Immunohematology

Journal of Blood Group Serology and Education Celebrating

2 Ye5ars Volume 25, Number 3, 2009

Immunohematology Journal of Blood Group Serology and Education Volume 25, Number 3, 2009 Contents

Letter to the Readers 89 Introduction to Vol. 25, No. 3, of Immunohematology J.P. AuBuchon

Invited Review: Milestones Series 90 Pioneers of blood group serology in the United States: 1950–1990 S.R. Pierce

Review 95 MNS blood group system: a review M.E. Reid

Review 102 Quality activities associated with hospital tissue services C.M. Hillberry and A.J. Schlueter

Review 107 Logistical aspects of human surgical tissue management in a hospital setting B.M. Alden and A.J. Schlueter

Review 112 Lewis blood group system review M.R. Combs

Review 119 Recognition and management of antibodies to human platelet antigens in platelet transfusion–refractory patients R.R. Vassallo

Review 125 Detection and identification of platelet antibodies and antigens in the clinical laboratory B.R. Curtis and J.G. McFarland

Review 136 Investigating the possibility of drug-dependent platelet antibodies J.P. AuBuchon and M.F. Leach

141 Announcements

142 Advertisements

146 Instructions for Authors Editors-in-Chief Managing Editor Sandra Nance, MS, MT(ASCP)SBB Cynthia Flickinger, MT(ASCP)SBB Philadelphia, Pennsylvania Philadelphia, Pennsylvania

Connie M. Westhoff, PhD, MT(ASCP)SBB Technical Editors Philadelphia, Pennsylvania Christine Lomas-Francis, MSc New York City, New York Senior Medical Editor Geralyn M. Meny, MD Dawn M. Rumsey, ART(CSMLT) Philadelphia, Pennsylvania Glen Allen, Virginia

Associate Medical Editors David Moolten, MD Ralph R. Vassallo, MD Philadelphia, Pennsylvania Philadelphia, Pennsylvania

Editorial Board

Patricia Arndt, MT(ASCP)SBB Brenda J. Grossman, MD Joyce Poole, FIBMS Pomona, California St. Louis, Missouri Bristol, United Kingdom

James P. AuBuchon, MD W. John Judd, FIBMS, MIBiol Mark Popovsky, MD Lebanon, New Hampshire Ann Arbor, Michigan Braintree, Massachusetts

Martha R. Combs, MT(ASCP)SBB Christine Lomas-Francis, MSc Marion E. Reid, PhD, FIBMS Durham, North Carolina New York City, New York New York City, New York

Geoffrey Daniels, PhD Gary Moroff, PhD S. Gerald Sandler, MD Bristol, United Kingdom Rockville, Maryland Washington, District of Columbia

Anne F. Eder, MD John J. Moulds, MT(ASCP)SBB Jill R. Storry, PhD Washington, District of Columbia Shreveport, Louisiana Lund, Sweden

George Garratty, PhD, FRCPath Paul M. Ness, MD David F. Stroncek, MD Pomona, California Baltimore, Maryland Bethesda, Maryland

Emeritus Editorial Board Delores Mallory, MT(ASCP)SBB Supply, North Carolina

Editorial Assistant Production Assistant Judith Abrams Marge Manigly Proofreader Copy Editor Lucy Oppenheim Electronic Publisher Mary L. Tod Wilson Tang

Immunohematology is published quarterly (March, June, September, and December) by the American Red Cross, National Headquarters, Washington, DC 20006.

Immunohematology is indexed and included in Index Medicus and MEDLINE on the MEDLARS system. The contents are also cited in the EBASE/Excerpta Medica and Elsevier BIOBASE/Current Awareness in Biological Sciences (CABS) databases.

The subscription price is $40.00 (U.S.) and $50.00 (foreign) per year.

Subscriptions, Change of Address, and Extra Copies: Immunohematology, P.O. BOX 40325, Philadelphia, PA 19106 Or call (215) 451-4902 Web site: www.redcross.org/immunohematology

Introduction to Vol. 25, No. 3, of Immunohematology

J.P. AuBuchon

The increased breadth and depth of the field of transfu- Thus, even if you are not directly involved in testing sion medicine is evident in this issue of Immunohematol- for platelet or HLA antibodies or defining platelet antigens’ ogy, and I am delighted and honored to have been asked to molecular or genomic bases, the platelet focus of this issue serve as its guest editor. will provide an exciting window on a world that a few years Those with accumulations of gray hair, myself includ- ago was far more limited in its understanding and capabili- ed, recall when blood banking meant primarily dealing with ties. So whether these articles will call you to rapt attention RBCs and their antibodies. Platelets presented collection, or offer an interesting insight while you sit back and relax, storage, and logistic problems, but when their transfusion the authors of the articles and I hope you will enjoy them. did not have the desired effect, we usually did not have much to offer the patient or the clinician. Indeed, this jour- James P. AuBuchon, MD nal was begun to facilitate knowledge transfer that would President and CEO, Puget Sound Blood Center be useful to those practicing RBC serology at the bench, Professor of Medicine and of Laboratory Medicine as immunohematology was regarded as synonymous with University of Washington RBC antibody identification. (An early meeting of the edito- 921 Terry Avenue rial board even debated whether to branch out into articles Seattle, WA 98104 about platelets.) How the times have changed! This issue covers the spectrum of what a clinical laboratory scientist may encounter today in a transfusion service, from tissue banking (now an integral part of some transfusion services) to molecular characterizations of RBC and platelet antigens. Functional competency in transfusion medicine today requires a wide range of knowledge that extends from the genomic to the clinical! This scope is evident in the articles on platelets, their antigens and antibodies, and their use in hemotherapy assembled for this issue. Refractoriness to platelet transfusion, and, indeed, the thrombocytopenia itself, may have an immunologic cause, and articles in this issue lay out strategies for identifying and coping with antibody specificities directed at HLA, platelet-specific, and drug-induced determinants. Knowledge of the molecular and genomic bases of platelet antigens is no longer of research interest only. Just as genomic typing has entered into practical RBC transfusion practice, these abilities offer the opportunity for us to become more specific in our attempts to match platelet donor and recipient and to understand the basis of the immunogenic discrepancy in the first place. Although not all of the methods described in these articles are likely to be found in every laboratory in the near future, information gained from their broader application will undoubtedly expand all of our capabilities.

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 89 Invited Review: Milestones in Immunohematology

Pioneers of blood group serology in the United States: 1950–1990

S.R. Pierce

The human blood groups were discovered in 1900 by in 1943 and was stationed at the newly commissioned U.S. Karl Landsteiner.1 The early decades of the 20th century Naval Hospital at Long Beach, California. It had all the lat- saw prodigious work on blood groups and their applica- est equipment, including a centrifuge that could be used to tion to transfusion, anthropology, forensics, genetics, and separate blood samples quickly into serum and RBCs for parentage testing. The discovery of Rh (reported by Levine ABO typing. Crossmatching was not yet routinely done; do- and Stetson2 in 1939 and in 1940 by Landsteiner and Wie- nors were merely chosen to match the patient’s ABO type. ner3, who may actually have found LW instead) and the After leaving the Navy, Shirley obtained a master’s degree subsequent application of the antiglobulin test by Coombs, in public health from Columbia University and was hired as Mourant, and Race4 in 1945 launched a new and exciting supervisor (later technical director) of the Blood Center of era in blood group serology. During the last half of the 20th Mt. Sinai Hospital in Chicago in 1951. In 1956, Shirley read century, many new blood group antigens were described, of a new certification planned by the AABB and the Ameri- largely through research done in transfusion services and can Society of Clinical Pathologists (ASCP). She was the first immunohematology reference laboratories. In the United to respond. Prerequisites were a master’s degree or higher States, the efforts were mainly those of medical technolo- and 3 years’ blood bank experience. The examination pro- gists. Tube shaking was the order of the day. Before the cess included written and practical components (first given explosion of viral testing and DNA assays, blood group in January 1957), an oral interview, and a written thesis. In serology dominated the scientific sessions of blood bank 1958, Shirley was the first to obtain this “Specialist in Blood meetings, such as the annual meeting of the American As- Banking” (SBB) certification.5,6 sociation of Blood Banks (AABB). The leaders of this era Shirley organized the Pre-Convention Seminar for of serology were widely recognized and respected, often technologists, which became a staple at the AABB Annual referred to simply by their first names by blood bank tech- Meetings; the first, held in 1959, addressed antibody de- nologists eager to keep abreast of the latest blood group tection and identification. With assistance from Marjory happenings. These “stars” were glad to oblige, giving Stroup, she conceived and organized an invitational meet- numerous workshops and lectures and hours of individual ing of leading immunohematology reference laboratory consultation on serologic problems. Most major cities experts; begun in 1968, and later renamed the Invitational had Antibody Clubs to discuss serologic problems and hear Conference of Investigative Immunohematologists (ICII), it guest lecturers. Blood bank technologists from all the ar- continues today (Table 1). With Drs. Douglas Huestis and ea’s hospitals flocked to these sessions—on their own time, Joseph Bove, she wrote the text Practical Blood Transfu- without expecting overtime pay. Workshops at regional and sion; the first edition appeared in 1969.7 national meetings drew large crowds. Times have changed, and the emphasis once placed on blood group serology has declined. Although still impor- Table 1. Attendees at the First Invitational Reference Laboratory Conference (later ICII), Chicago, 1968 tant in transfusion medicine and related fields, research on blood groups has shifted to specialized laboratories in Kay Beattie Gwen Jones which the function and molecular structure of blood group Oscar Behzad Delores McGuire antigens are studied. In both routine transfusion practice Charmaine Brice Grace Neitzer and cutting-edge research, a tremendous debt is owed to Sue Britten Tom Sasaki those serologists who laid the groundwork for our knowl- Shirley Busch Laima Sausais edge. Still, the adage “out of sight, out of mind” too often Mary Kay Curtin Polly Schmidt applies. As time moves on, these pioneers are often forgot- ten, unknown to newer workers. It is thus worthwhile to Dorothy Day Arden Stanbury recall some of those serologists whose work in the United Mary Lou Garris Majory Stroup States brought us to where we are today in the science of Bob Gilbert Jane Swanson immunohematology. Some made dramatic “first” discover- Ruth Guy Martha Thomas ies, while the contributions of many others were more in- Betty Hatcher Margaret Thompson cremental and cumulative. All inspired and educated us. Peter Issitt Mary Walker Shirley Busch received her degree in bacteriology in 1939 from the University of Illinois. She joined the Navy Elsa Jahn

90 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Pioneers of blood group serology in the U.S.

Kay Beattie began her career in the mid-1940s, setting Jane was the first recipient of the AABB’s Ivor Dunsford up the blood bank at Detroit’s Children’s Hospital in 1945. Memorial Award, initiated in 1968 for “research or devel- In 1961, she moved to the Michigan Community Blood Cen- opments of technics in immunohematology.” The Dunsford ter (acquired in 1975 by the American Red Cross), where she Award was the highest award given specifically for immuno- served as co-technical director with Grace Neitzer (promi- hematology, and the list of recipients constitutes an honor nent in the AABB’s Inspection and Accreditation program roll of blood group serologists. The AABB discontinued the and AABB president, 1985). Kay’s serologic career included award in 2004 (Table 2). many interesting cases, perhaps none more so than a blood Mary N. “Polly” Crawford was a physician, trained in donor found by Delores McGuire (later Mallory), then at pediatric medicine. When her husband, Pearson Cummin, Detroit Receiving Hospital. The donor presented as an died, she established the Pearson C. Cummin Memorial ABO typing problem, seemingly group A with no anti-B. He proved to be a dispermic mosaic with a small population of Table 2. Ivor Dunsford Memorial Award honorees B cells. The case was prophetic, for Kay became known for Year Honoree her work with weak subgroups of A and B. She also showed 1968 Jane Swanson that anti-M could often be enhanced by acidification of the serum, and she reported Gu, a weak expression of the G an- 1969 Grace Neitzer tigen. Kay died in 2006.8 1970 Shirley Busch Delores McGuire Mallory left Detroit to attend SBB 1971 Corinne Monroe school at the Irwin Memorial Blood Center in San Fran- 1972 Kay Beattie cisco (now Blood Centers of the Pacific). There, under the 1973 Marjory Stroup guidance of Bernice Hemphill, one of the founders of the 1974 Peter Issitt AABB, Delores became extensively involved in AABB activities, serving for many years on the Technical Workshop and 1975 W. L. Marsh other committees and on the Board of Directors. Delores 1976 Delores McGuire Mallory became technical director at the Community Blood Center 1977 Mary McGinniss in Dayton, Ohio (Ruth Mougey and this author were the first 1978 George Garratty students in her SBB program). Later, she became director 1979 Jane Haber of the American Red Cross National Reference Laboratory. 1980 Malcolm Beck She continued her extensive role in blood bank education and founded the journal Immunohematology, serving as its 1981 David Hatcher and Betty Francis Hatcher editor for many years. (Immunohematology evolved from 1982 John Judd a newsletter, the Red Cell Free Press, created by Sandy 1983 John Moulds Ellisor, Marion Reid, and Helen Glidden.) She reported 1984 Margaret Hanson a anti-Rb and anti-Gil and was editor-in-chief of Immuno- 1985 Jane Wilson hematology Methods and Procedures.9 1986 Jean Harris Jane Swanson graduated as a medical technologist in 1947, and in 1950 began work at the Minneapolis War Me- 1987 Margaret Treacy morial Blood Bank, joining Eleanor Amberg and Dr. Albin 1988 Ann Hoppe Matson. Matson’s PhD thesis had shown that unlike most 1989 Sandra Ellisor Native Americans, who tested as predominately group O, 1990 John Case members of the Blackfoot Nation had a high frequency of 1991 Susan Moore Steane the A gene. This led to anthropologic expeditions through- 1992 Margaret Wallace out Central and South America, collecting blood samples that were sent back for Jane to type. Later, Jane moved to 1993 Elizabeth Sebring the University of Minnesota, where she was the only tech- 1994 Susan Rolih nologist in the immunohematology reference laboratory. 1995 Marion Reid Her work at both facilities was prodigious. She discovered 1996 Marilyn Moulds k a a a a anti-P , -Co , -Do , -Gy , and -Mt . She found anti-Ch and 1997 Geoff Daniels anti-Kna and did extensive work with those nebulous 1998 Joann Moulds antibodies then termed “high titer, low avidity,” on hemag- glutinating HLA antibodies, and on early samples of anti- 1999 Virginia Vengelen-Tyler LW. She pioneered the use of chloroquine diphosphate to 2000 Joyce Poole strip HLA antigens from RBCs. Among those who trained 2001 Christine Lomas-Francis or worked with Jane were Ann Hoppe, longtime blood bank 2002 Steven Pierce expert for the Food and Drug Administration, and John 2003 Nancy Heddle Moulds.

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 91 S.R. Pierce

Laboratory in her home. Staffed only by herself, the labora- Vancouver to work for the Canadian Red Cross Transfusion tory nonetheless gained AABB accreditation as an approved Centre in 1955. There she met another British transplant, reference laboratory. Short on both space and reagents, Bill Pollack, who recruited her to join Ortho as a research Polly turned to micromethods, reviving Bruce Chown’s technologist. (Pollack’s work led to Ortho’s development of capillary tube techniques. To save electricity (and also to RhoGAM, Rh immune globulin.) Margaret later moved into be able to perform tests in her hotel room while attending education full time, directing a lecture and laboratory pro- meetings), Polly acquired a table-mounted, hand-cranked gram at Ortho that for years drew participants from around centrifuge of which she was quite proud. She had the rare the world. Margaret wrote and edited the newsletter Blood- pleasure, only dreamed of by most blood groupers, of find- lines. Ortho also developed a series of monographs, written ing herself to be of a rare blood type: Lu(a–b–), the first by Marjory and Margaret, which could be found in virtually such person known. Despite her ever-worsening deafness, every hospital transfusion service.12–14 Polly attended many blood bank meetings, sitting near Knickerbocker Biologicals, begun by Dr. Amos Cahan the front so she could read the speakers’ lips. Polly died in to market blood grouping reagents, also had a reference 1997.10 laboratory, the staff of which included Jim Jack, Allen Gelb, No account of U.S. immunohematology would be com- David Hatcher, and Julia Mann. Among the samples of in- plete without paying tribute to the many contributions of terest from the Knickerbocker reference laboratory were those working for commercial reagent companies. Dade Re- the first examples of anti-Xga and anti-Dib. In the 1950s, agents Company, founded by John Elliott in 1949, was one Knickerbocker created the first commercial red cell panel— of the first to offer blood grouping reagents commercially in an eight-cell set—and later antibody detection, or “screen- the United States.11 There have been many such companies ing” cells. The panels were initially a hard sell. Most trans- during the years, almost all offering “reference laboratory” fusion services considered them a research tool. They saw services to resolve serologic problems—often at no direct little reason to identify an antibody; donors were typically cost to customers. The large volumes of samples seen by crossmatched until a compatible unit was found. many of these laboratories meant they were well positioned Gelb, Hatcher, Jack, and others left Knickerbocker to to recognize new or unusual antibodies and blood groups. form their own company, Spectra Biologicals. One of Spec- They also developed many new procedures and reagents. tra’s major products was Spectrazyme, an enzyme additive One of the most prominent companies was Johnson & blend of ficin, papain, and bromelin. Spectra was sold to Johnson’s Ortho Clinical Diagnostics (now Ortho Diagnos- Becton-Dickinson in 1966. Knickerbocker, meanwhile, had tic Systems). Philip Levine, who became director of the bio- been sold to Pfizer Diagnostics; Julia Mann became the pri- logical division of the Ortho Research Foundation in 1944, mary technologist of the Pfizer reference laboratory. can be credited with steering Ortho into the manufacture With his wife Betty (B.J.) Francis, Dave Hatcher began of blood grouping reagents. Levine began soliciting speci- another company in 1970, Gamma Biologicals. In 1957, mens containing Rh antibodies (which he had first reported the Hatchers had helped create the first Antibody Club, in in 1939).2 The samples that poured in included many other Houston. When Dr. Fred Allen, Jr., of Boston was invited to unidentified antibodies as well; thus began the “consulta- speak at the Houston Antibody Club, he found the concept a tion laboratory,” later renamed the Philip Levine Laborato- capital idea and extolled it in an article in Transfusion.15 Anti- ries in his honor. Levine and the Ortho staff were involved body Clubs soon spread across the country. At Gamma Bio- in many key immunohematology discoveries, including the logicals, the Hatchers hired John and Marilyn Moulds from first anti-s, -k, -Tja (-PP1Pk), -Fyb, and -Dia. (The Levine Lab- Minneapolis. John was already active in immunohema- oratories closed in the early 1980s.) The Ortho Blood Group tology. Minneapolis War Memorial Blood Bank housed the Consultation Service was also greatly involved with blood AABB’s Rare Donor Program (managed for many years by bank education, efforts led primarily by Marjory Stroup Kathy Skradski), and John had proved diligent in procuring and Margaret Treacy. units of rare blood. In the early 1970s, in conjunction with Marjory Stroup received her medical technology degree ICII, John began a specimen exchange program among in 1947 and worked until 1951 at the University of Kansas reference laboratories that came to be called SCARF— Hospital. She recalls an Ortho salesman coming in to dem- Serum, Cell and Rare Fluids Exchange. Participating labo- onstrate their new product: antiglobulin reagent. To do so, ratories agreed to send suitable volumes of rare samples to he brought a sample with a “new” blood group antibody, each of the other members each year. It was a significant anti-K. The results convinced her of the significance of the program for reference laboratories, which could acquire antiglobulin test. In 1953, she joined Ortho, working first in sufficient exotic specimens to resolve many problems lo- Quality Assurance and manufacturing areas before moving cally and more quickly. Similar programs were soon begun into the Consultation Laboratory. She demonstrated that among the American Red Cross Reference Laboratories Sutter (Jsa) was in the Kell blood group system. (“Transferase”) and Blood Services laboratories (“Serum Margaret Treacy worked with Dr. Fred Stratton in Eng- Exchange”). Gamma’s reference laboratory, directed first by land (assisting in his work defining Du) before moving to John, later by Marilyn, was involved with such discoveries

92 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Pioneers of blood group serology in the U.S. as anti-McCa, -Tca and -Tcb, -Aub, and many other antibod- anemias. He joined the American Red Cross in Los An- ies to low-frequency antigens. Marilyn oversaw the Gamma geles in 1978. Sandy Nance worked for him for 10 years; Tutorial Program, an extensive educational series for tech- during this time, they pioneered the use of the monocyte nologists. monolayer assay and flow cytometry as immunohematologic Gamma Biologicals was purchased by Immucor in 1998. tools and described polyethylene glycol as an antibody Immucor had its own reference laboratory directed by Su- enhancement medium. san Rolih (previously with Biological Corporation of Ameri- Malcolm Beck worked at the North East London Blood ca, BCA). But by the 1990s market conditions had changed. Transfusion Service in Brentwood, under the direction of Thanks in large part to SCARF, local reference laboratories W. Laurence “Laurie” Marsh. When Laurie turned down a were able to handle most problem cases. The Immucor and job in America, he recommended Malcolm. Unluckily, after Gamma reference laboratories, the last of those sponsored moving to the United States, that job failed to materialize, by U.S. reagent manufacturers, were closed in 2004. and Malcolm found himself working at a hospital in Kala- By the end of World War II, the British had a well- mazoo, Michigan. When he published a paper about T- established blood donor program, augmented by several re- activation, it caught the attention of Dr. Richard Rosenfield nowned blood group serologists. None were better known at New York’s Mt. Sinai Medical Center, who arranged a po- than Rob Race and Ruth Sanger, whose Blood Group Unit sition for Malcolm at the New York Blood Center. New York at the Lister Institute (later headed by Patricia Tippett) was City, however, proved not to be to Malcolm’s taste, and he arguably the ultimate reference laboratory of the world. moved to the University of Michigan at Ann Arbor. In 1972, The British system turned out numerous blood group se- he became technical director of the Community Blood Cen- rologists, several of whom moved to the United States and ter in Kansas City. Malcolm became especially known for were highly influential in U.S. blood banking. Indeed, dur- his work with polyagglutination, and with snail agglutinins, ing the 1970s and 1980s, blood bank meetings were often leading to the development (with Fred Plapp, Lyle Sinor, dominated by speakers with distinct British accents. A core Jane Rachel, and the rest of the team at Kansas City) of the cadre of British expatriates became unofficially known as solid-phase technology used for antibody detection. the “British Mafia,” a name coined by Dr. Paul Schmidt. Laurie Marsh was eventually persuaded to move to the “British Nights” at the AABB annual meeting offered a time United States, replacing Peter Issitt at the New York Blood for the expatriates to gather among themselves; after an Center, where he stayed for the remainder of his career. A evening’s dinner and festivities, the Brits usually descended Welshman, Laurie was already a well-known serologist in upon the “hospitality suites” that then dominated the social England, noted for his work with cold agglutinins. He was scene at the annual meetings. There they would regale the involved in the discovery of anti-i, anti-Joa, and anti-Fy3, delighted crowds with British pub and rugger songs into the -Fy4, and -Fy5. With Ragnhild (Rock) Øyen and others, he wee hours. was particularly noted for his work with Lutheran and Kell. Peter Issitt had worked at London’s Hammersmith It was said that no matter what topic Laurie was asked to Hospital (in Professor Sir John Dacie’s Haematology De- lecture on, he would somehow turn the presentation to the partment16) and at St. Mary’s Hospital (with Professor Pat- Kell system, with which he was associated from serology to rick Mollison) before coming to the New York Blood Center isolating the glycoprotein to mapping and cloning its gene.20 in 1964. In 1968, he joined Spectra Biologicals, his primary Laurie died in 1997.21 assignment being to produce a book on blood groups. The John Judd, who replaced Malcolm at Brentwood, did first edition of Applied Blood Group Serology appeared 2 so again in Ann Arbor. He developed a reputation for chal- years later.17 Peter later worked at Cincinnati, Miami, and lenging some of the common U.S. serologic practices—such Durham, and became especially noted for his work with Rh as routine direct antiglobulin tests and autologous controls and MNS. He described anti-Wrb and anti-Duclos, and was for all transfusion candidates—policies not followed in Eng- known for his work with autoantibodies, especially those land. He fought the use of “pre-warm” testing to circumvent with mimicking specificities.18 cold agglutinins and championed the electronic crossmatch. George Garratty, also in Dacie’s department at Ham- He did extensive work on lectins and polyagglutination and mersmith Hospital, was persuaded by Peter’s letters showed (with Laurie Marsh) the use of AET to disrupt Kell from America also to move. In 1968, he joined Dr. Larry antigens on red cells. He wrote the manual Methods in Petz (who had spent 2 years in Dacie’s laboratory) in San Immunohematology, recently released in a third edition, Francisco to carry out research on complement. Finding with his name added to the title.22 that antiglobulin reagents in the United States were poor John Case had been a noteworthy blood grouper in in anticomplement activity, Garratty and Petz were drawn London before moving, in 1959, to New Zealand. In 1971, into a controversy (that lasted several years) on the impor- he joined Commonwealth Serum Laboratories in Australia tance of detecting RBC-bound complement in the study of and headed the National Blood Group Reference Labora- the hemolytic anemias and compatibility testing.19 George tories. He was then recruited to join Gamma Biologicals in became especially known for his work on immune hemolytic Houston. Officially, his duties were in regulatory affairs, in

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 93 S.R. Pierce which he argued forcefully that regulations should address 7. Huestis DW, Bove JR, Busch S. Practical blood transfusion. serologic realities rather than the whims of regulators. He 1st ed. Boston: Little, Brown & Co, 1969. always argued for high standards of performance. Serologi- 8. Mallory D. In memoriam. Katherine M. Beattie, BS, cally, he was known for his work with Rh and with antigens MT(ASCP)SBB 1922–2007. Immunohematology 2007;23:43. of low prevalence. He replaced Shirley Busch for the fourth 9. Mallory DM, ed. Immunohematology methods and pro- edition of Practical Blood Transfusion.23 cedures. Rockville, MD: American Red Cross National Debates between “the two Johns” (Judd and Case) were Reference Laboratory, 1993. a highlight of the chat rooms that appeared when AABB 10. No author given. Mary N. Crawford, MD (Polly) 1923– launched its presence on the Internet. Their good-natured, 1997. Immunohematology 1997;13:106. but often strongly worded bantering—some contrived but 11. Schmidt PJ. John Elliott and the evolution of American much of it real—sparked discussion and debate among blood banking. Transfusion 2000;40:608–12. blood bankers everywhere. Their “grammar police” argued 12. No author given. Blood group antigens and antibodies strongly for the use of correct terminology.24,25 as applied to compatibility testing. Raritan, NJ: Ortho Other notable British “imports” included Ed Steane (the Diagnostics, 1967. dynamics of antigen-antibody reactions and AABB presi- 13. No author given. Blood group antigens and antibodies dent, 1987), Margaret Nichols (monoclonal antibodies), as applied to hemolytic disease of the newborn. Rari- Marion Reid (another Brentwood alumna and winner of the tan, NJ: Ortho Diagnostics, 1968. 2006 Woman in Transfusion Award), and Christine Lomas- 14. No author given. Blood group antigens and antibodies Francis (Rh, and coauthor with Marion of the Blood Groups as applied to the ABO and Rh systems. Raritan, NJ: Or- Antigen FactsBook26). The careers of both Marion and Chris- tho Diagnostics, 1969. tine, who are both still active in immunohematology, have 15. Allen FH Jr. Editorial. On finding rare donors in the spanned the era from basic serology to molecular biology. hospital blood bank. Transfusion 1966;6:534–6. Many other U.S. serologists deserve equal commemo- 16. Garratty G. Sir John Dacie, MD, FRCP, FRCPath, FRS ration. Readers are encouraged to note, when perusing (1912–2005). Transfus Med Rev 2007;21:72–4. texts and published papers, who were the individuals who 17. Issitt PD. Applied blood group serology. Oxnard, CA: led our knowledge of blood groups to its current status, and Spectra Biologicals/Becton, Dickinson and Co, 1970. who today move that knowledge forward. Without them, 18. Garratty G, Mallory D. Symposium to honor the career much of what now seems so simple would remain mysteri- of Peter D. Issitt, PhD, FRCPath, FIBMS, FIBiol. Im- ously complex. munohematology 1999;15:1. 19. Petz LD, Garratty G. Acquired immune hemolytic ane- Acknowledgments mias. New York: Churchill Livingstone, 1980. The material included in this article is drawn largely 20. Marsh WL. Cold agglutinins to Kell: 35 years as a se- from interviews conducted by the author and by Marion rologist. In: Moores SB, ed. Progress in immunohema- Reid with many of the principals mentioned. The author tology. Arlington, VA: AABB, 1988:93–117. wishes to thank the many interviewees for their coopera- 21. Issitt PD, Garratty G. In memoriam. W. Laurence tion, assistance with this manuscript, and friendship over Marsh, PhD, FRCPath, FIBMS, FIBiol 1926–1997. Im- the past decades. munohematology 1998;14:42. 22. Judd WJ, Johnson ST, Storry JR. Judd’s methods in immu- References nohematology. 3rd ed. Bethesda, MD: AABB Press, 2008. 1. Landsteiner K. Zur Kenntnis der antifermentativen, 23. Huestis DW, Bove JR, Case J. Practical blood transfu- lytischen und agglutinierenden Wirkungen des Blutse- sion. 4th ed. Boston: Little, Brown & Co, 1988. rums und der Lymphe. Zbl Bakt 1900;27:357–62. 24. Mallory D, McGinniss M. In memoriam. John Case, 2. Levine P, Stetson R. An unusual case of intra-group ag- FIBMS, FIMLS 1926–2001. Immunohematology glutination. JAMA 1939;113:126–7. 2001;17:93. 3. Landsteiner K., Wiener AS. An agglutinable factor in 25. Ford DS. Vale John Case. Newsletter. Austral Soc Blood human blood recognized by immune sera for Rhesus Transfus Feb 2002:6. blood. Proc Soc Exp Biol Med 1940;48:223–4. 26. Reid ME, Lomas-Francis C. The blood group antigen 4. Coombs RRA, Mourant AE, Race RR. Detection of weak factsbook. 2nd ed. San Diego: Academic Press, 2004. and “incomplete” Rh agglutinins: a new test. Lancet 1945;2:15–6. Steven R. Pierce, SBB(ASCP), Retired, Kansas City, MO 5. No author given. Certification for blood bank workers. 64111. Am Assoc Blood Banks Bull 1957;10:93–4. 6. Busch S. The role of antibody screening as a safeguard in selecting blood for transfusion. Am Assoc Blood Banks Bull 1958;11:260–6.

94 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Review MNS blood group system: a review

M.E. Reid

(continued) Antigen ISBT number Prevalence The MNS blood group system is second only to the Rh blood group U MNS5 High system in its complexity. Many alloantibodies to antigens in the He MNS6 Low MNS system are not generally clinically significant although antibodies to low-prevalence and high-prevalence MNS antigens have Mia MNS7 Low caused hemolytic disease of the fetus and newborn. The MNS an- Mc MNS8 Low tigens are carried on glycophorin A (GPA), glycophorin B (GPB), Vw MNS9 Low or hybrids thereof, which arise from single-nucleotide substitution, unequal crossing over, or gene conversion between the gly- Mur MNS10 Low cophorin genes. Antigens in the MNS system are fully developed Mg MNS11 Low at birth. This review summarizes aspects of the MNS system, in- Vr MNS12 Low cluding the molecular basis of some antigens in the MNS blood e group system. Readers are referred to existing excellent reviews M MNS13 Low for background information.1–9 Throughout this document, infor- Mta MNS14 Low mation given without references can be found in the reviews listed Sta MNS15 Low previously, and the reader is referred to these reviews for refer- Ria MNS16 Low ences to original reports. Immunohematology 2009;25:95–101. Cla MNS17 Low Nya MNS18 Low Key Words: blood group, MNS, glycophorin, genes, hybrid Hut MNS19 Low glycophorins, molecular basis Hil MNS20 Low

v History M MNS21 Low fter the discovery of the ABO antigens, Landsteiner Far MNS22 Low and Levine reasoned that other variations among sD MNS23 Low Ahealthy humans existed. Thus, in 1927, they immu- Mit MNS24 Low nized rabbits with human blood and, after absorption, dis- Dantu MNS25 Low covered antibodies that detected the M or the N antigen.10 Hop MNS26 Low Because the antigens were clearly related but not exactly antithetical, they named them, respectively, after the sec- Nob MNS27 Low ond and fifth letters of the word immune. Later anti-M and Ena MNS28 High anti-N were found in human serum. In 1947, the S antigen ENKT MNS29 High was described; it was detected by an antibody in the serum ′N′ MNS30 High of a patient from Sydney, Australia. Three years later, the Or MNS31 Low antithetical antigen was described and, as was the con- DANE MNS32 Low vention at the time, given the name s. The U antigen was described in 1953 and named from “the almost universal TSEN MNS33 Low distribution” of the antigen. The other 41 antigens in the MINY MNS34 Low MNS blood group system are mostly named after the an- MUT MNS35 Low tibody maker or the person whose red blood cells (RBCs) SAT MNS36 Low 11 expressed the antigen. Table 1 lists the antigens, with their ERIK MNS37 Low ISBT numbers,12,13 and relative prevalence. In various stud- Osa MNS38 Low ies, M/N and S/s were shown to be in the same blood group system and to have linkage disequilibrium. The prevalence ENEP MNS39 High ENEH MNS40 High Table 1. Antigens of the MNS blood group system HAG MNS41 Low Antigen ISBT number Prevalence ENAV MNS42 High M MNS1 Polymorphic MARS MNS43 Low N MNS2 Polymorphic ENDA MNS44 High S MNS3 Polymorphic ENEV MNS45 High s MNS4 Polymorphic MNTD MNS46 Low

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 95 M.E. Reid of the various combinations of M, N, S, and s antigens in location of M/N on GPA is depicted in Figure 1, and the lo- Whites and Blacks is given in Table 2. cation of ‘N’/He at the N-terminus, and S or s closer to the lipid bilayer of GPB, are shown in Figure 2. Figures 1 and Table 2. Prevalence of various MNSs phenotypes* 2 also show the location of other antigens that result from Prevalence (% occurrence) a single amino acid change. Antigens carried on GPA and Phenotype Whites Blacks GPB hybrid molecules are listed in Table 5. The first 26 amino acids of GPAN and GPB are identical, M+N–S+s– 6 2 and the weak N antigen expression of M+N– RBCs (called M+N–S+s+ 14 7 ‘N’) is likely a consequence of the fewer GPB copies per RBC M+N–S–s+ 8 16 as compared with the number of copies of GPA per RBC. M+N+S+s– 4 2 GPB has a shorter amino-terminal exofacial domain than M+N+S+s+ 24 13 does GPA and little or no cytoplasmic domain. There is evi- M+N+S–s+ 22 33 dence that the exofacial domain of GPA interacts with RBC membrane skeletal proteins.19 M–N+S+s– 1 2 On intact RBCs, GPA is susceptible to cleavage by M–N+S+s+ 6 5 trypsin at amino acid 39; however, the α-chymotrypsin M–N+S–s+ 15 19 *The numbers are compiled from numerous sources.

Fig. 1 Schematic of 1 M/N/Mcg/M The antigens of low prevalence result from single ami- GPA showing the loca- 5 no acid changes or are on hybrid GPA/GPB molecules. For tion of antigens that many years, a number of low-prevalence antigens in the arise from simple amino MNS blood group system were grouped together in the acid changes. The MNTD–/MNTD+Thr17Arg trypsin and ficin/papain 26 so-called Miltenberger (Mi) subsystem.14,15 Originally, RBCs Ny(a–)/Ny(a+) Asp27Glu cleavage sites on intact ENEH/VW/Hut Thr28Met/Lys reactive with the serum containing anti-Mia were classified Or–/Or+ Arg31Trp RBCs are noted. De- Trypsin 31/39 into four classes on the basis of their different reactions with pending on the location a Vr–/Vr+ Ser47Thr four types of sera called Verweyst (Vw), Miltenberger (Mi ), of the antigenic determi- Os(a–)/Os(a+) Pro54Ser Ri(a–)/Ri(a+) Glu57Lys Murrell (Mur), and Hill (Hil). RBCs carrying some antigens nant, antibodies to high- Mt(a–)/Mt(a+) Thr58Ile Ficin/Papain 59/61 assigned to the Miltenberger subsystem reacted with one or prevalence antigens ERIK–/ERIK+ Gly59Arg ENEV+/ENEV– Val62Gly more of these three specific antisera yet failed to react with on GPA are generically ENAV/MARS Glu63Lys a ENEP/HAG Ala65Pro anti-Mia. For example, GP.Hil (Mi.V) RBCs reacted with classified as anti-En TS anti-Hil, but not with anti-Mia (Table 3). The Miltenberger (trypsin-sensitive), anti- EnaFS (ficin-sensitive), RBC Lipid subsystem grew to 11 classes, which were defined by one Lipid or anti-EnaFR (ficin- Bilayer Bilayer or more determinants reacting with type-specific antisera resistant). The line and (Table 3). As the complexity of the Miltenberger subsystem ball at amino acid 26 GPA increased, further expansion was no longer feasible, desir- represents the N-linked able, or relevant. As all the antigens are in the MNS system, glycan. 131 COOH the recommendation was made to use “GP” for the glycoprotein and GYP for the gene, with the abbreviated name of the propositus in whom the variant has been described, e.g., Mi.V became GP.Hil and the encoding gene is GYP.Hil. The Fig. 2 1 16,17 Miltenberger subsystem is now considered obsolete. Schematic of N/He M–VV/M+Thr3Ser 5 GPB showing the location Glycophorin A and Glycophorin B of antigens Antigens in the MNS system are carried on GPA, GPB, that arise or hybrids thereof. GPA and GPB are type I transmembrane from simple S/s Met29Thr sialoglycoproteins that traverse the RBC membrane lipid amino acid -chymotrypsin 32 Mit–/Mit+ Arg35His bilayer once and are orientated with their amino-termini changes. The Ficin/Papain to the outside of the RBC membrane. The amino-terminal α-chymotrypsin s–DD/s+Pro39Arg domains of GPA and GPB both carry O-glycans, whereas and ficin/papain only GPA carries an asparagine-linked–glycan (N-glycan), cleavage sites RBC Lipid which is attached to amino acid residue 26.18 The M and N on intact RBCs Bilayer blood group polymorphism is the result of a different ami- are noted. no acid sequence at the amino-terminus of GPA, whereas GPB 72 the S and s blood group polymorphism arises from a single COOH amino acid substitution at residue 29 of GPB (Table 4). The

96 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 MNS system: a review

cleavage site is protected by local Table 3. GP classification and associated antigens of the obsolete Miltenberger subsystem carbohydrate residues and, there- Reaction of RBCs with antiserum fore, is relatively resistant to this Mi enzyme. This explains why the M Classes Mia Vw Mur Hut Hil Hop Nob DANE TSEN MINY MUT and N antigens are trypsin-sensitive GP (ISBT #) (8) (9) (10) (19) (20) (26) (27) (32) (33) (34) (35) but -chymotrypsin–resistant. In α GP.Vw Mi.I + + 0 0 0 0 0 0 0 0 0 contrast, on intact RBCs, GPB is resistant to trypsin cleavage but GP.Hut Mi.II + 0 0 + 0 0 0 0 0 0 + sensitive to α-chymotrypsin cleav- GP.Mur Mi.III + 0 + 0 + 0 0 0 0 + + age (at amino acid 32), which ac- GP.Hop Mi.IV + 0 + 0 0 + 0 0 + + + counts for the trypsin-resistant, GP.Hil Mi.V 0 0 0 0 + 0 0 0 0 + 0 α-chymotrypsin–sensitive prop- GP.Bun Mi.VI + 0 + 0 + + 0 0 0 + + erties of its ‘N’, S, and s antigens GP.Nob Mi.VII 0 0 0 0 0 0 + 0 0 0 0 (Figs. 1 and 2). GP.Joh Mi.VIII 0 0 0 0 0 + + 0 0 0 0 Function of Glycophorin A and GP.Dane Mi.IX 0 0 + 0 0 0 0 + 0 0 0 Glycophorin B GP.HF Mi.X + 0 0 0 + 0 0 0 0 + + GPA (CD235A) and GPB GP.JL Mi.XI 0 0 0 0 0 0 0 0 + + 0 (CD235B) contribute most of the carbohydrate on the RBC membrane. The O-glycans on these two glycophorins carry most of the sialic acid and contribute to Table 4. Amino acids associated with M, N, S, and s antigens of GPA the net negative charge of the RBC. The negatively charged and GPB glycocalyx keeps RBCs from sticking to each other and Amino acid to the endothelial cells of the blood vessels2,18 and pro- Antigen Gene Glycophorin polymorphisms tects the RBC from invasion by bacteria and other patho- M GYPAM GPAM 1 Ser, 5 Gly 20,21 gens. GPA-deficient RBCs are more resistant to inva- N GYPAN GPAN 1 Leu, 5 Glu sion by Plasmodium falciparum merozoites because sialic S GYPBS GPBS 29 Met acid appears to be essential for adhesion of the parasite to s s the RBC.22–25 The sialic acid attached to GPA and GPB has s GYPB GPB 29 Thr been reported to be the target of the influenza virus20 and the encephalomyocarditis virus.26 It has also been demon- The high degree of sequence homology and organization strated that GPA plays a role in chaperoning band 3 trans- along the chromosome may be responsible for the relatively port to the RBC membrane.27,28 An absence of GPA or GPB, frequent occurrence of unequal homologous recombina- or GPA and GPB, from RBCs (En(a–), S–s–, and MkMk, tions and gene conversions among the three GYP genes. A phenotypes, respectively) does not result in altered RBC number of the resulting glycophorin products show variant morphology, and individuals with these phenotypes do not blood group phenotypes whose molecular studies, in turn, have significant physiologic abnormalities. led to the documentation of gene rearrangements (Tables 5 and 6).11,30 Glycophorin Gene Family The genes encoding GPA (GYPA) or GPB (GYPB) and GYP Gene Deletions Giving Rise to Null Phenotype RBCs GPE glycophorin (GYPE) are homologous and reside in The various deletions that give rise to the rare null phe- an approximately 350-kb gene cluster (5′-GYPA-GYPB- notypes, En(a–)Fin, S–s–U–, and MkMk, are summarized GYPE-3′) on 4q28–q31.29 GYPA has seven exons, GYPB in Table 7. It should be noted that deletion is not the only has six exons (of which exon 3 is a pseudo or a noncod- mechanism by which the En(a–) and S–s–U– phenotypes ing exon), and GYPE also has six exons (of which exons can arise.31–33 RBCs from individuals with the En(a–) Fin 3 and 4 are pseudo-exons). For GYPA and GYPB genes, phenotype lack GPA and, thus, all antigenic determinants exons 1 and 2 encode a leader sequence, exons 2 through 4 associated with GPA. RBCs from individuals with the S–s– encode the extracellular domains, and exon 5 encodes the U– phenotype lack GPB and, thus, all antigens carried on transmembrane domains. Exon 6 and part of the 5′ end of this glycophorin. RBCs from people with the MkMk pheno- exon 7 of GYPA encode the cytoplasmic domain of GPA. type lack both GPA and GPB, and thus lack M, N, Ena, S, s, If GYPE encodes a membrane-bound product (this is still and U blood group antigens. controversial), it would be a short sialoglycoprotein with the M antigen at its amino-terminus.

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 97 M.E. Reid

Molecular Bases of Glycophorin Variants and Antigens recombinants in reciprocal arrangement) can occur RBCs with altered glycophorin molecules may be de- between GYPA and GYPB if the chromosomes misalign tected in a variety of ways. The M, N, S, or s antigens are during meiosis. A chromosome carrying the hybrid GYP(A- expressed more strongly or more weakly than controls and B) gene does not have GYPA or GYPB genes (Lepore type), are unexpectedly sensitive or resistant to treatment by dif- whereas the reciprocal product is a chromosome carrying a ferent enzymes; there is a loss of a high-prevalence anti- hybrid GYP(B-A) gene as well as normal GYPA and GYPB gen or appearance of a low-prevalence antigen. Analysis by genes (anti-Lepore type). The novel amino acid sequences SDS-polyacrylamide gel electrophoresis is a useful tool to encoded by nucleotides at the junctions may be recognized reveal a change in mobility caused by either a new glyco- by antibodies in human serum. Examples of Lepore type phorin molecule or a change in glycosylation. and anti-Lepore type GYP genes, variant glycophorins, and Variant GYP genes arise from a nucleotide substitution the antigens expressed are given in Table 5.5 or substitutions, unequal crossover, or gene conversion. These GYP gene rearrangements have given rise to the anti- Double crossover genic diversity of the MNS blood group system.9,11,30 See the This mechanism also involves a mutual exchange of NCBI Web site ‘dbRBC’ (www.ncbi.nlm.nih.gov/gv/rbc) for nucleotides and generation of two reciprocal recombinant a list of reported alleles. genes. As a consequence of this mechanism, chromosomes Nucleotide substitutions giving rise to MNS antigens carrying a GYP(A-B-A) gene also carry a normal GYPB gene are shown in Figures 1 and 2. and the reciprocal chromosome bears a GYP(B-A-B) gene and a normal GYPA gene. Such genes can also arise through gene Single crossover conversion and although currently there is no way to prove A crossover (a mutual exchange of nucleotides between which mechanism gives rise to the GP(A-B-A) or GP(B-A-B) misaligned homologous genes with generation of two hybrid gene, the favored mechanism is gene conversion.

Table 5. Hybrid alleles, glycophorins, phenotypes, and associated low-prevalence antigens Allele Glycophorin Phenotype Antigens on hybrid GYP(A-B) GP(A-B) GP.Hil Hil, MINY GP.JL TSEN, MINY GP.TK SAT GYP(B-A) GP(B-A) GP.Sch (Mr) Sta GP.Dantu Dantu GYP(A-B-A) GP(A-B-A) GP.Mg Mg GP.KI Hil GP.SAT SAT GYP(B-A-B) GP(B-A-B) GP.Mur Mia, Mur, MUT, Hil, MINY GP.Bun Mia, Mur, MUT, Hop, Hil, MINY GP.HF Mia, MUT, Hil, MINY GP.Hop Mia, Mur, MUT, Hop, TSEN, MINY GP(A-B) GP.He (P2, NY, GL) He GYP(B-A-ψB-A) GP(A-A) GP.Cal He, Sta GYP(A-ψB-A) GP(A-B-A) GP.Vw Mia, Vw GP.Hut Mia, Hut, MUT GP.Nob Nob GP.Joh Nob, Hop GP.Dane Mur, DANE GP(A-A) GP.Zan (MZ) Sta GYPA 179G>A GPA GP.EBH ERIK encoded by one transcript GP(A-A) GP.EBH Sta encoded by a second transcript GYP(A-ψE-A) GP(A-A) GP.Mar Sta

98 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 MNS system: a review

Table 6. Molecular mechanisms and associated antigens carrying a GYP(A-B-A) gene will also carry a normal GYPB Molecular mechanism Associated antigens* gene (Table 8). During gene conversion, a splice sequence Single nucleotide substitution GPA: Vr, Mta, Ria, Nya, Or, ERIK, may be altered so that part of the GYPB pseudo-exon 3 is Osa, ENEP/HAG, ENAV/MARS, encoded. Gene conversion gives rise to hybrid genes that ENEV, MNTD encode novel glycophorin molecules carrying certain low- incidence antigens in the MNS blood group system (Tables GPB: S/s, Mv, sD, Mit 5 and 6).4,5,9 When studying samples with variant glyco- Two or more nucleotide substitution GPA: M/N phorins, it is important to consider the products of genes carried on the partner in trans chromosome, which usually GPB: ‘N’ is wild-type GYPA and GYPB genes. Unequal crossing over Sta, Dantu, Hil, TSEN, MINY, SAT Gene conversion He, Mia, Mc, Vw/Hut/ENEH, Mur, Antigens associated with altered glycosylation Mg, Me, Sta, Hil, Hop, Nop, DANE, Several antigens associated with MNS have not been MINY, MUT, ENDA assigned numbers by the ISBT Working Party. For exam- *Ena and U antigens are not listed. ple, M1 and Can, which are located within the N-terminal 31 amino acid residues of GPAM, and Tm, Sj, Hu, and Sext, which are located within the N-terminal domain of GPAN, Table 7. Molecular basis of null phenotype RBCs have not been assigned numbers yet. Antibodies to these Null phenotype RBCs Molecular basis antigens are not completely specific and exhibit weak affin- En(a–)Fin Deletion of GYPA (exon 2 to 7) and ity for, respectively, GPB and GPA. They react with excep- GYPB (exon 1) tional examples of, respectively, M–N+ and M+N– RBCs S–s–U– (deletion type) Deletion of GYPB (exon 2 to 6) with a high level of GlcNAc-containing O-glycans.35 and GYPE (exon 1) MkMk Deletion of GYPA (exon 2 to 7), Clinical Relevance GYPB (exon 1 to 6), and GYPE Antibodies to antigens in the MNS blood group system (exon 1) are often “naturally occurring” and if they do not react at 37°C can be ignored in transfusion practice. Rare examples of anti-M and anti-N have been reported to cause immediate Table 8. Mechanisms giving rise to variant GYP genes and their in cis partner genes* and delayed hemolytic transfusion reactions and hemolytic Mechanism for disease of the fetus and newborn (HDFN). Clinically signifi- variant GYP gene Variant GYP gene GYP gene(s) in cis cant anti-N are most likely found in people with the M+N– Single crossover GY(A-B) No GYPA or GYPB U– phenotype, and thereby detect GPB as well as N on GPA. (Lepore) Anti-M is often cold-reacting IgG and is enhanced by testing 36 Single crossover GY(B-A) GYPA and GYPB in an acidified environment. Some anti-M and anti-N only (anti-Lepore) react with RBCs previously exposed to glucose.37–39 In the Gene conversion GY(A-B-A) and GYPB late 1970s, as many as a third of kidney dialysis patients us- GY(A-ψB-A) ing formaldehyde-sterilized dialysis membranes had anti- 40,41 GYP(B-A-B) GYPA Nf, regardless of their MN phenotype. This phenomenon *Note that the in trans partner chromosome will bear whichever is rarely seen today because formaldehyde sterilization was GYP genes were inherited (most likely GYPA and GYPB). discontinued in dialysis practices after the reason for the presence of anti-Nf was delineated.42 Anti-S, -s, and anti- Gene conversion U have been implicated in hemolytic transfusion reactions Gene conversion can occur during the process of DNA (immediate and delayed) and in HDFN. Antibodies to many repair between homologous genes.34 This mechanism re- low-prevalence antigens in the MNS system have caused flects a directional transfer of nucleotides from one duplex HDFN; in fact, that is typically how they are detected. Of to another and does not result in a reciprocal product. In particular note is the clinical importance of allo-anti-Ena, eukaryotes, if adjacent, homologous genes misalign during which has caused hemolytic transfusion reactions,43,44 and meiosis, nucleotides from one gene can be inserted into one auto-anti-Ena, which has caused severe and fatal autoim- strand of DNA of the other gene. If this strand is the coding mune hemolytic anemia.45–47 strand, then a hybrid gene will be formed.4,5 Gene conversion can cause an insert of nucleotides from Concluding Remarks GYPA into the GYPB gene or an insertion of nucleotides The MNS blood group system, with 46 antigens, is sec- from GYPB into the GYPA gene. A chromosome carrying ond in its diversity only to the Rh system (with 50 antigens). a GYP(B-A-B) gene that is a result of this mechanism will Both systems are encoded by homologous genes: in the also carry a normal GYPA gene. Similarly, a chromosome MNS system, GYPA and GYPB, and in the Rh system, RHD

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 99 M.E. Reid and RHCE. Thus, the antigens are not only a consequence 14. Cleghorn TE. A memorandum on the Miltenberger of amino acid changes but also a result of novel sequences blood groups. Vox Sang 1966;11:219–22. induced by rearranged genes. The GYPA, GYPB, and GYPE 15. Dahr W. Miltenberger subsystem of the MNSs gene cluster is prone to mutation by radiation; there is a blood group system. Review and outlook. Vox Sang high incidence of somatic mutation events in atomic bomb 1992;62:129–35. survivors, in people exposed to accidental radiation, in pa- 16. Tippett P, Reid ME, Poole J, et al. The Miltenberger sub- tients with Bloom’s syndrome, and in patients receiving ra- system: is it obsolescent? Transf Med Rev 1992;6:170– diation therapy. The mutation events were dose dependent; 82. the greater the exposure, the greater proportion of RBCs 17. Reid ME, Tippett P. Review of a terminology pro- exhibited mutations.48–51 posed to supercede Miltenberger. Immunohematology 1993;9:91–5. Acknowledgment 18. Anstee DJ. The nature and abundance of human red cell I thank Robert Ratner for assistance in preparing the surface glycoproteins. J Immunogenet 1990;17:219– manuscript and figures. 25. 19. Chasis JA, Mohandas N, Shohet SB. Erythrocyte References membrane rigidity induced by glycophorin A-ligand 1. Anstee DJ. Blood group-active surface molecules of the interaction. Evidence for a ligand-induced association human red blood cell. Vox Sang 1990;58:1–20. between glycophorin A and skeletal proteins. J Clin 2. Blanchard D. Human red cell glycophorins: biochemical Invest 1985;75:1919–26. and antigenic properties. Transf Med Rev 1990;4:170– 20. Lublin DM. Functional roles of blood group antigens. 86. In: Silberstein LE, ed. Molecular and functional aspects 3. Cartron J-P, Rahuel C. Human erythrocyte glyco- of blood group antigens. Bethesda, MD: American As- phorins: protein and gene structure analyses. Transf sociation of Blood Banks, 1995;163–92. Med Rev 1992;6:63–92. 21. Moulds JM, Nowicki S, Moulds JJ, Nowicki BJ. Human 4. Reid ME. Some concepts relating to the molecular ge- blood groups: Incidental receptors for viruses and bac- netic basis of certain MNS blood group antigens. Transf teria. Transfusion 1996;36:362–74. Med 1994;4:99–111. 22. Cartron J-P, Prou O, Luilier M, Soulier JP. Susceptibil- 5. Huang C-H, Blumenfeld OO. MNSs blood groups and ity to invasion by Plasmodium falciparum of some hu- major glycophorins: molecular basis for allelic varia- man erythrocytes carrying rare blood group antigens. tion. In: Cartron J-P, Rouger P, eds. Molecular basis of Br J Haematol 1983;55:639–47. human blood group antigens. New York: Plenum Press, 23. Pasvol G, Chasis JA, Mohandas N, et al. Inhibition of 1995;153–88. malarial parasite invasion by monoclonal antibodies 6. Blumenfeld OO, Huang C-H. Molecular genetics of the against glycophorin A correlates with reduction in red glycophorin gene family. The antigens for MNSs blood cell membrane deformability. Blood 1989;74:1836–43. groups; multiple gene rearrangements and modulation 24. Hadley TJ. Invasion of erythrocytes by malaria para- of splice site usage result in extensive diversification. sites: a cellular and molecular overview. Annu Rev Mi- Hum Mutation 1995;6:199–209. crobiol 1986;40:451–77. 7. Daniels G. Human blood groups. 2nd ed. Oxford: Black- 25. Orlandi PA, Klotz FW, Haynes JD. A malaria invasion well Science Ltd, 2002. receptor, the 175-kilodalton erythrocyte binding anti- 8. Issitt PD, Anstee DJ. Applied blood group serology. 4th gen of Plasmodium falciparum recognizes the terminal ed. Durham, NC: Montgomery Scientific Publications, Neu5Ac(alpha 2-3)Gal- sequences of glycophorin A. J 1998. Cell Biol 1992;116(4):901–9. 9. Palacajornsuk P. Review: molecular basis of MNS blood 26. Allaway GP, Burness AT. Site of attachment of enceph- group variants. Immunohematology 2006;22:171–82. alomyocarditis virus on human erythrocytes. J Virol 10. Landsteiner K, Levine P. Further observations on indi- 1986;59:768–70. vidual differences of human blood. Proc Soc Exp Biol 27. Groves JD, Tanner MJ. Glycophorin A facilitates the Med 1927;24:941-2. expression of human band 3-mediated anion transport 11. Reid ME, Lomas-Francis C. Blood group antigen facts- in Xenopus oocytes. J Biol Chem 1992;267:22163–70. book. 2nd ed. San Diego: Academic Press, 2004. 28. Groves JD, Tanner MJ. The effects of glycophorin A 12. Daniels GL, Anstee DJ, Cartron J-P, et al. Blood group on the expression of the human red cell anion trans- terminology 1995. ISBT working party on terminology porter (band 3) in Xenopus oocytes. J Membr Biol for red cell surface antigens. Vox Sang 1995;69:265– 1994;140:81–8. 79. 29. Fukuda M. Molecular genetics of the glycophorin A 13. Daniels G, Flegel WA, Fletcher A, et al. International gene cluster. Semin Hematol 1993;30:138–51. Society of Blood Transfusion committee on terminology for red cell surface antigens: Cape Town report. Vox Sang 2007;92:250–3.

100 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 MNS system: a review

30. Huang C-H, Johe KK, Seifter S, Blumenfeld OO. Bio- 42. Dahr W, Moulds J. An immunochemical study on anti- chemistry and molecular biology of MNSs blood group N antibodies from dialysis patients. Immunol Commun antigens. Baillieres Clin Haematol 1991;4:821–48. 1981;10:173–83. 31. Rahuel C, London J, Vignal A, et al. Erythrocyte gly- 43. Furuhjelm U, Nevanlinna HR, Pirkola A. A second cophorin B deficiency may occur by two distinct gene Finnish En(a-) propositus with anti-Ena. Vox Sang alterations. Am J Hematol 1991;37:57–8. 1973;24:545–9. 32. Huang C-H, Reid ME, Blumenfeld OO. Remodeling of 44. Postoway N, Anstee DJ, Wortman M, Garratty G. A se- the transmembrane segment in human glycophorin by vere transfusion reaction associated with anti-EnaTS in aberrant RNA splicing. J Biol Chem 1994;269:10804– a patient with an abnormal alpha-like red cell sialogly- 12. coprotein. Transfusion 1988;28:77–80. 33. Storry JR, Reid ME, Fetics S, Huang C-H. Mutations in 45. Pavone BG, Billman R, Bryant J, et al. An auto-anti- GYPB exon 5 drive the S-s-Uvar+ phenotype in persons Ena, inhibitable by MN sialoglycoprotein. Transfusion of African descent: implications for transfusion. Trans- 1981;21:25–31. fusion 2003;43:1738–47. 46. Garratty G, Arndt P, Domen R, et al. Severe autoimmune 34. Maizels N. Might gene conversion be the mechanism of hemolytic anemia associated with IgM warm auto- somatic hypermutation of mammalian immunoglobu- antibodies directed against determinants on or associ- lin genes? Trends Genet 1989;5:4–8. ated with glycophorin A. Vox Sang 1997;72:124–30. 35. Dahr W, Knuppertz G, Beyreuther K, et al. Stud- 47. Garratty G. Specificity of autoantibodies reacting opti- ies on the structures of the Tm, Sj, M1, Can, Sext and mally at 37°C. Immunohematology 1999;15:24–40. Hu blood group antigens. Biol Chem Hoppe-Seyler 48. Jensen RH, Reynolds JC, Robbins J, et al. Glycophorin 1991;372:573–84. A as a biological dosimeter for radiation dose to the bone 36. Beattie KM, Zuelzer WW. The frequency and properties marrow from iodine-131. Radiat Res 1997;147:747–52. of pH-dependent anti-M. Transfusion 1965;5:322–6. 49. Jensen RH, Langlois RG, Bigbee WL, et al. Elevated fre- 37. Morel PA, Bergren MO, Hill V, et al. M and N specific quency of glycophorin A mutations in erythrocytes from hemagglutinins of human erythrocytes stored in glu- Chernobyl accident victims. Radiat Res 1995;141:129– cose solutions. Transfusion 1981;21:652–62. 35. 38. Reid ME, Ellisor SS, Barker JM, et al. Characteristics of 50. Langlois RG, Bigbee WL, Jensen RH, German J. Evi- an antibody causing agglutination of M-positive non- dence for increased in vivo mutation and somatic re- enzymatically glycosylated human red cells. Vox Sang combination in Bloom’s syndrome. Proc Natl Acad Sci 1981;41:85–90. U S A 1989;86:670–4. 39. Drzeniek Z, Kusnierz G, Lisowska E. A human anti- 51. Langlois RG, Bigbee WL, Kyoizumi S, et al. Evidence for serum reacting with modified blood group M determi- increased somatic cell mutations at the glycophorin A lo- nants. Immunol Invest 1981;10:185–97. cus in atomic bomb survivors. Science 1987;236:445–8. 40. Fassbinder W, Seidl S, Koch KM. The role of formaldehyde in the formation of haemodialysis-associated Marion E Reid, PhD, Director of Laboratory of Immuno- anti-N-like antibodies. Vox Sang 1978;35:41–8. hematology and Head of Laboratory of Immunochemistry, 41. Lynen R, Rothe M, Gallasch E. Characterization of New York Blood Center, 310 East 67th Street, New York, formaldehyde-related antibodies encountered in he- NY 10065. modialysis patients at different stages of immunization. Vox Sang 1983;44:81–9.

Free Classified Ads and Announcements

Immunohematology will publish classified ads and announcements (SBB schools, meetings, symposia, etc.) without charge. Deadlines for receipt of these items are as follows:

Deadlines 1st week in January for the March issue 1st week in April for the June issue 1st week in July for the September issue 1st week in October for the December issue

E-mail or fax these items to [email protected] or (215) 451-2538.

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 101 Review

Quality activities associated with hospital tissue services

C.M. Hillberry and A.J. Schlueter

Quality assurance is a vital component of a tissue service that aims distribution of any human cell or tissue, and the screening to meet regulatory requirements and provide safe and functional or testing of the cell or tissue donor.1 Hospital transfusion tissue for surgical procedures in the institution. Many hospital services that also provide tissue services commonly store transfusion services have or are in the process of assuming tis- purchased allograft bone, ligament, skin, heart valve, and sue service functions for their institutions, and the concepts of cornea for use within their facilities. tissue quality assurance requirements should be familiar to the transfusion service. This review discusses the key aspects of tissue Hospital tissue services are exempt from the FDA service quality assurance, including requirements for FDA regis- HCT/P (tissue) registration requirements if they only per- tration, selection and qualification of tissue suppliers, recordkeep- form the following tissue activities: ing, management of occurrences and tissue recalls, adverse event • Autologous bone (or other autologous tissue) is re- reporting, audits, and quality control. Comparing the similarities moved during a surgical procedure and then stored un- and differences between tissue and blood component regulatory til implantation into the patient at a future date within requirements suggests transfusion service processes can be read- the same facility.3 ily adapted to incorporate quality assurance for tissue service ac- • Autologous tissue is removed and stored at another tivities. establishment until future implantation at the facility Immunohematology 2009;25:102–106. that harvested the tissue.4 • Allograft tissue is received from a supplier and stored tissue service, quality assurance, records, re- Key Words: by the tissue service for use in surgical procedures only calls, adverse event reporting, supplier qualification at that facility.3 If the tissue service serves as a distributor to another ith the implementation of the US Food and Drug facility or to another physical location under the same man- Administration (FDA) current Good Tissue Prac- agement, tissue registration is required.3 For example, if an tices (cGTP) tissue regulations on May 25, 2005, W autologous bone removed during a surgical procedure at and the introduction of the Joint Commission accrediting one facility is transferred to another facility for implanta- standards on July 1, 2005, many transfusion services are tion, or if purchased allograft tissue is shipped directly to overseeing their hospital’s tissue service functions.1,2 Ad- another hospital, the shipping tissue service must register ditional accrediting agencies (e.g., AABB and College of with the FDA American Pathologists) may also be inspecting tissue service activities as part of their overall laboratory inspection. What is the registration process? Thus, quality assurance personnel in transfusion services, If the tissue service performs a manufacturing step that who are very experienced in activities related to qualifying is not one of the exempt activities, it must register with the suppliers, product tracking, recalls, and event reporting, FDA by using Form FDA 3356. The registration form can be are assuming similar tissue service responsibilities. Trans- obtained by writing to the Center for Biologics Evaluation fusion services can incorporate many of the tissue activities and Research (CBER) or downloading the form at www.fda. into their current procedures; however, some procedures gov/opacom/morechoices/fdaforms/cber.html.5 The tissue may need to be created or modified. From registration to service, like the transfusion service, is required to provide adverse event reporting, this review discusses aspects of facility contact information, types of tissues manufactured, quality assurance needed for a tissue service, with a focus manufacturing steps performed at the facility, and the on allograft tissue requirements, for comparison with simi- contact information of the reporting official. Even though lar functions required for a transfusion service. tissues may be handled or manufactured by various units within the hospital, one tissue registration may be submit- FDA HCT/P Registration ted that includes all tissue activities performed at the fa- Who must register? cility. For example, allograft surgical tissue, hematopoietic Similar to the requirement for blood components, the stem cells, and reproductive tissue may be managed inde- FDA requires all manufacturers of products containing or pendently in the hospital, but all could be reported to the consisting of human cells or tissues (HCT/Ps) intended for FDA in a single tissue registration. The tissue registration implantation, transplantation, infusion, or transfer into a must be updated annually, in December, or within 5 days of human recipient to register.3 Manufacturing is defined as initiating any new manufacturing function.3 any step in the recovery, processing, storage, labeling, or

102 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Quality activities of tissue services

Qualification and Selection of Tissue Suppliers Many companies have information on their Web sites, Hospital transfusion services have policies and proce- including state licenses, product information, package in- dures for selecting and qualifying blood component and re- serts, and accreditation certificates. A public query function agent suppliers that can be expanded to encompass tissue on the FDA Web sites (www.fda.gov/cber/tissue/tissreg- suppliers. A common difficulty encountered when quali- data.htm) allows anyone to search for a company’s tissue fying tissue suppliers is that many of the companies sell- registration information. Product recalls can also be found ing tissue products are not the manufacturer but simply a on the FDA Web site at www.fda.gov/opacom/Enforce. distributor of the product. This makes it difficult to deter- html. It may be useful to contact the manufacturer’s quality mine whom to qualify as the supplier, especially with the assurance department through a phone call or by sending a frequent buyouts and mergers that occur in the tissue in- supplier questionnaire to obtain any additional information dustry. In some cases the distributor can be qualified inde- needed to complete the qualification. In rare instances, the pendently, but in other instances the distributor will refer tissue service’s medical director may request an on-site au- the tissue service to the parent manufacturing company for dit of the manufacturer, after review of the Form FDA 483s much of the information necessary to complete the qualifi- received by the manufacturer or after a recall of any of its cation. When qualifying a supplier (either a distributor or products. However, some tissue manufacturers may not al- manufacturer), predetermined selection criteria must be low customers to observe some of their practices, to protect defined. Selection criteria are used to evaluate the quality proprietary processes. of the products produced by the manufacturer by review- Periodically after the initial selection and qualification, ing items described in Table 1. Additional secondary selec- the tissue supplier’s performance should be reevaluated. tion criteria may also be considered as described in Table Reevaluation may include items from the initial qualification 2. The secondary selection criteria assess factors that may as well as clinical feedback from the surgeons (e.g., ease of facilitate administrative interactions with the supplier and tissue use, clinical effectiveness of the tissue) and supplier can be useful in distinguishing among several companies performance issues (e.g., timely product delivery, product that provide products of similar quality. back orders).

Tissue Tracking Records A key task in tissue management is accurate tracking Table 1. Qualification criteria of tissue movement in and out of the inventory. The tissue Types of tissues manufactured service must be able to track the tissue from the time of FDA Form 483 citations and corrective actions receipt into the facility until the final disposition (implan- Manufacturing processes6 (including current FDA registration for the tation, return to the supplier, or disposal), analogous to the manufacturing steps performed) requirement for blood component tracking for transfusion 1 Qualifications of contract companies performing manufacturing steps services. Tissue tracking should also include documentation (e.g., tissue recovery, laboratory testing) of all instances in which the tissue was sent to patient care Written quality program areas under established storage conditions and returned to Nature of product recalls inventory after meeting acceptance criteria. Tissue storage and transport conditions are based on the manufacturer’s Clinical Laboratory Improvement Amendments (CLIA) certification storage and handling instructions and the storage time Licensed by the state if applicable (CA, FL, MD, NY) determined for the validated transport containers. This information must be communicated to the surgical areas and ideally incorporated into the nursing policies for tissue handling. The tissue service also establishes criteria for accepting returned unused tissues into inventory. The acceptance criteria may include the length of time in the Table 2. Additional selection criteria for tissue suppliers validated transport container, inspection of package seals, Nature of tissue tracking system tissue container integrity, and product temperature (e.g., Product cost and expiration dates product remains frozen). Customer service (product return policy, availability of product) The ability to track all tissue is essential in the event of a Method for monitoring contracted companies6 product recall that may require patient notification. Tissue Method of tissue recall notification tracking may use electronic or paper records, but any tracking method must contain several key pieces of information. Accreditation by American Association of Tissue Banks (AATB) or Eye Bank Association of America (EBAA) These include the unique identifier (donor identification number, lot number), final disposition, tissue supplier, and Method for adverse event reporting the patient name and identifier if the tissue was implanted. Allowance for customer audits The tissue service may receive several pieces of tissue with

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 103 C.M. Hillberry and A.J. Schlueter the same donor identifier and lot number from the same tissue service chooses not to release PHI, in the event of a supplier. Each of these tissues must be tracked individu- tissue recall it assumes responsibility for determining who ally. If the tissue service decides to assign new identification received the recalled tissue, as the manufacturer would be numbers, the original identification number must remain unable to provide patient information with the recall no- part of the tracking record. Alternatively, a subunit number tice. may be added to the original identification number so that The record retention time for tissue records is longer the original number remains part of the final tissue identifi- than the retention time for blood component records. The cation. Joint Commission Accreditation Program standards re- Commercial tissue tracking software is available to quire that several tissue records must be maintained for a assist the tissue service in tracking tissue inventory from minimum of 10 years or longer (if required by state or fed- tissue receipt to final disposition. Some programs have the eral laws) beyond the date of distribution, transplantation, capability of data entry through the use of bar code scan- disposition, or expiration, whichever is latest.2 These tissue ning. The data can be quickly retrieved from such programs records include the tissue supplier, the original numeric or in the event of product recall or adverse event. alphanumeric donor and lot identification, name(s) of the For tracking, recall, and billing purposes, the definition recipient(s) or the final disposition of each tissue, and the of an implanted tissue and a wasted tissue must be clear for expiration dates of all tissues. Other records that are re- both the surgical team and the tissue service. A conserva- quired to be retained for a minimum of 10 years are tissue tive approach is to track any tissue that has been exposed to storage temperatures, outdated procedures, manuals, and the patient (e.g., package opened in the surgical suite, tissue publications.2 placed on the patient and removed) similar to an implanted tissue because the patient may have been exposed to po- Management of Tissue Recalls tential infectious agents from such tissue either by direct Tissue recalls are unlike blood recalls in that they often contact with the tissue or indirectly from surgical staff or involve numerous tissues, and the tissues may have been instruments touching the tissue and then the patient during handled by multiple companies at various steps between the surgical procedure. If this conservative approach is used tissue procurement and distribution. Thus, for example, for tracking purposes, the tissue service must still distin- a tissue recalled by a parent company may have been dis- guish which tissues remained in the patient at the end of the tributed and labeled by a subsidiary, making the identity case, as patients can only be billed for these tissues. Tissues of the recalled item obscure. In the event of a tissue recall, to which the patient was simply exposed during the surgical the tissue manufacturer generally sends the recall notifi- procedure would be considered wasted for billing purposes. cation to the entity that ordered the tissue (e.g., the hospi- Thus, it is essential that the patient care staff communicate tal procurement service, the operating room, or the tissue clearly to the tissue service whether the implanted tissue service). The FDA also lists recalls at www.fda.gov/safety/ was left in the patient or the patient was only exposed to the recalls/default.htm and www.recalls.gov/. In addition, tissue. there are companies that compile recall notifications from A key element in tracking tissues that is different from multiple sources and distribute them to their subscribers. tracking blood component disposition is supplier notifica- Subscribing to such a service may be useful to help ensure tion of final tissue disposition. Many suppliers request that that the tissue service is aware of all pertinent recalls. The tissue services provide final tissue disposition information recall notification includes the reason for the recall, the to them so that proper recall notification may occur if it description of the recalled tissue (catalog number, product becomes necessary.2 Some manufacturers use carbon copy name), and the tissue identification number and may also notification forms, which facilitates retention of a copy of include how to handle any recalled tissue that may still the completed form for tissue service records; however, cur- be in inventory. Recalls are almost always voluntary on rently there is no standard method or format for providing the part of the manufacturer, either because it discovered final tissue disposition information to the manufacturer. A a problem or because the FDA raised concerns about the variety of information may be requested such as patient de- product. Only rarely does the FDA mandate a recall. mographics, surgeon name, date of final disposition, type of Regardless of whether the recall is voluntary or man- final disposition (e.g., implanted, wasted), facility contact dated, the same process should be followed by the tissue information, and tissue identifier. Health Insurance Porta- service. After receiving the recall notice, the tissue service bility and Accountability Act (HIPAA) regulations allow the should immediately check the current inventory for the re- release of protected health information (PHI) to the tissue called tissue and quarantine the tissue to prevent dispens- manufacturers without explicit patient consent for the pur- ing it for a patient. The quarantined tissue should then be pose of routine product tracking and ultimately to enable discarded or returned to the manufacturer as instructed in product recalls, including locating and notifying individuals the recall notification. After reviewing the available inven- who have received recalled tissues.2 However, there is no tory, the tissue service should determine whether any of requirement that the tissue service must release PHI when the recalled tissue was previously implanted. providing routine final tissue disposition information. If the

104 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Quality activities of tissue services

The tissue service should have a recall policy that out- to the FDA if they are fatal, are life-threatening, result in lines how to handle the recall information and potential permanent impairment of function or permanent damage to notification of the patients who have been affected. After body structure, or necessitate medical or surgical interven- reviewing the reason for the tissue recall, the tissue service tion.1 Unlike tissue adverse event reporting, blood donor medical director is responsible for determining when to or recipient adverse event reporting only involves notify the implanting surgeon. The medical director may fatality reports. Adverse events may be related to either determine that notification is not necessary in some cir- disease transmission or failure of intended tissue function. cumstances, for example, if the patient is deceased or a Examples of adverse reactions include bacterial or fungal minor error in labeling occurred that does not affect the surgical site infections, virus or cancer transmission, or tissue identity or integrity. Surgeon notification should increased intensity of medical care owing to the tissue include the reason for the recall, the patient information, implanted. recommendation regarding patient notification, and a The manufacturer’s quality assurance department may request for information related to any medical complica- request additional information from the tissue service in- tions from the implanted tissue. Patient notification may cluding patient laboratory tests, patient medical history, be deemed necessary by the tissue service medical direc- and current patient status to assist them in their internal tor in the event that the patient’s health is at risk from the investigation. Again, all of this may be provided without implanted tissue (e.g., positive microbial tested product, violation of HIPAA regulations. It is ultimately the manu- tissue donor tested positive for a communicable disease facturer’s responsibility to report the adverse event to the agent). If patient notification is deemed necessary and the FDA; however, the FDA encourages health-care profes- surgeon does not want to notify the patient or no longer is sionals to voluntarily report any adverse reactions related on staff at the institution, the tissue service medical to a communicable disease through MedWatch using Form director should take the responsibility for notifying the pa- 3500 in addition to notification of the tissue manufacturer. tients. Notifications can occur through phone calls, hand- If the tissue service is the tissue manufacturer (e.g., the har- delivered letters to individuals at clinic appointments, or vesting facility for autologous tissue that has been shipped registered letters, depending on the specific situation. If to another institution for implantation), the tissue service the letters are sent to the patient through registered mail, must submit a report within 15 days through MedWatch us- the tissue service should be aware that a patient’s family ing FDA Form 3500A.9 The reporting process for adverse member may sign for the letter and the patient may not be events related to blood components differs from the one for notified.7 tissue in that the reports for transfusion-related fatalities are not submitted on a standardized form. Adverse Event Reporting If the adverse event is related to the failure of a tissue The FDA defines adverse reaction to tissue as a noxious to function as expected, the tissue service should begin its and unintended response to any HCT/P for which there is a investigation by reviewing internal handling of the tissue. reasonable possibility that the HCT/P caused the response. The investigation may include review of records related to Every hospital needs a process for identifying and report- the tissue integrity at set time points, including each time ing adverse events. Because adverse events may be acute or the product was issued and received into the inventory, as delayed in presentation the adverse event reporting process well as storage equipment temperature records. In addition may involve several departments throughout the hospital.8 to conducting an internal investigation, the tissue service It is helpful to educate surgeons, the hospital epidemiology should notify the manufacturer of the problem through the department, or even the patient’s primary care provider to sales representative, a customer service line, or a manufac- encourage identification of potential tissue-related adverse turer complaint form located on the manufacturer’s Web events. Communication of the adverse event to the tissue site. service begins the investigation and reporting process. It is essential that all adverse reactions be investigated One or more designated services within the hospital and reported to the appropriate individuals. This information (e.g., hospital epidemiology, hospital quality assurance, tis- will allow the manufacturer to track recurring complaints sue service) must be responsible for accumulating reports about tissue problems and perform its own internal inves- and investigating all tissue adverse events. In some hospitals, tigations. The FDA uses the information reported through the tissue service is notified of any tissue-related adverse event MedWatch to determine whether a reaction is an isolated and performs the investigation with consultation with other event or an emerging problem. Currently the United Net- departments as needed. The tissue service medical director work of Organ Sharing (UNOS) under a contract with the would then be responsible for review of the data related to Centers for Disease Control and Prevention (CDC) is devel- the adverse event and determination of the likelihood that oping a system to allow easier reporting of adverse events the tissue may be implicated. If tissue involvement cannot and tracking of tissues between the tissue manufacturers reasonably be excluded, the manufacturer should be noti- and the hospitals. This system, called the Transplantation fied.2 In turn, all manufacturers must report adverse events Transmission Sentinel Network (TTSN), will include five

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 105 C.M. Hillberry and A.J. Schlueter key elements including registration of donors and recipients, routine activities of a transfusion service quality assurance reporting of adverse events, relay of information to regula- team. tory and public health agencies, and community education. An electronic mechanism for tracking and sharing informa- Acknowledgments tion between manufacturers and hospitals is being devel- We thank Judith Levitt for critical reading of the manu- oped for TTSN. Pilot testing for this electronic reporting script. system occurred in the spring of 2008.10 A similar program, called the Hemovigilance Module, under the US Biovigi- References lance Network, will track recipient adverse events associated 1. Current good tissue practice for human cell, tissue, and with blood component transfusions. cellular and tissue-based product establishments; inspection and enforcement; final rule. Federal register Other Tissue QA Functions 2004;69(226):68680–8. The transfusion service can easily incorporate tissue 2. The Joint Commission Accreditation Program. Hospi- into its current processes for auditing, occurrence manage- tal Standards, 2009. ment, and other quality control activities. Records can be 3. Human cells, tissues, and cellular and tissue-based audited for documentation of tissue receipt and acceptance products. Federal register 2001;66(13):5466–9. into inventory, tissue implant information in the patient’s 4. St. Martin LM, Witten CM. U.S. Food and Drug Admin- medical record, final disposition of the tissue (implanted, istration regulation of human cells, tissues, or cellular returned to the supplier, or wasted), and quality control of or tissue-based products (HCT/Ps) (presentation). Sep- tissue storage devices. Observation audits can be performed tember 6, 2007, www.docstoc.com/docs/542827/U-S- for tasks such as preparing tissue transport containers, re- Food-and-Drug-Administration-Regulation. turning product into inventory, and notifying the tissue 5. Human cells, tissues and cellular and tissue-based manufacturer of the tissue’s final disposition. products. Federal register 2008;73(13):3387. Occurrences involving tissues (e.g., equipment failures, 6. Humphries LK, Mansavage VL. Quality control in tis- deviations from standard operating procedures, supplier sue banking—ensuring the safety of allograft tissues. issues, documentation errors or omissions) can be tracked AORN J 2006;84:386–98. through the same occurrence management system current- 7. Steelman VM, Schlueter AJ. Managing a tissue recall in ly used by the transfusion service for blood components. If a large academic institution. Transfusion 2007;47:927– the facility manufactures tissues, tracking may also include 34. occurrences with product labeling, processing steps, and 8. Hospital tissue management: a practitioner’s hand- testing (microbial and infectious disease). book. 1st ed. AABB: Bethesda, MD, 2008. Quality control of equipment and testing is routine to 9. Guidance for Industry MedWatch Form FDA 3500A: the transfusion service, and similar functions are needed mandatory reporting of adverse reactions related to hu- for the tissue service. In particular, monitoring of tissue man cells, tissues, and cellular and tissue-based prod- storage conditions is essential as part of the verification that ucts (HCT/Ps). November 2005, www.fda.gov/cber/ high tissue quality is maintained. The tissue service must guidelines.htm. store the tissues under the appropriate storage conditions 10. Strong DM, AuBuchon J, Whitaker B, Kuehnert ML. as defined on the tissue package label.11 (An accompanying Biovigilance initiatives. ISBT Science Series 2008;3:77– review further discusses the intricacies of tissue storage.12) 84. Some tissue suppliers require the tissue service to sign a 11. Standards for tissue banking. 11th ed. American Asso- statement that the tissues were maintained at the accept- ciation of Tissue Banks (AATB): McLean, VA, 2006. able storage conditions. This acknowledgment may occur 12. Alden BM, Schlueter AJ. Logistical aspects of human at the time tissues are returned to the tissue supplier or surgical tissue management in a hospital setting. Im- through a supplier consignment agreement. munohematology 2009;25:107–11.

Conclusions Christine M. Hillberry, BS, MT (ASCP), CQA (ASQ) Tis- Tissue service quality assurance functions are generally sue Bank Coordinator, and Annette J. Schlueter, MD, PhD quite parallel to transfusion service quality assurance func- (corresponding author), Associate Professor, Department tions. They include tissue service registration with the FDA, of Pathology, University of Iowa, 200 Hawkins Drive, qualification of suppliers, ensuring a robust tissue tracking Iowa City, IA 52242. process, management of recalls, and reporting of adverse events. With attention to minor differences between tissue and blood component requirements, tissue service quality assurance activities can readily be incorporated into the

106 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Review

Logistical aspects of human surgical tissue management in a hospital setting

B.M. Alden and A.J. Schlueter

Many hospital transfusion services have assumed responsibility for tissue handling requirements greatly mirror those of blood. the coordinated management of human allograft tissue. This overview This overview outlines logistical aspects that should be con- summarizes logistical aspects of tissue management based on the ex- sidered in the operation of a hospital-based tissue service, perience of a centralized tissue service at a large academic hospital, with a focus on allograft tissue. in which tissue is stored in a location remote from patient care areas. Operational aspects include determination of which personnel classifications will perform the necessary functions, establishment and Personnel Requirements maintenance of the standing tissue inventory (including pros and cons When designing a comprehensive tissue service for a of alternative approaches to inventory acquisition), and necessary con- hospital, all aspects of tissue management must be identi- siderations for making tissue available for surgical cases in the hospital. fied, and the role of various service lines within the facility to The nature of communications regarding tissue orders for individual accomplish the necessary tasks must be clearly defined. The surgical cases is discussed, as well as mechanisms for storage of tissue functions that must be assigned include responsibility for and transportation and delivery of tissue to the surgical suites. Finally, inventory management; creation and revision of standard options for the disposition of tissue that has been dispensed from the operating procedures; acquisition, storage, dispensing, de- tissue service but was not used during the surgical procedure are sum- livery, and tracking of tissue; final preparation of tissue for marized. With attention to these details, a tissue service can provide reliable, high-quality tissue in a timely fashion. implantation; billing; and quality control and quality assur- Immunohematology 2009;25:107–111. ance activities. Multiple staff classifications both within and external to the tissue service may be qualified to perform Key Words: allografts, surgical tissue, tissue bank, tissue one or more of these functions. For example, tissues could service be transported within the facility by one or more of the following: patient care area or tissue service clerks, clinical ocumented tissue transplants were performed as laboratory scientists, nursing assistants, or even hospital early as the late 1800s, but tissue banking began volunteers. Similarly, billing could be the responsibility of Dwith the founding of the Navy Tissue Bank in 1949 purchasing or tissue service personnel. by Dr. George Hyatt. The purpose of this initial bank was After determining the tasks that will be performed by to store frozen bone for clinical and research purposes.1 the tissue service, the minimum staffing requirements nec- Hospitals soon developed their own tissue banks to store essary to perform these duties in an accurate and timely limited amounts of tissue to meet the needs of their sur- manner must be assessed. Factors to consider when determining geons, but by the late 1980s and early 1990s, the expanded minimum staffing requirements include the location of surgical use of tissue led to the growth of commercial tissue the tissue service in relation to the OR, the roles of the OR banks (hereafter referred to simply as tissue banks). At this and tissue service personnel in the tissue dispensing and time there was also a separation of the functions for retriev- delivery process, the number of surgeries per day requir- al, processing, and distribution to different organizations.1 ing tissue, and whether tissue is tracked by an electronic Surgeons relied on representatives of the tissue banks to system such as the laboratory information system (LIS) or ensure that the correct tissue was available for a surgery, by a paper system. In addition, quality assurance effort and the representatives transported tissue to the hospital, requirements should be considered (quality assurance con- arranged storage of the tissue before surgery, transported siderations are discussed further in an accompanying tissue to the operating room (OR), and maintained implant review5). Clearly, the necessary number of tissue service records. Thus, the responsibility for documentation and staff increases if they play a large role in tissue delivery, need verification of tissue acceptability was not in the hands of to travel a long distance to the OR, or must deliver tissue the hospital. Accumulating evidence of infectious disease multiple times per day. Also, if a large variety of tissues are transmission by tissue transplants and incidents of docu- handled with a tissue tracking and inventory system that is ment falsification led to much tighter regulation of human not part of the LIS, additional specially trained staff may be tissues by the US Food and Drug Administration (FDA).2,3 required. As the creation of a comprehensive tissue service The Joint Commission now also requires hospitals to estab- is often a response to increased regulations, it is likely that lish a centralized system to ensure that tissue is handled in the number of individuals who previously managed tissue a safe and efficient manner.4 The transfusion service has be- in the institution will be insufficient. Finally, consideration come the location for this system in many hospitals because should be given to designating a tissue bank coordinator to

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 107 B.M. Alden and A.J. Schlueter assist the medical director in the organization and oversight Table 1. Human surgical tissue inventory in a large teaching hospital of tissue activities. For example, at our institution, tissue setting bank personnel prepare and deliver an average of five tis- Number sues per day Monday through Friday (range, 1–10 tissues/ implanted Jan day) to ORs located approximately 5 to 10 minutes away Surgical Minimum 2009–June Tissues maintained in from the tissue bank. This level of activity necessitates min- service inventory 2009 inventory imum staffing levels of 1.5 full-time equivalents (FTE) for Cardiothoracic 20 3 Aortic valve (9 mm–30 mm) operations, 0.5 FTE at the supervisory level, and 0.5 FTE for 2 0 Hemi-pulmonary artery quality assurance. In addition, blood bank personnel cover 2 0 Pulmonary graft patch tissue deliveries during evenings and overnight, weekends, and holidays. This is an increase of approximately 1.5 FTE 20 1 Pulmonary valve (9 mm–30 mm) over the number of designated tissue handling staff before comprehensive tissue service was created. Dentistry/oral 2 1 Cancellous particles surgery Inventory Establishment and Maintenance 2 9 Cortical particles An initial task in creation of a tissue service is defining 8* 5 Iliac crest wedge the content of the standing inventory. Ideally, the standing (various sizes) inventory of tissue should be constructed to maximize qual- Neurosurgery 35* 42 Bone putty and paste ity patient outcomes while minimizing supply expense. The (various sizes) choice of tissues is influenced by the length of time needed 32* 16 Spinal bone spacers to obtain tissues from suppliers, surgeon preference, and (various types and sizes) information from tissue bank personnel. Over time, Ophthalmology 4 9 Amniotic membrane potential alterations to the standing inventory must be con- 2 2 Fascia lata strips sidered. Finally, the pros and cons of alternative financial 3 46 Sclera arrangements for tissue acquisition should be evaluated for Orthopedics 35* 186 Bone putty and paste each tissue. (various sizes) If there is a system by which surgeons can notify the 45 140 Cancellous bone chips tissue service in advance for elective cases (see next section, (various sizes) “Tissue Orders for Surgical Procedures”), the standing in- 20 61 Corticocancellous bone ventory need only include tissues required for emergency chips (various sizes) surgeries, unexpected situations in elective surgeries, and 6 3 Fascia lata (various sizes) tissues that are difficult to obtain quickly but might be 6 43 Femoral head needed for urgent cases. Minimal standing inventory may be determined by discussion with the surgeons to identify 12 17 Fibular and tibial cortical struts tissues deemed necessary for emergency use and identifi- 8* 2 Iliac crest wedge cation of commonly used tissues (ordered less frequently (various sizes) to minimize shipping and labor costs). As an example, the 16 74 Soft tissues (Achilles major components of the standing inventory at our institu- tendon, semitendinosus, tion, compared with the usage patterns during a 6-month patellar ligament, anterior period, are outlined in Table 1. Most allografts can usually tibialis) be obtained from tissue banks within 24 to 48 hours, but oc- 32* 28 Spinal bone spacers casionally they are back-ordered. However, the demand for (various types and sizes) certain tissues, such as pediatric cardiac valves, routinely 2 2 Whole long bones exceeds the supply. For such tissues, maintaining a range of Otolaryngology 10* 37 Acellular dermal matrix (vari- sizes to ensure availability may necessitate purchase when- ous sizes and thicknesses) ever they become available. Trauma/burn/ 10* 2 Acellular dermal matrix (vari- Maintaining a standing tissue inventory based on in- general surgery ous sizes and thicknesses) dividual surgeon preference requires a large capital invest- 30,000– 56,473 cm2 Cryopreserved skin ment. It is therefore worthwhile developing a consensus 40,000 (meshed and nonmeshed) among surgeons (including across surgical specialties, if cm2 possible) on the types of tissues stocked. For example, lim- Vascular 1 0 Femoral popliteal iting the types of bone putties in inventory or the tissue surgery artery—blood type O banks from which they are purchased can make the inven- 1 0 Saphenous vein— tory more manageable and reduce expenses. Tissue bank blood type O sales and technical representatives are a valuable resource *Minimum inventory for these items is shared between two or more for the surgeon and the tissue service. Their experience can services (listed in duplicate in the table).

108 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Logistics of hospital tissue management be useful in establishing par levels in the tissue inventory, disadvantage to loaner arrangements is the labor involved and their knowledge of impending supply limitations may in tracking incoming and outgoing tissues. Generally this assist in avoiding back orders. Nevertheless, it is imperative arrangement is only available for tissues stored at ambient that requests for acquisition of specific tissues come from temperatures, but some tissue banks allow frozen or cry- the surgeons to ensure there is no miscommunication opreserved tissues to be returned if the original shipping regarding what they intend to use for their patients. container is not opened and the tissue is returned within a New products or surgical staff may necessitate an in- designated time period. ventory adjustment. A centralized product evaluation committee, consisting of surgeons as well as representatives Tissue Orders for Surgical Procedures of hospital procurement and tissue services, can assist in A reliable communication system between the surgical determining the medical benefit and financial impact of ad- and tissue services is necessary to ensure that tissues need- ditional inventory items or replacement of existing tissues. ed for each case are available at the time of the procedure. It is also necessary to ascertain whether the new tissue is Tissue orders are placed for items from the standing inven- handled by a qualified supplier (for more details, see ac- tory or for noninventory items that in some cases must be companying tissue service quality assurance review5) and matched for the patient’s size. Important elements of the whether the tissue has FDA approval.6 If the tissue is not ordering process include routes and timing of communica- FDA approved, institutional review board approval may be tion, minimum necessary information identifying the tissue required by the hospital. and the patient, and notification of unexpected events be- Tissues can be obtained by direct purchase, consign- fore the surgery (e.g., rescheduled or cancelled cases, delays ment, and loaner arrangements. For commonly used tissues in tissue availability). with a long shelf life (often 2–5 years), a purchase arrange- Orders can be submitted to the tissue service electroni- ment is a viable option. Advantages include control of the cally, by paper, by phone call, or by a combination of these inventory and the ability to negotiate prices based on vol- approaches. Minimum basic information needed to fill the ume. The disadvantage is that most tissues, once purchased, order includes patient identification, requesting surgeon, are not returnable—especially refrigerated and frozen tis- tissue type and quantity, and scheduled surgical date. If the sues. If a new product becomes the surgeons’ preference or surgical date has not been set at the time the order is placed, if there is a turnover in surgical staff, a purchased product a mechanism should exist to communicate this date when it may no longer be implanted as frequently (or at all), leading is determined. Precise identification of the desired tissue is to expiration in inventory. Some companies will establish facilitated by establishing a uniform nomenclature for tis- exchange or return programs for tissue stored at ambient sue requests for use by both the surgical team and tissue temperature, if the tissue is a year or more from the expira- service personnel. A Web site or shared computer file can tion date, and if the exchange is for a similar product. provide detailed descriptions of unique tissues, for exam- Consignment agreements allow tissue to be stored at the ple, specific measurements and any flaws present in cardiac hospital, but only the tissue that is actually used is billed to valves or vessels. The creation of kits containing multiple the hospital. Some tissue banks offer to monitor the tissue sizes of the same tissue type that are usually ordered inventory, replace tissues close to expiration, and maintain together (e.g., bones needed for spinal surgeries) expedites documentation, either manually or via software programs. orders for such tissues. Allowing orders to be submitted by Disadvantages to consignment arrangements include po- a proxy (nurse, physician assistant, or secretary) may ex- tential increased tissue service labor costs if the consign- pedite rapid receipt of the order by the tissue service; how- ment inventory is managed by tissue service personnel. ever, requiring surgeon verification of tissue orders places This practice increases replacement tissue shipments and is responsibility for their accuracy at the highest level. accompanied by additional inventory documentation. Also, A minimum notification time (the time from when the some tissue banks charge higher prices for tissues that are tissue order is received to the scheduled surgery date) must brought in on consignment. be established based on the anticipated time needed to pur- Loaner arrangements are a third tissue acquisition op- chase or reserve tissue. If the tissue service is staffed 24/7 it tion. Typically, the loaner tissue is shipped to the hospital may be possible to receive orders and reserve tissues for the on the day before or the day of surgery, and any unused patient from the standing inventory even on the second or tissue is returned immediately after the surgery. As with third shift. For tissues not in inventory, order placement by consignment tissue, only loaner tissue that is actually used the surgical team 48 hours before surgery generally ensures is billed to the hospital. This is an attractive alternative for adequate time to receive tissue that a distributor routinely tissues used infrequently or when many sizes of a tissue carries in its inventory. On the opposite end of the spec- must be readily available to allow selection of the precise size trum, it is not always fiscally sound to purchase or reserve needed during the procedure, as for some neurosurgery cases. tissue for a patient as soon as the tissue order is received, It also minimizes the amount of tissue stored at the hospital, owing to potential changes in or cancellation of surgical which reduces labor-intensive inventory reconciliation. A dates. However, if tissue orders are placed as soon as the

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 109 B.M. Alden and A.J. Schlueter surgeons know that tissue is needed for a patient, any spe- tissues from one tissue bank require storage at <–135°C, cial requests or inventory issues can be dealt with in a me- whereas a similar tissue from another supplier calls for thodical rather than an emergent manner. One to 2 weeks storage at ≤–120°C. Some tissue banks offer to provide before the surgery is generally an adequate time frame storage devices for their tissues. If the tissue service must within which to reserve tissue for a particular patient. identify additional storage units, consideration should be A system that notifies the tissue service of changed given to using extra capacity in units that the transfusion or cancelled surgery dates helps to ensure that tissues are service uses for blood component storage. As with blood available by the appropriate date for specific patients and component storage units, tissue storage unit temperatures released or returned if the surgery is cancelled. Conversely, must be continuously monitored to ensure that temperature a method by which surgical teams can be notified of antici- ranges are not exceeded.4 Because equipment problems will pated tissue arrival dates and back-ordered tissue allows occasionally be encountered, it is imperative that backup surgeries to be rescheduled if necessary until tissue is avail- storage locations be available for all storage temperatures. able. Alarm settings on storage units should be set so there is Additional considerations apply to tissue orders for adequate time to move tissues to a new storage location if surgeries in which tissues that are in short supply must be needed. sized for the particular patient. In these situations, the sur- Designated storage areas should be established, into geon must supply the dimensions of the tissue needed or which tissues are placed after being reserved for specific send an x-ray or scan to be measured. In some instances patients, to ensure that adequate tissue is available for each these materials are submitted directly to the tissue bank, surgical patient. Moving reserved tissue to such locations which then supplies the surgeon with information on one or helps to identify which tissues need to be ordered to main- more potentially acceptable grafts. Identification of some of tain baseline standing inventory levels. these tissues requires weeks to months, and a short expiration date subsequently requires rapid implantation. Other Transportation and Delivery to Surgical Areas sized tissues are identified and available well in advance of Transport within the hospital may be defined as the the scheduled surgery date. The benefits of receiving such time required for tissue delivery to the OR, time spent in a tissues as soon as they are identified must be balanced holding area near the surgical area just before surgery or in against the risk of surgical cancellations and subsequent the OR during surgery, and return from the OR (if the tissue tissue wastage, as the tissue is usually not returnable. This is unused). Factors to consider when establishing a trans- latter outcome not only is fiscally undesirable, but prevents portation and delivery system include determining where another patient elsewhere from using this unique resource. tissue needs to be transported within the institution, how to maintain the designated storage temperature of each tissue, Tissue Acquisition and Storage and what information must accompany the tissue. Proce- Whether tissue is acquired to replace items in stand- dures using tissue occur within traditional ORs, in outpa- ing inventory or specifically ordered for certain patients, it tient minor procedure rooms, or on inpatient units (e.g., should be verified to be intact and at the appropriate tem- neonatal intensive care). Personnel available to transport perature on arrival and be handled and stored following the tissue to these destinations and distances from the tissue manufacturer’s directions. Direct shipment of tissue from storage location are variable. If tissue is stored remotely, the distributor to the tissue service (rather than transport delivery is facilitated by identification of a secure location in by the sales representative to the surgical suite) allows the or close to the areas of high tissue usage from which tissue tissue service to verify that tissues are received within ac- orders for part or all of a shift can be dispensed. Vending ceptable parameters and tracked continuously after receipt machine-type systems, similar to drug-dispensing systems, at the facility. (Tissue tracking is discussed in detail in the can be stocked with tissues that are dispensed to surgi- accompanying tissue service quality assurance review.5) To cal staff when the correct patient information is scanned. avoid inadvertent release of potentially compromised tis- These devices can provide substantial long-term savings sues, a quarantine area should be used to sequester tissues by minimizing transport time. The system should have a before verification of their acceptability on receipt from the mechanism that requires bar code scanning to track the vendor (or return from a patient care area). Tissue banks identification of the staff person obtaining the tissue, the supply handling and storage directions with each tissue. tissue itself, and the recipient. Bar codes on tissue packages These are located on the exterior packaging or a package are not standardized, and it may be necessary to print new insert and may change between lots of tissue from the same bar code labels that can be placed on the tissue package. supplier. These bar codes must be linked to the tissue bank’s original Tissues may be stored at ambient (or room) tempera- identifier. ture, refrigerated, frozen, or cryopreserved (≤–120°C).3 Tis- Tissue must remain within the tissue bank’s designated sues stored within the same general temperature range often storage temperature range at all times during transport.2 have slightly different limits. For example, cryopreserved Therefore, all containers used to transport tissue must be

110 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Logistics of hospital tissue management validated to maintain the desired temperature for at least Conclusion the length of time needed for the entire transport process. New FDA regulations and Joint Commission standards A validated period longer than the predicted transport time have increased the incentive for hospitals to centralize the is desirable to accommodate potential problems, such as handling of surgical tissues. The hospital transfusion ser- “lost” containers or weekend or holiday staffing, with less vice is often a logical place for this function. The transfusion frequent container returns to the tissue service. For ambi- service staff is already familiar with the processes for stor- ent temperature transport, most any container large enough ing, dispensing, and returning blood that are very similar to hold the tissue will suffice. For refrigerated or frozen tis- to those needed for tissue and thus possess unique skills sues, coolers or dry ice chests are needed. Cryopreserved for rapid development of a high-quality tissue service. The tissues require a Dewar or cryoshipping container that can increase in tissue regulations also encourages a different be charged with liquid nitrogen or that contains charged in- view of tissue within the hospital, one that emphasizes the serts that will maintain the proper temperature. limited nature of the resource as well as its potential for A copy of instructions for handling the tissue, the tissue transmission of infectious disease, and therefore the need specifications, and any available donor information must to manage it in a controlled, traceable manner. Coordinat- accompany the tissue to the surgical area.4 The date, the ing tissue services with transfusion service activities also time, and the personnel dispensing and receiving the tis- emphasizes this aspect of tissue handling and encourages sue must be documented. If the patient name and surgical similar practices on the part of the users of tissue. location are attached to the outside of each container when the tissue is dispensed, patient confirmation by both parties Acknowledgments can be made without opening the container. Accurate docu- We are grateful to Judith Levitt for critical reading of mentation must also occur of the patient into which the tis- the manuscript. sue was implanted and whether some tissue was wasted. References Return of Unused Tissue 1. Eisenbrey AB, Strong DM. Tissue banking in the hospital Before accepting a dispensed tissue back into invento- setting. In: Hillyer CD, Silberstein LE, Ness PM, Anderson ry, the tissue service must determine that the tissue meets KA, Roback JD, eds. Blood banking and transfusion certain criteria. These include maintenance of the appropri- medicine: scientific basis and clinical practice. 2nd ed. ate temperature and packaging integrity. Philadelphia: WB Saunders, 2006. Maintenance of proper temperature can initially be 2. Current good tissue practice for human cell, tissue, and assumed if the tissue transport container has been returned cellular and tissue-based product establishments; in- within the time limit for which it was validated. If there is spection and enforcement; final rule. Federal register any reason to suspect that the contents of a container returned 2004;69(226):68680–8. within the validated period are not at the appropriate tem- 3. Diconsiglio J. Keeping track of tissue in the sup- perature (e.g., the container might have been left open in ply chain. Materials Management in Health Care the OR), or if a container should be returned outside of 2008;17(10):30–2. the validated time frame, the internal temperature of the 4. The Joint Commission Accreditation Program, Hospi- transport container can be measured and tissues ac- tal Standards, 2009. cepted if the temperature is within the appropriate range. 5. Hillberry CM, Schlueter AJ. Quality activities associ- Packaging integrity can generally be determined by ated with hospital tissue services. Immunohematology verifying that the seals on the tissue packaging are intact. 2009;25:102–106. Many tissues are packaged in two containers to allow the 6. Recommended practices for surgical tissue banking. In: sterile inner package to be placed on the sterile field. As Perioperative standards, recommended guidelines and long as the inner package is intact the tissue remains sterile practices 2008. Denver, CO: AORN, 2008. and acceptable for use. The surgical staff must be queried before dispensing tissue that is missing the outer package, Beth M. Alden, BA, MT (ASCP), Lead Scientist, Tissue and to determine whether it is possible to aseptically remove the Cellular Therapies, and Annette J. Schlueter, MD, PhD tissue from the inner package. If not, the tissue must be dis- (corresponding author), Associate Professor, Department carded. of Pathology, University of Iowa, 200 Hawkins Drive, Multiple packages of similar tissues are sometimes Iowa City, IA 52242. simultaneously dispensed to an OR. If the packages do not have seals on the outer packaging that would allow detection of removal and replacement of the (nonsterile) inner tissue container and any of the tissue is returned, the tissue identifiers on the outer and inner packages should be checked to ensure that tissues were not inadvertently replaced into the wrong outer containers.

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 111 Review

Lewis blood group system review

M.R. Combs

Key Words: Lewis, Secretor, blood group, H, FUT2, FUT3 Le(a–b+) RBCs incubated with plasma from an Le(a–b–) donor. Mollison et al.13 demonstrated that the loss of Lewis The expression of Lewis blood group antigens depends on antigens may also occur in vivo. In 1969, Marcus and Cass14 the alleles inherited at two independent loci, FUT2 (Secre- confirmed that plasma Lewis antigens are glycosphingolip- tor [SE] gene) and FUT3 (Lewis [LE] gene). The Se and Le ids acquired by the RBC membrane from lipoproteins or alleles encode separate fucosyltransferases that interact aqueous dispersions. to form Lewis antigens in secretions and fluids. The Lewis For additional historic information and references, see antigens on RBCs are not integral to the membrane but the review by Watkins15 on the early investigations of gly- are passively adsorbed from the plasma. The antigens are coproteins that led to our current understanding of the ge- widely distributed in human tissue and fluids and are re- netic control, biosynthesis, and phenotypic interactions of ceptors for some pathogenic bacteria. Lewis antibodies are Lewis, ABO, H, and Secretor. rarely clinically significant, although there are rare reports of hemolytic transfusion reactions, hemolytic disease of the Genetics and Inheritance fetus and newborn, and renal transplant rejection. This re- The six antigens of the Lewis system are listed in view provides a general overview of the Lewis blood group Table 1. Expression of these antigens is determined by the system. An extensive overview by Daniels1 contains addi- alleles inherited at two independent loci, FUT2 (Secretor tional detailed information on the Lewis blood group sys- gene) and FUT3 (Lewis gene) Table 2. The FUT2 allele Se tem and related antigens. encodes a transferase that adds fucose to type 1 precursor chains in secretions and fluids to form type 1 H antigen. Be- History cause A and B terminal sugars may be added to type 1 H Anti-Lea was first described in 1946. Mourant2 reported chains, FUT2 also controls A and B antigen expression in a room temperature and 37°C directly agglutinating anti- secretions. The FUT2 allele se is nonfunctional. The FUT3 body in two women who delivered infants suspected of hav- allele Le encodes a different transferase that adds a fucose ing hemolytic disease of the fetus and newborn (HDFN). on type 1 precursor chains as well as on type 1 H chains. The The RBCs of one of the infants failed to react with the ma- FUT3 allele le is a nonfunctional allele. ternal serum; thus the antibodies were thought to be natu- Table 1. Lewis antigens rally occurring. This new antibody agglutinated 25 percent Antigen ISBT symbol ISBT No. of 96 RBC samples from English people. Two years later, Andresen3 described the original anti- Lea LE1 007001 b Leb LE2 007002 Le . This antibody only reacted with group O or A2, Le(b+) RBCs. Brendemoen4 later reported an anti-Leb that reacted Leab LE3 007003 with all Le(b+) RBCs regardless of ABO group. The two anti- LebH LE4 007004 bodies described by Andresen and Brendemoen were later ALeb LE5 007005 named anti-LebH and anti-LebL, respectively.5 BLeb LE6 007006 In 1948, Grubb6 observed the correlation between the Lewis blood groups and Secretor. Twenty Le(a+) adults were all nonsecretors of ABH substances and 41 of 42 Le(a–) Table 2. Lewis phenotypes, prevalence, and inherited alleles persons were secretors. The Lewis and Secretor loci were Prevalence Genotype later shown by family studies to be genetically independent.7 Grubb and Morgan8 first showed that Lewis substances are FUT2 Red cell (Secre- FUT3 Substances in present in secretions and serum, suggesting that Lewis phenotype Whites Blacks tor) (Lewis) secretions 9 was a system of saliva and plasma antigens. Furthermore, Le(a+b–) 22 23 sese LeLe or Lea Lewis antibodies could be neutralized by these soluble sub- Lele 6,10,11 12 stances. In 1955, Sneath and Sneath suggested that Le(a–b+) 72 55 Sese or LeLe or Lea, Leb, ABH the fundamental expression of Lewis antigens is in the SeSe Lele plasma and that RBCs simply adsorb the Lewis antigens Le(a–b–) 6 22 Sese or lele ABH in vivo. When Le(a+b–) donor RBCs were transfused to SeSe an Le(a–b+) patient, donor cells separated by differential sese lele Type 1precursor agglutination were found to be Le(a+b+). The transfor- Le(a+b+) Rare* Rare* SewSew or LeLe or Lea, Leb, ABH mation could also be reproduced in vitro. In vitro studies Sewse Lele also showed the loss of Lewis antigens from Le(a+b–) and *Present in 10–40% of some Asian populations.

112 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Lewis system review

The phenotype Le(a+b–) is found in non- Fig. 1 Lewis antigen Type 1 Precursor Lea secretors (sese), i.e., individuals homozygous biosynthesis. Le for a nonfunctional FUT2 allele but who have A = A transferase; β β1,3 Gal 1,3 GlcNAc Gal GlcNAc inherited at least one Le allele (Lele or LeLe), B = B transferase; R R α1,4 which results in the fucosylation of type 1 Fuc = fucose; Se Fuc precursor chains to form Lea antigen (Fig. 1). Gal = galactose; GalNAc = N-acetylga- The FUT2 gene controls only H antigen ex- Type 1 H b lactosamine; Le pression in secretions, not Lewis; therefore, Le GlcNac = N-acetyl- β1,3 β1,3 Gal GlcNAc R Gal GlcNAc R individuals with the Le(a+b–) phenotype glucosamine; also have Lea in their saliva, but no H. α1,2 α1,2 α1,4 Le = Le transferase; Fuc Fuc Fuc The most common Lewis phenotype, R = upstream Le(a–b+), is attributable to inheritance of carbohydrate; at least one Se allele at the FUT2 locus (SeSe Se = Se transferase. A B or Sese) and one Le allele at the FUT3 locus (LeLe or Lele). The Le fucosyltransferase Type 1 A Type 1 B adds fucose to type 1 H chains, formed as a α1,3 β1,3 α1,3 β1,3 GalNAc Gal GlcNAc R Gal Gal GlcNAc R b result of the inheritance of Se, to form Le α1,2 α1,2 antigen (Fig. 1). Although small amounts of Fuc Fuc Lea antigen are formed from type 1 chain pre- Le Le cursors, Lea is rarely detectable on the RBCs. a b Le , Le , and H antigens are in the saliva, and, ALeb BLeb depending on ABO group, A and B antigens α1,3 β1,3 α1,3 β1,3 as well. GalNAc Gal GlcNAc R Gal Gal GlcNAc R The Le(a–b–) phenotype is found in in- α1,2 α1,4 α1,2 α1,4 dividuals homozygous for a nonfunctional Fuc Fuc Fuc Fuc Lewis gene (lele). Lewis antigens are absent from RBCs and saliva. These individuals may be secretors (Sese or SeSe) or nonsecretors (sese) of ABH. The Lewis gene (FUT3), located on the short arm of The Le(a+b+) phenotype is rare in Europeans but oc- chromosome 19,20 was cloned in 1990.25 An intronless cod- curs in 10 to 40 percent of some Asian populations.16 These ing region encodes the fucosyltransferase. Silent base sub- individuals have at least one Le allele and at least one weak stitutions that result in a functional enzyme and single base or inefficient Se allele (SewSew or Sewse). Sew results in par- substitutions that cause the Le(a–b–) phenotype have been tial or weak secretion of ABH.17 reviewed.22 Most single base substitutions that cause the Le(a–b–) phenotype are enzyme-inactivating, although one Molecular Basis results in an enzyme with altered substrate specificity.22 As In 1995 Rouquier et al. 18 cloned FUT2 and the pseudo- with FUT2, many FUT3 mutations show ethnic specificity. gene Sec1. These genes as well as the closely linked, highly For more information on FUT2 and FUT3 alleles, see www. homologous FUT1 (H gene), which determines H anti- ncbi.nlm.nih.gov/gv/mhc/xslcgi.cgi?cmd=bgmut/home.26 gen on RBCs, are located on the long arm of chromosome 19.19,20 The FUT2 gene consists of two exons, with exon 2 Biochemistry encoding the fucosyltransferase.21 Numerous FUT2 mu- The FUT2 allele Se encodes a transferase that adds a tations with ethnic associations have been described and terminal α1,2-linked fucose to galactose on type 1 precur- reviewed.22 Some mutations result in a functional FUT2 sor to form type 1 H antigen (Fig. 1). The FUT3 allele Le allele (Se), whereas a nonsense mutation, 428G>A (Trp- encodes a transferase that adds a subterminal α1,4-linked 143Stop) causes a common nonsecretor allele (se428) in fucose to the N-acetylglucosamine (GlcNAc) of type 1 chain Europeans, West Asians, and Africans.19,22 Other non- precursor to form Lea antigen. The Le enzyme may also add secretor alleles with ethnic specificity are caused by single α1,4 fucose to the GlcNAc of type 1 H antigen to form Leb and multiple base deletions in FUT2. A missense mutation antigen (Fig. 1). Both Le and Se transferases prefer and 385A>T (Ile129Phe) causes a common enzyme-deficient compete for type 1 chain substrates in secretions and fluids. Se allele (Sew385) that is responsible for the Le(a+b+) phe- RBC Lewis antigen determinants are plasma glycosphingo- notype found in East and Southeast Asians.23 Three null lipids that are passively adsorbed onto cell membranes,14 FUT2 alleles attributable to gene recombination have been whereas Lewis antigens in secretions are glycoproteins.27 observed. One is the sedel allele, which is almost always The Le(a+b–) phenotype occurs in nonsecretors (sese) associated with an inactive FUT1 allele (h). Individuals who have at least one Le allele. Nonsecretors do not make homozygous for these alleles (sedelsedel and hh) have the type 1 H antigen; therefore the Le transferase adds α1,4 fu- classic Indian Bombay phenotype.24 cose to type 1 chain precursor to form Lea antigen only (Fig 1).

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 113 M.R. Combs

In the Le(a–b+) phenotype, the Le transferase adds activity increases with age, these infants will ultimately type α1,4 fucose on type 1 H chain formed as a result of the Se Le(a–b+). By 1 year of age, 50 percent of children express transferase. The presence of both α1,2 fucose and α1,4 fu- their adult phenotype,35 and by age 2, the Lewis phenotype cose on GlcNAc results in Leb antigen expression. Although of most children will reflect their inherited FUT2 and FUT3 the same α1,4 fucose when added to type 1 precursor results alleles.36 Lewis antigens are detectable in neonatal saliva. in Lea antigen, Lea is not expressed on the α1,4 fucosylated As in adults, secretors have Lea and Leb, nonsecretors have type 1 H structure (Fig. 1). Most type 1 chain is converted Lea.35 to Leb and adsorbed on the RBC membrane, resulting in the Le(a–b+) RBC phenotype. However, small amounts Pregnancy of Lea antigen are made as a result of the addition of α1,4 Lewis-positive women may become transiently Le(a– fucose by the Le transferase to type 1 precursor chains be- b–) during pregnancy and may produce Lewis antibod- fore Se transferase is able to add α1,2 fucose. Once the α1,4 ies.29,37 Decreased expression of Lewis antigens during preg- fucose is added to form Lea, steric hindrance prevents fur- nancy was first reported by Brendemoen.38 The increased ther fucosylation by Se transferase (to form Leb) or A and incidence of the Le(a–b–) phenotype during pregnancy B transferases.28 Some monoclonal or potent anti-Lea may may be a result of increased concentration of plasma lipo- react with this adsorbed Lea antigen on apparent Le(a–b+) proteins during pregnancy. In pregnant women, the ratio RBCs.29 of lipoprotein to RBC mass increases more than fourfold so In the Le(a–b–) phenotype, Lewis antigens are not de- that much more Lewis glycolipid is attached to plasma lipo- tectable in saliva or on the RBCs. Nonsecretors (sese and protein than is available for the RBC surface. Leb glycolipid lele) produce only type 1 chain precursors. In secretors, in plasma only decreases slightly during pregnancy; there- type 1 H chains will be produced, which can be further fuco- fore, decreased expression of Leb on RBCs is not caused by sylated by A and B transferases (Fig. 1). decreased blood levels of Leb glycolipid.39 In the Le(a+b+) phenotype, Lea is formed at the expense of Leb owing to ineffective competition of the Sew gene prod- Lewis Antigen Distribution uct for type 1 precursor. This results in more Lea antigen Lewis antigens are widely distributed in the human and less Leb antigen than is found in individuals with a nor- body and are often called “histo-blood group antigens.” mal Se gene.17 Lewis antigens are found on the pancreas, stomach, and Because Le transferase can add fucose to type 1 H chain large and small intestine mucosa; skeletal muscle; renal as well as to type 1 A and type 1 B chains, additional Lewis cortex; and adrenal glands. See Ravn and Dabelsteen40 for determinants such as ALeb, BLeb, and LebH may be expressed an extensive review of the tissue distribution of histo-blood depending on an individual’s ABO group (Fig. 1).30,31 These group antigens. antigens are defined by single antibodies, not separable In addition to saliva and plasma, Lewis antigens are mixtures of antibodies. LebH is present on cells with strong present in other fluids such as human milk,41 urine, gas- 42 43 H antigen expression, i.e., group O and A2, Le(b+) RBCs. trointestinal juices and seminal fluid, ovarian cyst fluid, The determinant involves type 1 H antigen and the Le α1,4 and amniotic fluid.44 As with RBCs, Lewis antigens on lym- fucose.16 ALeb is formed when Le transferase adds fucose to phocytes14 and platelets45 are acquired from the plasma. type 1 A chains; BLeb is formed when Le transferase adds The gastrointestinal tract may be the primary source of fucose to type 1 B chains (Fig. 1). plasma Lewis antigens.46 This is evident in a report of eight Leab is a determinant expressed on Le(a+) or Le(b+) patients with intestinal failure and resections of the ileum adult RBCs. The antigen is also present on cord samples and or jejunum, all with Le(a–b–) RBCs.47 Recipients of from infants who have inherited Le, suggesting that the an- bone marrow,48 kidney,49 and liver transplants50 maintain tigen is formed early in embryonic development.16 The com- their own RBC Lewis phenotypes, suggesting that these mon determinant is within the α1,4 fucose added by the Le organs are not primary sources of plasma Lewis antigens. transferase.32 Antibodies Infants Lewis antibodies are most often found in individuals Most newborns type Le(a–b–) during the first month with Le(a–b–) RBCs; their sera may contain a mixture of of life, although Lewis antigens can sometimes be detected anti-Lea, -Leb, and -Leab. Anti-Lea is usually found in Le(a– on cord RBCs with more sensitive techniques33,34 or the use b–) individuals who are ABH secretors.51 Le(a–b+) individ- of anti-Leab. If Le is inherited, only very low levels of the Le uals do not make anti-Lea because small amounts of Lea are fucosyltransferase are present at the site of production of present. plasma Lewis antigens. The Le fucosyltransferase becomes There are two major types of anti-Leb. The most com- active before the Se fucosyltransferase; therefore Lea devel- mon, anti-LebH, reacts preferentially with Le(b+) RBCs with ops first, and RBCs may type as Le(a+b–) followed by a tran- stronger H antigen expression, such as group O or A2 RBCs. sient Le(a+b+) phenotype if Se is inherited.33,35 As secretor Anti-LebL reacts with all Le(b+) RBCs regardless of ABO

114 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Lewis system review group. Anti-Leb is usually found in Le(a–b–) individuals are frequently detected and may cross the placenta. Spital- who are nonsecretors of ABH.51 Anti-Leb is rarely produced nik et al. 55 detected IgG anti-Lea by hemagglutination in 13 by people with Le(a+b-) RBCs. These individuals are typi- of 13 maternal samples and in 12 of the 13 cord samples by bH 4, 52,53 cally group A1 or A1B with anti-Le . enzyme-linked immunosorbent assay. They concluded that Other Lewis antibodies include anti-ALeb and -BLeb the rare incidence of HDFN is attributable to poor expres- which react with the compound antigens on group A or B sion of Lewis antigens on fetal cells instead of the frequently RBCs. Anti-Leab occurs mainly in Le(a–b–) secretors who cited low incidence of IgG Lewis antibodies. are group A1, B, or A1B. This antibody reacts with Lewis- Two cases of mild HDFN caused by Lewis antibodies positive RBCs from adults as well as cord RBCs from in- have been reported. The first reported case attributable to fants with an Le allele.16 anti-Lea caused a positive DAT with anti-Lea in the eluate Most Lewis antibodies are naturally occurring IgM. and hyperbilirubinemia. The infant’s RBCs initially typed Some may have an IgG component,54–56 and there are rare as Le(a–b+), which was speculated to be caused by blocking examples of pure IgG Lewis antibodies.57 Although most of the antigen by the maternal antibody. Four days later the Lewis antibodies are naturally occurring, some may be RBCs typed Le(a+b+) The 42-week gestation was speculat- stimulated by RBC transfusion. Cheng and Lukomskyi58 ed to have allowed stronger development of Lea antigen, and reported on two patients who exhibited Lewis antibodies the mildness was thought to be caused by partial neutraliza- after massive RBC transfusions of presumably Lewis- tion of the maternal antibody by Lea substance in the fetal positive units. One patient exhibited IgG anti-Lea and the plasma.72 Neonatal jaundice developed in a case of HDFN other had IgM and IgG anti-Leb. There was no evidence of caused by anti-Leb. The antibody was IgM and IgG and was hemolysis in either patient. hemolytic in vitro. An eluate prepared from the newborn’s In patients with 37°C reactive anti-Lea or anti-Leb, the DAT-positive RBCs contained anti-Leb.73 antibody titer may rise after RBC transfusion.13,59,60 In spite of this, Lewis antibodies are rarely implicated in hemolytic Clinical Significance transfusion reactions. Anti-Lea is more frequently associ- Lewis antigens are receptors for pathogenic bacteria. ated with acute hemolytic transfusion reactions61-64 than is In particular, Leb and type 1 H mediate attachment of Helico- anti-Leb.65,66 Three cases of delayed hemolytic transfusion bacter pylori,74 a gram-negative bacterium associated with reactions (DHTR) have been claimed.67–69 Hemolytic trans- gastritis, gastric and duodenal ulcers, adenocarcinoma,75 fusion reactions are rare because most Lewis antibodies are and immune thrombocytopenic pupura.76 Leb and type 1 not active at 37°C, transfused RBCs lose their Lewis anti- H are also receptors for Norwalk virus that causes gastro- gens into the recipient’s plasma, and there is neutralization enteritis.77 Conversely, lack of Lewis antigens, i.e., the of recipient Lewis antibodies by Lewis substance in donor Le(a–b–) phenotype, is associated with an increased suscep- plasma before the antibodies can bind to the RBCs of the tibility to infections by Candida spp,78 and uropathogenic recipient.13,59 Escherichia coli.79 Because most Lewis antibodies are IgM, they react best The Le(a–b–) phenotype is also associated with an in- in agglutination tests at room temperature and occasion- creased risk of cardiovascular disease.80 The mechanism for ally in agglutination tests at 37°C. Reactivity with anti- this association is unclear, but it may be mediated through human globulin (AHG) may be related to IgG or caused elevated triglycerides,80 insulin resistance syndrome,81 or by bound complement if polyspecific AHG reagent is used. obesity,82 each of which is also associated with the Le(a–b–) Lewis antibodies may demonstrate complement activation phenotype. by causing in vitro lysis of Lewis-positive RBCs, especially The Le(a–b–), Bombay phenotype occurs in patients with the use of enzyme-treated RBCs or fresh serum. To with leukocyte adhesion deficiency syndromes (LADII). aid Lewis antibody detection and identification, plasma A mutation in the GDP-fucose transporter results in and saliva Lewis substances or commercial sources of solu- hypofucosylation of glycoproteins. Clinically, these pa- ble substances may be used to neutralize Lewis antibodies, tients have leukocytosis, severe infections, and mental and and enzyme-treated RBCs may be used to enhance Lewis growth retardation, thereby demonstrating the broad im- antibody reactivity. pact of glycoprotein fucosylation.83,84 Transfusion services vary in their selection of RBC The role of Lewis in renal transplantation is contro- units for patients with Lewis antibodies. Some provide versial. Oriol et al.85 reported that Lewis-negative renal al- antigen-negative units if Lewis antibodies are hemolytic lograft recipients have a significantly lower graft survival in vitro, agglutinating at 37°C or reacting in the IAT with rate than do Lewis-positive recipients. Other investigators anti-IgG. In general, however, the selection of antiglobulin have found that renal transplant recipients who receive a crossmatch-compatible blood for patients with Lewis anti- Lewis-matched kidney have the best survival.86–88 Although bodies is recommended and considered safe.70,71 these reports suggest that cytotoxic Lewis antibodies may Lewis antibodies rarely cause HDFN. Although most play a role in renal transplant survival, other investigators are IgM and cannot cross the placenta, IgG Lewis antibodies have reported that Lewis phenotype compatibility does not

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 115 M.R. Combs affect survival.89,90 Detectable Lewis antibodies in renal 16. Reid ME, Lomas-Francis C. The blood group antigen transplant recipients, however, have been associated with factsbook. 2nd ed. San Diego, CA: Academic Press, antibody-mediated graft rejection.91-93 2004. In summary, the Lewis blood group system is inter- 17. Henry SM, Benny AG, Woodfield DG. Investigation of esting and complex; the antigens are widely distributed in Lewis phenotypes in Polynesians: evidence of a weak se- human tissue and fluid; and the expression of most Lewis cretor phenotype. Vox Sang 1990;58:61–6. antigens requires the interaction of transferases from more 18. Rouquier S, Lowe JB, Kelly RJ, et al. Molecular cloning of than one blood group system. Lewis antibodies are rarely a human genomic region containing the H blood group clinically significant; however, there is growing evidence α(1,2)fucosyltransferase gene and two H locus-related that the presence or absence of Lewis antigens in an indi- DNA restriction fragments: isolation of a candidate gene vidual can be clinically significant, and this is the focus of for the human Secretor blood group locus. J Biol Chem much of the current research and investigation involving 1995;270:4632–9. the Lewis blood group system. 19. Kelly RJ, Rouquier S, Giorgi D, et al. Sequence and expression of a candidate for the human secretor blood References α(1,2)-fucosyltransferase gene (FUT2). J Biol Chem 1. Daniels G. Human blood groups. 2nd ed. Oxford: Black- 1995;270:4640–9. well Science, 2002. 20. Reguigne-Arnould I, Couillin P, Mollicone R, et al. 2. Mourant AE. A ‘new’ human blood group antigen of fre- Relative positions of two clusters of human α-L- quent occurrence. Nature 1946;158:237-8. fucosyltransferases in 19q (FUT1-FUT2) and 19p (FUT6- 3. Andresen PH. The blood group system L: a new blood FUT3-FUT5) within the microsatellite genetic map of group L2. A case of epistasy within the blood groups. chromosome 19. Cytogenet Cell Gen 1995;71:158–62. Acta Path Microbiol Scand 1948;25:728–31. 21. Koda Y, Soejima M, Kimura H. Structure and ex- 4. Brendemoen OJ. Further studies of agglutination pression of the gene encoding secretor-type galacto- and inhibition in the Lea-Leb system. J Lab Clin Med side 2-α-L-fucosyltransferase (FUT2). Eur J Biochem 1950;36:335–41. 1997;246:750–5. 5. Ceppellini R, Dunn LC, Filomena I. Immunogenetica II. 22. Soejima M, Koda Y. Molecular mechanisms of Lewis an- Analisi genetica formale de caratteri Lewis con partico- tigen expression. Leg Med (Tokyo) 2005;7:266–9. lare riguardo alla natura epistatica della specificita sero- 23. Henry S, Mollicone R, Fernandez P, et al. Molecular ba- logica Leb. Fol Hered Path 1959;8:261–96. sis for erythrocyte Le(a+b+) and salivary ABH partial- 6. Grubb R. Correlation between Lewis blood group and se- secretor phenotypes: expression of a FUT2 secretor allele cretor character in man. Nature 1948;162:933. with an A→T mutation at nucleotide 385 correlates with 7. Sanger R, Race RR. The Lutheran-secretor linkage in reduced alpha(1,2) fucosyltransferase activity. Glycoconj man: support for Mohr’s findings. Heredity 1958;12:513– J 1996;13:985–93. 20. 24. Koda Y, Soejima M, Johnson PH, et al. Missense muta- 8. Grubb R, Morgan WTJ. The “Lewis” blood group char- tion of FUT1 and deletion of FUT2 are responsible for acters of erythrocytes and body fluids. Br J Exp Pathol Indian Bombay Phenotype of ABO blood group system. 1949;30:198-208. Biochem Biophys Res Commun 1997;238:21-5. 9. Grubb R. Some aspects of the complexity of the blood 25. Kukowska-Latallo JF, Larsen RD, Nair RP, Lowe JB. groups ABO. Rev Hematol 1950;5:268–75. A cloned human cDNA determines expression of a 10. Grubb R. Observations on the human group system mouse stage-specific embryonic antigen and the Lewis Lewis. Acta Path Microbiol Scand 1951;28:61–81. blood group α(1,3/1,4)fucosyltransferase. Genes Dev 11. Brendemoen OJ. Studies of agglutination and inhibition 1990;4:1288–303. in two Lewis antibodies. J Lab Clin Med 1949;34:538– 26. Blumenfeld OO, Patnaik SK. Allelic genes of blood group 42. antigens: a source of human mutations and cSNPs docu- 12. Sneath JS, Sneath PHA. Transformation of the Lewis mented in the Blood Group Antigen Gene Mutation Da- groups of human red cells. Nature 1955;176:172. tabase. Hum Mutat 2004;23:8–16. 13. Mollison PL, Polley MJ, Crome P. Temporary suppres- 27. Watkins WM. Blood-group substances. Science sion of Lewis blood-group antibodies to permit incom- 1966;152:172–81. patible transfusion. Lancet 1963;1:909–12. 28. Kobata A, Grollman EF, Ginsberg V. An enzymatic ba- 14. Marcus DM, Cass LE. Glycosphingolipids with Lewis sis for blood type A in humans. Arch Biochem Biophys blood group activity: uptake by human erythrocytes. Sci- 1968;124:609–12. ence 1969;164:553–5. 29. Issitt PD, Anstee DJ. Applied blood group serology. 4th 15. Watkins WM. A half century of blood-group antigen ed. Durham, NC: Montgomery Scientific Publications, research: some personal recollections. Trends Glycosci 1998. Glycotechnol 1999;11:391–411.

116 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Lewis system review

30. Clausen H, Levery SB, McKibbon JM, Hakomori S. Blood glycosphingolipids from human plasma. Arch Biochem group A determinants with mono- and difucosyl type 1 Biophys 1981;210:383–95. chain in human erythrocyte membranes. Biochemistry 47. Ramsey G, Fryer JP, Teruya J, Sherman LA. Lewis (a-b-) 1985;24:3578–86. red cell phenotype in patients undergoing evaluation for 31. Breimer ME, Karlsson K-A, Samuelsson BE. Character- small intestinal transplantation. (abstract) Transfusion ization of a human intestinal difucosyl heptaglycosylce- 2000;40(Suppl):114S. ramide with a blood group B determinant and a type 1 48. Oriol R, Le Pendu J, Sparkes RS, et al. Insights into carbohydrate chain. J Biol Chem 1982;257:1079–85. the expression of ABH and Lewis antigens through hu- 32. Schenkel-Brunner H. Human blood groups: chemical man bone marrow transplantation. Am J Hum Genet and biochemical basis of antigen specificity. 2nd ed. New 1981;33:551–60. York: Springer-Verlag, 2000. 49. Oriol R, Cartron JP, Cartron J, Mulet C. Biosynthesis 33. Cutbush M, Giblett ER, Mollison PL. Demonstration of of ABH and Lewis antigens in normal and transplanted the phenotype Le(a+b+) in infants and in adults. Br J kidneys. Transplantation 1980;29:184–8. Haematol 1956;2:210–20. 50. Dzik WH, Mondor LA, Maillet SM, Jenkins RL. ABO 34. Lin M, Shieh SH. Postnatal development of red cell and Lewis blood group antigens of donor origin in the Lea and Leb antigens in Chinese infants. Vox Sang bile of patients after liver transplantation. Transfusion 1994;66:137–40. 1987;27:384–7. 35. Lawler SD, Marshall R. Lewis and Secretor characters in 51. Miller EB, Rosenfield RE, Vogel P, et al. The Lewis blood infancy. Vox Sang 1961;6:541–4. factors in American Negroes. Am J Phys Anthropol 36. Oriol R. ABO, Hh, Lewis and secretions: serology, genet- 1954;12:427–43. ics and tissue distribution. In: Cartron JP, Rouger P, eds. 52. Garratty G, Kleinschmidt G. Two examples of anti-Leb Blood cell biochemistry: molecular basis of major blood detected in the sera of patients with the Lewis phenotype group antigens. New York: Plenum, 1994:37–73. Le(a+b-). Vox Sang 1965;10:567-71. 37. Kissmeyer-Nielsen F. Irregular blood group antibodies 53. Kornstad L. Anti-Leb in the serum of Le(a+b-) and in 200,000 individuals. Scand J Haematol 1965;2:331– Le(a-b-) persons: absorptions studies with erythro- 42. cytes of different ABO and Lewis phenotypes. Vox Sang 38. Brendemoen OJ. Some factors influencing Rh immu- 1969;16:124-9. nization during pregnancy. Acta Path Microbiol Scand 54. Spitalnik S, Cowles J, Cox MT, et al. A new technique 1952;31:579–83. in quantitative immunohematology: solid-phase kinetic 39. Hammar L, Mansson S, Rohr T, et al. Lewis phenotype ELISA. Vox Sang 1983;45:440–8. of erythrocytes and Leb-active glycolipid in serum of 55. Spitalnik S, Cowles J, Cox MT, Blumberg N. Detection of pregnant women. Vox Sang 1981;40:27–33. IgG anti-Lewisa antibodies in cord sera by kinetic ELISA. 40. Ravn V, Dabelsteen E. Tissue distribution of histo-blood Vox Sang 1985;48:235–8. group antigens. APMIS 2000;108:1–28. 56. Molthan L, Strohm PL, Gross BM, Paradis DJ. Frequen-

41. Watkins WM, Morgan WT. Specific inhibition stud- cies and immunoglobulin classes of the Lewis, P1, and ies relating to the Lewis blood-group system. Nature MN system antibodies. Lab Med 1983;14:422–6. 1957;180:1038–40. 57. Mollison PL, Engelfriet CP, Contreras M. Blood transfu- 42. McConnell RB. Lewis blood group substances in body sion in clinical medicine. 10th ed. Oxford: Blackwell Sci- fluids. Istituto GM (ed.) Proc 2nd Congr Hum Genet, ence, 1997. Rome, Italy 1961:858–61. 58. Cheng MS, Lukomskyi L. Lewis antibody following a 43. Morgan WT. A contribution to human biochemical ge- massive blood transfusion. Vox Sang 1989;57:155–6. netics; the chemical basis of blood-group specificity. 59. Hossaini AA. Neutralization of Lewis antibodies in vivo Proc R Soc Lond B Biol Sci 1960;151:308–47. and transfusion of Lewis incompatible blood. Am J Clin 44. Arcilla MB, Sturgeon P. Lewis and ABH substances in Pathol 1972;57:489–93. amniotic fluid obtained by amniocentesis. Pediatr Res 60. Holburn AM. Quantitative studies with (125I) IgM anti- 1972;6:853–8. Lea. Immunology 1973;24:1019–26. 45. Dunstan RA, Simpson MB, Rosse WF. Lea blood 61. de Vries SI, Smitskamp HS. Haemolytic transfu- group antigen on human platelets. Am J Clin Pathol sion reactions due to anti-Lewisa agglutinin. Br Med J 1985;83:90–4. 1951;1:280–1. 46. Hanfland P, Graham HA. Immunochemistry of the 62. Brendemoen OJ, Aas K. Hemolytic transfusion reac- Lewis-blood-group systems: partial characterization tion probably caused by anti-Lea. Acta Med Scand of Lea-, Leb-, and H-Type 1(LedH)-blood-group active 1952;141:458–60.

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 117 M.R. Combs

63. Roy RB, Wesley RH, Fitzgerald JD. Haemolytic transfu- 80. Hein HO, Sorensen H, Suadicani P, et al. The Lewis sion reaction caused by anti-Lea. Vox Sang 1960;5:546– blood group—a new genetic marker of ischaemic heart 50. disease. J Intern Med 1992;232:481–7. 64. Mollison PL, Cutbush M. Use of isotope-labeled red 81. Clausen JO, Hein HO, Suadicdani P, et al. Lewis phe- cells to demonstrate incompatibility in vivo. Lancet notypes and the insulin resistance syndrome in young 1955;268:1290–5. healthy white men and women. Am J Hypertens 65. Quiroga H, Leite A, Baia F, et al. Clinically significant 1995;8:1060–6. anti-Leb.(abstract) Vox Sang 2000;78(Suppl 1):P125. 82. Hein HO, Suadicani P, Gyntelberg F. The Lewis blood 66. Jesse JK, Sheek KJ. Anti-Leb implicated in acute hemo- group—a new genetic marker of obesity. Int J Obes lytic transfusion reaction: a rare occurrence. (abstract) (Lond) 2005;29:540–2. Transfusion 2000;40(Suppl):115S. 83. Hirschberg CB. Golgi nucleotide sugar transport and leu- 67. Pineda AA, Taswel HG, Brzica SM. Delayed hemolytic kocyte adhesion deficiency II. J Clin Invest 2001;108:3– transfusion reaction. An immunological hazard of blood 6. transfusion. Transfusion 1978;18:1–7. 84. Luhn K, Wild MK, Eckhardt M, et al. The gene defective 68. Weir AB, Woods LL, Chesney C, Neitzer G. Delayed in leukocyte adhesion deficiency II encodes a putative hemolytic transfusion reaction caused by anti-LebH anti- GDP-fucose transporter. Nat Genet 2001;28:69–72. body. Vox Sang 1987;53:105–7. 85. Oriol R, Opelz G, Chun C, Terasaki PI. The Lewis sys- 69. Contreras M, Mollison PL. Delayed haemolytic trans- tem and kidney transplantation. Transplantation fusion reactions caused by anti-LebH. Vox Sang 1980;29:397–400. 1989;56:290. 86. Oriol R, Cartron J-P, Yvart J, et al. The Lewis system: 70. Waheed A, Kennedy MS, Gerhan S, Senhauser DA. new histocompatibility antigens in renal transplanta- Transfusion significance of Lewis system antibodies. Am tion. Lancet 1978;1:574-5. J Clin Pathol 1981;76:294–8. 87. Salmon C, Cartron J-P, Rouger P. The human blood 71. Poole J, Daniels G. Blood group antibodies and their groups. New York: Masson, 1984. significance in transfusion medicine. Transfus Med Rev 88. Fischer E, Lenhard V, Romer W, et al. Influence of Lewis 2007;21:58–71. blood group system on clinical kidney transplantation. 72. Carreras Vescio LA, Torres OW, Virgilio OS, Pizzolato M. Proc Eur Dial Transplant Assoc 1979;16:377–82. Mild hemolytic disease of newborn due to anti-Lewisa. 89. Posner MP, McGeorge MB, Mendez-Picon G, et al. The Vox Sang 1993;64:194–5. importance of the Lewis system in cadaver renal trans- 73. Bharucha ZS, Joshi SR, Bhatia HM. Hemolytic disease of plantation. Transplantation 1986;41:474–7. the newborn due to anti-Leb. Vox Sang 1981;41:36–9. 90. Gratama JW, Hendriks GF, Persijn GG, et al. The in- 74. Boren T, Falk P, Roth KA, et al. Attachment of Helico- teraction between Lewis blood group system and bacter pylori to human gastric epithelium mediated by HLA-matching in renal transplantation. Transplanta- blood group antigens. Science 1993;262:1892–5. tion.1988;45:926–9. 75. Goodwin CS, Mendall MM, Northfield TC. Helicobacter 91. Williams G, Pegrum GD, Evans CA. Lewis antigens in re- pylori infection. Lancet 1997;349:265–9. nal transplantation. Lancet 1978;1:878. 76. Franchini M, Veneri D. Helicobacter pylori infection and 92. Spitalnik S, Pfaff W, Cowles J, et al. Correlation of immune thrombocytopenic purpura: an update. Helico- humoral immunity to Lewis blood group antigens bacter 2004;9:342–6. with renal transplantation rejection. Transplantation 77. Hutson AM, Atmar RL, Marcus DM, Estes MK. Norwalk 1984:37:265–8. virus-like particle hemagglutination by binding to H 93. Boratynska M, Banasik M, Halon A, et al. Blood group histo-blood group antigens. J Virol 2003;77:405–15. Lewis alloantibodies cause antibody-mediated rejec- 78. Hilton E, Chandrasekaran V, Rindos P, Isenberg HD. tion in renal transplant recipients. Transplant Proc Association of recurrent candidal vaginitis with in- 2007;39:2711–14. heritance of Lewis blood group antigens. J Infect Dis 1995;172:1616–19. Martha Rae Combs, MT(ASCP)SBB, Analytical Specialist, 79. Stapleton A, Nudelman E, Clausen H, et al. Binding of uro- Technical Director of Immunohematology, Duke Universi- pathogenic Escherichia coli R45 to glycolipids extracted ty Hospital, Transfusion Service, Box 2928, Room 1720N, from vaginal epithelial cells is dependent on histo-blood 2424 Erwin Road, Durham, NC 27710. group secretor status. J Clin Invest 1992;90:965–72.

118 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Review

Recognition and management of antibodies to human platelet antigens in platelet transfusion–refractory patients

R.R. Vassallo

Platelet transfusion refractoriness is a problem for parous and mul- is playing a role in poor 24-hour platelet count increments.7 tiply transfused patients, placing them at higher risk for mor- Because antibody-mediated destruction occurs rapidly, bidity and mortality when posttransfusion count increments are platelet recovery less than half of that expected within 1 significantly lower than expected. Although nonimmune causes of hour of transfusion suggests an immunologic component to transfusion refractoriness are very common, HLA alloantibodies are platelet transfusion refractoriness. Yankee et al.8 elegantly the most important of the less frequent immune factors responsible for inadequate count increments. As universal leukoreduc- demonstrated that immune-based platelet transfusion re- tion decreases the occurrence of HLA antibody formation, antibodies fractoriness is mediated primarily by alloantibodies to HLA to human platelet antigens (HPAs), an even less common immune class I determinants present on the platelet surface. Alloan- factor, may rise proportionately. Carefully matched apheresis tibodies to human platelet antigens (HPAs) are an impor- platelets can substantially improve platelet count increments in tant but less common cause of immune platelet transfusion the setting of HLA and HPA alloantibody-mediated transfusion refractoriness.1,6,9 Transfused platelet destruction by HLA refractoriness. An evidence-based HPA testing strategy is and HPA alloantibodies is potentially avoidable by the use described along with the incidence and specificity of HPA anti- of products from appropriately matched donors. Immune bodies in platelet transfusion refractoriness. Optimal strategies to destruction caused by drug-dependent platelet antibod- manage patients with HPA or combined HPA and HLA antibodies are presented. Ultimately, close cooperation between ordering ies is quite uncommon and the subject of another review 10 physicians and the blood provider is critical in choosing the in this issue of Immunohematology. Panreactive platelet correct tests and assuring platelet availability during intensive autoantibodies may also result in platelet transfusion re- support of these challenging patients. fractoriness. The diagnosis of autoimmune platelet destruc- Immunohematology 2009;25:119–124. tion is generally known before treatments that predictably worsen patients’ baseline thrombocytopenia are initiated. latelet transfusion refractoriness, defined as inappro- As we will see, however, transient de novo production of priately low platelet count increments 24 hours after autoantibodies may be observed with exquisitely sensitive Ptwo consecutive ABO-identical transfusions, is de- techniques in a significant percentage of platelet transfu- clining in frequency after the introduction of near-universal sion-refractory individuals. The clinical impact of transient leukoreduction.1,2 Its impact is significant, however, because autoantibodies appears to be minimal. Lastly, because ABO refractoriness is independently associated with bleeding determinants are present on the platelet surface, patients complications and decreased patient survival.3 with very high isoantibody titers can destroy up to 40 percent of major-mismatched platelets (e.g., group A platelets Causes of Platelet Transfusion Refractoriness in group O recipients) and thus benefit from ABO-identical Causes of poor posttransfusion platelet count incre- transfusions.11 Taking this into account, the classic defini- ments are classified according to the mechanism of platelet tion of platelet transfusion refractoriness assumes that loss, either from activation or by immune destruction. Al- ABO-identical platelets have already been tried without most 90 percent of platelet transfusion refractoriness oc- success in improving count increments. curs in the setting of shortened survival from accelerated platelet activation, which results in transfusion increments Matched Platelet Strategies less than half of those expected 24 hours after transfusion.4 Cellular product leukoreduction has been shown to re- These conditions include splenomegaly, infection, ampho- duce the incidence of HLA alloimmunization by more than tericin B therapy, disseminated intravascular coagulation, 50 percent (from 45% to 17 to 18% in the Trial to Reduce bleeding, stem cell transplantation, and posttransplant he- Alloimmunization to Platelets [TRAP] study).9 Despite patic venoocclusive disease or graft-versus-host disease.5,6 the mitigating effect of near-universal leukoreduction on Immune destruction occurs in approximately 25 percent of anti-HLA-mediated platelet transfusion refractoriness, unsuccessful transfusions, most frequently coexisting with approximately 5 percent of patients undergoing high-dose conditions that shorten platelet circulatory survival.4 chemotherapy or stem cell transplantation still require A 1-hour posttransfusion platelet count should be ob- matched platelets to achieve acceptable posttransfusion tained to determine whether immune platelet destruction platelet count increments.2 Several methods have been

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 119 R.R. Vassallo developed to provide HLA-matched platelets. These include count increments.14 If several weeks have passed since the (1) transfusing platelets from donors with the same HLA-A last HLA antibody identification test, it should be repeated. and -B antigens as the recipient (or least-incompatible mis- When HLA-identical or correctly matched antigen-negative matches inferred from known antigenic cross-reactivity or platelet transfusions have failed, in the absence of dramatic lists of immunogenic epitopes present on donor and recipi- worsening of the patient’s clinical status (e.g., new-onset ent HLA antigens), (2) providing antigen-negative prod- septic shock), testing for HPA antibodies is necessary. ucts that honor patients’ HLA antibody specificities, or (3) crossmatching recipient serum with donor platelets. Each of Human Platelet Antigens these approaches has strengths and weaknesses, described Human platelet antigen typing and antibody determi- in detail elsewhere.1,6 nation are described in detail in the accompanying review in Most matching algorithms for platelet transfusion- this issue of Immunohematology.15 All but one of the char- refractory patients, in recognition of the preeminent role acterized HPAs represent single nucleotide polymorphisms of HLA antibodies, begin with the documentation of HLA resulting in a lone amino acid change in one of four platelet antibodies with a screening assay. Anti-HLA is detected ap- surface glycoproteins (GPs). These include the platelet inte- proximately 15 to 40 percent of the time (dependent on the grins GP IIb-IIIa and GP Ia-IIa, the platelet von Willebrand vigor with which referring physicians have documented im- receptor (GP Ib-IX-V), and CD109 (a GPI-linked cell-cell mune causes of transfusion refractoriness with 1-hour post- interaction mediator). Genetic susceptibility is known to transfusion counts and the sensitivity of the HLA antibody play a role in HPA alloimmunization during pregnancy or detection assay).12 In the setting of poor 1-hour posttrans- after transfusion, so only a portion of antigen-negative infusion platelet increments, the likelihood of missed HLA dividuals are capable of forming antibodies on alloexposure antibodies (from nascent IgM antibodies in an IgG-only to common antigens.16,17 Antibodies have been reported to detection system or noncytotoxic antibodies in antiquated 11 antigens residing on these GPs with phenotypic frequen- assays) or the unusual presence of lone HPA antibodies cies in Caucasian blood donors between 13 percent and rises. Additional testing to search for missed HLA (i.e., 100 percent (Table 1).18 These have the potential to result more-sensitive testing) or as-yet-unsought HPA antibodies in poor responses in a similar percentage of platelet trans- is usually recommended. fusions given the right combination of transfused platelet The other circumstance necessitating further testing is antigen density, recipient antibody titer to avidity, and re- the failure of HLA-matched platelet support. Regardless of cipient reticuloendothelial system appetite for antibody- how HLA-matched platelets are chosen, 25 to 50 percent coated platelets. Another 11 formally recognized HPAs are of these platelet units do not result in acceptable incre- low-frequency antigens (<1%) against which antibodies are ments even 1 hour after the transfusion.13 Even a perfectly problematic only in pregnancy when the fetus carries the matched platelet unit is not guaranteed to maintain the father’s rare immunizing antigen.18 In the setting of platelet platelet count in an acceptable range for 24 hours as these transfusion, low-frequency antigens would hardly ever be patients often have comorbidities that shorten platelet cir- encountered and are thus highly unlikely to result in platelet culatory survival. Teasing out whether poor responses to transfusion refractoriness. matched platelets are caused by suboptimal platelet matching, worsening patient illness, or the development of new Table 1. Human platelet antigens (HPA) with Caucasian phenotypic antibody specificities has important consequences for on- prevalence >1% to which antibodies have been reported going platelet transfusion support. Subjecting the patient Platelet Caucasian to 1-hour posttransfusion platelet counts after each transfu- HPA Alternative membrane phenotypic nomenclature designation glycoprotein frequency* sion is mandatory in this setting to determine whether immune destruction is accounting for disappointing 24-hour HPA-1a PlA1 GP IIIa 98% posttransfusion results. A reexamination of the quality of HPA-1b PlA2 GP IIIa 29% HLA-matching is required. Approximately 60 to 80 per- HPA-2a Kob GP Ibα >99% cent of matched platelets supplied using only the recipi- HPA-2b Koa GP Ibα 13% ent’s HLA type (i.e., the HLA antibody screen is reactive, HPA-3a Baka GP IIb 81% but antibody specificities have not been determined) will be HPA-3b Bakb GP IIb 70% best-guess mismatches.1 The determination of what HLA antibody specificities are actually present can explain the HPA-4a Pena/Yukb GP IIIa >99% failure of many of these units and often results in dramati- HPA-5a Brb GP Ia 99% cally better results when antigen-antibody mismatches are HPA-5b Bra GP Ia 20% avoided. Even when the antigen-negative approach is used, HPA-15a Govb CD109 74% it is recognized that approximately 10 to 15 percent of pa- HPA-15b Gova CD109 81% tients broaden their anti-HLA repertoire with time, adding *From Norton et al.18 new specificities to those likely to produce poor platelet

120 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 HPAs in platelet transfusion refractoriness

Approximately 0.5 percent of Asian Americans and Af- destroy antibody-coated platelets, and the definition of re- rican Americans lack CD36 (a widely distributed scavenger fractoriness considers immediate destruction of less than receptor known on platelets as GP IV) on cells throughout half of transfused platelets as clinically insignificant. It is the body.19 They are at risk for isoimmunization to the so- not surprising that HPA antibodies would react similarly. called Naka antigen on alloexposure by pregnancy or trans- Second, the breadth of alloimmunization affects the likeli- fusion. Another 4.5 percent of individuals from these same hood that clinical transfusion refractoriness will result. For ethnic groups lack CD36 on their platelets (but express it our matching service, the median fraction of units predicted elsewhere) and unlike Caucasians, who almost uniformly to carry antigens incompatible with patients’ HLA antibod- express CD36 on platelets, can serve as donors for im- ies is approximately 65 percent. As will be demonstrated, munized patients. the corresponding value for our patients’ HPA antibodies is Table 2 shows the incidence of HPA antibodies in pa- less than half that. Thus, whereas HPA antibodies are prob- tients tested for the development of HLA and HPA antibod- ably less likely to result in clinical transfusion refractoriness ies after intensive episodes of transfusion. Several points than HLA antibodies, their clinical impact is not negligible deserve emphasis. The literature is divided regarding the and they are not particularly rare. appearance of HPA antibodies after transfusion. In studies testing only once or twice, the prevalence of HPA antibod- Table 2. Prevalence of HPA antibodies* ies is relatively low (4.5%; range, 0–13.6%), whereas the A. Single determination per patient, only identifiable HPA specificities prevalence reported with weekly testing is higher (13.8%; counted 20–28 range, 3.9–47.5%). In the study with the highest re- HPA only HPA+HLA HLA only Patients Reference ported prevalence, 70 percent of incident antibodies appeared during infection and lasted no more than 300 days, 1 1 21 117 Taaning et with a mean persistence of less than a month for partici- al.20 pants monitored for up to 70 days.27 Half of these antibod- 5 15 93 252 Kiefel et al.21 ies were reactive with autologous platelets, and the authors 0 6 58 293 Kickler et were unable to discern any effect of transient anti-platelet al.22 antibodies on platelet increments. The assay technology 0 3 12 145 Legler et used in this study also may not have clearly distinguished al.23 HPA from HLA antibodies, and there was only rudimentary 7 4 32 81 Kurz et al.24 estimation of the breadth of alloimmunization (using only 0 0 11 50 Godeau et four cells). Removing this study’s unusually high number of al.25 HPA antibodies, overall estimates for HPA antibody forma- 13 (1.4%) 29 (3.1%) 227 (24.2%) 938 tion become more uniform, averaging just under 5 percent, with approximately 16 percent of alloimmunized patients B. Multiple determinations per patient, identifiable and nonidentifiable manifesting HPA antibodies. This 5 percent figure derived specificities counted from the study of almost 1200 patients with various diag- HPA HLA noses is somewhat lower than the TRAP trial’s report of an only HPA+HLA only Patients Reference 8 percent incidence of HPA antibodies among its 530 pa- 9 tients with acute leukemia. 8 2 22 104 Uhrynowska et al.26 Although older literature suggests that HPA antibodies only rarely lead to significant refractoriness and that 18 10 9 59 McGrath et al.27 most instances of platelet transfusion refractoriness are limited to case reports, just over 3 percent of the patients 2 4 55 154 Murphy et al.28 our matching service supports annually have clinically significant HPA antibodies.27,29 This is lower than the expected overall anti-HPA detection rate of 5 percent. Our lower 28 16 (5.0%) 86 317 (8.8%) (27.1%) value of clinically significant HPA antibodies is not entirely unexpected for two reasons. First, it is well known that less *All studies identified HPA antibodies using monoclonal antibody im- than half of HLA alloantibodies cause strictly defined plate- mobilization of platelet antigens (MAIPA) tests or a GP-specific RIA let transfusion refractoriness (e.g., in the TRAP study, 17– except the studies of McGrath et al.27 and Murphy et al.,28 which used less-specific testing (comparing platelet immunofluorescence [PIF] 18% of patients receiving leukoreduced products exhibited before and after HLA antigen removal with chloroquine or comparing HLA antibodies, but only 3–4% had alloimmune transfu- PIF with HLA-dependent lymphocytotoxicity panels). Nonidentifiable sion refractoriness within the overall rate of refractoriness specificity represents GP reactivity in the MAIPA with platelets bear- 6,9 ing different HPA polymorphisms on GP IIb-IIIa, GP Ib/IX/V, or GP of 7–8%). Some antibodies may not be potent enough or Ia-IIa. find enough targets on the platelet surface to bind in significant numbers, some recipients may not be able to rapidly

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 121 R.R. Vassallo

In the cross-sectional prevalence studies from Table Providing HPA-Compatible Platelets 2A, more than two thirds of HPA antibodies appear in the When patients appear to have immune-mediated plate- presence of HLA antibodies. It has been proposed that HLA let transfusion refractoriness and either have no identifi- alloimmunization may identify individuals more suscep- able HLA antibodies or have had poor platelet increments tible to the formation of HPA antibodies.30,31 The additional with presumably properly HLA-matched platelets, testing restrictions that HPA antibodies place on HLA-selected for HPA antibodies is indicated. Detection and identifica- platelets help explain some of the failures of HLA-identical tion of HPA antibody specificities can be carried out with and HLA antigen-negative platelet units. These data also relatively rapid commercially available ELISAs. The older support the HPA antibody screening of individuals with- literature describes ELISAs as less sensitive, recommend- out apparent HLA antibodies when immune platelet trans- ing more time-consuming assays such as the gold-standard fusion refractoriness is highly likely in the setting of poor monoclonal antibody immobilization of platelet antigens 1-hour posttransfusion platelet count increments. It should (MAIPA) test (and its variants) or platelet immunofluores- be recognized, though, that the fraction of transfusion- cence (PIF) assays properly controlled for concurrent HLA refractory individuals who harbor HPA antibodies in isola- antibodies.39 More recent international workshops have tion is probably less than 2 percent. Case reports of isolated demonstrated relatively equal performance for ELISA and HPA antibodies exist and our matching service sees at least the MAIPA, attributed to variability in technique in labo- one such patient annually, usually with an HPA-1a (PlA1) or ratories using the technically complex MAIPA.40 MAIPA or CD36 (Naka) antibody.32,33 PIF testing is currently the only option to detect HPA-15 an- Table 3 shows the specificities of HPA antibodies iden- tibodies. Detection of CD36 isoantibodies is possible using tified in multiply transfused patients. In older studies with all three of the most commonly performed assays (ELISA, limited specificity testing, the majority of antibodies are MAIPA, and PIF). Because HPA antibody specificity is de- directed against HPAs with phenotypic frequencies less rived from the pattern of reactivity with cells or wells with than 30 percent (i.e., three quarters are directed at HPA-1b, known HPA antigens, it is helpful to confirm antibody iden- HPA-2b, and HPA-5b). All of these studies were conducted tity by demonstrating that the recipient’s genotype predicts using technologies unable to identify antibodies to CD109 that he or she does not express that particular antigen. or CD36. More recent studies capable of detecting CD109 Most blood suppliers do not have extensive panels of antibodies estimate that approximately 30 percent of HPA HPA-typed apheresis platelet donors. Because HPA-1a is antibodies target CD109, whereas almost 60 percent are the antibody most frequently responsible for neonatal al- still directed against antigens with phenotypic frequencies loimmune thrombocytopenia and posttransfusion purpura, less than 30 percent.35,36 Fortunately, for the majority of pa- larger blood centers often screen their apheresis donors to tients with HPA antibodies, less than 30 percent of platelet identify the approximately 2 percent who are HPA-1a (PlA1)- transfusions are at risk for inappropriately low increments, negative. Because HPA-1a is not commonly responsible for as noted above. platelet transfusion refractoriness, these donor panels are The TRAP study demonstrated that leukoreduction had usually not helpful. If a more broadly typed apheresis do- no effect on the incidence of HPA alloimmunization.9 This nor panel is available, choosing one whose HPA (and HLA) is because HLA presentation to recipients’ immune systems types avoid the recipient’s antibody specificities is prefer- occurs through transfused donor antigen-presenting cells able. On-demand donor typing is practical for recipients (APCs) and is thus ameliorable through leukoreduction, with antibodies to lower-frequency antigens because fewer whereas HPA presentation occurs by recipient APCs, which donors require screening. Genotyping is preferred, despite would be unaffected by leukoreduction of transfused blood rare incorrect phenotypic predictions, because it requires components.37,38 Accordingly, although HLA alloimmuniza- a smaller sample volume obtainable from the infectious tion is decreasing with the widespread use of leukoreduced disease testing tubes collected routinely at apheresis do- blood products, HPA alloimmunization will be unaffected nations, which can be scheduled as frequently as every 3 and could represent an increasing proportion of detected days. (The limited availability of antisera for phenotyping alloantibodies as HLA antibodies decline in prevalence. also makes genotyping more practical.) When patients have

Table 3. Identified HPA specificities 1a 1b 2a 2b 3a 3b 5a 5b 15a 15b 4 (7.1%)* 18 (32.1%) 1 (1.8%) 6 (10.7%) 4 (7.1%) 3 (5.4%) 2 (3.6%) 18 (32.1%) 5 (6.4%)† 15 (19.2%) 8 (10.3%) 1 (1.3%) 4 (5.1%) 22 (28.2%) 7 (9.0%) 16 (20.5%) *Composite of references 20–23, 26, 33, 31, and 34 (these studies only assayed antibodies to HPAs 1, 2, 3, and 5). †Composite of references 35 and 36 (assayed antibodies to HPAs 1, 2, 3, 4, 5, and 15).

122 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 HPAs in platelet transfusion refractoriness high-frequency antibodies, it is often helpful to begin HPA 2. Seftel MD, Growe GH, Petraszko T, et al. Univer- (geno)typing with family members. sal prestorage leukoreduction in Canada decreases Despite a lack of literature, in the absence of known platelet alloimmunization and refractoriness. Blood type-compatible donors, platelet crossmatching is often 2004;103:333–9. used to predict platelet compatibility. Crossmatching is 3. Kerkhoffs JL, Eikenboom JC, van de Watering LM, et al. most frequently performed using a commercially avail- The clinical impact of platelet refractoriness: correlation able solid-phase RBC adherence (SPRCA) assay. If HLA with bleeding and survival. Transfusion 2008;48:1959– antibodies are concurrently present, only HLA-compatible 65. units should be crossmatched. We have observed that re- 4. Doughty HA, Murphy MF, Metcalfe P, et al. Relative im- sults are more reliable for antibodies directed against HPAs portance of immune and non-immune causes of platelet with higher platelet surface densities, as greater amounts of refractoriness. Vox Sang 1994;66:200–5. bound antibody are more likely to identify incompatibility 5. Slichter SJ, Davis K, Enright H, et al. Factors affecting in the assay. Table 4 predicts that antibodies to HPAs ex- posttranfusion platelet increments, platelet refractori- pressed on GP IIb-IIIa or Ib-IX-V are more likely to yield ness, and platelet transfusion intervals in thrombocy- reliable SPRCA results. All too often, however, units ap- topenic patients. Blood 2005;105:4106–14. pearing compatible in vitro fail in vivo, and only after the 6. Hod E, Schwartz J. Platelet transfusion refractoriness. donor’s genotype is obtained days later is the platelet in- Br J Haematol 2008;142:348–60. crement failure explained. Donors whose transfused plate- 7. Daly PA, Schiffer CA, Aisner J, Wiernik PH. Platelet lets succeed in producing good count increments should transfusion therapy. One-hour posttransfusion incre- be called back to donate additional products. With high- ments are valuable in predicting the need for HLA- frequency recipient antibodies in the absence of compatible matched preparations. JAMA 1980; 243:435–8. donors, crossmatched units with the weakest incompatibil- 8. Yankee RA, Grumet FC, Rogentine GN. Platelet transfu- ity are preferred by some. There is, however, no evidence sion: the selection of compatible platelet donors for re- that decreased in vitro reactivity in any way predicts lim- fractory patients by lymphocyte HL-A typing. N Engl J ited in vivo platelet destruction. Unfortunately, more ro- Med 1969;281:1208–12. bust crossmatch methods (PIF, MAIPA) are generally too 9. The Trial to Reduce Alloimmunization to Platelets onerous to implement routinely. Ultimately, as economical, Study Group. Leukocyte reduction and ultraviolet B ir- high-throughput HPA genotyping methods come into wide- radiation of platelets to prevent alloimmunization and spread use, panels of donors pretyped for HPAs 1 through refractoriness to platelet transfusions. N Engl J Med 5 and 15 will assist in the choice of HPA antigen-negative 1997;337:1861–9. units and obviate the need for crossmatching. 10. AuBuchon JP, Leach MF. Investigating the possibility of We have observed impressive successes in broadly drug-dependent platelet antibodies. Immunohematol- transfusion-refractory patients when donors are carefully ogy 2009;25:136–140. chosen to avoid HLA and HPA antigens to which recipients 11. Cooling L. ABO and platelet transfusion therapy. Immu- are alloimmunized. Close cooperation is required among nohematology 2007;23:20–33. clinicians, the hospital transfusion service, and the blood 12. [No authors listed.] Detection of platelet-reactive anti- supplier to choose the right products and assure their avail- bodies in patients who are refractory to platelet transfu- ability throughout periods of intensive platelet support. sions, and the selection of compatible donors. Vox Sang 2003;84:73–88. Table 4. Platelet glycoprotein surface density 13. Heal JM, Blumberg N, Masel D. An evaluation of cross- Glycoprotein Platelet surface copy number* matching, HLA and ABO matching for platelet transfu- GP IIb-IIIa 50,000–80,000 sions to refractory patients. Blood 1987;60:23–30. 14. Petz LD. The HLA system and platelet transfusion. In: GP Ib-IX-V 25,000 GP Ib-IX; 12,000 GP V Murphy S, ed. The HLA system: basic biology and GP Ia-IIa 800–2,800 clinical applications. Bethesda, MD: American Associa- CD109 2,000 tion of Blood Banks, 1999. CD36 2,000 15. Curtis BR, McFarland JG. Detection and identification *From references 35, 41, and 42. of platelet antibodies and antigens in the clinical laboratory. Immunohematology 2009;25:125–135. References 16. Semana G, Zazoun T, Alizahed M, et al. Genetic suscep- 1. Vassallo RR Jr. New paradigms in the management of tibility and anti-human platelet antigen 5b alloimmu- alloimmune refractoriness to platelet transfusions. Curr nization: role of class II and TAP genes. Hum Immunol Opin Hematol 2007;14:655–63. 1996;46:114–19.

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 123 R.R. Vassallo

17. Blanchette VS, Chen L, de Friedberg ZS, et al. Alloimmu- 32. Pappalardo PA, Secord AR, Quitevis P, et al. Platelet nization to the PlA1 platelet antigen: results of a prospec- transfusion refractoriness associated with HPA-1a (PlA1) tive study. Br J Haematol 1990;74:209–15. alloantibody without coexistent HLA antibodies success- 18. Norton A, Allen DL, Murphy MF. Review: platelet al- fully treated with antigen-negative platelet transfusions. loantigens and antibodies and their clinical significance. Transfusion 2001;41:984–7. Immunohematology 2004;20:89–102. 33. Fujino H, Ohta K, Taniue A, et al. Primary refractoriness 19. Curtis BR, Ali S, Glazier AM, et al. Isoimmunization to platelet transfusion caused by Naka antibody alone. against CD36 (glycoprotein IV): description of four cases Vox Sang 2001;81:42–4. of neonatal alloimmune thrombocytopenia and brief re- 34. Lo SC, Chang JS, Lin SW, Lin DT. Platelet alloimmu- view of the literature. Transfusion 2002;42:1173–9. nization after long-term red cell transfusion in trans- 20. Taaning E, Sinonsen AC, Hjelms E, et al. Platelet alloim- fusion-dependent thalassemia patients. Transfusion munization after transfusion. Vox Sang 1997;72:238– 2005;45:761–5. 41. 35. Ertel K, Al-Tawil M, Santoso S, Kroll H. Relevance 21. Kiefel V, Konig C, Kroll H, Santoso S. Platelet alloanti- of the HPA-15 (Gov) polymorphism on CD109 in al- bodies in transfused patients. Transfusion 2001;41:766– loimmune thrombocytopenic syndromes. Transfusion 70. 2005;45:366–73. 22. Kickler T, Kennedy SD, Braine HG. Alloimmunization to 36. Berry JE, Murphy CM, Smith GA, et al. Detection of Gov system platelet-specific antigens on glycoproteins IIb-IIIa and antibodies by MAIPA reveals an immunogenicity similar to Ib/IX in multiply transfused thrombocytopenic patients. the HPA-5 alloantigens. Br J Haematol 2000;110:735– Transfusion 1990;30:622–5. 42. 23. Legler TJ, Fischer I, Dittmann J, et al. Frequency and 37. Semple JW, Speck ER, Milev YP, et al. Indirect allorecognition causes of refractoriness in multiply transfused patients. of platelet by T helper cells during platelet transfusions Ann Hematol 199;74:185–9. correlates with anti-major histocompatibility complex 24. Kurz M, Greinix H, Hocker P, et al. Specificities of antibody and cytotoxic T lymphocyte formation. Blood anti-platelet antibodies in multitransfused patients 1995; 86:805–12. with haemato-oncological disorders. Br J Haematol 38. Sayeh E, Aslam R, Speck ER, et al. Immune responsiveness 1996;95:564–9. against allogeneic platelet transfusions is determined by 25. Godeau B, Fromont P, Seror T, et al. Platelet alloimmu- the recipient’s major histocompatibility complex class II nization after multiple transfusions: a prospective study phenotype. Transfusion 2004;44:1572–8. of 50 patients. Br J Haematol 1992;81:395–400. 39. Lucas GF, Rogers SE. Evaluation of an enzyme-linked 26. Uhrynowska M, Zupanska B. Platelet-specific antibodies immunosorbent assay kit (GTi PakPlus) for the detection in transfused patients. Eur J Haematol 1996;56:248– of antibodies against human platelet antigens. Transfus 51. Med 1999;9:63–7. 27. McGrath K, Wolf M, Bishop J, et al. Transient platelet 40. Allen D, Ouwehand WH, de Haas M, et al. Interlabo- and HLA antibody formation in multitransfused patients ratory variation in the detection of HPA-specific al- with malignancy. Br J Haematol 1988;68:345–50. loantibodies and in molecular HPA typing. Vox Sang 28. Murphy MF, Metcalfe P, Ord J, et al. Disappearance of 2007;93:316–24. HLA and platelet-specific antibodies in acute leukaemia 41. Lucas GF, Metcalfe P. Platelet and granulocyte glycopro- patients alloimmunized by multiple transfusions. Br J tein polymorphisms. Transfus Med 2000;10:157–74. Haematol 1987;67:255–60. 42. Cserti-Gazdewich CM, Dzik WH, Dorn ME, et al. Quan- 29. McFarland JG. Alloimmunization and platelet transfu- titation of CD36 (platelet glycoprotein IV) expression on sion. Semin Hematol 1996;33:315–28. platelets and monocytes by flow cytometry: application 30. Schnaidt M, Northoff H, Wernet D. Frequency and to the study of Plasmodium falciparum malaria. Cytom- specificity of platelet-specific alloantibodies in HLA- etry B Clin Cytom 2009;76:127–34. immunized haematologic-oncologic patients. Transfus Med 1996;6:111–14. Ralph R. Vassallo, MD, FACP, American Red Cross Blood 31. Sanz C, Freire C, Alcorta I, et al. Platelet-specific anti- Services—Penn-Jersey Region, 700 Spring Garden Street, bodies in HLA-immunized patients receiving chronic Philadelphia, PA 19123. platelet support. Transfusion 2001;41:762–5.

Important Notice About Manuscripts Immunohematology is on the Web! for Immunohematology www.redcross.org/immunohematology

Please e-mail all manuscripts for consideration to For more information, send an e-mail message [email protected] to [email protected]

124 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Review Detection and identification of platelet antibodies and antigens in the clinical laboratory

B.R. Curtis and J.G. McFarland

As a result of the unique functional properties of platelets, more- linked to individual polystyrene beads have been developed robust methods were required for detection of antibodies raised that greatly increase assay throughput. These same anti- against them. Immunofluorescence detection by flow cytometry, body detection assays were also used to phenotype platelets solid-phase red cell adherence, and antigen capture ELISAs are for the human platelet antigens (HPAs) expressed. How- some of the current tests that have been developed to meet the ever, serologic typing required the availability of relatively challenges of platelet antibody detection and identification and antigen phenotyping. Recently developed protein liquid bead arrays large quantities of highly specific typing sera free of HLA are becoming the next-generation platelet antibody tests. Fueled class I antibodies. With the cloning and identification of the by development of PCR and determination of the molecular basis HPA-1a (PlA1) gene,19 serologic typing was soon replaced by of the PlA1 human platelet antigen (HPA), serologic platelet typing molecular methods for genotyping DNA.19,20 has now been replaced by genotyping of DNA. Allele-specific PCR, melting curve analysis, and 5′-nuclease assays are now evolving Platelet Glycoproteins and Antigens into more high-throughput molecular tests. Laboratory testing for Like RBCs, platelets have their own unique surface the diagnosis of immune platelet disorders has advanced consid- antigens. To date, 28 different platelet-specific antigens erably from its humble beginnings. carried on five different GPs have been completely char- Immunohematology 2009;25:125–135. acterized. The Platelet Antigen Nomenclature Working Party of the International Society of Blood Transfusion Historic Background and International Society of Thrombosis and Hemostasis21 evelopment of suitable tests for detection and iden- has assigned 23 HPA-designated systems (Table 1). With tification of platelet antibodies and antigens in the the exception of one (HPA-14bw), all consist of two alleles clinical laboratory diagnosis of immune platelet dis- D with the higher frequency allele given the “a” designation orders proved difficult compared with the relative ease with and the lower frequency allele the “b” designation. A “w” which simple tests were designed for detection of RBC an- (workshop) designation has been assigned to antibodies for tibodies and antigens. Homotypic platelet–platelet interac- which the antithetical allelic antigen has not yet been found. tions via cell surface receptors and their ligands caused non- Platelets also express other “non-HPA” antigens to which specific platelet agglutination after centrifugation, thereby antibodies are produced that have been implicated in im- preventing the application of standard RBC antibody tube mune platelet disorders including the blood group ABO(H) agglutination tests in the detection of platelet antibodies. In antigens, Nak(a) antigen on GPIV, and HLA class I (Table addition, isolation and preparation of platelets for testing 1). A, B, and H antigens are expressed on virtually all major resulted in elevated cell surface immunoglobulin levels that platelet membrane GPs, but the majority are expressed on significantly reduced assay sensitivity and specificity.1 GPIIb and platelet endothelial cell adhesion molecule (PE- Early assays for detection of platelet antibodies used CAM or CD31).22 HLA class I A, B, and C antigens are all the functional properties of platelets as assay end points expressed on platelets, which are the major source of HLA such as protein receptor–induced activation, alpha and class I antigen in the blood.23 dense granule release, and platelet aggregation and agglutination. These tests are often referred to as phase I or first- Immune Platelet Disorders Evaluated by Platelet generation assays and will not be discussed further. Later, Immunology Laboratories more sophisticated tests (phase II or second-generation as- Platelet antibody and antigen tests are primarily used says) included direct or indirect measurements of platelet in the laboratory investigation of six immune platelet dis- surface-bound immunoglobulins using secondary antibody orders: (1) autoimmune thrombocytopenia, often referred reagents in immunoblots,2 RIA,2–4 ELISA,5–7 immunofluo- to as ITP (idiopathic thrombocytopenic purpura), in which rescence,8–10 latex agglutination,11,12 and solid-phase RBC patients produce antibodies against their own platelets; (2) adherence.13,14 With the production of platelet glycoprotein neonatal alloimmune thrombocytopenia (NATP), in which (GP)-specific monoclonal antibodies throughout the early maternal IgG antibodies produced against a paternal plate- 1980s, phase III or third-generation assays were developed let antigen inherited by the child cause destruction of fetal that allowed for detection of antibodies bound to platelet and newborn platelets; (3) posttransfusion purpura (PTP), membrane GPs captured in the wells of microtiter plates.15–18 an acute, severe thrombocytopenia that develops 5 to 10 More recently, assays using purified platelet GPs covalently

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 125 B.R. Curtis and J.G. McFarland

Table 1. Platelet Glycoproteins and Antigens

Phenotypic Glycoprotein Location/ days after a blood transfu- Antigens Other Names Frequency* Amino Acid Change Nucleotide Substitution HPA-1a (PlA1) PlA, Zw 72% a/a GPIIIa / L33P T>C 196 sion in a patient previously HPA-1b (PlA2) 26% a/b sensitized to a platelet al- 2% b/b loantigen during pregnancy or prior transfusion HPA-2a (Kob) Ko, Sib 85% a/a GPIb / T145M C>T 524 α that leads to autoantibody HPA-2b (Koa) 14% a/b 1% b/b formation; (4) multiplate- let transfusion refractori- HPA-3a (Baka) Bak, Lek 37% a/a GPIIb / I843S T>G 2622 ness (MPTR), which oc- HPA-3b (Bakb) 48% a/b curs most frequently in 15% b/b oncology patients receiving HPA-4a (Pena) Pen, Yuk >99.9% a/a GPIIIa / R143Q G>A 526 platelet transfusion sup- HPA-4b (Penb) <0.1% a/b port after bone marrow <0.1% b/b transplant or chemotherapy who become highly al- b HPA-5a (Br ) Br, Hc, Zav 80% a/a GPIa / E505K G>A 1648 loimmunized, primarily to HPA-5b (Bra) 19% a/b 1% b/b HLA class I antigens; (5) drug-induced thrombocy- HPA-6bw Caa, Tu <1% b/b GPIIIa / R489Q A>G 1564 topenia (DITP), in which patients develop acute, b HPA-7bw Mo <1% b/b GPIIIa / P407A G>C 1317 severe thrombocytopenia HPA-8bw Sra <0.1% b/b GPIIIa / R636C T>C 2004 caused by drug-dependent antibodies (DDAbs) that HPA-9bw Maxa <1% b/b GPIIb / V837M A>G 2603 develop after exposure to various drugs—some of the HPA-10bw Laa 1% b/b GPIIIa / R62Q A>G 281 drugs that most commonly HPA-11bw Groa <0.5% b/b GPIIIa / R633H A>G 1996 cause DDAbs are quinine, vancomycin, and GPIIb/ HPA-12bw Iya 1% b/b GPIb / G15E A>G 141 β IIIa inhibitors; and (6) heparin-induced throm- HPA13bw Sita <1% b/b GPIa / M799T T>C 2531 bocytopenia with throm- HPA-14bw Oea 1% GPIIIa / De167K AAG 1929-31 bosis (HITT) or without it (HIT)—thrombocytopenia HPA-15a (Govb) Gov 35% a/a CD109 / Y703S A>C 2108 develops most often in pa- a HPA-15b (Gov ) 42% a/b tients exposed to unfrac- 23% b/b tionated heparin, resulting HPA-16bw Duva <1% GPIIIa / T140I C>T 517 in the production of antibodies to platelet-factor 4 HPA-17bw Vaa <1% GPIIIa / T195M C>T 622 (PF4)—heparin complexes HPA- ? Caba <1% GPIa / Q716H G>T 2235 that bind to platelet Fc receptors, leading to platelet HPA- ? Sta <1% GPIIIa / K137Q A>C 487 activation and clearance. Detailed descriptions of HPA- ? Kno <1% GPIIb / T619M C>T 1949 various methods for detec- HPA- ? Nos <1% GPIIIa / E628K G>A 1960 tion of platelet antibodies and allelic gene polymor- NA Naka 99.8% (Caucasian) CD36 (GPIV) phisms that encode plate- 97% (African) T>G 1264 let antigens are described 96% (Asian) C>T 478 in this article. The role of NA A, B, H 42% A GPIIb, PECAM, GPIIIa NA platelet antibody testing in 10% B GPIa, GPIIa, GPIV the diagnosis of the various 44% O (H) immune platelet disorders 4% AB is explored in other articles NA A, B, C NA HLA class I NA in this issue. *Phenotypic frequencies for the antigens shown are for the Caucasian population. Significant differences in gene frequencies may be found in African and Asian populations. NA = not applicable; ? = HPA number not yet assigned

126 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Platelet antibodies and antigens in the clinical laboratory

Platelet Antibody Tests Currently Used for the Laboratory Investigation of Immune Platelet A1 Disorders Assays using intact platelets Despite a good selection of platelet antibody detection methods available today, no single method is sufficient.24 Each method has limitations that make it necessary to

perform a combination of several methods to ensure a PF-CD41a+ Events thorough antibody workup. Specialized platelet antibody reference laboratories typically perform an assay that uses intact platelets to screen serum or plasma for antibodies, a GP antigen capture assay for identification of the antigen that antibodies target, and genotyping of patient DNA to A2 help confirm the HPA specificity of platelet alloantibodies identified.

Flow cytometry

The platelet suspension immunofluorescence test Events (PSIFT) using a fluorescence microscope developed by von dem Borne et al. in 19788 was the first widely adopted method using intact platelets for the detection of platelet-reactive antibodies. A flow cytometric version of the PSIFT was also described in von dem Borne’s original report, and as the instruments became more available, immunofluorescence by flow cytometry has now become one of the most popular B and sensitive methods for detection of platelet antibodies using intact platelets.9,10,25 A typical assay includes incubation of patient’s serum or plasma with platelets for 30 to 60 minutes, usually at room temperature. The platelets are then washed to remove unbound and some nonspecifically bound immunoglobulins, followed by incubation with fluo- Number of Events rescent labeled F(ab′)2 fragment of polyclonal anti-human IgG for about 20 minutes in the dark. After one wash, the platelets are suspended in buffer, and platelet-bound fluo- Fluorescence Intensity rescence is detected by a flow cytometer. Mean or median fluorescence intensity (MFI) values and fluorescence histograms are generated by the flow cytometer software for Fig. 1 Flow cytometry fluorescence dot plot A1( ) and histograms use in determining results (Fig. 1). Results can be expressed of platelet antibody test results (A2, and B). (A) Whole blood was as a fluorescence ratio (FLR) of MFI obtained with the incubated with a phycoerythrin (PE)-labeled monoclonal antibody patient’s serum divided by MFI obtained with a negative specific for platelet GPIIb (PE-CD41a), and normal or patient control serum from a normal healthy individual incubated serum with HPA-1a antibodies. Platelets were then washed, and with the same platelet target. An FLR cutoff for a positive bound human IgG platelet antibodies were detected by incubation with FITC-(Fab′) fragment of goat anti-human IgG and detection interpretation can be established by statistical analysis of 2 of platelet-bound fluorescence by flow cytometry. Platelets were FLR values obtained from testing 20 to 30 normal sera. Se- identified by gating on PE-CD41a+ fluorescent cells in dot plot rum or plasma positive control samples should include IgG (A1), and level of FITC-anti-human IgG fluorescence determined and IgM platelet-reactive antibodies detected in previously in histogram (A2) displayed for both normal serum (white) and tested patient samples. anti-HPPA-1a serum (black). MDFI = median fluorescence in- FITC is the most commonly used fluorescent label for tensity. (B) Histograms of drug-dependent platelet antibody test. the anti-IgG probe. The fluorochrome phycoerythrin (PE) Platelets were incubated with patient serum in the presence (white attached to an anti-human IgM probe is often used for si- histogram) and absence (gray histogram) of drug, and antibody detected with FITC-(Fab′) fragment of goat anti-human IgG. Fluo- multaneous detection of IgM platelet antibodies. The two 2 rescence is significantly stronger in the presence of drug, indicat- labeled probes can be added together during the final incu- ing the patient has drug-dependent platelet antibodies. bation, and both activated to fluoresce by the 480-nm laser. The fluorescent signals are detected separately

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 127 B.R. Curtis and J.G. McFarland

by visual examination of the wells. A positive test occurs because their emission wavelengths differ. F(ab′)2 fragments (antibody with Fc end removed) of the anti-immuno- when indicator RBCs bind to antibodies attached to the im- globulin probes are used to prevent nonspecific binding via mobilized platelets, forming a “carpet” of RBCs spread over Fc receptors on the target platelets. the bottom of the well. In the absence of platelet antibodies, Flow cytometry has proved to be a very sensitive meth- centrifugation forces indicator RBCs into a tight button at od for detection of platelet antibodies and is considered the the well bottom. best method for detection of drug-dependent platelet anti- SPRCA has been successfully used as a crossmatching bodies induced by drugs like quinine,26 sulfamethoxazole,27 test, because of its speed and ease of use.14,37 Shortcomings and vancomycin.28 For detection of drug-dependent anti- include low specificity, subjective end point, short expira- bodies the previously described procedure can be used with tion of indicator cell reagent, and inability to determine inclusion of a second tube of platelets and serum incubated HPA specificity of antibodies in the presence of non-HPA in the presence of the implicated drug at a concentration antibodies, e.g., HLA class I and ABO. Variations of this between 1 and 10 mM (Fig. 1).29,30 It is vital that the tube assay have been developed that use anti-human Ig-coated of platelets with patient’s serum and drug be washed with latex beads in place of indicator RBCs, which significantly drug solution to preserve platelet-bound drug-dependent extends the shelf life of the indicator reagent.11 antibodies. An FLR is calculated by dividing the MFI ob- The importance of including tests that use intact plate- tained with patient’s serum in the presence of drug by the lets, like flow cytometry and SPRCA, is that the antigenic MFI obtained with patient’s serum tested with only buffer. epitopes recognized by platelet-reactive antibodies are bet- FLR ratios of 2.0 or greater are usually considered positive ter preserved on intact platelets. Some HPA-3 antigens can for drug-dependent antibodies. only be detected using intact platelets because HPA-3a and Flow cytometry can also be used for detection of platelet- HPA-3b epitopes are labile in storage or apparently dis- associated IgG (PAIgG) and IgM in suspected cases of rupted during detergent solubilization steps required when ITP.31–33 The patient’s platelets are simply isolated, washed, using antigen capture assays,38,39 which are described in the and incubated with FITC-anti-IgG and PE-anti-IgM; fluo- next section. rescence is detected by flow cytometry.34 The method is sensitive, simple, and relatively quick to perform, but has Antigen-capture assays lower specificity owing to nonspecifically bound platelet im- Antigen-capture ELISA and modified antigen-capture munoglobulins like immune complexes that can be present ELISA on platelets of patients with nonimmune thrombocytope- The principle of antigen-capture assays is to adhere nias. individual platelet GPs, e.g., GPIIb/IIIa, to a solid support With the adaptation of performing platelet antibody (polystyrene bead or microtiter plate well) for detection of flow cytometry tests in 96-well microtiter plates, the use of platelet-specific antibodies (Fig. 2). GP captured from pan- frozen-stored platelets, and the availability of flow cytom- els of platelets with different HPA phenotypes can be used eters with high-throughput sampler (HTS) attachments, to determine the specific HPA (e.g., HPA-1a) recognized by flow cytometry has become a quick, high-volume method antibody present in serum samples. 15 for platelet antibody detection, including as a method for The antigen-capture ELISA (ACE) and modified platelet crossmatching to select compatible platelets for transfusion to highly alloimmunized thrombocytopenic patients.35,36 Disadvantages include the requirement for an expensive instrument, specialized reagents, and the inability to determine antibody HPA specificity in the presence of non-HPA antibodies, e.g., HLA class I.

Solid-phase RBC adherence Another method for detection of platelet antibodies that uses intact platelets is the solid-phase RBC adherence (SPRCA) or mixed-passive hemagglutination assay (MPHA), first developed by Shibata et al.13 and later made available as a commercial kit (Capture-P, Immucor, Inc., Norcross, GA).14,37 SPRCA uses intact target platelets adhered to the wells of a round-bottom 96-well microtiter plate. Patient test serum or plasma is added to the wells and incubated, followed by repeated washing of the wells. Fig. 2 Depiction of three different antigen-capture ELISAs for Platelet-bound antibodies are detected by addition of in- platelet antibody detection and identification:left , antigen-capture dicator RBCs coated with anti-human IgG followed by ELISA (ACE); center, modified ACE (MACE);right , monoclonal centrifugation at an optimal speed. Results are determined antibody immobilization of platelet antigens (MAIPA).

128 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Platelet antibodies and antigens in the clinical laboratory antigen-capture ELISA (MACE)17 are two platelet anti- the lysate by anti-mouse IgG polyclonal antibodies adhered gen-capture assays developed in 1987 and 1989, respectively. to the well of a microtiter plate (Fig. 2). A flow cytometry In the ACE, a platelet GP-specific monoclonal antibody is version of the MAIPA has been reported, in which a small adhered by simple electrostatic and hydrophobic interac- bead instead of a microplate well was used as a solid sup- tions to the bottom of a polystyrene microtiter plate well. port for monoclonal antibody capture of GP.42 Platelet GPs are released from the cell membrane when solubilized in a nonionic detergent such as Triton X-100. Protein bead arrays The platelet lysate is then incubated with the immobilized Antigen capture assays are being developed in which monoclonal antibody to capture the specific GP. The well is individual platelet GPs are attached to polystyrene micro- then washed and incubated with a protein solution, most beads.43,44 Protein bead arrays of this nature were first de- commonly 1 to 2 percent BSA in PBS, to block sites on the veloped and used for detection of antibodies to HLA class plastic not coated with antibody-GP complexes to prevent I and class II antigens.45 These highly sensitive assays are nonspecific binding of Ig from patient serum. Serum is now considered standard tests for HLA antibody detection added to the well and incubated for 30 to 60 minutes at either and identification. The microbeads used have different bead room temperature or 37°C. After multiple washes with buf- “addresses,” with different antigens attached to beads bear- fer, IgG platelet antibodies in the serum that bind to spe- ing variable amounts of flourescent dye incorporated into cific HPA on the GP are detected with an enzyme-labeled them at manufacture. For example, four individual purified anti-human IgG reagent after addition of enzyme substrate platelet GPs can be attached to each of four separate beads and measurement of the optical density in an ELISA plate containing different ratios of a green and a red fluorescent reader (Fig. 2). In a similar technique, designated the im- dye: GPIIb/IIIa on bead 1, GPIb/IX on bead 2, GPIa/IIa on munobead assay, a polystyrene bead instead of a microtiter bead 3, and GPIV on bead 4 (Fig. 3). The four beads are then well is used.18 The MACE is simply a modified version of the incubated in the same well of a microtiter plate or test tube, ACE in which platelets are first incubated with serum and together with patient’s serum containing antibodies, for ex- washed before detergent lysis (Fig. 2). The MACE was de- ample, against GPIIb/IIIa. The beads are then washed and veloped to prevent false-positive reactions that can occur in incubated with PE-labeled anti-human IgG. Fluorescence the ACE caused by naturally occurring human anti-mouse analysis of beads is performed on a Luminex instrument IgG antibodies (HAMA).40,41 HAMA are present at detectable (Luminex Corp., Austin, TX). The Luminex is a small flow levels in about 30 percent or more of human sera. HAMA cytometer with two lasers: one (reporter) laser activates the recognize sugar epitopes present on the mouse monoclonal fluorescent-labeled beads to identify their unique fluores- IgG antibodies used for GP capture.41 Incubation of patient cent addresses, and the other (detection) laser activates the serum with platelets first, followed by washing, removes PE-labeled anti-Ig for detection of antibody bound to GP on HAMA before addition of GP-antibody complexes to immo- the bead. Each bead emits a specific intensity of light based bilized mouse monoclonal IgG. Dilution of patient serum in on the amount of green fluorescence it contains (bead ad- buffer containing 10 percent mouse serum is often used to dress), allowing for separation and identification of each of inhibit interference from HAMA in the ACE. the four beads and simultaneous detection of PE fluores- The ACE assay not only is used for detection and iden- cence (patient antibody bound). Beads with PE fluorescence tification of platelet alloantibodies, but also has been used above an established threshold are interpreted as positive. successfully for detection of platelet autoantibodies in the Knowing which GP is attached to a specific, defined bead serum of patients with ITP.34 An acid eluate of the patient’s washed platelets can be incubated with captured GPIIb/ IIIa, GPIb/IX, and GPIa/IIa, the most common GP targets of platelet autoantibodies, as a direct test. Patient’s serum can be incubated with the same targets for indirect testing. Elution of autoantibodies from the patient’s platelets (breaks up immune complexes nonspecifically attached) and testing against platelet GPs significantly increases the specificity of the assay.34

Monoclonal antibody immobilization of platelet antigens The monoclonal antibody immobilization of platelet Fig. 3 Depiction of a protein bead array assay for platelet antibody antigens (MAIPA) assay was developed by Mueller- detection and identification using microbeads coated with differ- Eckhardt and colleagues,16 also in 1987, and is more popu- ent platelet GPs, e.g., GPIIb/IIIa, GPIb/IX, GPIa/IIa, and GPIV. lar outside the United States. As in the MACE, platelets are Patient’s serum with antibody against GPIIb/IIIa reacts only with first incubated with serum and washed, but then incubated GPIIb/IIIa beads and is detected with PE-labeled anti-human IgG with monoclonal antibody and washed again, followed by by fluorescence detection as depicted on the right side of the figure. detergent lysis. GP-antibody complexes are captured from

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 129 B.R. Curtis and J.G. McFarland

allows for the determination of GP and HPA specificity (Fig. PF4 ELISA 3). Modifications of this method have been reported using The PF4 ELISA was developed by Amiral et al.48 af- a standard flow cytometer and GPs attached to PE-labeled ter their discovery that PF4 complexed with heparin was beads.43 Protein bead arrays are highly sensitive and easily the target for most HDA. This finding led to the produc- adapted to high-throughput applications because multiple tion of several commercial kits for antibody detection. PF4 antibodies with different HPA specificities can be detected complexed with UFH or polyvinyl sulfate (PVS) is bound by multiplexing all GP-coated beads in a single well of the to the wells of a microtiter plate. Patient sera diluted from plate. Issues with assay specificity have been observed in 1:50 to 1:100 are added to the wells, and after incubation which antibodies are detected in sera from some normal, and washing, an enzyme-labeled polyclonal anti-human healthy people without previous history of alloimmuniza- IgG/A/M reagent is added for detection of HDA. Reactions tion. This reactivity is probably related to detection of anti- with patient sera are typically considered positive if they bodies in human sera that recognize neoantigens formed by exceed an optical density value of 0.400. Testing the pa- conformational changes that occur in GPs when attached to tient’s serum in parallel with 100 U/mL of added UFH to the beads (authors’ personal observations). inhibit HDA reactivity can be performed to confirm positive Recently, Fujiwara et al.44 applied this technology in the results. Although debate still exists about the value of this form of the immunocomplex capture fluorescence analysis added step,49 it can help distinguish clinically significant (ICFA) assay for detection of platelet and HLA class I anti- positives from the many modest positive results seen with bodies, but instead used GP-specific monoclonal antibodies these highly sensitive assays in patients with a low clinical attached to the beads to capture antibody-GP immune com- index for HIT.50,51 plexes from detergent lysates prepared after incubating the Another controversy is over the importance of detect- patient’s serum with platelets. In serum dilution studies, ing IgM and IgA antibodies in addition to IgG.52 IgG HDA the ICFA showed fourfold higher sensitivity compared with are thought to be more clinically significant because IgM the SPRCA for detection of anti-HPA5b and anti-GPIV, and and IgA HDA cannot crosslink and activate FcγRIIa recep- for detection of HPA-3a antibodies compared with MAIPA, tors on platelets, which is why IgA and IgM HDA are not but lower sensitivity for detection of HPA-3a when com- detected in the gold standard SRA. However, reports exist pared with SPRCA. of patients with high clinical suspicion for HITT in which HDA of only the IgA or IgM class could be detected.53,54 It is Tests for heparin-dependent antibodies likely in rare cases that some strong IgM and IgA antibodies Heparin-induced thrombocytopenia without (HIT) can cause HIT(T). or with thrombosis (HITT) is a serious immune-mediated complication of heparin therapy characterized by (1) sus- Platelet aggregation assay tained thrombocytopenia that occurs during heparin ther- The platelet aggregation assay using citrate platelet- apy, (2) a platelet count that recovers after discontinuation rich plasma (PRP) has been in use for detection of HDA of heparin, and (3) no other evident cause of thrombocy- since the early 1970s and is commonly used today.47 Platelet topenia.46 HIT is caused by heparin-dependent antibodies aggregation is evaluated in a standard platelet aggregom- (HDA) that target neoepitopes that form on PF4 owing to eter using PRP incubated 1:1 with patient’s citrate plasma conformational changes induced by association with hepa- and buffer or various concentrations of UFH (0.1–1.0 U/ rin. Immune complexes of PF4-heparin-HDA cross-link mL). Results are compared with those obtained with citrate FcγRIIa receptors on platelets, causing platelet activation plasma from a normal healthy donor. Performing the test and release of additional PF4 from platelet alpha gran- in a lumi-aggregometer allowing for simultaneous monitor- ules and formation of more PF4–unfractionated heparin ing of platelet activation and release of ATP may increase (UFH)–HDA complexes. Sensitive laboratory tests can reli- sensitivity of the method. Platelet aggregation is inexpen- ably confirm the diagnosis in patients suspected of having sive and instrumentation is readily available; however, the HIT.47 Two types of tests are available: immunoassays and test is time-consuming and lacks sensitivity and specificity functional assays. Immunoassays detect HDA in patient compared with other assays. sera that target the PF4 protein complexed with circulating heparin, whereas functional assays detect activation of SRA normal target platelets after exposure to patient antibody in The serotonin-release assay (SRA), first described by the presence of heparin. Several tests have been developed Sheridan et al.55 in 1986, continues to be considered the for detection of HDA including the PF4 ELISA, platelet ag- gold standard assay for HDA. The SRA uses washed plate- gregation, 14C serotonin-release assay (SRA), and heparin- lets. They are thought to increase sensitivity, because cal- induced platelet activation (HIPA) assay. The three most cium can be added to optimize platelet activation, and to frequently used of these assays (PF4 ELISA, SRA, platelet increase specificity, by removing plasma proteins that aggregation) will be discussed here. could cause unwanted platelet activation. 14C-serotonin

130 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Platelet antibodies and antigens in the clinical laboratory

is incubated with citrate PRP, which is taken up by dense Allele-specific PCR (ASP) granules inside the platelets. Platelets are washed once to Of the 21 biallelic platelet alloantigens described to remove plasma and residual radioactivity and suspended date, 20 are the result of a single nucleotide polymorphism in Tyrode’s buffer containing physiologic calcium. Labeled (SNP) in the genes that encode the proteins that express platelets, heat-inactivated test serum or citrate plasma, and them (Table 1).21,23 SNPs are single base changes in a gene 0.1 U/mL or 100 U/mL of UFH are incubated in wells of a that differentiate one allele from its antithetical allele. For microtiter plate with constant stirring. Sera with HDA acti- example, if the base T is at position 196 of the gene integrin vate platelets to release 14C-serotonin. Platelets are pelleted, beta III (ITGB3) that encodes GPIIIa, then an individual and released radioactivity is measured in the supernatant expresses the HPA-1a form of GPIIIa; however, if T is re- using a beta counter. Typically sera that release more than placed by a C at position 196 of ITGB3, then the person ex- 20 percent of the radioactivity with 0.1 U/mL UFH and less presses the HPA-1b form of GPIIIa on his or her platelets. than 20 percent at 100 U/mL UFH are considered positive Taking advantage of this simple difference, primers com- for HDA. Testing performed in the presence of blocking plementary to short sequences that flank the 196 position of monoclonal antibodies specific for FcγRIIa can also be per- ITGB3 are mixed with patient genomic DNA, individual de- formed to demonstrate inhibition of the reaction,47 but this oxynucleotide triphosphates (dNTPs), Taq polymerase, and step does not confirm that the reaction is heparin-dependent, buffer containing an optimum concentration of Mg2+. The and could actually mask the presence of a non–heparin- mixture is subjected to 20 to 45 cycles of high-temperature dependent antibody, e.g., HLA class I, HPA, or platelet denaturation of DNA to separate the double strands, fol- autoantibodies, if testing in the presence of 100 U/mL UFH lowed by lowering the temperature to allow the primers to is not also performed. Non-HDA can usually be differentiated attach to (anneal) the separated strands. Thermostable Taq from HDA because the former will not be inhibited in the polymerase then extends the primers to copy (amplify) the presence of 100 U/mL UFH. stretch of DNA across the area containing the base at posi- Despite being the gold standard, the SRA is a difficult tion 196. The resultant PCR amplicon is then separated out, assay to master. It also requires the use of relatively fresh usually on an agarose gel, stained with ethidium bromide, platelets, good controls, and, of course, radioactivity. and exposed to UV light for visual fluorescence detection of the gel band (Fig. 4). If the DNA being genotyped contains a Platelet Antigen Typing C instead of T at 196, then Taq cannot amplify the separated Serologic typing methods DNA strand with the primers complementary to DNA with The platelet antibody detection methods previously a T at 196, and thus no PCR amplicon forms and no band described have been adapted for use in serologic typing of appears in the gel. platelet antigens. However, serologic typing has declined dramatically in use because it is difficult and time-consuming to perform, and requires the availability of large volumes of highly specific anti-sera and the need to isolate sufficient platelets from whole blood.

Molecular typing methods With the development of PCR and determination of the molecular basis of HPAs, serologic platelet antigen typing was largely replaced by DNA typing. This shift was a highly significant event in improving laboratory testing and ad- vancing the ability to better diagnose alloimmune platelet disorders. Typing DNA for allelic variations that determine HPA was made possible with identification of the HPA- 1a/1b (PlA1/A2) polymorphism in the gene encoding GPIIIa (βIII integrin).19 This finding resulted in development of the first genotyping methods for HPA polymorphisms.19,20 Since then, the explosion in molecular diagnostics during Fig. 4 Ultraviolet light transillumination results of ethidium the last 15 years has resulted in the development of a variety bromide–stained 2 percent agarose electrophoresis gel showing of methods for platelet genotyping that are much improved separated allele-specific PCR products from genotyping assay for in their speed and accuracy. It is beyond the scope of this HPA-1. Lane 1: DNA size standards; lanes 2, 4, 6: HPA-1a allele article to describe every assay; therefore, we will describe typings; lanes 3, 5, 7: HPA-1b allele typings. One individual’s HPA-1a/1b genotyping is shown for each pair of lanes: lanes 2 some of the most common techniques in use today and di- 56,57 and 3, HPA-1a/1a; lanes 4 and 5, HPA-1a/1b; lanes 6 and 7, rect readers to more detailed reviews of the subject. HPA-1b/1b. Human growth hormone (HGH) PCR control band is denoted by arrow.

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 131 B.R. Curtis and J.G. McFarland

Allele-specific PCR is relatively simple and inexpen- tion. For MCA, the temperature is raised after PCR, and the sive to perform and continues to be a popular method for shorter reporter probes, which bind to the amplicon with HPA genotyping.58 The only drawbacks of the method are reduced stability, dissociate (melt) at lower temperatures, requirements for post-PCR processing and subjective resulting in loss of fluorescence signal that can be moni- interpretation, which combined make typing slower and tored as a melting curve. The thermal melting point (Tm) of increase the potential for human error. the reporter probes is sequence-specific, allowing for differentiation of two alleles, e.g., HPA-1a and HPA-1b. By taking Melting curve analysis the negative first derivative (–dF/dT) of the melting curve, Melting curve analysis (MCA) is a method to measure melting temperature histograms for individual probes can the dissociation (melting) attributes of double-stranded be created for easy visual comparison and discrimination of DNA (dsDNA) that occur during heating.59,60 At higher tem- homozygous wild-type and mutant and heterozygous geno- peratures, the two strands of dsDNA separate, and the heat types.57 The assay lends itself to multiplexing to increase required for the intrabase hydrogen bonds to be broken is throughput, and it is even possible to find new SNPs by rec- dependent on their length, GC content, and complemen- ognizing changes in the architecture of melting curves (Fig. 5). tarity. One popular method of MCA for platelet antigen genotyping uses the LightCycler instrument (Roche Diag- 5′-Nuclease or TaqMan assay nostics, Indianapolis, IN).59,60 The assay begins with PCR The 5′-nuclease or TaqMan assay is another popular amplification of patient DNA using primer pairs that flank genotyping method that also uses FRET chemistry.61,62 The the SNP of interest. As PCR product is formed, it is detected assay requires two sequence-specific probes, one that will with fluorescent-labeled oligonucleotide probes that anneal hybridize with the DNA sequence containing the SNP for to sequences adjacent to the SNP. One is a shorter allele- one allele (e.g., HPA-1a), and the other specific for the al- specific “reporter” probe and the other, a longer “anchor” ternate allele (e.g., HPA-1b). Probes are labeled at their 5′ probe. When both probes are bound, light from a diode end with fluorochromes having different fluorescent emis- activates the fluorescent labeled (usually fluorescein) an- sion spectra and a quencher molecule on their 3′ ends that chor probe and the fluorescence resonance energy trans- prevents fluorochrome fluorescence as long as the probe is fer (FRET) from the anchor probe to the reporter probe intact. Sequence-specific primers that flank the HPA SNP excites it to fluoresce. The fluorescence signal detected by of interest are used to amplify the DNA by PCR. Probes the LightCycler is proportional to the amount of amplicon and primers are added together with DNA for PCR, and as present, allowing for real-time monitoring and quantifica- Taq polymerase amplifies and extends the primers, the 5′-nuclease activity inherent in Taq degrades the probe hy- bridized to the SNP-containing DNA sequence. This releas- es the 5′ fluorochrome, separating it from the quencher and allowing it to fluoresce, which generates a detectable, quan- tifiable signal. The allele-specific probe that is not complementary does not bind and prevents Taq from acting on the probe to release its fluorescence. The 5′-nuclease and MCA methods lend themselves to moderately high throughput but are limited in their multiplexing capabilities by fluorochrome availability for probe labeling.

High-throughput platelet genotyping methods Conventional PCR-ASP assays with gel end point Fig. 5 Melting curve analysis by LightCycler for human platelet for platelet genotyping have begun to give way to high- antigen (HPA) genotyping. Shorter reporter fluorescence probes throughput methods. These methods are not only faster bind with reduced stability to the DNA template containing the and higher volume, but they are less tedious to perform HPA SNP of interest, and as the temperature of the reaction is and they reduce human errors. Several methods using slowly raised they dissociate or “melt” off with resultant loss of fluorescence signal that displays as a melting curve. Each peak different chemistries and detection platforms have been 63 represents a different allele (e.g., HPA-3a left peaks and HPA-3b reported, including glass slide and microplate arrays, a right peaks), which allows for genotype determination of homozy- liquid bead array,64 and mass spectrometry.65 Obstacles to gous wild-type, heterozygous, and homozygous mutant genotypes, high-throughput genotyping methods are the high costs of as denoted by arrows. Potential new SNPs can be detected with instrumentation and specialized reagents and the need for this method as “hybrid” peaks that fall between the two individual experienced individuals to implement, operate, and inter- allele peaks, as denoted in the figure. pret the data. High-throughput genotyping platforms by

132 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Platelet antibodies and antigens in the clinical laboratory nature could have important applications, to screening all 7. Horai S, Claas FH, van Rood JJ. Detection of platelet an- pregnant women for HPA-1a to identify those requiring antibodies by enzyme-linked immunosorbent assay (ELI- tenatal treatment to prevent NATP, and for screening large SA) on artificial monolayers of platelets. Immunol Lett numbers of blood donors to find matched platelet donors 1981;3:67–72. for highly alloimmunized patients. 8. von dem Borne AE, Verheugt FW, Oosterhof F, et al. Molecular tests for platelet genotyping are not with- A simple immunofluorescence test for the detection of out limitations. Rare cases have been reported of unknown platelet antibodies. Br J Haematol 1978;39:195–207. SNPs located near the HPA polymorphism being typed 9. Rosenfeld CS, Nichols G, Bodensteiner DC. Flow cyto- that prevent proper annealing of ASP PCR primers and metric measurement of antiplatelet antibodies. Am J probes, causing false-negative results.66 Silent HPA alleles Clin Pathol 1987;87:518–22. and mutations that affect GP conformation, which produce 10. Lazarchick J, Hall SA. Platelet-associated IgG assay discrepant genotype vs. phenotype results, have also been using flow cytometric analysis. J Immunol Methods reported.67,68 However, once identified these problems can 1986;87:257–65. be overcome, usually by redesign of primer and probe se- 11. Ramos RR, Curtis BR, Chaplin H. A latex particle assay quences. for platelet-associated IgG. Transfusion 1992;32:235–8. 12. Ogden DM, Asfour A, Koller C, Lichtiger B. Platelet cross- Summary matches of single-donor platelet concentrates using a la- Platelet antibody and antigen testing has advanced tex agglutination assay. Transfusion 1993;33:644–50. considerably since its inception with the use of first- 13. Shibata Y, Juji T, Nishizawa Y, et al. Detection of plate- generation assays that used the functional aspects of platelet antibodies by a newly developed mixed agglutination lets for antibody and antigen detection. Introduction of the with platelets. Vox Sang 1981;41:25–31. PSIFT assay paved the way for development of even more 14. Rachel JM, Summers TC, Sinor LT, Plapp FV. Use of a sophisticated flow cytometry immunofluorescence assays, solid phase red blood cell adherence method for pre- solid-phase assays, and monoclonal antigen-capture assays. transfusion platelet compatibility testing. Am J Clin Antigen-capture assays such as the ACE, MACE, and MAI- Pathol 1988;90:63–8. PA allowed for not only antibody detection but also identi- 15. Furihata K, Nugent DJ, Bissonette A, et al. On the as- fication of the specific HPA targeted by platelet antibodies sociation of the platelet-specific alloantigen, Pena, with and serologic typing. Identification of the HPA-1a/1b allelic glycoprotein IIIa. Evidence for heterogeneity of glyco- polymorphism in the ITB3 gene enabled development of protein IIIa. J Clin Invest 1987;80:1624–30. ASP-PCR genotyping assays that quickly replaced serologic 16. Kiefel V, Santoso S, Weisheit M, Mueller-Eckhardt C. typing and greatly advanced the field of platelet immunol- Monoclonal antibody-specific immobilization of platelet ogy and laboratory diagnosis of immune platelet disorders. antigens (MAIPA): a new tool for the identification of High-throughput platelet genotyping methods and protein platelet-reactive antibodies. Blood 1987;70:1722–6. bead arrays for platelet antibody detection and identifica- 17. Menitove JE, Pereira J, Hoffman R, et al. Cyclic throm- tion will further advance the field in the coming years. bocytopenia of apparent autoimmune etiology. Blood 1989;73:1561–9. References 18. McMillan R, Tani P, Millard F, et al. Platelet-associated 1. Pfueller SL, Cosgrove L, Firkin BG, Tew D. Relationship and plasma anti-glycoprotein autoantibodies in chronic of raised platelet IgG in thrombocytopenia to total plate- ITP. Blood 1987;70:1040–5. let protein content. Br J Haematol 1981;49:293–302. 19. Newman PJ, Derbes RS, Aster RH. The human platelet 2. Schwartz KA. Platelet antibody: review of detection alloantigens, PlA1 and PlA2, are associated with a leu- methods. Am J Hematol 1988;29:106–14. cine33/proline33 amino acid polymorphism in mem- 3. LoBuglio AF, Court WS, Vinocur L, et al. Immune throm- brane glycoprotein IIIa, and are distinguishable by DNA bocytopenic purpura. Use of a 125I-labeled antihuman typing. J Clin Invest 1989;83:1778–81. IgG monoclonal antibody to quantify platelet-bound 20. McFarland JG, Aster RH, Bussel JB, et al. Prenatal IgG. N Engl J Med 1983;309:459–63. diagnosis of neonatal alloimmune thrombocytope- 4. Shaw GM, Axelson J, Maglott JG, LoBuglio AF. Quantifi- nia using allele-specific oligonucleotide probes. Blood cation of platelet-bound IgG by 125I-Staphylococcal pro- 1991;78:2276–82. tein A in immune thrombocytopenic purpura and other 21. Metcalfe P, Watkins NA, Ouwehand WH, et al. No- thrombocytopenic disorders. Blood 1984;63:154–61. menclature of human platelet antigens. Vox Sang 5. Schiffer CA, Young V. Detection of platelet antibodies 2003;85:240–5. using a micro-enzyme-linked immunosorbent assay 22. Curtis BR, Edwards JT, Hessner MJ, et al. Blood group (ELISA). Blood 1983;61:311–17. A and B antigens are strongly expressed on platelets of 6. Horai S, Claas FH, van Rood JJ. Detection of plate- some individuals. Blood 2000;96:1574–81. let antibodies by enzyme-linked immunosorbent assay (ELISA). Comparative studies with the indirect immunofluorescence assay. Immunol Lett 1981;3:155–8.

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 133 B.R. Curtis and J.G. McFarland

23. Curtis BR, McFarland JG. Platelet immunology and al- 39. Harrison CR, Curtis BR, McFarland JG, et al. Severe neo- loimmunization. In: Simon TL, Dzik WH, Snyder E, et natal alloimmune thrombocytopenia caused by antibod- al., eds. Rossi’s principles of transfusion medicine. Phila- ies to human platelet antigen 3a (Baka) detectable only in delphia: Lippincott Williams & Wilkins, 2008:203–17. whole platelet assays. Transfusion 2003;43:1398–402. 24. Smith GA, Ranasinghe E, Ouwehand WH. The impor- 40. Hewitt J, Burton IE. Incidence of autoantibodies to tance of using multiple techniques for detection of plate- GPIIb/IIIa in chronic autoimmune thrombocytopenic let antibodies. Vox Sang 2007;93:306–8. purpura may be overestimated by the MAIPA. Br J Hae- 25. Garratty G, Arndt P. Applications of flow cytofluorome- matol 1994;86:418–20. try to transfusion science. Transfusion 1995;35:157–78. 41. Klee GG. Human anti-mouse antibodies. Arch Pathol 26. Visentin GP, Wolfmeyer K, Newman PJ, Aster RH. De- Lab Med 2000;124:921–3. tection of drug-dependent, platelet-reactive antibodies 42. Helmberg W, Folsch B, Wagner T, Lanzer G. Detection by antigen-capture ELISA and flow cytometry. Transfu- and differentiation of platelet-specific antibodies by flow sion 1990;30:694–700. cytometry: the bead-mediated platelet assay. Transfu- 27. Curtis BR, McFarland JG, Wu GG, et al. Antibodies in sion 1997;37:502–6. sulfonamide-induced immune thrombocytopenia rec- 43. Bakchoul T, Meyer O, Agaylan A, et al. Rapid detection ognize calcium-dependent epitopes on the glycoprotein of HPA-1 alloantibodies by platelet antigens immobilized IIb/IIIa complex. Blood 1994;84:176–83. onto microbeads. Transfusion 2007;47:1363–8. 28. Von Drygalski A, Curtis BR, Bougie DW, et al. Vancomycin- 44. Fujiwara K, Shimano K, Tanaka H, et al. Application of induced immune thrombocytopenia. N Engl J Med bead array technology to simultaneous detection of hu- 2007;356:904–10. man leucocyte antigen and human platelet antigen anti- 29. Aster RH, Bougie DW. Drug-induced immune thrombo- bodies. Vox Sang 2009;96:244–51. cytopenia. N Engl J Med 2007;357:580–7. 45. Pei R, Wang G, Tarsitani C, et al. Simultaneous HLA 30. Aster RH, Curtis BR, McFarland JG, Bougie DW. class I and class II antibodies screening with flow cytom- Drug-induced immune thrombocytopenia: pathogen- etry. Hum Immunol 1998;59:313–22. esis, diagnosis, and management. J Thromb Haemost 46. Warkentin TE. Clinical picture of heparin-induced 2009;7:911–18. thrombocytopenia. In: Warkentin TE, Greinacher A, eds. 31. Christopoulos CG, Kelsey HC, Machin SJ. A flow- Heparin-induced thrombocytopenia. New York: Marcel cytometric approach to quantitative estimation of plate- Dekker, 2000:43–80. let surface immunoglobulin G. Vox Sang 1993;64:106– 47. Warkentin TE. Laboratory testing for heparin- 15. induced thrombocytopenia. J Thromb Thrombolysis 32. Stockelberg D, Hou M, Jacobsson S, et al. Detection of 2000;10(Suppl 1):35–45. platelet antibodies in chronic idiopathic thrombocy- 48. Amiral J, Bridey F, Dreyfus M, et al. Platelet factor 4 topenic purpura (ITP). A comparative study using flow complexed to heparin is the target for antibodies gen- cytometry, a whole platelet ELISA, and an antigen cap- erated in heparin-induced thrombocytopenia. Thromb ture ELISA. Eur J Haematol 1996;56:72–7. Haemost 1992;68:95–6. 33. Hezard N, Simon G, Mace C, et al. Is flow cytometry ac- 49. Warkentin TE. Confirmatory procedure and other ma- curate enough to screen platelet autoantibodies? Trans- neuvers to assess pathogenicity of platelet factor 4 fusion 2008;48:513–18. (PF4)-dependent antibodies—distinguishing “signal” 34. Davoren A, Bussel J, Curtis BR, et al. Prospective evalu- from “noise”. Thromb Haemost 2008;100:523–4. ation of a new platelet glycoprotein (GP)-specific assay 50. Whitlatch NL, Perry SL, Ortel TL. Anti-heparin/platelet (PakAuto) in the diagnosis of autoimmune thrombocy- factor 4 antibody optical density values and the confir- topenia (AITP). Am J Hematol 2005;78:193–7. matory procedure in the diagnosis of heparin-induced 35. Kohler M, Dittmann J, Legler TJ, et al. Flow cytometric thrombocytopenia. Thromb Haemost 2008;100:678– detection of platelet-reactive antibodies and application 84. in platelet crossmatching. Transfusion 1996;36:250–5. 51. Dellaria SM, Greilich PE, Sarode R. Urgent cardiac sur- 36. Skogen B, Christiansen D, Husebekk A. Flow cytomet- gery in a patient with a remote history of heparin-induced ric analysis in platelet crossmatching using a plate- thrombocytopenia: use of the anti-PF4/heparin ELISA let suspension immunofluorescence test. Transfusion confirmatory test. Anesth Analg 2008;107:379–81. 1995;35:832–6. 52. Warkentin TE, Sheppard JA, Moore JC, et al. Laborato- 37. Rachel JM, Sinor LT, Tawfik OW, et al. A solid-phase red ry testing for the antibodies that cause heparin-induced cell adherence test for platelet cross-matching. Med Lab thrombocytopenia: how much class do we need? J Lab Sci 1985;42:194–5. Clin Med 2005;146:341–6. 38. Lin M, Shieh SH, Liang DC, et al. Neonatal alloim- 53. Amiral J, Wolf M, Fischer A, et al. Pathogenicity of IgA mune thrombocytopenia in Taiwan due to an antibody and/or IgM antibodies to heparin-PF4 complexes in against a labile component of HPA-3a (Baka). Vox Sang patients with heparin-induced thrombocytopenia. Br J 1995;69:336–40. Haematol 1996;92:954–9.

134 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Platelet antibodies and antigens in the clinical laboratory

54. McFarland JG, Curtis BR, Lochowicz AJ, et al. Cor- genotyping for blood groups: comparison of gene array relation of PF4 ELISA results and clinical likelihood of and 5′-nuclease assay techniques with human platelet HIT—can we improve on specificity? Blood 2008;(Sup- antigen as a model. Transfusion 2005;45:654–9. pl) 112:3423. 63. de Haasa M, van der Schoota CE, Beiboera SHW, et al. 55. Sheridan D, Carter C, Kelton JG. A diagnostic test for Red blood cell and platelet genotyping: from current heparin-induced thrombocytopenia. Blood 1986;67:27– practice to future high-throughput donor typing. Trans- 30. fus Med Hemother 2006;33:260–6. 56. Hurd CM, Cavanagh G, Schuh A, et al. Genotyping for 64. Curtis BR, Fick A, Lochowicz AJ, et al. Neonatal alloim- platelet-specific antigens: techniques for the detection of mune thrombocytopenia associated with maternal- single nucleotide polymorphisms. Vox Sang 2002;83:1– fetal incompatibility for blood group B. Transfusion 12. 2008;48:358–64. 57. Curtis BR. Genotyping for human platelet alloantigen 65. Garritsen HS, Fan AX, Bosse N, et al. Matrix-assisted la- polymorphisms: applications in the diagnosis of al- ser desorption/ionization time-of-flight mass spectrom- loimmune platelet disorders. Semin Thromb Hemost etry for genotyping of human platelet-specific antigens. 2008;34:539–48. Transfusion 2009;49:252–8. 58. Skogen B, Bellissimo DB, Hessner MJ, et al. Rapid deter- 66. Bertrand G, Bianchi F, Chenet C, et al. New mutation in mination of platelet alloantigen genotypes by polymerase the platelet beta3-integrin gene: implication for the di- chain reaction using allele-specific primers. Transfusion agnosis of fetomaternal alloimmunization. Transfusion 1994;34:955–60. 2006;46:2138–41. 59. Nauck MS, Gierens H, Nauck MA, et al. Rapid genotyp- 67. Skogen B, Wang R, McFarland JG, Newman PJ. A ing of human platelet antigen 1 (HPA-1) with fluorophore- dinucleotide deletion in exon 4 of the PlA2 allelic form labelled hybridization probes on the LightCycler. Br J of glycoprotein IIIa: implications for the correlation of Haematol 1999;105:803–10. serologic versus genotypic analysis of human platelet al- 60. Randen I, Sorensen K, Killie MK, Kjeldsen-Kragh J. loantigens. Blood 1996;88:3831–6. Rapid and reliable genotyping of human platelet antigen 68. Smith GA, Rankin A, Riddle C, et al. Severe fetomaternal (HPA)-1, -2, -3, -4, and -5 a/b and Gov a/b by melting alloimmune thrombocytopenia due to anti-human plate- curve analysis. Transfusion 2003;43:445–50. let antigen (HPA)-1a in a mother with a rare and silenced 61. Kjaer KM, Jaegtvik S, Husebekk A, Skogen B. Human ITGB3*0101 (GPIIIa) allele. Vox Sang 2007;93:325–30. platelet antigen 1 (HPA 1) genotyping with 5′ nuclease assay and sequence-specific primers reveals a single nu- Brian R. Curtis, MS, MT(ASCP)SBB (corresponding au- cleotide deletion in intron 2 of the HPA 1a allele of plate- thor), and Janice G. McFarland, MD, BloodCenter of Wis- let glycoprotein IIIa. Br J Haematol 2002;117:405–8. consin, Platelet & Neutrophil Immunology Laboratory, PO 62. Bugert P, McBride S, Smith G, et al. Microarray-based Box 2178, Milwaukee, WI 53201-2178.

Attention: State Blood Bank Meeting Organizers Notice to Readers Immunohematology, Journal of Blood Group Serology If you are planning a state meeting and would like cop- and Education, is printed on acid-free paper. ies of Immunohematology for distribution, please send request, 4 months in advance, to immuno@usa. redcross.org.

Attention SBB and BB Students Notice to Readers

You are eligible for a free 1-year subscription to All articles published, including communications and Immunohematology. book reviews, reflect the opinions of the authors and do not necessarily reflect the official policy of the American Ask your education supervisor to submit the name and Red Cross. complete address for each student and the inclusive dates of the training period to [email protected]. org.

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 135 Review

Investigating the possibility of drug- dependent platelet antibodies J.P. AuBuchon and M.F. Leach

Just about every physician has seen a case like this and unique form of drug-platelet interaction that will not be dis- then may call his or her laboratory-based colleagues for as- cussed here) is well known and often considered in the dif- sistance. ferential diagnosis of thrombocytopenia,6 other drugs (given Mabel Throckmorton, 67 years old, was admitted for the frequency of their use) probably account for far more treatment of community-acquired pneumonia. After obtain- cases of drug-induced thrombocytopenia. ing cultures, two appropriate antibiotics were begun, but her respiratory status deteriorated to require intubation and First Things First ventilation. Two days later, as her pulmonary status began However, before laying the blame for all thrombocy- to improve, a routine (automated) CBC noted that her plate- topenia on medications, let us first make sure that the prob- let count had dropped from 220,000/μL on admission to lem is not something else (Table 1). Is the patient’s plate- 15,000/μL. Her leukocyte count had peaked at 22,000/μL let count really low? Some patients’ platelets agglutinate and had now declined to 10,000/μL. Her hematocrit had de- in EDTA,7 and some patients have autoantibodies against creased from 38% to 33%. Her peripheral smear was unre- GPIIb/IIIa that display avidity in EDTA samples.8 In this markable except for the cytopenias. Her medications before case, the normal platelet count on admission makes these admission included a diuretic, digoxin, and a statin for cho- unlikely causes. Idiopathic autoimmune thrombocytopenia lesterol control. In addition to the antibiotics she was receiving, additional hospital medications included albuterol and Table 1. Why is the patient thrombocytopenic? omeprazole. Pseudothrombocytopenia The patient had a history of spontaneous intracranial EDTA-induced aggregation hemorrhage 30 years previously, which resolved without permanent impairment, and chemotherapy and transfu- Glycoprotein autoantibodies avid only in EDTA sion for an aggressive lymphoma 10 years previously. Mul- Production deficit tiple transfusions coupled with five pregnancies resulted in Primary marrow failure the patient’s being broadly alloimmunized to HLAs, and she Ineffective production, e.g., myelodysplasia required difficult-to-find matched platelets for transfusion Primary malignancy, e.g., leukemia support during that chemotherapy. Secondary malignancy (metastatic disease) Is the sudden and dramatic thrombocytopenia related to the primary illness, previous disease reappearance, or some- Antineoplastic medications thing nosocomial? If a drug caused the platelet count drop, Accelerated peripheral clearance which one is the culprit? Because the infection appears to be Idiopathic thrombocytopenic purpura (ITP) under effective treatment, the patient’s physician will under- Posttransfusion purpura standably be reluctant to change antibiotics at this juncture. Thrombotic thrombocytopenic purpura (TTP) However, given the patient’s previous history of intracra- Hemolytic uremic syndrome (HUS) nial hemorrhage, considerable anxiety would accompany the dramatic appearance of thrombocytopenia, particularly HELLP: hemolytic anemia, elevated liver enzymes, low platelets* when transfusion support would be difficult to obtain. How Disseminated intravascular coagulation (DIC) can the laboratory assist in the evaluation and treatment of Massive hemorrhage this patient? Drug-induced thrombocytopenia The frequency of drug-induced thrombocytopenia is Heparin-induced thrombocytopenia/thrombosis staggering. Although only 1 person in 100,000 encounters Autoantibodies (ex: procainamide) this annually in the United States,1 some medications such Fab antibodies (ex: abciximab) as trimethoprim-sulfamethoxazole or quinine/quinidine can murine increase this rate 1500-fold.2 Perhaps a quarter of all inten- Hapten-induced antibodies (ex: penicillin, cephalosporins) sive care patients exhibit thrombocytopenia as a result of Glycoprotein-complexing drug (ex: quinine) 3,4 drugs, and 6 percent of patients receiving the common an- Autoantibody recognition of glycoprotein conformational change tibiotic gentamicin make drug-dependent platelet antibodies (ex:fibans) (DDPAs).5 Although heparin-induced thrombocytopenia (a *A variant of preeclampsia.

136 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Drug-dependent platelet antibodies

(ITP) always has to be considered, but the sudden oc- and the platelets or RBCs involved in the process are inno- currence of platelet autoantibodies in this situation would cent bystanders that are removed from circulation after ad- be very unusual. A situation that is indistinguishable from sorbing or binding the complex onto their surface. However, ITP (with autoantibody formation) can occur with and well most drugs are too small to provoke an immune response, beyond the time of administration of some drugs, including and immune complexes of drug-immunoglobulin could not L-dopa, procainamide, and gold salts, among others, al- be demonstrated when sought. Furthermore, interactions of though none of these appears in the history of this case.9 The DDPAs with the cells that are removed from circulation are patient’s lymphoma may have returned to crowd out normal through binding with the antibodies’ (epitope-specific) Fab hematopoietic marrow elements, but the time course would domains rather than the Fc domain,12,13 suggesting that the likely be much more gradual. No drugs have been adminis- antibodies are interacting directly with the platelets in the tered recently that would suppress marrow function, and the presence of the drug. patient’s leukocyte response to the infection was appropriate. This theory has been supplanted by one involving (non- The story also does not fit the time course or situation for ei- covalent) interaction between the drug and a surface protein, ther disseminated intravascular coagulation (as the patient’s leading to a change in the conformation of the protein that infection appears to be resolving) or thrombotic thrombocy- would allow naturally occurring antibodies to now interact topenic purpura. The patient has not received a transfusion with the protein with sufficient avidity to provoke removal recently, so posttransfusion purpura is not a viable explana- of the cell.14,15 Most DDPAs appear to interact with GPIIb/ tion. IIIa, which binds fibrinogen and other ligands, or GPIb/IX, DDPAs can also cause reduced responsiveness to platelet the von Willebrand factor receptor, for reasons that are not transfusions. Serial posttransfusion platelet counts may be of currently well understood.16 some assistance in distinguishing the effects of DDPAs and A form of this mechanism has been invoked to explain alloantibodies: although most patients with HLA- or platelet- why first exposure to fibans can cause rapid and profound specific alloantibodies fail to show the expected posttransfu- thrombocytopenia in a small proportion of patients. Drugs of sion response in the first hour, patients with DDPAs often this class, such as tirofiban and eptifibatide, are administered have the expected 1-hour corrected count increment (CCI) after percutaneous transluminal coronary angioplasty in an but accelerated removal thereafter such that a good initial re- attempt to maintain arterial patency. These drugs have been sponse is followed by a poor CCI when measured at 18 to 24 engineered to bind tightly to a specific sequence of amino hours after transfusion. acids (arginine-glycine-aspartic acid) on GPIIb/IIIa to block Thus, on first blush, considering DDPAs as the cause of platelet activation. In doing so, slightly altered conforma- the thrombocytopenia in this case seems logical. Although a tions occur in the glycoprotein complex to allow the binding primary immune response to a drug may take several weeks of preexisting auto-antibodies, which would not have a target to much longer, an anamnestic response can occur in a few in the absence of the drug’s effect.17,18 Interestingly, although days to a week, and preformed antibodies can cause throm- both tirofiban and eptifibatide achieve their pharmacologic bocytopenia within hours.10 Given one’s multiple exposures presence through the same mechanism, the conformational to antibiotics throughout a lifetime and the potential for changes they induce in GPIIb/IIIa appear to be different be- cross-reactivity of an antibody across a class of drugs, a rapid cause antibodies capable of clearing platelets that have been or anamnestic response may occur even without document- treated with one of these drugs usually do not bind to plate- ed (or recalled) prior exposure to an antibiotic. lets treated with the other.19 Another drug in common use to inactivate circulating What’s Going On? platelets is abciximab. This is a chimeric monoclonal anti- There are multiple pathogenic mechanisms that may lie body, the murine portion of which is targeted to GPIIb/IIIa behind drug-induced thrombocytopenia, and understand- and blocks binding with fibrinogen. Some individuals have ing them can help explain the rationale for testing that may naturally occurring antibodies to the murine portion of the document the existence of DDPAs. antibody, and administration of the drug can lead to rapid Knowing that some drugs (including penicillins and profound thrombocytopenia as the anti-mouse antibody and cephalosporins) can bind covalently to proteins on bound to platelets clears the drug-platelet complex.20 Other cell membranes led to the theory of DDPA stimulation patients form such an antibody about a week after initial ex- through hapten, or neoantigen, formation. The drug anti- posure, and thrombocytopenia can develop at this time as body would thus actually be targeting a drug-protein struc- well. ture on the surface of cells, including RBCs or platelets.11 However, this is not the mechanism that explains the actions Identifying the Culprit of most DDPAs. How can the laboratory help identify the presence of a Another model that has been advanced is that of antigen- DDPA that might be causing the thrombocytopenia in this antibody complex formation. Here, the antibody is directed case? A review to ensure that other laboratory results and against the soluble form of the drug (or one of its metabolites), clinical information do not point toward a nondrug cause

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 137 J.P. AuBuchon and M.P. Leach should be performed first. Reviewing lists of most commonly via an enzyme immunoassay; the DDPA is thus detected by encountered causative agents of DDPAs would be a good conversion of a substrate to a colored product. Because stan- next step, as many clinicians may not be aware of the sub- dardized kits are not available in the United States for such stantial frequency with which some drugs induce DDPAs and testing, referral of specimens to a reference laboratory usual- thrombocytopenia. Certainly heparin should be on this list,6 ly is required. The resulting turnaround time renders the in- but the list extends far beyond this and includes some often- formation useful only in retrospect usually, because clinical implicated antibiotics (Table 2). decisions about drug discontinuation often have to be made in a much shorter time frame. Table 2. Partial listing of drugs causing immunologically mediated A commonly used approach for the detection of HLA- thrombocytopenia* and platelet-specific antibodies can be adapted for the detec- Abciximab Digoxin Omeprazole tion of drug antibodies. The solid-phase red cell adherence Acetaminophen Dipyridamole Orbofiban assay (SPRCA) technique can be adapted to detect DDPAs Acyclovir Enoxaparin Phenytoin stimulated by a wide variety of drug classes.23,24 This tech- Amikacin Eptifibatide Piperacillin nique can be learned by any skilled technologist and applied Amlodipine Famotidine Procainamide to yield clinically useful results in just a couple of hours, removing the guesswork from changing a patient’s drug regi- Amoxicillin Fluconazole Propranolol men. Validating such an assay should include comparison of Amphotericin B Furosemide Quinidine results with a laboratory having an established reputation in Ampicillin Gentamicin Quinine the field. Performance of the technique is aided by close co- Argatroban Glyburide Rifampin operation with the hospital pharmacy in acquiring the need- Atenolol Gold salts Salicylates ed “reagent” drug supply and choosing the appropriate drug Bortezomib Heparin Sulfonamides concentration to use in the system. Any assay for DDPAs will be hampered, of course, by the lack of ability to include Carbamazepine Hydrochlorothiazide Sulfonylureas drug metabolites in the test system (as these may, in some Cefazolin Imipenem Suramin cases, be the chemicals that are the true culprits14) and by Cefotetan Interferon Teicoplanin aqueous solubility limitations. If the patient also happens to Ceftazidime Linezolid Tirofiban be alloimmunized to HLA- or platelet-specific determinants, Ceftriaxone Lotrafiban Tobramycin performance of this assay will require use of reagent platelets Cimetidine Metoprolol Trimethoprim/ that lack the corresponding epitopes. This assay offers a nice sulfamethoxazole complement to a laboratory’s ability to identify antibodies Ciprofloxacin Metronidazole Valganciclovir causing heparin-induced thrombocytopenia through a solid- phase anti-platelet factor 4 (anti-PF4/heparin) enzyme im- Clarithromycin Naproxen Valproate munoassay.25 Clindamycin Nizatidine Vancomycin In cases in which the drug has caused a perturbation of Clopidogrel the immune system and the generation of a true autoanti- *This listing includes drugs reported in the literature as causes of im- body, laboratory testing will be unable to identify the culprit mune drug-induced thrombocytopenia or encountered in the authors’ experience.10 The reader is referred to a Web site provided by Dr. drug because the autoantibody will bind with all platelets. James George, University of Oklahoma, for an updated, well- Such antibodies usually target a structure in the GPIIb/IIIa characterized compendium of reports of drugs inducing thrombocytopenia: www.ouhsc.edu/platelets. Consideration should also be given complex and are indistinguishable from the autoantibod- to thrombocytopenia arising after ingestion of folk or herbal remedies. ies that arise in ITP. Further complicating assessment (and treatment) is the recognition that these autoantibodies may There are several types of assays that are applied by ref- persist long after discontinuation of the drug. Recognition of erence laboratories to detect DDPAs.3,21,22 Flow cytometry is the exact nature of the situation thus depends on identify- commonly used to detect an increase in binding of IgG to ing one of the drugs known to cause this kind of problem on platelets after incubation with drug and the patient’s serum. the patient’s medication list and demonstration of an ITP- (The order of addition—incubation of platelet with drug fol- like autoantibody (not dependent on the presence of drug) by lowed by addition of serum versus the incubation of drug with flow cytometry or SPRCA. serum followed by addition of this mixture to platelets—may depend on the pathophysiologic mechanism at work, and the What Then? alternative can be tried if the initial approach is not positive, If the drug that caused the thrombocytopenia as a result because variations are seen among patients and certainly of a DDPA can be identified and removed from the patient’s among drugs.) A similar approach of incubating normal regimen, an increase in the platelet count should be seen rel- platelets with drug and patient serum can be followed with atively quickly. The longer the serum half-life of the drug, of detection of IgG bound to platelets using goat anti-human course, the slower will be the recovery. The patient’s platelet IgG conjugated to alkaline phosphatase to detect the DDPA count usually will begin to recover (or responses to platelet

138 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Drug-dependent platelet antibodies transfusions return to normal) within 2 to 5 days. Although 7. Lau LG, Chng WJ, Liu TC. Transfusion medicine illus- antibodies to drugs can decline or disappear despite contin- trated: unnecessary transfusions due to pseudothrom- ued administration,26 discontinuation, especially in profound bocytopenia. Transfusion 2004;44:801. thrombocytopenia, would usually be advised, followed by 8. Braester A. Pseudothrombocytopenia as a pitfall in the avoidance of the drug in the future for fear of provoking an treatment of essential thrombocythemia. Eur J Haema- even stronger anamnestic response.10 Substitution of a drug tol 2003;70:251–2. in the same class as the offending drug usually does not result 9. Aster RH. Can drugs cause autoimmune thrombocy- in thrombocytopenia owing to cross-reactivity of the DDPA topenic purpura? Seminar Hematol 2000;37:229–38. despite chemical similarities of the pharmacologic agents.27 10. Kenney B, Stack G. Drug-induced thrombocytopenia. However, caution and close observation may be advisable. Arch Pathol Lab Med 2009;133:309–14. While the issue of DDPAs is being investigated, it may 11. Garratty G, Petz LD. Drug-induced immune hemolytic also be worthwhile to examine the patient’s RBCs. In the case anemia. Am J Med 1975;58:398–407. presented above, is the new anemia related to the DDPA, to 12. Christie DJ, Mullen PC, Aster RH. Fab-mediated bind- hydration status, or to a marrow-based problem? RBCs can ing of drug-dependent antibodies to platelets in quini- be subject to accelerated removal from circulation through dine- and quinine-induced thrombocytopenia. J Clin some of the same mechanisms as platelets, and, because of Invest 1985;75:310–14. the different surface proteins of RBCs, the prominence of 13. Smith ME, Reid DM, Jones CE, Jordan JV, Kautz CA, anemia versus thrombocytopenia may vary between situa- Shulman NR. Binding of quinine- and quinidine- tions. Nevertheless, recognition that drug-induced hemo- dependent drug antibodies to platelets is mediated by lysis is occurring or may occur could provide useful clinical the Fab domain of the immunoglobulin G and is not Fc information and impetus for a change in the patient’s drug dependent. J Clin Invest 1987;79:912–17. regimen. For example, in one series (preliminarily reported 14. Mueller-Eckhardt C, Salama A. Drug-induced immune in Leach et al.28), 10 of 17 selected patients with a DDPA that cytopenias: a unifying pathogenetic concept with special had been identified by SPRCA also had a positive direct an- emphasis on the role of drug metabolites. Transf Med tiglobulin test; the eluates from all 10 patients failed to react Rev 1990;4:69–77. in an RBC antibody panel, but 7 of these contained the same 15. Aster RH, Bougie DW. Drug-induced immune thrombo- DDPA that had been identified in the patient’s serum and re- cytopenia. N Engl J Med 2007;357:580–7. acted in a DDPA SPRCA. 16. Asvadi P, Ahmadi Z, Chong BH. Drug-induced thrombo- In conclusion, the very real possibility that a drug is caus- cytopenia: localization of the binding site of GPIX-specific ing a patient’s thrombocytopenia deserves investigation and, quinine-dependent antibodies. Blood 2003;102:1670–7. if substantiated, concerted action. Although not all situations 17. Aster RH. Immune thrombocytopenia caused by glyco- will be able to be resolved through laboratory testing, there protein IIb/IIIa inhibitors. Chest 2005;127(2Suppl):53S– is clearly assistance the laboratory can provide to help guide 9S. possible changes in the patient’s drug regimen. 18. Honda S, Tomiyama Y, Aoki T, et al. Association between ligand-induced conformational changes of integrin IIb- References beta3 and IIbbeta3-mediated intracellular Ca2+ signal- 1. van den Bemt PM, Meyboom RH, Egberts AC. Drug- ing. Blood 1998;92:3675–83. induced immune thrombocytopenia. Drug Safety 19. Bougie DW, Wilker PR, Wuitschick ED, et al. Acute 2004;27:1243–52. thrombocytopenia after treatment with tirofiban or epti- 2. Kaufman DW, Kelly JP, Johannes CB, et al. Acute throm- fibatide is associated with antibodies specific for ligand- bocytopenic purpura in relation to the use of drugs. occupied GPIIb/IIIa. Blood 2002;100:2071–6. Blood 1993;82:2714–18. 20. Curtis BR, Swyers J, Divgi A, McFarland JG, Aster RH. 3. Visentin GP, Liu CY. Drug-induced thrombocytopenia. Thrombocytopenia after second exposure to abciximab Hematol Oncol Clin N Am 2007;21:685–96,vi. is caused by antibodies that recognize abciximab-coated 4. Bougie DW, Wilker PR, Aster RH. Patients with qui- platelets. Blood 2002;99:2054–9. nine-induced immune thrombocytopenia have both 21. Von Drygalski A, Curtis BR, Bougie DW, et al. Vancomycin- “drug-dependent” and “drug-specific” antibodies. Blood induced immune thrombocytopenia. N Engl J Med 2006;108:922–7. 2007;356:904–10. 5. Leach MF, AuBuchon JP. Coincidence of gentamycin in- 22. Grossjohann B, Eichler P, Greinacher A, Santoso S, Kroll duced drug-dependent platelet antibodies. Transfusion H. Ceftriaxone causes drug-induced immune throm- 1998;38:29S. bocytopenia and hemolytic anemia: characterization 6. Warkentin TE. Heparin-induced thrombocytopenia: di- of targets on platelets and red blood cells. Transfusion agnosis and management. Circulation 2004;110:e454–8. 2004;44:1033–40.

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 139 J.P. AuBuchon and M.P. Leach

23. Leach MF, Cooper LK, AuBuchon JP. Detection of drug- 28. Leach MF, AuBuchon JP, Szczepiorkowski ZM. The independent, platelet-reactive antibodies by solid-phase cidence of positive direct antiglobulin tests and red cell red cell adherence assays. Br J Haematol 1997;97:755– drug antibodies in patients with drug-induced thrombo- 61. cytopenia. Transfusion 2005;45:134A. 24. Leach MF, Cooper LK, AuBuchon JP. Detection of drug- dependent antibodies by use of solid-phase red cell ad- James P. AuBuchon, MD (corresponding author), President herence techniques. Immunohematology 1995;11:143– and CEO, Puget Sound Blood Center, Professor of Medicine 9. and of Laboratory Medicine, University of Washington, 921 25. Visentin GP, Ford SE, Scott JP, Aster RH. Antibodies Terry Avenue, Seattle, WA, 98104, and Miriam F. Leach, from patients with heparin-induced thrombocytopenia/ MSAI, MSHIM, BSIS, MT(ASCP)SBB, QLI, LIS Specialist thrombosis are specific for platelet factor 4 complexed and Special Projects Coordinator, Blood Bank and Trans- with heparin or bound to endothelial cells. J Clin Invest fusion Medicine Service, Department of Pathology, Dart- 1994;93:81–8. mouth-Hitchcock Medical Center, Lebanon, NH. 26. Greinacher A, Kohlmann T, Strobel U, Sheppard JA, Warkentin TE. The temporal profile of the anti-PF4/ heparin immune response. Blood 2009;113:4970–6. 27. Christie DJ, Weber RW, Mullen PC, Cook JM, Aster RH. Structural features of the quinidine and quinine molecules necessary for binding of drug-induced antibodies to human platelets. J Lab Clin Med 1984;104:730–40.

Manuscripts The editorial staff of Immunohematology welcomes manuscripts pertaining to blood group serology and education for consideration for publication. We are especially interested in case reports, papers on platelet and white cell serology, scientific articles covering original investigations, and papers on new methods for use in the blood bank. Deadlines for receipt of manuscripts for consideration for the March, June, September, and December issues are the first weeks in November, February, May, and August, respectively. For instructions for scientific articles, case reports and review articles, see Instructions for Authors in every issue of Immunohematology or on the Web at www.redcross.org/immunohematology. Include fax and phone numbers and e-mail address with all articles and correspondence. E-mail all manuscripts to [email protected].

For information concerning For information concerning the National Immunohematology, Journal of Blood Reference Laboratory for Blood Group Group Serology and Education, or Serology, including the American Rare the Immunohematology Methods and Donor Program, please contact Sandra Procedures manual, contact us by e-mail at Nance, by phone at (215) 451-4362, by fax at [email protected] (215) 451-2538, or by e-mail at snance@usa. redcross.org

140 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Announcements

Masters (MSc) in Transfusion and Transplantation Sciences at The University of Bristol, England

Applications are invited from medical or science graduates for the Master of Science (MSc) degree in Transfusion and Transplantation Sciences at the University of Bristol. The course starts in October 2010 and will last for 1 year. A part-time option lasting 2 or 3 years is also available. There may also be opportunities to continue studies for PhD or MD following the MSc. The syllabus is organized jointly by The Bristol Institute for Transfusion Sciences and the University of Bristol, Department of Pathology and Microbiology. It includes:

• Scientific principles of transfusion and transplantation • Clinical applications of these principles • Practical techniques in transfusion and transplantation • Principles of study design and biostatistics • An original research project

Application can also be made for Diploma in Transfusion and Transplantation Science or a Certificate in Transfusion and Transplantation Science.

The course is accredited by the Institute of Biomedical Sciences.

Further information can be obtained from the Web site: http://www.blood.co.uk/ibgrl/MscHome.htm

For further details and application forms please contact:

Dr Patricia Denning-Kendall University of Bristol, Paul O’Gorman Lifeline Centre, Department of Pathology and Microbiology, Southmead Hospital, Westbury-on-Trym, Bristol BS10 5NB, England Fax +44 1179 595 342, Telephone +44 1779 595 455, e-mail: p.a.denning-kendall@bristol. ac.uk.

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 141 Announcements, cont., and advertisements

Monoclonal antibodies available at no charge The New York Blood Center has developed a wide range of monoclonal antibodies (both murine and humanized) that are useful for donor screening and for typing RBCs with a positive DAT. These include anti-A1, -M, -s, -U, -D, -Rh17, -K, -k, -Kpa, -Jsb, -Fya, -Fy3, -Fy6, Wrb, -Xga, -CD99, -Dob, -H, -Ge2, -Ge3, -CD55 (both SCR2/3 and SCR4), -Oka, -I, and anti- CD59. Most of the antibodies are murine IgG and require the use of anti-mouse IgG for detection (anti-K, -k, and -Kpa). Some are directly agglutinating (anti-A1, -M, -Wrb and -Rh17) and a few have been humanized into the IgM isoform (anti- Jsb). The antibodies are available at no charge to anyone who requests them. Please visit our Web site for a complete list of available monoclonal antibodies and the procedure for obtaining them.

For additional information, contact: Gregory Halverson, New York Blood Center, 310 East 67th Street, New York, NY 10021; e-mail: [email protected]; phone: (212) 570-3026; fax: (212) 737-4935; or visit the web site at www. nybloodcenter.org >research >immunochemistry >current list of monoclonal antibodies available.

Specialist in Blood Bank (SBB) Program Feb 20–21, 2010, SBB “Last Chance” Review. Gulf Coast Regional Blood Center, Houston, TX. This program is designed for individuals preparing to take the ASCP SBB or BB Registry examination. Physicians who are preparing for the Board examination in Blood Banking and individuals wishing a refresher in blood banking can also benefit from this program. Contact Clare Wong, (713) 791-6201, [email protected]. Details: www.giveblood.org/education/lastchance.htm.

National Platelet Serology Reference Laboratory National Neutrophil Serology Reference Laboratory Diagnostic testing for: Our laboratory specializes in granulocyte antibody detection • Neonatal alloimmune thrombocytopenia (NAIT) and granulocyte antigen typing. • Posttransfusion purpura (PTP) Indications for granulocyte serology testing include: • Refractoriness to platelet transfusion • Heparin-induced thrombocytopenia (HIT) • Alloimmune neonatal neutropenia (ANN) • Alloimmune idiopathic thrombocytopenia purpura (AITP) • Autoimmune neutropenia (AIN) • Transfusion related acute lung injury (TRALI) Medical consultation available Methodologies employed: Test methods: • Granulocyte agglutination (GA) • GTI systems tests • Granulocyte immunofluorescence by flow cytometry (GIF) — detection of glycoprotein-specific platelet antibodies • Monoclonal antibody immobilization of neutrophil antigens — detection of heparin-induced antibodies (PF4 ELISA) (MAINA) • Platelet suspension immunofluorescence test (PSIFT) TRALI investigations also include: • Solid phase red cell adherence (SPRCA) assay • Monoclonal immobilization of platelet antigens (MAIPA) • HLA (PRA) Class I and Class II antibody detection • Molecular analysis for HPA-1a/1b For further information contact: For further information, contact Neutrophil Serology Laboratory (651) 291-6797 Platelet Serology Laboratory (215) 451-4205 Randy Schuller(651) 291-6758 [email protected] Maryann Keashen-Schnell (215) 451-4041 office [email protected] Sandra Nance (215) 451-4362 American Red Cross Blood Services [email protected] Neutrophil Serology Laboratory American Red Cross Blood Services 100 South Robert Street Musser Blood Center St. Paul, MN 55107 700 Spring Garden Street Philadelphia, PA 19123-3594

CLIA licensed CLIA licensed

142 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Advertisements, cont.

Reference and Consultation Services IgA/Anti-IgA Testing

Antibody identification and problem resolution IgA and anti-IgA testing is available to do the HLA-A, B, C, and DR typing following: HLA-disease association typing • Identify IgA-deficient patients Paternity testing/DNA • Investigate anaphylactic reactions • Confirm IgA-deficient donors

For information, contact Our ELISA for IgA detects protein to 0.05 mg/dL. Mehdizadeh Kashi at (503) 280-0210, or write to: For additional information contact Cindy Flickinger at Pacific Northwest Regional Blood Services (215) 451-4909, or e-mail: [email protected], ATTENTION: Tissue Typing Laboratory or write to: American Red Cross 3131 North Vancouver American Red Cross Blood Services Musser Blood Center Portland, OR 97227 700 Spring Garden Street Philadelphia, PA 19123-3594 CLIA licensed, ASHI accredited ATTN: Cindy Flickinger

National Reference Laboratory for Blood Group Serology

Immunohematology Reference Laboratory CLIA licensed

AABB, ARC, New York State, and CLIA licensed Donor IgA Screening (215) 451-4901— 24-hr. phone number (215) 451-2538— Fax • Effective tool for screening large volumes of American Rare Donor Program donors • Gel diffusion test that has a 15-year proven track (215) 451-4900— 24-hr. phone number record: (215) 451-2538— Fax Approximately 90 percent of all donors [email protected] identified as IgA deficient by are confirmed by the more sensitive testing methods Immunohematology Journal of Blood Group Serology and Education For additional information, call Kathy Kaherl at: (860)678-2764, e-mail: [email protected] (215) 451-4902— Phone, business hours (215) 451-2538— Fax or write to: [email protected] Reference Laboratory Quality Control of Cryoprecipitated-AHF American Red CrossBiomedical Services Connecticut Region (215) 451-4903— Phone, business hours 209 Farmington Ave. (215) 451-2538— Fax Farmington, CT 06032

CLIA licensed

IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 143 Advertisements, cont.

Blood Group Antigens & Antibodies A guide to clinical relevance & technical tips

by Marion E. Reid & Christine Lomas-Francis This compact “pocketbook” from the authors of the Blood Group Antigen FactsBook is a must for anyone who is involved in the laboratory or bedside care of patients with blood group alloantibodies. The book contains clinical and technical information about the nearly 300 ISBT recognized blood group antigens and their corresponding antibodies. The information is listed in alphabetical order for ease of finding—even in the middle of the night. Included in the book is information relating to: • Clinical significance of antibodies in transfusion and HDN. • Number of compatible donors that would be expected Ordering Information to be found in testing 100 donors. Variations in different The book, which costs $25, can be ethnic groups are given. ordered in two ways: • Characteristics of the antibodies and optimal technique(s) • Order online from the publisher for their detection. at: www.sbbpocketbook.com • Technical tips to aid their identification. • Order from the authors, who • Whether the antibody has been found as an autoantibody. will sign the book. Send a check, made payable to “New York Blood Center” and indicate “Pocket Pocketbook Education Fund book” on the memo line, to: The authors are using royalties generated from the sale of this pocketbook for educational purposes to mentor people Marion Reid in the joys of immunohematology as a career. They will Laboratory of Immunochemistry accomplish this in the following ways: New York Blood Center • Sponsor workshops, seminars, and lectures 310 East 67th Street • Sponsor students to attend a meeting New York, NY 10065 • Provide copies of the pocketbook Please include the recipient’s (See www.sbbpocketbook.com for details to apply for funds) complete mailing address.

144 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 Advertisements, cont. Becoming a Specialist in Blood Banking (SBB)

What is a certified Specialist in Blood Banking (SBB)? • Someone with educational and work experience qualifications who successfully passes the American Society for Clinical Pathology (ASCP) board of registry (BOR) examination for the Specialist in Blood Banking. • This person will have advanced knowledge, skills, and abilities in the field of transfusion medicine and blood banking.

Individuals who have an SBB certification serve in many areas of transfusion medicine: • Serve as regulatory, technical, procedural and research advisors • Perform and direct administrative functions • Develop, validate, implement, and perform laboratory procedures • Analyze quality issues preparing and implementing corrective actions to prevent and document issues • Design and present educational programs • Provide technical and scientific training in blood transfusion medicine • Conduct research in transfusion medicine

Who are SBBs? Supervisors of Transfusion Services Managers of Blood Centers LIS Coordinators Educators Supervisors of Reference Laboratories Research Scientists Consumer Safety Officers Quality Assurance Officers Technical Representatives Reference Lab Specialist

Why be an SBB? Professional growth Job placement Job satisfaction Career advancement

How does one become an SBB? • Attend a CAAHEP-accredited Specialist in Blood Bank Technology Program OR • Sit for the examination based on criteria established by ASCP for education and experience

However: In recent years, a greater percentage of individuals who graduate from CAAHEP-accredited programs pass the SBB exam.

Conclusion: The BEST route for obtaining an SBB certification is … to attend a CAAHEP-accredited Specialist in Blood Bank Technology Program

Contact the following programs for more information:

Program Contact Name Phone Contact Email Contact Website On site or On line Program Walter Reed Army Medical Center William Turcan 202-782-6210 [email protected]. www.militaryblood.dod.mil On site MIL American Red Cross, Southern California Region Michael Coover 909-859-7496 [email protected] none On site

ARC-Central OH Region Joanne Kosanke 614-253-2740 x 2270 [email protected] none On site Blood Center of Southeastern Wisconsin Lynne LeMense 414-937-6403 [email protected] www.bcw.edu On site Community Blood Center/CTS Dayton, Ohio Nancy Lang 937-461-3293 [email protected] http://www.cbccts.org/education/ On line sbb.htm Gulf Coast School of Blood Bank Technology Clare Wong 713-791-6201 [email protected] www.giveblood.org/education/ On line distance/htm Hoxworth Blood Center, Univ. of Cincinnati Susan Wilkinson 513-558-1275 [email protected] www.hoxworth.org On site Indiana Blood Center Jayanna Slayten 317-916-5186 [email protected] www.indianablood.org On line

Johns Hopkins Hospital Jan Light 410-955-6580 [email protected] http://pathology2.jhu/ On site department/divisions/trnafusion/ sbb.cfm Medical Center of Louisiana Karen Kirkley 504-903-3954 [email protected] none On site NIH Clinical Center Dept.. of Transfusion Medicine Karen Byrne 301-496-8335 [email protected] www.cc.nih.gov/dtm On site Rush University Veronica Lewis 312-942-2402 [email protected] www.rushu.rush.edu/health/dept. On line html Transfusion Medicine Center at Florida Blood Marjorie Doty 727-568-5433 x 1514 [email protected] www.fbsblood.org On line Services Univ. of Texas Health Science Center at San Antonio Linda Myers 210-731-5526 [email protected] www.uthscsa.edu On site Univ. of Texas Medical Branch at Galveston Janet Vincent 409-772-3055 [email protected] www.utmb.edu/sbb On line

Univ. of Texas SW Medical Center Barbara Laird- 214-648-1785 [email protected] http://telecampus.utsystem.edu On line Fryer Additional Information can be found by visiting the following Web sites: www.ascp.org, www.caahep.org and www.aabb.org Revised August 2007 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009 145 ImmunohematologyImmunohematology JournalJOURNAL ofOF BloodBLOOD GroupGROUP SEROLOGY Serology AND and EDUCATION Education Instructions to the Authors

I. GENERAL INSTRUCTIONS b. Use short headings for each column needed and capitalize first letter Before submitting a manuscript, consult current issues of of first word. Omit vertical lines. Immunohematology for style. Double-space throughout the manuscript. c. Place explanation in footnotes (sequence: *, †, ‡, §, ¶, **, ††). Number the pages consecutively in the upper right-hand corner, beginning 8. Figures with the title page. a. Figures can be submitted either by e-mail or as photographs (5″×7″ II. SCIENTIFIC ARTICLE, REVIEW, OR CASE REPORT WITH glossy). LITERATURE REVIEW b. Place caption for a figure on a separate page (e.g. Fig. 1 Results of...), A. Each component of the manuscript must start on a new page in the ending with a period. If figure is submitted as a glossy, place first following order: author’s name and figure number on back of each glossy submitted. 1. Title page c. When plotting points on a figure, use the following symbols if 2. Abstract possible: � � � � � �. 3. Text 9. Author information 4. Acknowledgments a. List first name, middle initial, last name, highest degree, position held, 5. References institution and department, and complete address (including ZIP 6. Author information code) for all authors. List country when applicable. 7. Tables 8. Figures III. EDUCATIONAL FORUM B. Preparation of manuscript A. All submitted manuscripts should be approximately 2000 to 2500 1. Title page words with pertinent references. Submissions may include: a. Full title of manuscript with only first letter of first word capitalized 1. An immunohematologic case that illustrates a sound investigative (bold title) approach with clinical correlation, reflecting appropriate collaboration b. Initials and last name of each author (no degrees; all CAPS), e.g., M.T. JONES, J.H. BROWN, AND S.R. SMITH to sharpen problem solving skills c. Running title of ≤40 characters, including spaces 2. Annotated conference proceedings d. Three to ten key words B. Preparation of manuscript 2. Abstract 1. Title page a. One paragraph, no longer than 300 words a. Capitalize first word of title. b. Purpose, methods, findings, and conclusion of study b. Initials and last name of each author (no degrees; all CAPs) 3. Key words 2. Text a. List under abstract a. Case should be written as progressive disclosure and may include the 4. Text (serial pages): Most manuscripts can usually, but not necessarily, following headings, as appropriate be divided into sections (as described below). Survey results and i. Clinical Case Presentation: Clinical information and differential review papers may need individualized sections diagnosis a. Introduction ii. Immunohematologic Evaluation and Results: Serology and Purpose and rationale for study, including pertinent background references molecular testing b. Case Report (if indicated by study) iii. Interpretation: Include interpretation of laboratory results, Clinical and/or hematologic data and background serology/molecular correlating with clinical findings c. Materials and Methods iv. Recommended Therapy: Include both transfusion and Selection and number of subjects, samples, items, etc. studied and nontransfusion-based therapies description of appropriate controls, procedures, methods, equipment, v. Discussion: Brief review of literature with unique features of this reagents, etc. Equipment and reagents should be identified in case parentheses by model or lot and manufacturer’s name, city, and state. vi. Reference: Limited to those directly pertinent Do not use patient’s names or hospital numbers. vii. Author information (see II.B.9.) d. Results viii. Tables (see II.B.7.) Presentation of concise and sequential results, referring to pertinent tables and/or figures, if applicable IV. LETTER TO THE EDITOR e. Discussion A. Preparation Implication and limitations of the study, links to other studies; if 1. Heading (To the Editor) appropriate, link conclusions to purpose of study as stated in 2. Title (first word capitalized) introduction 3. Text (written in letter [paragraph] format) 5. Acknowledgments: Acknowledge those who have made substantial contributions to the study, including secretarial assistance; list any grants. 4. Author(s) (type flush right; for first author: name, degree, institution, 6. References address [including city, state, Zip code and country]; for other authors: a. In text, use superscript, Arabic numbers. name, degree, institution, city and state) b. Number references consecutively in the order they occur in the text. 5. References (limited to ten) 7. Tables 6. Table or figure (limited to one) a. Head each with a brief title; capitalize the first letter of first word (e.g., Table 1. Results of . . .) use no punctuation at the end of the title. Send all manuscripts by e-mail to [email protected]

IMMUNOHEMATOLOGY, VOLUME 22, NUMBER 4, 2006 215 146 IMMUNOHEMATOLOGY, Volume 25, Number 3, 2009

FIRST CLASS U.S. POSTAGE PAID SOUTHERN MD Musser Blood Center PERMIT NO. 4144 700 Spring Garden Street Philadelphia, PA 19123-3594