
ELISA, but none can be considered 100% sensitive. The possibility that serolog-
ical markers currently in use in blood banks (e.g., HBsAg, anti-HCV, -HIV,
-HTLV I/II, -HBc, ALT, or syphilis) could be used as surrogate markers for
T. cruzi was evaluated in 26,365 Brazilian donors (121). A very slight association
was observed only between syphilis in female donors (p = 0.005), but the low
number of cases (n = 4) precluded considering this to be a potential surrogate
once the same effect was not found in male donors or when the whole group was
considered. Thus, it seems that if T. cruziinfected donors are to be screened,
specific serological tests must be performed. Unfortunately, mandatory serologi-
cal screening is not a current approach in all Latin American countries, except for
Argentina, Brazil, Honduras, Paraguay, Uruguay, and Venezuela. Chile only
applies it in endemic regions. In Ecuador, although not mandatory, the majority
of Blood Services utilize testing.
3. Chemoprophylaxis
Because some countries still have a high prevalence of T. cruziinfected donors,
it is almost impossible to find enough noninfected donors to meet blood supply
needs. In addition, some of these areas are poor and devoid of a sophisticated
blood service. Also, some countries with a very low prevalence of infected donors
still do not consider Chagas disease a transfusion problem. Thus, it seems quite
reasonable to develop the sterilization of blood components, a procedure currently
under consideration in Bolivia, as advised by WHO (123). The first attempt to
promote sterilization of whole blood was by addition of thimerosal in 1952 (124).
Subsequently, other drugs were tested. An ideal model for screening drugs has
been developed, and more than 1000 drugs have been tested so far; among them,
only a few showed some trypanocidal effect. To have some value as a chemopro-
phylactic agent, a drug must fulfill the following criteria, irrespective of its mode
of action (125).
1. The drug must be active at pH 7.4 and at temperatures ranges from 30
to 22C, at which blood components are stored.
2. The drug must be active in undiluted blood and not adversely affect the
main metabolic and energetic pathways of red blood cells, platelets,
and plasma proteins throughout their storage period. In addition, no
interference must arise with the original antigenic composition of
human cells, nor should new antigens be generated.
3. The drug must not only be safe, but must also be used at a concentration
that, when transfused into the recipient, should not cause any pharma-
cological response for which it was originally designed.
Crystal Violet. This phenylmethanic dye totally eradicates T. cruzi viabil-

ity within 24 hours at a concentration of 1:4000 or 200 g/mL (0.6 mM) (54).
It comprises 96% of the composition of gentian violet, the main commercial salt
available (the other components are penta- and tetramethylparasaniline, the latter
known as brilliant green). The structural formula of crystal violet and its mecha-
nism of action are shown in Figure 7.
Mechanism of action. Crystal violet is reduced in the organism in the
presence of NADPH, forming carbon-centered free radicals, which are able to
remove hydrogen from other molecules, be added across unsaturated bonds, or
combine with themselves to form dimers (126,127). This carbon-centered free
radical can also auto-oxidize, producing a superoxide anion (2 ). This anion, in
the presence of superoxide dismutase (SOD), is converted into hydrogen peroxide
(H2O2). Because T. cruzi is deficient in catalase and reduced gluthatione (GSH),
which degrades H2O2 in H2O and O2, whenever there is an increase in H2O2, a
direct trypanocidal effect is observed; this can also be enhanced by the action of
Fe2+. The interaction between 2 and H2O2 generates the hydroxyl radical
(OH), one of the most toxic radicals, which targets the organisms mitochondria.
Since the action of crystal violet is not immediate, an incubation period is
necessary for complete parasite kill (24 hours at 4C). The presence of light
enhances this reaction 17-fold (126) additionally, reducing agents such as ascor-
bate acid will increase H2O2 generation (128). A combination of light (130 W/m2)
and ascorbate (10 mM) reduce both the time exposure (20 min) and the dye
concentration (1:16,000 or 0.4 mM) with the same efficacy (129,130). Generation
of H2O2 is not toxic to red or white blood cells because there is a high activity
of catalase and GSH in these cells. Furthermore, the dilution effect after transfu-
sion leads to a low concentration of H2O2 in the recipient.
Parasite sensitivity. A study of a patient in the acute phase whose blood
(treated with crystal violet at a concentration of 1:2000 for 48 hours) was
deliberately administered to a volunteer showed no evidences of transmission,
seroconversion, or xenodiagnosis in the recipient up to 90 days after the transfu-
sion; an untreated sample was highly infective when injected in experimental
mice (131). The effect of crystal violet has been extensively studied in blood from
382 Wendel
Figure 7 Mechanism of action from crystal violet against Trypanosoma cruzi. CAT, Catalase; AH,
ascorbate; GSH, reduced gluthatione; SOD, superoxide dismutase. (Adapted from Refs. 126129.)
donors belonging to the chronic phase. It has been proved to be effective against
all parasite stages (amastigotes, epimastigotes, and trypomastigotes) (126,132)
and several different strains (e.g., Y, FL, G, J, M, Peru, and Sonya). In an
experimental study with 18 recipients transfused with seropositive treated blood,
none developed infection (133). Rezende et al. (134) reported that among a group
of 774 recipients transfused with unscreened blood components, at least 300 of
which could be assumed to derive from infected donors, none developed acute or
chronic infection. The Brazilian experience using over 50,000 units has confirmed
the efficacy of crystal violet treatment (129).
Side effects. There are some mild effects on red blood cells (Rouleaux
formation), but no changes in hemoglobin, pO2, pCO2, pH, Na+, or K+ have been
observed. ATP and 2,3 DPG levels are slightly decreased, but the results are not
statistically different from controls. The major imbalance of crystal violet occurs
on platelets, perhaps by a direct action on the mitochondrial calcium metabolism

(135). In recipients, a slight purple color is observed that lasts for 24 hours, which
must not be considered an obstacle to use in highly endemic and remote areas,
particularly when developed blood services are not available. In addition, crystal
violet has been shown to be carcinogenic in rodents, but no effect was observed
in humans (136). The maximal infusion dosage into recipients is still uncertain,
although one report states it as 510 mg/kg (easily achieved with 2500 mL of
treated whole blood); its effect in massive transfusions is contradictory. Infusion
of crystal violet in infants is associated with white blood cell depression.
When used for other therapeutic purposes, it leads to gastrointestinal irritation
(nausea and vomiting), skin necrosis, thrombophlebitis, and keratoconjunctivitis.
The long-germ effect is also unknown, although some patients successfully
transfused with over 36,000 mL of treated blood over a 6-month period have been
reported (134).
Pc4. This phthalocyanin has also been discussed in Section II. A complete
inactivation (45 log 10 kill) of T. cruzi (Y strain) was observed at 2 M with
7.5 J/cm2 for FFP and 15 J/cm2 for red blood cells after a 5-minute treatment with
photoirradiation (137). As with crystal violet, the main target for T. cruzi inacti-
vation is the protozoan mitochondria, where a swelling and a shearing effect upon
the kinetoplast is observed; in the absence of light irradiation, less extensive
damage is observed. Unfortunately, no clinical studies dealing with survival rates
of transfused red blood cells have been published so far.
Other Agents. Phenolic antioxidants, amphotericin B, imidazole deriva-
tives, quinolines (WR6202), and gamma irradiation have been tested, but all
either failed to prove a superior effect over crystal violet or are still pending
further safety studies. Recently, the role of leukocyte filters in T. cruzi retention
has been studied; preliminary studies achieved a partial protection in mice heavily
transfused with the Y strain (a highly pathogenic strain) in doses 5001000 higher
than observed in chronic infected donors (138140).
K. Laboratory Screening
Several laboratory methods for blood donor screening are available.
1. Detection of Antibodies
Because of the easy application of the procedure and its relative low cost and high
sensitivity (though with variable specificity), methods aimed at T. cruzi antibody
detection are currently in use as a routine clinical diagnosis or as a blood donor
screening procedure. Several methods are commercially available in Latin Amer-
ica and the United States. The majority use whole parasites or crude lysates as
384 Wendel
the antigen source, although some research has been based on the development
of synthetic peptides or recombinant antigens. Trypanosoma cruzi antibody titer
is quite variable in an infected population and usually bears no relationship to
clinical symptoms; it also varies in the same individual during his or her lifetime.
The current serological methods are prone to cross-reacting with sera
containing antibodies against other infectious agents, such as yeasts, Leishmania
(Kala-azar and mucocutaneous leishmaniasis), T. rangeli (a nonpathogenic
agent found in Venezuela, Colombia, Peru, Ecuador, Panama, and Costa Rica),
T. gambiense (the agent of the sleeping sickness or African trypanosomiasis),
H. muscarum, L. seymoury, C. fasciculata, and other trypanosomatids. Addition-
ally, nonspecific IgM directed against phosphocoline (145), an antigenic fraction
widely present in several mycobacteria and parasites, is also responsible for
cross-reactivity, rendering the serological methods for T. cruzi antibody detection
still open for improvement. One has also to bear in mind that inconclusive results
are seen in approximately 35% of all blood donors, characterized by a reactive
result in only one assay (usually with clear negative or very low titer in a second
test). Even when a sample displays a clear reactive pattern, one cannot expect a
high positive predictive value. These events usually pose a major problem in
counseling of donors. For this reason, supplemental assays are also necessary
in order to confirm the screening tests. Finally, it is currently rather difficult
to obtain seroconversion panels as a result of the successful eradication of
disease in several countries, which makes the standardization of new tests a very
difficult task.
The main serological methods used for antibody detection are described in
the following sections (114,146).
Complement Fixation. This, the first serological test, was described in

1913. It shows low sensitivity in acute cases (<35%), reaching maximum sensi-
tivity (9095%) within 46 weeks after infection. Although very inexpensive, the
difficulties found in standardizing all components of the assay and low sensitivity,
render this test unsuitable for routine screening.
Latex Agglutination. This test is based on the principle of aqueous antigen

solution fixed onto latex particles, rendering a visual agglutination in the presence
of antibodies. It also has a variable sensitivity and specificity and is used only in
some remote areas of Latin America.
Direct Agglutination. This test is based on epimastigote forms treated

with trypsin and fixed in formaldehyde, which are directly agglutinated in the
presence of antibodies. Treatment of sera with 2-mercaptoethanol allows also a
distinction between IgG and IgM antibodies. Because of its high sensitivity in the
acute phase, it is particularly useful in those rare cases. However, false-positive
results are usually found in the presence of heterophil antibodies or nonspecific
agglutinins. Because of its relative high cost and the difficulty in finding suitable
commercially available reagents, this method is not used currently as blood
donor screening.
Indirect Hemagglutination Assay. Aqueous, soluble epimastigote extracts
can be easily fixed onto avian, sheep, or human red cells treated with tannic acid,
glutaraldehyde, or formaldehyde (147), maintaining satisfactory stability for
several months; several commercial kits are available. Although widely used
with easy performance, this method shows a lower sensitivity than IFAT or
ELISA (148150).
Indirect Immunofluorescence Assay. This test is based on fixed cultured
cells on slides and read by fluorescence microscopy (151,152). This very sensitive
test is usually the first to detect seroconversion samples (both IgM and IgG).
The practical performance of the test, its commercial availability, the long storage
period, relatively low cost, and high sensitivity were responsible for its wide use
as a screening test in Latin American countries since the mid-1970s. There are
two main problems with IFAT: the first is its nonapplicability to high-volume
screening. The second is its relatively high rate of false-positive results, since the
antibodies react mainly against intact membrane antigens, which are common to
other agents (especially Leishmania spp.) or epitopes displaying phosphocholine
(145). There is a wide array of autoimmune antibodies that also show consider-
able cross-reactivity (153).
ELISA. Since the early report by an immunoenzymatic test using crude
lysates (154) with some problems in its early phase (155,156) several important
improvements have been developed, including more purified antigens that en-
hanced the sensitivity and specificity of the test (157159). Recombinant proteins
or synthetic peptides show better specific results than parasite extracts (lower
cross-reactions) (160,161) but still have variable sensitivities according to the
patient and clinical manifestation of the disease, but still have to enhance the
sensitivity, an alternative would be to combine several recombinant antigens or
synthetic peptides to sensitize microwells. However, some interferences, such as
histeric hindrance, leads to poorer results for the mixed antigens than for
individual antigens (159,162), still leaving this alternative for the future. The pos-
sibility of automation and large use to several samples make this method as the
most reliable one for screening blood donors. Several kits are commercially
available, including in the United States, all based in epimastigote antigens.
2. Supplemental Tests
As previously discussed, there is still a need for good supplemental assays in
order to validate the original antibody screening tests used in the blood bank.
Currently, two assays are under evaluation.
386 Wendel
Western Blot and Line Immunoassay. Crude or purified antigens are run
in polyacrilamide gel electrophoresis and then transferred onto nitrocellulose
strips. All reactive bands are detected by an immunoenzymatic reaction. Although
this method is widely used for viral tests, there are still several problems when it
is applied to T. cruzi. The first one is the large number of bands present in the test,
according to each antigenic preparation, with some authors reporting up to 15 or
20 different bands ranging from 10 to 300 kDa; in addition, two-dimensional
electrophoresis reveals the presence of different proteins with the same molecular
weight (163). The second problem is that several antibodies are not quite specific
and when eluted from a single band and put again to react against a new
nitrocellulose strip, specimens will react against bands other than those originally
eluted (164). The sensitivity of Western blot has been reported to range from 86.8
to 100% (116,165,166). Highly purified antigens (natural or recombinant) or
synthetic peptides on nitrocellulose strips have been developed either in combi-
nation or individually, as single (167) or multiple line assays (168170).
Radioimmunoprecipitation Assay. Cultured epimastigotes, when placed
in a medium containing a radioisotope (125I or 35S), will incorporate radioactivity,
which will be demonstrated when lysed forms are reacted against specific
antibodies present in the serum. Two major glycoprotein bands are related to a
positive result: gp 72 and gp 90 (171). With some slight modifications in the
principle, there are three groups in the United States (87, 171,172) currently in
the process of validating the method; however, due to its complexity and high
cost, it will be restricted to only a few labs.
PCR. Chagas disease is a model of intense PCR efforts due to the relative
low sensitivity and technical difficulties involved with xenodiagnosis of and
hemoculture attained from chronic patients; in addition, both methods require as
long as 120 days to reach a conclusive result.
The main difficulty faced in applying PCR for the detection of Chagas
disease in chronic patients derives from the low and intermittent parasitemia
observed in an important fraction of chronic carriers. In order to obviate that,
investigators have focused on DNA sequences largely represented on the T. cruzi
genome, such as on the kinetoplast (141). Even though no methodology has
achieved 100% sensitivity compared to serology of chronic patients, the results
of all published comparative studies clearly favor the use of PCR for confirmation
of a serology-positive result for Chagas disease.
One problem still unsolved is the large blood volume that must be sampled
and extracted to detect T. cruzi DNA by PCR. PCR methods achieve a sensitivity
of about one parasite in 20 mL of blood, but chronic carriers may harbor a
parasitemia less than that amount.
There has been a report of a PCR method based on DNA extracted from
serum that achieved similar results when whole blood DNA was obtained
from the same samples, which led investigators to suggest that serum could
replace whole blood for PCR detection of T. cruzi (142). These results must be
interpreted with caution based on the fact that parasites tend to remain on the
buffy coat of separated blood, as observed by many groups.
PCR is currently applied for the diagnosis of Chagas disease in both
chronic and acutely infected patients and is under scrutiny as a confirmatory test
for blood banks. It is also the method of choice for testing chemotherapeutic
agents for T. cruzi on experimentally infected animals (143), and a quantitative
PCR method has been developed that can be used to monitor patients during
treatment (144).
L. Therapy
Treatment of acute cases is done by controlling the symptoms and by the use of
nifurtimox or benznidazole (173). These drugs are effective only in the acute
phase and should be used immediately after diagnosis; benznidazole might be
effective in the chronic phase in infected children (174), though it is not devoid
of severe side effects. The clinical management of chronic Chagas disease is
directed toward the control of signs and symptoms, but no permanent cure can
be achieved.
M. Conclusion
As in malaria, several problems related to the geographical distribution, pattern
of emigration, high number of asymptomatic donors, lack of medical knowledge
to correctly diagnose most of the cases, particularly in developed countries, and
the still pending technical refinements for laboratory screening result in Chagas
disease remaining a major public health problem in all American countries.
In addition, even with complete vectorial elimination, Chagas disease will re-
main as a zoonosis. Millions are already infected, requiring medical attention
throughout their lifetime. Finally, some asymptomatic patients are women of
childbearing age, who will continue to propagate the disease congenitally. Never-
theless, continuous efforts in several countries are yielding positive results;
complete elimination of the main vector is expected in fewer than 10 years. Also,
one can speculate that, given the association between exposure and age (120,121),
it is likely that no further vectorially contaminated donors younger than 30 years
of age will be found within 1020 years in Brazil (175). Thus, it seems that the
strategies currently developed to prevent transfusional transmission of Chagas
disease will certainly need a reassessment in a decade. This is a remarkable feat,
because only 100 years will have elapsed between the discovery and elimination
of Chagas disease, an unprecedented accomplishment in the history of human
infectious diseases.
388 Wendel
REFERENCES
1. Woolsey G. Transfusion for pernicious anaemia: two cases. Ann Surg 1911;
53:132135.
2. Turc JM. Malaria and blood transfusion. In: Westphal RG, Carlson KB, Turc JM,
eds. Emerging Global Patterns in Transfusion-Transmitted Infections. Arlington,
VA: AABB, 1990:3143.
3. International Forum: Which are the appropriate modifications of existing regulations
designed to prevent the transmission of malaria by blood transfusion, in view of the
increasing frequency of travel to endemic areas? Vox Sang 1987; 52:138148.
4. WHO. World malaria situation in 1990. Weekly Ep Rec 1992; 67:161168.
5. WHO. World malaria situation in 1990. Weekly Ep Rec 1992; 67:169176.
6. Gilles HM, Warrell DA. Bruce-Chwatts Essential Malariology. 3rd ed. London:
Edward Arnold, 1993.
7. Bruce-Chwatt LJ. Transfusion malaria revisited. Trop Dis Bull 1982; 79:827840.
8. Bruce-Chwatt LJ. Transfusion malaria. Bull WHO 1974; 50:337346.
9. Bruce-Chwatt LJ. Blood transfusion and tropical disease. Trop Dis Bull 1972;
69:825862.
10. Chataing B. Prevention du paludisme transfusionel. Rev Franc Transf Immuno-
hematol 1988; 31:8188.
11. Tabor E. Infectious Complications of Blood Transfusion. New York: Academic
Press, 1982.
selection criteria for the United States travellers to malarious areas. Transfusion
1991; 31:798804.
13. WHO. The biology of malaria parasites. Technical Report Series 743. Geneva:
World Health Organization, 1987.
14. Lpez-Antuano FJ, Schmunis G, eds. Diagnosis of malaria. Scientific Publication
No. 512. Washington, DC: Pan American Health Organization, 1990.
15. Fairley NH. Trans R Soc Trop Med Hyg 1947; 40:621676.
16. Council of Europe. Guide to the preparation, use and quality assurance of blood
components. Strasbourg, 1992: 17.
17. Wells L, Ala FA. Malaria and blood transfusion. Lancet 1985; i:13171318.
18. Andrade JCR, Wanderley DMV. Malria induzida no Estado de So Paulo, Brasil.
Rev Soc Bras Med Trop 1991; 24:157161.
19. Ferreira MV, Camargo LMA, Carvalho ME, Ninomia RT, Garcia LAV, Santos FR.
Prevalence and levels of IgG and IgM antibodies against Plasmodium falciparum
and P. vivax in blood donors from Rondonia, Brazilian Amazon. Mem Inst Oswaldo
Cruz 1993; 88(2):263269.
20. Kiesslich D, Torres KL, Fraiji NA. Mtodos de triagem para Malria em Doadores
de Sangue de rea Endmica. Bol Soc Bras hematol Hemot 1997; 19:2127.
21. Salutari P, Sica S, Chiusolo P, Micciulli G, Plaisant P, Nacci A, Antinori A, Leone
G. Plasmodium vivax malaria after autologous bone marrow transplantation: an
unusual complication. Bone M Transpl 1996; 18:805806.
22. Talabiska DG, Komar MJ, Wytock DH, Rubin RA. Post-transfusion acquired
malaria complicating orthotopic liver transplantation. Am J Gastroenterol 1996;

91:376379.
23. Tran VB, Tran VB, Lin KH. Malaria infection after allogeneic bone marrow
transplantation in a child with talassemia. Bone M Transpl 1997; 19:12591260.
24. Ibhanesebor SE, Otobo ES, Ladipo OA. Prevalence of malaria parasitemia in
transfused donor blood in Benin City, Nigeria. Ann Trop Ped 1996; 16:9395.
25. Standards for Blood Banks and Transfusion ServicesAABB, 15th ed. Bethesda,
MD: 1993.
26. Sazama K. Prevention of transfusional-transmitted malaria: Is it time to revisit the
standards (editorial). Transfusion 1991; 31:786788.
27. Chiodini PL, Hartley S, Hewitt PE, Barbara JAJ, Lalloo K, Bligh J, Voller A.
Evaluation of a malaria antibody ELISA and its value in reducing potential wastage
of red cell donations from blood donors exposed to malaria, with a note on a case
of transfusion-transmitted malaria. Vox Sang 1997; 73:143148.
28. BPAC considers malaria deferral, ICL study, plasma source. Blood Bank Week
10(25):1n4.
29. Choudhury N, Jolly JG, Mahajan RC, Dubey ML, Kalra A, Ganguly NK. Selection
of blood donors in malaria endemic countries. Lancet 1988; ii:972973.
30. Smith OM, Traul DL, McOlash L, Sieber F. Evaluation of merocyanine 540-
sensitized photoirradiation as a method of purging malarially infected red cells from
blood. J Infect Dis 1991; 163:13121317.
31. Smith OM, Dolan AS, Dvorak JA. Merocyanine 540-sensitized photoinactivation of
human erythocytes parasitized by Plasmodium falciparum. Blood 1992; 80:2124.
32. Lustigman S, Ben-Hur E. Photosensitized inactivation of Plasmodium falciparum
in human red cells by phthalocyanines. Transfusion 1996; 36:543546.
33. Ben-Hur E, Horowitz B. Advances in photochemical approaches for blood steriliza-
tion. Photochem Photobiol 1995; 62:383388.
34. Amato Neto V, SantAna EJ, Pinto PLS, Moura AAB, Duarte MFS, Campos R.
Estudo experimental sobre a possibilidade de preven da malria ps transfu-
sional atravs do uso de violeta de genciana. Rev Sade Publ S Paulo 1987;
21(6):497500.
35. Draper CC, Sirr SS. Serological investigation in retrospective diagnosis of malaria.
Br J Med 1980; 2:15751576.
36. Parra ME, Evans CB, Taylor DW. Identification of Plasmodium falciparum histi-
dine-rich protein 2 in the plasma of humans with malaria. J Clin Microbiol 1991;
29:16291634.
37. Taylor DW, Voller A. The development of a simple antigen detection ELISA for
Plasmodium falciparum malaria. Trans R Soc Trop Med Hyg 1993; 87:2931.
38. Namsiriponpun V, Wilde H, Pamsandang P, Tiersansern P. Field study of an
antigen-detection ELISA specific for Plasmodium falciparum malaria. Trans R Soc
Trop Med Hyg 1993; 87:3234.
39. Peyron F, martet G, Vigier JP. Dipstick antigen-capture assay for malaria detection.
Lancet 343:15021503.
40. Avila SLM, Leandro MC, Carvalho NB, Oliveira MS, Arruk VG, Sanchez MCA,
Boulos M, Ferreira AW. Evaluation of different methods for plasmodia detection
390 Wendel
in well defined population groups in an endemic area of Brazil. Rev Inst Med Trop
So Paulo 1994; 36:157162.
41. Hong VT, VanBe T, Tran PN, Thanh LT, Hien LV, OBrien E, Morris GE. Screening
donor blood for malaria by polymerase chain reaction. Trans R Soc Trop Med Hyg
1995; 89:4447.
42. Chikwem JO, Mohammed I, Okara GC, Ukwandu NCD, Ola TO. Prevalence of
transmissible blood infections among blood donors at the University of Maiducuri
teaching hospital, Maiduguri, Nigeria. East Afr Med J 1997; 74:213216.
43. Chagas C. Nova espcie morbida do homem, produczida por um Trypanosoma
(trypanosoma cruzi): Nota prvia. Brasil Md 1909; 23:161.
44. Chagas C. Neue Trypanosomen. Arch Schiffs Tropenhyg 1909; 13:120122.
45. Chagas C. ber eine neue Trypanosomiasis des Menschen. Arch Schiffs Tropenhyg
1909; 13:351353.
46. Chagas C. Nova tripanozomiase humana. Estudos sobre a morfolojia e o ciclo
evolutivo do Schizotrypanum cruzi n.g., n.s.p., ajente etioljico de nova entidade
mrbida no homem. Mem Inst Oswaldo Cruz 1909; 1:159218.
47. Chagas C. Nova entidade mrbida do homem. Resumo geral de estudos etioljicos
e clnicos. Mem Inst Oswaldo Cruz 1911; 3:219275.
48. Chagas C. The discovery of Trypanosoma cruzi and of American trypanosomiasis.
Mem Inst Oswaldo Cruz 1922; 15:111.
49. Chagas Filho C. Carlos ChagasOfcina Grfica da Universidade do Brazil, 1959.
50. Mazza S, Montana A, Benitez C, Juzin E. Transmisin de Schizotrypanum cruzi
al nio por leche de la madre con enfermedad de Chagas. Publ MEPRA 28:4146.
51. Pellegrino J. Transmisso da doena de Chagas pela transfuso de sangue. Primeiras
comprovaes sorolgicas em doadores e candidatos a doadores de sangue. Rev
Bras Med 1949; 6:297301.
52. Faria P. Sifilis, Maleita, Doena de Chagas e Transfuso. Folia Clin Biol 1951;
17:113117.
53. Freitas, JLP, Amato V, Sonntag R, Biancalana A, Nussenszweig V, Barreto JG.
Primeiras verfcaces de transmisso acidental da molstia de Chagas ao homem
por transfuso de sangue. Rev Paul Med 1952; 40:3640.
54. Nussenzweig V, Sonntag R, Biancalana A, Freitas JLP, Amato Neto V, et al. Ao
de corantes tri-fenil-metanicos sobre o Trypanosoma cruzi in vitro. Emprego da
Violeta de Genciana na proflaxia da transmisso da Molstia de Chagas por
transfuso de sangue. O Hospital 1953; 44:731744.
55. Geiseler PJ, Ito JI, Tegtemeier BR, Kerndt PR, Krance R. Fulminant Chagas disease
in bone marrow transplantation. 27th Intersc. Conf. on Antimicro Agents and
Chemoter. Washington: Am Soc for Microbiol 1987:418.
56. Grant I, Gold JWH, Wittnee M, et al. Transfusion associated acute Chagas disease
acquired in the United States. Ann Intern Med 1989; 111:849851.
57. Nickerson P, Orr P, Schoeder MC, Sekla L, Johnson J. Transfusion associated
Trypanosoma cruzi infection in a non endemic area. Ann Intern Med 1989;
111:851853.
58. Cimo PL, Luper WE, Scouros MA. Transfusion-associated Chagas disease in
Texas: report of a case. Tex Med 1993; 89:4850.
59. Lenes BA, Leiby DA, Tibbals MA, Olmedo M. A prospectively identified case of
transfusion transmitted Trypanosoma cruzi: a defining window period. Transfusion

1998; 10 (suppl):390.
60. Leiby DA, Yund J, Read EJ, Lenes BS, Dodd RY. Risk factors for Trypanosoma
cruzi infection in seropositive blood donors. Transfusion 1996; 36(suppl):228.
61. Galel S, Wolles S, Stumpf R. Evaluation of a selective donor testing strategy.
Transfusion 1997; 37(suppl):296.
62. Leiby DA, Jensen NC, Fucci MC, Stumpf RJ. Prevalence of Trypanosoma cruzi
antibodies in a blood donor population at low risk. Transfusion 1997; 37(suppl):297.
63. Schmunis GA. American trypanosomiasis as a Public Health Problem, in PAHO:
Chagas disease and the nervous system. Scientific Publication No. 547, Washington
DC: 1994:329.
64. Hayes RJ, Schofield CJ. Estimacin de las tasas de incidencia de infecciones y
parasitosis crncas a partir de la prevalencia: la enfermedad de Chagas en America
Latina. Bol Of Sanit Panam 1990; 108:308316.
65. Moncayo A. Chagas disease. Tropical Disease Research. Progress 199192. Elev-
enth Programme Report of the UNDP/World Bank/WHO Special Programme for
Research and Training in Tropical Diseases (TDR). Geneva: World Health Organi-
zation, 1993:6775.
66. World Bank. The World Bank Development Report 1993. Investing in Health. World
development indicators. Washington: University Press, 1993.
67. Wendel S, Brener Z, Camargo M, Rassi A, eds. Chagas disease (American trypano-
somiasis): its impact on transfusion and clinical medicine. ISBT, Brasil 92, SBHH,
1992.
68. Miles MA. Further enzymic characters of Trypanosoma cruzi and their evaluation
for strain identification. Trans Roy Sci Trop Med Hyg 1980; 74:221237.
69. Miles MA. The epidemiology of South American trypanosomiasisbiochemical
and immunological approaches and their relevance to control. Trans Roy Soc Trop
Med Hyg 1983; 77:523.
70. Morel C, Chiari E, Plessman-Camargo E, Mattei D, Romanha AJ, Simpson L.
Strains and clones of Trypanosoma cruzi can be characterized by pattern of
restriction endonuclease products of kinetoplast DNA minicircles. Proc Natl Acad
Sci USA 1980; 77:68106814.
71. Avila H, Gonalves AM, Nehme NC, Morel CM, Simpson L. Schizodeme analysis
of Trypanosoma cruzi stocks from South and Central America by analysis of
PCR-amplified minicircle variable region sequences. Mol Biochem Parasitol 1990;
42:175188.
72. PAHO-OPS. III Reunin de la Comisin Intergubernamental para la Eliminacin
del Triatoma infestans y la interrupcin de la tripanosomiais Americana Transfu-
sional. OPS/HPC/HCT/9437:121, Montevideo, Uruguay.
73. Nishioka AS, Ferreira MS, Rocha A, Burgarelli MKN, Silva AM, Duarte MIS,
Schmitt FC. Reactivation of Chagas disease successfully treated with benznidazole
in a patient with acquired immunodeficiency syndrome. Mem Inst Oswaldo Cruz
1993; 88:493496.
74. Chocair PR, Sabbaga E, Amato Net V, Shiroma M, Goes GM. Transplante de rim:
nova modadlidade de transmisso da doena de Chagas. Rev Med Trop S Paulo
1981; 23:282282.
392 Wendel
75. Ferraz AS, Figueiredo JFC. Transmission of Chagas disease through transplanted
kidney: occurrence of the acute form of the disease in two recipients from the same
donor. Rev Inst Med Trop S Paulo 1993; 35:461463.
76. Dias JCP, Brener Z. Chagas disease and blood transfusion. Mem Inst Oswaldo Cruz
1984; 79:139147.
77. Schmuis GA. Trypanosoma cruzi, the etiologic agent of Chagas disease: status in
the blood supply in endemic and non endemic countries. Transfusion 1991; 31:547
557.
78. Wendel S, Gonzaga AL. Chagas disease and blood transfusion: a new world
problem? Vox Sang 1993; 64:112.
79. Schmunis GA. Tripanosomase Americana: seu impacto nas Amricas e perspectivas
de eliminaco. In: Dias JCP, Coura JR, eds. Clinica e teraputica da Doena de
Chagas. Rio de Janeiro: Edotora Fiocruz, 1997:411427.
80. Wendel S. Doena de Chagas Transfusional. In: Dias JCP, Coura JR, eds. Clinica e
teraputica da Doena de Chagas. Rio de Janeiro: Edotora Fiocruz, 1997:411427.
81. Wendel S, Neitzert E, Lopes VHG, Batemarchi MV. Evaluation of different screen-
ing tests for T. cruzi antibodies and correlation with immunoblotting technique. Rev
Paul Med 1992; 110: Abstract TTD17.
82. Trujillo FC, Kasten FL, Gutirrez MMS, Gutirrez RH. Prevalencia de infeccin a
Trypanosoma cruzi en donadores de sangre en el estado de Jalisco, Mexico. Rev
Soc Bras Med Trop 1993; 26:8992.
83. Ache A. Prevalencia de infeccin humana por Trypanosoma cruzi en Bancos de
Sangre en Venezuela. Rev Inst Med Trop S Paulo 1993; 35:443448.
84. Kerndt PR, Waskin HA, Kirhhoff LV, Steurer F, Waterman SH, Nelson JM, Gellert
GA, Shulman IA. Prevalence of antibody to Trypanosoma cruzi among blood
donors in Los Angeles, California. Transfusion 1991; 31:814818.
85. Pan AA, Brashear J, Schur JD, Winkler MA, Hall H, Shih J, Decker R. Prevalence
study and confirmation of seropositive antibodies to Trypanosoma cruzi (Chagas
disease) in Hispanic and Non-Hispanic blood donors in the United States. Rev Paul
Med 1992; 110(5) (suppl.): Abstract TTD-18.
86. Read EJ, Leiby DA, Dodd RY. Seroprevalence of Trypanosoma cruzi (T. cruzi) in
blood donors with and without risk infection. Blood 1994; 84 (suppl 1): Abstract
1853.
87. Winkler MA, Brashear RJ, Hall HJ, Schur JD, Pan AA. Detection of antibodies
to Trypanosoma cruzi among blood donors in the southwestern and western
United States. IIEvaluation of a supplemental enzyme immunoassay and radio-
immuneprecipitation assay for confirmation of seroreactivity. Transfusion 1995;
35:219225.
88. Leiby DA, Read EJ, Lenes BA, Yund AJ, Stumpf RJ, Kirchhoff LV, Dodd RY.
89. Amato Neto V, Leonhardt H, Souza HBWT. Liofilizao do plasma: Medida capaz
de evitar a transmisso da Doena de Chagas em Bancos de Sangue. Rev Inst Med
trop So Paulo 1966; 8:122124.
90. Sullivan TS. Viability of Trypanosoma cruzi in citrated blood stored at room
temperature. J Parasitol 1944; 30:200.
91. Weinman D, MacAlister J. Prolonged storage of human pathogenic protozoa with
conservation of virulence. Am J Hyg 1947; 45:102121.
92. Filardi LS, Brener Z. Cryopreservation of Trypanosoma cruzi bloodstream forms.

J Protozool 1975; 22:398401.
93. Schlemper BR Jr. Estudos experimentais de quimioproflaxia de transmisso da
doena de Chagas por transfuso sangnea. Rev Patol Trop 1978; 7:55111.
94. Cerisola JA, Rabinovich A, Alvarez M, Di Corleho CA, Pruneda J. Enfermedad de
Chagas y la transfusion de sangre. Bol Of Sanit Panam 1972; 63:203221.
95. Lorca M, Lorca J, Child R, Atias A, Conales M, et al. Prevalencia de la infeccin
por Trypanosoma cruzi en pacientes politransfundidos. Rev Med Chile 1988;
116:112116.
96. Villalba R, Forns G, Alvarez MA, Romn J, Rubio V, Fernandez M, Garcia JM,
Vials M, Torres A. Acute Chagas disease in a recipient of a bone marrow transplant
in Spain: case report. Clin Infect Dis 1992; 14:594595.
97. World Health Organization. TDR News 1999; 59:10.
98. Wendel S. Transfusion-transmitted Chagas disease. Curr Opin Hematol 1998;
5:406411.
99. Kirchhoff LV, Neva FA. Chagas disease in Latin American immigrants. JAMA
1985; 254:30583060.
100. Kirchhoff LV. Trypanosoma cruzi: a new threat to our blood transfusion supply
(editorial). Ann Intern Med 1989; 111:773775.
101. Kirchhoff LV. American trypanosomiasis (Chagas disease)a tropical disease now
in the United States. N Engl J Med 1993; 329:639644.
102. Milei J, Mautner B, Storino R, Sanchez J, Ferrans VS. Does Chagas disease exist
as an undiagnosed form of cardiomyopathy in the United States? (editorial). Am
Heart J 1992; 123:17321735.
103. Wendel S. Chagas disease: an old entity in new places (editorial). Int J Art Organs
1993; 16:117119.
104. Schmunis GA, Zicker F, Pinheiro F, Brandling-Bennett D. Risk for transfusion-
transmitted infectious diseases in Central and South America. Emerg Inf Dis 1998;
4:511.
105. Coura JR, Nogueira ES, Silva JR. Indices de transmisso da doena de Chagas por
transfuso de sangue de doadores na fase crnica da doena. O Hospital 1966;
69:115122.
106. Rohwedder RW. Infeccin chagsica en doadores de sangre y las probabilidades de
transmitirla por medio de la transfusion. Bol Chile Parasitol 1969; 24:8893.
107. Zuna H, Recacoechea M, Bermudez H, Romero A, Castedo J. Transmissin de la
enfermedad de Chagas por via transfusional en Santa Cruz de la Sierra, Bolivia. Bol
Inf CENENTROP 1979; 5:4956.
108. Vamvakas EC, Taswell HF. Long-term survival after blood transfusion. Transfusion
1994; 34:471477.
109. Andrade ALLS, Martelli CMT, Loquetti AO, Oliveira OS, Silva AS, Zicker F.
Triagem sorolgica para o Trypanosmoa cruzi entre doadores de sangue do Brasil
central. Bol Of Sanit Panam 1992; 113:1927.
110. World Health Organization. Parasite antigens. Bull WHO 1975; 52:237249.
111. World Health Organization. (WHO) reportNewsletter 1982; 18:7.
112. Cura E, Wendel S, eds. Manual de Procedimientos de Control de Calidad para los
394 Wendel
Laboratorios de Serologia de los Bancos de Sangre. PAHO/HPC/HPT/94.21, Wash-

ington DC, 1994.
113. Brazilian Ministery of Health, Portaria No. 1376, 1993.
114. Takei K. Estudo da eficincia relativa dos diferentes testes serolgicos utilizados no
diagnstico da doena de Chagas. Resultados observados na anlise de 10.181
soros; doctoral thesis. So Paulo Departamento de Microbiologia e Immunologia do
Instituto de Cincias Biomdicas da Universidade de So Paulo, 1982.
115. Lackritz, EM, Satten GA, Aberle-Grasse J et al. Estimated risk of transmission of
the human immunodeficiency virus by screened blood in the United States. N Engl
J Med 1005; 333:17211725.
116. Wendel S, Siqueira RV, Lopes VHG, Batemarchi MV, Neitzert E, Antunes MF,
Mejias S. Assessing the performance of Trypanosoma cruzi antibody detection
currently in use in Brazilian blood centres: a few answers from a large prevalence
study. Rev Paul Med 1992; 110(5):16.
117. Souza HM. Chagas Infection Transmission Control: Situation of Transfusional
Transmission in Brazil and Other Countries of Latin America. Simsio Internacio-
nal Sobre Avanos do Conhecimento da Doena de Chagas 90 Anos aps sua
Descoberta. Rio de Janeiro, April 1999.
118. Gontijo ECD. Doena de Chagas Transfusional na Regio Metropolitana de Belo
Horizonte: aspectos clnico-epidemiolgicos e a questo institucional. Thesis. Belo
Horizonte, Universidade Federal de Minas Gerais, 1989.
119. Wendel S. Blood banking preventive approaches for Chagas disease. Mem Inst
Oswaldo Cruz 1993; 88(suppl):5960.
120. Zicker F, Martelli CMT, Andrade ALLS, Silva SA. Trends of T. cruzi infection based
on data from blood bank screening. Rev Inst Med Trop S Paulo 1990; 39:132137.
121. Wendel S, Biagini S. Absence of serological surrogate markers for Trypanosoma
cruzi infected blood donors. Vox Sang 1995; 69:4449.
122. Appleman MD, Shulman IA, Saxena S, Kirchhoff LV. Use of a questionnaire to
identify potential blood doors at risk for infection with Trypanosoma cruzi. Trans-
fusion 1993; 33:6164.
123. World Health Organization/Panamerican Health Organization. Iniciativa del Cono
Sur. VI Reunin Intergubernamental para la eliminacin de Triatoma infestans
y la interrupcn del la tripanosomiasis americana por transfusin. Documento
OPS/HPC/HCT 98/102, 1997.
124. Freitas JLP, Biancalana A, Amato Neto, Nussenzweig V, Barreto JG. Molstia de
Chagas em Bancos de Sangue na capital de S Paulo. O Hospital 1952; 41:99106.
125. Cover B, Gutteridge WE. A primary screen for drugs to prevent transmission of
Chagas disease during blood transfusion. Trans Roy Soc Trop Med Hyg 1982;
76:633635.
126. Docampo R, Moreno SNJ, Muniz RP, Cruz FS, Mason RP. Light-enhanced free
radical formation and trypanocidal action of gentian violet (crystal violet). Science
1983; 220:12921295.
127. Docampo R, Moreno SNJ. Free radicals metabolites in the mode of action of
chemotherapeutic agents and phagocytic cells on Trypanosoma cruzi. Ver Inf Dis
1984; 6:223238.
128. Docampo R, Moreno SNJ, Cruz F. Enhancement of the cytotoxicity of crystal violet
against Trypanosoma cruzi in the blood by ascorbate. Mol Biochem parasitol 1988;
27:241248.
129. Souza HM. The present state of chemoprophylaxis in transfusional Chagas disease
(editorial). Rev Soc Bras Med Trop 1989; 22:13.
130. Ramirez LE, Lages-Silva E, Painetti GM, Rabello RMC, Bordin JO, Souza HM.
Prevention of transfusion-associated Chagas disease by sterilization of Trypano-
soma cruzi-infected blood with gentian violet, ascorbic acid, and light. Transfusion
1995; 35:226230.
131. Amato Neto V, Mellone O. Estudo sobre a efccia da Violeta de Genciana na
proflaxia da transmisso da doena de Chagas em bancos de sangue: investigao
em voluntrio, receptor de sangue de caso agudo, ao qual foi adicionado o corante.
O Hospital 1959; 55:343346.
132. Schlemper BR Jr. Estudos experimentais de Quimio-proflaxia de transmisso da
doena de Chagas por transfuso sangnea. Rev Pat Trop 1978; 7:55111.
133. Amato Neto V. Contribuio ao conhecimento da forma aguda da Doena de
ChagasTese: Faculdade de Medicina da Universidade de So Paulo, 1958.
134. Rezende JM, Zupelli W, Bafutto M. O problema da transmisso da doena de
Chagas por transfuso de sangue. Emprego da Violeta de Genciana como medida
profltica. Ver Goiana Med 1965; 11:3547.
136. Docampo R, Moreno SNJ. The metabolism and mode of action of gentian violet.
Drug Met Rev 1990; 22:161178.
137. Gottlieb P, Shen LG, Chimezie E, Bahng S, Keeney ME, Horowitz B, Bem-Hur E.
Inactivation of Trypanosoma cruzi trypomastigote forms in blood component
by photodynamic treatment with phthalocyanines. Photochem Photobiol 1995;
62:869874.
138. Wendel S, Takaoka DT, Jancs MM, Mariano M, Fragata Filho AA, Pereira-
Chioccola Lourenco AM. A possible role of leukocyte filters to prevent transfusion-
transmitted Trypanosoma cruzi. Vox Sang 1994; 67(suppl 2):22.
139. Souza HM, Bordin JO, Bordossy L, MacPhereson DW, Blajchman MA. Prevention
of transfusion-associated Chagas disease: efficacy of white cell-reduction filters in
removing Trypanosoma cruzi from infected blood. Transfusion 1995; 35:723726.
140. Wendel S, Takaoka DT, Jancs MM, Mariano M, Leite KM, Chioccola VLP,
Loreno AM, Fragata Filho AA. Is there a role for leukocyte depletion filters to
remove Trypanosoma cruzi infection from donor blood? Bol Soc Bras Hematol
Hemot 1995; 17:2531.
141. Moser DR, Kirchhoff LV, Donelson JE. Detection of Trypanosoma cruzi by DNA
amplification using the polymerase chain reaction. J Clin Microbiol 1989; 27:1477
1482.
142. Russomando G, Figueredo A, Almirn M, Sakamoto M, Morita K. Polymerase
chain reaction-based detection of Trypanosoma cruzi DNA in serum. J Clin
Microbiol 1992; 30:28642868.
143. Urbina J, Payares G, et al. Cure of short and long-term experimental Chagas disease
using D0870. Science 1996; 273:969971.
144. Centurion-Lara A, Barrett L, van Voohis WC. Quantitation of parasitemia by
competitive polymerase chain reaction amplification of parasite kDNA minicircles
396 Wendel
during chronic infection with Trypanosoma cruzi. J Infect Dis 1994; 170:1334
1338.
145. Lal RB, Ottesen EA. Phosphocoline epitopes on helminth and protozoal parasites
and their presence in the circulation of infected human patients. Trans Roy Soc Trop
Med Hug 1989; 83:652655.
146. Ferreira AW. Serological diagnosistests for Chagas disease serodiagnosis: a
review. In: Wendel S, Brener Z, Camaro ME, Rassi A, eds. Chagas Disease
(American Trypanosomiasis): Its Impact on Transfusion and Clinical Medicine. So
Paulo: ISBT, 1992: 179193.
147. Sasaki AT, Hoshino-Shimizu S, Nakamura PM, Vaz AJ, Camargo ED, Silva MD.
Sorodiagnstico da doena de Chagas: novo reagente para o teste de Hema-
glutinao indireta (THAIIAL). Rev Soc Bras Med Trop 1996; 29:137144.
148. Luquetti AO. Use of Trypanosoma cruzi defined proteins for diagnosismulticentre
trial, serological and technical aspects. Mem Inst Oswaldo Cruz 1990; 85:497505.
149. Zicker F, Smith PG, Luquetti AO, Oliveira OS. Mass screening for Trypanosoma
cruzi infections using the immunofluorescence, ELISA and hemagglutination tests
on serum samples and on blood eluates from filter paper. Bull WHO 1990;
68:465471.
150. Andrade ALSS, Martelli CMT, Luquetti AO, Oliveira OS, Silva SA, Zicker F.
Serologic screening for Trypanosoma cruzi among blood donors in central Brazil.
Bull Pan Am Health Organ 1992; 26:157164.
151. Fife EH Jr, Mushel LH. Fluorescent antibody technique for serodiagnosis of
Trypanosoma cruzi infection. Proc Soc Exp Biol NY 1959; 101:540543.
152. Camargo ME, Souza SL. The use of filter paper blood smears in a practical
fluorescent test for American trypanosomiasis serodiagnosis. Rev Inst Med Trop S
Paulo 1966; 8:255258.
153. Velasquez L, Reyes L, Thors C, Miettinen A, Chinchella M, Linder E. Autoantibod-
ies give false positive reactions in the serodiagnosis of Trypanosoma cruzi. Trans
Roy Soc Trop Med Hyg 1993; 87:35.
154. Voller A, Draper C, Bidwell DE, Bartlett A. A micro-plate enzyme-linked im-
munosorbent assay (ELISA) for Chagas disease. Lancet 1975; 1:426429.
155. Wendel S, Lopes VHG, Batemarchi MV, Takaoka DT, Neitzert E. High levels of
false positive reaction for Trypanosoma cruzi antibodies tests used for blood donors
screening. Transfusion 1993; 33(suppl):41S.
156. Carvalho MR, Krieger MA, Almeida E, Oelemann W, Shikanai-Yasuda MA,
Ferreira AW, Pereira JB, Sez-Alquzar A, Dorlhiac-Llacer PE, Chamone DF,
Goldenberg S. Chagas disease diagnosis: evaluation of several tests in blood bank
screening. Transfusion 1993; 33;830834.
157. Ferreira AW, Belem ZR, Moura MEG, Camargo ME. Aspectos da padronizao de
testes sorolgicos para a Doena de Chagas: Um teste imunoenzimtico para a
triagem de doadores de sangue. Rev Inst Med Trop S Paulo 1991; 33:123128.
158. Brashear RJ, Winkler MA, Schur JD, Lee H, Burczak JD, Hall HJ, Pan AA.
Detection of antibodies to Trypanosoma cruzi among blood donors in the south-
western and western United States. I. Evaluation of the sensitivity and specificity of
an enzyme immunoassay for detecting antibodies to T. cruzi. Transfusion 1995;
35:213218.
159. Oelemann WMR, Teixeira MGM, Costa GCV et al. Evaluation of three commer-
cial enzyme-linked imunosorbent assays for diagnosis of Chagas disease. J Clin
Microbiol 1998; 36:24232427.
160. Houghton RL, Benson D, Skeiky Y, Sleath P, Lodes M, Badar R, Krettli A, Reed
SG. Multi-epitope peptide elisa for detection of serum antibodies to Trypanosoma
cruzi in patients with treated and untreated Chagas disease. Transfusion 1996;
36(suppl):137.
161. Hoffman A, Jaczko B, Jones J, Tursi N, Lee SR. Evaluation of a prototype screening
ELISA and supplemental analysis for the serological identification of Chagas
disease. Transfusion 1997; 37(suppl):181.
162. Umezawa ES, Bastos SF, Camargo ME, Yamauchi LM, Santos MR, Gonzales A,
Zingales B, Levin MJ, Souza O, Rangel-Ado R, Franco da Silveira J. Evaluation
of recombinant antigens for serodiagnosis of Chagas disease in South and Central
America. J Clin Microbiol 1999; 37:15541560.
163. Lanar DE, Manning JE. Major surface proteins and antigens on the different in vivo
and in vitro forms of Trypanosoma cruzi. Mol Biochem Paras 1984; 11:119131.
164. Rosfjord EC, Mikhael KS, Rowland EC, Powell MR. Analysis of antibody cross-
reactivity in experimental american trypanosomiasis. J Parasitol 1990; 76:698702.
165. Reiche EMV, Cavazzane M Jr., Okamura H, Tagata EC, Jankevicius SI, Jankevicius
JV. Evaluation of the Western blot in the confirmatory serologic diagnosis of
Chagas disease. Am J Trop Med Hyg 1998; 59:750756.
166. Teixeira MGM, Borges-Pereira J, Neitzert E, Souza MLNX, Peralta JM. Develop-
ment and evaluation of an enzyme linked immunotransfer blot technique for
serodiagnosis of Chagas disease. Trop Med Parasitolo 1994; 45:308312.
167. Pastini AC, Iglesias SR, Carricate VC, Guerin ME, Sanchez DO, Frasch AC.
Immunoassay with recombinant Trapanosoma cruzi antigens potentially useful
for screening donated blood and diagnosing Chagas disease. Clin Chem 1994;
40:18931897.
168. Sabino EC, Salles N, Stoops E, Santos ML, Saez-Alquezar A. Evaluation of Chagas
Innolia assay as a confirmatory test for Chagas disease. Mem Inst Oswaldo Cruz
1997; 92(suppl 1):266.
169. Costa GCV, Teixeira MGM, Borges-Pereira J, Castro JAF, Coura JR, Vanderborght
BOM, Stroops E, Zrein M, Peralta JM, Oelemann WMR. A recombinant antigen
and peptide line immunoassay (LIA) as an alternative diagnostic test for Chagas
disease. Mem Inst Oswaldo Cruz 1997; 92(suppl 2):267.
170. Oelemann WMR, Vanderborght BOM, Verssimo da Costa GC, Teixeira MGM,
Borges-Pereira J, Castro JAF, Coura JR, Stoops E, Hulstaert F, Zrein M, Peralta JM.
A recombinant peptide antigen line immunoassay optimized for the confirmation of
Chagas disease. Transfusion 1999; 39:711717.
171. Kirchhoff LV, Gam AA, Gusmo RA, Goldsmith RS, Rezende JM, Rassi A.
Increased specificity of serodiagnosis of Chagas disease by detection of antibody
to the 72- and 90-kilodalton glycoproteins of Trypanosoma cruzi. J Infect Dis 1987;
155:561564.
172. Leiby DA, Wendel S, Takaoka DT, Fachini R, Oliveira LC, Tibbals MA. Confirma-
tory testing for Trypanosoma cruzi antibodies: validation of radioimmuneprecipita-
tion (RIPA) testing. Transfusion 1997; 37(suppl 95):5301.
398 Wendel
173. Rassi A, Luquetti AO. Therapy of Chagas disease. In: Wendel S, Breuer Z, Camargo
ME, Rassi A, eds. Chagas Disease (American Trypanosomiasis): Its Impact on
Transfusion and Clinical Medicine. So Paulo: ISBT, 1992:237247.
174. Andrade ALLS, Zicker F, Oliveira RM, Silva AS, Luquetti AO, Travassos LR,
Almeida IC, Andrad SS, Andrade JG, Martelli CMT. Randomised trial of efficacy
of benznidazole in treatment of early Trypanosoma cruzi infection. Lancet 1996;
348:14071412.
175. Dias JCP, Schofield CJ. Controle da transmisso transfusional da doena de Chagas
na Iniciativa do Cone Sul. Rev Soc Bras Med Trop 1998; 31:373383.
17
Tickborne Infections
Ritchard G. Cable
American Red Cross Blood ServicesConnecticut Region,
Farmington, Connecticut
Jonathan Trouern-Trend
Epidemiology and Surveillance Program, American Red Cross
ARCNET, Farmington, Connecticut
I. INTRODUCTION
Tickborne diseases are achieving increased recognition in the United States as a

public health problem. Lyme disease has been identified to have occurred in 45
states and has attracted increased public interest. Babesiosis, transmitted by the
same tick, has important transfusion implications. Newly identified diseases such
as ehrlichiosis raise new questions about the role of the tick in human disease.
Much of this interest has been accelerated by increases in animal populations
(deer in particular), as well as increased movement of the human population into
outdoor environments, which are conducive to tick bites.
Simultaneously, heightened public scrutiny of blood safety in the wake of
the AIDS epidemic has resulted in a raising of the bar regarding potential
threats. Thus, new regulatory and public pressures are likely to focus on tickborne
diseases and their relationship to blood safety. The most striking example of this
is the recall of nearly 700 units of blood donated by Army reservists who trained
at Fort Chaffee, Arkansas, during the summer of 1997. The measure was initiated
at the request of the U.S. Food and Drug Administration (FDA) because the
exposure of the trainees to tick bites on maneuvers was high and a large number
of exposed troops reported illnesses following the exposure. At least some of the
illnesses were due to tickborne diseases such as Rocky Mountain spotted fever.
More than one type of tickborne disease was reported, none of which is classically
399
400 Cable and Trouern-Trend
thought of as transfusion transmitted (1,1a). Previously, this kind of event would

not have been considered a threat to the blood supply. It is clear that an increased
understanding of tickborne disease and the possibility of transfusion transmission
will be required by transfusion medicine experts and blood banking organizations.
Initially this chapter will discuss important ticks, their geographic distribu-
tion, and life cycles. Then an overview of tickborne human diseases will be
presented, with emphasis on the diseases likely to be (or to become) of interest
in transfusion medicine. Finally, we will discuss babesiosis, the most important
transfusion-transmitted tickborne disease, as well as the possibility that several
other tickborne diseases may be transfusion transmitted.
II. BIOLOGY OF TICKS

Ticks are small parasitic arachnids in the order Acari, which they share with the
mites. Of approximately 850 species described, all are epidermal parasites
throughout their life cycle and feed on the blood of reptiles, birds, or mammals.
The ticks are divided taxonomically into two main families, the Ixodidae, or hard,
ticks and the Argasidae, or soft, ticks. The Ixodidae is represented by 13 genera
including the medically important genus Ixodes (2,3).
III. LIFE CYCLE OF TICKS

The life cycle of a tick includes four stages: egg, larvae, nymph, and adult.
Copulation usually occurs on the host. A blood meal is normally required for egg
production. After feeding, the engorged female drops from the host and lays
between 100 and 20,000 eggs, depending on species and environmental condi-
tions (4). Larval ticks hatch from the eggs, then begin to search for a host. Four
alternate life cycles can be found among ticks. First is the one-host tick. The cycle
of feeding and molting until maturity is reached on one host. This reduces the
vector potential. The second type of life cycle is the two-host tick. In this
situation, a larval tick hatches on the ground and finds a host, usually a small
mammal or bird. The larva feeds on the host, then molts and becomes a nymph.
After the nymph feeds, it drops to the ground and molts into the adult stage.
A three-host tick feeds on a separate host for each stage. Finally, a many-host tick
feeds intermittently and on multiple hosts during its life cycle (2,4).
IV. TICK- AND MITEBORNE DISEASE

Ticks are important to human and veterinary medicine. They cause disease and
act as vectors and reservoirs for many serious pathogens. Several factors make
Tickborne Infections 401
these organisms efficient vectors of disease. First, the ingestion of host blood
allows collection of pathogens in the body of the tick. If the tick subsequently
moves to another host, transmission may occur. The long period of attachment on
a host increases the likelihood of transmission. High reproductive potential
ensures high populations of ticks and increases possibility of disease transmis-
sion. Pathogens are transmitted between life stages in some ticks (e.g., between
the larva and the nymph), a phenomenon known as transstadial transmission.
Pathogens may also be passed to the next generation by transovarial transmission.
Both types of transmission are important in the maintenance of pathogens in tick
populations (2).
Mites belong to the same order as ticks and share many of the same
characteristics and parasitic lifestyles. Ticks are really giants in the mite family,
most mites being much smaller than the smallest ticks (5). Scrub typhus, which
is miteborne, is discussed here since it may be transfusion-transmitted.
A great many tickborne diseases have been described. A number of these
have the potential for transfusion transmission. The more important tickborne
diseases of interest in transfusion medicine are summarized in Table 1. The re-
mainder of the chapter will expand upon these diseases.
V. VIRAL TICKBORNE DISEASES

A large number of viral agents related to human diseases have been identified as
being transmitted by ticks. Since some of these pathogens pass transovarially,
even one-host ticks are capable of transmitting disease via the next generation.
However, because many tickborne viruses have short incubation periods, the
likelihood of transmission by a blood donor is small. Donors are deferred for
fever, which should prevent collection of viremic blood in many cases. Some
tickborne viruses (especially those with longer incubation periods or with many
asymptomatic infections) have been transmitted by transfusion or have that
potential. The current deferral for blood donors who have traveled to malaria-
endemic areas reduces the risk of transmitting many of these tickborne viruses in
North America and Europe, since the ranges of the involved tick vectors overlap
those of malaria (6).
A. Flaviviruses
These enveloped single-stranded RNA viruses replicate in the cytoplasm. The
Flavivirus family contains nonarborviruses such as hepatitis C and hepatitis G
(7,8). Only one of the tickborne flaviviruses, the central European subtype of
tickborne encephalitis, has been documented as causing transfusion-transmitted
disease (9). Theoretically, others could do so. Two cases of tickborne viral en-
Table 1 Tickborne Diseases of Interest in Transfusion Medicine
Transfusion-
Agent Disease transmitted? Country Ref.
Viruses
Tickborne encepha- Kumlinge disease, Yes Finland 8,9
litis virus tickborne
encephalitis
Deer tick virus None known Possible United States 11
Colorado tick fever Colorado tick fever Yes United States 12,13
virus
Bacteria
Borrelia burgdorferii Lyme disease Possible United States 1719
Borrelia recurrentis Relapsing fever Possible Africa, China 2427
Rickettsia
Rickettsia rickettsi Rocky Mountain Yes United States 28
spotted fever
Orienta tsutsugamushi Scrub typhus Possible Asia, 30
Australia
Ehrlichia chaffiensis Human monocytic Possible United States 31
ehrlichiosis
Ehrlichia spp. Human granulo- Possible United States 31,32
cytic ehrlichiosis
Protozoa
Babesia microti Babesiosis Yes United States 3336,
6368
WA1 Babesiosis-like Yes United States 37,38
MO1 Babesiosis-like Possible United States 39
TW1 None known Possible Taiwan 40
cephalitis (Kumlinge disease) were reported from a single donor from Kumlinge,
Finland. The disease, endemic on this isolated island, was reportedly transmitted
by a blood donation a few hours before the donor became ill. The transfusion
illness was indistinguishable from the tickborne illness. No details on the blood
components storage or transfusion were published, and it is not clear how the
disease was recognized. The recipients recovered.
Another flavivirus, deer tick virus, was described in 1997. It has been
found in Ixodes scapularis, a vector of several human diseases, including Lyme
disease and babesiosis (10,11). This virus has not been demonstrated to cause
disease in humans. However, its close relationship to other viruses in the
tickborne encephalitis complex warrants some concern. In a serosurvey of
ticks from New England, a small percentage was found to be harboring this virus
(11). Further studies will be needed to determine whether this virus is a public
health problem.
B. Retroviruses
This family contains enveloped double-stranded RNA viruses that replicate in the
cytoplasm. Ticks spread several of these viruses to humans. Most are responsible
for only a few cases of disease, although Colorado tick fever virus causes several
hundred cases a year in North America.
Only Colorado tick fever virus has been demonstrated to cause transfusion-
transmitted disease. Cases of Colorado tick fever occur in the Rocky Mountains
and Pacific slope of the western United States and Canada. The vector is the wood
tick, Dermacentor andersoni. Up to 15% of humans at high risk (forest rangers,
loggers, back-country hikers) have antibodies to the virus. Most cases occur
between April and July, mirroring tick activity. Several hundred cases are reported
each year. Since reporting requirements vary by region, cases are thought to be
significantly underestimated. Cases have been imported to other parts of the
United States by people traveling in the endemic areas. Incubation is usually 35
days after a tick bite, sometimes up to 14 days. No prodromal manifestations
occur. Onset of symptoms is rapid. Fever, headache, and myalgia are the most
common symptoms. Encephalitis, thrombocytopenia, disseminated intravascular
coagulation (DIC), and a hemorrhagic state resembling Dengue hemorrhagic
fever have occurred. Many individuals, especially adults over 30 years, experi-
ence a prolonged convalescence (12).
Colorado tick fever was transmitted in Montana from a healthy blood donor
who developed fever 18 hours after blood donation and 4 days after removing a
tick (13). The blood was transfused during surgery to an 82-year-old with bowel
obstruction, who developed a prolonged febrile illness lasting 23 days. Both
donor and recipient blood were demonstrated to have Colorado tick fever virus
by mouse inoculation, confirming the diagnosis. This case illustrates again the
importance of reporting postdonation illness in order to recognize these rare
disease transmissions.
The virus can be isolated from plasma for up to 7 days after the onset of
symptoms and up to 20 weeks in the red cell fraction. An infection usually confers
immunity; however, rare cases of reinfection have been reported.
VI. BACTERIAL TICKBORNE DISEASES

A. Lyme Disease (Borrelia burgdorferi and Others)
Lyme disease is an important tickborne disease, with worldwide distribution
caused by a number of spirochetes in the genus Borrelia. Most cases are found
in the United States and western Europe. This disease accounts for more than 90%
of arthropodborne diseases reported in North America each year. In the United
States, endemic foci exist along the Atlantic seaboard from Massachusetts to
Maryland, in the upper Midwest (Wisconsin and Minnesota), and in the West
(California and Oregon). Cases have been reported from 45 states as well as the
Canadian provinces of British Columbia and Ontario (14). Elsewhere, cases have
been seen in the former Soviet Union, China, Japan, and Australia. The main
vectors are four host Ixodid ticks: I. scapularis (also called I. dammini) in the
Eastern and midwestern United States, I. pacificus in the western United States,
I. ricinus in Europe, and I. persulcatus in Asia.
In North America, the spirochete responsible for Lyme disease circulates
between the white-footed mouse (Peromyscus leucopus) and the white-tailed deer
(Odocoileus virginianus) via I. scapularis. The transmission cycle of the spiro-
chete starts with an infected mouse. Larval ticks hatch in the spring or early
summer and search out their first host, usually a mouse. After feeding for about
2 days, the tick drops off. The tick does not feed again until the next spring, after
it has molted to the nymphal stage. Usually the nymph feeds once, again usually
on a mouse, then molts to an adult. If the preferred host cannot be found, the
nymph will feed on other mammals, including humans. Nymphal ticks, infected
from their first blood meal, transmit the spirochete to humans. The spirochete
enters the hosts body at the site of the tick bite, then travels via the blood and
lymphatics to other organs of the body, where pathogenesis occurs. Adult ticks
feed on deer, then lay eggs. This association between Lyme disease and deer is
responsible for the dramatic increase in cases in this century. The abundance of
the vector correlates with the number of deer in the area.
Lyme disease can be divided into three stages. In the first stage there is a
localized skin lesion at the bite site called erythema migrans (EM). This lesion
can progress to a bulls-eyetype lesion with the inner part clear and the edges
red. About 60% of cases have this sign. Other symptoms of the disease can be
fever, malaise, fatigue, myalgia, and migratory arthralgias. A wide range of
neurological symptoms have been reported, including facial palsy, aseptic men-
ingitis, chorea, myelitis, and encephalitis. Cardiac symptoms may occur a few
weeks to months after EM, including atrioventricular block. Rarely acute myo-
pericarditis or cardiomegaly have been reported. A chronic infection may lead to
arthritis in the large joints, polyneuropathy, or leukoencephalitis.
When Borrelia burgdorferi was first described, it was thought to be the only
agent of Lyme disease (15). Research in the last decade has indicated that at least
eight species of Borrelia have been found in the Ixodes vectors, four of which
have been demonstrated to cause human disease. All four species are known to
cause EM. B. burgdorferi is present in North America and Europe but is absent
from Russia and Asia. This species is most often associated with polyarthritis.
In Eurasia, Borrelia garinii is associated with neurological symptoms; it has
been isolated from Japan to western Europe. Borrelia afzelii, another Eurasian
Borrelia, found from Japan and China to western Europe, is associated with the
chronic skin condition acrodermatitis chronica atrophicans. VS116 occurs in
Europe and has only been associated with EM (15a).
Though antibodies to B. burgdorferi are present in a significant fraction of
blood donors in endemic areas, no case of transfusion-transmitted Lyme disease
has been reported. B. burgdorferi has been cultured from the blood of patients
with early Lyme disease (16), which has led to the concern that Lyme disease may
be transmissible by transfusion (17). We have shown (18) that B. burgdorferi can
survive in stored blood products under frozen, refrigerated, or 2024C storage
conditions for the duration of the storage period of fresh frozen plasma (FFP), red
cells, and platelets, respectively. Because the characteristic EM rash is thought to
be caused by skin growth of spirochetes at the site of the tick bite, it is likely that
intravenously injected B. burgdorferi would not cause EM. Since the remainder
of the symptoms of Lyme disease are nonspecific, it is likely that, if transfusion-
transmitted Lyme disease exists, it would not be easily recognized.
Despite active surveillance for transfusion-transmitted Lyme disease, no
cases have been identified. However, we reported a case in which B. burgdorferi
DNA was detected in red cells and plasma from a donor who discovered EM the
day after blood donation and notified the blood center (19). He was confirmed
serologically to have Lyme disease. The involved blood products were cultured
but were negative for B. burgdorferi. Had this donor not called the blood center,
and had his blood products been transfused, it is not clear whether the recipients
would have been infected. It is also not clear whether the PCR-positive blood
units contained viable organisms.
In a case that illustrates the difficulty of transfusion transmission of Lyme
disease, we identified an implicated donor in a case of posttransfusion babesiosis.
Upon interview, the donor gave a history of having developed EM several days
following his blood donation and had been diagnosed and treated for Lyme
disease with a characteristic serial serological pattern for B. burgdorferi. The
recipient, who had a renal transplant and was on an immunosuppressive agent,
developed a typical serological response to babesiosis but did not seroconvert for
B. burgdorferi or become ill with Lyme disease (20). This was the case despite
the likelihood that both agents were transmitted to the donor by the bite of a
single, doubly infected, tick (21) 13 weeks before the blood donation. This case
suggests that the spirochetemic phase of B. burgdorferi after an infected tick bite
must be very short.
In an effort to estimate the risk of Lyme disease from transfusion, we
conducted serological follow-up on 155 open heart surgery patients who re-
ceived a total of 601 red cells and 371 platelet concentrates. None of the re-
cipients seroconverted for B. burgdorferi (95% CI, 00.5% for red cells and
00.8% for platelet concentrates) (22). At the same time, the study detected
one case of seroconversion for babesiosis and placed the per unit risk of exposure
to babesiosis from a red cell transfusion in Connecticut at 0.17% (95% CI,
0.0040.9%).
B. Relapsing Fever (Borrelia recurrentis and Others)

This disease, caused by a spirochete, is characterized by periods of fever lasting
29 days followed by afebrile periods lasting 24 days. The number of relapses
ranges from 1 to 10 or more. The initial febrile period may be accompanied by
transitory petechial rashes. Mortality ranges from 2 to 5% in untreated cases.
Antibiotics provide effective treatment.
A number of species of Borrelia are responsible for endemic or tickborne
relapsing fever (23). This disease is present in small endemic foci on every
continent except Australia. Rodents are thought to be the natural reservoir for this
disease. Argasid ticks in the genus Ornithodoros are the vector. Since the
spirochete can be transmitted transovarially, ticks can serve as long-term reser-
voirs, passing infection from generation to generation. Epidemics have occurred
in western Canada and the United States. Epidemics are usually associated with
sleeping in cabins infested with rodents and the Argasid tick vector. In 1973, an
outbreak in Arizona associated with cabins near the Grand Canyon resulted in 62
cases (24,25).
The spirochete is present in the blood only during the febrile stage,
particularly the first one. Cases of transfusion-transmitted relapsing fever have
been reported from both Africa and China (26,27). However, current practices of
excluding febrile donors and the low incidence of this disease should make this
a negligible transfusion risk.
VII. RICKETTSIAE
The rickettsia are aerobic gram-negative bacilli that are obligate intracellular par-
asites. Several genera (Rickettsia, Ehrlichia, Coxiella, and Orientia) are associ-
ated with human disease. Humans are most often accidental hosts. Infection
usually circulates in nature between animal hosts and arthropod vectors (ticks,
mites, fleas, lice). In some infections an asymptomatic rickettsemic stage may
precede symptoms. It is most likely during this period that rickettsial disease may
be transmitted by donated blood.
A. Rocky Mountain Spotted Fever (Rickettsia rickettsii)

This infection is found throughout the United States. Infections have also been
reported from Central and South America. The illness is characterized by a fever
of sudden onset, persisting for 23 weeks in untreated individuals. Fever is

accompanied by myalgia, fatigue, severe headache, chills, and conjunctival
injection. In about half the cases a maculopapular rash appears about the third
day. Hemorrhage is a common complication. Case fatality rates in untreated cases
are 1325%. In recent years, with treatment, case fatality rates in the United States
have been 35%. The disease is maintained in ticks by transovarian and trans-
stadial passage. In the eastern United States Ixodes variabilis is the primary
vector; in the western United States it is Dermacentor andersoni; in the south-
western United States it is Amblyomma americanum. In Central and South
America the primary vector is Amblyomma cajennense. The incubation period is
214 days.
Rocky Mountain spotted fever was transmitted in 1977 from a healthy
blood donor in Baltimore (28). Three days after his blood donation he developed
chills, headache, and a typical rash, and 7 days after donation he died. The in-
fected red cell was refrigerated for 9 days before transfusion. The recipient
developed fever 6 days after transfusion and was treated with antibiotics not
indicated for Rocky Mountain spotted fever. Four days later, after the donors
diagnosis was reported to the recipients physician, the recipient was treated with
chloramphenicol for Rocky Mountain spotted fever and eventually recovered.
Diagnosis in both donor and recipient was confirmed serologically and bacterio-
logically. The report comments on the importance of postdonation disease report-
ing by the donor to the blood center and then to the recipient hospital to allow
appropriate and timely treatment. This case of rickettsial transmission also
reinforces the possibility of transfusion transmission of other rickettsial diseases,
such as ehrlichiosis.
B. Scrub Typhus (Orientia tsutsugamushi)

This miteborne disease, also known as tsutsugamushi disease, is characterized
by a primary lesion, that develops at the site of a mite bite. Symptoms appear
after an incubation period of 621 days. Symptoms include headache, mal-
aise, fever, regional lymphadenitis, and a macular or maculopapular rash on
the trunk and the extremities. The case fatality rate varies between 1 and
60%. The reservoir and the vector for this disease are mites in the family
Trombiculidae. The disease is endemic in southern and southeastern Asia from
India to Indonesia, Japan, Korea, parts of China, New Guinea, and northern
Australia (29).
This rickettsia has recently been shown to persist in packed red cells, so it
appears there is potential for transmission (30). Large areas where the disease is
endemic are also in the malarial deferral zone, so travelers to these areas would
be deferred from donating blood in North America or Europe.
C. Human Monocytic Ehrlichiosis (Ehrlichia chafeensis)

This is an acute febrile illness found in the United States, primarily in the
southeastern and south central regions. Infections have been confirmed from 30
states. The pathogen infects mononuclear lymphocytes causing leukopenia and
thrombocytopenia. Sometimes referred to as Rocky Mountain spotted fever
without the spots, the disease usually presents as a fever with headache and
malaise. Nausea, vomiting, and myalgia are often present. Hepatitis and elevation
of serum hepatic aminotransferases are often noted. The case fatality rate is 25%,
with death resulting from renal failure or respiratory insufficiency. The vector for
this disease is the Lone Star tick, Amblyomma americanum. White-tailed deer are
thought to be a major reservoir for the disease. Since subclinical infections have
been reported, there is some possibility of transfusion transmission, although no
cases have been reported (31).
D. Human Granulocytic Ehrlichiosis (Ehrlichia spp.)

This disease has a clinical picture similar to human monocytic ehrlichiosis, with
a higher case fatality rate (710%). The vector of this disease is Ixodes scapularis,
the same vector as in Lyme disease and babesiosis. Ticks harboring multiple
pathogens have been recorded and antibodies to multiple tickborne diseases
reported (32). Most cases have come from the eastern and upper midwestern
United States. Other cases have been seen in Arkansas, California, and Florida.
White-tailed deer and rodents are suspected to be the reservoirs.
A single case of probable transfusion transmission of human granulocytic
ehrlichiosis (HGE) has been reported from Wisconsin in 1998 (32a). The recipient
was transfused with 2 units of red cells. Nine days later, the recipient was
hospitalized with fever, rigors, nausea, and vomiting. A pretransfusion blood
sample was negative for HGE by PCR. Testing 2 days posthospitalization was
positive for HGE by PCR and antibody. In addition, neutrophilic morulae could
be visualized by microscopy. The implicated donor had a history of Lyme Disease
and reported tick bites two months prior to blood donation. The donor tested
positive for HGE antibodies, however a test on the implicated unit was PCR
negative. The implicated unit had been stored for 30 days prior to transfusion.
This case illustrates the potential for asymptomatic donor transmission of HGE.
VIII. PROTOZOAL TICKBORNE DISEASES

A. Babesiosis
Babesia are intracellular sporozoan parasites that can resemble malarial parasites
both morphologically and clinically. More than 70 species have been described,
and some are serious veterinary pathogens. At least five species of Babesia or
Babesia-like organisms have been shown to infect humans, four with demonstra-
ble disease.
1. Babesia microti
Primarily a parasite of small rodents, particularly the white-footed deer mouse
(33), this species has caused several hundred infections in the United States,
primarily in the Northeast and upper Midwest (34). A significant number of these
cases have been asymptomatic. The vector Ixodes scapularis (Ixodes damini)
ranges from New Hampshire to Maryland, west to Wisconsin and Minnesota (14).
The ecology of B. microti infections in humans closely mirrors that of Lyme
disease and human granulocytic ehrlichiosis (HGE). Nymphs of the vector
I. scapularis bite humans and transmit the parasite. The ticks are infected as
larvae and acquire the infection from parasitic mice or voles. Deer, important to
the life cycle of I. scapularis and transmission of Lyme disease, are not infected
with B. microti and are not reservoirs for the parasite.
Infection follows a bite of an infected nymphal tick. This stage feeds
from May through September. The Babesia organism penetrates an erythrocyte.
The trophozoite multiplies by binary fission and forms tetrads. Lysis of the
erythrocyte releases the merozoites into the blood where they can reinfect other
cells, maintaining the infection or infecting a feeding tick. Transovarial passage
occurs in ticks allowing maintenance of the infection without a host (35).
Posttransfusion babesiosis is discussed below.
2. Babesia divergens and Babesia bovis

These two bovine Babesia species cause babesiosis in Europe and are more
often associated with fatal outcome than North American species of Babesia.
Babesia divergens is the most common Babesia found in European cattle.
The vector is Ixodes ricinus, the European vector of Lyme disease. Most patients
were splenectomized, a major risk factor for all forms of babesiosis. The mortality
rate is higher than 50%. Reported cases presented with fulminant, febrile,
hemolytic disease (34,36). No transfusion-associated cases have been reported for
this species.
3. WA1 (WAshington Case 1)

This Babesia-like organism has been isolated from human cases in northern
California, Oregon, and Washington. This organism is more closely related to
Babesia gibsoni, a parasite of dogs and coyotes, than B. microti. This species has
been implicated in transfusion transmission (see below). Seroprevalence rates up
to 16% have been reported from northern California. Infections have occurred in
both splenectomized persons and those with intact spleens (37,38).
4. MO1 (MissOuri Case 1)

This Babesia-like parasite, found to cause disease in a patient from Missouri, is
most closely related to the bovine parasite B. divergens (39). No transfusion cases
have been reported yet, although it is likely cases will appear.
5. TW1 (TaiWan Case 1)

A recently reported Babesia species similar to B. microti was isolated from an
asymptomatic Taiwanese patient (40). Local rodents also harbored the parasite.
This species has not yet been reported to cause any human disease.
B. Clinical Babesiosis (B. microti) (Table 2)

The incubation period is usually 13 weeks after a tick bite, sometimes up to 6
weeks (41). Transfusion-associated infections appear to have a longer incubation
period, from 2 to 8 weeks (see Table 3). Symptoms begin gradually and are
nonspecific. Common symptoms include malaise, fatigue, anorexia, arthralgias,
nausea, vomiting, abdominal pain, and dark urine. Also common are malaria-like
symptoms of fever, chills, myalgia, and sweating. The fever can be either
sustained or intermittent, reaching temperatures of 40C. Hemolytic anemia can
produce jaundice and renal failure. Disseminated intravascular coagulation can
be a serious complication.
B. microti infection differs from European Babesia infection in several
ways. First, European babesiosis is a much more serious, fulminant disease. It is
more often associated with asplenic patients and with fatal outcomes. Most
clinical B. microti infections, though serious, are not fatal. Only 30% of affected
patients in the United States have been asplenic. Of the spleen-intact patients with
clinical disease, most were over 50 years old, indicating that age is a factor in the
severity of disease (42).
The percentages of erythrocytes parasitized in clinical cases range from 10
to 85%. Antibiotic therapy (currently clindamycin and quinine) (43) is usually
curative: in more severe cases exchange transfusion has been utilized to reduce
the level of parasitemia (44).
Asymptomatic infections appear to be common and have been impli-
cated in some transfusion cases. Seroprevalence rates for B. microti have been
recorded as high as 6.9% in individuals with high risk of tick exposure on Shelter
Island, New York (45). Concurrent infection with B. burgdorferi has been
reported. In one study, 54% of patients with babesiosis had antibodies for
B. burgdorferi (46).
Table 2 Known Cases of Transfusion Babesiosis (B. microti) as of July 1999
Blood
Case Age, gender, year Evidencea component Patient factors Donor infection Ref.
1 70 yo M, 1979 Probable Platelets ITP; splenectomy after transfusion Nantucket, MA 44

2 23 yo F, 1980 Probable Frozen RBC Splenectomy; thalassemia Fire Island, NY 63
Tickborne Infections
3 8 Weeks F, 1980 Definite RBC Premature birth Fire Island, NY 51

4 79 yo F, 1982 Definite RBC Perforated diverticulum; colon Cape Cod, MA 64,65
resection
5 72 yo M, 1985 Probable RBC Myelodysplasia Marthas Vineyard, MA 66
6 46 yo M, 1989 Definite RBC Splenectomy Old Lyme, CT 67
7 57 yo M, 1989 Possible RBC Esophagectomy for adenocarcinoma Unknown 68
8 65 yo M, 1990 Probable RBC Splenectomy Cape Cod, MA Unpublished
9 56 yo F, 1992 Definite RBC Kidney transplant Southeast CT 69
10 73 yo M, 1992 Probable RBC CABG Southeast CT 70
11 64 yo M, 1993 Possible RBC CABG No donor id Unpublished
12 44 yo M, 1993 Possible RBC Splenic vein thrombosis Wisconsin Unpublished
13 73 yo F, 1994 Definite Platelets Aortic valve replacement Southeast CT Unpublished
14 51 yo M, 1996 Possible Pheresis platelets Lymphoma Southeast CT Unpublished
15 72 yo F, 1996 Definite RBC CHF, renal failure Southeast CT Unpublished
16 67 yo M, 1997 Probable Platelets Lung cancer Southeast CT Unpublished
17 Newborn, 1997 Definite Red cells Prematurity Suffolk County, NY 56
Long Island
Newborn, 1997 Red cells Newborn
70 yo F, 1997 Red cells None
aDefinitea donor has been identified with a positive smear, hamster inoculation, or PCR or with a clear-cut serological progression. Patient meets criteria
below. Probablea donor has been identified who is seropositive at some time after the implicated donation. Patient meets criteria below. Possibleno donor
has been identified, but the recipient has a clear-cut history of babesiosis with no other source.
411
Table 3 Transfusion-Transmitted BabesiosisPatient

Factors (N = 19)
Sex
Male 10
Female 7
Newbornunknown 2
Age
Newborn 3
120 years 0
2140 years 1
4160 years 5
Over 60 years 10
Risk factors
Splenectomy 4
Splenic vein thrombosis 1
Prematurity 2
Renal failure 2
Cancer/Hematological malignancy 4
None identified 6
Incubation period, (excluding case 14) 17 days8 weeks
Clinical outcome
Death associated with babesiosis 1
Death unrelated to babesiosis 2
Recovered with treatment 9
No information 7
IX. POSTTRANSFUSION BABESIOSIS

Babesia species, particularly B. microti, meet the requirements to be transfusion
transmitted. First, there is documented evidence for subclinical parasitemia in
healthy individuals (36) and for persistent parasitemia after acute babesiosis (47).
Donors with a history of babesiosis are deferred permanently from blood dona-
tion. Second, the Babesia parasite survives well under blood bank storage
conditions (48). B. microti survive well at refrigerated temperatures for up to 21
days, while they did not survive well at 25C for more than 3 days. Finally,
intraerythrocytic B. microti are infectious to hamsters when inoculated intrave-
nously or intraperitoneally (the basis for the classic bioassay of infectivity). From
the evidence of human transfusion cases, Babesia are also infectious to humans
by the intravenous route.
This discussion will focus on cases reported in the United States. Although
there have been human cases of babesiosis in Europe due to species other than
B. microti, no cases of transfusion-transmitted babesiosis have been reported
outside of the United States (34).
Until recently, all reported cases of transfusion babesiosis have been caused
by B. microti, which is endemic in New England, on Long Island, and in the upper
Midwest. Other recently reported species of Babesia and Babesia-like piroplasms
are discussed below.
The first case of transfusion babesiosis was reported in 1980 (44) (Table 2,
Case 1). The patient was a 70-year-old man who was on prednisone and
transfused in 1979 with 20 units of platelet concentrates for neutropenia, anemia,
and thrombocytopenia, diagnosed as ITP, then splenectomized 4 weeks later. Four
weeks after the splenectomy (8 weeks after the transfusion), he developed a fever
and was noted to have Babesia on a peripheral smear. He was resistant to
treatment with chloroquine and pentamidine and was treated successfully with
two exchange transfusions 6 weeks apart. This case is typical in that the patient
had had a splenectomy and was elderly and immunosuppressed. It was atypical
in that the only blood components received were platelet concentrates. However,
platelet concentrates can contain up to 0.5 mL of red cells.
By todays standards, the case for transfusion transmission in this first
report was not particularly compelling, since the only evidence for the alleged in-
fected donor, a summer resident of Nantucket, was an elevated Babesia serolog-
ical titer 3 months after donation. Hamster bioassay of a new sample from the
donor 3 months after the donation was negative. In addition, the incubation period
was 8 weeks, on the long side. A follow-up letter (49) correctly suggested that the
patient may have been already infected with unrecognized babesiosis and that the
splenectomy and/or prednisone, not the transfusion, may have been responsible
for the clinical observations. Although the patient had no history of a visit to
hyperendemic areas such as Nantucket or the New England off-shore islands,
much of coastal New England may have had Babesia-infected ticks. (Most
babesiosis patients do not report a specific tick bite.) However, the patient
seroconverted for Babesia antibody during the course of his illness, making
transfusion transmission more likely.
Confirmation of transfusion babesiosis requires the following evidence (50):
1. The patient must display clear-cut evidence of babesiosis with no other
source identified. Ideally the serologic and clinical evidence excludes
long standing subclinical infection.
2. In addition, a donor must be identified in the typical incubation period
of transfusion babesiosis who, at a minimum, has a history of travel or
residence in an endemic area and a positive Babesia antibody.
3. The donor should demonstrate recent infectivity by hamster inocula-
tion, PCR, or (less definitively) typical acute serologic progression.
For the purpose of analysis of case reports, we have used Possible to de-
scribe cases in which only criteria 1 are met, Probable for cases in which
criteria 1 and 2 are met, and Definite for cases in which all three criteria
are met.
The first definite case of transfusion babesiosis was reported in 1982 (51)
from a transfusion in 1980 (Table 2, Case 3). The recipient was a premature
infant. The donor was a healthy summer resident of Fire Island (Long Island,
NY), who was demonstrated to be parasitemic by hamster inoculation a year after
the original exposure to infected ticks.
In 1994, stimulated by an increase of tickborne babesiosis in Connecticut
(52) and by three transfusion case reports to the Connecticut Red Cross Blood
Center, the authors initiated a registry of posttransfusion babesiosis cases. Blood
centers and Babesia researchers in New England, New York City, Long Island,
and Minnesota were contacted and asked to provide case reports. The reports
were tabulated in a standardized format and returned to the reporter for verifica-
tion. An update to this survey was conducted in 1996.
The data that follow are a summary of the data retained in this registry as
of July 1999. In total, 17 cases of transfusion-transmitted babesiosis have been
reported: 10 published and 7 unpublished. Table 2 is a summary of those cases,
along with references to the published cases.
Table 3 is a summary of patient factors revealed by these case reports.
Transfusion babesiosis occurs in all age groups and both sexes. There appears to
be a preponderance of older recipients, but this observation may merely reflect
the transfused population. Splenectomy, which is a risk factor for tickborne
disease, also appears to be a risk factor for transfusion transmission. Other
possible risk factorsprematurity, cancer and immunosuppressive chemotherapy,
and renal failuremay or may not be real, since these are also prevalent in
transfused patients. In any event, a significant proportion of cases appear to have
no complicating medical illness, with transfusion occurring in the context of
surgery or trauma.
The transfusion incubation period appears to be 28 weeks, with case 14
being an outlier. Case 14 illustrates the difficulty of identifying an implicated
donor. The case involved 49 donors, many of whom could not be contacted.
However, 22 were tested for Babesia antibody. Since the seroprevalence of
babesiosis in Connecticut is reported to be 1.02.6% (53), there is a high
likelihood of falsely implicating a donor not responsible for transmission. The ap-
parently implicated donor was seropositive and had a clear risk history. However,
he donated platelets by apheresis, which are relatively free of red cells, and the
apparent incubation period was 4.5 months, much longer than previously reported
in the other cases.
Transfusion babesiosis is a treatable disease, with the current recommended
therapy being clindamycin and quinine (43). Case 17 illustrates drug resistance
to clindamycin and quinine in a newborn, who subsequently responded to
atovoquone. Exchange transfusion to remove infected erythrocytes has also been
used successfully in the past (44). Outcome in general is favorable, with only one
death attributed to the infection among 19 patients.
Early detection and a high index of suspicion are important in the manage-
ment of transfusion babesiosis. The disease is often missed, particularly in areas
not endemic for the disease. It is also confused with transfusion malaria, although
a skilled hematology technologist should be able to differentiate the intra-
erythrocytic parasites. Any patient who develops fever and a hemolytic anemia
after transfusion (otherwise not explained) should be suspected to have transfu-
sion babesiosis or malaria. They should be evaluated with review of a thick blood
smear by a hematology technologist experienced in such diagnoses.
Table 4 illustrates the geographic origin of implicated donors. Most were
from New England or Long Island, NY. The two cases from the Midwest reflect
the known endemic focus of infected ticks in this area. The predominance of cases
from Connecticut may, in fact, reflect the relatively high degree of exposure of
Connecticut blood donors. But it may also reflect an effort on the part of the blood
center and Connecticut health officials to educate physicians about babesiosis
(54). It is likely that a large number of cases are subclinical throughout the
endemic areas for babesiosis.
Implicated donors, in general, are asymptomatic and, typical of subclinical
babesiosis, often appear to have been infected for months to years. Some donors
have evidence for systemic disease, which may also reflect Lyme disease and
other tickborne illnesses (55).
Table 5 considers the blood components involved in the 17 reported cases.
Not surprisingly, most cases are transmitted by red cells. Platelets have been
Table 4 Transfusion-Transmitted Babesiosis

Geographic and Donor Factors (N = 17)
Donor origin
Southeast CT 8
Cape Cod, MA 2
Offshore islands, MA 2
Long Island, NY 3
Wisconsin 1
Minnesota 1
Period of donor infection
<1 month 3
112 months 3
>12 months 1
Donor symptoms
Chronic fatigue, hot flashes, night sweats 1
Lyme disease 1
None reported 15
Table 5 Transfusion-Transmitted BabesiosisBlood Product Implicated (N = 19)

Blood component
Red cells 14
Frozen red cells 1
Platelets/Platelet pheresis 4
Red cell storage
17 days 0
821 days 7
>21 days 2
No data 5
Platelet storage
13 days 2
45 days 1
No data 1
Infectivityco-components
Platelets from implicated RBC donor One case evaluated (Case 17)platelet
recipient not infected
Red cells from implicated platelet donor No data available
Red cells (from Case 17) 2/4 co-RBC components were
infectious
implicated in only four of the reported cases, and in only one of these cases is
there definite implication of the involved donor.
In only one case (Case 17) has it been possible to evaluate co-components.
This interesting case illustrates the variable reactivity of red cells, since four other
patients besides the index case were transfused with other aliquots of the red cells
and only two were infected. The two uninfected red cell recipients were new-
borns, who received smaller doses of infected red cells (56) compared with the
three infected patients. A platelet concentrate made from the same donation did
not transmit Babesia to the recipient.
These cases also provide information on the survival of the Babesia parasite
under blood bank storage conditions. Definite cases of transfusion babesiosis
have been reported with red cells as old as 35 days and platelets as old as 4 days.
This is considerably longer than predicted from laboratory storage conditions
(48). However, in the laboratory, storage was in EDTA whole blood in glass tubes.
It is likely that blood bank storage conditions favoring red cell and/or platelet
survival would also be more favorable for the survival and infectivity of the
Babesia parasite.
The transfusion-transmission of Babesia species other than B. microti that
affect humans is to be expected. However, there has been no report of transfusion
transmission of B. divergens, the clinical babesiosis in Europe. Recently, two or
three different human species in the United States have been describedin
Washington State (WA-1) (37,57), in Missouri (MO1) (39), and in northern
California (38). The reports of WA-1 infection have now been supplemented with
a report of a transfusion-transmitted case caused by a WA-1like organism (58),
the first transfusion babesiosis not caused by B. microti.
This case occurred in a multiply transfused, 76-year-old, spleen-intact man
with myelodysplasia and postcoronary artery bypass graft and aortic valve
replacement. Originally thought to have malaria, he was treated with chloroquine
and primaquine with initial resolution of symptoms. Later review at the Centers
for Disease Control and Prevention (CDC) identified the unusual piroplasm in his
red cells. The species was identified as very close to WA-1 serologically and
noncross-reactive with B. microti. His symptoms of fever and persistent parasite-
mia recurred, and he was treated with clindamycin and quinine; he appeared to
clear his parasitemia. Fifty-seven blood donors were investigated2 units of
whole blood, 15 red cells, 36 platelets, and 4 FFP. The implicated donor was a
34-year-old red cell donor with a titer against the recipients organism of 1:4096.
He lived in rural Washington State and was frequently exposed to deer. He was
parasitemic by hamster inoculation 7 months after his implicated donation.
No co-component or lookback donations could be evaluated because the recipi-
ents had died or were unavailable.
This case report is strikingly similar to many B. microti case reports:
misdiagnosis early (in this case, as malaria), chronic infection in both donor and
recipient, and a mild clinical course in the recipient. It cannot be determined from
the available case reports of tickborne illnesses and this transfusion case whether
any of the newly reported piroplasms will represent a significant public health
threat, either from tick transmission or blood transfusion. It is clear that apparent
transfusion babesiosis should be suspected outside of the endemic areas for
B. microti. The characteristic intraerythrocytic forms are the most reliable diag-
nostic tool. Serology to B. microti is likely to be negative or only weakly positive
in infections with these new organisms.
X. PREVENTION STRATEGIES AND FUTURE PROSPECTS

Of all the tickborne diseases, only babesiosis currently represents a substantial
risk for transfusion transmission. The disease is probably more common than
is currently appreciated, but many cases are subclinical. Further work is re-
quired to establish the frequency of transmission, the possibility of other, un-
recognized, clinical syndromes in infected donors and recipients, and appropriate
prevention strategies.
A history of tick bite is not an effective strategy to prevent babesiosis or
any other tickborne disease from entering the blood supply. More than 20% of
Connecticut blood donors who were residents of Babesia-endemic areas reported

a tick bite during a 5-month summer season (59), as did 0.79.0% of donors in a
postcard survey of geographically dispersed U.S. blood donors (60). Deferring
such a number of donors is not feasible. Furthermore, many patients with Lyme
disease and/or babesiosis do not report a tick bite.
The current indirect fluorescent antibody (IFA) screening test for babesiosis
is inadequate, even for large-scale blood donor population surveys, much less as
a potential donor-screening method. The test requires maintaining an infected
hamster population to serve as a source of infected red cells. It also is quite
subjective. This test has not been standardized for use in clinical laboratories,
although it is a useful research method (61). Development of an automated
microplate format test that could be applied to donor specimens is badly needed.
Unlinked studies would be needed to characterize the assay and establish confirm-
atory methods, followed by linked testing and lookback, similar to studies
currently being carried out using linked Chagas disease screening of blood
donors (62). Other Babesia-like organisms need further investigation, since it
appears that tests for B. microti will not detect these organisms.
Enhanced clinical suspicion of Babesia is warranted, particularly in en-
demic areas, but also in nonendemic areas, since early clinical suspicion, labora-
tory recognition, and appropriate treatment appear to be important to good clinical
outcome. Hematologists and blood bank professionals need to consider babesiosis
in any transfused patient with fever and hemolysis. Not all hemolytic episodes
after transfusion are the result of red cell antibodies! Similar recommendations
exist for transfusion malaria for the same reasons.
Although Lyme disease may be transmissible by blood transfusion, it is
unlikely this will represent an important means of transmission. It is likely to be
overlooked if transmission occurs in areas with limited Lyme awareness or if the
manifestations of transfusion transmission are found to be highly atypical for
tickborne disease. In these cases, the rare case may go undetected and untreated,
resulting in substantial and unnecessary morbidity. Thus, it is appropriate that
transfusion medicine consultants maintain a high index of suspicion that Lyme
disease could be transfusion transmissible.
Other tickborne diseases will need continued monitoring for their impor-
tance to the blood supply, but it does not appear today that they will represent
important threats.
Public health methods to reduce tickborne disease should certainly help
lessen exposure to infected ticks. As new exposure-prevention methods are
implemented, the resultant public awareness is likely to increase their expectation
that blood banks will increase their understanding of tickborne diseases and
respond appropriately to their potential threat to the blood supply.
REFERENCES
1. AABB. FDA issues statement on blood donors possibly exposed to tick-borne
pathogens. AABB FAXNET 1997; No. 341.
1a. Arguin PM, Singleton J, Rotz LD, Marston E, Treadwell TA, Slater K, Chamberland
M, Schwartz A, Tengelsen L, Olson JG, Childs JE. An investigation into the
possibility of transmission of tick-borne pathogens via blood transfusion. Transfusion
1999; 39:828833.
2. Harwood RF, James MT. Entomology in Human and Animal Health. New York:
Macmillan Publishing Company, 1979:371378.
3. Krantz GW. A Manual of Acarology. 2nd ed. Corvalis: Oregon State University,
1878:211215.
4. Schmidt GD, Roberts LS. In: Foundations of Parasitology. 4th ed. St. Louis: Times
Mirror/Mosby College Publishing, 1989:660663.
5. Schmidt GD, Roberts LS. In: Foundations of Parasitology. 4th ed. St. Louis: Times
Mirror/Mosby College Publishing, 1989:671679.
6. Shulman IA, Appleman MD. An overview of unusual diseases transmitted by blood
transfusion within the United States. In: Westphal RG, Carlson KB, Turc JM, eds.
Emerging Global Patterns in Transfusion-Transmitted Infections. Arlington, VA:
American Association of Blood Banks, 1990:111.
7. Murray PR, Drew WL, Kobayashi GS, Thompson JH. Medical Microbiology. C.V.
St. Louis: Mosby, 1990:628632.
8. Monath TP, Tsai TF. Flaviviruses. In: Richman DD, Whitley RJ, Hayden FG, eds.
Clinical Virology. New York: Churchill Livingston, 1997:11331134.
9. Wahlberg P, Saikku P, Brummer-Korvenkontio M. Tick-borne viral encephalitis in
Finland. The clinical features of Kumlinge disease during 19591987. J Intern Med
1989; 225(3):173177.
10. Costero A, Grayson MA. Experimental transmission of Powassan virus (Flaviviridae)
by Ixodes scapularis ticks (Acari:Ixodadae). Am J Trop Med Hyg 1996; 55(5):536546.
11. Telford SR 3rd, Armstrong PM, Katavolos P, Foppa I, Garcia AS, Wilson ML,
Spielman A. A new tick-borne encephalitis-like virus infecting New England deer
ticks, Ixodes dammini. Emerg Infect Dis 1997; 3(2):165170.
12. Wong JK. Colorado tick fever virus and other arthropod-borne Reoviridae. In:
Richman DD, Whitley RJ, Hayden FG, eds. Clinical Virology. New York: Churchill
Livingston, 1997:755764.
13. Transmission of Colorado tick fever virus by blood transfusion. MMWR 1975;
24:425427.
14. CDC. Lyme diseaseUnited States, 1996. MMWR 1997; 46(23):531535.
15. Johnson RC, Schmid GP, Hyde FW, Steigerwalt AG, Brenner DJ. Borrelia burg-
dorferi sp. nov.: etiological agent of Lyme disease. Int J System Bacteriol 1984;
34:496497.
15a. Baranton G, Postic D, Saint Girons I, Boerlin P, Piffaretti JC, Assous M, Grimont PA.
Delineation of Borrelia burgdorferi sensu stricto, Borrelia garinii sp. nov. and group
VS461 associated with Lyme borreliosis. Int J System Bacteriol 1992; 42:378383.
16. Benach JL, Bosler EM, Hanrahan JP. Spirochetes isolated from the blood of two
patients with Lyme disease. N Engl J Med 1983; 308:740742.
17. Aoki SK, Holland PV. Lyme diseaseanother transfusion risk? Transfusion 1989;
29:646650.
18. Badon SJ, Fister RD, Cable RG. Survival of Borrelia burgdorferi in blood products.
Transfusion 1989; 29:581583.
19. Badon SJ, Cable RG, Morse EE, Aslanzadeh J, Malawista SE. Evidence of Borrelia
burgdorferi in a blood donation. Blood 1995; 86(suppl 1):355a.
20. Cable RG, Krause P, Badon S, Silver H, Ryan R. Acute blood donor co-infection with
Babesia microti (Bm) and Borrelia burgdorferi (Bb) (abstr). Transfusion 1993;
33(suppl):50s.
21. Piesman J, Hicks TC, Sinsky RJ, Obiri G. Simultaneous transmission of Borrelia
burgdorferi and Babesia microti by individual nymphal Ixodes dammini ticks. J Clin
Microbiol 1987; 25:20122013.
170:231234.
23. Harwood RF, James MT. Entomology in Human and Animal Health. New York:
Macmillan Publishing Company, 1979:400403.
24. Boyer KM, Munford RS, Maupin GO, et al. Tick-borne relapsing fever: an interstate
outbreak originating at Grand Canyon National Park. Am J Epidemiol 1977; 105(5):
469479.
25. Pratt HD, Darsie RF. Highlights in medical entomology in 1974. Bull Entomol Soc
Am 1975; 21:173176.
26. Hira PR, Husein SF. Some transfusion-induced parasitic infections in Zambia. J Hyg
Epidemiol Microbiol Immunol 1979; 23(4):436444.
27. Wang CW, Lee CU. Malaria and relapsing fever following blood transfusion. Chin
Med J 1936; 50:241.
28. Wells GM, Woodward TE, Fiset P, Hornick RB. Rocky Mountain spotted fever
caused by blood transfusion. JAMA 1978; 239(26):27632765.
29. Benenson AS, ed. Control of Communicable Diseases Manual. 16th ed. Washington,
DC: American Public Health Association, 1995:392395.
30. Casleton BG, Salata K, Dasch GA, Strickman D, Kelly DJ. Recovery and viability
of Orientia tsutsugamushi from packed red cells and the danger of acquiring scrub
typhus from blood transfusion. Transfusion 1998; 38:680689.
31. Walker DH, Dumler JS. Emergence of the ehrlichioses as human health problems.
Emerg Infect Dis 1996; 2(1):1829.
32. Magnarelli LA, Dumler JS, Anderson JF, Johnson RC, Fikrig E. Coexistence of
antibodies to tick-borne pathogens of babesiosis, ehrlichiosis and Lyme borreliosis in
human sera. J Clin Microbiol 1995; 33:30543057.
32a. Eastlund T, Persing D, Mathiesen D. Human granulocytic ehrlichiosis after red cell
transfusion. Transfusion 1999; 39(suppl):117s.
33. Spielman A, Etkind P, Piesman J. Reservoir hosts of human babesiosis on Nantucket
Island. Am J Trop Med Hyg 1981; 30:560565.
34. Dammin GJ, Spielman A, Benach JL. The rising incidence of clinical Babesia microti
infection. Hum Pathol 1981; 12:398400.
35. Murray PR, Drew WL, Kobayashi GS, Thompson JH. Medical Microbiology.
St. Louis: Mosbey, 1990:375377.
36. Gelfand JA. Babesia. In: Mandell GL, Bennett JE, Dolin RA, eds. Principles and
Practice of Infectious Diseases. 4th ed. New York: Churchill Livingston, 1995:2497
2500.
37. Quick RE, Herwaldt BL, Thomford JW, Garnett ME, Eberhard ML, Wilson M, Spach
DH, Dickerson JW, Telford SR 3d, Steingart KR. Babesiosis in Washington State: a
new species of Babesia? Ann Intern Med 1993; 119(4):284290.
38. Persing DH, Herwaldt BL, Glaser C, Lane RS, Thomford JW, Mathiesen D, Krause
PJ, Phillip DF, Conrad PA. Infection with a Babesia-like organism in northern
California. N Engl J Med 1995; 332(5):298303.
39. Herwaldt B, Persing DH, Prcigout EA, Goff WL, Mathiesen DA, Taylor PW,
Eberhard ML, Gorenflot AF. A fatal case of babesiosis in Missouri: identification of
another piroplasm that infects humans. Ann Intern Med 1996; 124(7):643650.
40. Shih CM, Liu LP, Chung WC, Ong SJ, Wang CC. Human babesiosis in Taiwan:
asymptomatic infection with a Babesia microti-like organism in a Taiwanese woman.
J Clin Microbiol 1997; 35(2):450454.
41. Ruebush TK II, Juranek DD, Spielman A. Epidemiology of human babesiosis on
Nantucket Island. Am J Trop Med Hyg 1981; 30:937941.
42. Benach JL, Habicht GS. Clinical characteristics of human babesiosis. J Infect Dis
1981; 144:481.
43. Clindamycin and quinine treatment for Babesia microti infections. MMWR 1983;
32:6572.
44. Jacoby GA, Hunt JV, Kosinski KS, Demirjian ZN, Huggins C, Etkind P, Marcus LC,
Spielman A. Treatment of transfusion-transmitted babesiosis by exchange transfu-
sion. N Engl J Med 1980; 303:10981100.
45. Popovsky MA, Lindberg LE, Syrek AL, Page, PL. Prevalence of Babesia antibody
in a selected blood donor population. Transfusion 1988; 28:2961.
46. Benach JL, Coleman JL, Habicht GS. Serological evidence for simultaneous occuren-
ces of Lyme disease and babesiosis. J Infect Dis 1985; 152:473477.
47. Krause PJ, Spielman A, Telford SR III, Sikand VK, McKay K, Christianson D,
Pollack RJ, Brassard P, Magera J, Ryan R, Persing EH. Persistent parasitemia after
acute babesiosis. N Engl J Med 1998; 339:160165.
48. Eberhard ML, Walker EM, Steurer F. Survival and infectivity of Babesia in blood
maintained at 25C and 2-4C. J Parasitol 1995; 81:790792.
49. Przyjemski CJ. Was babesiosis transmitted by transfusion? N Engl J Med 1981;
304:733.
50. Badon SJ, Cable RG. Investigation of suspected cases of post-transfusion babesiosis
(abstr). Transfusion 1997; 37(suppl):56S.
51. Wittner M, Rowin KS, Tanowitz HB, Hobbs JF, Saltzman S, Wenz B, Hirsch R,
Chisholm E, Healy GR. Successful chemotherapy of transfusion babesiosis. Ann
Intern Med 1982; 96:601604.
52. State of Connecticut Department of Health Services. BabesiosisConnecticut 1990
1996. Conn Epidemiol 1997; 17:1819.
53. Krause PJ, Telford SR, Ryan R. Geographic and temporal distribution of Babesial
infection in Connecticut. J Clin Microbiol 1991; 29:14.
54. State of Connecticut Department of Health Services. Babesiosis update. Conn

Epidemiol 1992; 12:1516.
55. Meldrum SC, Birkhead GS, White DJ, Benach JL, Morse DL. Human babesiosis in
New York State: an epidemiologic description of 136 cases. Clin Infect Dis 1992;
15:10191023.
56. Dobroszycki J, Herwaldt BL, Boctor F, Miller JR, Linden J, Eberhard ML, Yoon JJ,
Ali NM, Tanowitz HB, Graham F, Weiss LM, Wittner M. A cluster of transfusion-
associated babesiosis cases traced to a single asymptomatic donor. JAMA 1999;
281:927930.
57. Thomford JW, Conrad PA, Telford SR III. Cultivation and phylogenetic characteri-
zation of a newly recognized human pathogenic protozoan. J Infect Dis 1994;
169:10501056.
58. Herwaldt BL, Kjemtrup AM, Conrad PA, Barnes RC. Transfusion-transmitted babe-
siosis in Washington State: first reported case caused by a WA-1 parasite. J Infect Dis
1997; 1275:12591262.
59. Trouern-Trend JJ, Leiby DA, Cable RG, Badon SJ. Self-reported tick exposure in a
Babesia-endemic area. (abstr) Transfusion 1997; 37(suppl):48S.
60. Chung APS, Leiby DA, Persing DH, Cable RG, Trouern-Trend JJ, McCullough J.
Seroprevalence of Babesia microti and Ehrlichia sp in US blood donors (abstr).
Transfusion 1998; 38(suppl):S392.
61. Krause P, Telford S, Ryan R, Conrad P. Antibody testing for babesiosis: standardiza-
tion of a serologic test for the diagnosis of human Babesia microti infection. J Infect
Dis 1994; 923926.
62. Leiby DA, Jensen NC, Fucci M-C, Stumpf RJ. Prevalence of Trypanosoma cruzi
antibodies in a blood donor population at low risk (abstr.) Transfusion 1997;
37(suppl):S297.
63. Grabowski EF, Giardina PJ, Goldberg D, Masur H, Read SE, Hirsch RL, Benach JL.
Babesiosis transmitted by a transfusion of frozen-thawed blood. Ann Int Med 1982;
96:466467.
64. Marcus LC, Valigorsky JM, Fanning W. A case of transfusion-induced babesiosis.
JAMA 1982; 248:465467.
65. Gordon S, Cordon RA, Mazdzer EJ, Valigorsky JM, Blagg NA, Barnes SJ. Adult
respiratory distress syndrome in babesiosis. Chest 1984; 86:633634.
66. Smith RP, Evans AT, Popovsky M. Transfusion-acquired babesiosis and failure of
antibiotic treatment. JAMA 1986; 256:27262727.
67. Mintz ED, Anderson JF, Cable RG, Hadler JL. Transfusion-transmitted babesiosis: a
case report from a new endemic area. Transfusion 1991; 31:365368.
68. Reddy RL, Dalmasso AP. Transfusion acquired babesiosis in Minnesota (abstr). In:
ISBT and AABB 1990 Joint Congress Abstract Book. Arlington, VA: American
Association of Blood Banks, 1990:11.
69. Cable RG, Krause P, Badon S, Silver H, Ryan R (abstr) Acute blood donor
co-infection with Babesia microti (Bm) and Borrelia burgdorferi (Bb). Transfusion
1993; 33(suppl.):50s.
170:231234.
Surveillance for Transfusion-Transmitted Infectious Diseases
18
Surveillance for
Transfusion-Transmitted
Infectious Diseases
Mary E. Chamberland and Rima F. Khabbaz
Centers for Disease Control and Prevention, Atlanta, Georgia
I. INTRODUCTION
The U.S. blood supply is among the safest in the world. The risk of transfusion-
transmitted infections has been markedly reduced as a result of improvements in
donor screening and serological testing, the development and implementation of
viral inactivation procedures for plasma-derived products, and changes in trans-
fusion practices. Nevertheless, since blood is a human tissue, it is a natural vehicle
for transmission of infectious agents. The transmission of human immunodefi-
ciency virus (HIV) through the blood supply in the early 1980s and the more
recent experiences with hepatitis C virus (HCV) transmission from intravenous
immune globulin (IVIG) (1) and bacterial sepsis from contaminated albumin (2)
point to the need for continued vigilance regarding both known and unrecognized
or uncharacterized threats to the blood supply.
An important component of the multifaceted approach to help ensure
the safety of the blood supply is surveillance. As part of its retrospective re-
view of the events of the early 1980s that led to the transmission of HIV
through blood and blood products, the Institute of Medicine developed a se-
ries of recommendations to help thwart future threats to the blood supply. One
of these was a recommendation to establish a surveillance system . . . [to] de-
tect, monitor, and warn of adverse effects in recipients of blood and blood
products (3).
423
424 Chamberland and Khabbaz
Surveillance is defined as the ongoing systematic collection, analysis, and

interpretation of health data, closely integrated with the timely dissemination of
these data both to those providing the data and to those who can apply the data
to control and prevention programs (4). While surveillance is distinct from
epidemiological research, the two are often complementary (5). Epidemiological
research studies usually start with a specific hypothesis to be tested, are conducted
in a well-defined population, and are often time-limited; in addition, the data are
complete. In contrast, surveillance data are often used to identify or describe a
problem, identify cases for epidemiological research, monitor temporal trends, or
estimate the magnitude of a problem. Surveillance usually encompasses many
more individuals than might be enrolled in a research study. As a consequence,
data collected by surveillance, which can either be active or passive, are usually
less complete, less detailed, and more open-ended than research data. Importantly,
surveillance programs provide an infrastructure or network, such that in the event
of an unusual case report or an acute problem, even if not related to the
surveillance program, an established methodology is in place for communicating
information directly to public health authorities.
This chapter will review the Centers for Disease Control and Preventions
(CDC) major systems of surveillance for transfusion-transmitted infectious dis-
eases in the United States. CDC has several different programs of surveillance for
current or potential risks related to transfusion of blood and blood products.
Some programs are disease-based, while others monitor donors and recipients
of blood and blood products. In addition, applied research conducted by CDC
and others to assess the risk of transmission of infections through transfusion will
be presented.
II. HUMAN IMMUNODEFICIENCY VIRUSES

A. National Surveillance for Acquired
Immunodeficiency Syndrome
National surveillance and reporting of acquired immunodeficiency syndrome
(AIDS) has been in place for nearly 15 years. All states and U.S. territories require
reporting of AIDS cases to local health authorities. State and local health
departments, in turn, transmit cases electronically to CDC (6). Health department
staff actively survey case reports submitted from physicians, hospitals, labora-
tories, and other medical care facilities and from record systems, such as death
certificates and tumor registries. Although completeness of reporting of diagnosed
AIDS cases varies by geographic region and patient population, reporting of
AIDS cases in most areas of the United States is more than 85% complete (7,8).
Overall, about 50% of all AIDS cases are reported to CDC within 3 months of
diagnosis (6).
Surveillance for Transfusion-Transmitted Infectious Diseases 425
AIDS surveillance captures data on cases resulting from transfusion of

blood and blood products and among persons with hemophilia. Such cases
increased dramatically during the mid-1980s and then stabilized or declined
slightly until 1993 (Figs. 1 and 2). AIDS case reports received after January 1993
were influenced by an expanded AIDS surveillance case definition and chiefly
represent reporting of persons who had CD4+ cell counts below 200/uL with or
without illness. This change greatly altered the pattern of case reports, and was
most pronounced in the first 3 months of 1993. Since that time, the number of
case reports each year among transfusion recipients and persons with hemophilia
has returned to pre-1993 levels. Most cases reported since 1985 reflect infections
that occurred before serological screening of donors and heat and chemical
inactivation of clotting factor concentrates were initiated and the long period
between infection with HIV and progression to AIDS. Since the institution of
these practices, AIDS surveillance has identified persons who report receipt of
blood and blood products screened negative for HIV antibody. Such reports
trigger epidemiological and laboratory investigations of case-patients and of
donors. Through December 1996, 36 adults and adolescents and 3 children had
developed AIDS after receiving blood screened negative for HIV antibody (i.e.,
during the window period before the development of detectable HIV antibod-
ies) (6). To assist in the identification of AIDS cases related to transfusion of
blood products, algorithms have been developed for follow-up of persons with
hemophilia and AIDS who report no other risk factors (P. Sullivan, personal
communication). In particular, AIDS cases are examined and investigated for
persons born after 1985 who should have received only screened blood products
and for persons who have evidence of a recent seroconversion. In addition, as part
of CDCs and state and local health departments investigation of AIDS cases with
no identified risk, information is sought about possible exposure to blood prod-
ucts. These investigations employ systematic and standardized procedures, in-
cluding review of pertinent records and interview of the patient and others (6).
B. Serosurveillance of Blood Donors for HIV

The national HIV serosurveillance program, which includes seroprevalence sur-
veys conducted in selected sentinel sites throughout the country, also contributes
to monitoring the safety of the blood supply. CDC has collaborated with the
American National Red Cross since 1985 to establish donor-based systems of
surveillance, using data from the routine testing of blood donors (9). HIV
prevalence data are available from nearly 50 American National Red Cross blood
centers, which account for approximately half of the blood collected in the United
States. Since monitoring began, the overall HIV prevalence rate has decreased
nearly fourfold, from one positive for every 4,500 persons tested in late 1985 to
less than one positive for every 15,000 persons tested (Fig. 3) (10). This decline
426
Figure 1 Adult and pediatric cases of acquired immunodeficiency syndrome in the United States attributed to receipt of blood transfusion,
Chamberland and Khabbaz
blood components, or tissue, by year of report, 19821996.

Surveillance for Transfusion-Transmitted Infectious Diseases
Figure 2 Adult and pediatric cases of acquired immunodeficiency syndrome in the United States among persons with hemophilia or other
427
coagulation disorders, by year of report, 19821996.

428
Figure 3 HIV seroprevalence in blood donors in the United States, by date of donation, November 1985December 1993. (From Ref. 10.)
likely reflects aggressive donor education programs, improved donor selection

and screening through interview and serological testing, the progressive elimina-
tion of HIV-infected donors among repeat donors (repeat donors comprise
approximately 80% of all donations), and increased awareness among potential
donors of their HIV infection status due to the widespread availability of HIV
testing and counseling (1113).
Incidence data from a subset of blood centers in this surveillance system
have been used to derive estimates of the residual risk of transmission of HIV by
screened blood; current risk estimates are one case of HIV transmission for every
450,000660,000 donations of screened blood in the United States (13). In addi-
tion, the National Institutes of Health coordinates a multidisciplinary epidemio-
logical research program, the Retrovirus Epidemiology Donor Study (REDS), to
study the incidence of various infectious agents among volunteer blood donors in
five blood centers (14). REDS found a similar estimate (one HIV transmission
per 493,000 donations) for HIV transmitted by transfusion of screened blood (15).
To further reduce the risk of transfusion-transmitted HIV-1 infection, p24 antigen
screening of all blood and plasma donations was instituted in March 1996. HIV-1
p24 antigen detects HIV infection an average of 6 days before antibody tests are
positive (16). Initial projections estimated that four to six infectious donations,
not captured by other screening tests, would be detected among the 12 million
blood donations collected annually. However, in the first 9 months, only one
antigen-positive/antibody-negative donation was detected among 6.1 million
blood donations evaluated (17).
Another important outgrowth of CDC and the American National Red
Cross serosurveillance of blood donors is evaluation of HIV-infected blood
donors. Since 1988, CDC, in collaboration with the American National Red Cross
and other major blood collection agencies, has conducted follow-up interviews
and serological testing of HIV-seropositive donors. Information from these stud-
ies has been used to develop epidemiological and behavioral profiles of HIV-
seropositive donors, learn more about their reasons and motivations for donation,
and help blood centers improve strategies for encouraging appropriate self-
deferral (18). Despite these efforts, some donors with risk behaviors that may put
them at increased risk for HIV and other infectious diseases continue to donate.
REDS used an anonymous mail survey of a sample of persons who had donated
blood in 1993 and found that 1.9% of respondents reported a deferrable risk that
was present at the time of their past donation (19).
C. Surveillance for Divergent Strains of HIV-1

The overwhelming majority of HIV infections throughout the world are caused
by HIV-1 viruses belonging to group M. HIV-1 viruses in group O (i.e., outlier
group) have been recently identified and characterized. The number of group O
infections reported worldwide is small, and most have occurred among persons
in West and Central Africa countries (20). However, because these infections are
inconsistently detected by current enzyme immunoassays (EIAs) for antibodies
to HIV-1, there are important implications for blood safety (21). CDC has an
ongoing program to evaluate HIV-seropositive blood donors for unusual variants;
previous studies that evaluated blood donors and other populations have not found
any evidence of infection with HIV-1 group O (22). In addition, CDC has
established active international and domestic surveillance for divergent HIV
strains to monitor their prevalence, evaluate the sensitivity of licensed HIV
serological tests for detecting these variants, and assist in the modification of
screening tests to improve their sensitivity. As part of this surveillance program,
CDC, in collaboration with state health departments, is undertaking epidemiolog-
ical and laboratory investigations of persons who were reported with HIV/AIDS
and who were born in a country where HIV-1 group O has been documented (i.e.,
West and Central Africa). Two HIV-1 group O infections among persons in the
United States have been reported through this sentinel surveillance; both persons
were originally from Africa, and neither was a blood donor (23,24). Efforts are
underway to modify existing HIV-EIAs to improve detection of group O strains,
without compromising sensitivity for the group M viruses. As an interim measure,
the U.S. Food and Drug Administration (FDA) has recommended that donors at
increased risk for HIV-1 group O infection (e.g., resident in areas where group O
strains are endemic) be deferred from donating blood or plasma.
D. HIV-2
HIV-2, another retrovirus that causes AIDS and is endemic in many countries in
West Africa, has been transmitted by transfusion of blood and blood products
in Europe (25). Epidemiological and donor testing data indicate that the preva-
lence of HIV-2 among blood donors in the United States is very low. Neverthe-
less, because of the potential for transmission by blood and blood products and
the availability of licensed combination HIV-1/HIV-2 EIAs, FDA recommended
in June 1992 that blood and plasma donors be screened for antibodies to HIV-2.
No cases of transfusion-acquired HIV-2 infection have been reported in the
United States, and only three HIV-2 seropositive units have been found among
whole blood and plasma donors (26).
III. HEPATITIS B AND HEPATITIS C VIRUSES

The Viral Hepatitis Surveillance Program (VHSP) provides nationwide informa-
tion about hepatitis (27). VHSP receives reports of hepatitis from state and local
health departments using a standardized case record that is either mailed or
transmitted electronically to CDC. The form collects clinical, laboratory, and

epidemiological data. Since VHSP is a passive system for surveillance, interpre-
tation of data is complicated by underreporting and incomplete reporting of
serologic and epidemiological data.
Consequently, beginning in 1979, CDC began a complementary, more
intensive program of surveillance for acute viral hepatitis in several sentinel
counties, which are representative of the United States as a whole (28). In the
Sentinel Counties Study of Acute Viral Hepatitis, attempts are made to stimulate
reporting by writing to physicians, making on-site visits to hospital infection-
control and laboratory personnel, and through publicity in local medical newslet-
ters (29). Patients are interviewed to identify risk factors for hepatitis, and
blood for serological testing is obtained from all patients and contacts when-
ever possible. The sentinel counties study has facilitated the identification of
new or emerging types of hepatitis viruses (e.g., hepatitis G virus), because it
incorporates state-of-the-art serological and nucleic acid-based testing. Both the
VHSP and sentinel counties systems collect data on transfusion of blood and
blood products.
The data collected in these systems have been useful in monitoring epide-
miological trends for the various types of hepatitis. For example, the VHSP and
the sentinel counties study have documented declines in transfusion-associated
hepatitis B virus (HBV) and HCV infection (Fig. 4) (27,28,30). The population-
based sentinel counties study found that most non-A, non-B hepatitis in the
United States is not transfusion-associated and that the number of cases of
transfusion-associated hepatitis C declined significantly after 1985 (31). Most of
this decline occurred before surrogate testing of blood donors began and was
temporally associated with changes in the donor population resulting from
exclusion of persons at high risk for HIV infection and HIV antibodypositive
donors, as well as with changes in transfusion practices.
No direct measures of transfusion-transmitted viral hepatitis are available.
The current incidence of transfusion-associated hepatitis infections is so low in
the sentinel counties study that the small sample size precludes further identifi-
cation of transfusion-associated cases. Current estimates for HBV infection range
from one infection in 500,000 units transfused (32), derived from the frequency
of HBV infection among donors and the sensitivity of serological assays for
identifying hepatitis B surface antigen (HBsAg) and antibody to hepatitis B core
antigen (anti-HBc), to one in 63,000, derived from a study of incident infections
among repeat blood donors, which estimated the infectious window period for
HBV prior to HBsAg seroconversion (15). The residual risk for transfusion-
acquired HCV infection is estimated to range from one in 10,000 (32) to one in
103,000 donations (15). This risk is higher than that for HBV and HIV infections
because of the lower sensitivity (9095%) of contemporary tests (i.e., second-
version EIAs) used to screen blood donors; a longer window period (average 82
432
Figure 4 Cases of hepatitis B and hepatitis C/non-A, non-B hepatitis attributed to blood transfusion from selected states, by year of report,
19831995. (From Ref. 27.)
days; range 54192 days) during which antibodies cannot be detected; and an
inability to reliably detect chronic infection (15,32). A third-version anti-HCV
test, recently approved for screening blood donors, may further improve the
ability to detect anti-HCV antibodies in a low-prevalence population, such as
blood donors (33).
Although viral inactivation procedures have virtually eliminated the risk of
HCV infection from plasma-derived products, the first outbreak of hepatitis C
associated with contaminated IVIG was reported in the United States in 1994 (1).
Infection was associated with lots of Gammagard (Baxter Healthcare Corpora-
tion, Deerfield, IL) produced from plasma screened by second-version anti-HCV
assays that were positive for HCV RNA. It has been hypothesized that the
second-version tests removed most of the complexing antibodies from plasma
used to make IGIV, ultimately resulting in more HCV in the IG fraction than in
the non-IG fraction. In addition, manufacture of this particular product did not
include a viral inactivation step as a further measure for product safety.
IV. HUMAN T-LYMPHOTROPIC VIRUSES

TYPE I AND TYPE II
Currently, there is no national program of surveillance for infections caused by
human T-lymphotropic viruses type I and type II (HTLV-I and HTLV-II). Epide-
miological studies conducted in the United States have found that HTLV-I
infection is concentrated largely among blacks in the southeastern states and
among immigrants from countries where HTLV-I is highly endemic (e.g., Japan,
countries in the Caribbean basin) (34). In contrast, HTLV-II is very prevalent
among injecting drug users and several American Indian populations in the
United States. Modes of transmission include sexual contact, sharing contami-
nated needles, from mother to child (primarily through breast-feeding), and blood
transfusion (34).
Both viruses are cell-associated and have been transmitted by transfusion
of cellular blood products (e.g., whole blood, red blood cells, and platelets);
noncellular blood components derived from plasma, such as cryoprecipitate, have
not been associated with transmission (34). A screening EIA to test blood donors
for HTLV-I was introduced in late 1988; however, because the two viruses are
quite homologous, significant cross-reactivity with HTLV-II occurs with both
screening and confirmatory tests. Among the more than 8.5 million donations
tested during 1989, 0.015% were confirmed as seropositive; based on polymerase
chain reaction (PCR) testing of a sample of the seropositive donations, approxi-
mately half were infected with HTLV-I and half with HTLV-II (35).
Estimates of the residual risk of transfusion-transmitted HTLV range from
one in 50,000 screened donated units (based on seroprevalence rates among
donors and screening test sensitivity) (36) to one in 641,000 (based on a study of
incident infections among repeat blood donors in the infectious window period)
(15). The former estimate is similar to that observed in a prospective study by
Nelson and coworkers of patients who received transfusions during cardiac
surgery between 1985 and 1991 (37). They found that the combined rate of
transfusion-transmitted HTLV-I and -II fell from about one in 8,500 units trans-
fused before donor screening to one in 69,272 units after institution of screening.
The relative insensitivity of combined tests, particularly in detecting HTLV-II (the
more common retrovirus of the two in the United States), has been a concern.
Although as many as 43% of HTLV-IIinfected donations may have gone
undetected when earlier EIAs were used (38), more recent assays have reduced
this proportion to about 34% (39).
V. CREUTZFELDT-JAKOB DISEASE
Creutzfeldt-Jakob disease (CJD) is a rare, invariably fatal neurodegenerative
disease believed to be caused by an unconventional, disinfection-resistant infec-
tious agent (40). The likely causative agent is a prion protein, which is a
confirmationally altered form of a normal plasma membrane protein. Iatrogenic
cases of CJD in recipients of human pituitary-derived growth hormone and dura
mater grafts have had incubation periods of as long as 20 years. The CJD agent
has been reported occasionally in the blood of CJD patients at low titer, and
inoculation of peripheral blood mononuclear cells and some derivatives of plasma
from CJD patients into rodent brains has resulted in CJD-like lesions (4043).
In contrast to these experimental data, there has been to date no confirmed reports
of CJD resulting from receipt of blood transfusion. Epidemiological case-control
studies show no increase in the frequency of transfusions among CJD patients
compared with controls (4447).
CDC has several systems of surveillance in place to help assess the risk, if
any, of transmission of CJD by transfusion of blood and blood products. CDC
conducts routine surveillance for CJD through ongoing review of national mor-
tality data. Results from 1979 through 1994 indicate that annual rates of CJD have
remained stable (at about one case per million population) (48). Thus, despite
regular blood donation by persons who subsequently develop CJD, blood trans-
fusions do not appear to be amplifying CJD infections in the U.S. population.
None of the 3642 reported cases of CJD was also reported to have hemophilia,
thalassemia, or sickle cell diseasediseases associated with increased exposure
to blood or blood products, such as clotting factor concentrates and/or cryopre-
cipitate or red blood cells (49). Two studies suggest that routine mortality
surveillance has good sensitivity to detect CJD cases. The first study, conducted
in 11 states, found that 80% of all neuropathologically confirmed cases of CJD
during 19861988 could be ascertained by review of death certificates (50).

The second study was conducted in April 1996 in four Emerging Infections
Program sites in three states and two metropolitan areas as part of active
surveillance for the newly reported variant of CJD and physician-diagnosed cases
of CJD (51). In these surveillance areas, greater than 90% of all pathologists,
neurologists, and neuropathologists were contacted. Of the 94 CJD deaths iden-
tified during 19911995, 81 (86%) were found from death certificate review.
In 1995, CDC began to supplement this routine surveillance for CJD by
actively soliciting more than 140 hemophilia treatment centers in the United
States for any case reports of CJD. This effort also included arranging for
neuropathological examinations of brain tissue from deceased hemophilic pa-
tients with neurological disorders to look for signs of CJD and the presence of
the agent thought to cause the disease. Despite active solicitation of treatment
centers, as well as efforts to increase providers awareness about CJD through
educational symposia at national hemophilia meetings, no center has reported a
patient with clinical CJD (B. Evatt, personal communication). Neuropathological
examination has been completed for 26 deceased persons with hemophilia; none
had evidence of CJD (B. Evatt, personal communication).
Finally, to further enhance the evidence derived from routine surveillance,
CDC is assisting in coordinating a long-term, follow-up study of recipients who
received blood components from donors who were subsequently reported to
have been diagnosed with CJD (52). The vital status of 179 recipients of trans-
fusable blood components from 14 donors who subsequently developed CJD
has been determined by using primarily the national death index through 1995;
none of these recipients were reported to have died of CJD. Among these
recipients, 41 persons lived 5 or more years after their transfusion, including 9
who lived as long as 1324 years. At least one additional follow-up study of
recipients of blood from donors who subsequently developed CJD has not
detected CJD in recipients (53).
VI. BACTERIAL CONTAMINATION

The incidence and range of adverse clinical outcomes from bacterial contamina-
tion of blood have been poorly characterized (54,55). From 1986 through 1991,
FDA received reports of 182 transfusion-associated fatalities, of which 29 (16%)
were caused by bacterial contamination of blood products (56,57). However, the
incidence of both fatal and nonfatal bacteria-associated transfusion-related com-
plications is likely underestimated (55,56). Recently, the U.S. General Accounting
Office estimated that the rate of bacteria-associated adverse reactions from
random donor platelet pools was 0.6 per 1000 pooled units (58). Transfusion of
red blood cells has been associated with bacterial sepsis, most often due to
Yersinia enterocolitica, which grows readily in refrigerated blood. From Novem-

ber 1985 through November 1996, CDC received 21 reports of sepsis, including
12 deaths, associated with transfusion of red blood cells contaminated with
Y. enterocolitica (56). Similar to what has been reported for platelet-related
bacteremia, the risk for sepsis increases with duration of storage of red blood
cells, and most episodes are associated with erythrocytes stored for 25 days
(54,55). The risk for red blood cellassociated sepsis has been estimated to be one
per 500,000 units transfused (59); similarly, AuBuchon has estimated that one of
every 1,000,000 units may be associated with endotoxin-induced septic shock
caused by gram-negative bacteria (60).
Because rates of bacteria-associated transfusion reactions in the United
States are unknown, CDC, in collaboration with national blood-collection organ-
izations, is initiating a prospective study to determine the rates of bacteria-
associated transfusion reactions from whole blood, red blood cells, and platelets
(56). This study will establish standardized definitions and systematic proce-
dures for the recognition, reporting, and clinical, epidemiological, and laboratory
evaluation of adverse transfusion reactions in recipients of contaminated blood or
blood components.
VII. TRANSFUSION-INDUCED PARASITIC DISEASES

A. Babesiosis
Babesia are intraerythrocytic parasites that can cause chronic asymptomatic
infection and hence pose a transfusion risk. Although there is no national
surveillance program for cases of babesiosis, hundreds of cases of babesiosis have
been reported in the United States, mostly in the Northeast. Of these, fewer than
25 cases of transfusion-associated babesiosis from asymptomatic infected blood
donors, mostly caused by Babesia microti but also by the more recently recog-
nized WA1-type Babesia parasite, have been reported (6163). The parasite
survives blood banking conditions and is transmissible by transfusion of red blood
cells and platelet concentrates (64). With the expansion of deer populations in the
northeastern United States, concerns exist that the incidence of transfusion-
transmitted babesiosis may increase (65). The tick vector and animal reservoir of
the Babesia more recently found in the northwestern United States remain to be
identified. Current strategies for prevention of transfusion-transmitted Babesia
rely on questioning donors for history of babesiosis and on performing a hema-
tocrit determination at the time of donation. Babesiosis classically manifests as a
febrile illness with hemolytic anemia, but infection can also cause chronic
asymptomatic or mildly symptomatic parasitemia. The risk for severe disease is
higher in the elderly, asplenic, and immunocompromised patients.
B. Chagas Disease
Chagas disease, a vectorborne disease caused by the parasite Trypanosoma cruzi,
is endemic in parts of Central and South America and Mexico. Transmission in
nature occurs by cutaneous or mucosal contact with the feces of reduvid bugs.
If untreated, Chagas disease can result in lifelong, asymptomatic parasitemia;
infected individuals can in turn transmit T. cruzi through transfusion. Receipt of
blood transfusions is the most common means of transmission in some disease-
endemic areas (66).
The immigration of thousands of persons from T. cruziendemic areas and
increased international travel have raised concerns about the increased potential
for transfusion-transmitted Chagas disease in North America (54). It has been
estimated that at least 100,000 persons with chronic T. cruzi infection reside in
the United States (66). To date, there have been three reported cases of Chagas
disease from transfusions in the United States and one in Canada; all four patients
were immunocompromised and developed acute, symptomatic disease, which
facilitated their recognition (54,66). The American National Red Cross has
conducted limited lookback studies of recipients of blood and blood products that
were seropositive for antibodies to T. cruzi and found no evidence of transmis-
sion, suggesting that the transmission rate in the United States may be no higher
than 10% (58,67). Until recently, reliable seroprevalence estimates among blood
donors have been hampered by the lack of serological tests with adequate
sensitivity and specificity. More recent studies using screening and confirmatory
testing schemas and conducted among selected blood donors who were more
likely to be at increased risk for T. cruzi infection because of birth in or travel to
disease-endemic countries have found approximately 0.1% of such donors to be
seropositive for antibodies to T. cruzi (58,6769). Extrapolating from these and
other studies, Dodd estimated that the overall risk for T. cruzi contamination is
one in 42,000 per unit of donated blood in the United States (58). Although there
is no national surveillance system in place to monitor transfusion-associated
Chagas disease, the drug (nifurtimox) that is used to treat acute infection in the
United States is available only through CDCs Drug Service. Hence, CDC would
likely learn of diagnosed, symptomatic cases of acute infection.
C. Leishmaniasis
Leishmaniasis is a parasitic disease, primarily found in the tropics and subtropics,
that is transmitted in nature by the bite of infected sand flies. Leishmaniasis can
cause skin lesions (cutaneous leishmaniasis) or mucosal lesions (mucosal leish-
maniasis) or invade internal organs (e.g., liver, spleen, bone marrow) of the body
(visceral leishmaniasis) and cause anemia, weakness, and sometimes death,
if untreated. Visceral leishmaniasis, caused by organisms in the Leishmania
donovani species complex, can be associated with parasitemia and therefore

with the potential for transmission by blood transfusion. Transmission of leish-
maniasis by blood transfusion has been reported only in rare instances (fewer
than 15 cases worldwide), and no case of any type of leishmaniasis acquired by
blood transfusion has been reported in the United States (B. Herwaldt, personal
communication).
Concern about the potential for transmission of leishmania through blood
transfusion occurred when a previously undescribed type of visceral leishmania-
sis was found to affect a small number of persons who had been deployed to the
Persian Gulf as part of Operation Desert Storm (70). To date, the Department of
Defense has parasitologically confirmed 12 cases of this unusual type of visceral
leishmaniasis, called viscerotropic leishmaniasis, among the almost 700,000
deployed personnel (A. Magill, personal communication) (70,71). Viscerotropic
leishmaniasis is unusual in that it is caused by a parasite (Leishmania tropica)
that typically causes cutaneous leishmaniasis, but in this instance the parasite
invades the internal organs of the body. All but one of the 12 detected cases of
viscerotropic leishmaniasis were in symptomatic persons; however, the clinical
manifestations were nonspecific. Parasitological confirmation of an active case
of viscerotropic leishmaniasis is difficult because it requires an invasive proce-
dure (e.g., a bone marrow or lymph node aspiration), and infected persons
generally have low levels of the parasite.
In November 1991, the U.S. Department of Defense and the American
Association of Blood Banks recommended that all persons who had traveled to
the Persian Gulf after August 1, 1990, be deferred as blood donors. This was a
precautionary measure because the prevalence of viscerotropic leismaniasis was
not known. In addition, L. tropica is known to survive in refrigerated components
and retain its infectivity (72). In January 1993, the American Association of Blood
Banks and the Department of Defense lifted the ban; no additional cases of
viscerotropic leishmaniasis had been identified during the nearly 14 months of
the ban (73).
D. Malaria
Malaria is a reportable disease in the United States, and cases confirmed by blood
smear are reported to local and/or state health departments by health care
providers and/or laboratories (74). CDC also directly obtains reports of other
cases from health care providers who request assistance in the diagnosis or
treatment of malaria. All reported cases of malaria that are acquired in the United
States are fully investigated to collect detained clinical and epidemiological data.
For cases in persons who report no history of international travel but who received
a blood transfusion, the blood bank is contacted to identify and collect blood
specimens from potentially implicated donors. Donors are serologically tested,
and any seropositive donors are investigated. Implicated donors are more likely
to be identified by indirect fluorescent-antibody detection than by blood smear
examination alone; most have subpatent parasitemia, and a positive blood smear
is found only in approximately 30% of donors (75).
On average, approximately three cases per year of transfusion-induced
malaria are reported in the United States (75). The overall incidence of transfu-
sion-transmitted malaria in the United States is about one case per every 4 million
units of blood collected (75). The infecting Plasmodium species for the 99 cases
of transfusion-induced malaria reported in 19581994 were: P. malariae, 32
(32%); P. falciparum, 31 (31%); P. vivax, 28 (28%); P. ovale, 7 (7%); and
P. falciparum and P. malariae, 1 (1%) (M. Parise, personal communication).
Of the 70 implicated donors, 39 (56%) were foreign nationals (M. Parise,
personal communication). At present, the optimal means of preventing transmis-
sion of malaria by blood transfusion is through deferral of donors with a history
of malaria or a relevant travel history. CDCs investigations of transfusion-
acquired malaria cases indicate that most commonly implicated donors incor-
rectly answer questions regarding travel to a malaria-endemic area.
VIII. HEMOPHILIA SURVEILLANCE ACTIVITIES

Another approach to monitoring the virological safety of blood products relies on
surveillance programs that focus on recipients of blood and plasma products.
CDC has two such systems of surveillance: the Hemophilia Surveillance System
(HSS) and the Universal Data Collection System (J. M. Soucie, personal commu-
nication). The HSS is a population-based program designed to identify all persons
with hemophilia in six states (Colorado, Georgia, Louisiana, Massachusetts, New
York, and Oklahoma). Cases are identified from various sources, including
hemophilia treatment centers, practicing hematologists, and others. To date,
approximately 4000 patients (or 25% of the estimated U.S. hemophilia popula-
tion) have been identified. Data are collected through retrospective chart abstrac-
tion and include available serological testing data for hepatitis A, B, and C viruses
and HIV. In addition, information about any infectious diseases diagnosed during
hospitalizations is collected. Data abstraction began in 1995 for the years 1993
and 1994 and is planned to continue through 1999. Analyses that are planned
include determination of prevalence and incidence rates of hepatitis and HIV
infection and examination of any unusual clinical illnesses.
To complement the HSS, CDC, in cooperation with health care providers
in comprehensive hemophilia treatment centers, has developed a new national,
prospective system called the Universal Data Collection System, which will be
fully implemented in 1998 (J. M. Soucie, personal communication). This system
gathers a uniform set of data about the health outcomes of an estimated 17,000
20,000 persons with hemophilia and related congenital blood clotting disorders
in the United States. Data about the nature and extent of joint and infectious
disease complications are collected in all 144 federally sponsored, specialized
hemophilia treatment centers throughout the country. In addition, a serum speci-
men is sent to CDC for serological testing for hepatitis and HIV infection and for
storage in a national serum bank for use in future investigations related to blood
safety issues. The system also has an acute illnessreporting system that facilitates
identification and investigation of potential infection sources and outbreaks and
the development of intervention strategies to prevent further disease occurrence.
IX. CONCLUSION
Since blood is a biological product, it is unlikely that the risk of transfusion-
transmitted infections will ever be reduced to zero. Nonetheless, the blood supply
in the United States in the 1990s is among the safest in the world and is safer now
than it has ever been. Factors contributing to the reduced risk of transfusion-
transmitted infections include improvements in donor selection, donor screening
by serological tests and other markers of infection, and viral inactivation proce-
dures for plasma-derived products. These approaches, coupled with surveillance
to monitor for trends in infection, are the mainstay of public health strategies to
protect the blood supply from known pathogens and to monitor for the emergence
of new infectious agents. CDC views surveillance as a critical component of its
overall mission. One of the four goals of CDCs strategic plan to help address
emerging infectious threats is the improvement and expansion of surveillance
capabilities for infectious diseases in the United States and internationally (76).
Enhanced surveillance can play an important role in helping to ensure the safety
of blood and plasma products.
REFERENCES
1. Bresee JS, Mast EE, Coleman PJ, Baron MJ, Schonberger LB, Alter MJ, Jonas MM,
Yu MW, Renzi PM, Schneider LC. Hepatitis C virus infection associated with admin-
istration of intravenous immune globulin. A cohort study. JAMA 1996; 276:1563
1567.
2. Centers for Disease Control and Prevention. Bacterial sepsis associated with receipt
of albumin. MMWR 1996; 45:866867.
3. Committee to Study HIV Transmission Through Blood and Blood Products, Division
of Health Promotion and Disease Prevention, Institute of Medicine. Leveton LB, Sox
HC Jr, Stoto MA, eds. HIV and the Blood Supply. An analysis of crisis decisionmak-
ing. Washington, DC: National Academy Press, 1995.
4. Thacker SB, Berkelman RL, Stroup DF. The science of public health surveillance.
J Public Health Policy 1989; 10:187203.
5. Thacker SB, Berkelman RL. Public health surveillance in the United States.
Epidemiol Rev 1988; 10:164190.
6. Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report 1996;
8(no. 2):139.
7. Buehler JW, Berkelman RL, Stehr-Green JK. The completeness of AIDS surveillance.
J AIDS 1992; 5:257264.
8. Rosenblum L, Buehler JW, Morgan MW, Costa S, Hidalgo J, Holmes R, Lieb L,
Shields A, Whyte BM. The completeness of AIDS case reporting, 1988: A multistate
collaborative surveillance project. Am J Public Health 1992; 82:14951499.
9. Petersen LR, Dodd R, Dondero TJ Jr. Methodologic approaches to surveillance of
HIV infection among blood donors. Pub Health Rep 1990; 105:153157.
10. Centers for Disease Control and Prevention. National HIV Serosurveillance Sum-
mary: Update1993. Vol 3. Atlanta: U.S. Department of Health and Human Ser-
vices, 1995.
11. Centers for Disease Control and Prevention. National HIV Serosurveillance Sum-
mary: Results through 1992. Vol 3. Atlanta: U.S. Department of Health and Human
Services, 1993.
12. Johnson ES, Doll LS, Satten GA, Lenes B, Shafer AW, Kamel H, Casanova RJ,
Petersen LR. Direct oral questions to blood donors: the impact on screening for
human immunodeficiency virus. Transfusion 1994; 34:769774.
Lewis WF, Notari EP, Petersen LR. Estimated risk of transmission of the human
333:17211725.
14. Zuck TF, Thomson RA, Schreiber GB, Gilcher RO, Kleinman SH, Murphy EL,
Ownby HE, Williams AE, Busch MP, Smith JW, Nass CC, Hollingsworth CG, Nemo
GJ for the REDS Group. The Retrovirus Epidemiology Donor Study (REDS);
rationale and methods. Transfusion 1995; 35:944951.
15. Schreiber GB, Busch MP, Kleinman SH, Korelitz JJ, for the Retrovirus Epidemiology
Donor Study. The risk of transfusion-transmitted viral infections. N Engl J Med 1996;
334:16851690.
16. Busch MP, Lee LLL, Satten GA, Henrard DR, Farzadegan H, Nelson KE, Read S,
preceding human immunodeficiency virus type 1 seroconversion: implications for
17. Lackritz EM, Stramer SL, Jacobs TA, Strauss D, Busch MP, Schable CA. Results of
national testing of U.S. blood donations for HIV-1 p24 antigen [Abstract 751]. In:
Program and Abstracts of the 4th Conference on Retroviruses and Opportunistic
Infections, Washington DC, January 2226, 1997.
18. Doll LS, Petersen LR, White CR, Ward JW, HIV Blood Donor Study Group. Human
immunodeficiency virus type 1-infected blood donors: behavioral characteristics and
reasons for donation. Transfusion 1991; 31:704709.
SH, Hollingsworth CG, Nemo GJ, for the Retrovirus Epidemiology Donor Study.
Estimates of infectious disease risk factors in US blood donors. JAMA 1997;
277:967972.
20. Hu DJ, Dondero TJ, Rayfield MA, George JR, Schochetman G, Jaffe HW, Luo C-C,
Kalish ML, Weniger BG, Pau C-P, Schable CA, Curran JW. The emerging genetic
diversity of HIV. The importance of global surveillance for diagnostics, research and
prevention. JAMA 1996; 275:210216.
21. Schable C, Zekeng L, Pau C-P, Hu D, Kaptue L, Gurtler L, Dondero T, Tsague J-M,
Schochetman G, Jaffe H, George JR. Sensitivity of United States HIV antibody tests
22. Pau CP, Hu DJ, Spruill C, Schable C, Lackritz E, Kai M, George JR, Rayfield MA,
Dondero TJ, Williams AE, Busch MP, Brown AE, McCutchan FE, Schochetman G.
Surveillance for human immunodeficiency virus type 1 group O infections in the
United States. Transfusion 1996; 36:398400.
23. Sullivan PS, Jones J, Schable CA, Schochetman G, Pau C, Rayfield M, Do A, Tetteh
C, Ward JW. National surveillance for HIV-1 group O infections: Preliminary results
[Abstract 159]. In: Program and Abstracts of the 4th Conference on Retroviruses and
Opportunistic Infections, Washington, DC, January 2226, 1997.
24. Sullivan PS, Do AN, Robbins K, Kalish M, Subbarao S, Pieniazek D, Schable C,
Afaq G, Markowitz J, Myers R, Joseph JM, Benjamin G. Surveillance for variant
strains of HIV: subtype G and group O HIV-1 [letter]. JAMA 1997; 278:292.
25. OBrien TR, George JR, Holmberg SD. Human immunodeficiency virus type 2
infection in the United States: epidemiology, diagnosis, and public health importance.
JAMA 1992; 267:27752779.
26. Sullivan M, Williams A, Guido E, Metler R, Schable C, Stramer S. Detection and
characterization of an HIV type 2 antibody positive blood donor in the U.S. [abst
S229]. Transfusion 1997; 37(suppl):58S.
27. Centers for Disease Control and Prevention. Hepatitis Surveillance Report No. 56.
Atlanta: U.S. Department of Health and Human Services, 1995.
28. Alter MJ, Hadler SC, Margolis HS, Alexander WJ, Hu PY, Judson FN, Mares A,
Miller JK, Moyer LA. The changing epidemiology of hepatitis B in the United States.
Need for alternative vaccination strategies. JAMA 1990; 263:12181222.
29. Alter MJ, Mares A, Hadler SC, Maynard JE. The effect of underreporting on the
apparent incidence and epidemiology of acute viral hepatitis. Am J Epidemiol 1987;
125:133139.
30. Centers for Disease Control. Public Health Service inter-agency guidelines for
screening donors of blood, plasma, organs, tissues, and semen for evidence of
hepatitis B and hepatitis C. MMWR 1991; 40 (RR-4):117.
31. Alter MJ, Hadler SC, Judson FJ, Mares A, Alexander WJ, Hu PY, Miller JK, Moyer
LA, Fields HA, Bradley DW, Margolis HS. Risk factors for acute non-A, non-B
hepatitis in the United States and association with hepatitis C virus infection. JAMA
1990; 264:22312235.
32. Alter MJ. Residual Risk of Transfusion-Associated Hepatitis. NIH Consensus Devel-
opment Conference on Infectious Disease Testing for Blood Transfusions. Bethesda,
National Institutes of Health, 1995.
33. Gretch D. Diagnostic tests for hepatitis C. In Program and Abstracts of NIH
Consensus Development Conference on Management of Hepatitis C. Bethesda,
National Institutes of Health, 1997.
34. Khabbaz RF, Heneine W, Kaplan JE. Testing for other human retroviruses: HTLV-I
and HTLV-II. In: Schochetman G, George JR, eds. AIDS Testing. 2d ed. New York:
Springer-Verlag, 1994:206223.
35. Centers for Disease Control. Human T-lymphotropic virus type I screening in
volunteer blood donorsUnited States, 1989. MMWR 1990; 39:915, 921924.
36. Dodd RY. The risk of transfusion-transmitted infection [editorial]. N Engl J Med
1992; 327:419421.
37. Nelson KE, Donahue JG, Munoz A, Cohen ND, Ness PM, Teague A, Stambolis VA,
Yawn DH, Callicott B, McAllister H, Reitz BA, Lee H, Farzadegan H, Hollingsworth
CG. Transmission of retroviruses from seronegative donors by transfusion during
cardiac surgery: a multicenter study of HIV-1 and HTLV-I/II infections. Ann Intern
Med 1992; 117:554559.
38. Hjelle B, Wilson C, Cyrus S, Bradshaw P, Lo J, Schammel C, Wiltbank T, Alexander
S. Human T-cell leukemia virus type II infection frequently goes undetected in
contemporary U.S. blood donors. Blood 1993; 81:16411644.
39. Gallo D, Yeh ET, Moore ES, Hanson CV. Comparison of four enzyme immunoassays
for detection of human T-cell lymphotropic virus type 2 antibodies. J Clin Microbiol
1996; 34:213215.
40. DeArmond SJ, Prusiner SP. Etiology and pathogenesis of prion diseases. Am J Pathol
1995; 146:785811.
41. Manuelidis EE, Kim JH, Mericangas JR, Maneulidis L. Transmission to animals of
Creutzfeld-Jakob disease from human blood [letter]. Lancet 1985; II:896897.
42. Tamai Y, Kojima H, Kitajima R, Taguchi F, Ohtani Y, Kawaguchi T, Miura S, Sata
M, Ishihara Y. Demonstration of the transmissible agent in tissue from a pregnant
woman with Creutzfeldt-Jakob disease [letter]. N Engl J Med 1992; 327:649.
43. Tateishi J. Transmission of Creutzfeldt-Jakob disease from human blood and urine
into mice [letter]. Lancet 1985; II:1074.
44. Davanipour Z, Alter M, Sobel E, Asher DM, Gajdusek DC. A case-control study of
Creutzfeldt-Jakob disease. Dietary risk factors. Am J Epidemiol 1985; 122:443451.
45. Esmonde TFG, Will RG, Slattery JM, Knight R, Harries-Jones R, DeSilva R,
341:205207.
46. Kondo K, Kuroiwa Y. A case control study of Creutzfeldt-Jakob disease: association
with physical injuries. Ann Neurol 1982; 11:377381.
47. Little BW, Mastrianni J, DeHaven AL, Brown P, Goldfarb L, Gajdusek DC. The
epidemiology of Creutzfeldt-Jakob disease in eastern Pennsylvania [Abstract 614P].
Neurology 1993; 43:A316.
48. Holman RC, Khan AS, Belay ED, Schonberger LB. Creutzfeldt-Jakob disease in the
United States, 19791994: using national mortality data to assess the possible
occurrence of variant cases. Emerg Infect Dis 1996; 2:333337.
49. Schonberger LB, Sullivan M, Holman RC, Austin H, Dodd R, Khan A. Creutzfeldt-
Jakob Disease (CJD) and blood safety [Abstract PW 40-4]. In: Abstracts of Xth
International Congress of Virology, Jerusalem, Israel, August 1116, 1996.
50. Davanipour Z, Smoak C, Bohr T, Sobel E, Liwnicz B, Chang S. Death certificates:
an efficient source for ascertainment of Creutzfeldt-Jakob disease cases. Neuro-
epidemiology 1995; 14:16.
51. Centers for Disease Control and Prevention. Surveillance for Creutzfeldt-Jakob
diseaseUnited States. MMWR 1996; 45:665668.
52. Sullivan MT, Schonberger LB, Kessler D, Williams AE, Dodd RY. Creutzfeldt-Jakob
disease (CJD) investigational lookback study [Abstract S6]. Transfusion 1997;
37(suppl):2S.
53. Heye N, Hensen S, Muller N. Creutzfeldt-Jakob disease and blood transfusion [letter].
Lancet 1994; 343:298299.
54. Sloand EM, Pitt E, Klein HG. Safety of the blood supply. JAMA 1995; 274:1368
1373.
Micro Rev 1994; 7:290302.
56. Centers for Disease Control and Prevention. Red blood cell transfusions contami-
nated with Yersinia enterocoliticaUnited States, 19911996, and initiation of a
national study to detect bacteria-associated transfusion reactions. MMWR 1997;
46:553555.
57. Hoppe PA. Interim measures for detection of bacterially contaminated red cell
components [editorial]. Transfusion 1992; 32:199201.
58. U.S. General Accounting Office. Blood Supply. Transfusion-Associated Risks.
GAO/PEMD-97-2. Washington, DC: Government Printing Office, 1997.
59. Dodd RY. Adverse consequences of blood transfusion: quantitative risk estimates. In:
Nance ST, ed. Blood Supply Risks, Perceptions and Prospects for the Future.
Bethesda, MD: American Association of Blood Banks, 1994:124.
60. AuBuchon JP. Blood transfusion options. Improving outcomes and reducing costs.
Arch Pathol Lab Med 1997; 121:4047.
170:231234.
62. Herwaldt BL, Kjemtrup AM, Conrad PA, Barnes RC, Wilson M, McCarthy MG,
Sayers MH, Eberhard ML. Transfusion-transmitted babesiosis in Washington State:
first reported case caused by a WA1-type parasite. J Infect Dis 1997; 175:12591262.
63. Mintz ED, Anderson JF, Cable RG, Hadler JL. Transfusion-transmitted bebesiosis: a
case report from a new endemic area. Transfusion 1991; 31:365368.
64. Eberhard ML, Walker EM, Steurer FJ. Survival and infectivity of Babesia in blood
maintained at 25C and 24C. J Parasitol 1995; 81:790792.
65. Popovsky MA. Transfusion-transmitted babesiosis [editorial]. Transfusion 1991;
31:296298.
66. Schmunis GA. Trypanosoma cruzi, the etiologic agent of Chagas disease: status in
the blood supply in endemic and nonendemic countries. Transfusion 1991; 31:547
557.
67. Leiby DA, Read EJ, Lenes BA, Yund AJ, Stumpf RJ, Kirchoff LV, Dodd RY.
68. Brashear RJ, Winkler MA, Schur JD, Lee H, Burczak JD, Hall HJ, Pan AA. Detection
of antibodies to Trypanosoma cruzi among blood donors in the southwestern and
western United States. I. Evaluation of the sensitivity and specificity of an enzyme
immunoassay for detecting antibodies to T cruzi. Transfusion 1995; 35:213218.
69. Schulman IA, Appleman MD, Saxena S, Hiti AL, Kirchhoff LV. Specific antibodies
to Trypanosoma cruzi among blood donors in Los Angeles, California. Transfusion
1997; 37:727731.
70. Magill AJ, Grogl M, Gasser RA Jr, Sun W, Oster CN. Visceral infection caused by
Leishmania tropica in veterans of Operation Desert Storm. N Engl J Med 1993;
328:13831387.
71. Magill AJ, Grogl M, Johnson SC, Gasser RA. Visceral infection due to Leishmania
tropica in a veteran of Operation Desert Storm who presented 2 years after leaving
Saudi Arabia. Clin Infect Dis 1994; 19:805806.
72. Grogl M, Daugirda JL, Hoover DL, Magill AJ, Berman JD. Survivability and
infectivity of viscerotropic Leishmania tropica from Operation Desert Storm partic-
ipants in human blood products maintained under blood bank conditions. Am J Trop
Med Hyg 1993; 49:308315.
73. Gunby P. Desert Storm veterans now may donate blood; others called for discussion
of donor tests. JAMA 1993; 269:451452.
74. Centers for Disease Control and Prevention. Malaria surveillanceUnited States,
1994. MMWR 1997; 46(SS-5):118.
selection criteria for United States travelers to malarious areas. Transfusion 1991;
31:798804.
76. Centers for Disease Control and Prevention. Addressing Emerging Infectious Disease
Threats. A Prevention Strategy for the United States. Atlanta: U.S. Department of
Health and Human Services, 1994.
19
Alternatives to Allogeneic Blood
and Strategies to Avoid Transfusion
Lawrence T. Goodnough
Washington University School of Medicine, St. Louis, Missouri
I. INTRODUCTION
Approximately 12 million red blood cell (RBC) units are transfused to nearly 4
million patients annually in the United States (1). The conservation of blood has
historically arisen from awareness that the inventory of this resource is limited
(2) as well as the knowledge that blood transfusion carries a risk (3). In addition,
emphasis on the costs of health care has raised issues related to the costs of
blood transfusion (46). Recent guidelines have emphasized that in the elective
transfusion setting, no blood transfusion is a desirable outcome (7). Further-
more, consensus conference recommendations (8) have emphasized that if blood
is to be transfused, autologous (the patients own) blood is preferable to al-
logeneic (from an anonymous, volunteer donor) blood. Thus the costs of blood
conservation, for which an increasing array of technological procedures and
products have become available, have also become an issue. The purpose of this
review is to provide an overview of emerging trends in blood transfusion and
blood conservation interventions in order to help identify areas important for
future investigation.
II. STRATEGIES TO AVOID BLOOD TRANSFUSION

Indications for blood transfusion have been impaired by the lack of data quanti-
tating the benefits of blood for reducing the morbidity or mortality of anemia in
otherwise untransfused patients (9). A randomized trial to determine at what level
447
448 Goodnough
of anemia blood transfusion can reduce morbidity or mortality cannot be done for
ethical reasons. Attempts to reach conclusions from the considerable experience
obtained in managing Jehovahs Witnesses, who refuse blood on the basis of
religious beliefs, have been problematic (10). This is illustrated by the random-
ized prospective study of the synthetic oxygen carrier Fluosol-DA compared to
placebo in such a population, in which patients generally did well if their lowest
hematocrit (HCT) level exceeded 12% but did not do well at a HCT level of less
than 12%, irrespective of treatment group (11). One cannot reach conclusions
from these data regarding the transfusion trigger appropriate for populations
who would accept blood. Moreover, it is recognized that patients are heteroge-
neous, so that patients with medical (chronic, due to underproduction) anemias
are transfused differently from patients with surgical (acute, due to blood loss)
anemias (9). A more recent study (12) of surgical mortality in Jehovahs Witness
patients was able to show that preoperative hemoglobin levels and subsequent
blood loss were related to mortality, especially in patients known to have
cardiovascular disease. Attempts to link transfusions with outcomes such as
length of stay (13,14) have been unsuccessful. Two studies have found that
females are more likely to be transfused than are their male counterparts in the
elective surgical setting (15,16). The latter study found, even when using gener-
ous criteria to define excessive transfusions in which transfusion to replace losses
of less than 15% of the initial RBC volume was deemed inappropriate, that 21%
of patients undergoing joint replacement surgery received blood unnecessarily.
III. UTILIZATION REVIEW: IS IT EFFECTIVE?

A recent review concluded that transfusion audits can improve transfusion
practices if they are performed in a timely manner and are combined with
education of the individual ordering physician (17). Plasma and platelet products
are particularly amenable to this approach. Two studies using concurrent educa-
tion or consultation reduced plasma usage by 46% and 77%, respectively (18,19).
Another study using a retrospective audit was able to reduce inappropriate plasma
use from 53 to 22% of units transfused (20). Similarly, use of platelet transfusions
were reduced by 56% and 14% in two studies that used consultation (21) and
audit (22), respectively. Another study followed all requests for transfusions with
a nonrequested consultation. The authors report a reduction in transfusion of
platelets by 44%, and plasma and cryoprecipitate by 57%, but a reduction in red
blood cells (RBC) of only 19% over a 4-year period (23). Other studies cast doubt
on whether utilization review is really an effective process. Hoeltge et al. (24)
used a combination of indicators to evaluate transfusions on medical and surgical
services and concluded that only 4% of transfusions were unjustified. Renner
et al. (25) found the percentage of unjustified transfusions to be 1.4% before and
Blood Conservation Interventions 449
0% after an educational intervention; while this could be interpreted to be a

triumph for the educational intervention, a more reasonable conclusion is that the
identification of unjustified transfusions, before and after the intervention, was
flawed (26). At our own institution in 1994, during which 23,002 RBC units were
transfused, 48 transfusion events underwent peer review by the transfusion
committee, and, of these, only 2 cases were felt to be unjustified.
These extraordinary low rates of inappropriate transfusions may be a
consequence of several factors. First, RBC transfusion audits in circumstances of
hemorrhage are difficult, if not impossible, to perform accurately. These settings
would include the emergency room/trauma unit, operating rooms, and intensive
care units. For this reason, our institutional process of utilization review does not
include transfusions administered intraoperatively. Yet, studies of transfusions
practices in orthopedic surgery indicate that at least 25% of RBC transfusions in
this setting can, in retrospect, be identified as inappropriate (16). Second, the
clinical indicators that define appropriate transfusion practice may be too
generous. In the study that concluded that 96% of transfusions were appropri-
ate, a posttransfusion hemoglobin (Hgb) concentration of 11 g/dL was used as a
threshold to distinguish appropriate from inappropriate (24). Third, the
medical chart audit has substantial limitations. Clearly documented information
as to why the transfusion was administered is commonly unobtainable. We found
that in orthopedic surgical patients, only 68% of postoperative transfusion events
on the day of surgery had chart documentation of blood loss and/or change in
vital signs (27). Nevertheless, these patients can be inferred to have been
transfused during periods of substantial blood loss, since recorded hematocrits
were 33.5 0.9% before transfusion and 31.3 0.5% after transfusion. Further-
more, the rationale for transfusion was recorded in only 16% of day-of-surgery
transfusions and in only 27% of transfusions administered on postoperative days.
A recent multiinstitution study concluded that retrospective utilization
review was not effective at altering transfusion practice (28). One alternative
approach to retrospective chart audit is the prospective use of transfusion algo-
rithms, in which the transfusion decision process is coupled with information that
can serve as clinical indicators for transfusion.
IV. TRANSFUSION ALGORITHMS

Studies have evaluated the impact of point-of-care testing (29,30), in which
intraoperative assays (whole blood prothrombin time, activated partial thrombo-
plastin time and platelet count results available within 4 minutes) were linked
with a transfusion algorithm (Fig. 1) for plasma and platelet products in cardiac
surgical patients with a diagnosis of microvascular bleeding (MVB). Patients
were randomized to either standard therapy, in which blood products were
450 Goodnough
Figure 1 An algorithm approach for hemostatic therapy in cardiac patients determined

to have microvascular bleeding after heparin neutralization. Platelets = Platelet transfusion
(6 units of random-donor or apheresis unit equivalent); PLT RX = platelet therapy (platelet
transfusion and/or DDAVP therapy at physicians discretion); FFP = plasma therapy
(2 units of fresh-frozen plasma); [+] MVB = continued MVB; PT:APTT = whole blood
prothrombin time and activated partial thromboplastin time control values (values/mean
values from a normal reference population); PLT count = platelet count ( 103/mm3).
(From Ref. 30.)
transfused at the discretion of the physician according to any laboratory-based

tests requested, or to an algorithm group, in which platelet and plasma therapy
were given according to an algorithm based initially on platelet count, followed
by branch pathways determined by PT and aPTT results. Both intraoperative
and initial postoperative chest tube drainage were less in the algorithm group.
One patient in the algorithm group required later surgical reexploration, compared
to five patients in the standard group. The more effective therapy in the algorithm
group was reflected in the lower RBC transfusion needs in the algorithm group
compared with the standard therapy group (5.9 3.8 vs. 9.8 9.4 units,
respectively). The improved patient care, along with reduced blood transfusions,
resulted in substantial economic savings. This approach has been described as a
powerful engine of change (31).
Approaches to red cell transfusion therapy in the surgical patient need to
acknowledge that patients are heterogeneous for risks related to anemia. In one
such approach to risk stratification for cardiac surgical patients has been pub-
lished (32), clinical variables determine whether a patient is considered at
standard or at increased risk for morbidity associated with cardiac surgery.
While complications related to anemia represent only one aspect of morbidity/
mortality in this setting, the physiological changes known to accompany acute
anemia (33) and the potential for myocardial tissue injury (34) suggest that risk
stratification for RBC transfusion decisions would be prudent. A recent analysis
of over 2000 cardiac bypass patients led to a model that calculated a transfusion
risk score, which was then validated prospectively in over 400 additional patients
(35). Now that cardiac surgery can be prospectively stratified, not only for
surgical risk but also for transfusion likelihood, the role of algorithms may be
especially productive in the standardization of blood transfusion and blood
conservation practices.
If transfusion outcomes can be predicted from patient-related factors, then
blood transfusion and blood conservation algorithms can be utilized to minimize
the variability of transfusion outcomes related to institutional (procedural) and
physician (transfusion practices) factors. Algorithms could take into account
patient heterogeneity by using the Higgins et al. (32) and McGovern et al. (35)
stratification of standard and increased risk patients. Blood transfusions and
blood-conservation strategies could be administered according to algorithms that
are incorporated into the daily practice of coronary revascularization (36). Such
an algorithm is illustrated in Figure 2. This approach can then enable physicians
to analyze transfusion outcomes, and the relationship of these outcomes, to
transfusion triggers, autologous blood procurement, pharmacological interven-
tions, and even emerging blood substitutes. The knowledge that we learn from
the standardization of these practices, and the comparison of transfusion out-
comes of different institutions or different therapeutic approaches, would form an
important database that could serve as a reference for the continuous improve-
ment of care in the surgical setting (37).
V. ALTERNATIVES TO ALLOGENEIC BLOOD

A. Autologous Blood Donation
For transfusion settings such as elective surgery, preoperative autologous blood
donation (PAD) represents an attractive alternative to allogeneic blood transfu-
452 Goodnough
Figure 2 An algorithm approach for red cell transfusion in cardiac surgical patients
postoperatively. After establishing that the patients volume status is adequate, decisions
to transfuse would be based on hemoglobin/hematorcrit level, rate of blood loss, and
hemodynamic parameters [mean arterial pressure (MAP) and cardiac index (CI)]. Thresh-
olds for transfusion would differ for patients determined to be a low risk and high risk
for perisurgical complications. Mixed venous oxygen percent saturation (SVO2) could
serve as a physiological indicator of the balance between oxygen supply and demand for
transfusion decision making. (From Ref. 36.)
sion. This previously underutilized practice (38) has now become a standard of
care in elective surgical procedures such as orthopedic and urological surgery
(39,40). Guidelines on the criteria for utilization of this intervention, along with
selection of patients suitable for the technique, have been published (42,42).
Potential candidates for PAD prior to surgery include any patient scheduled for a
procedure for whom blood type and crossmatch is requested. Using this approach,
studies have found that, overall, 9% of autologous donors undergoing elective
surgery receive allogeneic blood (43). Under conditions of standard collection
(i.e., one unit weekly) the likelihood of exposure to allogeneic blood for autolo-
gous blood donors has ranged from 10% for radical prostatectomy (39) to 17%
for elective orthopedic surgery (44) to 56% for patients undergoing coronary
revascularization (45).
The efficacy of PAD is dependent on the degree to which the patients bone
marrow is stimulated to increase the production of red blood cells in order to
replace blood donated. Several studies have shown that the endogenous erythro-
poietin response is suboptimal at the level of mild anemia achieved during
collection of autologous blood (46,47). Under standard conditions of one
autologous unit donated weekly, a computer model demonstrated that if the
erythropoietic response to autologous blood phlebotomy is not able to maintain
hematocrit level during the donation interval, the predeposit of autologous blood
may actually be harmful (48).
In contrast to autologous blood donation under standard conditions,
studies of aggressive autologous blood phlebotomy (twice weekly for 3 weeks,
beginning 2535 days before surgery) have demonstrated that endogenous eryth-
ropoietin levels do rise when orthopedic surgical patients undergo a blood
donation (blood loss) of up to 1000 mL weekly during PAD. In a controlled
clinical trial of aggressive phlebotomy (49), serial erythropoietin levels showed a
logarithmic rise even in the Hgb concentration range of 110114 g/dL. This was
accompanied by significant erythropoiesis, in which a RBC volume equivalent to
three allogeneic blood units [at 200 mL each (50)] was generated over a
preoperative interval of 28 days (51). The erythropoietin response to blood loss
and the subsequent erythropoietic response have been shown to be independent
of patient gender or age (52). The ability of recombinant human erythropoietin
(EPO) to further accelerate the erythropoietic recovery from blood loss during
PAD has been demonstrated in several controlled trials (51,53,54). When aggres-
sive autologous phlebotomy was accompanied by simultaneous EPO administra-
tion, the equivalent of an additional two blood units (nearly five total) was
generated for subsequent blood conservation. Subsequent clinical trials have
demonstrated that EPO therapy during autologous blood donation can reduce the
need for allogeneic blood in patients undergoing elective orthopedic surgery
(55,56), even in patients with anemia of chronic inflammatory disease such as
rheumatoid arthritis (57).
Guidelines for autologous blood transfusion are controversial. Some pub-
lished guidelines recommend that the same criteria be used for utilization review
for autologous as are used for allogeneic blood units (57). This view is supported
by the risks of an immediate transfusion reaction, in which administrative error
and bacterial contamination account for nearly 50% of fatalities from transfusions
(58). These risks can occur for autologous blood units as well as for allogeneic
blood units. The alternative view holds that the risk/benefit relationship is lower
for autologous blood than for allogeneic blood, since the risk for disease trans-
missible by autologous blood has been reduced; this viewpoint argues for
different standards for transfusion of autologous than for allogeneic blood. While
transfusion services may choose to establish different guidelines for transfusion
454 Goodnough
of autologous and allogeneic blood (59), physicians should not retransfuse

autologous blood simply because it is available (60).
B. Acute Preoperative Hemodilution

Acute isovolemic hemodilution is the removal of blood from a patient shortly
before surgery and replacement with crystalloid and colloid solutions. Blood is
collected in standard blood bags containing anticoagulant, stored at room temper-
ature to preserve platelet function, and reinfused after major blood loss has ceased
or sooner, if indicated. The rationale for the use of hemodilution is that if
intraoperative blood loss is relatively constant with or without preoperative
normovolemic hemodilution, it is better to lose blood at a lower hematocrit.
With moderate hemodilution, blood is withdrawn until the hematocrit
level is a certain value, e.g., 2730% (61). Recently, the efficacy of moderate
hemodilution has been questioned (62,63). For example, the removal of three
blood units in a 100 kg man with a reduction in preoperative hematocrit level
from 44% to 32% and who subsequently undergoes radical prostatectomy with
an estimated blood loss of 2600 mL, results in savings in surgical RBC volume
lost of only 215 mL, or the equivalent of one blood unit (62). The safety and
efficacy of more extensive hemodilution remains unproven (64).
The procurement of autologous blood by acute hemodilution has substantial
cost advantages over autologous blood predeposit; costs of inventory and testing
are not incurred for autologous blood procured by hemodilution, since this blood
does not leave the operating suite. A recent study demonstrated that moderate
hemodilution is a cost-effective alternative to preoperative autologous blood
donation in patients undergoing radical prostatectomy (65). If hemodilution is
coupled with emerging but still experimental artificial oxygen carriers or hemo-
globin solutions (66), in which these blood substitutes are used to replace blood
removed by hemodilution, the conservation benefit of this technique may be
greatly enhanced.
C. Intraoperative Blood Recovery

The term intraoperative blood recovery describes the technique of salvaging and
reinfusing blood lost by a patient during surgery. The oxygen-transport properties
of recovered red blood cells are equal to or better than stored allogeneic red cells,
and the survival of recovered red blood cells appears to be at least comparable to
that of transfused allogeneic red cells (67). The incidence of adverse events
resulting from reinfusion of recovered blood is not known but is thought to be an
infrequent (<1:15,000) occurrence. Fatal air embolism has been reported with
recovered blood infused under pressure (68). Hemolysis of recovered blood can
occur during suctioning from the surface instead of deep pools of shed blood,
particularly when blood is aspirated at vacuum settings greater than 100 torr.
The clinical importance of free hemoglobin in the concentrations usually seen has
not been established, although excess free hemoglobin may indicate inadequate
washing. Dilutional coagulopathy may occur if large volumes of salvaged blood
are administered (69).
As with preoperative autologous blood donation and acute preoperative
hemodilution, intraoperative autologous blood recovery should undergo scrutiny
concerning both safety and cost-effectiveness. A controlled study recently dem-
onstrated a lack of efficacy for intraoperative blood recovery when transfusion
requirements and clinical outcome were followed (70). A second study found that
only a minority of patients undergoing major orthopedic and cardiac surgery
achieved cost equivalence with intraoperative recovery using semi-automated
instruments compared with banked blood (71). While the recovery of a minimum
of one blood unit equivalent is possible for less expensive methods (with
unwashed blood), it is generally agreed that at least two blood unit equivalents
would need to be recovered using an automated recovery device (with washed
blood) in order to achieve cost-effectiveness (72). A recent study of this approach
in patients undergoing aortic aneurism repair concluded that this strategy is
particularly valuable in patients with at least 1000 mL surgical blood loss (73).
D. Postoperative Autologous Blood Recovery

In the postoperative orthopedic surgical setting, a number of reports have de-
scribed the successful recovery and reinfusion of washed (74,75) and unwashed
(7678) wound drainage from patients undergoing arthroplasty or spinal proce-
dures. The volume of reinfused drainage blood has been reported to be as much
as 3000 mL, with an average of more than 1100 mL in patients undergoing
cementless knee replacement (77). The corresponding red cell volume can be
substantial, ranging from 174 to 704 mL red cell volume or 0.75 to 3.5 units of
blood (79). Perioperative blood loss has been shown to be greater in patients
undergoing modern cementless total joint replacement when compared with cases
using cemented fixation technique (74,80), making postoperative red cell recov-
ery especially attractive for cementless procedures.
The safety of reinfused unwashed orthopedic wound drainage has been
controversial. Theoretical concerns have been expressed regarding infusion of
potentially harmful materials in salvaged blood, including free hemoglobin,
red cell stroma, marrow fat, toxic irritants, tissue or methacrylate debris, fibrin
degradation products, and activated coagulation factors and complement (81).
Two small series have reported complications (82,83); the etiology of the com-
plications is not clearly identified. Several larger studies have reported no
serious adverse effects when drainage was passed through a standard 40 m
blood filter (76,84).
456 Goodnough
E. Designated (Directed) Blood Transfusions

This transfusion practice (in which the blood donor is known to the transfusion
recipient) is commonly requested by patients as an alternative to allogeneic
volunteer blood transfusion. This transfusion practice, unlike autologous blood
transfusion, is controversial; no evidence to date indicates that designated blood
is safer than allogeneic blood (85). Reports of fatal graft-versus-host disease
(GVHD) in surgical patients who received blood from blood relatives serve to
emphasize the potential risks of this practice; directed donations from blood
relatives must be irradiated in order to prevent GVHD (86). While all allogeneic
blood and blood components, including designated donations, undergo testing
before transfusion, an additional mechanism to ensure blood safety is the confi-
dential donor screening process (87). Here the allogeneic blood donor is given
information identifying high-risk behaviors indicating that people should not
donate blood; high-risk donors can withdraw from blood donation or identify
themselves as high risk, confidentially, by checking a box indicating that the
blood donated by them should not be transfused. This screening mechanism is
compromised in designated blood donation, since the blood donor is known to
the transfusion recipient and is therefore arguably nonvoluntary. Furthermore,
directed blood donation has a potentially adverse effect on blood donor recruit-
ment, since individuals who save themselves for the unexpected transfusion
needs of family and friends might be less inclined to become allogeneic volunteer
blood donors (88). Finally, procurement of designated blood does not provide
additional protection against allogeneic blood exposure in patients who qualify
for autologous blood donation (89), reinforcing that the superior transfusion
practice of autologous blood donation should not be replaced by a controversial
transfusion practice.
An analysis of our own directed donor program illustrates how this practice
can create problems for patients and physicians (90). In 220 consecutive requests
for directed donors, 29 (29%) of the 101 patients were in nonelective transfusion
settings, such as hospitalized patients on medical services or surgical patients
within 24 hours of surgery. These requests potentially delayed transfusion therapy
because of the time (824 h) required for donor screening, phlebotomy, and unit
testing. Only 40% (46) of 115 units ultimately collected were of benefit to the
designated recipient. Directed donor programs need to better reflect their purpose
of providing an alternative to allogeneic volunteer blood in the elective, rather
than nonelective, transfusion setting.
VI. CONCLUSION
The procurement of autologous blood has become a standard of care for elective
surgery in the United States. However, the costs and potential complications, as
well as the potential benefits, of autologous blood-procurement strategies need to

be considered. Autologous blood donation under routine conditions of one
blood unit donation weekly is not accompanied by any significant erythropoietic
response and therefore results simply in chronic hemodilution. Interventions
that stimulate preoperative erythropoiesis, via either aggressive phlebotomy
or EPO therapy, represent more effective approaches to blood conservation
than conventional autologous blood predeposit. Second, point-of-care autologous
blood procurement strategies such as intraoperative blood recovery and
acute normovolemic hemodilution represent more cost-effective approaches.
Acute hemodilution will be utilized as a building block for emerging pharmaco-
logical strategies such as EPO therapy, blood substitutes, and artificial oxygen
transport solutions.
REFERENCES
1. Wallace EL, et al. Collection and transfusion of blood in the United States, 1992.
Transfusion 1995; 35:802812.
2. Roche JK, Stengle JM. Open heart surgery and the demand for blood. JAMA 1973;
225:15161521.
3. Goodnough LT, Shuck JM. Review of risks, options, and informed consent for blood
transfusion in elective surgery. Am J Surg 1990; 159:602609.
4. Schwartz WB, Mendelson DN, Hospital cost containment in the 1980s. N Engl J Med
1991; 324:10371042.
5. Forbes JM, Anderson MD, Anderson GF, et al. Blood transfusion costs: a multicenter
study. Transfusion 1991; 31:319323.
6. Goodnough LT, Bodner MS, Martin JW. Blood transfusion and blood conservation:
cost and utilization issues. Amer J Med Qual 1994; 9:172183.
7. American College of Physicians. Strategies for elective red cell transfusion. Ann
Intern Med 1992; 116:403406.
8. National Institute of Health Consensus Conference. Perioperative red cell transfusion.
JAMA 1988; 260:27002703.
9. Welch GH, Meehan K, Goodnough LT. Prudent strategies for elective red cell
transfusion. Ann Intern Med 1992; 116:393403.
10. Mann MC, Votto J, Kambe J, McNamee MJ. Management of the severely anemic
patients who refuses transfusion: lessons learned during the care of a Jehovahs
witness. Ann Int Med 1992; 117:10421048.
11. Gould SA, Rosen AL, Sehgal LR, et al. Fluosol-DA As a red cell substitute in acute
anemia. N Engl J Med 1986; 314:19531956.
12. Carson JL, Duff A, Poses RM, et al. Effect of anamia and cardiovascular disease on
surgical mortality and morbidity. Lancet 1996; 348:10551060.
13. Kim DM, Brecher ME, Estes TJ, Morrey BF. Relationship of hemoglobin and
duration of hospitalization after total hip arthroplasty: implications for the transfusion
target. Mayo Clin Proc 1993; 68:3741.
14. Goodnough LT, Vizmig K, Riddell J IV, Soegiarso RW. Discharge haematocrit as
458 Goodnough
clinical indicator for blood transfusion audit in surgery patients. Transfusion Med
1994; 4:3544.
15. Friedman BA, Burns TL, Schork MA. An analysis of blood transfusion of surgical
patients by sex: a quest for the transfusion trigger. Transfusion 1980; 20:179188.
16. Goodnough LT, Verbrugge D, Vizmrg K, Riddell J. Identifying elective orthopaedic
surgical patients with amounts of blood in excess of need; the transfusion trigger
revisited. Transfusion 1992; 32:648653.
17. Toy PTCY. Effectiveness of transfusion audits and practice guidelines. Arch Pathol
Lab 1994; 118:435437.
18. Ayoub MM, Clark JA. Reduction of fresh frozen plasma use with a simple education
program. Am Surg 1989; 55:563565.
19. Shanberge JN. Reduction of fresh-frozen plasma use through a daily survey and
education program. Transfusion 1987; 27:226227.
20. Barnette RE, Fish DJ, Eisenstaedt RS. Modification of fresh-frozen transfusion
practices through educational intervention. Transfusion 1990; 30:253257.
21. McCullough J, Steeper TA, Connelly DP, Jackson B, Huntington S, Scott EP. Platelet
utilization in a university hospital. JAMA 1988; 259:24142418.
22. Simpson MB. Prospective-concurrent audits and medical consultation for platelet
transfusions. Transfusion 1987; 27:192195.
23. Lichtiger B, Fisher HE, Huh YO. Screening of transfusion service requests by the
blood bank pathologist: impact on cost containment. Lab Med 1988; 19:228230.
24. Hoeltge GA, Brown JC, Herzig RH, et al. Computer-assisted audits of blood
component transfusion. Cleve Clin J Med 1989; 56:267269.
25. Renner SW, Howanitz JH, Fishkin BG. Toward meaningful blood usage review: com-
prehensive monitoring of physician practice. QRB Qual Rev Bull 1987; 13:7680.
26. Goodnough LT. Retrospective utilization reviewAre we just going through the
motions? Arch Pathol Lab Med 1996; 120:802.
27. Audet AM, Goodnough LT, Parvin CA. Evaluating the appropriateness of red blood
cell transfusions: the limitations of retrospective medical record reviews. Int J Qual
Health Care 1996; 8:4149.
28. Lam HT, Schweitzer SO, Petz L, et al. Are retrospective peer-review transfusion
monitoring systems effective in reducing red blood cell utilization. Arch Pathol Lab
Med 1996; 120:810816.
29. Despotis GJ, Grishaber JE, Goodnough LT. The effect of an intraoperative treatment
algorithm on physician transfusion behavior in cardiac surgery. Transfusion 1994;
34:290296.
30. Despotis GJ, Santoro SA, Spitznagel E, Kater KM, Cox JL, Barnes P, Lappas DL.
Prospective evaluation and clinical utility of on-site coagulation monitoring in cardiac
surgical patients. J Thorac Cardiovasc Surg 1994; 107:271279.
31. Reinersten JL. Algorithms, guidelines, and protocols: can they really improve what
we do? Transfusion 1994; 34:281228.
32. Higgins TL, Estafanous FG, Loop FD, Beck GJ, Blum JM, Paranand L. Stratification
of morbidity and mortality outcome by preoperative risk factors in coronary artery
bypass patients. JAMA 1992; 267:23442348.
33. Rao TLK, Montoya A. Cardiovascular, electrocardiographic and respiratory changes

following acute anemia with volume replacement in patients with coronary artery
disease. Anesth Dev 1985; 12:4954.
34. Parksloe MRJ, Wuld R, Fox M, Reilly CS. Silent myocardial ischemia in a patient
with anaemia before operation. Br J Anaesth 1990; 64:634637.
35. Magovern JA, Sakert T, Benckart DH, et al. A model for predicting transfusion after
coronary artery surgery. Ann Thorac Surg 1996; 61:2735.
36. Goodnough LT, Despotis GJ, Hogue CW, Ferguson TB. On the need for improved
transfusion indicators in cardiac surgery. Ann Thorac Surg 1995; 60:473480.
37. Hannan EL, Kilburn H, Racz M, Shields E, Chassin MR. Improving the outcomes of
coronary artery bypass surgery in New York State. JAMA 1994; 271:761766.
38. Toy PTCY, Strauss R, Stehling L, et al. Predeposit autologous blood for elective
surgery: a multicenter study. NEJM 1987; 316:517520.
39. Goodnough LT, Shafron D, Marcus RE. The impact of preoperative autologous blood
donation on orthopaedic surgical practice. Vox Sang 1990; 59:6569.
40. Goodnough LT, Grishaber JE, Birkmeyer JD, Monk TG, Catalona WJ. Efficacy and
cost-effectiveness of autologous blood predeposit in patients undergoing radical
prostatectomy procedures. Urology 1994; 44:226231.
41. British Committee for Standards in Haematology Blood Transfusion Task Force.
Guidelines for autologous donation: preoperative autologous donation. Transfusion
Med 1993; 3:307316.
42. NHLBI.Transfusion alert: use of autologous blood. Transfusion 1995; 35:703711.
43. Renner SW, Howanitz PJ, Bachner P. Preoperative autologous blood donation in 612
hospitals. Arch Pathol Lab Med 1992; 116:613619.
44. Goodnough LT. The implications of cost-effectiveness for autologous blood procure-
ment. Arch Pathol Lab Med 1994; 118:471472.
45. Surgenor DM, Churchill WH, Wallace EL, et al. Determinants of red cell, platelet,
plasma, and cryoprecipate transfusions during coronary artery bypass graft surgery.
Transfusion 1996; 36:521532.
46. Kickler TS, Spivack JL. Effect of repeated whole blood donations on serum im-
munoreactive erythropoietin levels in autologous donors. JAMA 1988; 260:6567.
47. Goodnough LT, Brittenham G. Limitations of the erythropoietic response to serial
phlebotomy: implications for autologous blood donor programs. Lab Clin Med 1990;
115:2835.
48. Cohen JA, Brecher ME. Preoperative autologous blood donation: benefit or detri-
ment? A mathematical analysis. Transfusion 1995; 35:640644.
49. Goodnough LT, Price TH, Parvin CA, et al. Erythropoietin response to anemeia is not
altered by surgical or recombinant human erythropoietin therapy. Br J Haematol
1994; 87:695699.
50. Goodnough LT, Bravo J, Hsenh Y, Keating L, Brittenham GM. Red blood cell volume
in autologous and homologous blood units: implications for risk/benefit assessment
for autologous blood crossover and directed blood transfusions. Transfusion 1989;
29:821822.
51. Goodnough LT, Price TH, Rudnick S, Soegiarso RW. Preoperative red blood cell
460 Goodnough
production in patients undergoing aggressive autologous blood phlebotomy with and

without erythropoietin therapy. Transfusion 1992; 32:441445.
52. Goodnough LT, Price TH, Parvin CA. The endogeneous erythropoietin response and
the erythropoietic response in blood loss anaemia: the effects of gender and age. J Lab
Clin Med 1995; 126:5764.
53. Goodnough LT, Rudnick S, Price TH, et al. Increased collection of autologous blood
preoperatively with recombinant human erythropoietin therapy. N Engl J Med 1989;
321:11631167.
54. Goodnough LT, Price TH, Friedman KD, et al. A phase III trial of recombinant human
erythropoietin therapy in non-anemic orthopaedic patients subjected to aggressive
autologous blood phlebotomy: dose, response, toxicity and efficacy. Transfusion
1994; 34:6671.
55. Price TH, Goodnough LT, Vogler W, et al. The effect of recombinant erythropoietin
on the efficacy of autologous blood donation in patients with low hematocrits.
56. Mercuriali F, Zanella A, Barosi G, et al. Use of erythropoietin to increase the volume
of autologous blood donated by orthopedic patients. Transfusion 1993; 33;5560.
57. Silberstein LE, Kruskall MS, Stehling LC, et al. Strategies for the review of
transfusion practices. JAMA 1989; 262:19931997.
58. Sazama K. Report of 355 transfusion associated deaths: 1976 through 1985. Trans-
fusion 1990; 30:583590.
59. Stehling L, Luban NLC, Anderson KC, et al. Guidelines for blood utilization review.
Transfusion 1994; 34:438448.
60. Welch GH, Meehan K, Goodnough LT. Prudent strategies for elective red cell
transfusion. Ann Intern Med 1992; 116:393403.
61. Messmer K, Kreimeier M, Intagliett A. Present state of intentional hemodilution. Eur
Surg Res 1986; 18:254263.
62. Goodnough LT, Grishaber J, Monk TG, Catalona WJ. Acute preoperative hemodilu-
tion in patients undergoing radical prostatectomy: a case study analysis of efficacy.
Anesth Analg 1994; 78:932937.
63. Brecher ME, Rosenfled. Mathematical and computer modeling of acute normovole-
mic hemodilution. Transfusion 1994; 34:176179.
64. Weiskopt RB. Mathematical analysis of isovolemic hemodilution indicates that it can
decrease the need for allogeneic blood transfusion. Transfusion 1995; 35:3741.
65. Monk TG, Goodnough LT, Birkmeyer JD, Brecher ME, Catalona WJ. Acute norm-
ovolemic hemodilution is a cost-effective alternative to preoperative autologous
donation in patients undergoing radical retropubic prostatectomy. Transfusion 1995;
35:559565.
66. Winslow RM. Blood substitutes. Curr Status Transfusion 1989; 29:753754.
67. Ray JM, Flynn JC, Bierman AH. Erythrocyte survival following intraoperative
autotransfusion in spinal surgery: an in vivo comparative study and 5-year update.
Spine 1986; 11:879882.
68. Linden JV, Kaplan HS, Murphy MT. Fatal air embolism due to perioperative blood
recovery. Anesth Analg 1997; 84:422-6.
69. Horst HM, Dlugos S, Fath JJ, et al. Coagulopathy and intraoperative blood salvage.
J Trauma 1992; 32(5):646653.
70. Bell K, Stott K, Sinclair CJ, Walker WS, Gillon J. A controlled trial of intraoperative
autologous transfusion in cardiothoracic surgery measuring effect on transfusion
requirements and clinical outcome. Transfusion Med 1992; 2:295300.
71. Solomon MD, Rutledge ML, Kane LE, Yawn HD: Cost comparison of intraoperative
autologous versus homologous transfusion. Transfusion 1988; 28:379382.
72. Bovill DF, Moulton CW, Jackson WS, Jensen JK, Barcellos RW. The efficacy of
intraoperative autologous transfusion in major orthopaedic surgery: a regression
analysis. Orthopedics 1986; 9:14031407.
73. Goodnough LT, Monk TG, Sicard G, et al. Intraoperative salvage in patients
undergoing elective abdominal aortic aneurism repair: an analysis of costs and
benefit. J Vasc Surg 1996; 24:213218.
74. Semkiew LB, Schurman OJ, Goodman SB, Woolson ST. Postoperative blood salvage
using the cell saver after total joint arthroplasty. J Bone J Surg 1989; 71A:823827.
75. Flynn JC, Price CT, Zink WP. The third step of total autologous blood transfusion in
scoliosis surgery. Spine 1991; 16:S328329.
76. Faris PM, Ritter MA, Keating EM, Varleri CR: Unwashed filtered shed blood
collected after knee and hip arthroplasties. J Bone J Surg 1991; 73A:11691177.
77. Martin JW, Whiteside LA, Milliano MT, Reedy ME. Postoperative blood retrieval
and transfusion in cementless total knee arthroplasty. J Arthroplasty 1992; 7:205210.
78. Ayers DC, Murray DG, Duerr DM. Blood salvage after total hip arthroplasty. J Bone
Oint Surg 1995; 77A:13471351.
79. Goodnough LT, Verbrugge D, Marcus RE. The relationship between hematocrit,
blood lost, and blood transfused in total knee replacement: implications for post-
operative blood salvage and reinfusion. Am J Knee Surg 1995; 8:8387.
80. Mylod AG, France MP, Muser DE, Parsons JR. Perioperative blood loss associated
with total knee arthroplasty: a comparison of procedures performed with and without
cementing. J Bone J Surg 1990; 72A:110112.
81. McCarthy JC, Turner RH, Reuten JJ, Valeri CR, Ragno GM, Korten KW. The effect
of cell washing on the quality of shed blood in major reconstructive surgery. Orthop
Trans 1992; 16:484.
82. Clements DH, Sculco TP, Burke SW, Mayer K, Levine DB. Salvage and reinfusion
of postoperative sanguineous wound drainage. J Bone J Surg 1992; 74A:646651.
83. Woda R, Tetzlaff JE. Upper airway oedema following autologous blood transfusion
from a wound drainage system. Can J Anesth 1992; 39:290292.
84. Blevins FT, Shaw B, Valeri RC, Kasser J, Hall J. Reinfusion of shed blood
after orthopaedic procedures in children and adolescents. J Bone J Surg 1993;
75A:363371.
85. Kruskall MS, Umlas J. Acquired immunodeficiency syndrome and directed blood
donations: a dilemma for American medicine. Arch Surg 1988; 123:2325.
86. Thaler M, Shamiss A, Orgad S, et al. The role of blood from HLA-homozygous
donors in fatal transfusion-associated graft-vs-host disease. N Engl J Med 1989;
321:2528.
462 Goodnough
87. Perkins HA, Samson S, Busch MP. How well has self-exclusion worked? Transfusion
1988; 28:601602.
88. Chambers LA, Kruskall MS, Leonard SS, Ellis AM. Directed donor programs may
adversely affect autologous donor participation. Transfusion 1988; 28:645.
89. Goodnough LT. Predeposit of designated blood does not protect against homologous
blood exposure in patients who predeposit autologous blood for elective surgery. Am
J Clin Pathol 1989; 92:484487.
90. Ali S, Goodnough LT. An analysis of a directed donor program in 1995. Blood 1995;
86:852A.
20
Leukoreduction
Anne B. McDonald
Beth Israel Deaconess Hospital, Boston, Massachusetts
Walter H. Dzik
Beth Israel Deaconess Hospital and Harvard Medical School, Boston,
Massachusetts
I. INTRODUCTION
Transfusion of allogeneic donor leukocytes present in cellular blood components

results in a number of both proven and perceived adverse effects in recipients.
Over the last decade, a greater understanding of these possible adverse effects has
been the impetus for improving methods of leukoreduction. Advances in technol-
ogy have allowed more widespread use of leukoreduced components and have
prompted active investigation into the potential benefits of these products. This
chapter will describe the methods, indications for, and efficacy of leukoreduction.
Cellular blood components contain a large number of residual donor
leukocytes (Table 1). The point at which a component can be considered ade-
quately leukoreduced has varied in the past, with the level of 5 106 white blood
cells (WBCs) per unit of red blood cells (RBCs) being established by the
American Association of Blood Banks in 1991 and extension of the standard to
leukoreduced platelets in 1996.
II. METHODS OF LEUKOREDUCTION

A. Centrifugation
The use of centrifugal force to separate blood components by density has been
used for many years. The removal of the buffy coat from whole blood is widely
463
464 McDonald and Dzik
Table 1 Approximate Number of Leukocytes

in Blood Components
Fresh whole blood 109
Red cell concentrate 108109
Buffy coat depleted red cells 108
Washed red cell concentrate 107
Frozen deglycerolized red cells 106107
Platelet concentrate 107108
Apheresis platelets 105108
Fresh frozen plasma <104
practiced. The process of centrifugation separates the least dense leukocytes from
the packed red cells, with more efficient removal of lymphocytes and monocytes
than granulocytes (1). In a variation of this processthe top and bottom
methodwhole blood is centrifuged at relatively high centrifugal force to com-
press both platelets and leukocytes into the buffy coat. The supernatant plasma is
then expressed out of the top of the primary bag into one satellite bag, while the
red cells are removed via the bottom into another. The buffy coat, composed of
donor white cells and platelets, is then pooled with further buffy coats and
centrifuged again at a lower centrifugal force. This allows separation of the
platelet-rich plasma from the white cell component, which is then discarded.
This method removes 7080% of leukocytes, a reduction sufficient to prevent
many febrile nonhemolytic transfusion reactions, but not other complications
whose prevention requires a higher degree of leukoreduction.
A variation of this method involves manipulation of the temperature of
whole blood prior to centrifugation. It was found that if the whole blood was first
cooled and then warmed prior to centrifugation, the removal of leukocytes could
be increased to 90%, with no apparent evidence of damage to red cells or
coagulation factors (2).
Other methods used in the past include washing and use of frozen/thawed
red cells. Washing is not the most efficient way to remove leukocytes because of
the wide variation in the number removed (3). Significant red cell loss may occur.
Because preparation takes place in an open system, washed red cells must
be used within 24 hours. Frozen deglycerolized red cells reduce the risk of
cytomegalovirus (CMV) transmission (4), but, because of the presence of viable
lymphocytes (5), they are inadequate to prevent graft-versus-host disease in
immunodeficient recipients.
Leukoreduction 465
B. Filtration
Advances in filtration technology have made consistent production of leuko-
reduced components a viable option. Table 2 summarizes the changes in filtration
over recent years.
1. Mechanism of Filtration
Removal of leukocytes by filtration results primarily from retention by a barrier.
The interior of most modern leukocyte removal filters consists of layers of
synthetic mesh made of nonwoven fibers. The design is such that the blood
entering the filter encounters a large surface area of the medium. Optimally, the
blood cannot bypass the filtration medium and the amount of blood retained in
the filter is minimal.
Adsorption of leukocytes to the synthetic fibers also contributes to leuko-
reduction. Modification of synthetic surfaces, by increasing wettability to
decrease the surface tension or by alteration in surface charge of the fibers (6),
has improved the performance of filters.
Biological mechanisms also influence leukoreduction. Electron micro-
scopic analysis of filters demonstrates that platelets undergo activation and
spreading on the surface of filter fibers during filtration of red blood cells.
Leukocytes, especially polymorphonuclear leukocytes, adhere to the activated
platelets and are removed (7). Platelets may also contribute by forming micro-
aggregates that trap cells (8).
2. Factors Affecting Filtration Performance

As leukoreduction methods have evolved, the factors that influence the residual
number of leukocytes have become increasingly important as variables that can
Table 2 History of Filters for Blood Transfusion
Generation Removal target Filter material
First Clots Screen filter; pore size:

170240 m
Second Micro-aggregates Screen/depth woven filter;
pore size: 40 m
Third Leukocytes Nonwoven web of microfibers;
fiber diam: 220 m
be controlled to improve the quality of the end product. These factors are
summarized in Table 3.
The capacity of the filter used is of prime importance in the degree of
leukoreduction achieved. Current high-performance filters can reduce the residual
WBC content by 34 logs. The input load of the WBCs to be filtered has a direct
relation to the postfiltration WBC content. Manufacturers guidelines provide
filter capacity, or number of units of RBCs or platelets, for the average WBC load.
The cellular composition of the input product will affect the filtration process in
other ways, too. For example, elevated hematocrit in red cell concentrate may
result in slightly decreased leukocyte removal performance because of inhibited
leukocyte approach to adsorption points (9). The nature of the erythrocyte may
also affect filter performance. Filtration of hemoglobin AS (sickle cell trait) blood
results in poorer WBC removal than does filtration of hemoglobin AA blood
(10,11). It is thought that sickling of the red cells within the filter results in
obstruction of flow, preventing adequate contact of the leukocytes with the
filtration medium. Platelets in red cell concentrates also influence filtration.
Filters have decreased retention for fresh RBCs in the presence of platelets; this
has been attributed to biological interaction between platelets and WBCs that
promotes retention of the latter (8).
The flow rate through the filter is also a factor in leukoreduction; acceler-
ated blood flow (100 mL/min) is associated with decreased reduction (9). This
may result from partial detachment of adsorbed leukocytes due to accelerated
shear stress or from decreased contact time with the filtration medium. Alterna-
tively, slow filtration rate may result in insufficient pressure to ensure adequate
contact of leukocytes with the filtering medium or may allow an increase in
temperature resulting in increased leukocyte deformability (12). Studies have
shown that the efficiency of filtration is improved at refrigerated temperatures
(13). The reason for this may relate to decreased plasticity of leukocytes at lower
temperatures. The medium in which cells are suspended also influences the
efficiency of leukoreduction. The addition of small volumes of plasma to RBCs
Table 3 Factors Affecting Filter Performance

Capacity of the filter
Input number of leukocytes
Flow rate, pressure, priming, rinsing
Temperature, viscosity
Number and function of platelets
Holding time between blood collection and filtration
Erythrocyte and leukocyte deformability
Plasma content of cell suspension media
Leukoreduction 467
suspended in SAGMAN (saline/adenine/glucose/manitol)-additive solution re-

sulted in a marked improvement in filtration efficiency (14). This may result from
adhesive proteins present in plasma contributing to leukocyte retention.
3. Timing of Filtration
Leukocyte reduction can be performed before storage, before issuance from the
blood bank, or at the bedside (Table 4) (15). Prestorage leukoreduction is gaining
acceptance as the use of leukoreduced components becomes more routine. In-line
Table 4 Timing of Filtration
Prestorage In blood bank At bedside
Advantages
May prevent accumula- Red cell units filtered in Conveniencefilter
tion of cytokines of short time span?im- added to intravenous
donor origin during proved leukoreduction. line at time of filtration.
storage.
May decrease risk of bac- Smaller no. of staff per-
terial contamination. forming procedure
less variability.
Leukoreduced units Easier to control quality
readily available on compared with bedside
request. filtration.
Easier to control quality Record of special needs
compared with bedside of particular patients.
filtration.
Disadvantages
Transfusion of Additional labor cost in Filtration occurs slowly
leukoreduced compo- blood bank. (24 h)red cell unit
nents to patients with May not remove leuko- warmsmay lead to
no defined clinical cyte fragments that less efficient filtration.
indication. accumulate during More difficult to institute
storage. quality control
Does not remove cyto- measures.
kines that may Units which fail quality
accumulate during control have already
platelet storage. been transfused.
Cost of filter. More staff involved in
processmore diffi-
cult to institute
standard procedures.
Cost of filter.
filtration (a blood bag system with integrated filter) allows for separation of
components and filtration without the need for opening the system, thereby
avoiding contamination and bacterial overgrowth during subsequent storage.
Prestorage leukoreduction has theoretical advantages of standardization and easy
availability of filtered units. Prestorage leukoreduction may decrease the inci-
dence of febrile nonhemolytic transfusion reactions (FNHTRs) from platelet
transfusion that results from passive transfer of cytokines. In an animal model
(16), the incidence of platelet refractoriness was decreased by prestorage leuko-
reduction. This may have resulted from removal of leukocyte fragments that
may not be removed during poststorage filtration. The issue of the importance
of leukocyte fragments has not been resolved. Some studies in an animal
model indicate a greater degree of platelet refractoriness following transfusion of
plasma supernatant of blood leukoreduced after storage when compared with that
leukoreduced before storage (17). However, it has also been demonstrated
that leukoreduction had no effect on soluble class I human leukocyte antigen
(HLA) substance, and that deliberately prepared leukocyte fragments bearing
HLA antigens did not seem to stimulate an alloimmune response in vitro (18).
There have not yet been any clinical trials in humans to determine the effect
of prestorage versus poststorage leukoreduction on prevention of alloimmuniza-
tion. Therefore, the issue remains unresolved.
The choice of filtration in the laboratory prior to issuance by the blood bank
versus filtration at the bedside would appear to many to be a minor issue in
leukoreduction. However, recent investigations have confirmed that bedside
filtration may be less than optimal in many cases for a variety of reasons. Bedside
filtration carries the advantage that expensive filters are restricted to units
transfused to patients for whom leukoreduction is indicated. However, concerns
have arisen regarding the quality of leukoreduction during bedside filtration.
Ledent and Berlin (12) found a 78% failure rate (>5 106 WBC/unit) when
leukoreduction was performed under conditions mimicking those of bedside
filtration. The failure rate when performed as in a blood bank setting was less
than 1%. The major difference in conditions was flow rate, with the rate at the
bedside very much slower than that in the laboratory. Slow filtration allows time
for warming of the unit, perhaps allowing increased leukocyte deformability and
thus less efficient removal of leukocytes. Deliberate warming of blood to 27C
resulted in a 20-fold increase in the number of leukocytes passing through the
filter (19), adding further weight to the importance of temperature control.
Therefore, the ideal time span in which to filter a unit of red cells would appear
to be slow enough so as not to generate excess shearing forces on the white cells
trapped in the filter, yet fast enough to minimize any potential warming of the
unit. Filtration at a relatively fast flow rate (by gravity over 1015 minutes), rather
than the slow flow at the bedside (over 23 hours), would appear to be the
most appropriate.
Leukoreduction 469
Familiarity with the process of leukoreduction and adequate staff training

may be further variables in the process of leukoreduction. Sirchia et al. (20)
initially found that, in a hematology outpatient setting with staff trained in the use
of bedside filters, the quality of leukoreduced product was not significantly
different from that filtered in the laboratory setting. However, subsequent studies
at the same center (19) found that, even under ideal conditions for bedside
filtration, there was a 5% filtration failure rate. Again, the issue of rise in
temperature during filtration was felt to be important, with improved results
if transfusion was completed within 100 minutes of removal of the blood from
the refrigerator.
In addition to the important issue of quality control, other considerations
may influence the decision concerning when to leukoreduce blood. One issue
investigated recently is the effect of leukoreduction on the likelihood of bacterial
overgrowth of blood products. Investigation of the potential role of leukoreduc-
tion has involved experiments using deliberate inoculation of units. Units of blood
are spiked with varying concentrations of bacteria and then split into pairs, one
of which is then filtered. Comparison of bacterial growth is then made at variable
times. Most RBC experiments have focused on Yersinia enterocolitica, the most
commonly encountered serious bacterial contaminant in RBCs. A significant
reduction in bacterial concentration occurs during the time between inoculation
and filtration (7 hours at room temperature)an effect attributed to the natural
antibacterial properties of blood (21). In addition, it was found that if the
initial inoculation was of low concentration, filtration was successful in prevent-
ing bacterial overgrowth at 42 days. However, overgrowth during storage was
not prevented when higher initial concentrations (greater than 3 colony-forming
units/mL) were used. Whether or not routine prestorage filtration of white cells
may be of benefit in reducing bacterial overgrowth in the clinical setting is not
known. The optimal time to filter is not clear, but it appears there may be a
benefit in an early delay to allow bacteriocidal activity of polymorphonuclear
phagocytes in the first few hours after collection (22). However, the application
of chemical sterilants may, in the future, make discussion of leukoreduction for
this reason unnecessary.
Prevention of the storage lesion has also been proposed as an argument in
favor of widespread use of prestorage leukoreduction. It was initially felt that
leukocyte degeneration during storage would result in release of lysosomal
enzymes that could damage the red cell or platelet. However, comparison studies
have shown little difference in the degree of hemolysis of red cells stored with or
without prestorage leukoreduction, and in vivo survival studies failed to show any
advantage to prestorage filtration. Prestorage filtration of platelet concentrates
also failed to have any major impact on either in vitro measures of effects of
storage or posttransfusion survival of radiolabeled platelets.
C. Low Leukocyte Apheresis Platelets

In recent years, improvements in apheresis technology have led to increased
collection of platelets with decreased concentration of residual donor leukocytes
in the final product (Table 5). Further reduction has been achieved by some
manufacturers through the use of in-line filters in the disposable tubing used in
the process. One large study evaluated the performance of the MCS+ system
(Haemanetics Corporation, Braintree, MA). With the use of an in-line filter, it was
found that 98.3% of 432 collections contained less than 5 106 leukocytes.
Collection variables, such as use of donors with platelet count less than 200,000
per L, lipemic samples, anticoagulant citrate dextrose solution (ACD) reactions,
or red cell contamination, were found to predict those collections in which
leukoreduction was not adequate (23).
Manipulation of the path of the blood flow during apheresis has also been
used by manufacturers to improve separation of cellular elements. In the COBE
Table 5 Methods of Leukoreduction: Advantages and Disadvantages
Method of %
leukoreduction removal Advantages Disadvantages
Centrifugation 7080 Simple inexpensive Variable degree of leuko-

method reduction
Adequate for prevention Inadequate for many of
of FNH transfusion clinical indications for
reactions leukoreduction
Freeze/thaw/wash 9095 Relatively simple method Time-consuming
Adequate for most clini- Significant red cell loss
cal indications for may occur
leukoreductiona Red cells must be trans-
fused within 24 h
because of risk of
contamination
Filtration 99.9 Simple method Expense
Highly efficient Potential red cell/platelet
Adequate for most loss
clinical indications for
leukoreductiona
Low-WBC 99.9 Leukoreduced at source Expense
apheresis Adequate for most
platelets clinical indications for
leukoreductiona
aInadequate for prevention of transfusion-associated graft-versus-host disease.
Leukoreduction 471
BCT Spectra system (Lakewood, CO), a platelet-rich interface is separated from

the denser red cells by centrifugation, the interface then being drawn towards a
separation dam. Platelets, being more buoyant, rise over the dam and continue
to flow in the same direction to the outlet line and platelet collection bag.
Leukocytes, being less buoyant, are drawn in the opposite direction toward the
red cell return line. This counterdirectional flow of platelets improves separation
of these cellular components. In recent times, further modifications have been
made by the same company with the introduction of the Spectra Leukocyte
Reduction System (Spectra LRS) based upon the principle of fluidized particle
bed separation. A cone-shaped chamber is added to the disposable set at the
platelet collection line. The platelet concentrate enters at the narrow base of
the conical chamber. Flow patterns that develop in the widening portion of the
chamber result in slowing of particle velocity. The deceleration is greatest for
heavy particles, such as leukocytes, while lighter particles, such as platelets,
advance to the higher chamber level and subsequently escape to the collection
bag. The width of the conical chamber increases in stepwise gradations. The
design is such that any white cells that advance to a higher level are directed back
into the center of the chamber and away from the outlet by the spinning forces.
The conical chamber narrows at the top, resulting in acceleration of the platelets
prior to entering the outlet tubing. Preliminary results indicate that this method
achieves leukoreduced products without the need for secondary filtration (24).
An alternative method used in the CS3000 series of Baxter Biotech (Round
Lake, IL) involves the initial separation of platelet-rich plasma from the leuko-
cytes and red cells in a separation chamber. The leukocytes and red cells are
returned to the patient, while the platelet-rich plasma enters a second chamber
and is centrifuged with return of the plasma to the patient. The manufacturers
have again used the geometry of the separation chamber to minimize the number
of leukocytes remaining in the platelet-rich plasma. In addition, a system of optics
monitors the density of the platelet-rich plasma and can be adjusted to achieve
consistently leukoreduced platelet collections.
III. COUNTING METHODS FOR LOW

LEUKOCYTE NUMBERS
It is important in terms of the clinical, financial, and technical aspects of
leukoreduced components to verify that the end product meets expectations for
leukocyte content. The low concentration of white cells now achieved in many
products is below the threshold of routine methods of counting cells. In some
circumstances it may be adequate to ensure that a product has passed or failed
in terms of the accepted standard of leukoreduction. However, the lower limit of
leukoreduction at which all the various adverse effects of transfused white cells
can be seen has not yet been determined. As such, development of improved
counting methods has paralleled the development of improved leukoreduction.
This is important in order to evaluate new technology critically and to gain a
better understanding of clinical trials involving leukoreduced components.
Methods to count residual leukocytes have used light or fluorescent micros-
copy, radioimmunoassays, flow cytometry, volumetric capillary cytometry, and
the polymerase chain reaction (PCR) (25). The most widely used method involves
a large volume (50 L) hemocytometer (Nageotte chamber). The sample to be
counted is mixed with a red cell or platelet lysing agent and a nuclear stain,
allowed to settle undisturbed in the counting chamber, and then examined under
200 magnification. When viewed by a trained observer, the method is simple
and has been shown to be accurate to approximately 1 WBC per L (26).
By concentrating a larger volume prior to sampling, the lower limit of detection
of Nageotte-based methods may be reduced even further (27).
Flow cytometry, by which a larger volume of sample can be analyzed,
achieves a lower limit of accurate detection of 0.1 WBC per L. Most flow
cytometric methods stain leukocytes with a fluorescent nuclear stain and analyze
the emission of light. The major disadvantage of this method is the need for
expensive instrumentation.
Other methods that have not yet achieved widespread use include volumet-
ric capillary cytometry and PCR-based counting methods. PCR-based methods
have an acceptable lower limit of detection, but have disadvantages of labor
intensity, equipment cost, and sample contamination.
IV. CLINICAL INDICATIONS FOR LEUKOREDUCTION

A. Prevention of Febrile Nonhemolytic
Transfusion Reactions
Febrile nonhemolytic transfusion reactions (FNHTRs) can be a troublesome and
often frightening aspect of transfusion to both patient and clinician. Modest
reduction in the WBC count was found to be effective in reducing this complica-
tion with RBC transfusion, and prevention of FNHTRs is an established indica-
tion for transfusion of leukoreduced components.
The cause of FNHTRs may be multifactorial, with the final common
pathway being elaboration of inflammatory cytokines that react with receptors in
the thermoregulatory centers of the brain. Proposed mechanisms for FNHTRs
(Table 6) include recipient antibodies reacting with donor leukocytes, release of
inflammatory cytokines by recipient cells in response to antigen-antibody com-
plexes between recipient antibody and donor antigenic material (28), and the
passive transfer of inflammatory cytokines that may accumulate in platelets
during storage (29). Filtration of RBCs has been found to be effective in
Leukoreduction 473
Table 6 Mechanism of FNH Transfusion Reactions
Mechanism Source of cytokine Clinical situation
Classic Donor WBCs Recipient with leukocyte antibody attacks

donor WBCs. Donor cells release cytokine.
Prevented by leukoreduction.
Passive cytokine Donor WBCs Cytokines released in components stored at
room temperature and passively infused
into recipient.
Prevented by leukoreduction prior to storage.
Immune complex Recipient WBCs Recipient with antibody to cells (or proteins)
in donor unit attacks donor material after
transfusion. Resulting immune complex
triggers recipient immune system to
release cytokines.
Leukoreduction confers incomplete
protection.
prevention of FNHTRs in multitransfused patients. Conversely, prevention of

FNHTRs with platelet concentrate transfusion has been found to be more difficult
(30), favoring the hypothesis that passive transfer of cytokines accumulating
during storage is a causative factor. Prestorage leukoreduction of platelet concen-
trates has been found to moderate the increase in cytokine concentration that
occurs with storage (31). Further studies have shown a lower incidence of
reactions to plasma-reduced concentrates compared with poststorage leuko-
reduced units (32).
B. Prevention of HLA Alloimmunization

Formation of HLA alloantibodies has been recognized as one of the causes of
refractoriness to platelet transfusions. Prevention of this immunization has been
attempted with the use of leukoreduced components. It has been shown that the
use of leukoreduced components has decreased the overall incidence of HLA
sensitization among multiply transfused patients (33). Analyses of particular
subgroups of patients have demonstrated in some studies that, for patients
previously exposed to HLA antigens, such as females with history of pregnancy,
the use of leukoreduced components did not as effectively prevent immunization
or delay time to development of refractoriness (34). The National Institutes of
Health Trial to Reduce Alloimmunization to Platelets found that transfusion
support with leukoreduced pooled platelets, leukoreduced apheresis platelets, or
ultraviolet Btreated platelets resulted in significantly less HLA alloimmunization
and platelet refractoriness compared with unmodified pooled platelets. This trial
provides the best clinical evidence to date that leukoreduction prevents HLA
alloimmunization in patients undergoing cytoreductive chemotherapy.
A major issue for consideration is that HLA alloimmunization may be only
a poor surrogate marker for bleeding complications due to platelet refractoriness.
Because HLA alloimmunization is only one cause of refractoriness to platelet
transfusions, its elimination may not necessarily prevent this problem. Further-
more, serious bleeding complications are infrequent even when unmodified
components are used, and it may be difficult to justify the widespread use of a
costly procedure, such as leukoreduction, for a potentially small gain. Analysis of
this issue continues.
C. Prevention of Transmission of Cytomegalovirus and

Other Leukotropic Viruses
The human leukotropic virusescytomegalovirus (CMV), Epstein-Barr virus
(EBV), and human T-cell lymphotropic virus (HTLV-I/II)reside within leuko-
cytes of infected individuals. In terms of pathogenicity, CMV is the most
important and is capable of causing significant complications in immuno-
compromised patients. HTLV-I/II is potentially a serious problem, but screening
measures have minimized the risk of transmission by transfusion. EBV has high
prevalence in the community, but transfusion-transmitted disease has not been a
significant problem.
Early studies demonstrated that leukoreduction could prevent transfusion-
transmitted CMV (35). In one large study, Bowden et al. (36) randomized 502
CMV seronegative patients undergoing bone marrow transplantation to receive
either CMV-seronegative blood components or leukoreduced components from
CMV-unscreened donors. Infection occurring more than 21 days after the day of
transplant was considered related to transfusion. Infections occurring prior to day
21 were considered to be due to infection prior to enrollment in the study.
There was no significant difference between the two study arms in either proba-
bility of infection or survival. However, the occurrence of a small number of
infections in the filtered arm when compared to the complete absence of cases in
the seronegative arm is of concern (Table 7).
Problems identified with this study include the fact that filtration was done
at the bedside with no assessment of the adequacy of leukoreduction and that
the platelet filters used in the early phase (PL50) had a lower performance
rating than filters currently in use. Therefore, it may be that a proportion of the
patients received inadequately filtered units. A further smaller study by van
Prooijen et al. (37) using blood center leukoreduced units showed no evidence of
CMV transmission with filtered units, supporting the current recommendation
that adequately leukoreduced components can be regarded as equivalent to
Leukoreduction 475
Table 7 Filtered Versus Seronegative Study Arms
Filtered Seronegative p-value
Probability CMV disease (d1-100) 2.4% 0% 0.03

Probability CMV disease (>d21) 1.2% 0% 0.25
Deaths attributed to CMV 5 0 0.56
CMV-seronegative, and can be used to prevent CMV transmission to a sero-

negative recipient.
It has been suggested that exposure of seropositive recipients to allogeneic
donor leukocytes may promote activation of latent recipient virus. In vitro studies
have suggested that this effect is not seen with exposure to other allogeneic
cellular elements (38). There has also been analysis of the possibility that there
may be transmission through transfusion of a second strain of CMV to a
seropositive recipient. This phenomenon has been seen with solid organ trans-
plantation (39) but not yet with blood transfusion.
V. CONCLUSION
Leukoreduction has become an important consideration for choosing appropriate
blood components for transfusion. The major disadvantages of leukoreduction are
additional cost and loss of cellular components intended for transfusion. The po-
tential advantages of adequately leukoreduced components have been outlined.
Emphasis on quality control is essential to ensure that leukoreduced components
confer maximal benefit.
REFERENCES
1. Nakajo S, Chiba S, Takahashi TA, Sekiguchi S. Comparison of a top and bottom
system with a conventional quadruple-bag system for blood component preparation
and storage. In: Sekiguchi S, ed. Clinical Applications of Leukocyte Depletion.
Oxford: Blackwell Scientific Publications, 1982:1830.
2. Takahashi TA, Nakajo S, Chiba S, Hoseda M, Sekiguchi S. Cooling and warming
method for preparation of 1-log leukocyte-removed red cell concentrates using the
top and bottom method. In: Hogman CF, ed. Leukocyte Depletion of Blood Compo-
nentsPresent Trends and the Future. Amsterdam: VU University Press, 1994:145
148.
3. Wenz B. Clinical and laboratory precautions that reduce the adverse reactions,
allo-immunization, infectivity and possible immunomodulation associated with ho-
mologous transfusions. Trans Med Rev 1990; 4(suppl 1):37.
4. Tolkoff-Rubin NE, Rubin RH, Keller EE, et al. Cytomegalovirus infection in dialysis
patients and personnel. Ann Intern Med 1978; 89:625628.
5. Kurtz SR, Van Deinse WH, Valeri CR. The immunocompetence of residual leuko-
cytes at various stages of red cell cryopreservation with 40% w/v glycerol in an ionic
medium at 80C. Transfusion 1978; 18:441447.
6. Nishimura T, Kuroda T, Mizoguchi Y, et al. Advanced methods for leukocyte removal
by blood filtration. In: Brozovic B, ed. The Role of Leukocyte Depletion in Blood
Transfusion Practice. Oxford: Blackwell Scientific, 1989:3540.
7. Steneker I, van Luyn MJA, van Wachen PB, et al. Electron microscope examination
of white cell depletion on four leukocyte depletion filters. Transfusion 1992; 32:450
457.
8. Steneker I, Prins HK, Florie M, Loos JA, Biewenga J. Mechanism of leukocyte
depletion of red cell concentrates by filtration. The effect of the cellular composition
of the red cell concentrate. Transfusion 1993; 33:4250.
9. Nishimura T, Oka S, Yamawaki N. Technical aspects of leukocyte depletion. In:
Hogman CF, ed. Leucocyte Depletion of Blood ComponentsPresent Trends and
the Future. Amsterdam: VU University Press, 1994:5158.
10. Mijovic V, Kruse A. Filtration of blood from donors with HbAS: an unexpected
problem. In: Brozovic B, ed. The Role of Leucocyte Depletion in Blood Transfusion
Practice. Oxford: Blackwell Scientific, 1989:4850.
11. Bodensteiner D. White cell reduction in blood from donors with sickle cell trait
(letter). Transfusion 1994; 34:84.
12. Ledent E, Berlin G. Inadequate white cell reduction by bedside filtration of red cell
concentrates. Transfusion 1994; 34:765768.
13. Steneker I, Pieterz RNI, Reesink HW. Leukocyte depletion capacity in relation to
filtration temperature (abstr). Vox Sang 1994; 67(suppl 2):71.
14. Ledent E, Berlin G. Does plasma influence the efficiency of leukocyte filtration
(abstr)? Vox Sang 1994; 67(suppl 2):20.
15. Dzik WH. Leukoreduced components: you can filter now or you can filter later.
Transfus Sci 1992; 13:207210.
16. Engelfreit CP, Diepenhorst P, Gissen MVD, von Riesz E. Removal of leukocytes from
whole blood and erythrocyte suspensions by filtration through cotton wool. Vox Sang
1975; 28:8189.
17. Blajchman MA, Bardossy L, Carmen RA, Goldman M, Heddle NM, Singal DP. An
animal model of allogeneic donor platelet refractoriness: the effect of time of
leukoreduction. Blood 1992; 79:13711375.
18. Dzik S, Szulflad P, Eaves S. HLA antigens on leukocyte fragments and plasma
proteins: prestorage leukoreduction by filtration. Vox Sang 1994; 66:104111.
19. Sirchia G, Rebulla P, Sabbioneda L, Garcea F, Greppi N. Optimal conditions for white
cell reduction in red cells by filtration at the patients bedside. Transfusion 1996;
36:322327.
20. Sirchia G, Rebulla P, Parravicini A, Marangoni F, Cortelezzi A, Stefania A. Quality
control of red cell filtration at the patients bedside. Transfusion 1994; 34:2630.
Leukoreduction 477
21. Buchholz DH, AuBouchon JP, Snyder EL, et al. Removal of Yersinia enterocolitica
from AS-1 red cells. Transfusion 1992; 32:667672.
22. Hogman CF, Gong J, Eriksson L, Hambraeus A, Johansson CS. White cells protect
donor blood against bacterial contamination. Transfusion 1991; 31:612626.
23. Fitzpatrick JE, Orsini F, Pantano J. Residual leukocyte content of plateletpheresis
concentrates produced on the Haemonetics MCS+: evaluation of a method to predict
high post filter white cell counts (abstr). Transfusion 1995; 35(suppl 1):2S.
24. Riggert J, Zingsem J, Fourmel JJ, et al. European multicenter evaluation of the COBE
Spectra LRS leukoreduction system (abstr). Vox Sang 1996; 70(S2):11.
25. Dzik S. Principles of counting low numbers of leukocytes in leukoreduced blood
components. Trans Med Rev 1996;
26. Rebulla P, Dzik WH. Multicenter evaluation of methods for counting residual white
cells in leukocyte depleted red blood cells. Vox Sang 1994; 66:2532.
27. Sadoff BJ, Dooley DC, Kapoor V, Law P, Friedman LI, Stromberg RR. Methods for
counting a 6 log 10 white cell depletion in red cells. Transfusion 1991; 31:150156.
28. Dzik WH. Is the febrile response to transfusion due to donor or recipient cytokine
(letter)? Transfusion 1992; 32:594.
29. Heddle NM, Klama L, Singer J, et al. The role of plasma from platelet concentrates
in transfusion reactions. N Engl J Med 1994; 331:625628.
30. Goodnough LT, Riddell JV, Lazarus H, et al. Prevalence of platelet transfusion
reactions before and after implementation of leukocyte-depleted platelet concentrates
by filtration. Vox Sang 1993; 65:103107.
31. Aye MT, Palmer DS, Giulivi A, Hashemi S. Effect of filtration of platelet concentrates
on the accumulation of cytokines and platelet release factors during storage. Trans-
fusion 1995; 35:117124.
32. Heddle NM, Klama L, Kelton JG, Walker I, Meyer R, Levine M. Investigations of
two interventions to prevent acute reactions to platelets (abstr). Proceedings of the
24th congress of the ISBT, Mukuhari, Japan, 1996, p. 27.
33. Heddle NM. The efficiency of leukodepletion to improve platelet transfusion re-
sponse: a critical appraisal of clinical studies. Transfus Med Rev 1994; 8:1528.
34. Sintnicolaas K, van Marwijk KM, van Prooijen HC, et al. Leukocyte depletion of
random single-donor platelet transfusions does not prevent secondary human leuko-
cyte antigen-alloimmunization and refractoriness: a randomized prospective study.
Blood 1995; 85:824828.
35. Brady MT, Milam JD, Anderson DC, et al. Use of deglycerolized red blood cells to
prevent posttransfusion infection with cytomegalovirus in neonates. J Infect Dis
1984; 150:334339.
86:35993603.
37. van Prooijen HC, Visser JJ, van Oostendorp WR, de Gast GC, Verdonck LF.
Prevention of primary transfusion-associated cytomegalovirus infection in bone
marrow transplant recipients by the removal of white cells from blood components
with high-affinity filters. Br J Haematol 1994; 84:144147.
38. Busch MP, Lee TH, Heitman J. Allogeneic leukocytes but not therapeutic blood
elements induce reactivation and dissemination of latent human immunodeficiency
virus type 1 infection: implications for transfusion support of infected patients. Blood
1992; 80:21282135.
39. 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; 70:187194.
21
Viral Inactivation
Bernard Horowitz
VITEX (V.I. Technologies, Inc.), New York, New York
Ehud Ben-Hur
Consultant in Photomedicine, New York, New York
I. INTRODUCTION
The emergence of human immunodeficiency virus (HIV) as a transfusion-

transmitted virus in the early 1980s caused great concern about the safety of the
blood supply. Since that time the introduction of improved donor screening and
testing has reduced the risk of developing a transfusion-associated HIV infection
in the United States to a very low level. The risk of transmission of hepatitis B
virus (HBV) and hepatitis C virus (HCV) is similarly very low today, although it
is 5- to 10-fold higher than the risk of HIV (1).
For pooled blood products the risk of viral transmission is increased
proportionately to the number of pooled donations: probability of infection =
1-[(1-risk per donor)no.donors]. As a result, between 1979 and 1985, 70% of
patients with severe hemophilia in the United States were infected with HIV,
which contaminated the coagulation factor concentrates used in their treatment
(2), and most adult hemophiliacs are infected with HBV, HCV, or both (3,4).
This situation was the driving force behind the efforts to develop methods for
virus inactivation in blood products. These efforts were successful, and the
emphasis is now on developing methods that will be applicable to blood compo-
nents. The established methods and those still under development will be re-
viewed in this chapter.
479
480 Horowitz and Ben-Hur
II. HISTORICAL BACKGROUND

During World War II there was an unprecedented need for large quantities of
plasma derived from pooled sources. Sterilization efforts to prevent the increase
of hepatitis were focused on the use of short wavelength ultraviolet light (UVC,
254 nm), with initially promising results (5). However, this approach was
abandoned when it became clear that adequate inactivation of viruses in plasma
could be achieved only by UVC doses that inactivated plasma components (6).
Serum albumin, a relatively heat-stable protein that could be stabilized further by
addition of fatty acids, was the only blood product available in a virally sterile
form for a long time. It was heated at 60C for 10 hours, a process that remains
in use today.
The concerns about the spread of hepatitis viruses, and subsequently HIV,
through blood transfusion stimulated development of other virus-inactivation
methods with remarkable success (7). However, these methods have certain
limitations. They are applicable only to plasma and its fractions but not to the
cellular components of blood. In addition, these methods can inactivate effec-
tively lipid-enveloped viruses such as HIV, HBV, and HCV, but some nonenvel-
oped viruses [hepatitis A virus (HAV) and parvovirus B19)] are not completely
inactivated. The last two viruses are of minor concern and remain a challenge for
the future.
In addition to reducing or eliminating the risk of virus transmission,
adopting viral inactivation procedures presents numerous advantages. The win-
dow period will no longer be of concern, and efforts to reduce it by direct
detection of viral nucleic acid will no longer be needed. Errors in testing or the
inadvertent release of blood that tests positive will not result in viral transmission.
Viruses that are not tested for, including new ones, will be eliminated, obviating
the need to introduce new tests. Of course, the virus-inactivation procedure
should not affect the blood component in a way that reduces its therapeutic
activity. In addition, the approach should not pose a health risk from, for example,
new immunogenic structures or toxic residues. The process should also be
efficient with respect to yield and ease of implementation and thus be economical.
In this age of managed care, costs are not a minor consideration.
III. ESTABLISHED METHODS

A. Wet Heat (Pasteurization)
The use of heat in the liquid state (60C, 10 h) is the oldest method of sterilization
and is termed after its inventor, Pasteur. It is based on the sensitivity of most
proteins to heat. Heat-labile proteins in the product to be treated have to be
stabilized prior to pasteurization by addition of low molecular weight solutes,
Viral Inactivation 481
such as sugars, amino acids, and salts (7,8). These additives prevent unwanted
protein denaturation and loss of biological activity. The stabilization of the virus
by the added solute is much lower than that of the clotting factors (Table 1).
Treatment that achieves a sufficiently high level of virus elimination (over 6
log10) results in 6080% recovery of clotting factors. Nonenveloped viruses,
however, which tend to be heat-stable, are killed to a lesser extent by this process.
Other advantages of pasteurization are its ease of implementation in a factory
setting, avoidance of potentially toxic chemicals, and homogeneity of the treat-
ment (i.e., virus is inactivated at the same rate throughout the treated product).
B. Dry Heat
Virus inactivation by heating of lyophilized blood proteins occurs at higher
temperatures and takes a longer time than pasteurization because proteins are
stabilized in the absence of water. HIV can be eliminated from lyophilized Factor
VIII by heating at 68C for 96 hours (10). Recovery of Factor VIII activity can
be high under these conditions (11). However, because of heterogeneity of
lyophilized cakes with respect to solute and moisture content, heating has to be
conducted at 80C for 72 hours to achieve reproducible viral safety (12). Recov-
ery of Factor VIII activity exceeds 90%. As with pasteurization, noneveloped
viruses are more resistant to dry heat than are enveloped viruses, and their titer
(e.g., that of parvovirus) is reduced but not eliminated (13). A particular advantage
of dry heating is that it can be performed in the final container, eliminating the
possibility of posttreatment recontamination.
C. Solvent-Detergent
The use of the solvent tri(n-butyl)phosphate (TNBP), typically with Tween 80 or
other detergents, disrupts the viral lipid envelope with concomitant inactivation
of lipid-enveloped viruses (14). The treatment is rapid (46 h at 2430C), plasma
Table 1 Inactivation Velocity, K, of Lipid-Enveloped

Viruses at 60C With or Without Stabilization
K (ln/hr)
Virus A. Stabilized AHF B. Buffer B/A
VSV 1.61 175 109

Sindbis 0.30 230 769
Sendai 0.74 276 375
Source: Adapted from Ref. 9.
proteins are not affected, and recovery of clotting factors can reach 100%.
The added solvent and detergent are removed after treatment, either simply in the
course of purifying the protein or with hydrophobic chromatography using a C18
resin. The safety of solvent-detergent (SD)treated blood products with respect
to HBV, HCV, and HIV is supported by studies in chimpanzees, 14 independent
clinical trials, and the preparation of HIVIG, a hyperimmune gamma globulin to
HIV prepared from HIV-infected donors (1517). A summary of viral safety from
clinical trials conducted with products treated by these methods is given in
Table 2. Advantages of SD treatment include its ease of implementation in a
factory setting and its very high level of virucidal action under conditions where
virtually all proteins are unaffected.
The excellent safety record of coagulation factors treated with solvent-
detergent encouraged the development of SD-treated plasma as a substitute for
fresh frozen plasma (FFP) (18). The procedure involves pooling of up to 2500
units of FFP, treatment with 1% TNBP and 1% Triton X-100 at 30C for 4 hours,
and removal of the reagents by hydrophobic chromatography. The final product
is sterile-filtered and frozen. Under these conditions, the rate of inactivation of
the model lipid-enveloped viruses, vesicular stomatitis virus (VSV) and Sindbis
virus, exceed those observed with Factor VIII concentrates. In addition, more than
106 infectious doses (ID50) of HBV, more than 105 ID50 of HCV, and more than
107.2 ID50 of HIV are killed. Approximately 15% of donor units contain anti-HAV
antibody. We have shown that more than 104.5 ID50 of HAV are neutralized in the
process (19). Because of pooling, SD-treated plasma has 30-fold more anti-HAV
antibody than intramuscular immune globulin, which is known to prevent HAV
spread, and has approximately the same quantity of antiparvovirus antibody as
intravenous immune globulin, a preparation used therapeutically to treat parvo-
Table 2 Safety of Virally Sterilized Coagulation Factor Concentrates
No. of patients infected/No. treated

Quantity tested
Method (U 106) HBV HCV HIV
Wet heat (60C, 10 h) 18.8 2/? 2/95 0/237

Dry heat (80C, 72 h) 0.1 0/16 0/32 0/32
Dry heat and vapor 1.1 4/46 0/70 0/110
(60C, 10 h)
Solvent-detergent 17.6 0/55 0/449 0/524
aResults from patients with hemophilia who received standard therapy with coagulation factor
concentrates virally sterilized by the indicated method. Following infusion, patients were monitored
for 612 months using standard serological assays.
Source: Data from Refs. 15, 17.
virus infections in immunocompromised patients. The coagulation factor content

is similar to that of the start pool (20,21) and is more consistent than that found
in individual donor units. There is no activation of coagulation factors during
treatment, and the level of other proteins is normal. Toxicological studies indicate
that the tiny amounts of solvent and detergent that remain (below 3 ppm) are safe.
SD-treated plasma has been extensively evaluated in the United States and
Europe (2224) and has been approved for use in most European countries, the
United States, and Canada. In the United States, more than 20 clinical study sites
took part in the clinical trials that preceded licensure. The principal efficacy
endpoints were the correction of coagulation factor deficiencies and the treatment
of thrombotic thrombocytopenic purpura. SD-treated plasma behaved like FFP in
all cases. Enveloped viruses have not been transmitted in studies cited above or
in more than 4 million units of SD-treated plasma infused to date. Studies being
conducted in the United States indicated that parvovirus, a nonenveloped virus
found in approximately one of 1000 blood donors, could be transmitted by SD
plasma despite the presence of antiparvovirus antibody in the unit being trans-
fused. Consequently, in the United States, SD plasma is now tested for parvovirus
by polymerase chain reaction (PCR).
IV. NEW APPROACHES

A. Plasma and Blood Proteins
Additional techniques for virus inactivation are being studied because the estab-
lished methods do not eliminate completely the nonenveloped and heat-resistant
HAV and parvovirus B19. Given the extensive history of safety achieved by
currently employed methods and the limitations of laboratory and preclinical
virus validation studies, it is likely that any new procedure will be added to rather
than replace existing processes. Combining methods that act by independent
mechanisms has the advantage of inactivating a broader spectrum as well as a
higher quantity of viruses.
1. Antibody Affinity Purification

During purification of Factor VIII with immunoaffinity column chromatography,
HIV infectivity is reduced 104-fold. This approach has been combined with either
SD (24) or heat treatment (25) for enhanced safety.
2. Nanofiltration
Viruses can be removed by newly developed filters. The term nanofiltration
comes from the ability of the process to remove viruses as small as 15 nm (26,27).
The use of nanofiltration has the advantage of easy addition to existing processes.
However, 35 nm filters do not remove HAV or parvoviruses. Although 15 nm

filters are effective for this purpose (28), recovery of high molecular weight
proteins is unacceptably low. In addition, the question of manufacturing consis-
tency of each filter needs to be addressed. As a result, nanofiltration would be
limited to sterilization of only some blood-derived proteins.
3. UVC Light Irradiation

UVC light targets the viral nucleic acid, producing photochemical modifications
of the pyrimidines (29). As a result, a wide variety of viruses are inactivated
irrespective of the nature of their envelope. Viruses containing single-stranded
nucleic acids are more sensitive. In addition, sensitivity is correlated with the size
of the nucleic acid (30). The former is due to the inability to repair damage in the
absence of a complementary strand, while the latter reflects the fact that a larger
target is hit more often.
Photochemical modification of nucleic acids by UVC proceeds via direct
reactions of the excited pyrimidines, whereas damage to proteins involves free
radical reactions. This difference in mechanism has been used to enhance the
specificity of virus inactivation by UVC in protein solutions by adding scavengers
of free radicals. The most effective scavenger found so far is the plant flavonoid
rutin, an efficient quencher of reactive oxygen species (ROS). Rutin has no effect
on the inactivation kinetics of various viruses but protects several coagulation
factors against UVC-induced damage (31). Rutin also protects fibrinogen, albu-
min, and IVIG against UVC irradiation of fibrin sealant for viral inactivation
(32,33). Fibrin sealant sterilized with SD and UVC is now in clinical trials
evaluating its role in hemostasis and wound healing during surgery. It has been
concluded that addition of UVC treatment to existing processes used in the
manufacture of blood derivatives will provide an added margin of safety. This is
especially true with respect to nonenveloped viruses.
4. Starch-Bound Iodine
Iodine is a strong oxidizing agent and, as a result, is a powerful microbicide.
However, in its free form iodine is not sufficiently selective. When bound to
polymers such as polyvinylpyrrolidone (34) and in particular crosslinked starch
(35), the virucidal action of iodine is more controlled. Thus, starch-bound iodine
at a concentration of 1.05 mg/mL resulted in more than 7 log10 inactivation of
model lipid enveloped and nonenveloped viruses, while more than 70% of
clotting factors activity in plasma was retained (35). Additional research is needed
to determine the efficacy of crosslinked starch-iodine on human pathogenic
viruses, as well as the effect on plasma proteins.
5. Methylene Blue and Visible Light

Methylene blue (MB) is a photosensitizer, i.e., in conjunction with light it can
inactivate biological systems. Because the presence of oxygen is required, MB is
a photodynamic agent. The virucidal action of MB is well known (36), but the
mechanism of action is not entirely clear. Nucleic acid damage is usually
produced as a result of MB photosensitization but was ruled out as the cause of
virus kill in one case (37) but not in others (38). Recently, a procedure has been
developed in which individual plasma units are treated with 1 M MB and white
fluorescent light for one hour at 60,000 lux (39). The individual units are refrozen
and stored for later use. Model enveloped viruses and cell-free HIV are inacti-
vated effectively; cell-associated HIV and nonenveloped viruses are less affected
(40,41). Complete virus studies, including hepatitis viruses and a demonstration
that infectious units can be rendered noninfectious, have yet to be reported.
The advantage of this approach compared with SD treatment is that pooling is
not required (i.e., recipients would received plasma from individual donations,
rather than from a plasma pool made from hundreds or thousands of donations).
On the other hand, treatment of individual units does not allow for careful
monitoring and control of procedures as can be achieved with plasma pools
processed in a factory setting.
MB photodynamic treatment of plasma resulted in no adverse reactions in
a controlled clinical study (42), and neoantigens were not produced in the treated
plasma (43). The in vitro coagulation capacity of MB-treated plasma is reduced
mainly because of reduced fibrinogen and Factor VIII activity (44). It was
therefore recommended that such plasma not be used for patients with severely
reduced ability for synthesizing clotting factors (44). Recently it has been
reported that MB is mutagenic in a cultured mammalian cell system (45).
Considerable investigation will be required, therefore, to assess the genotoxic
potential of MB-treated plasma prior to its clinical use in the United States. This
may be the reason that MB-treated plasma has been withdrawn from the market
in Germany.
B. Platelet Concentrates
The cellular components of blood are more difficult to sterilize than protein
solutions, because cell structure and function are disrupted more easily than
protein structure and function. In addition, infectious virus or its nucleic acid can
be harbored intracellularly. The challenge is eased somewhat because red blood
cells (RBCs) and platelets lack a nucleus and are nonreplicating. A decontaminat-
ing process must leave cellular function intact during both the treatment period
and subsequent storage. The storage period for platelets is 5 days, and the critical
functions require adhesion to subendothelial matrix proteins, aggregation, and
secretion of intracellular organelles. After transfusion, sufficient numbers of

platelets should persist in the circulation. In vitro measures of platelet function,
such as aggregation response to agonists, are most commonly used to assess
platelet concentrate quality. However, in vitro assays do not adequately predict
posttransfusion platelet recovery and survival in vivo (46).
The use of psoralens and UVA light (PUVA) is a promising approach for
inactivation of pathogenic organisms in platelet concentrates and is now in
clinical trials. The ability of PUVA to target nucleic acids is an obvious advantage
for sterilizing platelets, which lack a nucleus. Psoralens preferentially bind to
nucleic acids in the dark and upon exposure to UVA light form adducts with the
pyrimidines, which effectively inhibit nucleic acid replication, transcription, and
translation (47). As a result, PUVA inactivates not only pathogens but also
leukocytes. The inactivation of the latter is beneficial, since transfused leuko-
cytes may lead to alloimmunization (48), nonhemolytic febrile transfusion reac-
tions (49), and graft-versus-host disease (50). In addition to covalent binding
to nucleic acids upon exposure to UVA light, psoralens can also produce reac-
tive oxygen species (ROS), such as singlet oxygen, and induce photodamage
in lipids and proteins. Platelet damage can, therefore, occur under treatment
conditions that result in greater than 6 log10 virus inactivation. This problem has
been dealt with by adding the plant flavonoid rutin as a quencher of ROS to
eliminate photodynamic damage during treatment with 4-aminomethyl-4,5,8-
trimethylpsoralen (AMT) and UVA (51). Others use psoralens with reduced
photodynamic activity (52).
The advantages of AMT over other psoralens as an agent for virus inacti-
vation are that (a) it is water soluble and (2) because it is cationic, it binds more
tightly to nucleic acids and is highly effective for photoinactivation of single-
stranded RNA viruses. The disadvantage of AMT is that it is mutagenic in the
dark after metabolic activation with some of the Ames tester strains (53). To
circumvent this potential problem with the clinical use of AMT, we developed a
procedure to remove it after light exposure. The method employs a hydrophobic
resin (C18), which adsorbs >99% of AMT and is effective in reducing mutagen-
icity below detection level without affecting platelets aggregation response (53).
In addition to quenchers, the exclusion of the shorter UVA wavelengths ( < 340
nm) is also helpful in enhancing the specificity of platelet decontamination by
PUVA (54).
In addition to viruses, PUVA can inactivate bloodborne parasites (55) and
bacteria (56) in platelet concentrates. Because the risk of bacterial contamination
is currently the reason for limiting the storage of platelet concentrates to 5 days,
inactivation of bacteria may extend the allowable storage time to 7 days. It should
be noted that the genomes of parasites and bacteria are usually inactivated at
lower doses of PUVA. It is also important to stress that treatment of platelet
concentrates with AMT-UVA appears to result in fully functional platelets in vivo

under conditions resulting in inactivation of free and cell-associated HIV (57,58).
Other psoralens being studied for use in decontamination of platelet con-
centrates include brominated psoralens, which were claimed to possess improved
efficiency and selectivity for viral inactivation (59). Psoralens with undisclosed
structure are reported to be highly virucidal, nonmutagenic, and lacking photo-
dynamic activity (52). In the absence of published data, the latter claims are
difficult to evaluate, but one of these psoralens, termed S59, is in clinical trials.
C. Red Blood Cell Concentrates

The use of PUVA is not applicable for the sterilization of RBCs because of the
strong absorption of UVA light by hemoglobin. Only red light ( > 600 nm) can
effectively penetrate RBCs, and for this reason sensitizers that absorb maximally
in the red are being studied. These compounds do not target the viral nucleic acid
and are therefore less specific than PUVA. Their virucidal action requires oxygen;
they are thus defined as photodynamic agents. Many classes of photosensitizers
have been tested over the years for their virucidal activity; however, only a few
are being seriously studied for RBC sterilization.
1. Methylene Blue
The absorption maximum of methylene blue (MB), 665 nm, is favorable for
sterilization of RBCs, and there is some clinical experience with its use for ster-
ilizing FFP (see above). However, there are problems associated with the use of
MB in RBCs. The main problem is the lack of inactivation by MB of cell-
associated HIV (40). Moreover, MB can photosensitize induction of HIV in
latently infected cells (60), and at a dose range in which sufficient virus elimina-
tion is achieved, the treatment causes RBC membrane damage (61). Other MB
derivatives are being studied that may circumvent these problems (62).
2. Benzoporphyrin Derivative
Benzoporphyrin derivative (BPD) is a photosensitizer undergoing clinical trials
for photodynamic treatment of skin cancer and other indications. BPD has an
absorption band at 692 nm and at concentrations of 24 g/mL plus 57 J/cm2 was
able to inactivate both cell-free and cell-associated HIV in whole blood (63).
There was only minimal hemolysis during 2 days of storage. Interestingly,
ziduvidine-resistant and -sensitive strains of HIV appear to be equally sensitive
to BPD photoinactivation. More work is required to evaluate RBC quality
following this virucidal treatment.
3. Hypericin
Initial work with hypericin suggested that this plant pigment may be an anti-HIV
agent (64). However, later studies indicate that most of the antiretroviral activity
of hypericin is light-dependent (65). This is in agreement with a large body of
literature on hypericin as a photodynamic agent. Maximal absorption of hypericin
occurs at 590 nm, and it is therefore not ideal for sterilization of RBCs. Even so,
there are efforts to optimize its virucidal potential by studies of structure-activity
relationships of several hypericin derivatives (66).
4. Phthalocyanines
Arguably, these are the most promising photosensitizers for sterilization of RBCs.
Phthalocyanines are porphyrin-like synthetic dyes. Because of their expanded
macrocycle, they absorb intensely at 660700 nm. Aluminum phthalocyanine
(AIPc) and its sulfonated derivatives are effective in photosensitizing inactivation
of lipid-enveloped viruses, including HIV (67). Other phthalocyanines were
shown to inactivate nonenveloped viruses (68). While tetrasulfonate AIPc caused
the least RBC damage under virucidal conditions, there was a need to eliminate
residual damage to erythrocytes, evidenced by reduced circulatory survival (69).
This was achieved by adding quenchers of ROS such as mannitol, glutathione,
and trolox (6971) prior to light exposure. While protecting RBCs, these quench-
ers had no effect on virus inactivation.
Other ways to reduce RBC damage while maintaining the virucidal potency
of phthalocyanines are the use of an appropriate light source to achieve high
irradiance (72) at a selective wavelength (73) and a special delivery vehicle (74).
These additional procedures to increase the specificity of the treatment were
required following the observation that inactivation of HIV in all its forms, as
well as inactivation of bloodborne parasites, was achieved only with the silicon
phthalocyanine Pc 4 (7577). Pc 4, however, caused more RBC damages in the
absence of these special precautions.
When RBCs are treated with Pc 4 and red light, taking into consideration
all of the above, virus sterilization can be achieved with little or no hemolysis
during storage (78). In vivo circulatory survival of rabbit RBCs is also close to
normal. Toxicological studies of Pc 4 are underway prior to evaluation of this
procedure in clinical trials.
5. Inactine
Inactine has recently been reported to inactivate both enveloped and nonenvel-
oped viruses. RBCs were reported to store well following treatment (79). The
structure of inactine has not bee disclosed, but it is said to covalently modify DNA
and to be mutagenic prior to but not following neutralization.
V. CONCLUSIONS
The inactivation of viruses in blood proteins and plasma has made the transfusion
of these products absolutely safe with respect to transmission of HBV, HCV, and
HIV. The addition of nanofiltration, where applicable, or UVC irradiation to the
currently established methods (heat and SD treatment) should make these prod-
ucts safe also with respect to HAV and parvovirus B-19, the two nonenveloped
viruses reported to be transmitted by plasma derivatives. The remaining challenge
is the sterilization of red cell and platelet concentrates. The use of photosensitizers
for this purpose appears to be promising. PUVA is in clinical trials for platelets,
and Pc 4 is about to enter clinical trials for red cells. Table 3 summarizes the
Table 3 Comparison of the Approaches for Virus Inactivation in Blood
Blood
Approach component Advantages Disadvantages
Wet heat Purified Convenientall viruses Protein activity recov-

proteins are susceptible ery is medium
Plasma Nonenveloped virused
killed to a lesser extent
Plasma must be pooled
Dry heat Purified Can be performed in the Nonenveloped viruses
proteins final container not completely killed
High protein recovery
Solvent-detergent Purified Enveloped viruses very Nonenveloped viruses
proteins sensitive not inactivated
Plasma Recovery of protein Plasma must be pooled
activity is close to
100%
Nanofiltration Purified Easy to add to existing Limited to proteins of
proteins process lower molecular weight
UVC light Purified Inactivates all virus types Rutin must be added to
proteins protect protein activity
Plasma Specialized equipment
required
Photosensitization Plasma Compatible with the ster- Not yet commercially
ilization of cellular available, except for
components methylene blue for
plasma in Germany
and Switzerland
Platelets Plasma does not have to Efficacy for viruses yet
be pooled to be proven
Red cells
relative merits of these methods. Since none of them is perfect, currently

employed screening methods are unlikely to be discontinued. On the other hand,
adoption of virus inactivation procedures may make the addition of new screening
tests unnecessary.
ACKNOWLEDGMENT
This work was supported in part by grant No. 2 RO1-HL 412221 from the
National Heart, Lung and Blood Institute.
REFERENCES
2. Goedert JJ, Kessler CM, Aldort LM, Biggar RJ, Andes WA, White GC, Drummond
JE, Vaidya K, Mann DL, Eyster ME, Ragni MV, Lederman MM, Cohen AR, Bray
GL, Rosenberg PS, Friedman RM, Hilgartner MW, Blattner WA, Kroner B, Gail MH.
A prospective study of human immunodeficiency infection and the development of
AIDS in subjects with hemophilia. N Engl J Med 1989; 321:11411148.
3. Brettler DB, Alter HJ, Dienstag JL, Frosberg AD, Levine PH. The prevalence of
antibody to hepatitis C in a cohort of hemophilic patients. Blood 1990; 76:254.
4. Troisi CL, Hollinger FB, Hoots WK, Contant C, Gill J, Ragni M, Parmley R, Sexauer
C, Gomperts E, Buchanan G, Schwartz B, Adair S, Fields H. A multi-center study of
viral hepatitis in a United States hemophilic population. Blood 1993; 81:412428.
5. Oliphant JW, Hollander A. Homologous serum jaundice: experimental inactivation
of etiologic agent in serum by ultraviolet irradiation. Public Health Rep 1946;
61:598600.
6. Murray R, Oliphant JW, Tripp JT, Hampi B, Ratner F, Diefenbach WCL. Effect of
ultraviolet irradiation on the infectivity of ictrogenic plasma. J Am Med Assoc 1955;
157:814.
7. Hilfenhaus J, Herrman A, Mauler R, Prince AM. Inactivation of the AIDS-causing
retrovirus and other human viruses in antihemophilic plasma protein preparations by
pasteurization. Vox Sang 1986; 50:208211.
8. Hilfenhaus J, Weidmann E. Pasteurization as an efficient method to inactivate blood
borne viruses in factor VIII concentrates. Arzneimittelforschung 1986; 36:621625.
9. Horowitz B, Weibe ME, Lippin A, Vandersand J, Stryker MH. Inactivation of viruses
in labile blood derivatives. II. Physical methods. Transfusion 1985; 25:523527.
10. Levy JA, Mitra GA, Wong MF, Mozen MM. Inactivation by wet and dry heat of
AIDS-associated retroviruses during factor VIII purification from plasma. Lancet
1985; 1:14561457.
11. Hollinger B, Dolana G, Thomas W, Gyorkey F. Reduction in risk of hepatitis
transmission by heat treatment of a factor VIII concentrate. J Infect Dis 1984;
150:250262.
12. Winkelman L, Owen NE, Evans DR, Evans H, Haddan ME, Smith JK, Prince PJ,
Williams JD, Lane RS. Severely heated therapeutic factor VIII concentrate of high
specific activity. Vox Sang 1989; 57:97103.
13. Mannucci PM. Clinical evaluation of viral safety of coagulation factor VIII and IX
concentrates. Vox Sang 1993; 64:197203.
14. Horowitz B, Wiebe ME, Lippin A, Stryker MH. Inactivation of viruses in labile blood
derivatives. I. Disruption of lipid-enveloped viruses by tri(n-butyl) phosphate deter-
gent combinations. Transfusion 1985; 25:516522.
15. Horowitz B, Prince AM, Hamman J, Watklevicz C. Viral safety of solvent/detergent-
treated blood products. Blood Coagul Fibrinol 1994; 5(suppl):S21S28.
16. Prince AM, Horowitz B, Baker L, Shulman RW, Ralph H, Valinsky J, Cundell A,
Brotman B, Boehle W, Rey F, Piet M, Reesink H, Lelie N, Tersmette M, Miedema
F, Barbosa L, Nemo G, Nastala CL, Allan JS, Lee DR, Eichberg JW. Failure of human
immunodeficiency virus (HIV) immune globulin to protect against experimental
challenge with HIV. Proc Natl Acad Sci USA 1988; 85:69446948.
17. Horowitz B. Inactivation of viruses found with plasma proteins: In: Goldstein J, ed.
Biotechnology of Blood. Stoneham: Butterworth-Heinemann, 1991:417430.
18. Horowitz B, Bonomo R, Prince AM, Chin SN, Brotman B, Shulman RW. Solvent/
detergent-treated plasma: a virus inactivated substitute for fresh frozen plasma. Blood
1992; 79:826831.
19. Lemon SM, Murphy PC, Smith A, Zou J, Hamman J, Horowitz B. Removal/neutral-
ization of hepatitis A virus during manufacture of high purity, solvent/detergent
treated plasma. Blood 1992; 79:826831.
20. Piquet Y, Janvier G, Selosse P, Doutremepuich C, Journeau J, Nicolle G, Platel D,
Vezon G. Virus inactivation of fresh frozen plasma by a solvent-detergent procedure:
biological results. Vox Sang 1992; 63:251256.
21. Hellstern P, Sachse H, Schwinn H, Overfrank K. Manufacture and in vitro character-
ization of solvent/detergent-treated human plasma. Vox Sang 1992; 63:178185.
22. Solheim BG, Svennevig JL, Mohr B, et al. The use of Octasplas in patients
undergoing open heart surgery. In: Muller-Berghanns G, et al., eds. DIS: Pathogene-
sis, Diagnosis and Therapy of Disseminated Intravascular Fibrin Formation. Amster-
dam: Elsevier Science Publishers, 1993:253262.
23. Moake J, Chintagumpala M, Turner J, McPherson P, Nolasco L, Steuber C, Santiago-
Borrero P, Horowitz M, Pehta J. Solvent/detergent-treated plasma suppresses shear-
induced platelet aggregation and prevents episodes of thrombotic thrombocytopenic
purpura. Blood 1994; 84:490497.
24. Pehta JC. Clinical studies with solvent-detergent-treated products. Transfusion Med
Rev 1996; 10:303311.
25. Piszkeiwicz D, Sun CS, Tondreau SC. Inactivation and removal of human
immunodeficiency virus in monoclonal purified antihemophilic factor (human)
(Hemophil M). Thromb Res 1989; 55:627634.
26. Hamamoto Y, Harada S, Kobayashi S, Yamaguchi K, Iijima H, Manabe S, Tsurumi
T, Aizawa H, Yamamoto N. A novel method for removal of human immunodeficiency
virus: filtration with porous polymeric membranes. Vox Sang 1989; 56:230236.
27. Burnouf-Radosevich M, Appourchaux P, Huart JJ, Burnouf T. Nanofiltration, a new
specific virus elimination method applied to high purity factor IV and factor XI
concentrates. Vox Sang 1994; 67:132138.
28. Horowitz B, Zou J, Hamman J, Gottlieb P, Chin S, Marx G, Golyakhovsky V.
Elimination of small, heat and solvent/detergent (S/D)-resistant viruses in factor VIII
concentrate and fibrin sealant. Throm Haemost 1995; 73:1460.
29. Patrick MH, Rahn RO. Photochemistry of DNA and polynucleotides: photoproducts.
In: Wang SY, ed. Photochemistry and Photobiology of Nucleic Acids, Vol. 2. New
York: Academic Press, 1976:3595.
30. Rauth AM. The physical state of viral nucleic acid and the sensitivity of viruses to
ultraviolet light. Biophys J 1965; 5:257273.
31. Chin S, Williams B, Gottlieb P, Margolis-Nunno H, Ben-Hur E, Hamman J, Jin R,
Dubovi E, Horowitz B. Virucidal UVC treatment of plasma and factor VIII concen-
trate: protection of proteins by antioxidants. Blood 1995; 86:43314336.
32. Marx G, Mou X, Freed R, Ben-Hur E, Yang C, Horowitz B. Protecting fibrinogen
with rutin during UVC irradiation for viral inactivation. Photochem Photobiol 1996;
63:541546.
33. Chin S, Jin R, Wang XL, Hamman J, Marx G, Mou X, Andersson I, Lindquist LO,
Horowitz B. Virucidal treatment of blood protein products with UVC. Photochem
Photobiol 1997; 65:432435.
34. Highsmith FA, Caple M, Walthall B, Shanbron E, Drohan WN. Viral inactivation
of vesicular stomatitis virus in normal human serum by crosslinked polyvinyl-
pyrrolidone. J Infect Dis 1993; 167:10271033.
35. Highsmith FA, Xue H, Caple M, Walthall B, Drohan WN, Shanbrom E. Inactivation
of lipid-enveloped and non-lipid-enveloped model viruses in normal human plasma
by crosslinked starch-iodine. Transfusion 1994; 34:322327.
36. Wallis C, Melnick JL. Photodynamic inactivation of animal viruses: a review.
Photochem Photobiol 1965; 4:159170.
37. Specht K. The role of DNA damage in PM2 viral inactivation by methylene blue
photosensitization. Photochem Photobiol 1994; 59:506514.
38. Abe H, Wagner SJ. Analysis of viral DNA, protein and envelope damage after
methylene blue, phthalocyanine derivative or merocyanine 540 photosensitization.
Photochem Photobiol 1995; 61:402409.
39. Lambrecht B, Mohr H, Knuver-Hopf J, Schmitt H. Photoinactivation of viruses in
human fresh plasma by phenothiazine dyes in combination with visible light. Vox
Sang 1991; 60:207213.
40. Wagner SJ, Robinette D, Storry J, Chen XY, Shumaker J, Benade L. Differential
sensitivities of viruses in red cells suspensions to methylene blue photosensitization.
Transfusion 1994; 34:521526.
41. Mohr H, Lambrecht B, Selz A. Photodynamic virus inactivation of blood compo-
nents. Immunol Invest 1995; 24:7385.
42. Wieding JU, Neumeyer H. Erste Erfahrungen mit Methylenblau-virus inaktiviertem
Fresh-frozen-plasma: Ergebnisse einer klinischen und einer in vitro Studie. Infusion
Ther 1992; 19:8490.
43. Mohr H, Knuver-Hopf J, Lambrecht B, Scheidecker H, Schmitt H. No evidence for
neoantigens in human plasma after photochemical virus inactivation. Ann Hematol
1992; 65:224228.
44. Zeiler T, Riess H, Wittman G, Hintz G, Zimmerman R, Muller C, Geuft HG, Huhn
D. The effect of methylene blue phototreatment on plasma proteins and in vitro
coagulation capability of single-donor fresh-frozen plasma. Transfusion 1994;
34:685689.
45. Wagner SJ, Cifone MA, Murli H, Dodd RY, Myhr B. Mammalian genotoxicity
assessment of methylene blue in plasma: implications for virus inactivation. Transfu-
sion 1995; 35:407413.
46. Murphy S, Rebulla P, Bertolini F, Holme S, Moroff G, Snyder E, Stromberg R. In
vitro assessment of the quality of stored platelet concentrates. Transfusion Med Rev
1994; 8:2936.
47. Ben-Hur E, Song PS. Photochemistry and photobiology of furocoumarins
(psoralens). Adv Radiat Biol 1984; 11:131171.
48. Kao KJ. Effects of leukocyte depletion and UVB irradiation on allogenicity of major
histocompatibility antigens in platelet concentrates: a comparative study. Blood 1992;
80:29312937.
49. Heddle NM, Klama L, Singer J, Richards C, Fedak P, Walker I, Kelton JG. The role
of plasma from platelet concentrates in transfusion reactions. N Engl J Med 1994;
331:625628.
50. Lehmann PV, Shumm G, Moon D, Hurtenbach U, Falcioni F, Muller S, Zagy ZA.
Acute lethal graft-versus-host reaction induced by major histocompatibility complex
class II reactive T helper cell clone. J Exp Med 1990; 171:14851496.
51. Margolis-Nunno H, Robinson R, Ben-Hur E, Horowitz B. Quencher enhanced
specificity of psoralens photosensitized viral inactivation in platelet concentrates.
Transfusion 1994; 34:802810.
52. Lin L, Cook DN, Wiesehahn GP, Alfonso R, Behrman B, Cimino GD, Corten L,
Damonte PB, Dikeman R, Dupuis K, Fang YM, Hanson CV, Hearst JE, Lin CY,
Londe HF, Metchette K, Nerio AT, Pu JT, Reames AA, Rheinschmidt M, Tessman J,
Isaacs ST, Wollowitz S, Corash L. Photochemical inactivation of viruses and bacteria
in platelet concentrates by use of a novel psoralen and long wavelength ultraviolet
light. Transfusion 1997; 37:423435.
53. Margolis-Nunno H, Robinson R, Ben-Hur E, Chin S, Orme T, Horowitz B. Elimina-
tion of potential mutagenicity in platelet concentrates that are virally inactivated with
psoralens and ultraviolet A light. Transfusion 1995; 35:855862.
54. Margolis-Nunno H, Robinson R, Horowitz B, Geacintov NE and Ben-Hur E.
Psoralen-mediated virus photoinactivation in platelet concentrates: enhanced speci-
ficity of virus kill in the absence of shorter UVA wavelengths. Photochem Photobiol
1995; 62:917922.
55. Gottlieb P, Marolis-Nunno H, Robinson R, Shen LG, Chimezie E, Horowitz B,
Ben-Hur E. Inactivation of Trypanosoma cruzi trypomastigote forms in blood
components with a psoralen and ultraviolet A light. Photochem Photobiol 1996;
63:562565.
56. Lin L, Londe H, Janda JM, Hanson CV, Corash L. Photochemical inactivation of
pathogenic bacteria in human platelet concentrates. Blood 1994; 83:26982706.
57. Benade L, Shumaker J, Xu Y, Shen X, Dodd RY. Inactivation of free and cell-
associated human immunodeficiency virus in platelet suspensions by aminomethyl
psoralen and ultraviolet light. Transfusion 1994; 34:680684.
58. Margolis-Nunno H, Bardossy L, Robinson R, Ben-Hur E, Horowitz B, Blajchman

MA. Psoralen-mediated photodecontamination of platelet concentrates: inactivation
of cell-free and cell associated forms of human immunodeficiency virus and assess-
ment of platelet function in vivo. Transfusion 1997;
59. Goodrich RP, Yerram NR, Tay-Goodrich BH, Forster P, Platz MS, Kasturi C, Park
SC, Aebischer JN, Rai S, Kulaga L. Selective inactivation of viruses in the presence
of human platelets: UV sensitization with psoralen derivatives. Proc Natl Acad Sci
USA 1994; 91:55525556.
60. Piet B, Legrand-Poels S, Sappety C, Piette J. NF-KB transcription factor and human
immunodeficiency virus type 1 (HIV-1) activation by methylene blue photosensitiza-
tion. Eur J Biochem. 1995; 228:447455.
61. Wagner SJ, Storry JR, Mallory DA, Stromber RR, Benade LE, Friedman LI. Red cell
alterations associated with virucidal methylene blue phototreatment. Transfusion
1993; 33:3036.
62. Skripchenko A, Robinette D, Wagner SJ. Comparison of methylene blue and methy-
lene violet for photoinactivation of intracellular and extracellular virus in red cell
suspension. Photochem Photobiol 1997;
63. North J, Coombs R, Levy J. Photodynamic inactivation of free and cell-associated
HIV-1 using the photosensitizer benzoporphyrin derivative. J AIDS 1994; 7:891898.
64. Meruelo D, Lavie G, Lavie D. Therapeutic agents with dramatic antiretroviral activity
at effective doses: aromatic polycyclic diones hypericin and pseudohypericin. Proc
Natl Acad Sci USA 1988; 85:52305234.
65. Carpenter S, Kraus GA. Photosensitization is required for inactivation of equine
infectious anemia virus by hypericin. Photochem Photobiol 1991; 53:169174.
66. Lavie G, Mazur Y, Lavie D, Prince AM, Pascual D, Liebes L, Levin B, Meruelo D.
Hypericin as an inactivator of infectious virus in blood products. Transfusion 1995;
35:392400.
67. Horowitz B, Williams B, Rywkin S, Prince AM, Pascual D, Geacintov N, Valinsky
J. Inactivation of viruses in blood with aluminum phthalocyanine derivatives. Trans-
fusion 1991; 31:102108.
68. Gaspard S, Tempete C, Werner GH. Studies on photoinactivation by various phtha-
locyanines of a free or replicating non-enveloped viruses. J Photochem Photobiol
B:Biol 1995; 31:159162.
69. Rywkin S, Lenny L, Goldstein J, Geacintov NE, Margolis-Nunno H, Horowitz B.
Importance of type I and type II mechanisms in the photodynamic inactivation of
viruses in blood with aluminum phthalocyanine derivatives. Photochem Photobiol
1995; 56:463469.
70. Rywkin S, Ben-Hur E, Reid ME, Oyen R, Ralph H, Horowitz B. Selective protection
against IgG binding to red cells treated with phthalocyanines and red light for virus
inactivation. Transfusion 1995; 35:414420.
71. Ben-Hur E, Rywkin S, Rosenthal I, Geacintov NE, Horowitz B. Virus inactivation in
red cell concentrates by photosensitization with phthalocyanines: protection of red
cells but not of vesicular stomatitis virus with a water-soluble analog of vitamin E.
Transfusion 1995; 35:401406.
72. Ben-Hur E, Geacintov NE, Studamire B, Kenney ME, Horowitz B. The effect of
irradiance on virus sterilization and photodynamic damage in red blood cells sensi-
tized by phthalocyanines. Photochem Photobiol 1995; 61:190195.
73. Ben-Hur E, Zuk MM, Kenney ME, Oleinick NL, Mulvihill J, Horowitz B. Action
spectra (660-700 nm) for virus inactivation and red cell damage photosensitized by
the silicon phthalocyanine Pc 4. Lasers Med Sci 1996; 11:221225.
74. Ben-Hur E, Zuk MM, Chin S, Banerjee D, Kenney ME, Horowitz B. Biodistribu-
tion and virus inactivation efficacy of the silicon phthalocyanine Pc 4 in red blood
cell concentrates as a function of delivery vehicle. Photochem Photobiol 1995;
62:575579.
75. Margolis-Nunno H, Ben-Hur E, Gottlieb P, Robinson R, Oetjen J, Horowitz B.
Inactivation of multiple forms of human immunodeficiency virus in red cell concen-
trates by phthalocyanine photosensitization. Transfusion 1996; 36:743750.
76. Gottlieb P, Shen LG, Chimezie E, Bhang S, Kenney ME, Horowitz B, Ben-Hur E.
Inactivation of Trypanosoma cruzi trypomastigote forms in blood components by
photodynamic treatment with phthalocyanines. Photochem Photobiol 1995; 62:869
874.
77. Lustigman S, Ben-Hur E. Photosensitized inactivation of Plasmodium falciparum in
human erythrocytes by phthalocyanines. Transfusion 1995; 36:562565.
78. Ben-Hur E, Oetjen J, Zuk MM, Chan WS, Horowitz B. Photochemical decontamina-
tion of red cell concentrates with silicon phthalocyanine (Pc 4) and red light.
79. Zhang Q-X, Edson C, Budowsky E, Purmal A. Inactine: a method for viral inactiva-
tion in red blood cell concentrate. Transfusion 1998; 38(suppl):75S.
22
The Role of Quality in Blood Safety
Lucia M. Berte
Quality Systems Consultant, Elmhurst, Illinois
David E. Nevalainen
Quality Consultant, Health Care, Baileys Harbor, Wisconsin
Assuring blood transfusion safety means having error-free processes that begin
with donor selection and continue through blood component administration.
High-quality blood products and safe patient outcomes are best assured when
errors are prevented along the entire range of blood banking activity. Two goals
are desirable: (a) reduce to zero the number of blood components made from
unsuitable donations that result in any finished, labeled blood product that could
be distributed for clinical use, and (b) get the right blood products in the right
quantity to the right patient in the right place at the right time (1). These two broad
collections of processes and subprocesses involve different groups of people in
different locations at different times. However, a logical approach to apply
error-prevention initiatives across the entire range of activities can have the
desired end results.
I. BLOOD COMPONENT MANUFACTURING

The objective of the blood manufacturing processas it is regarded by the U.S.
Food and Drug Administration (FDA)is to put safe, finished, labeled blood
products into stock at the point of distribution for future clinical use (1). The
major steps of this process include obtaining the whole blood (raw material) from
qualified blood donors, testing the blood for infectious diseases, preparing blood
components, and distributing the components to entities who prepare them for
transfusion. To underscore the importance of this objective, FDA has chosen to
497
498 Berte and Nevalainen
aggressively treat blood bank facilities as manufacturers of pharmaceutical prod-

ucts and thus applies strict pharmaceutical manufacturing regulations to blood
collection and testing activitiesan environment distinctly different from the
hospital clinical laboratory in which many technical blood banking personnel had
their training. The transition from a service-oriented culture, as blood banks have
always perceived themselves, to a manufacturing culture, as FDA perceives them,
occurred slowly but steadily during the 1990s.
To achieve their respective manufacturing objectives, highly regulated
pharmaceutical and medical device manufacturers must operate within the bound-
aries of current good manufacturing practice (cGMP) described in the Code of
Federal Regulations (CFR). Meeting CFR requirements fosters an environmental
culture known as total process control (TPC). By practicing TPC, blood banks
can safely operate within the applicable CFR regulations found in 21 CFR Parts
210, 211, 606, 640, and 820 (24), as shown in Table 1.
II. CLINICAL TRANSFUSION PROCESS

The major steps of the clinical transfusion process include obtaining and labeling
a patient sample, submitting it to the transfusion service with an indication of
requirements including the degree of emergency, testing for compatibility, deliv-
ering the correct blood component to the correct location within the required time,
correctly transfusing the component to the patient for whom it was originally
intended, and monitoring the outcome (1). Transfusion services are also under the
ultimate purview of FDA, but because of a memorandum of understanding
between FDA and the Health Care Financing Administration (HCFA) to reduce
duplicative inspections, most transfusion services are routinely inspected by
laboratory accreditation agencies authorized to inspect by HCFA. This has
resulted in a misconception that CFR requirements do not apply; they do apply
and can be found in CFR Part 606 (3). The principles of TPC are equally
appropriate to assuring a safe transfusion process, particularly because there are
so many interdepartmental interactions involved.
III. PRINCIPLES OF TPC

Total process control satisfies CFR requirements because it reduces the variability
in process performance and outcome that leads to errors. In blood banking terms,
this translates to reducing errors that allow unsuitable blood donors being drawn,
positive disease marker test results being overlooked, untested or test-positive
components in the blood supply, mislabeled specimens for compatibility testing,
transfusion of the wrong patient, or lack of patient monitoring. TPC methods take
variation out of the process, thus ensuring a more predictable outcome. From this
Table 1 Contents of Applicable Parts of the Code of Federal Regulations (CFR) for Blood Banks
Part 211 Part 606 Part 640 Part 820

Part 210 Subparts: Subparts: Subparts: Subparts:
210.1 Status of cGMP A: General provisions A: General provisions A: Whole blood A: General provisions
210.2 Applicability of B: Organization and B: Organization and B: Red blood cells B: Quality system requirements
cGMP personnel personnel
210.3 Definitions C: Buildings and C: Plant and facilities C: Platelets C: Design controls
facilities
D: Equipment D: Equipment D: Plasma D: Document and record
controls
E: Control of E: (Reserved) E: (Reserved) E: Purchasing controls
components . . .
F: Production and F: Production and F: Cryoprecipitate F: Identification and

process controls process controls traceability
G: Packaging and G: Finished product G: Source plasma G: Productions and process
labeling control control controls
H: Holding and H: Laboratory controls H: Albumin H: Acceptance activities
distribution
I: Laboratory controls I: Records and reports I: Plasma protein fraction I: Nonconforming product
J: Records and J: Immune globulin J: Corrective and preventive
reports action
K: Returned and sal- K: Alternative procedures K: Handling, storage, distribu-
vaged drugs . . . tion and installation
L: Packaging and label control
M. Records
N: Servicing
O: Statistical techniques
499
Source: Refs 24.

perspective, blood banks cannot afford not to practice TPC if we are true to our
objective of providing safe, efficacious blood transfusions.
The essential principles of TPC are summarized in these statements from
an FDA guidance document (5):
Quality, safety, and effectiveness are built into a product.
Quality cannot be inspected or tested into a product.
Each step in the process must be controlled to meet quality standards.
These three statements provide facilities with a roadmap for what to do to assure
quality blood products and patient outcomes but do not describe how to do it.
Facilities determine for themselves what resources and methods they need to
apply to accomplish the intent of GMP. The important features of TPC are shown
in Figure 1.
Figure 1 The elements of Total Process Control (TPC) form a circular flow of process
analysis, SOP development, training, and assessment for which records document all
necessary activities. (From Ref. 14.)
The Role of Quality in Blood Safety 501
To best explain how the elements of TPC work, examples will be used.
An example for blood processing is a new screening test for a transfusion-
transmitted disease to be performed on each unit of donated blood. The example
for clinical transfusion is the implementation of a new armband system to
specifically identify transfusion recipients and link them with their compatibility
testing specimens. The following sections briefly describe the elements of TPC
in relation to these two examples and provide references for readers to access
important how-to details.
A. Validation
Before a new test, a new computer system or a new process is implemented, or
if there has been a significant change in an existing test method, instrument,
software, or process, an activity known as validation must take place. Process
validation is a requirement of cGMP Parts 211 (2) and 820 (4) for manufactured
pharmaceuticals and medical devices. It establishes documented evidence that a
specific process will consistently produce a product meeting its predetermined
specifications and quality characteristics. The facility must prepare a written
validation protocol that specifies what procedures and tests are to be conducted
and the data to be collected during practice runs of the new process. Only when
the data demonstrate that the process functions reliably and invariably can
standard operating procedures (SOPs) be finalized, remaining personnel trained,
and actual operations begun.
Table 2 outlines the activities of the major phases of a validation protocol
that are briefly described in the following paragraphs. Some details may vary
between processes to be validated, but the major concepts remain unchanged.
Validation examples specific to blood banking have been published (6).
1. Installation Qualification
In the example of a new blood donation screening test, all equipment used must
undergo installation qualificationactivities that demonstrate and document that
the equipment is suitable for its intended purpose, has been installed properly, and
is functioning as intended. For an enzyme immunoassay screening test, equip-
ment to be qualified would include specimen pipettors and dilutors, incubators,
tray washers, and spectrophotometric readers. Draft procedures for equipment
calibration, maintenance, quality control, and troubleshooting are first derived
from the directions included in the manufacturers equipment manuals, to which
are then added facility-specific details. Respective forms are designed to capture
measurements, data, and observations for the ongoing quality control program.
The new armband in our clinical transfusion example does not require installation
502
Table 2 Validation Activities
Installation qualification (IQ) Operational qualification (OQ) Performance qualification (PQ) Revalidation
Identification of items requiring Evaluation of process capabilities Development of testing plan Development of protocol
calibration
Determination of calibration Coordination of multiple pro- Predetermination of process/ Reaffirmation of IQ
method and schedule for cesses and operations product specifications
each item
Identification of items requiring Consideration of process variables Performance of process by Reaffirmation of OQ
maintenance operations personnel
Determination of maintenance Quantification of process variables Comparison of process out- Review of performance
methods and schedules for comes to specifications for history, OR
each item acceptability
Development of operating Determination of acceptable oper- Training and documentation Performance of a process run
procedures, including ating limits and variations
adjustments
Development of troubleshoot- Development of process Implementation of process Comparison of outcomes
ing procedures procedures to specifications for
acceptability
Development of monitoring and Retraining, if needed
control procedures
Listing of equipment parts Recommendation for
improvements
Training and documentation
Source: Ref. 7.
Berte and Nevalainen
qualification unless barcode readers and/or electronic label printers are part of the
new armband-generation process.
2. Operational Qualification
This second phase of a validation protocol consists of evaluating the capability
of the processin this case, that the new blood screening test performs ac-
ceptably according to the manufacturers procedure using the newly qualified
equipment. The effectiveness and reproducibility of the new test is vigorously
challenged with conditions simulating those expected to be encountered during
actual production. In our testing example, challenges would include both strongly
and weakly reactive test samples, borderline reactive samples, and previous
proficiency testing material. In our armbanding example, the process would be
tested to see that barcodes and labels are readable and that the identification
linkages are traceable throughout specimen collection, compatibility testing, and
transfusion. The challenges should be repeated often enough to assure that the
results are consistent and meaningful. The draft testing procedures and related
forms are then readied for the validation runs.
3. Process Qualification
This phase of the validation protocol requires that selected operations personnel
(a) perform the new process as it would be done during actual operations,
(b) capture data about how the process worked, and (c) compare data to prede-
termined acceptance criteria. For the donor blood test, this final phase should not
proceed until operations and quality assurance personnel are confident that the
validation will be successful because of the high cost of performing full test runs.
In the testing example, a preset number of test runs would be performed by staff
members according to procedures derived from the manufacturers package insert
and facility-specific activities that were qualified in the preceding operational
qualification phase. In the armband example, all patients in a specified loca-
tion only would receive the new armband for a predetermined time. Any prob-
lems in the new processes are documented and followed up. Results of the
validations are reviewed by quality assurance personnel who reach a formal
conclusion as to the acceptability of the new process. Acceptance includes all
appropriate reviews and signatures and the finalization of, training in, and
activation of all related calibration, maintenance, quality control, troubleshooting,
and operating procedures.
4. Revalidation
Revalidation of manufacturing processes should be performed periodically and
whenever there are changes in raw materials, equipment, or processes that could
affect the quality of the process outcome. Revalidation should also be considered
when there are significant changes in the processs quality performance history
as determined by statistical process control measurements. Revalidation consists
of reaffirming installation qualification and operational qualifications and either
performance of a validation run or comprehensive review of the processs quality
history records. In our testing example, revalidation would be required if the
manufacturer should change any test equipment, the test methodology, significant
reagents, or other variables of the blood screening test that could potentially affect
the quality of the test results. In our armband example, revalidation would be
required if there were changes to barcoding software or symbology or to label-
printing equipment.
B. Calibration and Maintenance

Initial calibration and maintenance procedures and schedules were established as
part of installation qualification in the validation process. Recalibration is re-
quired periodically and after repairs or adjustments (3). Routine and preventive
maintenance procedures prolong the operating life of equipment, which mini-
mizes equipment problems and changes that lead to a requirement for revalida-
tions. In our testing example, all calibrations and maintenance are performed
according to the established procedures and schedules and the results documented
and periodically reviewed to assure continued acceptable function. TPC requires
an identification mechanism, tracking process, and cumulative file of all records
generated for each piece of equipment used in the new donor test process.
C. Standard Operating Procedures

Standard operating procedures (SOPs) are the key feature of TPC. Properly
written, controlled SOPs used by trained personnel are a blood bank facilitys best
assurance of quality and safety in the blood supply. In the donor screening test
example, SOPs are drafted from instructions provided in the manufacturers
equipment manuals and package inserts for the specific test. In the armband
example, existing SOPs for specimen collection, compatibility testing, and blood
administration are modified to include the new identification steps. After the
successful validation, draft SOPs are finalized. When validated SOPs are used to
train employees, processes become less variable, errors are reduced, and the
desired quality is more consistently achieved. Document control is an additional
process that promotes consistent coordinated handling of a facilitys many
procedures (7). In our examples, procedures would be developed using a pre-
established format for completeness and consistency that featured an SOP iden-
tification mechanism essential to the timely review, revision, removal, and
archiving of all the processs SOPs and related forms. SOP management includes
the use of an SOP for how to write SOPs and another for how to change them.
Many external inspection deficiency citations involve lack of, outdated, or
incomplete SOPs or personnel found not to be following existing SOPs. These
problems lead to personal variations that can and have been shown to cause errors
that reduce blood safety and compromise the benefit of transfusion.
D. Training and Competence Assessment

When TPC is not the basic operating philosophy, personal deviations can signif-
icantly contribute to the numbers and types of errors that occur. Establishing
current, validated SOPs is one aspect of an approach to minimize variation.
The complementing aspect is to train personnel in the contents, exclusive use of,
and unwavering adherence to approved SOPs.
To minimize the effect of personnel being the most variable entity, ap-
proved SOPs are used as the basis for developing task-specific training guides.
Qualified trainers use the guides to train personnel and assess demonstrated
competence in the specific SOPs for the new process. All training and competence
outcomes are documented. In the donor testing example, training activities and
schedules can be more easily controlled because personnel work in the testing
department. In the armband example, trainees would include personnel who
collect compatibility testing specimens, perform compatibility testing, or ad-
minister blood transfusions and would include laboratory phlebotomists, non-
laboratory specimen collection personnel, medical technologists and technicians,
nurses, and physicians. Only a comprehensive, coordinated training process with
well-written SOPs will assure that all involved personnel understand and fol-
low the procedures for blood transfusion identification traceability. At this level
of complexity, it is easy to see how variations can be made that lead to
mistransfusions.
E. Monitoring
After the new procedures are validated, personnel trained, and a start-up date
selected, TPC requires ongoing monitoring of process variables to assure that the
process remains in control. A number of monitoring activities should be ongoing.
1. Quality Control
Laboratories have a long history of using quality control (QC) methods to assure
the accuracy of test results. For every test method or piece of equipment, there
are general and specific regulations and accreditation requirements for labora-
tories for QC monitoring and requirements for follow-up action when the results
are out of acceptable range. The types of QC commonly performed in blood banks
are shown in Table 3. QC schedules are to be established and all results
documented and reviewed for trends and patterns that could suggest out-of-con-
trol performance. In the donor testing example, QC monitoring would include
positive and negative controls with every test run, blank, and background controls
for spectrophotometric measuring accuracy, monitoring of all temperature-regu-
lated equipment, and periodic recalibrations of pipettors and dilutors, among
other activities. The armband system needs no QC.
Table 3 Blood Bank Quality Control
Equipment
Item Testing
Blood drawing scales Check with standard weight

Hemoglobinometers Calibration to standards
Shipping containers Ability to hold temperature
Centrifuges Functional calibration, RPM, timer
Serologic
Cell-washer
Refrigerated
Microhematocrit
Cell-washer centrifuge Tube fill and decant
Refrigerated centrifuge Internal temperature
Temperature of component after
centrifugation
Refrigerators: Internal temperature
Blood storage
Reagent storage
Refrigerator: Continuous temperature recording;
Blood storage comparison of graph to thermometer;
Freezer: (mechanical/liquid nitrogen) alarm activation; nitrogen level
Component storage (freezer)
Platelet incubator
Plateletsopen storage area Temperature every 4 hours
pH meter Calibration to standard
Thermometers Compare against NIST standard
Water baths Single point temperature check
Dry baths Several point temperature check
Incubators
View boxes
Table 3 Continued
Equipment
Item Testing
Component thawing devices Temperature, Cleanliness

Blood Irradiators Decay of source
Radiation leaks
Dosage delivery
Blood warmers Plate/bath temperature
Effluent temperature
Alarm activation
Components
Item Testing
Red blood cells Hematocrit

Cryoprecipitate Factor VIII level
Granulocytes Granulocyte count
Platelets, random Platelet count, pH, volume
Platelets, apheresis
Fresh frozen plasma Volume
Leukocyte-reduced red blood cells Removal of leukocytes
Red blood cell recovery
Intraoperatively recovered red blood cells Contamination
Free hemoglobin
Frozen-deglycerolized red blood cells Osmolality
Free hemoglobin
Red cell recovery
Red cell viability
Reagents
Item Testing
Reagent antisera (anti-A, anti-B, anti-D, etc) Positive control

Reagent red blood cells (A1 & B cells, Positive control
Screening Cells, Coombs Control Cells)
Antiglobulin serum Reactivity
Tests for syphilis, HIV-1-Ag, anti-HIV-1, Control testing of each lot, each test run
anti-HIV-2, anti-HCV, anti-HTLV-I, anti-
HTLV-II, anti-HBc, HBsAg
Source: Ref. 3 and Ziebell L, Kavemeier K, eds. Quality Control: A Component of Process
Control in Blood Banking and Transfusion Medicine. Bethesda, MD: American Association of Blood
Banks, 1999.
2. Statistical Process Control

Using statistical process control (SPC) tools such as control charts, histograms,
bar charts, and other tools, QC and other collected data are analyzed and
converted into valuable information about the status of the facilitys processes.
The information provided by these tools helps identify if there are process
problems and where they may be occurring. When the causes of system problems
are removed, both quality and productivity are improved. Use of the graphical
tools shown in Table 4 would inform where the facility is currently, where the
variations are, the relative importance of the identified problems, and whether the
changes made had the desired impact (8). These tools facilitate the use of a
structured problem-solving process to achieve better solutions. In our testing
example, a Pareto chart could show the number, time of day, and operator for
failed test runs. A bar graph would depict the number of repeat tests by shift or
by operator or by analyzer used. In the armband example, the number and source
of specimens without armband-traceable identification could be visualized on a
bar chart. Direct observation of specimen collection and blood administration
personnel would provide information on whether SOPs were understood and
followed. The information obtained from using SPC tools helps the facility
visualize the data and prioritize the types and sources of problems so that
improvement efforts can be better organized.
3. Occurrence Reporting and Follow-Up

One of the best methods to identify problems is to have the staff who perform
the daily work report occurrences whenever an expected outcome is not real-
ized. Whatever names are given to occurrencesincidents, errors, deviations,
accidents, nonconformances, events, complaints etc.the most important issue
is to encourage the reporting and make it nonpunitive to employees. The objective
is to identify wherever something is not working as it should and take action
to eliminate the cause of the problem. Process-improvement methods are then
applied, and follow-up is performed by monitoring selected indicators and
continuing analysis of occurrence reports. Occurrence analysis should lead to
identification of problems as either system (process), knowledge (training),
or behavior (discipline), each of which requires a different form of corrective
action. In the donor-testing example, operators would complete an occurrence
report form for each failed test run. In the armband example, each instance
when a patient to be transfused was not wearing the proper blood bank
armband would result in an occurrence report. The reports are analyzed to
determine whether (a) the root cause was in the process, (b) involved per-
sonnel were insufficiently trained, or (c) personnel merely did not follow
directions.
Table 4 Statistical Process Control Tools and Uses
Tool Use Manifestation
Flowchart To identify the actual path that a product or service Symbols to depict process steps and decisions
follows to identify problems that lead to deviations
and errors
Check sheet To gather data based on sample observations to begin to Plot of defects vs. dates or times
detect patterns
Pareto chart To display relative importance of all problems or condi- Vertical bar graph with bars in descending
tions to a) identify basic cause of a problem, b) priori- order
tize problem solving, c) monitor success
Cause and effect diagram To identify, explore, and display the possible causes of a Fishbone
problem or condition
Run chart To display trends with observation points over a speci- Plot of measurement vs. time
fied time period
Histogram To discover and display the distribution of data by bar Bar graph in frequency of distribution curve
graphing the number of amounts in each category
Scatter diagram To display what happens with one variable when another Plot of one variable on the x axis and the
variable changes to test a theory that the two variables other variable on the y axis
are related
Control chart To discover how much variability in a process is due to Run chart with statistically determined upper
random variation and how much is due to unique and lower control limit lines on either side
events to determine whether a process is in statistical of the process average
control
Process capability chart To determine whether the process, given its natural varia- Distribution curve showing allowable spread
tion, is capable of meeting established specifications of specification limits and measure of
actual process variation
Source: Ref. 8.
509
4. Internal Quality Audit

Another assessment tool is the quality system audit. The facilitys quality system
consists of its policies and procedures for the critical quality processes in TPC
such as those already described above and others. The audit compares the
facilitys stated quality policies, processes, and procedures to a predetermined
reference standard such as cGMP, the American Association of Blood Banks
(AABB) Standards for Blood Banks and Transfusion Services (9), the Clinical
Laboratory Improvement Amendments of 1988 (CLIA 88) (10), and other
regulations and accreditation requirements for laboratories. The auditor deter-
mines the conformity of the facilitys actual operations with the documented
system. To avoid any conflicts of interest, quality system audits should be
conducted by personnel who do not have responsibility for the areas being
audited. The audit results are reported to the facilitys top management for
corrective action. In the examples, the facilitys quality auditors compare the
policies, processes, and procedures of the involved departments with both the
reference standards for blood donation testing and specimen/patient identification
and their actual observations of documents and personnel performing testing,
specimen collection, and blood administration. Any discrepanciesknown in
auditing language as nonconformances are reported to the facilitys manage-
ment for corrective action. Follow-up audits may be conducted to determine if
the necessary corrective action took place and was successful.
F. Corrective Action
The monitoring activities of QC, SPC, occurrence reporting, and quality audit all
provide an overview of where the facilitys problems lie. The root cause of each
problem must be identified using a tool such as a cause-and-effect diagram (8)
and the problems categorized as system, knowledge, or behavior prior to taking
any corrective action. The chosen long-term corrective action must be appropriate
for the type of problem or the fix will not work. For example, training solutions
are often applied to system problems because it is easier to schedule a retraining
session than to dismantle a multistep process and rebuild it with error-prevention
steps to eliminate the problems root cause. In this scenario, follow-up monitoring
would, unfortunately, demonstrate that personnel have been further trained in a
process that still does not work.
G. Process Improvement
The time and resources expended in correcting quality problems are best utilized
when following a systematic approach for solving problems and improving
processes. The facility should adopt one of the common specialized problem
solving models, such as those shown in Table 5, train personnel in its use, and
use the model for each performance improvement effort it undertakes. Most of
the improvement methods use multidisciplinary teams to work through the
successive steps.
Process improvements should not only be generated for problem areas.
When facilities truly practice continuous quality improvement, they will period-
ically reassess current processes to determine whether there may be a way to
perform them faster, better, or more cost-effectively. At any given time, there may
be several process-improvement efforts simultaneously underway as the facility
works through its prioritized list of problems and scheduled routine reviews.
IV. CONNECTION OF TPC TO A QUALITY SYSTEM

As more blood donor screening questions and tests were added in the years after
anti-HIV testing was first implemented in 1985, the complexity of managing
testing information, donor deferrals, and computer software increased dramati-
cally. Because rising numbers of untested and unsuitable blood components were
entering the blood supply, FDA began actively to apply the quality and process
control provisions of CFR Parts 210-211 (2) for pharmaceutical manufacturing to
blood banking facilities. The various TPC activities must be coordinated to truly
Table 5 Process Improvement Models
Source Model name
Shewharta Plan - Do - Check - Act (PDCA)

JCAHOb Plan - Design - Measure - Assess - Improve
Juranc Diagnostic and Remedial Journeys
HCAd FOCUS - PDCA
Scholtese Team Process
ODIf Focus - Analyze - Develop - Execute (FADE)
aFrom Shewhart WA. Economic Quality Control of Manufactured
Product. New York: Van Nostrand, 1931.
bFrom Comprehensive Accreditation Manual for Hospitals. Oakbrook
Terrace, IL: Joint Commission on Accreditation of Healthcare Organi-

zations, 1999.
cFrom Juran JM. Juran on Leadership for Quality. New York: The Free
Press, 1989.
dFrom Hospital Corporation of America, Nashville, TN.
eFrom Scholtes PR. The Team Handbook: How to Use Teams to
Improve Quality. 2d ed. Madison, WI: Joiner Associates, Inc., 1996.

fFrom Quality Action Teams. Burlington, MA: Organizational Dy-
namics, Inc., 1991.

build quality into all processes and monitor the outcomes; current CFR cGMP
and laboratory regulations alone do not facilitate this coordination. To assist blood
banks in this effort, FDA introduced the Guideline for Quality Assurance in Blood
Establishments (11) as a draft in 1993; the document was finalized in 1995 after
considerable public input.
Concurrently with revision of the FDA draft guideline, the AABB produced
The Quality Program (12), a detailed expansion of the FDA guideline combining
the elements of TPC with the operational functions of blood banking from donor
selection through blood administration. Both FDA and AABB require blood
banks and transfusion services to have a quality program in place, though not
necessarily theirs. The AABB program, however, has the comprehension and
simplicity necessary for facilities of any size.
The framework of the AABB quality program is shown in Figure 2. The
elements of TPC are organized onto the left side of the quality program frame
and are collectively called the quality system. AABB refers to the TPC
elements as quality system essentials (QSEs). The important feature of the QSEs
is that they are to be uniformly applied to the operational functions shown across
the top of the frame. For example, training is to be given to all personnel in each
operational function and should minimally include organizational, safety, quality,
computer, and job-specific training. Likewise, the process for changing an SOP
should be the same in all the facilitys operational functions.
Whatever the size or scope of a blood bank facility (i.e., however many or
few activities or personnel in the operational functions at the top of the frame),
the QSEs remain unchanged because they are so fundamental to the facilitys
ability to build quality and safety into blood banking products and services.
To emphasize the importance of this concept, the AABB requires its institutional
members to have written quality policies that state what the facility intends to do
to implement the QSEs (TPC elements) in its operational functions. The quality
policies are supported by quality processes and procedures that describe how the
facility implements its stated intentions.
V. QUALITY SYSTEM DOCUMENTATION

A model exists for how a facility can organize the documentation for its quality
system and operational functions. The model comes from the ISO 9000 quality
standards compendium, which is a collection of definitions, guidelines, standards,
and requirements for quality management (13). The section of the compendium
that describes how to develop a quality manual (ISO 10013) provides a document
hierarchy for a quality system in three levels: quality policy, quality procedures,
and work instructions.
Figure 2 The AABB Quality Program grid showing cross-functional total process control elements (quality system essentials) applied to
513
all blood bank operational functions. (From Ref. 15.)

A. Level A Documentation
As mentioned previously, the AABB requires that institutional members have a
written quality plan that states the facilitys quality objectives and policy. The
quality plan describes in a narrative or outline format what the facilitys policy is
with regard to each QSE (TPC element). The plan need not be lengthy, as other
documents will provide more detail for how the facility implements its quality
objectives. Sample quality plans are available to help facilities draft their own
(1315).
B. Level B Documents
Quality system processes describe the what, when, where, who of the TPC
elements/QSEs on the left side of the quality program frame. These documents
describe the processes for generating an SOP, training personnel, reporting an
occurrence, developing a validation protocol, and other tasks as shown in Table 6.
Quality system processes are cross-functionalthat is, they are applied across all
facility operating functions. For example, a person reporting an occurrence in the
component-processing area would follow the same procedure and use the same
reporting form as a person reporting an occurrence in the donor-testing area.
Important to note is that quality system processes often cross departmental lines
as to actions taken, responsibilities, and documentation to be generated and thus
have broad applicability. They are not written with the outline and detail of the
technical operating procedures used for the task specifics of employees jobs.
Flowcharts and tables are often used to describe processes.
C. Level C Documents
Level C documents are the technical operating procedures and related forms used
by employees in the facilitys operational functions. They are the work instruc-
tions that provide the detail necessary for personnel to perform their specific job
tasks. Using the donor testing example, a Level B quality process for SOP
development would be used to draft the Level C operating procedures for
calibration and maintenance of the spectrophotometric equipment, performance
of the immunoassay tests, reporting test results in the computer, and documenting
daily quality control. In the patient-identification example, the format described
in the same Level B SOP for SOPs would be used to write all the operating
procedures for patient specimen collection, compatibility testing, and blood
administration. Each Level C procedure may have one or more associated forms
that must be linked to it in some prescribed fashion as described in the Level B
process for the facilitys document control. All Level C procedures in one facility
Table 6 Sample Contents of a Quality Manual
Level A: Quality Policies Level B: Quality Processes
Quality Plan Overview

Quality Program Organization Policy Review of the Quality Program
Personnel Policy Training Process
Staff Competence Assessment
Continuing Education
Validation Policy Validation Protocol Preparation
Calibration and Preventive Maintenance Laboratory Equipment Calibration
Policies
Laboratory Equipment Maintenance
Proficiency Testing Policy Proficiency Testing Process
Vendor Qualification Policy Vendor Qualification Process
Contract Review Process
Receipt/Inspection of Incoming Re-
agents and Supplies
Process Control Policy Change Control Process
Quality Control Program
Documents, Records, Reviews Writing Standard Operating Procedures
Writing Training Documents
Controlling Documents
Incidents, Errors, Accidents Policy Occurrence Reporting and Follow-Up
Internal Assessment Policy Internal Assessment Process
Process Improvement Policy Corrective Action Process
Process Improvement Using CQI Tools
Source: Ref. 15.
would be written in prescribed formats. A Level B document describing the

process for change control would be used to request any Level C procedure
changes in any operational function.
Figure 3 depicts the levels of the quality documentation pyramid showing
the written quality plan at Level A, the TPC/QSE cross-functional elements at
Level B, and the operational functions at Level C. In any operational function,
one is guided by the facilitys Level A quality policy, applies the appropriate
Level B cross-functional quality processes, such as validation, training, and
document control, and performs the resulting Level C operating procedures
without personal deviations. The quality system approach provides the means to
build quality, safety, and effectiveness into blood products and services, monitor
the outcomes, and make improvements where needed.
Figure 3 A modification of the quality system documentation pyramid showing the

quality policies, blood bank cross-functional quality processes, and operational functions.
(Courtesy of Abbott Quality Institute, Abbott Park, IL.)
VI. QUALITY SYSTEM ASSESSMENTS

As previously stated, quality cannot be inspected or tested into a product or
service. Experience in manufacturing has shown this philosophy to be sound.
Therefore, the historical approach of external inspections to detect a facilitys
deficiencies, such as those conducted by laboratory regulatory and accreditation
agencies, must change into a process that more effectively evaluates the facilitys
ability to detect its own weaknesses and take action to prevent errors.
In well-designed quality systems, total process control is the corporate
culture. Continuous process improvement is fervently practiced to reduce errors
and satisfy customers, which, by definition, increases productivity and decreases
cost. Self-assessment is the means for blood banks and transfusion services to
determine their current state of performance so that they can prioritize and target
their improvement efforts.
A. Self-Assessment
What should be assessed? The AABB has divided the scope of blood banking
activity into the operational functions shown across the top of the frame in Figure
2. For each operational function, processes and subprocesses have been identified.
For each process, one or more quality indicators have been identified that can be
used by facilities to monitor their performance over time (7). Operations person-
nel collect data and compare actual performance to preset expectations. Areas
needing improvement are prioritized, and the chosen continuous process improve-
ment model is applied. Follow-up monitoring of the quality indicators informs
the facility of its progress. Periodic reporting provides information for executive
management to make important decisions about resource allocations.
Self-assessment is an ongoing activity for which data should be collected
as part of the work process wherever possible. As an example from processing,
the log of unsuitable components should be updated as the components are
identified and quarantined. Occurrence reports should be generated as soon as
problems are discovered. Computers can be used to update blood donation,
deferral, and incomplete unit statistics as donor records are entered. As a
transfusion service example, persons performing compatibility testing can add to
the log of unacceptable specimens on specimen receipt in the laboratory. Com-
puters can be used to capture specimen receipt-to-result turnaround times for
emergent compatibility testing. Incomplete or incorrect transfusion report forms
can be logged on return to the transfusion service.
B. Quality Audits
Internal audits assess the conformance of the facilitys written quality system to
the standard and compare the actual activities to the written system. First-party
audits are performed by the facilitys own auditors as part of TPC monitoring.
Second-party audits are performed by facilities on their suppliers and take place
external to the facility. For example, a hospital blood bank may choose to perform
a second-party audit on the supplier of the blood components it procures for
transfusion. Third-party external quality audits are conducted by contracted
auditors to assess the facilitys quality system with respect to national or interna-
tional quality system standards and to provide registration or certification of
acceptable quality systems. The most common type of third-party audit is an ISO
9000 registration audit.
Table 7 is a comparison of the AABB QSEs to the ISO 9001 quality system
conformance model that demonstrates their compatibility. Some U.S. blood banks
have totally embraced the TPC and quality system philosophies and are preparing
for ISO 9000 registration audits. Two U.S. blood bank quality systems (16,17)
and several in other countries (18) have already become registered. Both the
Table 7 Comparison of the AABB Quality System Essentials to the ISO 9001 Quality
System Conformance Model
AABB Quality System Essentials ISO 9001 Quality System Conformance Model
Organization 4.1 Management responsibility

4.2 Quality system
Personnel 4.18 Training
Equipment 4.11 Control of inspection, measuring and test
equipment
Supplier issues 4.3 Contract review
4.6 Purchasing
Process control 4.4 Design control (N/A for 9002)
4.7 Control of customer supplied product
4.8 Product identification and traceability
4.9 Process control
4.10 Inspection and testing
4.12 Inspection and test status
4.15 Handling, storage, packaging, preservation
and delivery
4.19 Servicing (if applicable)
Documents and records 4.5 Document and data control
4.16 Control of quality records
Incidents, errors, accidents 4.13 Control of nonconforming product
Internal and external assessment 4.17 Internal quality audits
Process improvement 4.14 Corrective and preventive action
4.20 Statistical techniques
Facilities and safety [4.9(b) Process control]
Source: Courtesy of Abbott Quality Institute, Abbott Park, IL.
AABB and FDA have redesigned their inspection programs to better evaluate
whether the facilitys quality system conforms to regulations and accreditation
requirements and whether the facility has demonstrated compliance with and
effective implementation of its own quality policies, processes, and procedures.
VII. SUMMARY
The desirable and necessary patient outcome of safe, efficacious blood transfu-
sion requires a paradigm shift to the fervent belief that quality, safety, and
effectiveness must be built into all of a facilitys blood banking activities. In
manufacturing facilities, personnel are usually working in a single management
structure, which affords much more control in cooperation and coordination for
process development, validation, monitoring, and improvement. In contrast, the

clinical transfusion process involves the interaction of individuals who work in
numerous management settings for which the provision of safe blood transfusion
is just one of many priorities that may not be recognized as an important
responsibility. The challenges to building TPC in each environment are many, but
the rewards outweigh the efforts.
Total process control fosters processes and procedures that prevent errors,
thereby building safety into the blood supply and blood transfusion. To manage
TPC throughout the various activities of blood donation, component processing,
testing, and transfusion, a quality system is needed. Quality models for transfu-
sion medicine combine the elements of the ISO 9000 quality standards with the
elements of blood bank and laboratory technical standards to accomplish a truly
comprehensive quality system.
REFERENCES
1. McClelland DBL, McMenamin JJ, Moores HM, Barbara JAJ. Reducing risks in
blood transfusion: process and outcome. Transfusion Med 1996; 6:110.
2. Food and Drug Administration, Department of Health and Human Services. 21 CFR
Parts 200-299. Washington, DC: U.S. Government Printing Office, revised annually.
Parts 600-799. Washington, DC: U.S. Government Printing Office, revised annually.
Part 820. Washington, DC: U.S. Government Printing Office, revised annually.
5. Food and Drug Administration, Department of Health and Human Services. Guide-
line on general principles of process validation. Washington, DC: U.S. Government
Printing Office, 1987.
6. Holliman S, ed. Validation in blood establishments and transfusion services.
7. Nevalainen DE, Berte LM, Callery MF. Quality systems in the blood bank environ-
ment, 2d ed. Bethesda, MD: American Association of Blood Banks, 1998.
8. Brassard M, Ritter D. The memory jogger II. Methuen, MA: GOAL/QPC, 1994.
9. Standards for Blood Banks and Transfusion Services, 19th ed. Bethesda, MD:
American Association of Blood Banks, 1999.
10. Department of Health and Human Services. 42 CFR Parts 430 to end. Washington,
DC: U.S. Government Printing Office, revised annually.
11. Guideline on quality assurance in blood establishments; FDA Docket #91N-0450.
Bethesda, MD: U.S. Food and Drug Administration, 1995.
12. The quality program. Bethesda, MD: American Association of Blood Banks, 1994.
13. ISO Standards Compendium: ISO 9000 Quality Management, 6th ed. Geneva:
International Organization for Standardization, 1996.
14. Nevalainen DE, Callery MF. Module VIII: Quality system documentation. In: Quality
systems in the blood bank and laboratory environment. Bethesda, MD: American
Association of Blood Banks, 1994.
15. AABB Transfusion Service Quality Assurance Committee. A model quality system
for the transfusion service. Bethesda, MD: American Association of Blood Banks,
1997.
16. South Texas Blood and Tissue receives ISO accreditation. AABB Weekly Report
1996; 2(44):4.
17. Denver blood center granted ISO 9000 certification. AABB Weekly Report 1998;
4(26):1.
18. Nevalainen DE, Lloyd HL. ISO 9000 quality standards: a model for blood banking?
Transfusion 1995; 35(6):521524.
23
Cost-Effectiveness Analysis of
Risk-Reduction Strategies
James P. AuBuchon
Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire
I. INTRODUCTION
From the time of Hippocrates, the intention of doing no harm to ones patient has
been codified as a physicians duty. Viral transmission through transfusion clearly
represents an untoward outcome that both physician and patient would like to
avoid, and both parties would certainly prefer to minimize the risks of a
transfusion. Laudable and effective efforts have been directed at reducing trans-
fusion risks through donor screening, infectious disease marker testing, and
viral-inactivation efforts, but the risks of transfusion can only be lowered or
replaced, not eliminated. In addition to an ongoing balancing of risks and benefits,
the medical community is now also asked by society to achieve more with fewer
resources. The desire to reduce transfusion risk must be accomplished in an era
of finite resources. This situation is an appropriate one for application of decision
analysis toolsnot only to determine the relative return on investments in new
approaches to this health care problem, but also to illustrate where blood safety
efforts are most likely to yield substantial payoffs.
II. A PRIMER IN DECISION ANALYSIS

Cost-effectiveness analyses (CEAs) have been used widely in medicine to quan-
titate the benefits and costs of interventions. Their construction and reporting
follow some general rules, and further standardization should improve their
521
522 AuBuchon
applicability and comparisons. A brief introduction is offered here for those

unfamiliar with the techniques of CEAs.
There are several basic assumptions on which cost-effectiveness analyses
are predicated. The first is that there is a desire to maximize the health of the
population. This outcome is usually measured in terms of longevity, such as
additional years of life saved for a group of patients to which an intervention has
been applied. This measure is often modified by a quality-of-life factor. This
factor is meant to represent the perceived quality of life in the presence of residual
disease symptoms, morbidity from treatment side effects, or other factors that
prevent the patient from enjoying perfect health. The exact quality adjustment
factor assigned to various states of morbidity can be determined by an analysis
of patients preferences conducted in one of a variety of manners, such as
determining how much time they would be willing to trade from their lifespan
for amelioration of symptomatology or what risk of death they would accept in a
therapeutic maneuver to be free of symptoms (14). Multiplying the quality-of-
life factor for an outcome state (perfect health = 1.0; death = 0) by the longevity
yields a commonly used outcome measure, the quality-adjusted life-year (QALY).
This measurement of outcome is particularly useful because it is not linked to a
specific disease or treatment, and outcomes expressed in a generally applicable
term such as this can be readily compared.
A second precept of CEAs is that a choice must be made between interven-
tions that are competing for a finite pool of resources. Therefore quantitating the
difference in benefits obtained between two interventions and contrasting those
with the difference in the costs of their use should define the relative cost-
effectiveness of the approach under study. Thus the two interventions, such as
medical vs. surgical treatment of angina, are seen as competing for the limited
resources available to care for patients with this condition.
Practically, the construction of a CEA begins with a decision tree. In laying
out all the possibilities that may develop following selection of a particular course
of action, the tree represents all the various courses that a patient may follow after
an intervention is begun and the various outcomes that are possible. The tree
would include representations of all side effects of treatments and their outcomes,
partial or complete cures of a condition, etc. Once all the possibilities have been
embodied in the tree, the probabilities that each might occur need to be deter-
mined, either through a review of the literature, consultation with experts, or
performance of experiments. Similarly, the costs of each individual step must be
determined and the total expenditures for a patient reaching each outcome state
be summed. Also, the longevity of a patient in each outcome state and the quality
of life in those states must be quantified. Finally, mathematically rolling back
the tree allows calculation of the total outcome achieved with each intervention;
similarly, the costs associated with the intervention are tallied and compared
between interventions. The marginal cost-effectiveness is then calculated as:
Cost-Effectiveness Analysis of Risk-Reduction Strategies 523
Costeffectiveness(A vs. B) = (costsAcostsB)/(outcomesAoutcomesB)

There are, of course, many nuances and pitfalls in the construction and
evaluation of these decision trees and a variety of mathematical tools and
statistical models that can be applied. For example, the preference for spending
money later rather than in the present is usually captured by application of a
discount rate to distant costs of an intervention; similarly, future benefits are also
discounted, and the rate(s) applied must be reasonable representations of societys
time preferences (5). Models can also include tools [e.g., a Markov model (6)]
that represent conditions involving a variety of health states emerging over time.
The perspective chosen for the analysis can also be critical for its results. Usually,
a societal perspective is selected to capture globally the expenses and benefits
of an intervention. However, in some circumstances, the decision maker or the
party most directly affected by a decision can be identified more precisely, and
this may lead to inclusion of a different set of costs (7).
Properly constructed, a cost-effectiveness analysis can yield a concise
estimation of relative costs and benefits expressed in a common measure, dollars
per QALY, that can readily be compared between interventions. The reader is
referred to standard texts and consensus statements to gain additional information
on the construction and interpretation of these analyses (712).
III. SAMPLES OF COST-EFFECTIVENESS ANALYSIS

APPLICATIONS IN TRANSFUSION MEDICINE
A. Autologous Transfusion Strategies
A number of options in transfusion medicine have been evaluated for their
cost-effectiveness. These analyses are valuable from two points of view: determi-
nation of the cost-effectiveness of a technique and identification of factors having
an effect on the cost-effectiveness estimation (13,14). While the former applica-
tion may be of most interest to health economists and those charged with
health-planning decisions for society, it is the latter application that is most
meaningful and useful to those actually practicing transfusion medicine. Further-
more, application of CEA techniques to blood safety decisions has helped to
identify the importance of avoiding new complications as attempts are made to
decrease the likelihood of those that are already diminishing.
1. Preoperative Autologous Donation

Preoperative autologous donation (PAD) is an excellent example of how an
analysis can be used to highlight several aspects of an interventions use.
Although the benefits of PAD are self-evident, growing concern has been
expressed about the costs associated with this form of hemotherapy and the
524 AuBuchon
wastage that is inevitable since the need for transfusion is so difficult to predict.
Through CEA, the magnitude of the benefits and costs have been quantitated for
specific situations and for general application (Table 1). The fact that PAD has
much poorer cost-effectiveness than most medical interventions (1921) in many
applications is not surprising given the relatively high level of safety of allogeneic
transfusion at present (15,22,23) and the need to collect many more units than are
necessary for transfusion in order to have autologous units available to cover the
potential need.
Through these analyses, however, those charged with delivering this option
to patients can gain valuable insights into how best to apply the technique. For
example, while it may be intuitive that collecting blood for a surgical procedure
that rarely results in transfusion is likely to be wasteful, the magnitude of the
impact of such use of PAD is clearly evident in the calculations of Etchason et al.,
who documented that cost-effectiveness varied exponentially with the likelihood
Table 1 Cost-Effectiveness of Transfusion Risk Reduction Strategies
Cost-effectiveness
Intervention (cost/year of life extended)
Preoperative autologous donation

Coronary artery bypass surgery (2 units) (19) $500,000
Primary unilateral hip arthroplasty (20) $557,000
Primary unilateral knee arthroplasty (20) $1.3 million
Bilateral or revision arthroplasty (20) $140,000
Hysterectomy (20) $1.4 million
Transurethral resection of prostate (20) $23.6 million
Alanine aminotransferase testing (for NANB hepatitis
detection) (15)
Before anti-HCV test availability (cost-saving)
In addition to anti-HCV testing $7.9 million
HIV antibody testing (vs. no testing) (16) $3,600
Additions to HIV antibody testing
p24 antigen (16) $2.2 million
PCR for HIV genome (16) $2.0 million
Hepatitis B core antibody (17) $2.3 million
Solvent-Detergent Treatment of Plasma (62,63)
Overall (based on conservative plasma usage) $210 million
Published estimates of cost-effectiveness of a variety of interventions designed to improve transfusion
safety. In comparison, most medical/surgical interventions have a cost-effectiveness of <$50,000/year
of life extended.
NANB = Non-A, non-B; TTP = thrombotic thrombocytopenic purpura; anti-HCV = hepatitis C
antibody test; PCP = polymerase chain reaction testing.
of discard (21) (Fig. 1). Examining factors related to cost-effectiveness in

particular situations in more detail, Birkmeyer et al. (19,20) showed that not
only was collection of units without anticipated need likely to result in poor
cost-effectiveness, but transfusion of autologous units in situations where there
was less likelihood of clinical benefit also contributed to poor cost-effectiveness
(Table 2). Also, both of these groups have defined the importance of applying the
technique in those patients having greater expectations of longevity: the longer a
patient is expected to survive after surgery, the greater the time period in which
the patient could enjoy the benefits of having avoided viral infection through
allogeneic transfusion. Thus, pursuing PAD more aggressively in younger pa-
tients may offer more benefit than when PAD is applied in older patients.
These analyses have also highlighted the costs of certain decisions that have
been made. For example, testing autologous units for infectious disease markers
provides no benefit to the donor/recipient but adds cost to the system. This testing
has been explained as necessary (24) because of the potential for mistransfusion
of autologous units, i.e., transfusion to the wrong patient. While mistransfusion
remains a large problem in transfusion medicine, placing some of the burden for
it on autologous units distorts their true cost. Placing suitable, unused autologous
units in the allogeneic inventory would reduce some of the economic burden of
overcollection inherent in a PAD program and justify some of the testing costs.
However, crossover is not widely practiced for a variety of ethical and practical
reasons (23). CEAs can help identify the costs and implications associated
with policy decisions such as these and may prompt reconsideration of choices
or additional documentation of the rationale for their having been made in the
first place.
On reflection, the poor cost-effectiveness of PAD in many situations was
predictable. These calculations did provide useful assistance, however, in that
their quantitations assisted in providing imperatives for improvement in ordering
practices, for example (26), and they also yielded other, unexpected insights that
could help redefine the provision of PAD services. For example, the PAD analyses
mentioned above all included the assumption that the donation process was
risk-free. However, this is certainly not the case. Approximately 1 in 17,000
autologous donations is followed by reactions so severe that hospitalization is
required (27). The frequency of a serious postdonation reaction is even higher in
patients/donors not meeting usual allogeneic donation criteria; about 1 in 220 may
develop a serious reaction, and the identification of those at greatest risk is
problematic (28). These statistics are alarming when applied to patients scheduled
for CABG: removal of the assumption of safety of donation for these patients
demonstrated that only 1 in 101,000 donations by CABG patients would need to
result in a fatal peri-donation reaction before all the health benefits of having
autologous blood available were entirely negated (19). This poses the obvious
question: Can (this many) units be collected safely from a group of patients with
526
Figure 1 Effects of patients age (15, 40, 60, or 70 years) and likelihood of the use of autologous units on
the cost-effectiveness of PAD. Note the exponential relationship between cost-effectiveness and proportion
of PAD units actually transfused. Patient age (i.e., projected longevity) also has a large impact on
cost-effectiveness calculations for PAD. Except in cases of long postoperative lifespan and high likelihood
AuBuchon
of use, PAD cost-effectiveness is far poorer than that of most medical and surgical interventions (which
usually have cost-effectiveness estimates better than $50,000/QALY). (Data adapted from Ref. 20.)
Table 2 Factors Affecting the Cost-Effectiveness of PAD Applications
Patient variables Practice variables
Reaction risk Proportion of units discarded

Reaction consequences Transfusion threshold with autologous vs. allogeneic units
Longevity expectation
coronary artery disease so severe that surgery is necessary? While no study has
been of sufficient size to answer the question directly, a prospective donor also
needs to take into account that just delaying coronary revascularization in
order to allow for the collection of several units may have substantial mortality
riskapproximately 0.52% per month (29). These concerns call into question
whether PAD is the safest course of action for all patients.
Safety and efficacy questions extend to patients in more healthy states as
well. For example, 10% of patients donating for themselves before hysterectomy
ultimately needed to be transfused in a recent study, a transfusion rate 12 times
higher than that of patients who did not donate for themselves; the subset of
patients who donated preoperatively had lower admission hemoglobin values, and
autologous donation was an independent risk factor for transfusion (30). Obvi-
ously no benefit (at some cost and some, albeit small, risk) is achieved through
PAD when there is only a small likelihood of significant blood loss. Furthermore,
judicious use of erythropoetin to stimulate regeneration or expansion of red cell
mass and application of iron-replacement therapy may minimize complications
of repetitive phlebotomies (3133); however, any universal application of a
pharmacological adjunct or, indeed, any single PAD strategy cannot be supported.
Rather, judicious, patient-specific decisions must be made to identify the most
appropriate course of action based on the patients condition and the likelihood
of transfusion need, among other factors.
The cost-effectiveness analyses of PAD detailed above focused on preven-
tion of viral transmission as the primary benefit of using ones own blood for
surgery. Concerns have arisen over the last several years about the potential
immunosuppressive effect of allogeneic blood exposure and the possibility of
decreased tumor cell surveillance and increased risk of postoperative infection
following allogeneic transfusion (3437). Understanding the clinical significance
of these studies has been difficult because of the many confounding factors
embedded in the analyses and because of frankly contradictory results from
similarly designed studies. One group of researchers did include an estimate of
substantially increased risk of postoperative infection as a consequence of al-
logeneic transfusion in a CEA model of PAD before orthopedic surgery (38).
Inclusion of this complication led to the conclusion that PAD would be cost-
528 AuBuchon
effective or even cost-saving. On the other hand, a study performed for the
Canadian health care system concluded that even if the immunosuppressive
effects of allogeneic exposure were clinically significant, these would have a
relatively small impact on health care costs (39). Once the immunomodulatory
effects of allogeneic transfusion are well characterized and their clinical conse-
quences defined, any immunosuppressive effects of allogeneic transfusion could
be included in recalculation of a PAD model previously focused on viral trans-
mission. However, indications that leukoreduction may provide an alternative that
avoids this complication of allogeneic transfusion (40) raises another approach to
be considered: since a leukoreduction filter for a red cell unit generally costs
about half of the additional amount that a hospital must pay for an autologous
unit (41), leukoreduction may be a less costly and simpler approach to avoiding
this complication than PAD (42). For example, use of leukoreduced red cell units
in colorectal surgery resulted in a postoperative infection rate and length of
hospital stay similar to that of patients not requiring a transfusion and signifi-
cantly better than those receiving leukocyte-replete units. Additional studies to
define more precisely the relative costs and benefits of options to avoid compli-
cations such as this are necessary to clarify their cost-effectiveness.
2. Acute Normovolemic Hemodilution

Acute normovolemic hemodilution (ANH) is another technique that makes
autologous blood available. As the units remain with the patient and are not tested,
a simple, low-cost system is often envisioned that should be cost-effective.
However, as several mathematical models have demonstrated (43,44), the volume
of red cells saved by the procedure is quite small, often less than one units
worth, unless multiple units (four to six) can be collected from the patient and
significant blood loss actually occurs. In circumstances such as these, ANH can
supply at least as much autologous hemotherapy support as PAD and may do so
at approximately half the cost (45). As preoperative administration of erythro-
poietin may increase a patients red cell mass and allow more extensive hemodilu-
tion, or since use of a blood substitute may allow a lower hematocrit to be
tolerated during ANH, greater application of ANH may be seen in the future.
Again, there are insufficient data today to conduct a formal CEA, but additional
work toward reduction of the dose (and cost) of erythropoietin used in this
approach would improve cost-effectiveness estimates (46). Similarly, addition of
an oxygen-carrying solution to the hemotherapy protocol would allow for greater
reduction in a patients hematocrit, the collection of a larger number of units, and,
potentially, avoidance of a larger number of allogeneic units. As with other
autologous hemotherapy options, selection of patients who are most likely to need
transfusion and from whom adequate quantities of blood can be collected will be
central to determining the yield (and thus the cost-effectiveness) of the approach (47).
3. Intra- and Postoperative Red Cell Recovery

Intraoperative and postoperative recovery of red cells are also widely practiced
means of avoiding or reducing allogeneic transfusion. For surgical procedures in
which large volumes of blood are shed, recycling the red cells can dramatically
reduce the volume of allogeneic red cells required (48,49). The devices are often
used in cardiac surgery, although blood loss may be minimal in a simple
primary operation performed by a careful operator. In such cases, the devices are
used more as hemoconcentrators for blood left in the extracorporeal circuit at the
end of the procedure than as a tool to salvage red cells that would otherwise have
been discarded. Applications in vascular surgery and in orthopedic procedures
where several liters of blood may be lost have resulted in reductions in allogeneic
blood usage by over half. As the cost of disposables, amortization of the capital
equipment expense, and support of the operator can total $250500 per case, the
most commonly quoted wisdom is that a transfusion need of at least two to three
units of red blood cells must be anticipated before intraoperative recovery is less
expensive than allogeneic transfusion (50,51). [The distributed costs of infectious
disease transmission through transfusion are approximately $1/unit, and the
summed cost of all complications of allogeneic transfusion has been estimated at
less than $10/unit (46). Therefore, there is little room for providing an alternative
approach to allogeneic transfusion that reduces overall cost!]
Care must be taken in using intraoperative red cell recovery techniques
that complications of their use do not develop that outweigh the potential benefits.
For example, aspiration of tumor cells or bacteria may lead to an inadvertent
intravenous bolus of the cells since neither are completely removed by
even washing techniques (52,53). Operators must remain aware that only red
cells are being returned and that other components of blood may be needed
by the patient undergoing a massive transfusion. Finally, the potential for ac-
tivating a disseminated intravascular coagulation-like scenario when un-
washed red cells are reinfused must be considered, although this is a rare
phenomenon.
Of more questionable benefit is the application of technology to recover red
cells shed postoperatively. As with preoperative collection, a primary difficulty is
predicting who will bleed and who will need transfusion. In orthopedic surgery,
for example, the usual volume collected after arthroplastic surgery is approxi-
mately that of one unit of whole blood (300500 mL), but the hematocrit is much
lower (1525%) (54). As a result, only about 5% of patients receive a units worth
of red cells when postarthroplasty recovery is practiced. Therefore, the cost-
effectiveness of postoperative recovery of red cells usually provides little real
benefit to most patients at an extraordinarily poor cost-effectiveness estimate
(55,56). However, reducing the cost of postoperative salvage by piggybacking the
collection on disposables already used for intraoperative collection and reinfusion
530 AuBuchon
may make salvage of red cells lost postoperatively not only feasible but also
cost-effective.
B. Infectious Disease Marker Testing

As the risk of viral transmission through allogeneic transfusion continues to
fall, additional measures to reduce this risk further become increasingly less
cost-effective. Implementation of human immunodeficiency virus (HIV) antibody
testing in 1985 almost paid for itself, since it culled so many potentially infective
units from the allogeneic supply (16) (Table 1). On the other hand, starting from
a base of infrequent transmission (approximately 1/453,000 units with the most
current form of HIV antibody testing), addition of HIV p24 antigen testing
reduces the risk only a small amount (to approximately 1/676,000); this decreased
the projected number of annual HIV transmissions through transfusion in the
United States from 32 to 24. In consequence of this relatively small yield, its
cost-effectiveness was very poor (16). Recalculation of the cost-effectiveness of
HIV p24 antigen testing based on updated estimates of its yield suggest that its
true cost-effectiveness is more in the range of $20 million/QALY. Addition of
nucleic acid testing (NAT) techniques to close the window for HIV infection
would have even poorer cost-effectiveness. (The higher residual risk of HCV even
with version 3.0 HCV EIA testing would suggest that it would be a more
cost-effective target for NAT.) Similarly, use of alanine aminotransferase testing
as a surrogate marker for non-A, non-B hepatitis in the mid-1980s actually saved
the health care system money despite low sensitivity and specificity because of
the relatively high rate of hepatitis transmission. Implementation of a specific test
(anti-HCV) was also cost-saving since neither of the surrogate tests in place were
highly effective. However, once a very sensitive anti-HCV test was implemented,
the yield from surrogate testing plummeted, resulting in poor cost-effectiveness
estimates (15) and recommendations to drop this approach to avoiding non-A,
non-B hepatitis (57).
However, despite the advances in allogeneic transfusion safety that have
occurred in the last decade, the public, through the media and through their
congressional representatives, still clamors for increased levels of safety. While
regulatory agencies may be forced for political reasons to require additional
screening and testing measures, it is clear that these will have limited yields and
will result in increasingly astronomical cost-effectiveness estimations (58).
Beyond the economic perspective, there are potential safety drawbacks to
implementing new blood-testing efforts. A significant proportion (1525%) of
blood donors found to be HIV antibody positive admit on interview that they had
been aware of their exposure risks and donated, at least in part, to have HIV
testing performed in a free and socially acceptable venue (59). Media publicity
about introduction of a new and improved HIV test could attract more persons
to donate for the purposes of discerning their HIV status rather than altruism.
Because no test is 100% sensitive (and some window period of negativity will
always exist following infection and infectivity), implementation of a new test to
make the blood supply safer could have its impact negated or ever reversed (60).
The greatest risk of this magnet effect is in situations where the incremental
sensitivity gained from implementation is small and the probability of infection
is relatively high. This combination (plus a high proportion of false positives that
would have erroneously been interpreted by donors as indicative of AIDS) led
blood bankers generally to avoid use of hepatitis B core antibody testing as a
surrogate marker for HIV infection prior to availability of HIV antibody testing
(61). The potential for the magnet effect to yield an unexpected, unfortunate, and
unintended result must be kept in mind when new interventions to improve blood
safety are being considered.
C. Microbial Inactivation
Additional increments in blood safety may be more readily achievable, however,
through microbial inactivation of blood components. These techniques would
have to be applied to all units, the vast majority of which were already free from
infectious agents, and thus cost-effectiveness would be expected to be rela-
tively poor unless the invervention were quite inexpensive. [For example, if
solvent-detergent (SD) treatment of frozen plasma (FP) cost $20 per unit, the
cost-effectiveness of this intervention would be close to $300,000/QALY (18).
Updating the cost-effectiveness to current infectious risks and prevailing costs
brings the cost-effectiveness of SD FP to approximately $210 million/QALY
(62,63).]
Beyond the obvious economic concern, the potential side effects of a
microbial inactivation treatment must also be considered. Any of the treatments
under consideration at present have the potential for leaving a small chemical
residual in the unit or generating a reactive intermediate molecule. This raises the
potential for chemical toxicity or mutagenicity following transfusion. Undoubt-
edly, toxicological tests will be performed prior to licensure documenting that
such untoward effects are exceedingly unlikely and/or small. However, when the
risk of transmission of a viral agent is also extremely small, just a minute
toxicological risk could outweigh all the benefit obtained from viral inactivation.
Manufacturers are clearly searching for ways to minimize residual agent follow-
ing inactivation, such as through adsorption filtration to remove methylene blue
from plasma (64), but this concern will have to be evaluated individually for each
proposed inactivation method.
The SD treatment process raises another issue of unintended negative
effects. For manufacturing logistics and economic reasons, this technique is
performed on pools of plasma rather than on individual units. If the technique
532 AuBuchon
inactivated all microbes, the pooling would be of no concern. However, the SD

technique does not inactivate nonenveloped viruses, and the pooling might
provide access for a nonenveloped microbe from a single donor to be distributed
to a large number of recipients. While the nonenveloped viruses known to be
capable of parenteral spread, including hepatitis A and parvovirus B19, are of
limited concern for most transfusion recipients, a nonenveloped virus could
certainly arise with clinical consequences analogous to those of HIV infection.
Were such a virus to evolve, estimates from a CEA have suggested that such a
virus would have to be present at only an undetectably low frequency in the donor
population (1/71,000,000) before all the benefits of avoiding HIV, HBV, and HCV
through SD treatment of FP were negated (18). Concerns such as this have
prompted suggestions that the pool size for SD treatment should be made smaller,
but in most circumstances reducing pool sizes will have only a minimal impact
on avoidance of contamination (65).
Avoiding disease agents that are not readily inactivated, such as in
Creutzfeld-Jakob disease (CJD), would also seem prudent. While transmission of
CJD via transfusion has never been documented, even the remote possibility of
this event still generates fear. Because of a lack of a simple test for the disease
and the stability of CJD prions in a wide variety sterilizing solutions, some have
suggested deferring donors over the age of 50 or 60 to reduce the possibility of
donation by a person in the presymptomatic phase of CJD. Deferral of long-time
donors with a history of travel to the United Kingdom is another mechanism
designed to reduce the risk of transmissible spongiform encephalopathy. How-
ever, because these segments of the donating population are the least likely to be
recently infected with a blood-transmissible agent, their units represent a smaller
risk of transfusion-transmitted viral disease. Replacing these multiply tested
donors with first-time donors, who are also younger, may actually decrease the
safety of the donor pool (66). This is another situation where an unintended side
effect of a safety measure may actually produce new risk of a magnitude greater
than the original risk to be avoided.
D. Mistransfusion
Successful reduction of infectious disease risks through allogeneic transfusion
might have been predicted to lead to a redirection of attention to other transfusion
safety issues. Rather than this, however, additional effort has been directed toward
reducing these risks even further, ignoring estimates of low yield and concerns
about availability of financial support to match the desires for safer blood. As a
result, larger risks remain that might be addressed through interventions that are
more cost-effective than yet another infectious disease screening test.
An obvious example is mistransfusion and the risk of an ABO-related
fatality. The importance of ABO compatibility and the potentially catastrophic
consequences of an error have been appreciated for a century. However, the

frequency of fatal ABO errors has remained relatively constant (6769). The
majority of these are due to systems problems. The safety of the process
depends on human adherence to a protocol that is imperfect, and human error
can result in the transfusion of a unit of blood to someone other than the intended
recipient (67,7072). The magnitude of this problem is not trivial. European
studies have shown that errors surrounding transfusions occur with surprising
frequency. Misidentification occurred with 0.74% of patients and 0.2% of units,
and other major errors occurred with 0.5% of patients (73). A thorough, recent
study of transfusions in New York State documented that the same kinds of
problems that have been reported before are still occurring at the same frequency
despite safeguards such as unit labeling, double-checking of information, and so
forth (69). Extrapolating the frequency of mistransfusion (1/12,000 units) and
ABO-related fatalities (1/600,000 unit) to the country as a whole, approximately
two dozen patients die each year as a result of a fatal mistransfusion. This fatality
rate exceeds the number of HIV transmissions annually (22). However, the public
remains extremely fearful of HIV transmission through transfusion and ignorant
of the greater risks of mistransfusion; as a consequence, additional efforts are
directed toward viral safety and few are targeted at reducing mistransfusions.
This apparent contradiction is reflected in cost-effectiveness analyses as
well. A barrier system capable of preventing >99.99% of mistransfusions due
to labeling and patient identification errors is commercially available (74).
When its cost-effectiveness was evaluated from a societal perspective, it was
somewhat less cost-effective than most medical interventions (approximately
$200,000/QALY)although more cost-effective than most blood safety initia-
tives; however, when evaluated from a hospitals perspective (where the costs of
legal ramifications are logically included), use of the device was found to reduce
a hospitals overall costs (75). Thus society has directed the blood banking system
to focus resources toward reducing the risk of HIV transmission, although the
same resources directed at preventing mistransfusion would yield 10 times the
benefit. The reasons for this apparent inconsistency are explored in a later section.
E. Bacterial Contamination
The reappearance of concern regarding bacterial contamination (at least among
transfusion medicine professionals) highlights another dilemma faced by those
charged with providing a safe blood supply. While the risk of septic or endotoxic
shock after a red cell transfusion is on the order of 1 in 19,000,000 units, the
result is frequently fatal; far more platelet transfusions (perhaps 1/1,000113,000
platelet units) may harbor significant concentrations of bacteria, but because they
are more likely skin contaminants, their potential for inducing morbidity or
mortality is less (76). The source of these contaminants has been well described
534 AuBuchon
asymptomatic bacteremia of skin flora escaping surface decontaminationbut no

practical, generally feasible technological solutions have as yet been identified.
Gram staining and/or culture prior to unit release from the transfusion service has
been discussed most frequently, but the poor sensitivity of the former and the time
required for and expense of the latter have precluded widespread implementation
of these techniques. Genetic probe assays are under development that could
provide improved sensitivity (77), but the time and reagents involved may well
result in this bacterial detection system having a significant cost attached to its
use. Should such a system be used routinely when it is available? Given the
relatively short longevity of most transfusion recipients [due to their underlying
illness (78)] and the rarity of significant red cell contamination, even a detection
system of modest cost will probably be associated with astronomical cost-
effectiveness estimates when applied to red blood cells. Because of the higher
probability of bacterial contamination of platelets, such a system would provide
a higher safety yield when applied to this component. However, the mortality
associated with contamination of platelet units appear lower than for red cells,
and the even shorter longevity of the average recipient of platelets might still yield
a very poor cost-effectiveness estimate. If documentation of sterility allowed a
lengthened storage period, however, cost savings might result.
The previous paragraph outlines some of the major features of a CEA for a
bacterial detection system; once the details of such a practical system were
available, including its sensitivity, specificity, and cost, a formal CEA could
be readily constructed. Would such an analysis determine whether such an
intervention were warranted? Would the results help shape policy regarding its
implementation?
IV. DECISIONS, DECISION MAKERS, AND CEAs

The increasing constraints on health care expenditures in the United States have
made all in the health care field keenly aware of the importance of spending
resources wisely. Nevertheless, those charged with the direct care of patients have
not relinquished their duty to act as a patients advocate, seeking the best possible
treatment and outcome for a patient, regardless of its cost. At the same time, those
who administer health care plans and institutionssome of whom also have
patient care responsibilities in other rolesmust make difficult decisions about
what treatments to offer to whom under what circumstances in order to maximize
the benefit provided to the population served.
Judgments of decision makers in modern health care frequently employ
some type of formal or informal decision analysis tool. The questions Does it
work? or Is it better? always come first to a physician. These are followed by
What does it cost? and at least a quick mental check on the presumed return
for the investment. These questions may be followed by a formal CEA, but other
factors also need to be considered.
Avoidance of HIV is a good example. Why is the public scared to death
of HIV transmission through transfusion yet apparently complacent about the
risks of mistransfusion?
A key factor may be the inevitability of premature death associated with
HIV infection. Surety of death as an outcome will prompt extreme measures
of avoidance (79,80). Although an intervention such as PAD will provide ac-
tual benefit to only a tiny minority of those using this transfusion option, all
patients donating blood for themselveseven those ultimately not needing a
transfusionreceive the psychological benefit of having taken steps to avoid a
dreaded outcome. Not only is the intensity of this drive significant, but it is poorly
captured in CEAs. As only tangible benefits are readily counted in the outcomes
(unless lingering psychic stress would alter quality of life or require medical
intervention), the importance of avoiding HIV for many people would not be fully
captured in an enumeration of the years of life extended by prevention of a certain
number of HIV transmissions.
Another facet of fear leads to a parallel problem. Activities in which
the untoward outcome is obvious and predictable are associated with less dread
than risks that occur in an unseen fashion and that occur indiscriminately,
that is, are not abetted by the actions of the victim. A transfusion-transmitted
infection would thus fall into the quadrant of highest dread since it occurs without
warning to an unsuspecting and undeserving victim and manifests only after a
period of latency (81). This situation, coupled with inordinate fear of HIV in
particular, as detailed above, may help explain the publics irrational fear of
HIV transmission via transfusion and the continued push for further reductions in
that risk.
Finally, the rule of rescue may apply to transfusion safety decisions (82).
In considering various therapeutic interventions, there is a human tendency to
favor implementation of approaches that result in preventing an imminent death
over those that would prolong the life of a patient beginning at some time in the
future. Although most blood safety initiatives are preventive rather than ther-
apeutic in nature, the concept of the rule of rescue may still be applicable since
the outcome of a blood safety initiative is often seen as preventing a premature
death. In addition, although measures taken to improve the safety of the blood
supply ultimately benefit only a small number of recipients, this group of patients
is readily identifiable and highly visible. In most situations, both lay and profes-
sional people will opt for the intervention that benefits a small, clearly identifiable
group of recipients over one that provides a similar or even greater aggregate
benefit spread in small increments over a large number of people (83,84). Thus,
the benefit of a blood safety initiative may have more psychological value than
as measured in QALYs.
536 AuBuchon
Therefore, there are several psychological factors driving decisions to

increase the safety of the blood supply that are not readily embodied in a decision
tree or in the calculations of a CEA. What is the role, then, of decision analysis
techniques in transfusion safety decisions? Are they helpful at all? If properly
constructed and carefully reported, they can offer important information even if
they are not the primary decision-making tool.
In situations lacking high emotional content, CEAs might be expected to
provide immediately applicable guidance. For example, should leukocyte reduc-
tion or seronegativity be used to avoid transfusion-transmitted cytomegalovirus
(CMV) infection? If the two techniques have equivalent capabilities of interdict-
ing infectious units (85) but different associated costs, a simple comparative cost
analysis may direct future decisions. If additional studies reveal some difference
in their capabilities, then a CEA would reveal which approach provided the best
use of resources. In the current climate where there is competition for fixed-fee
contracts for marrow transplant patients and there is a strong focus on outcomes
(including length of hospital stay as well as mortality), a well-constructed CEA
could provide useful guidance for a CMV hemotherapy strategy.
In other situations, the conclusion of a CEA may not be congruent with the
final course of action selected. FDAs decision to recommend HIV p24 antigen
testing of donated blood was made with the knowledge of its very poor cost-
effectiveness but after congressional input that clearly indicated that blood supply
safetyand not costneeded to be the driving motivation. Nevertheless, CEA
can play a useful role even in these situations. To begin, their quantification of
costs and benefits afford a clearer understanding of the implications of the final
decision. Because a CEA must include a determination of costs, these data should
highlight what resources will be consumed if the intervention is implemented.
In a situation of finite resources, this information can be translated to changes in
the outcomes of other aspects of medical care since the resources available for
their use will necessarily be decreased. Alternatively, it might be hoped that those
persons or institutions requiring the interventions use would be in analogous
position of power to ensure that the required resources would be supplied.
For the individual health care provider, the economic output from a CEA is
probably only of academic interest. Physicians should continue to apply all the
resources available to their disposal to provide the best care possible consonant
with their patients wishes. However, these same CEAs can help guide physicians
decisions to provide information and insight so that they can indeed provide the
best care possible. The quantification of risks and the likelihood of various
outcomes may be key information for a patient faced with a difficult decision
without clear choices. Recognition that a particular intervention may be more
cost-effective in certain situations may prompt the health care provider to
investigate this aspect of the models implications carefully and lead him or her
to be more aggressive in applying the technique in situations where the benefits

are greatest and the risks minimized.
V. FUTURE BLOOD SAFETY DECISIONS

Many CEAs have already been completed addressing key questions in blood
safety. The successes of a variety of blood safety initiatives of the past decade
improved donor selection, expanded blood testing, availability of hemotherapy
alternativeshave left little room for improvement to be accomplished by future
interventions. Inevitably, as we progress on the safety curve to increasingly
flatter portions, smaller and smaller returns can be expected for the investment of
limited health care resources. Those having decision-making authority in these
situations must be made aware of the implications of their choices.
At the same time, those charged with implementing these options or directly
caring for patients will have to deal with tighter budgets and an increasing list of
options. CEAs can be helpful in highlighting trade-offs that must occur or options
that might be pursued in order to provide the greatest service to patients.
The tools of CEA cannot be expected always to identify the path that should
be taken, but they help illuminate the way that has been chosen.
REFERENCES
1. McNeil BJ, Pauker SG, Sox HC, Jr, Tversky A. On the elicitation of preferences for
alternative therapies. N Engl J Med 1982; 306:12591262.
2. Nease RF, Jr. Risk attitudes in gambles involving length of life: aspirations, variations
and ruminations. Med Decis Making 1994; 14:201203.
3. Llewellyn-Thomas HA. Patients health-care decision making: a framework for
descriptive and experimental investigations. Med Decis Making 1995; 15:101105.
4. Sutherland HJ. Assessing patients preferences (editorial). Med Decis Making 1995;
15;286287.
5. Chapman GB, Elstein AS. Valuing the future: temporal discounting of health and
money. Med Decis Making 1995; 15:373386.
6. Beck JR, Pauker SG. The Markov process in medical prognosis. Med Decis Making
1983; 3:419458.
7. Russell LB, Gold MR, Siegel JE, Daniels N, Weinstein MC, for the Panel on
Cost-Effectiveness in Health and Medicine. The role of cost-effectiveness analysis in
health and medicine. JAMA 1996; 276:11721177.
8. Weinstein MC. Clinical Decision Analysis. Philadelphia: Saunders, 1980.
9. Eisenberg JM. Clinical economics. A guide to the economic analysis of clinical
practices. JAMA 1989; 262:28792886.
10. Weinstein MC, Stason WB. Foundation of cost-effectiveness analysis for health and
medical practices. NEJM 1977; 276:716721.
538 AuBuchon
11. Weinstein MC, Siegel JE, Gold MR, Kamlet MS, Russell LB, for the Panel on
Cost-Effectiveness in Health and Medicine. Recommendations of the Panel on
Cost-Effectiveness in Health and Medicine. JAMA 1996; 276:12531258.
12. Siegel JE, Weinstein MC, Russell LB, Gold MR, for the Panel on Cost-Effectiveness
in Health and Medicine. Recommendations for reporting cost-effectiveness analyses.
JAMA 1996; 276:13391341.
13. AuBuchon JP. Lessons learned from decision analysis. Transfusion 1996; 36:755
760.
14. AuBuchon JP. The role of decision analysis in transfusion medicine. Vox Sang 1996;
71:15.
15. Busch MP, Korelitz JJ, Kleinman SH, Lee SR, AuBuchon JP, Schreiber GB. Declin-
ing value of alanine aminotransferase in screening of blood donors to prevent
posttransfusion hepatitis B and C virus infection. The Retrovirus Epidemiology
Donor Stuedy. Transfusion 1995; 35:903910.
16. AuBuchon JP, Birkmeyer JD, Busch MP. Cost-effectiveness of expanded HIV test
protocols for donated blood. Transfusion 1997; 37:4551.
17. AuBuchon JP, Birkmeyer JD, Busch MP, Dodd RY, Lackritz EM, Petersen LR and
HIV Blood Donor Study Group. Cost-effectiveness of anti-HBc testing to reduce HIV
transmission risk (abstr). Transfusion 1996; 36:43S.
18. AuBuchon JP, Birkmeyer JD. Safety and cost-effectiveness of solvent-detergent
treated plasma. In search of a zero-risk blood supply. JAMA 1994; 272:12101214.
19. Birkmeyer JD, AuBuchon JP, Littenberg B, OConnor GT, Nease RF, Nugent WC,
Goodnough LT. The cost-effectiveness of preoperative autologous donation in coro-
nary bypass surgery. Ann Thoracic Surg 1994; 57:161169.
20. Birkmeyer JD, Goodnough LT, AuBuchon JP, Noordsij PG, Littenberg B. The
cost-effectiveness of preoperative autologous blood donation in total hip and knee
replacement. Transfusion 1993; 33:544551.
21. Etchason J, Petz L, Keeler E, Calhoun L, Kleinman S, Snider C, Fink A, Brook R.
The cost effectiveness of preoperative autologous blood donations. N Engl J Med
1995; 332:719724.
Lewis WF, Notari EP, IV, Petersen LR. Estimated risk of HIV transmission by
screened blood in the United States. N Engl J Med 1995; 333;17211725.
24. Silvergleid AJ. All blood collected should be tested for infectious disease markers.
In: Maffei LM, Thurer RL, eds. Autologous blood transfusion: current issues.
Arlington, VA: American Association of Blood Banks, 1988:177182.
25. AuBuchon JP, Dodd RY. Analysis of the relative safety of autologous blood units
available for transfusion to homologous recipients. Transfusion 1988; 28:403405.
26. AuBuchon JP, Gettinger A, Littenberg B. Determinants of physician ordering of
preoperative autologous donations. Vox Sang 1994; 66:176181.
27. Popovsky MA, Whitaker B, Arnold NL. Severe outcomes of allogeneic and autolo-
gous blood donation: frequency and characterization. Transfusion 1995; 35:734737.
28. AuBuchon JP, Popovsky MA. The safety of preoperative autologous blood donation
in the non-hospital setting. Transfusion 1991; 31:513517.
29. Rankin JS, Hennein HA, Keith FM. The heart: I. Acquired diseases. In: Way LW, ed.
Current Surgical Diagnosis and Therapy. 10th ed. Norwalk, CT: Appleton & Lange,
1994:358382.
30. Kanter MH, van Maanen D, Anders KH, Castro F, Mya WW, Clark K. Preoperative
autologous blood donations before elective hysterectomy. JAMA 1996; 276:798801.
31. Brugnara C, Chambers LA, Malynn E, Goldberg MA, Kruskall MS. Red blood cell
regeneration induced by subcutaneous recombinant erythropoietin: iron-deficient
erythropoiesis in iron-replete subjects. Blood 1993; 81:956964.
32. Canadian Orthopedic Perioperative Erythropoietin Study Group. Effectiveness of
perioperative recombinant human erythropoietin in elective hip replacement. Lancet
1993; 341:12271232.
33. de Andrade JR, Jove M, Landon G, Frei D, Guilfoyle M, Young DC. Baseline
hemoglobin as a predictor of risk of transfusion and response to epoetin alfa in
othopedic surgery patients. Am J Orthop 1996; 25:533542.
34. Vamvakas EC, Taswell HF. Blood transfusion as a predictor of long-term mortality
(abstr). Transfusion 1993; 33:28S.
35. Vamvakas EC, Moore SB. Blood transfusion and postoperative septic complications.
Transfusion 1994; 34:714727.
36. Vamvakas E, Moore SB. Perioperative blood transfusion and colorectal cancer
recurrence: a qualitative statistical overview and meta-analysis. Transfusion 1993;
33:754765.
37. Vamvakas EC, Moore SB. Confounding and the effect of allogeneic transfusion on
survival (letter). Vox Sang 1995; 69:142143.
38. Healy HC, Frankforter SA, Graves BK, Reddy RL, Beck JR. Preoperative autologous
blood donation in total-hip arthroplasty. A cost-effectiveness analysis. Arch Pathol
Lab Med 1994; 118:465470.
39. Canadian Coordinating Office for Health Technology Assessment. Leukoreduction:
the techniques used, their effectiveness and costs. Ottawa: Canadian Coordinating
Office for Health Technology Assessment (CCOHTA), 1998.
40. Jensen LS, Kissmeyer-Nielsen P, Wolff B, Qvist N. Randomised comparison of
leucocyte-depleted versus buffy-coat-poor blood transfusion and complications after
colorectal surgery. Lancet 1996; 348:841845.
41. Forbes JM, Anderson MD, Anderson GF, Breecker GC, Rossi EC, Moss GS. Blood
transfusion costs: a multicenter study. Transfusion 1991; 31:318323.
42. Jensen LS, Grunnet N, Hanberg-Sorensen F, Jorgensen J. Cost-effectiveness of blood
transfusion and white cell reduction in elective colorectal surgery. Transfusion 1995;
35:719722.
43. Cohen JA, Brecher ME. Preoperative autologous blood donation: benefit or detri-
ment? A mathematical analysis. Transfusion 1995; 35:640644.
44. Goodnough LT, Monk TG, Brecher ME. Autologous blood procurement in the
surgical setting: Lessons learned in the last 10 years. Vox Sang 1996; 71:133141.
45. Monk TG, Goodnough LT, Birkmeyer JD, Brecher ME, Catalona WJ. Acute normo-
volemic hemodlution is a cost-effective alternative to preoperative autologous dona-
tion in patients undergoing radical retropubic prostatectomy. Transfusion 1995;
35:559565.
540 AuBuchon
46. AuBuchon JP. Blood transfusion options: improving outcomes and reducing costs.
Arch Pathol Lab Med 1997; 121:4047.
47. Brecher ME, Goodnough LT, Monk T. The value of oxygen-carrying solutions in
the operative setting, as determined by mathematical modeling. Transfusion 1999;
39:396402.
48. Solomon MD, Rutledge ML, Kane LE, Yawn DH. Cost comparison of intraoperative
49. Popovsky MA, Devine PA, Taswell HF. Intraoperative autologous transfusion. Mayo
Clin Proc 1985; 60:125134.
50. AuBuchon JP. Autologous transfusion and directed donations: current controversies
and future directions. Transf Med Rev 1989; 3:290306.
51. Solomon MD, Rutledge MI, Kane LE, Yawn DH. Cost comparison of intraoperative
52. Rumisek JD, Weddle RL. Autotransfusion in penetrating abdominal trauma. In: Hauer
JM, Thurer RL, Dawson RB, eds. Autotransfusion. New York: Elsevier/North Hol-
land, 1981:105113.
53. Karczewski DM, Lema MJ, Glaves D. The efficiency of an autotransfusion system
for tumor cell removal from blood salvaged during cancer surgery. Anesth Analg
1994; 78:11311135.
54. Umlas J, Foster RR, Dalal SA, OLeary SM, Garcia L, Kruskall MS. Red cell loss
following orthopedic surgery: the case against postoperative blood salvage. Transfu-
sion 1994; 34:402406.
55. Umlas J, Sager A. Cost-effectiveness of recovering blood post-arthroplasty (abstr.)
Transfusion 1994; 34:27S.
56. Jackson BR, Umlas J, AuBuchon JP. The cost-effectiveness of postoperative red cell
recovery for preventing transfusion-associated viral transmission following orthope-
dic surgery. Transfusion (in press).
57. Infectious disease testing for blood transfusions. NIH Consensus Development Panel
on Infectious Disease Testing for Blood Transfusions. JAMA 1995; 274:13741379.
58. AuBuchon JP, Birkmeyer JD, Busch MP. Blood Safety: Challenges and opportunities.
Ann Int Med 1997; 127:904909.
59. Leitman SF, Klein HG, Melpolder JJ, Read EJ, Estaban JI, Leonard EM, Harvath L,
Shih W-K, Nealon R, Foy R, Darr F, Alter HJ. Clinical implications of positive tests
for antibodies to human immunodeficiency virus type 1 in asymptomatic blood
donors. N Engl J Med 1989; 321:917924.
60. Korelitz JJ, Busch MP, Williams AE for the Retrovirus Epidemiology Donor Study.
Antigen testing for human immunodeficiency virus (HIV) and the magnet effect: Will
the benefit of a new HIV test be offset by the numbers of higher-risk, test-seeking
donors attracted to blood centers. Transfusion 1996; 36:203208.
61. Zuck TF, Eyster ME. Blood safety decisions, 1982-1986: perceptions and misconcep-
tions. Transfusion 1986; 36:928931.
62. Pereira A. Cost-effectiveness of transfusing virus-inactivated plasma instead of
standard plasma. Transfusion 1999; 39:479487.
63. Jackson BR, AuBuchon JP, Birkmeyer JD. Update of cost-effectiveness analysis for
solvent-detergent treated plasma. JAMA 1999; 282:329.
64. AuBuchon JP, Pickard C, Herschel L, OConnor JL, Williams S. Preparation of
methylene blue (MB) plasma with reduction in residual MB content by adsorption
filtration (abstr). Transfusion 1996; 36:65S.
65. Lynch TJ, Weinstein MJ, Tankersley DL, Fratantoni JC, Finalyson JS. Considerations
of pool size in the manufacture of plasma derivatives. Transfusion 1996; 36:770775.
66. Busch M, Glynn S, Schreiber G for the Retrovirus Epidemiology Donor Study.
Increased risk of viral transmission due to exclusion of older donors because of
Creutzfeld-Jakob disease (CJD) (abstr). Transfusion 1996; 36:59S.
fusion 1990; 30:583590.
68. Honig CL, Bove JR. Transfusion-associated fatalities: review of Bureau of Biologics
reports. Transfusion 1980; 20:653661.
69. Linden JV, Paul B, Dressler KP. A report of 104 transfusion errors in New York State.
Transfusion 1992; 32:601606.
70. Renner SW, Howanitz PJ, Bachner P. Wristband identification error reporting in 712
hospitals. A College of American Pathologists Q-Probes study of quality issues in
transfusion practice. Arch Pathol Lab Med 1993; 117:571577.
71. Reason J. Human Error. Cambridge: Cambridge University Press, 1990.
72. Voelker R. Treat systems, not errors, experts say. JAMA 1996; 276:15371538.
73. Baele PL, De Bruyere M, Deneys V, Dupont E, Flament J, Lambermont M, D,
Steensens L, van Camp B, Waterloos H. Bedside transfusion errors. A prospective
study by the Belgium SAnGUIS group. Vox Sang 1994; 66:117121.
74. Wenz B, Burns ER. Improvement in transfusion safety using a new blood unit and
patient identification system as part of safe transfusion practice. Transfusion 1991;
31:401403.
75. AuBuchon JP, Littenberg B. A cost-effectiveness analysis of using a mechanical
barrier system to reduce the risk of mistransfusion. Transfusion 1996; 36:222226.
Microbiol Rev 1994; 7:290302.
77. Brecher ME, Hogan JJ, Boothe E, Kerr A, McClannan L, Jacobs MR, Yomtovian R,
Chongokolwatana V, Tegtmeier G, Henderson S, Pineda A, Halling V, Kemper M,
Kuramato K, Holland PV, Longiaru. Platelet bacterial contamination and the use of
a chemiluminescence-linked universal bacterial ribosomal RNA gene probe. Trans-
fusion 1994; 34:750755.
78. Vamvakas EC, Taswell HF. Long-term survival after blood transfusion. Transfusion
1994; 34;471477.
79. Fowler FJ, Jr, Cleary PD, Massagli MP, Weissman J, Epstein A. The role of reluctance
to give up life in the measurement of the values of health states. Med Decis Making
1995; 15:195200.
80. Schneiderman LJ, Kaplan RM. Fear of dying and HIV infection vs hepatitis B
infection. Am J Publ Health 1992; 82:584586.
81. Morgan MG. Risk analysis and management. Sci Am 1993; 269:3235, 3841.
542 AuBuchon
82. Hadorn DC. Setting health care priorities in Oregon. Cost-effectiveness meets the rule
of rescue. JAMA 1991; 265:22182225.
83. Hux JE, Leviton CM, Naylor CD. Prescribing propensity: influence of life-expectancy
gains and drug costs. J Gen Intern Med 1994; 9:195201.
84. Rose G. Sick individuals and sick populations. Int J Epidemiol 1985; 14:3238.
86:35993603.
24
Red Blood Cell Substitutes
Zbigniew M. Szczepiorkowski and Christopher P. Stowell

Harvard Medical School and Massachusetts General Hospital, Boston,
Massachusetts
I. INTRODUCTION
A. Background
During the last 15 years there has been a rapid development in the number of
materials with the ability to transport and deliver oxygen and other gases. While
many of them were conceived as substitutes for red blood cells (RBC),* it is
becoming apparent that their unique characteristics suit them for new applications
beyond the scope of RBC. In this chapter, the different approaches that are being
taken to develop non-RBC oxygen carriers will be discussed and the status of the
clinical trials in progress as of June 2000 will be reviewed.
B. Desirable Characteristics for a RBC Substitute

The major factors driving the development of RBC substitutes have been con-
cerns about the adequacy of the blood supply and the infectious complications of
transfusion. Although the elimination of infectious risks has been a primary goal,
*The terms blood substitute and RBC substitute are used commonly, including in medical
literature databases. The term RBC substitute will be used in this chapter recognizing that these
oxygen-carrying materials are not intended to replicate all of the functions of the many elements of
whole blood and that they have characteristics that suit them for clinical roles beyond those possible
for the erythrocyte.
543
544 Szczepiorkowski and Stowell
a number of other favorable characteristics are highly desirable in a RBC

substitute, some of which are listed in Table 1.
A fundamental property for any RBC substitute, of course, is the ability to
carry physiologically useful amounts of oxygen and carbon dioxide in a volume
small enough to minimize the occurrence of volume overload or excessive
dilution of other elements of the blood. Not only must the substitute carry an
adequate amount of oxygen, it must also be able to load and unload gases under
the appropriate conditions. Since the capacity and oxygen-loading characteristics
of some RBC substitutes can be manipulated, the possibility of designing substi-
tutes for specific applications exists. For example, a high-affinity substitute may
be desirable to enhance oxygen loading in patients with impaired pulmonary
function, whereas a lower-affinity substitute may improve oxygen delivery to
tissue in shock.
In addition, the substitute must have a useful intravascular half-life. Most
of the RBC substitutes in development have half-lives in the range of 1224 hours
as compared to 4060 days for an erythrocyte. These half-lives may suffice in the
setting of acute bleeding but would obviously be inadequate for the support of
patients with chronic anemia.
At least initially, it was implicitly assumed that the oncotic pressure and
osmolarity of a blood substitute should be as close as possible to that of blood in
order to preserve biocompatibility and that low viscosity would be an asset by
reducing resistance to flow. More recent work with various oxygen carriers
indicates that the effect of these properties is more complex than was initially
appreciated. For example, the vascular endothelium is sensitive to changes in
Table 1 Desirable Characteristics for RBC Substitutes
Characteristics Requirement
Efficacy High capacity for O2 and CO2

Physiological gas exchange
Suitable intravascular half-life
Approximately isoncotic and isosmotic
Favorable rheological properties
Safety Minimal infectious risk
Minimal immunogenicity (e.g., neoantigens, xenoantigens)
Lack of toxicity
Limited extraneous physiological effects
Logistics Stability under a wide range of conditions
Immediate availability
Abundance
Low cost
Red Blood Cell Substitutes 545
shear stress, a property of a flowing fluid directly proportional to viscosity.

Decreased viscosity, uncompensated for by increased flow velocity, lowers the
shear stress, in response to which the vascular endothelium decreases the output
of various endothelium-derived relaxing factors such as endothelin and pro-
stacyclin (1,2). The resulting vasoconstriction paradoxically limits blood flow,
despite the improved hemodynamic characteristics of the RBC substitute.
Since a major driving force for the development of RBC substitutes has
been safety, a material intended for this purpose must essentially eliminate the
transmission of enveloped and nonenveloped viruses and pose a risk of bacterial
contamination no greater than the intravenous solutions in current use. It is
presumed that such materials would pose less of a risk, if one exists, of parenteral
transmission of prion-associated diseases.
Ideally, the substitute would be immunologically inert. Hypersensitivity
reactions or the development of a humoral or cellular response to the substitute
would obviously limit its utility. These concerns are particularly pertinent to
substitutes based on animal hemoglobin or chemically modified human hemoglo-
bins that may bear neoantigens. Preferably, the substitute would lack allo-
antigens altogether.
As with any therapeutic agent, the ideal RBC substitute must also have
minimal toxicity and lack biological effects other than those desired. Certainly
the untoward effects of the substitute must pose no greater hazard than the
transfusion of conventional blood components.
There are several practical considerations that have influenced the develop-
ment of RBC substitutes as well. The RBC substitute must be stable under
reasonable storage conditions, preferably at room temperature, and for prolonged
periods of time. Immediate availability without prolonged or complicated prepa-
ration is also an asset, particularly for use out of the hospital, such as at a trauma
site or on the battlefield. If it is to be clinically useful, the substitute must also be
available in large quantities and at a reasonable cost.
The potential advantages of RBC substitutes lie in the areas of safety and
convenience. The developmental goal has been to maximize these features while
retaining as much of the functional efficacy of the erythrocyte as possible.
C. The Major Approaches

Three quite different approaches have been taken in the development of RBC
substitutes. One line of investigation has pursued the use of the perfluorocarbons,
compounds in which gases, including oxygen, are highly soluble. The second
major approach has been to modify hemoglobin to minimize the problems created
when this molecule is free in the plasma while taking advantage of its favorable
and well-understood abilities to acquire and deliver oxygen. Progress towards
clinical application has been much greater with these two approaches than
with the development of liposome-encapsulated hemoglobin (LEH), or neo-

hematocytes, the third avenue of investigation.
Some key features of these three types of RBC substitutes are represented
in Figure 1 and compared to the native erythrocyte, whose characteristics are
diagrammed in the panels to the right. As each of the main types of RBC
substitutes is discussed, a comparison of its characteristics to those of the
erythrocyte will be made by referring to this figure.
II. PERFLUOROCARBON RBC SUBSTITUTES

A. Physiology and Background
The perfluorocarbons were first developed as part of the Manhattan Project in the
course of a search for an inert fluid in which highly reactive uranium species
could be handled. These cyclic or linear molecules consist of a carbon backbone
extensively substituted with fluorine atoms (Fig. 1). They are chemically and
biologically inert and do not support the growth of any known microorganism.
The characteristic that initially attracted interest to the perfluorocarbons is their
high solubility for gases. Volume for volume, perfluorocarbons can dissolve at
least 20 times more gas than can water. The oxygen content of perfluorocarbon
emulsion is linearly proportional to the oxygen tension in its environment, as
shown by the solid line in Figure 1. Oxygen loading and unloading is driven
by mass action, diffusing from areas of high oxygen tension to areas where it is
low, e.g, in vivo from the lungs to the perfluorocarbon and from the per-
fluorocarbon to the tissue. The dashed line in Figure 1 represents the oxygen
saturation curve for blood and demonstrates the important point that, at physio-
logical oxygen tensions, blood is much more readily saturated with oxygen and
has a greater oxygen-carrying capacity. The oxygen-carrying capacity of per-
fluorocarbons only approaches that of blood at very high oxygen tensions, levels
reached in vivo only when supplemental oxygen is provided. Nonetheless, the
oxygen capacity of the perfluorocarbons is substantially greater than that of
plasma or other intravenous replacement fluids.
The perfluorocarbons are not miscible in water, but they can be induced to
form stable emulsions with the addition of surfactants and stabilizers, such as
lecithin. Following intravenous injection into mammals, the droplets of the
emulsion are removed from the circulation by the reticuloendothelial system
(RES), usually in a matter of hours, as illustrated in Figure 1, in which the
intravascular clearance of a perfluorocarbon emulsion (solid line) is compared to
that of the erythrocyte (dotted line). The perfluorocarbons are eventually exhaled
via the lungs, but they may remain in the RES for prolonged periods before
excretion. Some of the earlier perfluorocarbons studied remained in the RES for
months; the newer formulations are excreted within several days.
HEMOGLOBIN SOLUTIONS
PFC UNMODIFIED MODIFIED LEH ERYTHROCYTES
F F
c F c F
F c F c
F F
STRUCTURE
Red Blood Cell Substitutes
150 300 450 50 100 150 50 100 150 50 100 150 50 100 150
O2 SATURATION
CONCENTRATION
CONCENTRATION
CONCENTRATION
CONCENTRATION
CONCENTRATION
CLEARANCE
12 24 60 120 12 24 60 120 12 24 60 120 12 24 60 120 12 24 60 120
hours days hours days hours days hours days hours days
Figure 1 Comparison of several characteristics of three types of RBC substitutes to one another and to erythrocytes: structure,
oxygen binding, and intravascular clearance. The dotted lines in the panels for the RBC substitutes represent the behavior of
the erythrocyte and are shown for comparison. The RBC substitutes are perfluorocarbon (PFC), unmodified and modified
547
(polymerized/cross-linked) hemoglobin, and liposome-encapsulated hemoglobin (LEH).

Some of the potential advantages and disadvantages of perfluorocarbons as

RBC substitutes are listed in Table 2. Because of their synthetic source, the
composition of these materials can be strictly controlled and could be modified
to meet specific needs. Most of the perfluorocarbon emulsions have extended
shelf lives, and production of huge quantities at relatively low cost is possible.
Infectious and immunological complications are virtually nonexistent, although
the same might not be true for other components of the preparation such as the
surfactant or the stabilizer, particularly if they are of biological origin. The
viscosity of perfluorocarbon emulsions is low, which might favor its use of
oxygenating low flow or obstructed areas not accessible to erythrocytes.
One of the drawbacks of using perfluorocarbons in biological systems is
the requirement for preparing them as emulsions. While this technology is
reasonably well established, reproducibly obtaining emulsions with uniform
characteristics and guaranteed stability is a challenge. The sizes of the emulsion
particles may be quite heterogeneous with unknown biological effects. Clearance
of the different perfluorocarbons is highly variable and the consequences of their
extended half-lives in the RES are unknown. The need for elevated oxygen
tensions in order to load these preparations with useful amounts of oxygen also
imposes limitations.
The first perfluorocarbon to reach clinical trials was Fluosol-DA, a mixture
of two perfluorocarbons, a surfactant, and an emulsion stabilizer that required
storage as a frozen emulsion (Table 3). In early clinical trials, Fluosol-DA proved
to be relatively well tolerated. Although 510% of people receiving this per-
fluorocarbon emulsion experienced rash and blood pressure lability, these effects
were eventually attributed to the emulsion stabilizer. Hepatosplenomegaly was
also observed accompanied by mild, transient changes in liver function tests,
presumably reflecting the uptake of perfluorocarbons by the RES. Mild elevations
in serum creatinine and blood urea nitrogen (BUN) were noted in some patients
as well. Phase II clinical trials were conducted administering Fluosol-DA to
Table 2 Potential Advantages and Disadvantages of Perfluorocarbon RBC Substitutes
Advantages Disadvantages
1. Control of composition 1. Requirement for emulsification/stabilization

2. Possibility of specific modification 2. Heterogeneous particle size
3. Large-scale production 3. Variable (long) RES clearance
4. Low production costs 4. High FiO2 required
5. Prolonged shelf life 5. Low O2 capacity at physiological PO2
6. Minimal infectious risk 6. Rapid plasma clearance
7. Minimal immunogenicity
8. Low viscosity (?)
Table 3 Perfluorocarbon RBC Substitutes Studieda
Product
(manufacturer) Perfluorocarbon Trial level Application
Fluosol-DA Perfluorodecalin Phase II Acute blood loss

(Green Cross/ Perfluoropropylamine (discontinued)
Alpha)
Approved PTCA
(withdrawn)
Oxygent Perflubron Phase II CABGANH,
(Alliance) surgeryacute
blood loss
Phase III SurgeryANH
Imagent GI Perflubron Approved GI imaging
(Alliance)
Liquivent Perflubron (neat) Phase Ib/II Liquid ventilation
(Alliance) Phase II/III IRDS
(discontinued) Liquid ventilation
pedi and adult
ARDS
Oxyfluor Perfluorodichlorooctane Phase II Surgery, Neuro-
(HemaGen/PFC) (discontinued) protectant/bypass
aInformationcurrent to 6/00.
PCTA, Percutaneous, transluminal coronary angioplasty; CABG, coronary artery bypass graft; ANH,
acute normovolemic hemodilution; GI, gastrointestinal; IRDS, infant respiratory distress syndrome;
ARDS, adult respiratory distress syndrome.
patients with acute, life-threatening bleeding; however, other than a temporary

improvement in hemodynamics, the performance of the preparation was disap-
pointing (3). The poor results were due, at least in part, to the low dose of
perfluorocarbon administered, the low oxygen capacity of this particular prepa-
ration, and the dire condition of the patients receiving it. After this experience,
interest in Fluosol-DA as a RBC substitute waned rapidly and clinical trials for
this application ceased. Fluosol-DA was eventually licensed by the U.S. Food and
Drug Administration (FDA) for oxygenation of the distal vascular bed in percu-
taneous transluminal coronary angioplasty (PTCA) (4), but it was removed from
the market because of poor sales and improvements in angioplasty catheter
design. However, the shortcomings of Fluosol-DA were instrumental in highlight-
ing where improvements were necessary and guiding the development of new
perfluorocarbons.
B. The Next Generation of Perfluorocarbon

Oxygen Carriers
1. Perflubron
Two manufacturers have conducted clinical trials of new perfluorocarbon oxygen
carriers (see Table 3). Alliance Pharmaceutical Corporation has developed an
emulsion of perfluorooctyl bromide (Perflubron) stabilized with lecithin that has
a much higher oxygen capacity than earlier perfluorocarbon preparations and can
be stored as an emulsion at room temperature (5). Physiological studies of
Perflubron emulsion in animal model systems showed that it was capable of
delivering oxygen to several target tissues with improvement in their metabolic
condition and function (68). In human safety studies, it was well tolerated with
the principal side effects being a mild thrombocytopenia (nadir in the range of
100150 109/L) that resolves within a few days and a transient, flu-like
syndrome, which begins within a few hours of the infusion. The latter has been
attributed to the release of cytokines and arachidonic acid metabolites from the
cells of the RES as the perfluorocarbon is phagocytosed (9).
One formulation of perflubron, named Oxygent, is being developed as a
means of facilitating oxygen delivery during acute normovolemic hemodilution
(ANH). Instead of replacing the withdrawn blood with crystalloid solution, an
oxygen carrier such as Oxygent is used, thereby minimizing any potential
problems created by the iatrogenically produced anemia. Oxygent has been tested
as a replacement fluid in ANH in general surgery patients and in patients
undergoing cardiopulmonary bypass in phase II and III clinical trials.
Although the use of Oxygent as an extender in ANH is being pursued
vigorously, other applications have been studied and illustrate the potential
versatility of the perfluorocarbons. The ability of the perfluorocarbon emulsions
to deliver oxygen is being exploited to sensitize solid tumors to radiation. Oxygen
potentiates the effects of radiation and chemotherapy on malignant cells, and its
absence in hypoperfused and hypoxic areas of a tumor limits the effectiveness of
these therapies (10). In animal studies (11) and phase II studies in patients with
solid tumors, Oxygent augmented the effect of radiation therapy, presumably by
improving oxygen delivery to the tumor.
Perflubron, unlike most of the perflurocarbons, contains a bromine substit-
uent that renders it radiopaque. A Perflubron preparation named Imagent has been
studied as a contrast medium for several imaging applications and was approved
(as Imagent GI) for use in gastrointestinal radiography, although it is not being
actively promoted. Imagent US is a Perflubron formulation that has reached phase
II trials for use as an ultrasound contrast medium (5) for the detection of tumors
and metastases as well as for cardiac applications. However, Alliance is presently
not pursuing these imaging applications.
One of the most novel applications of the perfluorocarbons is in partial

liquid ventilation. In this application, perfluorocarbon is instilled, neat, into a
patients lungs, partially filling the alveoli. The instilled perfluorocarbon serves
as a source of readily available oxygen and as a surfactant as well, helping to
expand the alveoli and improve gas diffusion. In preliminary studies, Perflubron
(Liquivent) was instilled into the lungs of infants with severe infant respiratory
distress syndrome (12). Based on the improvement in the clinical condition of
many of the patients, phase II and subsequently phase III studies were conducted
in infants, children, and adults with respiratory distress syndrome. However,
patient accrual in the phase III trials was discontinued because of unexpectedly
high mortality rates in the treatment arm.
2. Oxyfluor
HemaGen/PFC formulated a perfluorocarbon emulsion named Oxyfluor consist-
ing of perfluorodichlorooctane stabilized with lecithin and safflower oil, which
may be stored for a year at room temperature. Animal studies showed that
Oxyfluor was capable of delivering oxygen in models of shock resuscitation and
surgical bleeding and was well tolerated. Lung hyperinflation was noted in the
treated animals, as has been observed with many of the perflurocarbons, although
it has not emerged as a significant problem in humans. Safety studies in humans
have demonstrated that thrombocytopenia and a flu-like syndrome occur with this
perfluorocarbon preparation, as was the case with Oxygent.
Although Oxyfluor underwent phase II trials in surgery patients,
HemaGen/PFC chose to exploit the ability of perfluorocarbons to dissolve gases
by using Oxyfluor to remove the microbubbles that form in the circulation of
patients undergoing cardiopulmonary bypass. These microbubbles are thought to
embolize to the microvasculature of the brain, where they produce the neurolog-
ical and neuropsychiatric changes seen in some patients following bypass (13,14).
In animal cardiopulmonary bypass model systems, Oxyfluor was effective in
scavenging microbubbles and reducing the formation of these microemboli (15).
HemaGen/PFC ceased operations for financial reasons. Hence, Oxyfluor is no
longer in development.
III. HEMOGLOBIN-BASED RBC SUBSTITUTES

A. Physiology and Background
The concept of using cell-free hemoglobin solutions to replace RBC has an
obvious rationale and was first demonstrated to be feasible by Mulder, who in
1934 showed that cats could survive and function normally for a brief period of
time after their blood was replaced with stroma-free hemoglobin. Some of the
advantages of using hemoglobin as an RBC substitute are listed in Table 4, among

them good oxygen-carrying capacity and the ability to function at physiological
oxygen levels. The potential advantages and disadvantages of low viscosity were
discussed previously. The oncotic pressure of hemoglobin solutions is high, which
may be particularly desirable in the treatment of acute bleeding where volume
resuscitation is, if anything, more important than the restoration of RBC mass.
The absence of erythrocyte antigens eliminates the need for compatibility testing,
and the stability of hemoglobin solutions, even at room temperature, reduces
many of the logistic problems entailed in red cell transfusion. In addition to the
processes for isolating and purifying the source hemoglobin, the technology for
inactivating and removing viruses from protein solutions has the potential of
reducing the infectious risks to the levels associated with plasma derivatives such
as albumin and intravenous immunoglobulin.
There are, however, a number of problems that can arise when cell-free
hemoglobin is present in the plasma (Table 4). As a result, considerable effort has
been expended to modify cell-free hemoglobin so as to minimize these potential
problems. The extent to which these strategies have been successful is in the
process of being evaluated in clinical trials.
The first of the problems with cell-free hemoglobin is that the tetramer
rapidly dissociates to dimer in the circulation and is filtered out by the kidney,
producing hemoglobinuria and having toxic effects on the proximal tubular
cells. The other consequence, of course, is that the infused hemoglobin disap-
pears within a matter of a few hours (see Fig. 1) via glomerular filtration and
haptoglobin-mediated clearance.
The loss of hemoglobin from the plasma and its toxic effect on the kidneys
can be avoided if dissociation to the dimer is reduced. One of the strategies that
has been effective in minimizing dissociation is to crosslink the dimers forming
stable tetramers or higher order oligomers. The reagents used to form stable,
higher molecular weight hemoglobin species include bis(3,5-dibromosalicyl)
Table 4 Advantages and Disadvantages of Unmodified Hemoglobin
1. High capacity for O2 and CO2 1. Rapid clearance

2. Functions at physiological PO2 level 2. Renal toxicity
3. Low viscosity (?) 3. Vasoactivity
4. High oncotic pressure 4. Increased oxygen affinity
5. Absence of RBC antigens 5. Autoxidation
6. Prolonged shelf life 6. Immunogenicity (modified or nonhuman)
7. Purification/viral inactivation possible 7. Potentiation of sepsis (?)
fumarate, glutaraldehyde, and open-chain raffinose. Another approach to mini-

mize rapid clearance of hemoglobin is to increase its molecular weight by
attaching oligomers of polyoxyethylene (POE) or polyethylene glycol (PEG).
A second potential problem with hemoglobin in solution is that it is
vasoactive, a property that manifests as a hypertensive effect in both animals and
humans. Hemoglobin binds nitric oxide (NO), thereby releasing the constitutive
relaxing effect it exerts on the smooth muscle in the vascular wall and producing
vasoconstriction (1618). In general, the hemoglobin-based RBC substitutes with
the smallest proportions of dimers and tetramers also exhibit the least vasoactiv-
ity. Other mechanisms undoubtedly also influence the vasoactivity of hemoglo-
bin-based RBC substitutes, among them the autoregulatory response to oxygen
delivery (19,20) and the effect of viscosity and shear stress on endothelin-regu-
lated vascular tone (21).
A third potential problem with hemoglobin in solution is the tendency for
its oxygen affinity to increase, primarily in response to the loss of 2,3-diphospho-
glycerate (2,3-DPG). As oxygen affinity increases and the oxyhemoglobin disso-
ciation curve shifts to the left, as shown in Figure 1, a greater degree of tissue
hypoxia is required to induce oxygen unloading. Although there may be situations
in which a high-affinity oxygen carrier would be useful, the general assumption
has been that hemoglobin-based oxygen carriers should replicate the characteris-
tics of intraerythrocytic hemoglobin. Hence, restoration of the native oxyhemo-
globin dissociation curve has been achieved by several methods, among them
chemical modification. For example, pyridoxylation and di-aspirin crosslink
formation also have the salutary effect of shifting the dissociation curve to the
right. Another way to avoid the left shift is through the use of bovine hemoglobin,
whose activity is modulated by chloride ion in a manner analogous to the
allosteric response of human hemoglobin to 2,3-DPG. When released into a
solution with a physiological chloride ion concentration, bovine hemoglobin does
not exhibit the left shift seen with human hemoglobin. Other chemical modifica-
tions, as well as the use of site-directed mutagenesis to introduce chloride-binding
sites into recombinant human hemoglobin, are possible strategies to minimize
this change.
In solution, extraerythrocytic hemoglobin tends to autoxidize to methemo-
globin and eventually to irreversibly altered metabolites (22). In the erythrocyte,
methemoglobin reductase operates to reclaim methemoglobin, but it is obviously
not available to cell-free hemoglobin. Autoxidation might result in the loss of
functional hemoglobin and, even more concerning, lead to the formation of
free radicals. Although methemoglobin does form during the storage of hemoglo-
bin in solution, it represents only a small proportion (<10%) of most of the
hemoglobin-based oxygen carriers that have been studied.
The possibility that animal hemoglobin, or even human hemoglobin that
has been chemically altered, might stimulate an immune response must also be
considered as a potential drawback to the use of these materials. The tendency to

stimulate immune reactions must be established for each of the hemoglobin
preparations, particularly if repeat use is anticipated. The coupling of PEG to
hemoglobin may have the effect of reducing its antigenicity.
The concern that cell-free hemoglobin might exacerbate sepsis was raised
by the results of the experiments in a murine model system where the infusion of
autologous hemoglobin was associated with increased mortality following bacte-
rial challenge (23). This potentiating effect might be related to the ability of
hemoglobin to promote the growth of microorganisms or to impair RES clearance
of bacteria. Cell-free hemoglobin may also have a direct effect on sepsis, since it
has been shown to bind bacterial endotoxin and enhance its activity in vitro (24).
In vivo, exogenous hemoglobin seems to potentiate the lethal effect of endotoxin
administered to mice (25), although there is also a report of the failure of
polymerized hemoglobin to increase mortality in another murine infection model
system (26). Until the basis for these observations of in vitro and animal systems
is better understood, potentiation of sepsis must be regarded as a possible hazard
of infusion of hemoglobin-based RBC substitutes.
Some of the potential problems with cell-free hemoglobin, and the ap-
proaches that have been taken to minimize them, are summarized in Table 5.
B. Human Hemoglobin-Based RBC Substitutes

Seven hemoglobin-based RBC substitutes have reached various stages of devel-
opment in the past decade, although only five are still being actively pursued.
Four of these use human hemoglobin; two are based on bovine hemoglobin; and
Table 5 Approaches Used to Solve Problems Associated with

Cell-Free Hemoglobin
Problem Possible modifications
Rapid clearance Polymerization

Increase molecular weight
Stabilization of tetramer (di- recombinant)
Vasoactivity Polymerization
Increase molecular weight
High oxygen affinity Bovine hemoglobin
Chemical modification
Site-directed mutagenesis
Autoxidation Site-directed mutagenesis
Chemically linked methemoglobin reductase
Immunogenicity Attach polyethylene glycol
one takes advantage of recombinant technology to produce a modified human

hemoglobin in vitro. Some of the features of each of these RBC substitutes, which
are summarized in Table 6, are as follows:
Table 6 Modified Hemoglobin-Based RBC Substitutesa
Product Hemoglobin
(manufacturer) source Trial level Application
PolyHeme Human Phase III Trauma, surgery

(Northfield)
Hemolink Human Phase II Cardiopulmonary bypass
(Hemosol) ANH, orthopedic
surgeryacute blood
loss, dialysis
Phase III Cardiac surgery
HemAssist Human Phase II Septic shock, hemodialysis,
(Baxter) hemorrhagic shock, car-
diopulmonary bypass
Phase III Acute blood losssurgery,
(all trials trauma
terminated)
PHP Human Phase Ib/II NO-induced shock
(Ajinomoto/Apex)
PEG Hemoglobin Bovine Phase Ib Radiosensitizer solid tumors
(Enzon)
Hemopure Bovine Preclinical Erythropoiesis
(Biopure) Phase II Sickle cell crisis, oncology,
surgeryorthopedic,
urological, vascular,
cardiac surgery
Phase III Surgerycardiac,
orthopedic
Oxyglobin Bovine Approved Veterinaryanemia, acute
(Biopure) blood loss
Optro Recombinant Phase I ErythropoiesisESRD,
(Somatogen refractory anemia
Baxter)
Phase II ANH, Acute blood loss
surgery
(all trials
terminated)
aInformation
current to 6/00.
ESRD; End-stage renal disease.
1. PolyHeme is a preparation of hemoglobin from outdated human blood

that has been polymerized with glutaraldehyde in a process developed
by Northfield Laboratories. In human safety studies, it was well toler-
ated with a notable lack of the pressor effect seen with many hemoglo-
bin-based RBC substitutes (27). Northfields goal for this product is to
develop it as an RBC substitute i.e., as a true equivalent to units of
donor blood that may be used in similar circumstances. PolyHeme has
been used successfully to maintain adequate total hemoglobin levels
(intraerythrocytic plus cell-free hemoglobin) in trauma patients, where
it supplanted conventional RBC transfusions (28). Phase III testing in
trauma patients and patients undergoing surgery is in progress.
2. Hemolink is the primary hemoglobin derivative being developed by
Hemosol, Ltd. and is prepared by polymerizing human hemoglobin
using open-chain raffinose. The products subsequently undergo exten-
sive purification and viral-inactivation procedures and are stable for at
least 60 days at room temperature. This material has undergone phase
I and phase II testing and has initiated phase III trials for acute blood
loss in surgery. The eventual goal is to develop this material as an
extender in acute normovolemic hemodilution as well as in acute
surgical or traumatic blood loss.
3. HemAssist, also known as di-aspirin crosslinked hemoglobin (DCLHb),
is a product of Baxter Healthcare, which was produced by crosslinking
the two chains of human hemoglobin using the reagent bis(3,5-
dibromosalicyl) fumarate. The cross-linked hemoglobin was purified,
heat inactivated, and stored frozen. During early animal experiments,
it was noted that this material had a modest hypertensive effect,
which was originally attributed to binding of NO, although endo-
thelin release from vascular endothelial cells (29) and adrenergic ef-
fects probably also contribute to the pressor activity. It was particularly
noteworthy that the vasoactivity of HemAssist countered the effects
of hemorrhagic shock in an animal model system (30). Baxter sought
to exploit this property in phase II trials, where HemAssist was
administered to patients with septic shock, a condition associated
with elevated NO levels. Baxter also pursued the use of HemAssist
in hemodialysis to reduce the frequency of hypotensive episodes
and in patients with hemorrhagic, hypovolemic shock. However,
phase III studies in surgery and trauma patients were terminated volun-
tarily in 1998 because of safety considerations (30a). Baxter, which had
acquired Somatogen in 1998, also terminated trials with its recombi-
nant hemoglobin-based RBC substitute, Optro. They have announced
plans to develop a second-generation substitute based on recombinant
hemoglobin.
4. Pyridoxylated hemoglobin polyoxyethylene (PHP) is produced by a

Japanese company, Ajinomoto, but it is being developed in this coun-
try by Apex Bioscience. This material is prepared by pyridoxylating
human hemoglobin to adjust the oxyhemoglobin dissociation curve
back to the right, followed by coupling it to polyoxyethylene to reduce
renal clearance. Phase I and II testing did not reveal any clinically
important side effects but did demonstrate that this preparation is
vasoactive. The clinical role for this substitute, which is being targeted
by Apex, is septic shock, based on the supposition that PHP will bind
excess NO and exert a pressor effect.
C. Bovine Hemoglobin-Based RBC Substitutes

Two RBC substitutes based on bovine hemoglobin are presently in clinical trials.
Bovine hemoglobin, which shares 90% amino acid sequence homology with
human hemoglobin, has a few advantages over its human counterpart, the first
being its allosteric response to chloride ion and its independence from 2,3-DPG.
Bovine hemoglobin is potentially available in large quantities and can be obtained
from closed herds with well-documented and regulated health histories, unlike
human source hemoglobin. The relative scarcity of human hemoglobin, which
must be obtained from outdated donor RBC, may result in higher production
costs for human hemoglobinbased substitutes compared to those based on
bovine hemoglobin.
Another advantage of bovine hemoglobin is that there is obviously no risk
of transmission of human pathogens. On the other hand, public concern about the
possibility that viruses or prions might cross over from one species to another,
even when they are distantly related, may slow acceptance. The attention attracted
to the possible relationship between human new variant Jakob-Creutzfeldt disease
and bovine spongiform encephalitis certainly attests to the public wariness about
possible new pathogens.
Given the high degree of amino acid sequence homology between human
and bovine hemoglobins, immunogenicity would be expected to be low. Whether
or not it would exceed the immunogenicity of modified human hemoglobin will
have to be determined during clinical trials.
1. PEG hemoglobin is prepared by Enzon by coupling multiple molecules
of PEG to bovine hemoglobin followed by purification. The coupled
PEG increases the molecular weight so that renal clearance is reduced
as well as immunogenicity. PEG hemoglobin persists in the circulation
for about 48 hours, which is longer than for most of the hemoglobin-
based RBC substitutes, which are generally cleared within 2436
hours. The oncotic pressure and viscosity of PEG hemoglobin are
higher than for many of the hemoglobin-based RBC substitutes. Its

volume-expanding properties may make it particularly suitable for use
in acute blood loss, ANH, and hypovolemic shock.
PEG hemoglobin was found to be safe and effective in an animal
exchange transfusion model (31,32) but is being developed primarily
as a sensitizing agent for radiotherapy. Phase I safety studies have been
completed and a multicenter, dose-escalation, phase II trial is planned
to evaluate PEG hemoglobin as a sensitizer in patients receiving
palliative radiation therapy for solid tumors.
2. Hemopure is a bovine hemoglobin preparation polymerized with glu-
taraldehyde and purified to remove tetramers and dimers. Hemopure
can be stored for two years at room temperature. The first formulation,
which contained albumin and less highly polymerized hemoglobin,
was found in early safety studies to produce gastrointestinal dys-
motility and pain, presumably as a result of smooth muscle spasm
induced by NO binding (15,16). These experiences led Biopure, the
company developing this product, to more thoroughly eliminate the
unpolymerized dimers and tetramers. The present formulation has a
much lower incidence of gastrointestinal symptoms, primarily mild
dysphagia, which is similar to several of the other hemoglobin-based
RBC substitutes.
In phase I studies in volunteers and three different groups of
patients (sickle cell crisis, orthopedic surgery, prostatectomy), the
product was well tolerated (33). Improved exercise tolerance was noted
in the sickle cell patients (34). Phase II trials, using Hemopure for
perioperative transfusion in patients undergoing cardiac and vascular
surgery, as well as in sickle cell crisis, have been carried out. Phase III
trials in which patients undergoing noncardiac surgery may receive up
to 10 units of Hemopure are in progress. A veterinary formulation,
Oxyglobin, was licensed by FDA in 1997.
An interesting and novel property of Hemopure is its apparent
ability to stimulate erythropoiesis. This property may be due in part
to a hematinic effect, since elevations in serum iron, ferritin, and
erythropoietin levels were seen after infusion of this hemoglobin
preparation (35).
D. Recombinant Hemoglobin-Based RBC Substitutes

One company, Somatogen, harnessed recombinant technology to produce a
modified human hemoglobin, which was expressed in Escherichia coli (36). The
recombinant was designed to produce a single, di- chain and conventional
chains that combine to form a crosslinked tetramer. In addition, a mutation was
introduced to correct the leftward shift that occurs when hemoglobin is removed
from the influence of 2,3-DPG.
Some of the advantages and disadvantages of the recombinant hemoglobins
are summarized in Table 7. While the opportunity for engineering recombinant
hemoglobin and the potential of tailoring it for different applications are exciting,
Somatogen has faced unprecedented challenges in scaling up the production and
purification processes. They have, however, reported success in operating a
50,000 L production-level fermenter.
A series of phase I and II trials were completed with this recombinant
hemoglobin preparation, Optro, which was generally well tolerated. Mild gastro-
intestinal symptoms were reported with this hemoglobin preparation as with
others (37).
One of the more novel properties of this product is an erythropoietic effect,
which is similar to that observed with Hemopure. In studies of human and
murine bone marrow, Optro stimulated the proliferation of burst-forming units
erythroid (BFU-E), and was able to overcome zidovudine suppression (38). Since
erythropoietin stimulates a later stage of RBC development, the colony-forming
uniterythroid (CFU-E), their effects should potentiate one another.
Early trials of Optro in end-stage renal disease, surgical blood loss, and
ANH were conducted, and a larger-scale trial of the use of this product as an RBC
substitute for intraoperative blood loss was mounted. All clinical trials with this
product were halted when the parent company, Baxter, terminated their trials with
HemAssist.
IV. LIPOSOME-ENCAPSULATED HEMOGLOBIN

The third approach that has been taken to develop RBC substitutes has been the
use of liposomes composed chiefly of egg, vegetable, or synthetic phospha-
tidylcholine, to which other molecules, such as cholesterol and gangliosides, may
be added (39). Liposomes may also be formed from other bipolar lipids, which
have the virtue of being less expensive, but development of these nonphospho-
lipid liposomes is not as advanced. Liposomes may have a variety of structures,
Table 7 Potential Advantages and Disadvantages of Recombinant Hemoglobins

Compared to Modified Hemoglobins
1. Identity to native human hemoglobin 1. Production scale-up

2. Ability to engineer 2. Requirement for high level of purification
3. Production expense
but for this application, a spherical structure delimited by a bilamellar lipid mem-
brane, as shown in Figure 2, serves as a Trojan horse for transporting exogenously
supplied hemoglobin through the circulation. Liposome-encapsulated hemoglo-
bin (LEH) offers a number of potential advantages over perfluorocarbon- and
hemoglobin-based RBC substitutes, which are listed in Table 8. Many of the
problems created by the presence of cell-free hemoglobin in the plasma can be
averted, among them the need for crosslinking or other means of protecting
hemoglobin from clearance. The native oxyhemoglobin dissociation curve could
easily be preserved by encapsulating 2,3-DPG along with hemoglobin in the
liposome, and methemoglobin reductase could be added to limit autoxidation.
Encapsulated hemoglobin would presumably have minimal vasoactive effects.
The primary disadvantage of this approach lies in the complexity of
the technology for making liposomes and the difficulty of producing uniform
preparations, which might also have the effect of driving up the cost of produc-
tion. In addition, the delivery of large volumes of LEH raises the question of the
safety and particularly the impact of loading the RES with exogenous phos-
pholipids. In animal systems, several transient effects have been noted following
Hemoglobin
Phospholipids
Figure 2 Liposome-encapsulated hemoglobin (LEH).
Table 8 Potential Advantages and Disadvantages of

Liposome-Encapsulated Hemoglobin
1. Hemoglobin modification not required 1. Complex technology

2. Native oxygen affinity (2,3-DPG) 2. Cost (?)
3. Oxidation protection (methemoglobin reductase)
4. No vasoactivity
administration of LEH, including liver function test abnormalities, thrombo-

cytopenia, leukocytosis, and complement activation (40). Nonetheless, a consid-
erable body of experimental data from animal studies attests to the interest in
liposomes as a means of delivering not only oxygen, but a variety of other
therapeutic agents as well.
V. IMPACT OF RBC SUBSTITUTES ON TRANSFUSION

SERVICES AND BLOOD BANKS
The impact of the availability of RBC substitutes for clinical use will depend on
several factors, the foremost being which substitutes are licensed and for which
applications. Many of the RBC substitutes and oxygen carriers are being devel-
oped for applications where RBC have not had a role (e.g., liquid breathing) and
will clearly have little effect on conventional transfusion practice. The availability
of an RBC substitute licensed for use in the acute transfusion setting obviously
does have the potential of affecting RBC utilization. At least initially, use of these
products may be limited to patients with religious objections to conventional
blood transfusion, military applications, and trauma, particularly in the field, and
would have only a small impact on most transfusion services. However, the
potential for use in the emergency ward and the operating room is extensive and
could substantially decrease the volumes of RBC transfused. If RBC substitutes
with evenly modestly prolonged half-lives could be developed, they would make
even further inroads into the transfusion of RBCs, particularly in the perioperative
setting. While the transfusion service could manage the inventory of RBC
substitutes, it is also quite possible that the permissive storage conditions and
universal compatibility may suit them to be handled by the pharmacy. The
transfusion service would still be required to manage inventories of platelets,
fresh frozen plasma, and cryoprecipitate.
Several factors, particularly safety and cost, may mitigate this progression,
however. As the risk of infectious disease transmission by the conventional blood
supply decreases, the safety margin offered by RBC substitutes narrows and the
tolerance for their side effects diminishes. In addition, cost becomes an increas-
ingly important factor as the safety advantage shrinks. For the products based on
human hemoglobin, another possible limiting factor may be the availability of
source material. The supply of outdated RBC originating from volunteer donors
is presently only great enough to provide hemoglobin sufficient to supplant the
RBC supply by about 5%. Other sources of human hemoglobin would have to be
sought. The development of the technology to collect multiple units of RBCs at
one time may enhance collections, but probably not enough to significantly
augment the supply. One possible solution would be to permit collection of red
cells for further manufacture from donors who might not otherwise meet the usual
criteria or who might be paid. Blood donor centers, finding the demand for their
RBC units decreasing, may begin to develop a second category of source
hemoglobin donors, as well as emphasizing apheresis collections of platelets
and plasma. Obtaining source material may not present the same difficulties
for substitutes based on perfluorocarbons, animal hemoglobin, or recombinant
hemoglobin, assuming, for the latter, that adequate production scale-up can
be achieved.
RBC substitutes may also find a niche in the developing world where the
majority of transfusions are for acute anemia. In countries where bloodborne
pathogens are endemic, substitutes may represent a substantially safer alternative
to donor RBC. In some situations, a substitute may be less expensive to supply
than developing and maintaining the infrastructure required for a volunteer donor
blood system.
VI. CONCLUSION
In addition to the approaches taken to developing clinically useful RBC substi-
tutes described above, a number of other means of enhancing oxygen delivery,
listed in Table 9, are being explored. Most of these are much farther from clinical
Table 9 Other Alternatives to Transfusion

of Donor RBC in Development
In vitro RBC culture
Persistent oxygen bubblesa
Protein (hemoglobin) bubblesb
Intravenous allosteric modifiers
Hemoglobin from transgenic swinec
aFrom Ref. 41.
bFrom Ref. 42.
cFrom Ref. 43.
studies than the products already mentioned, and one promising approach,
obtaining human hemoglobin from transgenic pigs, is not being actively pursued
by its developer.
The animal and human studies of the materials originally conceived of as
RBC substitutes have led to an appreciation for the unique and sometimes
unexpected properties of these products and has suggested applications well
beyond the role of donor red cells. Some of these potential applications are listed
in Table 10.
The search for a RBC substitute is reaching a critical juncture; the final
demonstration of safety and efficacy in broad-based phase III clinical trials.
Demonstrating the efficacy of an RBC substitute is not entirely straightforward,
since there is no uniform or generally accepted method for assessing the efficacy
of a RBC transfusion (44). Replacement of allogeneic red cell transfusions and
reduction of mortality are likely to be the criteria used to judge the effectiveness
of the RBC substitute for transfusion applications.
The demonstration of safety also faces some challenges. The risks of the
transfusion-transmitted infections, which gave such impetus to the development
of RBC substitutes, have become extremely low. The aggregate risk for the
transmission of human immunodeficiency virus (HIV), hepatitis B virus (HBV),
hepatitis C virus (HCV), and human T-lymphotropic virus type I (HTLV-I) was
estimated to be 1:34,000 units in 1996 (45) and may be even lower at present.
Fatal transfusion events are even more rare (46). The RBC substitutes will have
to exhibit exemplary side effect profiles to provide this degree of safety. They
will, however, have the advantage of being less susceptible to the emergence of
new infectious agents in the blood supply. This fear of unknown but dreaded
Table 10 Applications for RBC Substitutes

Acute blood losstrauma, surgery
Acute blood lossJehovahs Witnesses, multiple red cell alloantibodies/rare blood type,
endemic infection in donor blood supply
Extender in acute normovolemic hemodilution
Septic shock
Anti-ischemicsickle cell crisis, PTCA, MI, cardiopulmonary bypass, vaso-occlusive
stroke
Ex vivo organ/tissue preservation
Neuroprotectantcardiopulmonary bypass
Sensitizer for chemo- and radiotherapy
Imaging
Partial liquid ventilationIRDS, ARDS, near-drowning, smoke inhalation, infection
Erythropoiesis
outcomes, which tends to exaggerate the apparent risks of donor blood, may also
add to the perceived relative safety of a substitute.
Despite some difficulties, it is likely that one or more of these oxygen-
carrying materials will be licensed for clinical use within the next few years,
either as an RBC substitute or for some other application. At that point, it will
have taken a century from the time that the discovery of the ABO blood groups
made red cell transfusion feasible to the time when a RBC substitute becomes a
clinical reality.
REFERENCES
1. Intaglietta M, Johnson PC, Winslow RM. Microvascular and tissue oxygen distribu-
tion. Cardiovasc Res 1996; 32:632643.
2. Karmaker N, Dhar P. Effect of steady shear stress on fluid filtration through the rabbit
arterial wall in the presence of macromolecules. Clin Exp Pharmacol Physiol 1996;
23:299304.
3. Gould SA, Rosen AL, Sehgal LR, Sehgal HL, Langdale LA, Krause LM, Rice CL,
Chamberlin WH, Moss GS. Fluosol-DA as a red cell substitute in acute anemia. N
Engl J Med 1986; 314:16531656.
4. Kerins DM. Role of perfluorocarbon Fluosol-DA in coronary angioplasty. Am J Med
Sci 1994; 307:218221.
5. Andre M, Nelson T, Mattrey R. Physical and acoustical properties of perfluorooctyl-
bromide, an ultrasound contrast agent. Invest Radiol 1990 25:983987.
6. Braun RD, Linsenmeier RA, Goldstick TK. New perfluorocarbon emulsion improves
tissue oxygenation in cat retina. J Appl Physiol 1992; 72:19601968.
7. Hogan MC, Kurdak SS, Richardson RS, Wagner PD. Partial substitution of red blood
cells with free hemoglobin solution does not improve maximal O2 uptake of working
in situ dog muscle. Adv Exp Med Biol 1994; 361:375378.
8. Keipert PE, Faithfull NS, Roth DJ, Bradley JD, Batra S, Jochelson P, Flaim KE.
Supporting tissue oxygenation during acute surgical bleeding using a perfluorochemi-
cal-based oxygen carrier. Adv Exp Med Biol 1996; 388:603609.
9. Flaim SF, Hazard DR, Hogan J, Peters RM. Characterization and mechanism of
side-effects of Imagent BP (highly concentrated fluorocarbon emulsion) in swine.
Invest Radiol 1991; 26(suppl):S122S128.
10. Suit H. Tumor oxygenation and radiosensitivity. In: Winslow RM, Vandegriff KD,
Intaglietta M, eds. Blood Substitutes: Physiological Basis of Efficacy. Boston:
Birkhuser, 1995:187199.
11. Rockwell S, Kelley M, Irvin CG, Hughes CS, Yabuki H, Porter E, Fischer JJ.
Preclinical evaluation of Oxygent (TM) as an adjunct to radiotherapy. Biomater Artif
Cells Immobil Biotechnol 1992; 20:883893.
12. Leach CL, Greenspan JS, Rubenstein DS, Shaffer TH, Wolfson MR, Jackson JC,
DeLemos R, Fuhrman BP. Partial liquid ventilation with perflubron in premature
infants with severe respiratory distress syndrome. N Engl J Med 1996; 335:761767.
13. Blauth CI, Arnold JV, Schulenberg WE, McCartney AC, Taylor KM. Cerebral
microembolism during cardiopulmonary bypass. Retinal microvascular studies in

vivo with fluorescein angiography. J Thorac Cardiovasc Surg 1988; 95:668676.
14. Blauth CI, Smith P, Newman S, Arnold J, Siddens F, Harrison MJ, Treasure T, Klinger
L, Taylor KM. Retinal microembolism and neuropsychological deficit following
clinical cardiopulmonary bypass: comparison of a membrane and a bubble oxygen-
ator. A preliminary communication. Eur J Cardiothorac Surg 1989; 3:135139.
15. Spiess BD, Braverman B, Woronowicz AW, Ivankovich AD. Protection from cerebral
air emboli with perfluorocarbons in rabbits. Stroke 1986; 17:11461149.
16. Alayash AI, Cashon RE. Reactions of nitric oxide and hydrogen peroxide with
hemoglobin-based blood substitutes. Ann NY Acad Sci 1994; 738:378381.
17. Rioux F, Petitclerc E, Audet R, Drapeau G, Fielding RM, Marceau F. Recombinant
human hemoglobin inhibits both constitutive and cytokine-induced nitric oxide-
mediated relaxation of rabbit isolated aortic rings. J Cardiovasc Pharmacol 1994;
24:229237.
18. Rioux F, Drapeau G, Marceau F. Recombinant human hemoglobin (rHb1.1) selec-
tively inhibits vasorelaxation elicited by nitric oxide donors in rabbit isolated aortic
rings. J Cardiovasc Pharmacol 1995; 25:587594.
19. Gulati A, Sharma AC, Burhop KE. Effect of stroma-free hemoglobin and diaspirin
cross-linked hemoglobin on the regional circulation and systemic hemodynamics.
Life Sci 1994; 55;827837.
20. Ulatowski JA, Nishikawa T, Matheson-Urbaitis B, Bucci E, Traystman RJ, Koehler
RC. Regional blood flow alterations after bovine fumaryl -crosslinked hemoglobin
transfusion and nitric oxide synthase inhibition. Crit Care Med 1996; 24:558565.
21. Garcia-Sepulcre ME, Carnicer F, Mauri M, Prieto A, Perez-Mateo M. Increased
plasma endothelin in liver cirrhosis and response to plasma volume expansion (letter).
Am J Gastroenterol 1996; 91:24522453.
22. Vandegriff KD. Stability and toxicity of hemoglobin solutions. In: Winslow RM,
Vandegriff KD, Intaglietta M, eds. Blood Substitutes: Physiological Basis of Efficacy.
Boston: Birkhuser, 1995:105131.
23. Griffiths E, Cortes A, Gilbert N, Stevenson P, MacDonald S, Pepper D. Hemoglobin-
based blood substitutes and sepsis. Lancet 1995; 345:158160.
24. Kaca W, Roth RI, Levin J. Hemoglobin, a newly recognized lipopolysaccharide
(LPS)-binding protein that enhances LPS biological activity. J Biol Chem 1994;
269(40):2507825084.
25. Su D, Roth RI, Yoshida M, Levin J. Hemoglobin increases mortality from bacterial
endotoxin. Infect Immun 1997; 65:12581266.
26. Langermans JA, van Vuren-van der Hulst M, Bleeker WK. Safety evaluation of a
polymerized hemoglobin solution in a murine infection model. J Lab Clin Med 1996;
127:428434.
27. Gould SA, Sehgal LR, Sehgal HL, Moss GS. The development of hemoglobin
solutions as red cell substitutes: hemoglobin solutions. Transfus Sci 1995; 16:517.
28. Gould SA, Moore EE, Hoyt DB, Burch JM, Haenel JB, Garcia J, DeWoskin R, Moss
GS. The first randomized trial of human polymerized hemoglobin as a blood
substitute in acute trauma and emergent surgery. J Am Coll Surg 1998; 187:113120.
29. Gulati A, Sharma AC, Singh G. Role of endothelin in the cardiovascular effects
of diaspirin crosslinked and stroma reduced hemoglobin. Crit Care Med 1996;
24:137147.
30. Schultz SC, Powell CC, Burris DG, Nguyen H, Jaffin J, Malcolm DS. The efficacy
of diaspirin crosslinked hemoglobin solution resuscitation in a model of uncontrolled
haemorrhage. J Trauma 1994; 37:408412.
30a. Sloan EP, Koenigsberg M, Gens D, Cipolle M, Runge J, Mallory MNRG Jr. Diaspirin
cross-linked hemoglobin (DCLHb) in the treatment of severe traumatic hemorrhagic
shock. A randomized controlled efficacy trial. JAMA 1999; 282:18571864.
31. Conover CD, Malatesta P, Lejeune L, Chang CL, Shorr RG. The effects of hemodilu-
tion with polyethylene glycol bovine hemoglobin (PEG-Hb) in a conscious porcine
model. J Investig Med 1996; 44:238246.
32. Nho K, Linberg R, Johnson M, Gilbert C, Shorr R. PEG-hemoglobin: an efficient
oxygen delivery system in the rat exchange transfusion and hypovolemic shock
models. Art Cells, Blood Subs Immob Biotech 1996; 22:795803.
33. Hughes GS, Francom SF, Antal EJ, Adams WJ, Locker PK, Yancey EP, Jacobs EE
Jr. Effects of a novel hemoglobin-based oxygen carrier on percent oxygen saturation
as determined with arterial blood gas analysis and pulse oximetry. Ann Emerg Med
1996; 27:164169.
34. Hughes GS, Yancey EP, Albrecht R, Locker PK, Francom SF, Orringer EP, Antal EJ,
Jacobs EE Jr. Hemoglobin-based oxygen carrier preserves submaximal exercise
capacity in humans. Clin Pharmacol Ther 1995; 58:434443.
35. Hughes GS, Francom SF, Antal EJ, Adams WJ, Locker PK, Yancey EP, Jacobs EE
Jr. Hematologic effects of a novel hemoglobin-based oxygen carrier in normal male
and female subjects. J Lab Clin Med 1995; 126:444451.
36. Looker D, Abbott-Brown D, Cozart P, Durfee S, Hoffman S, Mathews A, Miller-
Roehrich J, Shoemaker S, Trimble S, Fermi G, Komiyama N, Nagai K, Stetler GL.
A human recombinant hemoglobin designed for use as a blood substitute. Nature
1992; 356:258260.
37. Murray JA, Ledlow A, Launspach J, Evans D, Loveday M, Conklin JL. The effects
of recombinant human hemoglobin on esophageal motor functions in humans.
Gastroenterology 1995; 109:12411248.
38. Schick MR, Rosenthal GJ. A comparative evaluation of hemin and recombinant
hemoglobin in the stimulation of erythropoietic precursors (abstr). Blood 1996;
88(suppl 1):104b.
39. Rudolph AS. Encapsulation of hemoglobin in liposomes. In: Winslow RM,
40. Szebeni J, Wassef NM, Rudolph AS, Alving CR. Complement activation in human
serum by liposome-encapsulated hemoglobin: the role of natural anti-phospholipid
antibodies. Biochem Biophys Acta 1996; 1285:127130.
41. Burkard ME, Van Liew HD. Oxygen transport to tissue by persistent bubbles: theory
and simulations. J Appl Physiol 1994; 77(6):28742878.
42. Sakai H, Hamada K, Takeoka S, Nishide H, Tsuchida E. Physical properties of
hemoglobin vesicles as red cell substitutes. Biotechnol Prog 1996; 12:119125.
43. ODonnell JK, Martin MJ, Logan JS, Kumar R. Production of human hemoglobin in
transgenic swine: an approach to a blood substitute. Cancer Detect Prev 1993;
17:307312.
44. Fratantoni J. Demonstration of efficacy of a therapeutic agent. In: Winslow RM,
45. Schreiber GB, Busch MP, Kleinman SH, Korelitz JJ, for the Retrovirus Epidemiology
Donor Study. The risk of transfusion-transmitted viral infections. NEJM 1996;
334:16851690.
46. Linden JV, Tourdalt MA, Scribner CL. Decrease in frequency of transfusion fatalities
(letter). Transfusion 1997; 37:243244.
25
Professional Standards and
Voluntary Accreditation
Jay E. Menitove
Community Blood Center of Greater Kansas City, Kansas City, Missouri
Hillary V. Schaeffler
American Association of Blood Banks, Bethesda, Maryland
The American Association of Blood Banks (AABB) is the professional society

for almost 8500 individuals involved in blood banking and transfusion medicine.
It also represents more than 2000 institutional members, including community
and Red Cross blood collection centers, hospital-based blood banks, and transfu-
sion services that collect, process, distribute, and transfuse blood and blood
components. Members are responsible for virtually all of the blood collected and
more than 80% of the blood transfused in the United States. Throughout its
50-year history, the AABBs highest priority has been to maintain and enhance
the safety of the nations blood supply. To ensure the safety and adequacy of the
nations blood supply, the AABB sets voluntary standards and accredits institu-
tions that implement the standards.
To ensure harmonization and consistency in all of AABBs standards and
accreditation activities, in 1998 the Board of Directors created an umbrella
committee, the Standards Program Committee. This committee, whose primary
role is to oversee the development and/or revision of all AABB standards,
comprises a chairperson, chairs of two subcommittees (Quality Management
Subcommittee and Communications Subcommittee), and chairs of five specialty
program units (Blood Banks/Transfusion Services, Hematopoietic Progenitor Cell
Services, Immunohematology Reference Laboratories, Parentage Testing Labora-
tories, and Perioperative Collection and Transfusion Services). The Quality
Management Subcommittee develops/revises quality standards and ensures that
569
570 Menitove and Schaeffler
all Standards quality concepts are consistent, and the Communications Subcom-
mittee disseminates guidance related to the standards. The program units are
responsible for creating or revising the technical/scientific standards, which are
specific for each standard-setting activity (Fig. 1).
In 1997, an Accreditation Program Committee was created to oversee all
accreditation activities. This committee comprises a chairperson, chairs of two
subcommittees (Quality Systems Subcommittee and Education Advisory Sub-
committee), and chairs of seven specialty program units (Donor Centers, Trans-
fusion Services, Hematopoietic Progenitor Cell Services, Immunohematology
Reference Laboratories, Parentage Testing Laboratories, Perioperative Collection
and Transfusion, and SBB Schools). The Quality Systems Subcommittee works
with the Accreditation Program Committee to fully implement quality principles
into AABBs Accreditation Program and to provide quality assessment tools for
AABBs institutional members. The Education Advisory Subcommittee serves as
the educational advisory group and coordinator for the Accreditation Programs
educational and training activities. Each program unit oversees and coordinates
the specific program unit activities, including coordinating and reviewing the
program assessment tools and assessor guidance (Fig. 2).
I. STANDARDS PROGRAM
The AABB entered the voluntary standard-setting arena in 1958 when it pub-
lished the first edition of Standards for a Blood Transfusion Service, the purpose
of which was to improve the quality and safety of human blood transfusions.
In 1990, the first edition of the Standards for Parentage Testing Laboratories was
published. In 1991, the first hematopoietic progenitor cell standards were added
to the Standards for Blood Banks and Transfusion Services, and 5 years later the
first edition of the Standards for Hematopoietic Progenitor Cells was published.
In 1999 the first edition of the Standards for Immunohematology Reference
Laboratories was published, and currently the first edition of Standards for
Perioperative Collection and Transfusion is being developed (Table 1).
Voluntary standards provide a benchmark for judging performance in a
specific field. For many years professionals in the legal, accounting, and engi-
neering fields have set minimum voluntary standards to determine a baseline for
compliance. Although higher standards of practice may be targeted, defining the
minimum provides a basis for measurement. When voluntary standards are
implemented, the public and current and potential customers may judge an
organizations compliance. Further, as international trade and communication
increase, voluntary standards form the basis for an objective method of distin-
guishing between those who implement the standards and those who do not.
The first edition of the blood bank Standards focused on technical specifi-
cations and followed a path of workflow approach for the blood bank and
Professional Standards and Voluntary Accreditation 571
Figure 1 Standards Program Committee.

572
Figure 2 Accreditation Program Committee.

Menitove and Schaeffler
Table 1 AABB Standards, Editions, and Their Effective Dates
Effective Revision
Standards Edition date cycle
Standards for Blood Banks and Transfusion Services 19th June 1999 18 months
Standards for Hematopoietic Progenitor Cell 2nd March 2000 18 months
Services
Standards for Immunohematology Reference 1st July 1999 2 years
Laboratories
Standards for Parentage Testing Laboratories 4th July 1999 2 years
Standards for Perioperative Collection and 1st July 2001 2 years
Transfusion
transfusion service, i.e., beginning with donor selection and ending with blood
administration. Over the last 40 years, 19 editions of blood bank/transfusion
service Standards have been published, each edition building on the technical
specifications from the prior edition.
Over the last decade, the AABB leadership and Standards committees have
recognized the importance of quality systems to ensure consistency in the
provision of appropriate components and services. In 1991, the requirement for a
program of quality assurance was incorporated into the 14th edition of Stand-
ards for Blood Banks and Transfusion Services. In 1996, a written quality plan
was required. The next year, in response to questions about what items should be
included in a quality plan, an association policy (Association Bulletin #97-4)
identified 10 Quality System Essentials (QSEs) as minimum requirements of a
quality system. The premise of the QSEs is that a quality management system
ensures that operations are controlled to produce consistent products and services.
To achieve and maintain this control, it is required that policies, processes, and
procedures are documented and implemented appropriately, and, when appropri-
ate, records of the activities must be created and maintained. Internal assessments
are conducted at regular, scheduled intervals, and if problems are identified,
corrective action plans must be developed.
In 1999, the 10 QSEs were fully incorporated into the 19th edition of
Standards (Table 2). In a further evolution, the framework of the 20th edition
of the Standards for Blood Banks and Transfusion Services, published in the fall
of 2000, is the 10 QSEs. The Standards publication is now organized into 10
sections, each section corresponding with a QSE. The first chapter is Manage-
ment, the second Resources, the third Equipment, etc. The technical requirements
appear in the appropriate chapter, creating a matrix of technical and quality
requirements. For example, instead of repeating that equipment must be main-
tained throughout the standards, the requirement is stated once in Section 3,
Table 2 Quality System Essentials

Organization
Resources
Equipment
Supplier and customer issues
Process control
Documents and records
Incidents, errors, accidents; nonconforming products and services; complications
Assessments: internal and external
Process improvement through corrective and preventive action
Facilities and safety
Equipment. Then the specific requirements for equipment (e.g., blood warmers,
alarms) are placed in that section. With this model, the technical standards are as
important as the quality requirements, but this system incorporates them into a
quality management system.
II. DEVELOPING AABB STANDARDS

Since the first issue of standards was developed, there have been extensive
changes to the standard setting process. These changes include modifying the
scope of expertise of standards program unit members, modifying the process for
reviewing and revising existing standards, standardizing the interval between
editions, providing the opportunity for member and public comment, and estab-
lishing criteria for new standards.
The AABB Board of Directors appoints the chair and members of the
program unit. Criteria for selection and appointment of the chair include a broad
knowledge of the program unit topic, a scholarly publication record, and appro-
priate interpersonal skills. Program unit selection reflects particular expertise in
the relevant field and experience in developing applicable policies. Program unit
members may serve for two editions of standards. Members are appointed so that
half of the members of the latest edition remain to provide a link from one
program unit to the next. One or more public members may also be appointed to
a program unit. This person represents the interests of the general public.
Liaisons from other organizations, including federal agencies, are also
included in the standard setting process. Liaison invitations are based on the
standard setting activity. Liaisons participate fully in committee discussions and
work assignments, but they do not vote.
Each new standards program unit reviews the prior edition and with
consensus deletes, adds, clarifies, or proposes new standards. In order to provide
confidence in the standards setting process, the AABB Standards Committee base
their decisions on scientific evidence, when available, good medical practice,
technological changes, public policy issues, and government regulations and
guidelines. When there are differences of opinion, committee decisions are made
by a majority vote of the members present.
Once the revision phase of the process is completed, the Standards are
presented to the Standards Program Committee and the Board of Directors for
approval to publish as proposed Standards. Once approved, the proposed changes
are published for comment on the public section of the AABB web site for 60
days. Notice of the proposed Standards is printed in a nationally circulated
newspaper and AABB newsletters. Following the comment period, the program
unit reviews each comment, and final wording is determined. The final draft
undergoes a thorough legal, regulatory, and technical review and is submitted to
the Standards Program Committee and Board for final approval. Once the final
Standards are issued, institutional members have 60 days to implement them.
If it is determined that a new standard must be set or a current standard must
be revised before the next edition is published, a program unit may add or revise
a standard if there is scientific evidence for doing so. In this instance, a proposed
interim standard is posted for an abbreviated comment period. After consideration
of the comments and with Standards Program Committee and Board approval, an
interim standard is set and must be implemented by accredited members. In rare
circumstances, when an accelerated review and approval process is required, an
emergent standard is set with Board approval but without a comment period.
If an institution has data to prove that it can meet the intent of a standard
through another mechanism, it may apply to the relevant program unit for a
variance or an exemption. In this case, the program unit considers the request,
and if the variance is approved the Communications Subcommittee communi-
cates the decision. If an institution disagrees with a variance decision, it may
appeal the decision to a Standards Review Committee. Variances apply for one
edition of Standards, although institutions may reapply for variances.
III. GUIDANCE DOCUMENTS

Separate guidance documents have been created for each set of standards.
Because the standards contain requirements, a need was identified for the
explanation of a rationale or intent of specific standards, examples of implementa-
tion, and approved variances to standards. Accordingly, the guidance for Stand-
ards for Blood Banks and Transfusion Services and Standards for Hematopoietic
Progenitor Cell Services is contained in Standards Source, a subscription that
consists of six quarterly issues. Separate guidance documents are published for
Standards for Parentage Testing Laboratories and Standards for Immunohemato-
logy Reference Laboratories. It is anticipated that a guidance document will also

be developed for the Standards for Perioperative Collection and Transfusion.
IV. REGIONAL AND COUNTRY-SPECIFIC STANDARDS

Since 1997, upon request the AABB has been working with regional organiza-
tions outside of the United States to develop regional or country-specific volun-
tary blood banking standards that incorporate specific technical requirements that
are appropriate for the region. Regional requirements may differ as a result of
federal regulations, standard of care, resources, or regional infectious disease
markers. To be consistent with the ISO 9000 standards, which is an internationally
recognized set of standards, the AABB expanded the 10 QSEs to 20 quality
essentials. The result is a set of standards that is compatible with QSEs and ISO
9000 concepts and merges quality standards with technical requirements.
V. ACCREDITATION PROGRAM
Following publication of the first edition of standards in 1958, AABBs Inspec-
tion and Accreditation program was developed to identify whether inspected
organizations met the requirements of the standards (a detection-oriented model).
Inspectors used a checklist during on-site inspections to determine whether the
specific requirements in the standards had been met. Following a review of the
inspectors completed checklist, summary reports of the inspections were written
by area chairs.
In 1997 a new process was developed to incorporate a systems approach
(a prevention-oriented model). Checklists transitioned to assessment tools, and
the terminology of inspectors transitioned to assessors. With this new model,
summary reports are left on-site at the conclusion of the assessment and assessors
identify nonconformances that may be an isolated incident or a systems-related
problem. Assessors evaluate both the quality and operational activities performed
within an institution. The quality system evaluation is based on the same criteria
for every facility; the operational systems are identified by the activities per-
formed by the institution. The result is that each assessment is customized for
the facility.
At the core of the accreditation program are hundreds of volunteer asses-
sors. Assessors must have appropriate expertise, participate in assessor continuing
education, and perform a minimum number of assessments each year. Assessors
and their organizations benefit from free training in quality concepts and auditing
techniques, networking with other individuals in their profession, and learning
about best practices used in other facilities. In 1999 the AABB also hired three
lead assessors who have expertise in the field and are well trained in quality and
quality assessments.
As the Food and Drug Administration, Health Care Financing Administra-
tion, and Joint Commission on Accreditation of Healthcare Organizations focus
on the importance of quality assurance and systems evaluation, the AABB is
confident that its standards and accreditation program is preparing AABB-
accredited institutions for the changing environment. Institutions that are accred-
ited by the AABB send a signal to their customers and the public that they
implement voluntary industry standards, which require control of operations
through a quality management system and implementation of the blood banking
communitys agreed-upon technical standards. After more than 40 years in the
standard-setting arena, the AABBs standards and accreditation program has
played a significant role in improving patient safety, and institutions with AABB
accreditation are respected around the world.
26
The Role of Federal Regulation in
Blood Safety
Jay S. Epstein and Mary Gustafson
U.S. Food and Drug Administration, Rockville, Maryland
I. STATUTORY HISTORY
The modern blood industry dates back more than 50 years. Establishments known
as blood banks began to appear in the 1930s, and widespread use of blood and
its derivatives began during World War II. Federal regulatory control was exerted
over blood and blood products almost from their inception based on the preexist-
ing Biologics Control Act of 1902 (also known as the Virus-Toxin Act). This law
requiring licensing of biologics was consolidated with other public health laws in
1944 to become the Public Health Service (PHS) Act. The first blood product,
Normal Human Plasma, was licensed for medical use in 1940. The first blood
bank was licensed for the manufacture of whole blood in 1946. The first
blood grouping reagent for serological testing of red blood cells was licensed in
1949. The PHS Act was implemented by the National Institutes of Health (NIH)
during the time the first blood licenses were issued. Blood was also considered
to be a drug subject to the Federal Food, Drug and Cosmetic (FDC) Act; however,
the full scope of regulatory controls available under the FDC Act was not
implemented until regulatory control for biologics was transferred from the NIH
to the U.S. Food and Drug Administration (FDA) in 1972 (1). Today, blood, blood
components, blood-derived and analogous products, and in vitro diagnostic blood
screening tests and other medical devices used in the manufacture of blood and
blood components are regulated under federal statutes implemented by FDA.
The primary statutes covering blood regulation are the PHS Act (42 USC
202 et. seq.) and the FDC Act (21 USC 302 et. seq.). Both laws have predecessors
that date to the early 1900s. Both were enacted following tragic events that
579
580 Epstein and Gustafson
motivated the U.S. Congress to insist that medicines and other medical use
products be controlled. The impetus for the enactment of the first biologics statute
was the death of several children in the fall of 1901. Those children had been
given a diphtheria antitoxin contaminated with tetanus toxin. The source of the
antitoxin was a horse that later developed tetanus. The event was deemed
preventable had controls been in place throughout the procurement and process-
ing of the antitoxin. Congress therefore enacted the Biologics Control Act, which
required the licensing of manufacturing establishments as well as the biological
products they manufactured. Thus, the Act, which later was incorporated into
Section 351 of the PHS Act, serves as the legal basis of licensing biologics (2).
The original PHS Act did not specifically address blood and blood derivative
products, but these products were regulated as being analogous to a therapeutic
serum, a category specifically included in the language of the statute. This
interpretation was challenged in 1968 in Blank v. United States (3). Mrs. Maxine
Blank had been convicted of misdemeanor charges under the PHS Act for
violations incurred in operating a commercial blood bank located in Dallas,
Texas. She was found guilty but appealed her conviction. The grounds for appeal
were based upon the absence of any reference to blood or blood products in the
statute. The court agreed with Mrs. Blank and reversed the conviction. However,
it was not the intention of Congress to omit the regulation of blood and
blood-derived products from the authority of the PHS Act, and in 1970 Con-
gress amended the Act specifically to include, blood, blood components and
derivatives (4).
The FDC Act was preceded by the Pure Food and Drug Act of 1906. This
early law was passed following disclosure of unsanitary and uncontrolled prac-
tices in the meat-packing industry in the popular book The Asphalt Jungle by
Upton Sinclair. The early law addressed the purity of foods and drugs and was
silent with respect to their safety and efficacy. The drug act was strengthened in
1938, however, after more than 100 people died after ingesting elixir of sulfanil-
amide in which the drug was dissolved in diethylene glycol, an extremely toxic
substance, rather than alcohol or water. The recodified 1938 law was the forerun-
ner of the modern FDC Act and included the requirement for safety as well as
purity (5). The law was further amended in 1962, after the thalidomide tragedy,
to include efficacy as well as safety, and again in 1976, to encompass the
regulation of medical devices. Until the Food and Drug Administration Modern-
ization Act of 1997, the only other major change to the FDC Act was strengthen-
ing of device regulation and clarification of the regulation of combination drug
and device products by the Safe Medical Device Act of 1990 (6,7).
Key provisions of the PHS Act and the FDC Act are outlined in Table 1.
Besides differing in structure, the two statutes differ in objective, scope, focus,
and enforcement mechanism. The PHS Act was enacted to promote public health
and disease prevention by assuring safe and effective vaccines and other biolog-
Federal Regulation and Blood Safety 581
Table 1 Key Provisions of Statutes Enforced by FDA
Section Provision
PHS Act
351(a) Requires licenses for biological products
351(b) Prohibits false labeling
351(c) Authorizes inspections
351(a) Authorizes suspension/revocation of licenses
361 Provides authority to control spread of communicable disease
FDC Act
301, et seq Outlines prohibited acts and penalties
501 Prohibits adulteration
502 Prohibits misbranding
510 Requires registration of producers of drugs and devices
704 Authorizes inspections
ical products intended to treat and prevent diseases of humans. The FDC Act was
intended to guard the consumer against adulterated and misbranded foods, drugs,
devices, and cosmetics. The PHS Act provides for administrative actions of
license suspension and revocation as the primary remedies for violating the
statute. In contrast, the FDC Act is far more enforcement-oriented in that it
outlines prohibited acts and describes remedies that are judicial, e.g., product
seizure, prosecution, and injunction. However, since biological products under the
PHS Act concurrently are drugs or devices under the FDC Act, the enforcement
provisions of the FDC Act can be applied.
II. REGULATION AND GUIDANCE

The past decade and a half have brought sweeping changes to practices within
the blood industry and to the regulation of that industry. The AIDS epidemic in
particular brought a new focus of public concern to blood and blood-derived
products. The industry and the regulators of the industry have been faulted for
their response to the AIDS crisis (8). Partly in response, activities once thought
of as medical services are now considered as pharmaceutical manufacturing by
FDA. This change in outlook itself has caused major reorganization of the blood
industry.
The blood industry encompasses a broad range of types of product. The
industry includes blood donations from individuals processed by physical means
only and transfused as single donor blood components or small pools (usually
fewer than 10 units); blood-derived products consisting of large pools of dona-
tions (thousands) that are extensively manufactured by physical and chemical

means to result in specialized products with consistent dosage and intended uses
distinct from simple components; blood-related products manufactured by bio-
technology to be analogous to a naturally derived blood product; medical devices
to test the safety and stability of blood products; medical devices used in the
processing of blood components; and drugs used to anticoagulate, provide
nutrients during storage, and extend the shelf life of blood components. Each area
has different regulatory issues and involves different regulatory approaches. All
facets of the blood industry are within the regulatory purview of FDA, except that
FDA does not set or enforce medical practice standards.
Agencies tasked with implementing and enforcing laws are given the power
to promulgate regulations that more fully interpret the laws from a practical
standpoint. These regulations have the force of law. Promulgation of regulations
follows a strict procedural path that usually includes public notification of a
proposed rule and a period of comment before the rule is issued as final.
Regulations affecting blood, blood derivatives, and medical devices used in blood
banking are found in Title 21, Code of Federal Regulations (21 CFR), Parts 200
(drugs); 300 (investigational drugs); 600 (biologics, including blood); and 800
(medical devices). Table 2 lists provisions of each Part of 21 CFR referenced.
Biological products, including blood, are subject to Part 600 and, depending
on the intended use of the biological product, are subject to additional pro-
visions included in Parts 200 and 300 for biological drugs and Part 800 for
biological devices.
In addition to regulations, which have the force of law, agencies also
generate guidance documents. These documents, which include guidelines, points
to consider, memoranda, and reviewers guidances, represent the agencys policy
on a particular topic. They are intended to supplement or interpret existing
regulations. Guidance documents do not have the force of law, as do regulations.
They are not binding on the industry or the agency and are not enforceable, as are
regulations. Generally, however, the guidance documents provide the agencys
view of how to fulfill the requirements of a regulation. There may be other
equivalent ways to fulfill a regulatory requirement. But if a firm adopts the
recommendation addressed in a guidance document and makes it part of its
standard operating procedures (SOPs), the firm is obliged as a matter of current
good manufacturing practices (cGMP) to follow its own procedures.
Guidances, in the form of memoranda addressed to registered blood estab-
lishments, have been issued frequently since the mid-1980s. Many of the guid-
ances have been issued upon approval of new test kits to screen the blood supply.
The guidances have advised the industry how to use the tests in determining the
suitability of donor units and have further clarified such issues as blood compo-
nent labeling, blood quarantine and disposition, donor deferral, procedures for
reinstating deferred donors based on supplemental testing, and lookback
Table 2 Relevant Provisions of Title 21 Code of Federal Regulations
Cite Addresses
21 CFR, Part 200Drugs

201 Drug labeling
202 Prescription drug advertising
207 Drug registration
210, 211 Current good manufacturing practice
21 CFR, Part 300Drugs
312 Investigational new drug applications
21 CFR, Part 600Biologics
601 Licensing
606 Current good manufacturing practice for blood and components
606.120-122 Blood and component labeling
607 Blood registration
610 General product standards, including labeling and testing
640680 Additional standards
21 CFR, Part 800Medical Devices
801 Device labeling
803 Medical device reporting
807 Device registration
809 In vitro diagnostics, including labeling
810 Device recall
812 Investigational device exemptions
814 Premarket approval
820 Quality system regulation
860 Medical device classification
(i.e., product retrieval or recipient notification related to previous donations).

Other guidances have discussed manufacture and use of autologous blood
components; standards for donor suitability; licensing criteria for specific prod-
uct processing steps, e.g., irradiation and leukoreduction; and error/accident
reporting. A particular subject may be addressed in more than one guidance
document. Blood memoranda and other guidance documents are published
and available in hard copy from the Division of Communication and Public
Affairs (HFM-43), Center for Biologics Evaluation and Research, Food and
Drug Administration, 1401 Rockville Pike, Rockville, MD 20852-1448. These
guidance documents are also available in a variety of electronic forms, including
on the Internet (www.fda.gov.cber.guidelines.htm), as noted in the listing in
Appendix A.
Although not directly enforceable, guidance documents, once promulgated,

carry weight in court. Once a guidance is issued, industry, particularly the blood
industry, believes it is compelled to follow the guidance whether or not it is
thought to be in its best interest. Because of concerns about the way in which
guidance documents are established and issued, the agency has prepared an
internal procedure for the promulgation of guidance documents, entitled Good
Guidance Practices (GGP) (9). This procedure stratifies guidance into two levels
depending on the nature, significance, and controversiality of the guidance. The
GGP requires the agency to publish for public comment any guidance falling into
the level one category (significant interpretation of statute or regulation), and
unless the guidance addresses an urgent public health issue the guidance will not
be effective until it is reissued for implementation after a comment period. Level
2 guidances are deemed to be less significant and less controversial and are
effective upon publication, although they remain open for comment. It is import-
ant to note that the public may comment on any guidance at any time during
its existence.
III. LICENSING AND PRODUCT APPROVAL

Products and manufacturers of products regulated under the purview of FDA are
subject to several reporting and approval requirements. First is the requirement
for all manufacturers of drugs and medical devices to register with the agency
and list their products. This requirement is part of the FDC Act in Section 510.
Regulations covering registration and product listing are found in 21 CFR, Part
207, for drugs; Part 607 for blood and blood components; and Part 807 for
medical devices. Registration fulfills a manufacturers obligation to advise the
agency of its existence, location, and scope of manufacturing. Information
obtained during the registration process is used primarily to assure inspectional
oversight.
Products also are subject to premarket approval or clearance prior to being
introduced into interstate commerce. Prior to receiving marketing approval, a
product can be distributed for the purpose of generating and collecting clinical
information to support its intended marketing claim. Generally, distribution of the
product requires an exemption from the requirements that only approved products
be distributed in interstate commerce. Such an exemption is approved by FDA
following a sponsors filing a request for an exemption. Requests for exemption
from the requirements for licensure to permit distribution of biological drugs and
devices are filed subject to the provisions for investigational new drugs (IND) in
21 CFR, Part 312. Although legally medical devices, biological devices subject
to license are procedurally covered under the IND regulations. Other medical
devices regulated by CBER are subject to investigational device exemptions
(IDE) regulations in 21 CFR, Part 812. Just as the preapproval processes differ
depending on whether the product is a drug, device, or biological drug or device,
the approval processes differ as well. Products that are deemed to be biologics
are regulated under the licensing provisions of Section 351 of the PHS Act.
Products regulated in this manner include blood and blood components, blood-
derived and related products, and in vitro tests required or recommended for
testing the blood supply. Other medical devices used to process, test, store, or
administer blood are regulated under the medical device amendments to the FDC
Act. Depending upon the classification of the device, permission to market it is
granted by approval of a premarket approval application (PMA) or clearance of
a premarket notification (510k). Blood anticoagulants, additive and rejuvenation
solutions, and colloidal plasma volume expanders are subject to the drug-approval
mechanisms under Section 505 of the FDC Act. In accordance with intercenter
agreements between the Center for Biologics Evaluation and Research (CBER),
the Center for Drug Evaluation and Research, and the Center for Devices and
Radiological Health, CBER has jurisdiction for the regulation of blood as well as
drugs and devices used in the processing of blood regardless of the regulatory
mechanism (10,11).
The majority of blood and blood-related products are licensed as biologics
under the PHS Act. Therefore, the remainder of this chapters discussion will
focus on licensing and the related regulatory requirements. Prior to the im-
plementation of the Food and Drug Administration Modernization Act of 1997,
licensing of blood products required that the product and the establishment
responsible for the manufacture of the product each be licensed. That is, separate
licenses were issued for the biological product (product license) and the facility
manufacturing the product (establishment license). Since February 1998, a single
biologics license has been issued. On July 30, 1998, FDA published a proposed
rule to amend the biologics regulations to eliminate the filing of a separate
establishment license application and a product license application in order to
market a biological product in interstate commerce. The separate license applica-
tions are being replaced by a single biologics license application (12). The impact
of this change is discussed later in this chapter.
IV. POSTLICENSURE REQUIREMENTS

After licensure, a manufacturer of a licensed biological product is responsible for
certain postapproval requirements. These include annual updates to registration
and listing, compliance with current good manufacturing practices in the manu-
facture of the product, lot release requirements for some products, and certain
reporting requirements. The reporting requirements include reporting changes to
the approved application in accordance with 21 CFR, Part 601.12. This regulation
was revised in 1997 to reduce reporting requirements and is discussed in greater

detail later in this chapter (13).
Adverse experiences associated with use of biological products are report-
able under 21 CFR 600.81. Manufacturers of licensed biologics, with the ex-
ception of blood and blood components and in vitro diagnostics, are required
to report serious, unexpected adverse events as outlined in the regulation. Man-
ufacturers of in vitro diagnostics are subject to medical device reporting require-
ments in 21 CFR, Part 803. Both types of reports may be filed using
FDAs MedWatch reporting program. MedWatch reports are facilitated by
using Form 3500A (Appendix B, accessible from the FDA home page
www.fda.gov/medwatch/report/hcp.htm). Error and accident (biological product
deviation) reporting (21 CFR 600.14) is required for licensed manufacturers and
is recommended but is voluntary for manufacturers of unlicensed (intrastate)
blood and blood components. FDA has published a proposed rule that would
require error and accident reporting for all manufacturers of blood and blood
components whether or not the products are in interstate commerce (14). All
manufacturers of blood and blood components, whether licensed or not, are
required to report fatalities associated with the collection or transfusion of such
products (21 CFR 606.170). Fatal events should be reported by telephone
(301-827-6220) within 24 hours with a full report submitted within 7 days. An
electronic mail account is also available to facilitate preliminary reporting. The
e-mail address is <fatalities2@cber.fda.gov>. Additional requirements for report-
ing of serious adverse reactions to blood and blood components are being
considered by FDA.
V. INSPECTION AND ENFORCEMENT

Inspection has been a focus of biologics regulation since the earliest biologics
law. Inspections of biologics firms have been accomplished by officers responsi-
ble for the review of biological products applications. In recent years, the field
investigations force has played greater role in inspections of biological products,
including blood. Routine (biennial) surveillance inspections for most blood and
blood components have been conducted by FDA field personnel since the late
1970s. As part of an initiative termed Team Biologics, lead responsibility for
periodic inspections has been transferred to the FDA field force for all biological
products, including licensed fractionated blood products and in vitro diagnostic
blood screening reagents (15). As with blood and blood components, Compliance
Program Guidance Manuals have been developed to provide guidance for uni-
form inspections and consistent application of regulatory requirements applicable
to plasma derivatives (16).
Inspectional observations are communicated to firms being inspected on
FDA Form 483. Observations that are judged by the investigator to be potentially
violative of regulations are listed on the form. Further compliance action, such as
issuance of a warning letter, follows only after the results of the inspection,
including the Form 483 observations, are reviewed by compliance officers at the
district level and sometimes at the Center level. Administrative and legal actions
such as license suspension or revocation, injunction, seizure, and prosecution are
infrequent occurrences. Voluntary compliance is the usual outcome and a goal
of FDA.
Except in situations of imminent threat to health, if an establishment is
found violating any of the laws that FDA enforces, it usually is given a chance to
correct the problem voluntarily before FDA pursues an enforcement action. When
an establishment cannot or will not correct a violative situation, FDA can invoke
administrative or legal sanctions. If a firm is licensed and deficiencies present are
grounds for revocation (21 CFR 601.5), the firm can be advised that the agency
will proceed to revoke the firms license. Further, if a danger to health exists, the
firms license can be summarily suspended. Legal actions against a firm include
injunction and prosecution. Mandatory injunction (specifying corrective action),
as opposed to prohibitory injunction or license revocation (causing cessation of
operations), is the remedy most frequently sought when there is a need to maintain
the blood supplied by a firm operating out of compliance. Violative products can
be removed from the market by voluntary removal or correction of labeling by
the manufacturer. These actions are classified as recalls when FDA otherwise
would take action against the violative product, i.e., by seizure (17). See appendix
C for a summary of enforcement actions in recent years.
VI. CURRENT TRENDS IN BLOOD SAFETY

A. Role of DHHS
Federal public health agencies, i.e., FDA, NIH, and the Centers for Disease
Control and Prevention (CDC) are components of the Department of Health and
Human Services (DHHS). These PHS agencies report to the Assistant Secretary
for Health within DHHS. Each has a unique role in the nations public health. The
July 1995 Institute of Medicine (IOM) Report recommended that the depart-
ments role in blood safety be strengthened and formalized (8). Following a
review of decisions made from 1982 to 1986, the IOM panel recommended that
the Secretary of Health and Human Services designate a Blood Safety Director
at a high level to be responsible for the federal governments efforts to maintain
the safety of the nations blood supply. Additionally, the report recommended that
the PHS should establish a Blood Safety Council to assess current and potential
future threats to the blood supply, to propose strategies for overcoming these
threats, to evaluate the response of the Public Health Service to these proposals,
and to monitor the implementation of these strategies.
DHHS responded by establishing the Assistant Secretary for Health as the

Blood Safety Director. In addition, a Blood Safety Committee was formed
consisting of PHS agency heads, their deputies, and certain other members of
DHHS. The role of the Blood Safety Committee is to bring to the Blood Safety
Director issues that require department attention and coordination. The PHS
Advisory Committee on Blood Safety and Availability (ACBSA) also was formed
to advise the Blood Safety Director on matters relating to public health and to
provide public communication about blood risks. The advisory committee has
met periodically since April 1997 and has addressed safety and availability issues
such as hepatitis C lookback, transfusion risk from Creutzfeldt-Jacob disease
(CJD), and the availability of blood products such as immune globulins and
components for transfusion.
Lines of responsibility necessarily were drawn between the long-standing
Blood Products Advisory Committee (BPAC), an FDA scientific advisory com-
mittee, and the newly created PHS ACBSA. The BPACs role is to provide
scientific input and advice to FDA regarding regulatory matters such as determi-
nations of product safety and efficacy. The ACBSA has a broader composition
and advises the Blood Safety Director on global aspects of public health issues,
taking into account social, legal, economic, and ethical concerns as well as current
science. For example, while the BPAC may recommend to FDA that a newly
developed donor screening test be approved based on the performance character-
istics of the test and its demonstrated efficacy in clinical trials, the PHS ACBSA
may consider additional factors relevant to its implementation as a blood
donor screen, such as the risk/benefit ratio and the overall societal impact of
national testing.
B. Changes in Licensing
Section 351 of the PHS Act requires that a biological license be in effect for
commercial interstate distribution of a biological product. Regulations (21 CFR,
Section 601) implementing the PHS Act in the past required that licensing be
accomplished by the application for and issuance of separate licenses for the
establishment preparing the biological product and the product itself. The require-
ment for separate licensing of an establishment and each product manufactured
has been changed. The agency has eliminated the establishment license filing for
all biological products, including blood, blood components, and derivatives (12).
The current establishment license application and multiple product license appli-
cations have been replaced with a single form. Applications for product approval
(i.e., new drug and biologics licenses) are accomplished by the submission of one
harmonized application form for all drugs and biological products, and one
biologics license approval will issue covering the biological product and its
manufacturing facilities and processes. The harmonized application (FDA Form
356h, see Appendix D or http://forms.psc.gov/forms/FDA/fda.html) should be

included with all application submissions. Guidance documents have been pre-
pared to provide instructions on filing information required in the chemistry,
manufacturing, and controls (CMC) section and the establishment section of the
application. Conversion to the biologics license application (BLA) is voluntary at
present but will be required after October 20, 2000 (12).
The requirements and procedures for reporting changes to an approved
license application were revised in 1997 (13). Previously, the regulation affecting
reporting of changes to an approved application (21 CFR 601.12) required that
every important, proposed change be reported and that every change in manu-
facturing and labeling be approved by the Director of the CBER prior to
implementation of the change. In an effort to reduce the reporting burden of
industry and to more effectively manage limited government resources, the
regulation was revised to stratify reporting responsibilities based on a risk
assessment according to defined criteria. Changes judged to pose a significant
risk, to adversely affect the product must be reported in a license application
supplement and reviewed and approved prior to distribution of product prepared
with the change. Changes judged to have a moderate risk of adversely affecting
the product must be reported at least 30 days prior to distribution of product
produced with the change and remain subject to approval even after implementa-
tion. Changes with a minimal risk to the product may be implemented and
reported to the agency in an annual report. Changes in labeling are reported in a
similar fashion.
In addition to the two changes in blood and blood product licensing noted
above, work is progressing towards reinventing the mechanism for oversight of
blood products. These efforts may modify licensing requirements further, partic-
ularly for blood and blood components. FDA intends to pilot development of
monograph-type standards for conventional blood components as an alternative
to review SOPs and other supporting information in license applications. In this
regulatory scheme, license applicants would certify their compliance with the
published standards as a basis for licensure. Inspections would assess the validity
of applicants certifications and their adherence to published component standards.
C. Changes in the FDA Inspection Program

Inspections of licensed biologics establishments have been the responsibility of
agencies tasked with implementing the PHS Act since the beginning of this
century. When the NIHs Division of Biologics Standards joined FDA in 1972 as
the Bureau of Biologics (BoB), it retained its independent responsibility for
inspections, even though FDA had an established inspection program for other
products regulated under the FDC Act. However, in the late 1970s the BoB (now
CBER) transferred routine surveillance inspections of blood and plasma estab-
lishments to FDAs inspection force located within FDAs Office of Regulatory

Affairs (ORA) because of the increased inspection inventory resulting from the
registration of a large number of intrastate blood establishments and the promul-
gation of cGMP regulations for blood (21 CFR 606 et. seq.) that applied to
intrastate as well as interstate blood establishments. Experience over the past two
decades has proved this strategy to be practical and successful. However, recent
criticisms from industry also reflected in a General Accounting Office (GAO)
audit have focused on inconsistency among investigators and among FDA re-
gions; inadequate technical and policy training of investigators; and lack of a
centralized monitoring program to oversee investigator performance, identify
industry trends, and determine areas needing policy development (18). FDA has
taken steps to remedy these problems by creation of a specialized cadre of
investigators who inspect blood and plasma collection establishments.
With the exception of blood and plasma establishment inspections, both
prelicensing and routine surveillance inspections of all other biologics, including
plasma derivatives, until recently were the responsibility of FDAs CBER. How-
ever, reviews by oversight organizations, such as GAO, IOM, the DHHS Inspec-
tor General and Congress, indicated a need for change. Key criticisms were that
CBER inspections, although science based and product oriented, lacked strin-
gency in assessment of compliance cGMP, evidence development, and follow-up
enforcement actions. To address these concerns and as part of FDAs continuing
review of its practices under the National Performance Review (NPR), CBER
transferred the lead responsibility for periodic inspections of biologics to ORA.
This was accomplished within a partnership between FDAs ORA and CBER
called Team Biologics (15). The goal of Team Biologics is to improve the
compliance of biologics firms by combining the different strengths of ORA
and CBER in the inspection process. Part of this initiative involves creation of
a dedicated group of biologics investigators in ORA and the formation of
ORA/CBER teams to direct, conduct, and monitor inspections and compliance
actions for biologics firms. Most inspections now routinely are conducted by
teams consisting of both ORA investigators and CBER product specialist inspec-
tors. The transfer to ORA of the lead role for blood fractionator inspections was
accomplished in the fall of 1997, and the transfer of CBER-regulated in vitro
diagnostic products occurred in the spring of 1998.
The Team Biologics inspectional approach is being evaluated periodically
by a steering committee charged with oversight of the program.
D. Gene-Based Testing
In September 1994, FDA sponsored a workshop entitled Conference on the
Feasibility of Genetic Technology to Close the HIV Window in Donor Screen-
ing. The purpose of the workshop was to gather data on ways to close the
window of infectivity (i.e., the period in which infectious virus is present in

donated units of blood but not detected by current test methodology) for HIV.
Information presented at this workshop and in later publications showed that
nucleic acid testing, if applied to individual units of donated blood, could
significantly reduce the number of infectious donations in the blood supply. It also
was recognized that screening donors for the HIV-1 p24 antigen could be
beneficial as an interim measure. Such testing later was recommended by FDA
in August 1995 (19). This policy is controversial, however, because of its high
cost and limited benefits (low rate of detections and remaining risk). Recently,
estimates have indicated that nucleic acid testing of individual units of blood
could prevent approximately 12 HIV infectious donations per year, 84 HCV
infectious donations per year, and 81 HBV infectious donations per year (20).
Unfortunately, the technology available at present makes it impractical to test
individual units of blood; however, concurrent testing of small pools of donor
samples (minipools) is being studied as an interim measure.
Initially, manufacturers of fractionated blood products approached FDA
with proposals to utilize various nucleic acid (NAT) technologies to detect nucleic
acid sequences of HIV, HCV, and HBV viruses in preproduction pools of donated
plasma. While supportive of this initiative as a safety measure in the manufacture
of plasma derivatives, FDA took the view that testing of pooled plasma samples
de facto also was donor screening. FDA therefore requested that the proposals
under study contain algorithms for testing pools in a manner that allows identifi-
cation of the individual positive unit(s) within the pool. Test sponsors further were
advised to trace donors and notify them about test information that could affect
their health and their status for future donations. More recent studies of NAT
include not only plasma intended for fractionation, but also testing of transfusible
blood and blood components.
This area of testing is the first time that FDA has permitted use of pooled
sample testing for determining individual donor suitability, and the concept has
presented some novel challenges in addressing regulatory concerns. Among
those concerns are: test sample issues (sample quality and impact of pool size on
the sensitivity of the screen as a whole); test issues per se (sensitivity and
specificity of the test, manufacturing consistency, and test reproducibility); the
test environment (operator training and proficiency, adherence to cGMPs, and
control of contamination); and logistics (retrieving or removing the infected
units, time required for test performance, product retrieval, effect on short-dated
transfusible products, ability to trace positive results back to the donor, and
notification of recipients later found to have been transfused with investigational
NAT-positive blood).
FDA foresees the possibility that gene-based tests could replace some
current test methodologies. The first such replacement may be substitution of HIV
NAT testing for the currently approved HIV-1 p24 antigen test if pool testing is
shown to be sufficiently sensitive.
E. Precautions Related to Creutzfeld-Jakob Disease

Between 1983 and 1992, the FDA received four reports of persons who had
donated blood and, subsequently, were diagnosed with CJD. Implicated products
were managed ad hoc. However, subsequent to publication in 1993 of an FDA
guidance document on reporting of postdonation information, increasing numbers
of such cases became known. As a result, CJD and its safety implications for
blood and blood products were discussed at meetings of the BPAC in December
1994 and then again in June 1995 before a special advisory committee chartered
as the Transmissible Spongiform Encephalopathies Advisory Committee
(TSEAC). Based on advice of the TSEAC, FDA issued recommendations on
August 8, 1995, and again on December 11, 1996, regarding donor deferral
criteria for CJD and risk of CJD, and the quarantine and destruction of in-date
source plasma and plasma derivatives and in-date transfusible blood and blood
components from such donors (21,22). These recommendations identified several
categories of donors at increased risk of developing CJD based on iatrogenic and
familial risk. Iatrogenic risk applies to persons who have had injections with
human pituitary-derived growth hormone or have had dura mater transplants.
Familial risk applies to persons with one or more genetically related family
members with CJD who were told that their families have increased risk for CJD.
The FDA recommended permanent deferral of donors with CJD or CJD risks
unless, for cases of genetic risk, the donor underwent genetic testing that did not
reveal a familial CJD-associated abnormality of the prion protein gene.
While beneficial as a precautionary measure to reduce the theoretical risk
of transmission of CJD from blood products, FDAs policy on withdrawal of
plasma derivatives affected by a CJD-implicated donor has unintended conse-
quences in exacerbating shortages of plasma derivatives. The shortage situation
became a serious problem in the fall of 1997, leading to extensive reexamination
of the policy. Additionally, the emergence of a novel variant of CJD necessitated
further scientific assessment of the policy.
In 1996, a previously unrecognized variant of CJD was described in the
United Kingdom. The disease is referred to as new variant CJD (nvCJD) (23).
Laboratory and epidemiological studies have linked nvCJD to an outbreak of
bovine spongiform encephalopathy (BSE) in the United Kingdom. BSE infection
in cattle in the United Kingdom appeared in 1980, peaked in 1992, and fell to low
levels by 1996. To date, no cases of nvCJD or BSE have been identified in the
United States. The risk for transmission of nvCJD by blood and blood products
is unknown. However, theoretical considerations suggest that the risk may be
greater than for classical CJD.
Laboratory and epidemiological studies have been conducted to assess the

risk of transmission of classical CJD through blood and blood products, and
accumulating information indicates that the transmission of the CJD infectious
agent is highly unlikely (24,25). In contrast, epidemiological studies of nvCJD
are too small to provide useful risk assessments, and laboratory studies are only
starting to be conducted.
On January 19, 1998, the PHS ACBSA reviewed available information
concerning CJD transmissibility by blood as well as the impact of CJD-related
withdrawals upon the supply of medically necessary plasma derivatives. Based
upon this review, the committee recommended that FDA consider revising its
guidance to the extent necessary to relieve shortages of medically necessary
plasma derivatives. The recommendation subsequently was considered by the
PHS ACBSA. At its July 23, 1998, meeting, Assistant Secretary for Health and
Surgeon General David Satcher, M.D., Ph.D., announced that plasma derivatives
should be withdrawn and intermediates quarantined only if a blood donor
developed nvCJD and that previously recommended withdrawals and quarantines
be discontinued for classical CJD and CJD risk factors. FDA made a consistent
recommendation available on the internet on September 8, 1998 (26).
It is recognized that residents of the United Kingdom have an increased risk
of developing nvCJD; however, the extent of the risk is not yet known. Prompted
by a February 1998 decision in the United Kingdom to import all plasma for
fractionation, the TSEAC met on December 18, 1998, and June 2, 1999, to
consider whether donors who have traveled to or resided in the United Kingdom
should be deferred. At these meetings, the TSEAC recommended that such donors
be deferred until more is known about the scope of the epidemic of nvCJD and the
risk of transmission of nvCJD by blood. A policy decision by the PHS that donors
should be deferred indefinitely if they have spent 6 months or more cumulatively
in the United Kingdom between January 1, 1980, and December 31, 1996, was
announced at the June 1999 meeting of the BPAC, and a guidance document
recommending this deferral was published by FDA in November 1999 (27).
F. Hepatitis C Virus Lookback

Non-A, non-B hepatitis, now known to be caused predominantly by HCV, was
first identified in studies of posttransfusion hepatitis in the 1970s (28). Once
thought to be a benign form of hepatitis, HCV is now known to be a major cause
of chronic liver disease with significant risk of progression to cirrhosis, liver
failure, and hepatocellular carcinoma decades after the initial infection. Donor
testing for HCV infection was not possible until May 1990, following FDAs
approval of an antibody-detection test with sensitivity between 50 and 80% based
on a single recombinant antigen, FDA recommended at that time that units of
blood from donors testing reactive on screening not be used for transfusion. On
April 23, 1992, the recommendation was revised to extend testing to both
products for transfusion and for further manufacturing into injectable products.
This decision coincided with FDA approval of the first multiantigen test for
antibodies to HCV, a test with sensitivity greater than 90%. A supplemental, more
specific test was licensed (available previously only as an investigational test) in
mid-1993.
Tracing of recipients who received blood components from donors later
found reactive for antibodies to HCV has been termed targeted lookback.
Although targeted lookback for HCV had been discussed in public meetings since
October 1989, it was not recommended for a variety of reasons. Most importantly,
in the absence of supplemental testing, the positive predictive value of the first
HCV screening test was low and the significance of a reactive test in terms of
infectivity was not known. Additionally, few HCV-infected individuals were
eligible for investigational treatment of their liver disease, and the long-term
effect of treatment was unknown.
Over time, it became known that individuals reactive for anti-HCV in a
supplemental assay are likely to be chronically infected with HCV. More specif-
ically, in studies of blood donors, 7395% of supplemental testpositive and
1421% of supplemental testindeterminate blood donors had detectable HCV
RNA by PCR (2931). It also is now recognized that negative or unscreened units
from donors later found reactive for anti-HCV may have been contaminated with
HCV. In parallel with these improved understandings, there have been im-
provements in the management and treatment of HCV infections that were
recognized by publication of treatment guidelines developed at a NIH Consensus
Conference in March 1997. Driven in large measure by congressional concerns
over the emerging epidemic of liver disease from chronic HCV infections, the
issue of targeted lookback for HCV was brought before the PHS ACBSA. The
Committee discussed the topic at its April 2425, 1997, meeting and its August
1112, 1997, meeting. After careful consideration, the Committee recommended
that the PHS initiate a targeted lookback program extending back 10 years to
identify recipients of previously donated units from donors who have tested
positive for antibody to HCV following screening by a multiantigen screening
test [enzyme immunoassay (EIA)] since 1992.
In March 1998 and September 1998, FDA issued guidance consistent with
this recommendation. At public meetings on November 24, 1998, and January 28,
1999, the PHS ACBSA reconsidered the issue of recipient notification related to
repeatedly reactive results on the EIA 1.0 test. Following acceptance by the
committee, FDA issued guidance on June 22, 1999, to replace earlier guidances
(32). This current guidance provides for lookback to EIA 1.0 testing and also
recommends that the search of records of prior donations from donors with
repeatedly reactive screening tests for HCV extend back indefinitely to the extent
that electronic or other readily retrievable records exist.
G. Donor Recruitment and Quality Issues

Recruitment and selection of quality donors are critical steps in ensuring the
safety and availability of the nations blood supply. Existing regulations address
suitability of the donor (21 CFR 640.3 and 640.63) and labeling of products based
on donor source, either from a paid or volunteer donor [21 CFR 606.121(c)(5)].
In recent years, there have been renewed concerns about donor quality. The issue
of donor incentives and their impact on blood donor motivation was highlighted
at a workshop co-sponsored by FDA and the NIHs National Heart, Lung, and
Blood Institute (NHLBI) in September 1996. Since then, data from the Retrovirus
Epidemiology Donor Study (REDS) has indicated that approximately 2% of
blood donors in the sample reported at least one or more risk behaviors that
should have resulted in their deferral at the time of the screening interview. Within
this group approximately 0.4% reported a deferrable risk behavior within 3
months before the donation, i.e., within the window period of many bloodborne
infections (33). The need to study donor behavior and motivation issues and
validate donor history questions was discussed at the March 2526, 1999, BPAC
meeting. Information was presented to the PHS Advisory Committee on Blood
Safety and Availability at its April 2930, 1999, meeting suggesting that the
demand for blood is increasing faster than the supply. Strategies to address donor
motivation and quality issues with respect to increasing a safe blood supply are
under study. This is particularly important at this time since it is anticipated that
the deferral of donors who have traveled to or resided in the United Kingdom for
6 months or more cumulatively from January 1980 through December 1996 will
reduce the donor population by 2.2% nationally.
The federal role in ensuring blood safety continues to evolve under public
pressure and scrutiny. Likewise, the blood industry is being held to higher
standards for adherence to cGMP by FDA. The goal of many current efforts is to
prevent the possibility for another bloodborne epidemic similar to AIDS. The
American public expects and deserves the safest blood supply that is achievable.
This goal lies at the heart of federal regulation of blood safety.
REFERENCES
1. Solomon JM. The evolution of the current blood banking regulatory climate. Trans-
fusion 1994; 34:272277.
2. Noguchi PD. From Jim to Gene and beyond: an odyssey of biologics regulation. Food
Drug Law 1996; 51:367373.
3. Blank v. United States. Federal Reporter, 2d Series 1968; 400:302306.
4. Public Law No. 91-515, 58 Stat. 702 (1970).
5. Beatrice MG. Regulation, licensing and inspection of biological products. Phar-
maceut Eng 1991; 10:2935.
6. The Food and Drug Administration Modernization Act of 1997. Public Law No.
105-115, 1997.
7. Compilation of Laws Enforced by the U.S. Food and Drug Administration and
Related Statutes. Vol 1. Washington, DC: U.S. Government Printing Office, 1996.
8. Leveton LB, Sox HC Jr., Stoto MA, eds. HIV and the Blood Supply: An Analysis of
Crisis Decision Making. Washington, DC: National Academy Press, 1995.
9. The Food and Drug Administrations Development, Issuance, and Use of Guidance
Documents. Fed Reg 1997; 62:89618972.
10. Intercenter Agreement Between the Center for Drug Evaluation and Research and the
Center for Biologics Evaluation and Research. Rockville, MD: Food and Drug
Administration, 1991.
11. Intercenter Agreement Between the Center for Drug Evaluation and Research and the
Center for Devices and Radiological Health. Rockville, MD: Food and Drug Admin-
istration, 1991.
12. Biological Products Regulated under Section 351 of the Public Health Service
Act: Implementation of Biologics License Elimination of Establishment License
and Product License. Fed Reg 1999; 64:S6441S6454.
13. Changes to an approved application. Fed Reg 1997; 62:3989039906.
14. Biological products; reporting of errors and accidents in manufacturing: proposed
rule. Fed Reg 1997; 62:4964249648.
15. Team Biologics: A Plan for Reinventing FDAs Ability to Optimize Compliance of
Regulated Biologics Industry. Rockville, MD: Food and Drug Administration, 1997.
16. Compliance Program Guide for Plasma Fractionation Facilities. Rockville, MD: Food
and Drug Administration, 1997.
17. Tourault MA. Modern principles of blood banking compliance with Food and Drug
Administration regulations, In: Harmening DM, eds. Modern Blood Banking and
Transfusion Practices. 3rd ed. Philadelphia: F.A. Davis Company 1994:288302.
18. Blood Supply: FDA Oversight and Remaining Issues of Safety. Washington, DC:
General Accounting Office, 1997; PEMD-97-1.
19. Recommendations for Donor Screening with a Licensed Test for HIV-1 Antigen.
Food and Drug Administration, 1995.
21. Precautionary Measures to Further Reduce the Possible Risk of Transmission of
Creutzfeldt-Jakob Disease by Blood and Blood Products. Rockville, MD: Food and
Drug Administration, 1995.
22. Revised Precautionary Measures to Reduce the Possible Risk of Transmission of
Creutzfeldt-Jakob Disease (CJD) by Blood and Blood Products. Fed Reg 1997;
62:4969449695.
23. Will RG, Ironside JW, Zeidler M. A new variant of Creutzfeldt-Jakob disease in the
U.K. Lancet 1996; 347:921925.
24. Sullivan MT, Schonberger LB, Kessler D, Williams A, Dodd R. Creutzfeldt-Jakob
disease (CJD) investigational lookback study. Transfusion 1997; 37(suppl):2S.
25. Heye N, Hensen S, Muller N. Creutzfeldt-Jakob disease and blood transfusion.
Lancet 1994; 343:298299.
26. Change to the Guidance Entitled Revised Precautionary Measures to Reduce the
Possible Risk of Transmission of Creutzfeldt-Jakob Disease (CJD) by Blood and

Blood Products,Information Sheet. Rockville, MD: Food and Drug Administra-
tion, 1998.
27. Guidance for Industry: Revised Precautionary Measures to Reduce the Possible Risk
of Transmission of Creutzfeldt-Jakob Disease (CJD) and New Variant Creutzfeldt-
Jakob Disease (NVCJD) by Blood and Blood Products; Availability. Fed Reg 1999;
64:6571565716.
28. Public Health Service Interagency Guidelines for Screening Donors of Blood,
Plasma, Organs, Tissues, and Semen for Evidence of Hepatitis B and Hepatitis C.
MMWR 1991; 40:117.
29. Sayers MH, Gretch DR. Recombinant immunoblot and polymerase chain reaction
testing in volunteer whole blood donors screened by a multi-antigen assay for
hepatitis C virus antibodies. Transfusion 1993; 33:809813.
30. Kleinman SH, Alter H, Busch M, Holland P, Tegtmeier G, Nelles M, Lee S, Page E,
Wilber J, Polito A. Increased detection of hepatitis C virus (HCV)-infected blood
donors by a multiple-antigen HCV enzyme immunoassay. Transfusion 1992; 32:805813.
31. Yun Z, Lindh G, Weiland O, Johansson B, Sonnerborg A. Detection of hepatitis C
virus (HCV) RNA by PCR related to HCV antibodies in serum and liver histology
in Swedish blood donors. J Med Virol 1993; 39:5761.
32. Draft guidance for industry: Current Good Manufacturing Practice for blood and
blood components: (1) quarantine and disposition of prior collections from donors
with repeatedly reactive screening tests for hepatitis C virus (HCV); (2) supplemental
testing, and the notification of consignees and transfusion recipients of donor test
results for antibody to HCV (anti-HCV). Notice of availability. Fed Reg 1999;
64:3330933313.
SH, Hollingsworth CG, Nemo G. Estimates of infectious disease risk factors in U.S.
blood donors. J Am Med Assoc 1997; 277:967972.
APPENDIX A: U.S. FOOD AND DRUG ADMINISTRATION

CENTER FOR BIOLOGICS EVALUATION AND RESEARCH
INFORMATION SOURCES
As a service to the public, FDAs Center for Biologics Evaluation and Research
(CBER) provides information in a variety of ways.
Internet
CBERs Internet site is located at: http://www.fda.gov/cber/
The site contains a myriad of information including:
Product information, e.g., recall/withdrawal/safety issues, product
approvals, information sheets, adverse event reporting information
(www.fda.gov/cber/products.htm);
On line documents, e.g., guidelines/guidances, product approval docu-
ments, establishment and product files, Federal Register notices, informa-
tion sheets, letters to industry and healthcare providers, memoranda to
blood establishments, points to consider, and general and administrative
information about CBER (www.fda.gov/cber/publications.htm);
Documents available electronically under the Freedom of Information Act
(FOI) and how to submit an FOI request (www.fda.gov/cber/efoi.htm);
Information on FDA Modernization Act (www.fda.gov/cber/fdama.htm)
and Prescription Drug User Fee Act (www.fda.gov/cber/pdufa.htm)
Additional Internet Sources

Office of Regulatory Affairs (ORA) Web Site (www.fda.gov/ora) This
site has direct links to a variety of information including ORA Field
Contacts
FDA downloadable forms, e.g., Biologics License Application, registration
and label submittal forms (http://forms.psc.dhhs.gov/fdaforms.htm)
Fax Information System

Direct toll free access in the U.S.: 1-888-CBER-FAX or 1-888-223-
7329
Outside the U.S. & local to Rockville, MD: 301-827-3844
Updated August 23, 1999

Callers outside the U.S. must call this system from a FAX machine with a
touch-tone telephone attached or built in.
When prompted to enter your FAX number, please enter the entire 10 digit
number, including the 3 digit area code, even if you are local to Rockville, MD.
Do not include a 1 for long distance dialing.
To obtain a complete list of documents available from this system,
select document 9999
To obtain a list of documents added to the system in the last 30 days,
select document 9998
To obtain a list of Recall/Withdrawal/Safety notifications, select docu-
ment 9997
Up to 5 documents can be ordered at a time. Please enter the document
number(s) listed on the index.
E-Mail
Consumer questions about biological products can be sent by e-mail
to: OCTMA@CBER.FDA.GOV
Manufacturers assistance questions can be sent by e-mail to: MATT
@CBER.FDA.GOV
Documents can be requested by e-mail from: CBER_INFO@CBER.
FDA.GOV
Telephone
Voice Information System
Direct access to Consumer Safety Officers or Public Affairs Specialists:
1-800-835-4709 or 301-827-1800
Blood and Plasma Products Information
Recall and market withdrawal notices for fractionated plasma products:
1-888-CBER-BPI or 301-827-4604
Division of Blood Applications 301-827-3543 (for questions about
your application)
Blood and Plasma Branch (fax: 301-827-2857)
Regulatory Project Management Branch (fax: 301-827-3534)
Devices Review Branch (fax: 301-827-3535)
Automated E-Mail List

CBER has established three automated e-mail lists to distribute infor-
mation within CBER and to the public.
FPRECALL
Members of this list will receive notices of recalls and market
withdrawals of fractionated products.
BLOODINFO
Members of this list will receive all FPRECALL notices and other
blood-related documents.
CBERINFO
Members of this list will receive FPRECALL and BLOODINFO
documents, and notification of ALL new documents, including
guidelines, points to consider, Whats New on CBERs web site, and
other CBER information.
Subscribing to multiple lists will result in receiving multiple copies.

To subscribe, send an e-mail message to FDALISTS@ARCHIE.FDA.GOV
THE FIRST LINE OF THE MESSAGE MUST CONTAIN:

subscribe <listname> <e-mail address>
Example: If Joan Smith, at company.com, wanted to subscribe to the
CBERINFO list, the first line of her message would look like this:
subscribe CBERINFO JSmith@company.com
Printed Copy
Single copies of documents are available. Requests may be sent in writing to:
Office of Communication, Training and Manufacturers Assistance

(HFM-40),
Center for Biologics Evaluation and Research (CBER),
Food and Drug Administration
1401 Rockville Pike
Rockville, MD 20852-1448
Documents may also be obtained by calling CBERs Voice Information System at:
1-800-835-4709 or 301-827-1800
For a complete list of guidelines, guidance, points to consider and other docu-
ments available in hard copy, request document D9001 (hard copy ID number)
or document 9001 from the FAX Information System.
For a complete list of memoranda and related documents pertaining to blood and
blood products available in hard copy, request document D9002 (hard copy ID
number ) or document 9002 from the FAX Information System.
APPENDIX B
Form Approved: OMB No. 0910-0291 Expires: 8/31/00

See OMB statement on reverse
For use by user-facilities, Mfr report #
distributors and manufacturers for

MANDATORY reporting UF/Dist report #
Page ____ of ____ FDA Use Only
A. Patient information C. Suspect medication(s)

1. Patient identifier 2. Age at time 3. Sex 4. Weight 1. Name (give labeled strength & mfr/labeler, if known)
of event:
or female lbs #1
or
Date
of birth: male kgs
#2
In confidence
2. Dose, frequency & route used 3. Therapy dates (if unknown, give duration)
B. Adverse event or product problem from/to (or best estimate)
1. Adverse event Product problem (e.g., defects/malfunctions) #1 #1

and/or
2. Outcomes attributed to adverse event
(check all that apply) disability #2 #2
death __________________ congenital anomaly 4. Diagnosis for use (indication) 5. Event abated after use
stopped or dose reduced
(mo/day/yr)
required intervention to prevent #1
life-threatening permanent impairment/damage #1 yes no doesn't
apply
hospitalization initial or prolonged other: #2
___________________
6. Lot # (if known) 7. Exp. date (if known)
#2 yes no doesn't
apply
3. Date of 4. Date of
#1 #1 8. Event reappeared after
event this report
(mo/day/yr) (mo/day/yr) reintroduction
#2
5. Describe event or problem #2
#1 yes no doesn't
apply
9. NDC # for product problems only (if known)
#2 yes no doesn't
apply
10. Concomitant medical products and therapy dates (exclude treatment of event)
D. Suspect medical device

1. Brand name
2. Type of device
3. Manufacturer name & address 4. Operator of device

health professional
lay user/patient
other:
________________
5. Expiration date
6. (mo/day/yr)
model #________________________________________
7. If implanted, give date
6. Relevant tests/laboratory data, including dates catalog #_______________________________________ (mo/day/yr)
serial # ________________________________________
8. If explanted, give date
lot # ___________________________________________ (mo/day/yr)
other #
9. Device available for evaluation? (Do not send to FDA)
yes no returned to manufacturer on _________________
(mo/day/yr)
10. Concomitant medical products and therapy dates (exclude treatment of event)
7. Other relevant history, including preexisting medical conditions (e.g., allergies,

race, pregnancy, smoking and alcohol use, hepatic/renal dysfunction, etc.)
E. Initial reporter
1. Name & address phone #
2. Health professional? 3. Occupation 4 Initial reporter also

Submission of a report does not constitute an sent report to FDA
admission that medical personnel, user facility, yes no
distributor, manufacturer or product caused or yes no unk
contributed to the event.
FDA Form 3500A
Medication and Device Submission of a report does not constitute U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES
Public Health Service Food and Drug Administration
an admission that medical personnel, user
Experience Report facility, distributor, manufacturer or product
(continued) caused or contributed to the event.
Refer to guidelines for specific instructions Page ____ of ____ FDA Use Only
F. For use by user facility/distributordevices only H. Device manufacturers only

1. Check one 2. UF/Dist report number 1. Type of reportable event 2. If follow-up, what type?
user facility distributor death correction
3. User facility or distributor name/address serious injury additional information
malfunction (see guidelines) response to FDA request
other: ________________________ device evaluation
3. Device evaluated by mfr? 4. Device manufacture date

not returned to mfr.
(mo/yr)
4. Contact person 5. Phone Number yes evaluation summary attached

no (attach page to explain why not) 5. Labeled for single use?
or provide code: yes no
6. Date user facility or distributor 7. Type of report 8. Date of this report ________________________
became aware of event (mo/day/yr)
(mo/day/yr) initial 6. Evaluation codes (refer to coding manual)
follow-up # _____
method
9. Approximate 10. Event problem codes (refer to coding manual)
age of device
patient results
code
device conclusions
code
11. Report sent to FDA? 12. Location where event occurred
yes ___________________ hospital outpatient 7. If remedial action initiated,
check type
8. Usage of device
no
(mo/day/yr)
home diagnostic facility

nursing home ambulatory recall notification
initial use of device
13. Report sent to manufacturer?
outpatient
surgical facility reuse
repair inspection
yes ___________________
treatment facility unknown
no (mo/day/yr) other: replace patient monitoring
9. If action reported to FDA under
specify
14. Manufacturer name/address relabeling modification/ 21 USC 360i(f), list correction/removal
adjustment reporting number:
other:
10. Additional manufacturer narrative and/or 11. Corrected data
G. All manufacturers
1. Contact office name/address (& mfring site for devices) 2. Phone number
3. Report source
(check all that apply)
foreign
study
literature
consumer
health
4. Date received by manufacturer 5. professional
(mo/day/yr) (A) NDA # ___________
user facility
6. If IND, protocol #
IND # ___________ company
representative
PLA # ___________
distributor
pre-1938 yes
other:
7. Type of report
(check all that apply) OTC
product
yes
5-day 15-day
8. Adverse event term(s)
10-day periodic
Initial follow-up # ____
9. Mfr. report number
The public reporting burden for this collection of information has been estimated to average one- DHHS Reports Clearance Office An agency may not conduct or sponsor,
hour per response, including the time for reviewing instructions, searching existing data sources, Paperwork Reduction Project (0910-0291) and a person is not required to respond to,
Please DO NOT RETURN this
gathering and maintaining the data needed, and completing and reviewing the collection of infor- Hubert H. Humphrey Building, Room 531-H a collection of information unless it displays form to this address.
mation. Send comments regarding this burden estimate or any other aspect of this collection of 200 Independence Avenue, S.W. a currently valid OMB control number.
information, including suggestions for reducing this burden to: Washington, D.C. 20201
FDA Form 3500A - back

APPENDIX C
CBER Enforcement Actions

Action FY95 FY96 FY97 FY98 FY99
Revocationsa 5 0 3 5 5
Suspensions 3 1 3 3 0
Injunctions 0 1 3 1 0
Seizures 0 0 3 0 1
Warning letters 19 21 46 31 27
aIncludes Notices of Intent to Revoke.
Recalls Classified
Type of
product FY95 FY96 FY97 FY98 FY99
Blood 592 669 1423 1524 1199

Source plasma 19 23 27 38 36
Devicea 29 7 30 23 12
Vaccineb 0 1 8 4 14
Therapeuticc 3 4 29 4 15
Tissue 5 3 2 5 19
Total 648 707 1519 1598 1295
aIncludes in vitro diagnostic test kits and blood bank software.
bIncludes allergenic products.
cIncludes blood and plasma derivatives.
APPENDIX D
Form Approved: OMB No. 0910-0338

DEPARTMENT OF HEALTH AND HUMAN SERVICES Expiration Date: March 31, 2003
FOOD AND DRUG ADMINISTRATION See OMB Statement on page 2.
APPLICATION TO MARKET A NEW DRUG, BIOLOGIC, FOR FDA USE ONLY

OR AN ANTIBIOTIC DRUG FOR HUMAN USE APPLICATION NUMBER
(Title 21, Code of Federal Regulations, Parts 314 & 601)
APPLICANT INFORMATION
NAME OF APPLICANT DATE OF SUBMISSION
TELEPHONE NO. (Include Area Code) FACSIMILE (FAX) Number (Include Area Code)
APPLICANT ADDRESS (Number, Street, City, State, Country, ZIP Code or Mail Code, AUTHORIZED U.S. AGENT NAME & ADDRESS (Number, Street, City, State,
and U.S. License number if previously issued): ZIP Code, telephone & FAX number) IF APPLICABLE
PRODUCT DESCRIPTION
NEW DRUG OR ANTIBIOTIC APPLICATION NUMBER, OR BIOLOGICS LICENSE APPLICATION NUMBER (If previously issued)
ESTABLISHED NAME (e.g., Proper name, USP/USAN name) PROPRIETARY NAME (trade name) IF ANY
CHEMICAL/BIOCHEMICAL/BLOOD PRODUCT NAME (If any) CODE NAME (If any)
DOSAGE FORM: STRENGTHS: ROUTE OF ADMINISTRATION:
(PROPOSED) INDICATION(S) FOR USE:
APPLICATION INFORMATION
APPLICATION TYPE
(check one) NEW DRUG APPLICATION (21 CFR 314.50) ABBREVIATED NEW DRUG APPLICATION (ANDA, 21 CFR 314.94)
BIOLOGICS LICENSE APPLICATION (21 CFR Part 601)
IF AN NDA, IDENTIFY THE APPROPRIATE TYPE 505 (b)(1) 505 (b)(2)

IF AN ANDA, OR 505(b)(2), IDENTIFY THE REFERENCE LISTED DRUG PRODUCT THAT IS THE BASIS FOR THE SUBMISSION
Name of Drug Holder of Approved Application
TYPE OF SUBMISSION (check one) ORIGINAL APPLICATION AMENDMENT TO A PENDING APPLICATION RESUBMISSION
PRESUBMISSION ANNUAL REPORT ESTABLISHMENT DESCRIPTION SUPPLEMENT EFFICACY SUPPLEMENT
LABELING SUPPLEMENT CHEMISTRY MANUFACTURING AND CONTROLS SUPPLEMENT OTHER
IF A SUBMISSION OF PARTIAL APPLICATION, PROVIDE LETTER DATE OF AGREEMENT TO PARTIAL SUBMISSION:
IF A SUPPLEMENT, IDENTIFY THE APPROPRIATE CATEGORY CBE CBE-30 Prior Approval (PA)
REASON FOR SUBMISSION
PROPOSED MARKETING STATUS (check one) PRESCRIPTION PRODUCT (Rx) OVER THE COUNTER PRODUCT (OTC)
NUMBER OF VOLUMES SUBMITTED THIS APPLICATION IS PAPER PAPER AND ELECTRONIC ELECTRONIC
ESTABLISHMENT INFORMATION (Full establishment information should be provided in the body of the Application.)
Provide locations of all manufacturing, packaging and control sites for drug substance and drug product (continuation sheets may be used if necessary). Include name,
address, contact, telephone number, registration number (CFN), DMF number, and manufacturing steps and/or type of testing (e.g. Final dosage form, Stability testing)
conducted at the site. Please indicate whether the site is ready for inspection or, if not, when it will be ready.
Cross References (list related License Applications, INDs, NDAs, PMAs, 510(k)s, IDEs, BMFs, and DMFs referenced in the current application)
FORM FDA 356h (4/00) Created by Media Arts/USDHHS: (301) 443-2454 EF
PAGE 1
This application contains the following items: (Check all that apply)
1. Index
2. Labeling (check one) Draft Labeling Final Printed Labeling
3. Summary (21 CFR 314.50 (c))

4. Chemistry section
A. Chemistry, manufacturing, and controls information (e.g., 21 CFR 314.50(d)(1); 21 CFR 601.2)
B. Samples (21 CFR 314.50 (e)(1); 21 CFR 601.2 (a)) (Submit only upon FDAs request)
C. Methods validation package (e.g., 21 CFR 314.50(e)(2)(i); 21 CFR 601.2)
5. Nonclinical pharmacology and toxicology section (e.g., 21 CFR 314.50(d)(2); 21 CFR 601.2)
6. Human pharmacokinetics and bioavailability section (e.g., 21 CFR 314.50(d)(3); 21 CFR 601.2)
7. Clinical Microbiology (e.g., 21 CFR 314.50(d)(4))
8. Clinical data section (e.g., 21 CFR 314.50(d)(5); 21 CFR 601.2)

9. Safety update report (e.g., 21 CFR 314.50(d)(5)(vi)(b); 21 CFR 601.2)
10. Statistical section (e.g., 21 CFR 314.50(d)(6); 21 CFR 601.2)
11. Case report tabulations (e.g., 21 CFR 314.50(f)(1); 21 CFR 601.2)
12. Case report forms (e.g., 21 CFR 314.50 (f)(2); 21 CFR 601.2)
13. Patent information on any patent which claims the drug (21 U.S.C. 355(b) or (c))
14. A patent certification with respect to any patent which claims the drug (21 U.S.C. 355 (b)(2) or (j)(2)(A))
15. Establishment description (21 CFR Part 600, if applicable)

16. Debarment certification (FD&C Act 306 (k)(1))
17. Field copy certification (21 CFR 314.50 (k)(3))
18. User Fee Cover Sheet (Form FDA 3397)
19. Financial Information (21 CFR Part 54)

20. OTHER (Specify)
CERTIFICATION
I agree to update this application with new safety information about the product that may reasonably affect the statement of contraindications,
warnings, precautions, or adverse reactions in the draft labeling. I agree to submit safety update reports as provided for by regulation or as
requested by FDA. If this application is approved, I agree to comply with all applicable laws and regulations that apply to approved applications,
including, but not limited to the following:
1. Good manufacturing practice regulations in 21 CFR Parts 210, 211 or applicable regulations, Parts 606, and/or 820.
2. Biological establishment standards in 21 CFR Part 600.
3. Labeling regulations in 21 CFR Parts 201, 606, 610, 660, and/or 809.
4. In the case of a prescription drug or biological product, prescription drug advertising regulations in 21 CFR Part 202.
5. Regulations on making changes in application in FD&C Act Section 506A, 21 CFR 314.71, 314.72, 314.97, 314.99, and 601.12.
6. Regulations on Reports in 21 CFR 314.80, 314.81, 600.80, and 600.81.
7. Local, state and Federal environmental impact laws.
If this application applies to a drug product that FDA has proposed for scheduling under the Controlled Substances Act, I agree not to market the
product until the Drug Enforcement Administration makes a final scheduling decision.
The data and information in this submission have been reviewed and, to the best of my knowledge are certified to be true and accurate.
Warning: A willfully false statement is a criminal offense, U.S. Code, title 18, section 1001.
SIGNATURE OF RESPONSIBLE OFFICIAL OR AGENT TYPED NAME AND TITLE DATE
ADDRESS (Street, City, State, and ZIP Code) Telephone Number
( )
Public reporting burden for this collection of information is estimated to average 24 hours per response, including the time for reviewing
instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of
information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing
this burden to:
Department of Health and Human Services Food and Drug Administration
Food and Drug Administration CDER, HFD-94 An agency may not conduct or sponsor, and a
CBER, HFM-99 12420 Parklawn Dr., Room 3046 person is not required to respond to, a collection
1401 Rockville Pike Rockville, MD 20852 of information unless it displays a currently valid
Rockville, MD 20852-1448 OMB control number.
FORM FDA 356h (4/00)
PAGE 2
Index
Page numbers in italic indicate figures or tables.
ABO antibodies, 50 Anaphylactic reactions, 7476

Accreditation by AABB, 576577 causes of, 75
Acquired immune deficiency syndrome Antibodies:
(AIDS) (see Human immuno- passive transfer, 51
deficiency virus) thermal amplitude, 51
Acute hemolytic transfusion reactions, Anti-hemophilic factor (AHF)
5260 concentrates:
clinical presentation, 5657 contamination with hepatitis A, 175
management, 5860 contamination with HIV, 251252,
pathophysiology, 5758 479
signs and symptoms, 56 Anti-hepatitis B core (anti-HBc),
162163, 167, 186, 210,
Acute preoperative hemodilution, 454
280281
Adult T-cell leukemia, 299
Anti-hepatitis C virus testing, 209
Adverse reaction reporting to FDA, 586
Anti-HIV-1 testing, 202205
Advisory Committee on Blood Safety
Anti-HTLV-I/II testing, 206208,
and Availability, 588, 595
302304
Air embolism, 81, 454 Arrhythmia, transfusion-associated, 80
Alanine aminotransferase (ALT), 162, Assessments of quality systems,
167, 186, 193, 283 516518
cost effectiveness of, 530 Autologous blood donation, 451454
tests for, 210211 cost effectiveness of, 523528, 535
Algorithms for transfusion, 449451
Alloimmunization: Babesiosis, 2, 174, 408417, 436
to HLA antigens, 7778, 473474 agents, 409410
to red cell antigens, 50, 7677 age of implicated blood, 415, 416
American Association of Blood Banks clinical presentation, 410411
(AABB) standards, 569577 deferral criteria, 23
607
608 Index
[Babesiosis] Chelation therapy, 79

duration of infectivity, 415, 417 Circulatory overload, 7879
incubation period, 412, 414 Citrate toxicity, 80
patients at risk, 412 Coagulation factor concentrates:
prevention of, 417418 contamination of, 175, 251252
survival in blood, 412 safety of, 482483
testing for, 417418 Commercial donors, 5
transfusion transmission, 411, and hepatitis risk, 5
412417 Compatibility testing, 6467
vector, 409410 Confidential unit exclusion, 2931,
Bacterial contamination, 2526, 262263
171172, 221243 Confirmatory testing for infectious dis-
of autologous units, 225 ease markers, 195200
clinical presentation, 223226 Controls for infectious disease testing,
detection of, 226233 188190
incidence of, 221222, 435436, 533 Correctional institutions, 15
and leukoreduction, 237238, 469 Cost effectiveness, 521537
organisms associated with, 223 Creutzfeldt-Jakob disease, 2, 335349
prevention of, 233240, 534 animal models for transmission,
reporting of, 241 340341
deferral criteria, 2425, 344345,
Bacterial infection, and immuno-
532, 592593
modulation, 143147
familial, 339
Blood Products Advisory Committee,
and FDA, 592593
588
iatrogenic, 337338
Blood salvage, 454455
inactivation/removal, 347
Blood substitutes, 543563 (see also
and leukoreduction, 345346
Red blood cell substitutes)
lookback studies, 342343
Borrelia burgdorferi (see Lyme disease)
and mad cow disease, 339340
new variant, 15, 335, 338340
Callback, 3132 and plasma pool size, 347348
Cancer recurrence, and immuno- prevention of, 343348
modulation, 140143 prion protein, 335337
CD4+ lymphocytes, 257259 risk of transmission by blood,
Chagas disease, 2, 366387 170171, 340343, 434435
agent and life cycle, 367371 sporadic, 337
chemoprophylaxis for, 380383 testing for, 346347
clinical presentation, 378379 and United Kingdom travel, 345
deferral criteria, 23, 380 Crystal violet, 381383
prevalence in donors, 372, 373 Cut-off, calculation of, 188189
prevention, 379383 Cytomegalovirus (CMV), 170, 315328
risk of transmission, 170, 173174, epidemiology of, in blood donors,
375378, 437 324325
testing for, 383387 latency, 316318
transmission by blood, 372, 374375 leukoreduction to prevent, 325328
treatment of, 387 testing for, 320323
Index 609
[Cytomegalovirus (CMV)] Food and Drug Administration (FDA):

transfusion transmission epidemi- adverse reaction reporting, 586,
ology, 318320 601602
application form, 604605
D antigen, immunogenicity of, 51 Blood Products Advisory Committee,
Delayed hemolytic transfusion reac- 588
tions, 6064 and Creutzfeldt-Jakob disease,
clinical presentation, 6162 592593
management of, 62 fatality reporting, 586
Designated/directed donation, 456 and gene-based testing, 590592
Disseminated intravascular coagulation guidance documents, 581584
(DIC), 59 inspections, 586587, 589590
Donor deferral registries, 4, 2728 licensing and product approval,
Donors: 584585 , 588589
first-time, 6 and lookback for HCV, 593594
paid, 5 and nucleic acid testing, 590592
recruitment of, 6 regulations, 581583
repeat, 7 statutory authority, 579581
website information, 598
Ehrlichiosis, 409
Emerging infections, 173174
Glucose-6-phosphate dehydrogenase
Enzyme-linked immunosorbent assay
(G6PD) deficiency, 52
(ELISA), 187188
Graft-versus-host disease, 109119
Error management, 87106
diagnosis of, 116
accident analysis, 8890
causal analysis, 101105
patients at risk for, 111114
chart audit, 90
pathogenesis of, 114116
event reporting, 9093
prevention of, 117119
observation/audit, 88
therapy for, 116
simulation, 90
Errors, laboratory, 172173
Errors, transfusion, 55 Health history interview, 23, 8, 9, 11
classification of, 9399 direct questioning regarding HIV
costs of, 532533 risk, 16
reporting to FDA, 586 Hearsay, 1920
Erythropoietin administration, 457 Hemolysis:
External controls, 188190 extravascular, 53
Extravascular hemolysis, 53 intravascular, 53
non-immune, 5152
Fatalities, reporting of, to FDA, 586 Hemolytic reactions, 4768
Febrile, nonhemolytic transfusion reac- Hemophilia, persons with, 251,
tions, 7274, 472473 439440, 479
First-time donors, 6 Hemovigilance, 160
Flaviviruses, 401403 Hepatocellular carcinoma, 279
Fluorescent treponemal antibody absorp- Hepatitis, 279289
tion test (FTA-ABS), 211212 A, 21, 169, 175
610 Index
[Hepatitis] [Human immunodeficiency virus (HIV)]

B core antibody (anti-HBc), 162163, health history screening for risk,
167, 186, 280281 1420
B immune globulin (HBIG), 21 and hemophilia patients, 251252,
B surface antigen, 208209 479
B virus, 167, 279282, 431432 HIV-2, 204205, 256, 264265, 430
tests for, 208210 lookback for, 261262
close contact with, 20, 22 nucleic acid testing for, 268269
course of, 285 proteins of, 254255
C virus, 167169, 282285, 431433 p24 antigen, 267268, 429, 530, 536
lookback for, 593594 tests for, 205206
tests for, 209 risk factors for, 1420
G virus, 169, 286287 risk of transmission, 165, 166,
and hepatocellular carcinoma, 279 259261, 429
history of, 2021 structure of, 252253
non-ABC, 285286 surveillance for, 161, 424430
nucleic acid testing for, 281282, 284 testing for, 202206, 263269
quasispecies, 284 window period for seroconversion,
risk of transmission, 167169, 281, 201, 265267, 429
431433 Human leukocyte antigen system (see
surveillance for, 161, 430433 HLA system)
TT virus, 169, 287288 Human monocytic Ehrlichiosis, 408
window period for seroconversion,
Human T-lymphotropic virus (HTLV),
201, 431432
types I and II, 169, 295304
High-risk donors, 3, 1420
and adult T-cell leukemia, 299
HLA system, 77
characterization, 295297
and graft-versus-host disease, 113114
diseases associated with HTLV-I,
and immunomodulation, 149150
298300
diseases associated with HTLV-II, 301
Homosexual behavior, 1617
donor notification and counseling, 304
Human granulocytic Ehrlichiosis, 408
epidemiology of, 297298
Human immunodeficiency virus (HIV),
genome, 296
251271
and CD4+ lymphocytes, 257259 and HTLV-I-associated myelopathy,
and confidential unit exclusion, 299300
2930, 262263 risk of transmission by blood,
direct questioning regarding, 16 301302, 433434
and disease, 257259 testing for, 206208, 302304
donation by infected donors, 17 and tropical spastic paraparesis,
donor notification and counseling, 299300
269270 Hypotensive platelet reactions, 8182
genes of, 254255 Hypothermia, 80
genetic variation in, 253255
geographic distribution of, 256 IgA deficiency, 7576
group O, 15, 19, 256, 264265, 270, Immunoblot assays, 199
429430 Immunofluorescence assays, 198199
Index 611
Immunomodulation, 139154 Isovolemic hemodilution, 454

and bacterial infection, 143147 cost effectiveness of, 528
and cancer recurrence, 140143
clinical effects, 140 Leishmaniasis, 437438
mechanisms of, 150153 Leukocyte counting, 471472
and recurrent spontaneous abortion, Leukoreduction, 463475
148149 by apheresis, 470471
Incarceration, history of, 15 and bacterial contamination,
Incentives for donation, 47 237238, 469
Infections, emerging, 173174 by centrifugation, 463464
Infectious disease risks, 159177 and CMV, 325328, 474475
Infectious disease testing, 185215 and febrile nonhemolytic transfusion
algorithms for, 193194, 202215 reactions, 73, 472473
analytes, 186 by filtration, 465470
confirmatory testing, 195200 and graft-versus-host disease, 118119
cost effectiveness of, 530531 and HLA immunization, 473474
EIA/ELISA, 187188 indications for, 472475
EIA neutralization tests, 199200 methods for, 463471
for hepatitis B, 208210 timing of, 467469
for hepatitis C, 209 Lookback, 3335
for HIV-1, 202206, 263269 for HCV, 593594
for HIV-2, 202205, 264 for HIV, 261262
for HTLV-I/II, 206208, 302304 and nucleic acid testing, 215
immunoblot assays, 199 Lyme disease, 403405, 415, 418
immunofluorescence assays, 198199 Lymphocyte infusions and graft-versus-
methods for, 187188 host disease, 119
nucleic acid testing, 200201,
590592
for p24 antigen, 205206 Magnet effect, 7, 531
quality control for, 190191 Malaria, 355366
sensitivity of, 191193 agents, 356358
specificity of, 191193 clinical presentation, 360361
supplemental testing, 195200 deferral criteria, 2223, 361362
for syphilis, 211212 diagnosis of, 366
Western blot assays, 198 inactivating agents, 363
Institute of Medicine (IOM) report, 423, risk of transmission, 170, 438439
587 testing for, 362365
Interview, donor, 8, 9, 11, transmission by blood, 358360
Intraoperative blood recovery, 454455 Male-to-male sex, 1617
cost effectiveness of, 529530 Malignancy, history of, 26
Intravascular hemolysis, 53 Market withdrawals, 3132
Intravenous drug use, 18
Iron overload, 79 Non-replacement fees, 6
Irradiation, 117118 Normovolemic hemodilution, 454
ISO 9000, 517518 cost effectiveness of, 528
612 Index
Nucleic acid testing (NAT), 168, Questionnaire, donor, 23, 8, 9

200201, 212214
cost effectiveness of, 530 Rapid plasma reagin test (RPR), 211
and FDA, 590592 Rare Donor Registry, 76
for HCV, 212214
Reactions:
for HIV, 212214, 268269
acute hemolytic, 5260
and pooling of specimens, 193,
anaphylactic, 7476
212213
circulatory overload, 7879
delayed hemolytic, 6064
p24 antigen, 267268
febrile, nonhemolytic, 7272,
cost effectiveness of testing for, 530,
472473
536
hemolytic, 4768
Paid donors, 5
hypotensive, 8182
and hepatitis risk, 5, 185
laboratory evaluation of, 6264
Parvovirus B19, 173, 176, 483
pseudo-hemolytic, 5354
Pasteurization, 480481
prevention of, 6768
Perfluorocarbons, 546551
Plasmodia (see Malaria) red eye syndrome, 8283
Plateletpheresis, paid donors, 5 treatment of nonhemolytic, 82
Polymerase chain reaction (see Nucleic urticarial, 74
acid testing) Recalls, 3233
Pooled plasma products, 175176 Recombinant immunoblot assay
Pooled specimens for nucleic acid test- (RIBA), 209
ing, 193, 212213 Recruitment of donors, 6
Positive predictive value, 192193 Recurrent spontaneous abortion, and im-
Postdonation information, 4, 3132 munomodulation, 148149
Postoperative autologous blood recov- Red blood cell substitutes, 543564
ery, 455 applications of, 563
cost effectiveness of, 529530, 535 bovine hemoglobin-based, 557558
Preoperative autologous blood donation, characteristics, 543545
451454 human hemoglobin-based, 551557
cost effectiveness of, 523528 liposome-encapsulated hemoglobin,
Preoperative hemodilution, 454 559561
Pretransfusion testing, 6467 perfluorocarbons, 546551
Primers for NAT testing, 201 recombinant hemoglobin-based,
Prison, donors in, 4 558559
Red eye syndrome, 8283
Quality, 497519 Repeat donors, 7
Quality-adjusted life-year (QALY), 522 Retrovirus Epidemiology Donor Study
Quality control for infectious disease (REDS), 3, 7, 1011, 165, 266,
testing, 190191 429, 595
Quality System Essentials, 517518, Risks, infectious disease, 159177
573574 bacterial contamination, 171172,
Quality systems, 497519 221222, 435436, 533
Index 613
[Risks, infectious disease] Surveillance for infectious diseases,

Chagas disease, 170, 173174, 160161, 423440
375378, 437 for babesiosis, 436
Creutzfeldt-Jakob disease, 170171, for bacterial contamination, 435436
340343, 434435 for Chagas disease, 437
estimating, 163 for CJD, 434435
hepatitis, 167169, 281, 431433 for hepatitis, 430433
HIV, 165, 166, 259261, 429 for HIV/AIDS, 424430
HTLV, 169170, 301302, 433434 for HTLV I/II, 433434
malaria, 170, 438439 for malaria, 438439
reduction of, 176177 Survival after transfusion, 174175
Rocky Mountain spotted fever, Syphilis:
406407 tests for, 211212
transmission of, 240241
Scrub typhus, 408
Sepsis, 171172, 223226
Tattoos, 21
Serious Hazards of Transfusion pro-
Testing for transmissible agents (see
gram, 160
Infectious disease testing)
Seroconversion:
Thalassemia, chronic transfusion for,
for hepatitis, 201, 431432
79
for HIV, 201, 265267, 429
Tickborne infections, 399418, 402
and risk behavior, 595
Ticks, life cycle, 400
Serologic tests for syphilis, 211212
Total process control, 498510
Sex with another male, 1617
Shelf life and bacterial contamination, Transcription-mediated amplification
235236 (TMA), 201, 213214
Sickle cell anemia, chronic transfusion Transfusion reactions, 49 (see also
for, 79 Reactions)
Solvent/detergent plasma, 175176 Transfusion-related acute lung injury
Solvent/detergent viral inactivation, (TRALI), 125134
481483 clinical presentation of, 125126
cost effectiveness of, 531532 complications of, 126127
Spontaneous abortion, recurrent, im- diagnosis of, 127, 132133
munomodulation and, 148149 implicated components, 129132
Standard operating procedures, incidence of, 128129
504505 mechanism of, 129130
Standards, professional, 569577 prevention of, 133134
Storage temperature and bacterial con- treatment of, 133
tamination, 236237 Transmissible spongiform encephalo-
Sub-Saharan Africa, 18 pathies (TSEs), 335349 (see
Supplemental testing for infectious also Creutzfeldt-Jakob disease)
disease markers, 195200 Treponema pallidum:
Surrogate markers for hepatitis, 162, testing for, 211212
210211, 283 transmission of, 240241
614 Index
Trial to Reduce Alloimmunization Viral Hepatitis Surveillance Program,

to Platelets (TRAP), 77, 430431
473474 Volunteer donors, 5
Tropical spastic paraparesis, 299300
Trypanosoma cruzi (see Chagas disease) Western blot testing, 198, 202208,
TT virus, 169, 287288 264, 302304
for HIV-1, 202205
Urticaria, 74 for HTLV I/II, 207208
Utilization review, 448449 for p24 antigen, 205206
Window period, 2
Vaccination, deferral for, 26 for hepatitis, 201, 431432
Validation, 501504 for HIV, 201, 265267, 429
Viral inactivation, 175176, 252, 433, and risk behavior, 595
479489
cost effectiveness of, 531532 Yersinia enterocolitica, 171172,
of platelets, 485487 224225, 436
of red blood cells, 487488 and leukoreduction, 469

Blood Safety and SURVEILLANCE PDF

Uploaded by

Copyright:

Available Formats

Blood Safety and SURVEILLANCE PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Blood Safety and SURVEILLANCE PDF

Uploaded by

Copyright:

Available Formats

BLOOD

Marcel Dekker, Inc. New York Basel

Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved.

This book is printed on acid-free paper.

Eastern Hemisphere Distribution

World Wide Web

Copyright 2001 by Marcel Dekker, Inc. All Rights Reserved.

Current printing (last digit):

PRINTED IN THE UNITED STATES OF AMERICA

The blood contains the soul . . .

Needless to say, these experimentshowever apparently successful in the

Although other blood transfusion experiments were done in the eighteenth

Claude Lenfant, M.D.

transfusion of incompatible blood. The following section describes various infec-

Foreword Claude Lenfant iii

1. Impact of Blood Donor Screening Procedures on

II. Immunological Complications

2. Hemolytic Transfusion Reactions 47

7. Clinical Effects of the Immunomodulation

III. Infectious Complications

8. Overview of Infectious Disease 159

10. Bacterial Contamination 221

11. HIV and Blood Transfusion 251

15. Creutzfeldt-Jakob Disease, Other Human Transmissible

17. Tickborne Infections 399

IV. Strategies to Reduce Risk

18. Surveillance for Transfusion-Transmitted

Kenneth C. Anderson, M.D. Department of Adult Oncology, Dana-Farber

James P. AuBuchon, M.D. Professor of Pathology and Medicine, Dartmouth-

James B. Battles, Ph.D. Associate Professor, Office of Medical Education,

Ehud Ben-Hur, Ph.D. Consultant in Photomedicine, New York, New York

Lucia M. Berte, M.A., MT(ASCP)SBB, DLM, CQA(ASQ),CQMgr. Quality

Morris A. Blajchman, M.D., F.R.C.P.(C.) Professor, Department of Pathology

Jos O. Bordin, M.D., Ph.D. Associate Professor, Hematology and Transfusion

Mark E. Brecher, M.D. Director, Transfusion Medicine Service, and Professor,

Ritchard G. Cable, M.D. Medical Director, American Red Cross Blood

Mary E. Chamberland, M.D., M.P.H. Assistant Director for Blood Safety,

Roger Y. Dodd, Ph.D. Head, Transmissible Diseases Department, Holland

Lawrence T. Goodnough, M.D. Professor of Medicine and Pathology, Depart-

Christopher J. Gresens, M.D. Associate Medical Director, Sacramento Medi-

Mary Gustafson, M.S., MT(ASCP)SBB Director, Division of Blood Applica-

Bernard Horowitz, Ph.D. VITEX (V.I. Technologies, Inc.), New York,

Harold S. Kaplan, M.D. Professor of Clinical Pathology and Director, Trans-

Rima F. Khabbaz, M.D. Deputy Director, Division of Viral and Rickettsial

Elizabeth J. Kicklighter, M.D. Warren G. Magnuson Clinical Center, National

Harvey G. Klein, M.D. Chief, Department of Transfusion Medicine, Warren G.

Steven Kleinman, M.D. Clinical Professor, Department of Pathology, Univer-

Anne B. McDonald, M.D. Fellow, Transfusion Medicine, New England Dea-

Jay E. Menitove, M.D. Executive Director and Medical Director, Community

David E. Nevalainen, Ph.D., MT(ASCP), CQA(ASQ) Quality Consultant,

Mark A. Popovsky, M.D. Executive Director and Chief Medical Officer,

Alfred M. Prince, M.D. Member/Head, Laboratory of Virology, Lindsley F.

Hillary V. Schaeffler Director, Standards and International Affairs, American

Laura L. Spinelli, M.D. Fellow in Hematopathology, Department of Pathology,

Christopher P. Stowell, M.D., Ph.D. Director, Blood Transfusion Service,

Zbigniew M. Szczepiorkowski, M.D., Ph.D. Associate Director, Blood Trans-

Gary E. Tegtmeier, Ph.D. Director, Viral Testing Laboratories, Community

Jonathan Trouern-Trend Epidemiology and Surveillance Program, American

Jay E. Valinsky, Ph.D. Vice President, Laboratory and Technical Services,