Immunohematology - American Red Cross...mefenamic acid mefloquine melphalan mephenytoin methoin...

ImmunohematologyJ O U R N A L O F B L O O D G R O U P S E R O L O G Y A N D E D U C A T I O N

V O L U M E 2 0 , N U M B E R 3 , 2 0 0 4

This Issue ofImmunohematology

Is Supported bya Contribution From

Dedicated to Education in the Field of

Blood Banking

ImmunohematologyJ O U R N A L O F B L O O D G R O U P S E R O L O G Y A N D E D U C A T I O N

V O L U M E 2 0 , N U M B E R 3 , 2 0 0 4

C O N T E N T S

137Letter to the readers

Introduction to the review articlesS.T. NANCE

138Review: drug-induced immune hemolytic anemia—the last decade

G.GARRATTY

147Review: what to do when all RBCs are incompatible—serologic aspects

S.T. NANCE AND P.A.ARNDT

161Review: transfusing incompatible RBCs—clinical aspects

G. MENY

167Review: evaluation of patients with immune hemolysis

L.D. PETZ

177Case report: exacerbation of hemolytic anemia requiring multiple incompatible RBC transfusions

A.M. SVENSSON, S. BUSHOR,AND M.K. FUNG

184Delayed hemolytic transfusion reaction due to anti-Fyb caused by a primary immune response: a case study

and a review of the literatureH.H. KIM,T.S. PARK, S.H. OH, C.L. CHANG, E.Y. LEE,AND H.C. SON

187Maternal alloanti-hrS—an absence of HDN

R. KAKAIYA, J. CSERI, B. JOCHUM, L. GILLARD,AND S. SILBERMAN

190 193C O M M U N I C A T I O N S Letter to the Editor-in-Chief

Letter to the Editors Immunohematology to be listed in Index MedicusHAMA (Human Anti-Mouse Antibodies) do not and MEDLINE

Cause False Positive Results in PAKPLUS S.G. SANDLER

L.A.TIDEY, S. CHANCE, M. CLARKE,AND R.H.ASTER

Reply to letterM.F. LEACH AND J.P.AUBUCHON

195 196Letters From the Editor-in-Chief S P E C I A L S E C T I O N

Ortho dedication Exerpts from the American Red Cross ReferenceThe final 20th anniversary issue Laboratory Newsletter—1976

198 199 202A N N O U N C E M E N T S A D V E R T I S E M E N T S I N S T R U C T I O N S F O R A U T H O R S

EDITOR-IN-CHIEF MANAGING EDITORDelores Mallory, MT(ASCP)SBB Mary H. McGinniss,AB, (ASCP)SBB

Supply, North Carolina Bethesda, Maryland

TECHNICAL EDITOR SENIOR MEDICAL EDITORChristine Lomas-Francis, MSc Scott Murphy, MD

New York, New York Philadelphia, Pennsylvania

GUEST EDITORSandra T. Nance, MS, MT(ASCP)SBB

Philadelphia, Pennsylvania

ASSOCIATE MEDICAL EDITORSDavid Moolton, MD S. Gerald Sandler, MD Geralyn Meny, MD Ralph Vassallo, MD

Philadelphia, Pennsylvania Washington, District of Columbia Philadelphia, Pennsylvania Philadelphia, Pennsylvania

EDITORIAL BOARD

EDITORIAL ASSISTANT PRODUCTION ASSISTANTLinda Berenato Marge Manigly

COPY EDITOR ELECTRONIC PUBLISHER PROOFREADERLucy Oppenheim Paul Duquette George Aydinian

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

The contents are cited in the EBASE/Excerpta Medica and Elsevier BIOBASE/Current Awareness in Biological Sciences (CABS) databases.

The subscription price is $30.00 (U.S.) and $35.00 (foreign) per year.

Subscriptions, Change of Address, and Extra Copies:Immunohematology, P.O. Box 40325

Philadelphia, PA 19106Or call (215) 451-4902

Web site: www.redcross.org/pubs/immuno

Copyright 2004 by The American National Red CrossISSN 0894-203X

Patricia Arndt, MT(ASCP)SBBLos Angeles, California

James P.AuBuchon, MDLebanon, New Hampshire

Geoffrey Daniels, PhDBristol, United Kingdom

Richard Davey, MDNew York, New York

Sandra Ellisor, MS, MT(ASCP)SBBAnaheim, California

George Garratty, PhD, FRCPathLos Angeles, California

Brenda J. Grossman, MDSt. Louis, Missouri

W. John Judd, FIBMS, MIBiolAnn Arbor, Michigan

Christine Lomas-Francis, MScNew York, New York

Gary Moroff, PhDRockville, Maryland

Ruth Mougey, MT(ASCP)SBBCarrollton, Kentucky

John J. Moulds, MT(ASCP)SBBRaritan, New Jersey

Marilyn K. Moulds, MT(ASCP)SBBHouston, Texas

Sandra Nance, MS, MT(ASCP)SBBPhiladelphia, Pennsylvania

Paul M. Ness, MDBaltimore, Maryland

Mark Popovsky, MDBraintree, Massachusetts

Marion E. Reid, PhD, FIBMSNew York, New York

Susan Rolih, MS, MT(ASCP)SBBCincinnati, Ohio

David F. Stroncek, MDBethesda, Maryland

Marilyn J.Telen, MDDurham, North Carolina

I M M U N O H E M A T O L O G Y, V O L U M E 2 0 , N U M B E R 3 , 2 0 0 4 137

Introduction to the review articlesWelcome to the third issue of 2004, celebrating

Immunohematology’s 20 years of publication. As withthe first two issues of 2004, it contains four invitedreview articles. This issue’s focus is on diagnostic usesof immunohematology serologic testing.

The first review is a recurring favorite, “Drug-Induced Immune Hemolytic Anemia—The LastDecade,” by George Garratty, PhD. Over the years, theeditors of Immunohematology have invited Dr.Garratty to write review articles on drug-inducedhemolytic anemia and he has kept our readers current.This review continues the tradition with anotherexcellent review.

The second and third reviews focus on what to dowhen all units of blood are incompatible. Serologicaspects are covered by Sandra Nance and PatriciaArndt; these aspects include the testing oftenperformed in the immunohematology referencelaboratory (IRL) and monocyte monolayer assaysperformed as a specialty test in two IRLs in the UnitedStates. This second review also includes cases thatdemonstrate the points under discussion. Dr. GeralynMeny reviews clinical responses when all units ofblood are incompatible and reminds us of all theelements that should be considered when this situationarises in your facility.

Dr. Larry Petz completes the invited review sectionwith “Evaluation of Patients With Immune Hemolysis.”This review takes the reader through the differentialdiagnoses of a patient that presents with immunehemolysis and discusses the importance of the precisediagnosis in prognosis and therapy.

This third issue is a good mix of reviews thatdiscuss serologic testing and clinical use of results fordiagnostic and therapeutic measures. These reviewsare written by authors with a great deal of experiencein their field. I sincerely hope you enjoy the issue asmuch as I have enjoyed being the guest editor.

Sandra T. Nance, MS, MT(ASCP)SBBMember of the Editorial Board and

Guest Editor of this issue,American Red Cross Blood Services

Penn-Jersey Region700 Spring Garden StreetPhiladelphia, PA 19123

L E T T E R T O T H E R E A D E R S

138 I M M U N O H E M A T O L O G Y, V O L U M E 2 0 , N U M B E R 3 , 2 0 0 4

Review: drug-induced immunehemolytic anemia—the lastdecadeG. GARRATTY

I have written three previous reviews on drug-induced immune hemolytic anemia (DIIHA) for thisjournal.1–3 The last one was written in 1994.3 This year,I would like to review what has happened in the lastdecade.

When Dr. Petz and I published the first edition ofour book (Acquired Immune Hemolytic Anemias) in1980,4 we found that there was reasonable evidence tosupport that about 30 drugs could cause DIIHA. Onedrug,methyldopa,was by far the most common drug todo this. Almost 70 percent of DIIHAs referred to ourlaboratory in the 1970s were associated with methyl-dopa. Antibodies associated with methyldopa weredrug-independent and the patients showed all theserologic and hematologic characteristics of “warmtype” autoimmune hemolytic anemia (WAIHA).Twenty-three percent of the DIIHAs were associatedwith high-dose IV penicillin therapy; only about 10percent of DIIHAs were associated with other drugs(e.g., quinine, rifampicin, and hydrochlorothiazide).

In the next 20 years, methyldopa and high dose IVpenicillin were used less and less; we have not seen acase of DIIHA associated with these two drugs formany years. By 1994 about 71 drugs had beenimplicated in DIIHA.5 Recently published resultsreflect a changing picture in the spectrum of DIIHA inthe last 25 years.6,7 There are now approximately 100drugs associated with DIIHA (see Table 1), andmethyldopa and penicillin have been replaced by asingle group of drugs, the cephalosporins (93% ofcases), with cefotetan alone accounting for 83 percentof the DIIHAs we have encountered in the last 10years. Table 2 shows the drugs causing DIIHA that wehave encountered in the past 26 years (1978–2003).Methyldopa is probably underrepresented as, by 1978,cases of autoimmune hemolytic anemia (AIHA) inpatients taking methyldopa were not usually sent tospecialist laboratories such as ours for investigation.

Because of these statistics, I will be emphasizingthe cephalosporins in this review.

Cephalosporin-Induced Immune HemolyticAnemia

There are about 70 individual published casereports of cephalosporin-induced immune hemolyticanemia (CIIHA),8–64 but many more are contained inreviews or tables without case histories65–67 (see Table3). Most patients have had severe hemolytic anemia(HA), often with intravascular lysis, and 40 percentwere associated with fatal HA. It is not known if this isthe tip of the iceberg and there are many more cases of milder HA or positive DATs that are not reported;the same questions apply to cephalosporin-inducedthrombocytopenia. Tables 4 and 5 summarize theclinical and serologic findings associated withcefotetan- and ceftriaxone-induced immune HA. Itshould be emphasized that cefotetan antibodies alwaysreact with cefotetan-coated RBCs and almost alwaysreact with untreated RBCs in the presence of cefotetan(“immune complex” method), and about one-third willreact with RBCs without the presence of drug (i.e.,willappear to be autoantibodies). The latter findings canlead to problems in the blood transfusion service. If apatient receives cefotetan prophylactically for surgery,receives a blood transfusion during or after surgery,andthen develops HA 7 to 10 days afterwards, a delayedhemolytic transfusion reaction is often suspected. Thehematologic findings can also mimic AIHA. If the HA isdue to cefotetan, the DAT will be positive (although wehave reported one case where the DAT was negative).68

Sometimes the serum will react with all untreatedRBCs, mimicking an alloantibody to a high-frequencyantigen, or a mixture of alloantibodies or autoantibody,and many hours may be wasted investigating thesepossibilities. If there is a history of cefotetan


administration, I recommend testing an eluate from thepatient’s RBCs against untreated and cefotetan-coatedRBCs to help with the differential diagnosis. If theeluate is reactive with cefotetan-treated RBCs and notthe same untreated RBCs, the diagnosis is obvious.Unfortunately, in about 15 percent of cases we havestudied, the eluate also reacted with untreated RBCs

Review: drug-induced hemolytic anemia

Table 1. Drugs causing immune hemolytic anemia and/or positive DATs*

aceclofenacacetaminophen/paracetamolaminopyrine/pyramidonamphotericin Bampicillinantazoline apazone/azapropazoneapronalidebutizide carbenicillincarbimazole carboplatincarbromalcatergen/cyanidanolcefamandolecefazolincefiximecefotaximecefotetancefoxitinceftazidimeceftizoximeceftriaxonecephalexincephalordinecephalothinchaparralchlordiazepoxide chlorinated hydrocarbonschlorpromazinechlorpropamidecianidanolcisplatincladribineclavulanate potassium (e.g.,Timentin)cyclofenil diclofenac

diethylstilbestrol diglycoaldehyde (INOX)diphenylhydantoindipyroneerythromycinetodolacfenfluramine fenoprofenfludarabinefluorescein fluoroquinolonefluorouracil (5-FU)furosemideglafeninehydralazinehydrochlorothiazide9-hydroxy-methyl-ellipticiniumibuprofenindene derivatives (e.g., sulindac)insulininterferon interleukin-2isoniazidlatamoxeflevodopa mefenamic acidmefloquinemelphalanmephenytoin methoin6-mercaptopurinemethadonemethicillinmethotrexate methyldopamethysergidenafcillin

nomifensineoxaliplatinp-aminosalicylic acidpenicillin Gphenacetinphenytoinpiperacillinprobenecidprocainamidepropyphenazonequinidine quinineranitidine rifampicinrituximabsodium pentothal/thiopental stibophenstreptomycinsulbactam sodium (e.g., in Unasyn)sulindacsulfonamidessulfonylurea derivatives (e.g., chlorpropamide)suprofensuramintazobactam (e.g., in Zosyn)teicoplanintemafloxacinteniposidetetracyclineticarcillintolbutamide tolmetintriamterenetrimellitic anhydridezomepirac

*Drugs in publications containing reasonable evidence for an immune etiology were the only drugs included (many more are in the literature).

Table 2. Drugs associated with DIIHA investigated at American RedCross Blood Services, Los Angeles, in the last 26 yrs.(1978–2003)

Drug Number of patients

cefotetan 74ceftriaxone 12piperacillin 5tazobactam 5clavulanate 3fludarabine 2penicillin 2probenecid 2rifampicin 2cefotaxime 1sulbactam 1ticarcillin 1mefloquine 1cefoxitin 1chlorpropamide 1nafcillin 1phenacetin 1procainamide 1erythromycin 1tolmetin 1oxaliplatin 1

Total 119

Table 3. Cephalosporins reported to cause immune hemolytic anemia(up to 2003)

Drug Number of case reports References

cephalothin 5 8–11cefazolin 1 11cephalexin 2 12, 13cefamandole 1 14cefoxitin 2 15, 16cefotaxime 3 17–19ceftriaxone 19 20–37cefotetan 35* 37–59ceftizoxime 4 60–62cefixime 1 19ceftazidime 2 63, 64

Total 75*Many other cases are reported, without individual case reports, in references 65 (43

cases) and 66 (85 cases)


(Arndt, Leger, Garratty, unpublished observations). Thediagnosis may then rest on the presence of a high-titer(> 100) cefotetan antibody in the patient’s serum.Some illustrative case histories associated with suchproblems have been published in this journal.36

Technical hints when investigating CIIHA1. When preparing cefotetan-coated RBCs, use a

buffer with a pH of 6 to 7 (or normal saline) insteadof the high pH buffer (pH 8 to 10) used forpreparation of penicillin-coated RBCs.69 The lower

pH does not reduce sensitivity but does reducenonspecific uptake of protein leading to falselypositive antihuman globulin (AHG) tests. Someinvestigators have suggested that drugs can besolubilized more efficiently in 1% albumin.70 We donot advise this when testing for cefotetanantibodies, as albumin can cause reduced bindingof drug to drug-treated RBCs.71

2. Most individuals, including healthy donors, haveantibodies in their sera that can directly agglutinatecefotetan-coated RBCs.72 In addition,almost all sera(including those from healthy individuals) willyield positive IATs with cefotetan-treated RBCs.When testing a patient’s serum with cefotetan-coated RBCs, dilute the serum 1 in 100 in saline.36

This will avoid agglutination and nonspecificprotein uptake onto cefotetan-treated RBCs;clinically significant antibodies always have titers > 100 (e.g., in the thousands). Some texts stillrecommend a dilution of 1 in 20, which is what werecommended when using cephalothin-coatedRBCs. We have found that 1 in 20 is sufficient toexclude nonspecific adsorption of proteins bycefotetan-treated RBCs, but is not sufficient toexclude agglutination of cefotetan-treated RBCs byabout 10 percent of sera (even from healthydonors).72

3. The lesson to be learned by the above is: neverreport cefotetan antibodies to be present based onreactions of undiluted sera. As with penicillin-induced immune HA,the diagnosis should be basedon an eluate from the patient’s DAT-positive RBCsreacting with cefotetan-coated RBCs, but notuntreated RBCs; unlike penicillin, the cefotetanantibody in the eluate may sometimes react withuntreated RBCs (see reference 65). Anotherproblem is that the last wash of the patient’s RBCsmay be reactive.73 The presence of a high titer (>100) cefotetan antibody in the serum confirms thediagnosis.

4. Ceftriaxone-coated RBCs cannot be made, thus theserologic diagnosis is based on the results of the“immune-complex” method. Nonspecific ad-sorption of proteins is not a problem by thismethod, thus undiluted sera can be used.Sometimes, positive reactions are only obtained byusing enzyme-treated RBCs, or metabolites of thedrug (e.g., in the presence of urine from a patienttaking ceftriaxone).29,32,35

G. GARRATTY

Table 4. Clinical* and serologic** findings associated with cefotetan-induced IHA

• Approx. 80 percent of patients received cefotetan for surgery; usually asingle dose of 2 g was used.

• A history of previous cefotetan therapy was not common.

• HA was obvious in less than 1 day to 13 days after receiving cefotetan.Only two patients had HA in < 1 day; the mean of the other 29 was 9days.

• Patients’ nadir Hb after receiving cefotetan = 2.6 g/dL (mean = 4.8 g/dL).

• Most patients had signs of intravascular lysis (hemoglobinemia/hemoglobinuria).

• Fatal HA and renal failure occurred in 19 percent of patients.

• Patients always had a positive DAT:– 100% had RBC-bound IgG– 86% had RBC-bound C3– 44% had RBC-bound IgA– 7.4% had RBC-bound IgM

• All sera agglutinated cefotetan-treated RBCs (median titer 512) andreacted by IAT (median titer 16,000); all normal sera also reacted withthe cefotetan-treated RBCs, but were nonreactive when diluted ≥ 1 in100.

• All but one of the sera reacted with untreated RBCs in the presence ofcefotetan (“immune complex” mechanism).

• 33 percent and 40 percent of sera reacted with RBCs without drugbeing present, with saline-suspended RBCs or in the presence of PEG,respectively.65

*Clincal data from references 38–40, 42–49, 50–53, 56, 57** Serologic data from reference 65

Table 5. Clinical and serologic findings associated with ceftriaxone-induced IHA

• HA is more acute and severe in children compared to adults. Inchildren, HA started 5 minutes to 7 days after receiving drug; for sixchildren there was a mean of 27 minutes and for three children a meanof 6 days.22,26–29,31,34,37 In adults, HA started 30 minutes to 34 days.One patient started after 30 minutes; the mean for nine others was 9 days.20,21,27,28,30,32,33,35,36

• Fatal HA occurred in 67 percent of 9 children and 30 percent of tenadults.

• There is a history of previous ceftriaxone therapy.

• Positive DATs associated with complement are present in all cases andadditional IgG is present in 75 percent; none had detectable RBC-boundIgA or IgM.65

• Antibodies are detected only by immune complex method65 (serum +drug + RBCs). Some patients’ sera only react with a metabolite ofceftriaxone (ex vivo antigen).29,32,35



DIIHA Associated With NonimmunologicAdsorption of Proteins Onto RBCs

RBCs treated with the first-generation cephalo-sporin, cephalothin, were found to adsorb manyproteins nonimmunologically when incubated in vitrowith normal plasma/sera.74 Such adsorbed proteins ledto a positive AHG test, but it was thought that this wasclinically insignificant, and indeed there were only fivecases of DIIHA due to cephalothin reported in morethan 30 years of use.8–11 Some other drugs have beenfound to show a similar phenomenon (see Table 6), butsome of these drugs have been thought to cause DIIHAmore commonly than cephalothin.

Causes for the uptake of protein are not alwaysclear. Garratty and Petz75 suggested that cephalothinchanged the RBC membrane in a way that created sitesfor protein binding. Later this was challenged byBranch and Petz,76 who suggested that no membranechange occurred and that protein attached to freeprotein binding sites on the cephalothin that hadcovalently bonded to proteins on the RBC membrane.Garratty and Leger77 were stimulated by an old finding,that cefotetan-treated RBCs were “PNH-like” (e.g., had apositive acidified serum test).78 To reexamine the issue,Garratty and Leger77 confirmed these old findings byshowing, using flow cytometry, that cephalothin- andcefotetan-treated RBCs had markedly decreasedamounts of CD55 and CD58 on the RBC membrane;CD59 was only slightly decreased. Thus, the resultswere different from the changes associated with PNHbut confirmed the original hypothesis of Garratty andPetz75 that cephalosporins can modify the RBCmembrane. Whether this is the cause of thenonimmunologic uptake of protein remains to beproved.

Other drugs have been found to cause non-immunologic adsorption of proteins onto RBCs (Table6). Some drugs containing aldehyde groups can bind toRBC proteins and adsorb proteins from the plasma,leading to positive AHG tests, but HA has not beenassociated with such drugs.79,80

Drugs belonging to the platinum family (e.g.,cisplatin, carboplatin, and oxaliplatin) have beenassociated with positive DATs and HA.81–92 Someinvestigators81 have suggested that cisplatin causesnonimmunologic adsorption of protein and that theHA is coincidental; others believe the drug can causeDIIHA. We believe that some patients can developantibodies to the drug and that the DIIHA may occurbecause of this, regardless of nonimmunologic proteinadsorption. We have described a patient who hadantibodies to oxaliplatin but no HA.92

Nevertheless, I now believe that nonimmunologicadsorption of protein may lead to HA. This is mainlybased on work we have performed on another familyof drugs, the beta lactamase inhibitors. Examples ofthese commonly used drugs are sulbactam (containedin Unasyn), clavulanate (contained in Augmentin andTimentin), and tazobactam (contained in Zosyn). Theyare used together with the beta lactam antibioticsampicillin, ticarcillin, and piperacillin, respectively.

Lutz and Dzik93 reported that 39 percent of patientsreceiving Unasyn developed weakly positive DATs butno HA. We showed later that RBCs treated in vitro withthe beta lactamase inhibitors contained in Unasyn,Augmentin,Timentin,and Zosyn would adsorb proteinsnonimmunologically, leading to positive IATs, and thatthis may be the cause of positive DATs associated withsuch drugs. It should be mentioned that patients maymake antibodies to the antibiotics also present in thedrug (e.g., ampicillin,piperacillin), and these antibodiescan cause positive AHG tests and HA. Garratty andArndt94,95 suggested that HA could also occur inpatients who have no antibodies to the antibiotics andthat this may be associated with nonimmunologicadsorption of protein. This suggestion was supportedby case histories and in vitro experiments where it wasshown that RBCs treated with the beta lactamaseinhibitors, washed, and incubated in normal plasma,washed, and then added to a monocyte monolayeryielded results suggesting that the RBCs would haveshortened survival.95 Broadberry et al.96 described aDIIHA associated with tazobactam; the HA was thoughtto be due to nonimmunologic adsorption of protein.

Unfortunately, we cannot prove in the laboratorythat this nonimmunologic adsorption of proteins iscausing the HA in a particular patient; one can onlysuggest it to the physicians and refer them to theappropriate literature (e.g., references 6 and 94–96).

Table 6. Drugs associated with nonimmunologic adsorption of proteinsonto RBCs

cephalosporinscisplatin/oxaliplatin/carboplatindiglycoaldehyde (INOX)suraminsulbactam (contained in Unasyn)clavulanate (contained in Augmentin and Timentin)tazobactam (contained in Zosyn)


G. GARRATTY

Mechanisms of DIIHAThere are two well-accepted mechanisms for

DIIHA, namely, that individuals can produce eitherdrug-dependent antibodies or drug-independentantibodies (Table 7). The latter can be classicautoantibodies in that they do not require drug to be

present for their reaction with RBCs in vitro or in vivo;drug is only needed to initiate the production ofantibodies. It is still unclear how this occurs, but themost popular concept is that certain drugs given tocertain patients somehow affect the immune system toproduce pathogenic autoantibodies, as do certaininfections in certain patients. The mechanisms mayturn out to be similar. The AIHA often continues forsome time after the drug is discontinued. Some drugantibodies (e.g., cefotetan) sometimes react with RBCswithout drug being present, but I do not think thatthese are “classic” autoantibodies (in contrast to thoseassociated with methyldopa). Drug-independentantibodies (e.g., cefotetan antibodies) are never seenwithout the presence of drug-dependent antibodiesand when the drug is stopped, the “autoantibodies”become weaker, eventually disappearing (see Table 7).I believe that these antibodies are not producedbecause of the drug’s effect on the immune system,butrather are part of the spectrum of antibodies producedas a result of the immune response to a hapten (e.g.,antibodies to the drug, carrier [RBC protein], and drug+ carrier).97 See Figure 1.

Drug-dependent antibodies can be of two types.One type of antibody can be inhibited by the drug(hapten inhibition) and reacts only with washed, drug-coated RBCs (prototype drug is penicillin). The othertype, which relates to most drugs, is not inhibited bythe drug and reacts with untreated RBCs in thepresence of the drug. We still do not know why. Therehave been two main suggestions:

• The drug antibody combines with the drug,forming immune complexes which attach to theRBC (transiently) and activate complement.

• The drug changes the RBC membrane,producinga new antigen (“neoantigen”) as a new epitope.This neoantigen may be a chemically modifiedmembrane protein or part drug, part RBCmembrane protein (“compound” epitope).

The latter part of the latter concept is the basis ofthe unifying hypothesis originally suggested byHabibi,98 and supported by Mueller-Eckhardt andSalama,99 which I feel has much merit. Unfortunately,like other concepts, it has some failings (these arediscussed in references 5, 6, and 99). Although theimmune complex hypothesis is out of favor, the clinicalfindings and some serologic data fit much better withthe “immune complex”hypothesis. Shulman and Reid100

tried to take some of the best of both hypotheses, andsuggest some explanations for the discrepancies, butpersonally I do not think that, in 2004, we have anybetter explanations that we did in 1994!

Table 7. Drugs that have been reported to induce RBC drug-independentantibodies (i.e., autoantibodies)*

Group 1cladribinefludarabinelevodopamefenamic acidmethyldopaprocainamide

Group 1acatergen chaparralchlordiazepoxidecianidanolcyclofenildiphenylhydantoinfenfluramineibuprofeninterleukin-2 methoin methysergiderituximab

Group 2azapropazonecarbimazoleceftazidimecefoxitincefotetancefotaxime chlorinated

hydrocarbonsdiclofenacglafenine latamoxefnomifensinephenacetinstreptomycinteniposidetolmetinzomepirac

*Drugs in groups 1 and 1a have been reported to induce drug-dependent antibodiesonly. More evidence is needed to prove that drugs in group 1a really can induce RBCautoantibodies. Drugs in group 2 induce drug-independent antibodies together withantibodies reacting by different mechanisms.

Fig. 1. Proposed unifying theory of drug-induced antibody reactions.97

The thicker, darker lines represent antigen-binding sites on theF(ab) region of the drug-induced antibody. Drugs (haptens) bindloosely or firmly to cell membranes, and antibodies may be madeto the drug (producing in vitro reactions typical of a drugadsorption [penicillin-type] reaction; the membrane compo-nents, or mainly membrane components (producing in vitroreactions typical of autoantibody); or part-drug, part-membranecomponents (producing an in vitro reaction typical of the so-called immune-complex mechanism).

Antibody to Drug

Antibody to(Mainly) MembraneComponents Antibody to Drug

and MembraneComponents

Drug

Red Cell Membrane



References1. Garratty G. Drug induced immune hemolytic

anemia and/or positive direct antiglobulin tests.Immunohematology 1985;2:1-8.

2. Garratty G. Current viewpoints on mechanismscausing drug-induced immune hemolytic anemiaand/or positive direct antiglobulin tests.Immunohematology 1989;5:97-106.

3. Garratty G. Immune hemolytic anemia and/orpositive direct antiglobulin tests caused by drugs.Immunohematology 1994;10:41-50.

4. Petz LD, Garratty G. Acquired immune hemolyticanemias. New York: Churchill Livingstone, 1980.

5. Garratty, G. Drug-induced immune hemolytic an-emia. In: Garratty G, ed. Immunobiology of trans-fusion medicine. New York: Dekker, 1994:523-51.

6. Petz LD, Garratty G. Immune hemolytic anemias.2nd ed. Philadelphia: Churchill Livingstone,2004:261-317.

7. Garratty G, Arndt P, Leger R. The changingspectrum of drug-induced immune hemolyticanemia over the last 25 years (abstract). Blood2003;102:560a.

8. Gralnick HR, McGinniss M, Elton W, et al.Hemolytic anemia associated with cephalothin.JAMA 1971; 217:1193-7.

9. Jeannet M, Bloch A, Dayer JM, et al. Cephalothin-induced immune hemolytic anemia. ActaHaematol 1976;55:109-17.

10. Rubin RN, Burka ER. Anti-cephalothin antibodyand Coombs’-positive hemolytic anemia. AnnIntern Med 1977; 86:64-5.

11. Moake JL, Butler CF, Hewell GM, et al. Hemolysisinduced by cefazolin and cephalothin in a patientwith penicillin sensitivity. Transfusion 1978;18:369-73.

12. Forbes CD, Mitchell R, Craig JA, et al. Acute intra-vascular haemolysis associated with cephalexintherapy. Postgrad Med J 1972;48:186-8.

13. Manoharan A, Kot T. Cephalexin-induced haemo-lytic anaemia (letter). Med J Aust 1987;147:202.

14. Branch DR, Berkowitz LR, Becker RL, et al.Extravascular hemolysis following the adminis-tration of cefamandole. Am J Hematol 1985;18:213-9.

15. Toy E, Nesbitt R, Savastano G, et al. Warmautoantibody following plasma apheresis, compli-cated by acute intravascular hemolysis associatedwith cefoxitin-dependent antibody resulting infatality (abstract).Transfusion 1989;29(Suppl):51S.

16. DeTorres OH. Hemolytic anemia and pancytopeniainduced by cefoxitin. Drug Intell Clin Pharm1983;17:816-8.

17. Shulman IA, Arndt PA, McGehee W. Cefotaxime-induced immune hemolytic anemia due toantibodies reacting in vitro by more than onemechanism.Transfusion 1990;30:263-6.

18. Salama A, Göttsche B, Schleffer T, et al. ‘Immunecomplex’ mediated intravascular hemolysis due to IgM cephalosporin dependent antibody.Transfusion 1987;27:460-3.

19. Malaponte G, Arcidiacono C, Mazzarino C, et al.Cephalosporin-induced hemolytic anemia in aSicilian child. Hematology 2000;5:327-34.

20. Garratty G, Postoway N, Schwellenbach J, et al. Afatal case of ceftriaxone (Rocephin)-inducedhemolytic anemia associated with intravascularimmune hemolysis. Transfusion 1991;31:176-9.

21. Lo G, Higginbottom P. Ceftriaxone inducedhemolytic anemia (abstract). Transfusion 1993;33:(Suppl):25S.

22. Bernini JC, Mustafa MM, Sutor LJ, et al. Fatalhemolysis induced by ceftriaxone in a child withsickle cell anemia. J Pediatr 1995;126:813-5.

23. Lascari AD, Amyot K. Fatal hemolysis caused byceftriaxone. J Pediatr 1995;126:816-7.

24. Borgna-Pignatti C, Bezzi TM, Reverberi R. Fatalceftriaxone-induced hemolysis in a child withacquired immunodeficiency syndrome. PediatrInfect Dis J 1995;14:1116-7.

25. Scimeca PG, Weinblatt ME, Boxer R. Hemolysisafter treatment with ceftriaxone (letter). J Pediatr1996;128:163.

26. Moallem HJ, Garratty G, Wakeham M, et al.Ceftriaxone-related fatal hemolysis in anadolescent with perinatally acquired humanimmunodeficiency virus infection. J Pediatr 1998;133:279-81.

27. Longo F, Hastier P, Buckley MJM, et al. Acutehepatitis, autoimmune hemolytic anemia, anderythroblastocytopenia induced by ceftriaxone.Am J Gastroenterol 1998;93:836-7.

28. Maraspin V, Lotric-Furlan S, Strle F. Ceftriaxone-associated hemolysis. Wien Klin Wochenschr1999;111(9):368-70.

29. Meyer O, Hackstein H, Hoppe B, et al. Fatalimmune haemolysis due to a degradation productof ceftriaxone. Br J Haematol 1999;105:1084-5.


G. GARRATTY

30. Punar M, Özsü H, Eraksoy H, et al. An adult case offatal hemolysis induced by ceftriaxone. ClinMicrobiol Infect 1999;5:585-6.

31. Viner Y, Hashkes PJ, Yakubova R, et al. Severehemolysis induced by ceftriaxone in a child withsickle cell anemia. Pediatr Infect Dis J 2000;19(1):83-5.

32. Seltsam A, Salama A. Ceftriaxone-induced immunehaemolysis: two case reports and a concise reviewof the literature. Intensive Care Med 2000;26:1390-4.

33. Falezza GC, Piccoli PL, Franchini M, et al.Ceftriaxone-induced hemolysis in an adult (letter).Transfusion 2000;40:1543-5.

34. Citak A, Garratty G, Ücsel R, et al. Ceftriaxone-induced haemolytic anaemia in a child with noimmune deficiency or haematological disease.J Pediatr Child Health 2002;38:209-10.

35. Kim S, Song KS, Kim HO, et al. Ceftriaxoneinduced immune hemolytic anemia: detection ofdrug-dependent antibody by ex-vivo antigen inurine. Yonsei Med J 2002;43:391-4.

36. Arndt P. Practical aspects of investigating drug-induced immune hemolytic anemia due tocefotetan or ceftriaxone—a case study approach.Immunohematology 2002;18:27-32.

37. Eastlund T,Mulrooney D,Neglia J, et al. Self-limitedimmune hemolysis in a child after six days ofceftriaxone therapy (abstract). Transfusion 2002;42(Suppl):96S.

38. Garratty G, Nance S, Lloyd M, et al. Fatal immunehemolytic anemia due to cefotetan. Transfusion1992;32:269-71.

39. Weitekamp LA, Johnson ST, Fueger JT, et al.Cefotetan-dependent immune hemolytic anemiadue to a single antibody reacting with both drugcoated and untreated red cells in the presence of drug. Book of Abstracts. ISBT/AABB JointCongress. Arlington, VA: American Association ofBlood Banks 1990:33.

40. Wojcicki RE, Larson CJ, Pope ME, et al. Acutehemolytic anemia during cefotetan therapy.Book of Abstracts. ISBT/AABB Joint Congress.Arlington,VA: American Association of BloodBanks 1990:33.

41. Gallagher MT, Schergen AK, Sokol-Anderson ML,et al. Severe immune mediated hemolytic anemiasecondary to treatment with cefotetan.Transfusion 1992;32:266-8.

42. Ehmann WC. Cephalosporin-induced hemolysis: acase report and review of the literature. Am JHematol 1992;40:121-5.

43. Chenoweth CE, Judd WJ, Steiner EA, et al.Cefotetan-induced immune hemolytic anemia.Clin Infect Dis 1992;15:863-5.

44. Wagner BKJ, Heaton AH, Flink JR. Cefotetandisodium-induced hemolytic anemia. AnnPharmacother 1992;26:199-200.

45. Dhawan M, Kiss JE, DiAddezzio NA, et al. Fatalcefotetan-induced immune hemolysis (abstract).Blood 1993;82:582a.

46. Eckrich RJ, Fox S, Mallory D. Cefotetan-inducedimmune hemolytic anemia due to the drug-adsorption mechanism. Immunohematology1994;10:51-4.

47. Peano GM, Menardi G, Quaranta L, et al. A rapidlyfatal case of immune haemolytic anaemia due tocefotetan.Vox Sang 1994;66:84-5.

48. Ogburn JR, Knauss MA,Thapar K, et al. Cefotetan-induced immune hemolytic anemia (IHA)resulting from both drug adsorption and immunecomplex mechanisms (abstract). Transfusion1994;34(Suppl):27S.

49. Mohammed S, Knoll S, vanAmburg A III, et al.Cefotetan-induced hemolytic anemia causingsevere hypophosphatemia. Am J Hematol 1994;46:369-70.

50. Garratty G, Leger RM, Arndt PA. Severe immunehemolytic anemia associated with prophylacticuse of cefotetan in obstetric and gynecologicprocedures. Am J Obstet Gynecol 1999;9:337-42.

51. Badon SJ, Cable RG. Hemolysis due to cefotetan(abstract).Transfusion 1999;39(Suppl):42S.

52. Johnson ST, Fueger JT, Gottschall JL. Cefotetan-dependent antibody cross-reacting with ceftri-axone treated RBCs (abstract). Transfusion 1999;39(Suppl):81S.

53. Stroncek D, Proctor JL, Johnson J. Drug-inducedhemolysis: cefotetan-dependent hemolytic anemiamimicking an acute intravascular immunetransfusion reaction. Am J Hematol 2000;64:67-70.

54. Naylor CS,Steele L,Hsi LR,et al. Cefotetan-inducedhemolysis associated with antibiotic prophylaxisfor cesarean delivery. Am J Obstet Gynecol 2000;182:1427-8.

55. Moes GS, MacPherson BR. Cefotetan-inducedhemolytic anemia: a case report and review of theliterature. Arch Pathol Lab Med 2000;124(a):1344-6.



56. Marques MB, Carr KD, Brumfield CG, et al. Apregnant patient with sickle cell disease andcefotetan-induced immune hemolysis. Lab Med2000;31(10):541-3.

57. Ray EK, Warkentin TE, O’Hoski PL, et al. Delayedonset of life-threatening immune hemolysis afterperioperative antimicrobial prophylaxis withcefotetan. Can J Surg 2000;43:461-2.

58. Chai L,Pomper GJ,Ross RL,et al. Severe hemolyticanemia and liver damage induced by prophylacticuse of cefotetan (abstract). Transfusion 2001;41(Suppl 1):58S.

59. Afenyi-Annan AN, Judd WJ. Cefotetan inducedimmune-mediated hemolysis complicated bythrombocytopenia: alloimmune or thrombotic(abstract)? Transfusion 2002;52(Suppl):46S.

60. Shammo JM, Calhoun B, Mauer AM, et al. First twocases of immune hemolytic anemia associatedwith ceftizoxime. Transfusion 1999;39:838-44.

61. Endoh T,Yagihashi A, Sasaki M, et al. Ceftizoxime-induced hemolysis due to immune complexes:case report and the epitope responsible forimmune complex-mediated hemolysis. Transfusion1999;39: 306-9.

62. Calhoun BW, Junsanto T, Donoghue MT, et al.Ceftizoxime-induced hemolysis secondary tocombined drug adsorption and immune-complexmechanisms.Transfusion 2001;41:893-7.

63. Chambers LA, Donovan LM, Kruskall MS.Ceftazidime-induced hemolysis in a patient withdrug-dependent antibodies reactive by immunecomplex and drug adsorption mechanisms. Am JClin Pathol 1991;95:393-6.

64. Fueger JT, Bell JA, Gottschall JL, et al. Ceftazidime-dependent antibody reacting with untreated redcells in the presence of drug (abstract).Transfusion 2001;41(Suppl 1):104S.

65. Arndt PA, Leger RM, Garratty G. Serology ofantibodies to second- and third-generationcephalosporins associated with immune hemo-lytic anemia and/or positive direct antiglobulintests. Transfusion 1999;39:1239-46.

66. Viraraghavan R, Chakravarty AG, Soreth J.Cefotetan-induced haemolytic anaemia. A reviewof 85 cases. Adverse Drug React Toxicol Rev2002;21:101-7.

67. Bateman ST, Hu E, Lane C, et al. Antibody toceftriaxone in HIV pediatric patients and potentialimplications for RBC hemolysis (abstract). Blood2002;100(11):48b.

68. Leger RM,Arndt PA, Garratty G. A negative or veryweakly positive DAT should not exclude theinvestigation of drug-induced immune hemolyticanemia. Transfusion 2003;43:1640-1.

69. Petz LD, Branch DR. Drug-induced immunehemolytic anemia. In: Chaplin H Jr., ed. Immunehemolytic anemias. New York: ChurchillLivingstone, 1985:47-94.

70. Osbourne SE, Johnson ST, Weitekamp LE, et al.Enhanced detection of drug-dependent antibodiesreacting with untreated red blood cells in thepresence of drug (abstract). Transfusion 1993;34(Suppl):20S

71. Arndt P, Garratty G. Use of albumin solutions forsolubilizing certain drugs can decrease binding ofdrugs to RBCs (abstract). Transfusion 2002;42(Suppl):105S.

72. Arndt PA, Garratty G. Is severe immune hemolyticanemia, following a single dose of cefotetan,associated with the presence of “naturally-occurring” anti-cefotetan? (abstract). Transfusion2001;41(Suppl):24S.

73. Arndt PA, Leger RM, Garratty G. Reactivity of “lastwash” elution controls in investigations ofcefotetan antibodies (abstract). Transfusion 1999;39:(Suppl):47S.

74. Spath P, Garratty G, Petz LD. Studies on theimmune response to penicillin and cephalothin inhumans. II. Immunohematologic reactions tocephalothin administration. J Immunol 1971;107:860-9.

75. Garratty G, Petz LD. Drug-induced immunehemolytic anemia. Am J Med 1975;58:398-407.

76. Branch DR, Sy Siok Hian AL, Petz LD. Mechanismof nonimmunologic adsorption of proteins usingcephalothin-coated red cells (abstract).Transfusion 1985;24:415.

77. Garratty G, Leger R. Red cell membrane proteins(CD55 and CD58) are modified followingtreatment of RBCs with cephalosporins (abstract).Blood 1995;86:68a.

78. Sirchia G, Murcuriali F, Ferrone S. Cephalothin-treated normal red cells: a new type of PNH-likecells. Experientia 1968;24:495-6.

79. Jamin D, Demers J, Shulman I, et al.An explanationfor nonimmunologic adsorption of proteins ontored blood cells. Blood 1986;67:993-6.

80. Jamin D, Shulman I, Lam HT, et al. Production of apositive direct antiglobulin test due to Suramin.Arch Pathol Lab Med 1988;112:898-900.


G. GARRATTY

81. Zeger G, Smith L, McQuiston D, et al. Cisplatin-induced nonimmunologic adsorption of immuno-globulin by red cells.Transfusion 1988; 28:493-5.

82. Getaz EP, Beckley S, Fitzpatrick J, et al. Cisplatin-induced hemolysis. N Engl J Med 1980;302:334-5.

83. Levi JA, Aroney RS, Dalley DN. Haemolyticanaemia after cisplatin treatment. Br Med J (ClinRes Ed) 1981;282:2003-4.

84. Nguyen BV, Jaffe N, Lichtiger B. Cisplatin-inducedanemia. Cancer Treat Rep 1981; 65:1121.

85. Cinollo G, Dini G, Franchini E, et al. Positive directantiglobulin test in a pediatric patient followinghigh-dose cisplatin. Cancer Chemother Pharmacol1988;21:85-6.

86. Weber JC, Couppie P, Maloisel F, et al. Anémiehémolytique au cisdiamino-dichloroplatinum. LaPress Médicale 1990;19:526-7.

87. Marani TM, Trich MB, Armstrong KS, et al.Carboplatin-induced immune hemolytic anemia.Transfusion 1996;36:1016-8.

88. Desrame J, Broustet H, de Talilly PD, et al.Oxaliplatin-induced haemolytic anaemia. Lancet1999;354:1179-80.

89. Garufi C, Vaglio S, Brienza S, et al. Immuno-hemolytic anemia following oxaliplatin adminis-tration (letter). Ann Oncol 2000;11:497.

90. Earle CC, Chen WY, Ryan DP et al. Oxaliplatin-induced Evan’s syndrome (letter). Br J Cancer2001;84:441.

91. Sørbye H, Bruserud Ø, Dahl O. Oxaliplatin-induced haematological emergency with animmediate severe thrombocytopenia andhaemolysis. Acta Oncol 2001;40:882-3.

92. Arndt P, Garratty G. Positive direct antiglobulintests associated with oxaliplatin can be caused byantibody and/or nonimmunological proteinadsorption (abstract). Transfusion 2003 43:102a.

93. Lutz P, Dzik W. Very high incidence of a positivedirect antiglobulin test (+DAT) in patientsreceiving Unasyn® (abstract). Transfusion 1992;32(Suppl):23S.

94. Garratty G, Arndt PA. Positive direct antiglobulintests and haemolytic anaemia following therapywith beta-lactamase inhibitor containing drugsmay be associated with nonimmunologicadsorption of protein onto red blood cells. Br JHaematol 1998;100:777-83.

95. Arndt PA, Leger RM, Garratty G. Positive directantiglobulin tests and haemolytic anaemiafollowing therapy with the beta-lactamaseinhibitor, tazobactam, may also be associated withnon-immunologic adsorption of protein onto redblood cells (letter). Vox Sang 2003;85:53.

96. Broadberry RE, Farren TW, Kohler JA, et al.Tazobactum-induced haemolytic anaemia possiblycaused by non-immunological adsorption of IgGonto patient’s red cells. Transfus Med 2004;14:53-7.

97. Garratty G. Target antigens for red-cell-boundautoantibodies. In: Nance SJ, ed. Clinical andbasic science aspects of immunohematology.Arlington, VA: American Association of BloodBanks, 1991:33-72.

98. Habibi B. Drug induced red blood cellautoantibodies co-developed with drug specificantibodies causing haemolytic anaemias. Br JHaematol 1985;61:139-43.

99. Mueller-Eckhardt C, Salama A. Drug-inducedimmune cytopenias: a unifying pathogeneticconcept with special emphasis on the role of drugmetabolites.Transfus Med Rev 1990;4:69-77.

100. Shulman NR, Reid DM. Mechanisms of drug-induced immunologically mediated cytopenias.Transfus Med Rev 1993;VII:215-29.

George Garratty, PhD, FRCPath, American Red CrossBlood Services, Southern California Region, 1130 S.Vermont Avenue, Los Angeles, CA 90006. e-mail: [email protected].

Phone, Fax, and Internet Information: If you have any questions concerning Immunohematology,Journal of Blood Group Serology and Education, or the Immunohematology Methods and Proceduresmanual,contact us by e-mail at [email protected] information concerning the National ReferenceLaboratory for Blood Group Serology, including the American Rare Donor Program, please contact SandraNance, by phone at (215) 451-4362, by fax at (215) 451-2538, or by e-mail at [email protected]


A serologic workup to investigate the presence ofirregular antibodies in a patient’s sample includestesting the patient’s serum (or plasma) against a panelof reagent RBCs and the patient’s autologous RBCs. Ifall panel RBCs and the autologous RBCs are equallyreactive, then the most likely conclusion is that thepatient’s serum contains a warm autoantibody and/or acold autoantibody. If all panel RBCs are reactive andthe autologous RBCs are nonreactive, then the patienthas either multiple alloantibodies and/or an antibodydirected against a high-frequency antigen. Using a casestudy approach, the first part of this review will coverwhat steps can be used to detect and identify anyalloantibodies underlying autoantibodies and how todifferentiate multiple alloantibodies from antibodiesdirected against high-frequency antigens. If it isdetermined that an alloantibody is directed against ahigh-frequency antigen, then it becomes important toknow the likelihood of that antibody being clinicallysignificant, because antigen-negative RBCs may bedifficult to obtain for transfusion. The second part ofthis review will focus on in vitro cellular methods usedto determine the clinical significance of alloantibodies.

It is 2 a.m. on the overnight shift and you arealone in the blood bank with a specimen drawnfrom a patient in the emergency department whohas a gunshot wound. This patient has a historyof previous admissions and transfusions at yourhospital. The antibody screen is 3+ withscreening cells I, II, and III by the gel method. Theemergency department is notified of the positivescreen and gives an indication that blood isneeded for transfusion; the patient’s hemoglobinis holding at 5g/dL but he is symptomatic.

The antibody’s specificity should be identified;howthis is done depends on the resources available at the

facility. The patient’s sample should be tested against apanel of reagent RBCs, either at the facility or at animmunohematology reference laboratory (IRL). Sincethis was a three-cell screen and all RBCs tested werereactive, the chance of this being a simple antibody islow and reactions of the three cells on the gel cardshould be reviewed to see if there are differences in theappearance of the reactions with each of the three RBCsuspensions. If there are differences, then multipleantibodies might be present; if all reactions appear thesame, then there may be an autoantibody or anantibody to a high-frequency antigen present.

A panel is performed and all RBCs are reactive; theautologous RBCs are nonreactive. The gel cards shouldbe reviewed to determine if the reactions are equal. Inthis case, the gel reactions are equal. The case isreferred to the local IRL and an anti-Yta is identified. Inthe interim period between sample submission andresolution, the patient receives four units ofincompatible blood. Over the next 3 weeks, the Hbgradually falls and the patient receives two units ofYt(a–) blood.

This review will focus on further investigation ofsituations involving cases that present serologicallywith all reagent RBCs reactive, including or notincluding the autologous RBCs. The starting point willbe in the hospital to highlight when, in eachcircumstance, it would be advisable to send the sampleto the IRL.

AutoantibodiesThe presence of an autoantibody is usually

determined by a reactive autologous control in apatient who has not been recently transfused. Thisreactivity could be primarily present at roomtemperature (probable cold autoantibody) or primarilypresent at 37°C and detected at the antihuman globulin

Review: what to do when all RBCsare incompatible—serologicaspectsS.T. NANCE AND P.A.ARNDT



(AHG) phase of testing (probable warm autoantibody).It is “best practice” to confirm that the reactivity is dueto an autoantibody by performing an autoadsorptionand showing that the reactivity is removed; theautoadsorbed sample can then be tested to determineif alloantibodies are also present. Autoadsorptions areperformed using autologous RBCs that have beentreated (e.g., with heat, enzyme, or ZZAP [acombination of enzyme and DTT]) to remove some ofthe globulin coating on the RBCs and/or to increase thecapacity of the RBCs to adsorb autoantibody from theserum. If the patient has been transfused recently, e.g.,in the last 3 or 4 months, then autologous RBCs mustbe obtained to determine if the reactivity isautoantibody prior to reporting it as such. Twomethods used to obtain autologous RBCs fromposttransfusion samples are reticulocyte separation byweight-based gradient separation1 and separation ofhemoglobin S positive RBCs by hypotonic wash.2

Once the autologous RBCs are obtained, they may beinsufficient in volume to use for routine adsorption butthey can be cleared of coating globulins and testedwith the patient’s serum and eluate to determine ifthere is reactivity directed against the patient’s RBCs.

If the patient has been transfused in the last 3 to 4months, then an autoadsorption should not beperformed to determine if underlying alloantibodiesare present,because a small amount of transfused RBCscan adsorb alloantibody.3 One method that can beused to determine if alloantibodies are present in atransfused patient with autoantibodies is allogeneicadsorption (adsorption of the patient’s serum/plasmawith allogeneic RBCs).4 The selection of RBCs for theadsorptions is critical; there must be adsorption ofautoantibody onto RBCs expressing some antigens andlacking others so that the commonly encountered,potentially clinically significant antibodies can beidentified in the adsorbed serum. The inherent riskwith allogeneic adsorptions is that, if there is anantibody to a high-frequency antigen, it will beadsorbed by the allogeneic RBCs and not bedetectable. Usually, the following RBC samples arechosen: R1, R2, and rr; at least one sample lacking K; randomly selected RBCs negative for Jka, Jkb, S, s;and RBCs treated with enzyme or ZZAP so that they are effectively negative for Fya,Fyb,M,and N. In general,consideration is usually not given to Lea, Leb, or P1

unless there is a suspicion that antibodies to thoseantigens are present. Variations of allogeneic adsorptionsinclude utilizing untreated cells (if the autoantibody is

directed at an enzyme-sensitive antigen) or adding LISSor PEG to the serum-RBC mixture.5–11 The use of PEGfor adsorptions has been the topic of controversybecause, in some studies, loss of alloantibody reactivityhas been reported.7,8,12 Adsorptions can be performedat 37°C or at 4°C to remove primarily warm or coldautoantibodies. A combination of temperatures is oftenused for those antibodies showing reactivity in roomtemperature (RT) and 37°C phases (sometimes referredto as warm/cold or combination autoantibodies). Theselection of the number of adsorptions that oneperforms varies between IRLs; some do one moreadsorption than the strength of the serum reactivity(e.g., two adsorptions for an antibody reacting 1+;someuse more adsorptions if testing is performed withPEG), and some use the strength of the DAT todetermine the number of adsorptions. Others monitorthe strength of the DAT on the adsorbing cells or testthe adsorbed serum against fresh adsorbing cells todetermine if more adsorptions need to be done. Thishas the disadvantage of delaying adsorptions andmakes the turnaround time longer.

37°C reactive autoantibodiesCase #1: A 68-year-old African American man is

admitted to the emergency room (ER) with a Hb of4.5 g/dL. He is pale and short of breath. Theadmitting resident requests four units of leuko-reduced RBCs as soon as possible. The antibodyscreen is 3+ positive at the AHG phase with all threescreening RBCs. Table 1 shows the initial panelresults, with reactivity with a few RBCs at the 37°Cphase, corresponding to anti-E specificity, andreactivity with all RBCs at the AHG phase, includingthe autologous control. The hospital refers the sampleto the local blood center IRL. The ER later calls to saythat the patient was transfused a month ago atanother hospital. The hospital blood bank calls thetechnologist at the IRL, who is just finishing thetesting on the autoadsorbed serum. Anti-E is presentin LISS at 37°C and AHG phases (see results in Table2). Because the patient has been recently transfused,the IRL repeats the adsorptions using enzyme-treatedallogeneic RBCs. Four allogeneic adsorptions areperformed using the RBC samples listed in the bottomof Table 3. The results of the allogeneic adsorptionsare shown in Table 3. The results show that thepatient has an antibody reactive with all RBCs tested(from the initial panel, Table 1), the suspected anti-E,and something else, possibly anti-K (reactive against


Review: all RBCs incompatible—serologic aspects

RBC suspension #2). Another possibility is that thisreactivity is due to an HLA antibody. Further testsshow that this is an anti-K. The anti-K was notpresent in the autoadsorbed serum, perhaps due toremaining transfused RBCs in the patient’s sampleadsorbing the weak reactivity. The reactivity with all RBCs tested (Table 1) is presumed to be due to an

autoantibody even though autoadsorption was notvalid due to the presence of transfused RBCs.

Case Comments: Results in this case suggest thatthe patient’s broadly reactive antibody has auto-specificity; the proof of an autoantibody would beadsorption of the antibody onto autologous RBCs. Inthe case of a transfused patient, autologous cell

Table 1. Initial panel results—case #1

Panel cells D C E c e f K k Fya Fyb Jka Jkb M N S s Lua Lub P1 IS 37° LISS Anti-IgG

1 + + 0 0 + 0 0 + 0 + + + + 0 + + 0 + + 0 0 3+

2 + 0 0 + + + + 0 + + 0 + 0 + 0 + 0 + + 0 0 3+

3 + 0 + + 0 0 0 + + 0 + + + + + + 0 + 0 0 2+ 3+

4 0 + 0 + + + 0 + 0 + + 0 0 + + 0 0 + + 0 0 3+

5 0 0 + + + + 0 + 0 + 0 + + + 0 + 0 + + 0 2+ 3+

6 + 0 0 + + + 0 + 0 + 0 + + + 0 + 0 + 0 0 0 3+

7 + 0 + + + + 0 + + + 0 + + + + 0 0 + + 0 2+ 3+

8 0 0 0 + + + + + + + + 0 + + + 0 0 + + 0 0 3+

9 0 0 0 + + + + + 0 + + 0 + + + + 0 + + 0 0 3+

10 0 0 0 + + + 0 + + 0 + + + 0 0 + 0 + + 0 0 3+

11 0 0 0 + + + 0 + 0 0 + + + 0 0 + 0 + + 0 0 3+

Auto 0 0 4+

Table 2. IRL testing on the autoadsorbed serum—case #1

Panel cells D C E c e f K k Fya Fyb Jka Jkb M N S s P1 IS 37°C LISS Anti-IgG

1 + 0 + + + + 0 + 0 + + 0 0 + + + 0 0 2+ 2+

2 + + 0 0 + 0 + 0 + 0 0 + + + + + + 0 0 0

3 + + 0 0 + 0 0 + 0 + 0 + + 0 + + 0 0 0 0

4 + 0 + + 0 0 + + 0 + + 0 + 0 + + + 0 2+ 2+

5 + + 0 + + + 0 + + + + + 0 + 0 + + 0 0 0

6 0 + 0 + + + 0 + + 0 + + + + + 0 0 0 0 0

7 0 0 + + + + 0 + 0 + + 0 + + 0 + + 0 2+ 2+

Table 3. The results of four allogeneic adsorptions—case #1

LISS LISS LISSPanel cells D C E c e f K k Fya Fyb Jka Jkb M N S s P1 37°C 37°C 37°C

1 + 0 + + + + 0 + 0 + + 0 0 + + + 0 0 2+ 0 0 0 2+

2 + + 0 0 + 0 + 0 + 0 0 + + + + + + 0 1+ 0 1+ 0 1+

3 + + 0 0 + 0 0 + 0 + 0 + + 0 + + 0 0 0 0 0 0 0

4 + 0 + + 0 0 + + 0 + + 0 + 0 + + + 0 2+ 0 0 0 2+

5 + + 0 + + + 0 + + + + + 0 + 0 + + 0 0 0 0 0 0

6 0 + 0 + + + 0 + + 0 + + + + + 0 0 0 0 0 0 0 0

7 0 0 + + + + 0 + 0 + + 0 + + 0 + + 0 2+ 0 0 0 2+

R1* 0 0 0 0 0

R2* 0 0 0 0 0

rr* 0 0 0 0 0 0

*enzyme-treated RBCs used for the allogenic adsorptions

An

ti-I

gG

An

ti-I

gG

An

ti-I

gG

Serum adsorbed withR1 R2

rr



separations would need to be done to isolate theautologous RBCs. Tests with autologous RBCs and thepatient’s serum or eluate are expected to be positive ifan autoantibody is present. If they are negative, then itshould be determined if an antibody to an antigen ofhigh frequency is present. The allogeneic adsorptionswould have adsorbed an antibody to a high-frequencyantigen. If the unexplained reactivity (with RBCsuspension #2) of the adsorbed serum had not beenshown to be anti-K, then tests with chloroquine-treatedRBCs would have been performed to determine if thereactivity was due to HLA antibodies.

Cold-reactive autoantibodiesSome laboratories approach the detection of

underlying alloantibodies, in the presence of suspectedcold-reactive antibodies, by using a prewarmingmethod. This is not recommended until it has beenproved by autoadsorption that the antibody is anautoantibody, as incorrect use of the technique hasbeen reported to yield incorrect results.13 Excellentdiscussions of the pros and cons of prewarming werepublished in 1995 and 1996.14–16 Once the auto-reactivity of the broadly reactive antibody is defined,prewarmed testing may be of value in the assessmentof underlying alloantibodies and in the characterizationof the clinical significance of the reactivity. Caution isadvised, however, as the loss of potentially clinicallysignificant alloantibody reactivity in prewarmed testinghas been described.17–19 Autoadsorption at cold

temperatures (4°C) is the method of choice to resolvecases of suspected cold autoantibodies. If the patienthas been transfused, allogeneic adsorption in coldtemperatures as well as use of rabbit RBCs foradsorption can be used. It is essential to read packageinserts for use of rabbit RBCs or stroma, as otherantibodies may also be adsorbed.

Case #2: A sample is received in the blood bankfrom a patient scheduled for cardiac surgery the nextday. The antibody screen by gel shows strongreactivity with many RBCs in the top layers of the geland some RBCs streaming through the gel in all tests.The panel gives similar results (all RBCs are reactive)and the sample is sent to the local IRL, as the hospitaldoes not test in tubes and has exhausted its resources.The IRL’s routine approach involves tube testing inalbumin (immediate spin [IS], RT for 15 minutes,37°C for 30 minutes with albumin, and AHG phaseusing anti-IgG) and against ficin-treated cells (37°Cfor 15 minutes, AHG phase with anti-IgG). The IRLobserves 1+ reactivity at IS, 3+ at RT, 1+agglutination and a trace of hemolysis at 37°C, and+w reactivity at the AHG phase that could reflectcarryover from 37ºC (see Table 4). The ficin-treatedRBC panel was completely hemolyzed at 37°C and noRBCs were left to read at the AHG phase. The DAT was3+ with anti-C3. Four adsorptions with autologousRBCs treated with ZZAP were performed at 4°C. Inthis case, the patient’s sample was sufficient to allowfour aliquots of treated autologous RBCs to be

Table 4. IRL results of panels using serum from case #2

Panel cells D C E c e f K k Fya Fyb Jka Jkb M N S s P1 IS RT 37°C 37°C

1 + + 0 0 + 0 0 + 0 + + + + 0 + + + 1+ 3+ 1+th* +w h† nc‡

2 + 0 0 + + + + 0 + + 0 + 0 + 0 + + 1+ 3+ 1+th +w h nc

3 + 0 + + 0 0 0 + + 0 + + + + + + 0 1+ 3+ 1+th +w h nc

4 0 + 0 + + + 0 + 0 + + 0 0 + + 0 + 1+ 3+ 1+th +w h nc

5 0 0 + + + + 0 + 0 + 0 + + + 0 + + 1+ 3+ 1+th +w h nc

6 + 0 0 + + + 0 + 0 + 0 + + + 0 + 0 1+ 3+ 1+th +w h nc

7 + 0 + + + + 0 + + + 0 + + + + 0 + 1+ 3+ 1+th +w h nc

8 0 0 0 + + + + + + + + 0 + + + 0 + 1+ 3+ 1+th +w h nc

9 0 0 + + + + + + 0 + + 0 + + + + + 1+ 3+ 1+th +w h nc

10 0 0 0 + + + 0 + + 0 + + + 0 0 + + 1+ 3+ 1+th +w h nc

11 0 0 0 + + + 0 + 0 0 + + + 0 0 + + 1+ 3+ 1+th +w h nc

Auto 1+ 3+ 1+th +w h nc

*Trace hemolysis†Completely hemolyzed‡No cells remaining

An

ti-I

gG

An

ti-I

gGAlbumin Ficin



prepared. In some cases, insufficient RBCs areavailable, due either to the patient’s Hb or to thesample size submitted. In those cases, it may benecessary to use the same aliquot of autologous RBCsover and over and re-treat the aliquot betweenadsorptions. The autoadsorbed serum wasnonreactive when tested with a set of three antibodyscreening cells. When the report of a coldautoantibody was given to the hospital, the cardiacsurgeon requested a cold agglutinin titer andthermal amplitude study to determine if this wouldinterfere with the surgery and the plans to use cellsalvage techniques. A thermal amplitude study wasperformed with dilutions of the patient’s serum insaline with the following results (titers): 30°C = 8,22°C = 64, 4°C > 2048. The surgery could not bepostponed. After consultation with the medicaldirector of the IRL, hospital blood bank physician,and surgeon, it was decided that the apparent coldagglutinin might be a problem with cell salvage at RTand cold cardioplegia might result in agglutinatedautologous RBCs in the coronary arteries. In thesurgery, the blood was first drained from the heartand then cold was applied. Cell salvage was not usedand the temperature was not lowered below normalRT. The patient did well and received two units ofblood, which were given through a blood warmer.

Case Comments: This patient had serologicresults consistent with a diagnosis of cold agglutininsyndrome; the patient’s Hct was 33%. This had notbeen detected prior to scheduling surgery. Thepatient’s pathologic cold agglutinin was more of a

concern for cardiac surgery than for other surgeriesbecause the temperature of the patient is lowered incardiac surgery.20

Multiple Antibody SpecificitiesOften the presence of multiple antibodies can be

discerned by the differences seen in serologic phasesor strengths of reactivity with panel RBCs in tubetesting and in the visual appearance of positivereactions in gel testing. Depending upon theconfiguration of the panel and the antibodies presentin the patient’s sample, nonreactive cells may beobtained or reactivity indicating dosage effects orseparate antibodies may be discerned. These featurescan be used to differentiate the presence of multipleantibody specificities from an antibody to a high-frequency antigen. The use of enzyme-treated RBCscan also be helpful in differentiating antibodyspecificities.

Case #3: A sample from a patient with sickle celldisease was received with a request for two units ofblood for outpatient transfusion the next day. Theantibody screen was positive. Cell I was 2+, cell IIwas 1+, and cell III was 3+ in the AHG phase. A panelwas tested and there was variable reactivity with 16of 16 RBC suspensions tested. The blood bankrequested that the IRL evaluate the sample, as it wasapparent the hospital would not have sufficient RBCpanels to evaluate the sample, even though theysaved the most recently expired panels. At the IRL, thepatient’s RBCs typed as D+C–E–c+e+ in the routineRh phenotype that is performed on all new patients.

Table 5. IRL special panel configured for case #3

Panel cells D C E c e f K k Fya Fyb Jka Jkb M N S s Lua Lub P1 IS 37°C 37°C

1 + 0 0 + + + 0 + + 0 0 + + + + + 0 + + 0 0 2+ 0 0√

2 + 0 0 + + + 0 + + 0 0 + + + 0 + 0 + 0 0 0 2+ 0 0√

3 0 0 0 + + + 0 + 0 + + 0 + + 0 + 0 + + 0 0 1+ 0 3+

4 0 0 0 + + + 0 + 0 + + 0 0 + + + 0 + + 0 0 1+ 0 3+

5 0 0 0 + + + 0 + 0 0 0 + + 0 + 0 0 + + 0 0 0√ 0 0√

6 + 0 0 + + + 0 + 0 0 0 + 0 + 0 + 0 + 0 0 0 0√ 0 0√

7 + 0 0 + + + + + 0 0 0 + + 0 + + 0 + + 0 1+ 3+ 2+ 3+

8 0 0 0 + + + + + 0 + 0 + + + + 0 0 + + 0 1+ 3+ 2+ 3+

9 + + 0 0 + 0 0 + 0 + 0 + + + + + 0 + + 0 1+ 2+ 2+ 3+

10 + + 0 0 + 0 0 + 0 + 0 + + + + + 0 + 0 0 1+ 2+ 2+ 3+

11 + 0 + + 0 0 0 + 0 + 0 + + + 0 + 0 + + 0 0 1+ 2+ 4+

12 + 0 + + 0 0 0 + 0 + 0 + + + 0 + 0 + + 0 0 1+ 2+ 4+

An

ti-I

gG

An

ti-I

gG

LISS Ficin


The IRL panel was configured by the staff so that thefirst six RBC suspensions were C–E–K– and the rest ofthe panel (six RBC suspensions) was designed to rulein anti-C, -E, or -K if they were present. The routinetesting at the IRL included a LISS panel and a ficin-treated RBC panel run in parallel. The results areshown in Table 5. The use of enzyme-treated RBCsdemonstrated the likelihood of anti-Fya. Subsequentstudies proved that anti-C, -E, -K, -Fya, and -Jka werepresent in this patient’s sample.

Case Comments: Utilizing information obtainedfrom the patient’s Rh phenotype made antibodyidentification easier. Additionally, the IRL studied thepanel that had been performed at the hospital and sawthat there was variation in reactivity. With theknowledge that the first antibodies formed often haveRh specificity, that the K antigen is very immunogenic,and that the hospital panel indicated those specificitiescould cause the reactivities seen, a custom panel wasdesigned. Since the IRL obtains panels from allmanufacturers, it was easy to make a “custom” panel toperform the testing. Custom panel configuration isfacilitated with computer software called Antigen PlusAb-ID (Rowny Systems, Inc., Gaithersburg, MD), whichallows the user to select the RBCs (by phenotype) fortesting. This program utilizes electronic informationfrom the panel manufacturer (Immucor/Gamma,Norcross, GA; Olympus America, Melville, NY), whichallows mistake-free entry into Antigen Plus. Ortho-Clinical Diagnostics (Raritan, NJ) does not yet providethis information; the user must manually enter thephenotype information from their panels. Then theuser requests the desired phenotype of RBCs and thesoftware peruses the database and presents choices tothe user. In this way, only RBCs that are of the neededphenotype are tested, thus conserving serum (and thereagent RBCs); the phenotype information on theselected RBCs is made into a panel worksheet by thesoftware, thus preventing handwriting and associatederrors. The information is presented in a panel formatwith space for recording test results.

By using all the resources available, the IRL wasable to resolve the case by testing two panels in about3 hours. If the IRL does not have access to RBCs withthe phenotypes needed for testing, allogeneicadsorptions of the patient’s serum may be useful toadsorb out some specificities and leave others behind.It should be remembered that very strongly reactiveantibodies may require high numbers of adsorptions tobe completely removed. Increasing the number of

adsorptions does increase the risk of diluting the serumand thus missing weak antibodies.

Antibodies to High-Frequency AntigensCase #4: A 59-year-old Caucasian man was on

dialysis for kidney failure. Two units of blood wererequested for transfusion with his next dialysisprocedure (EPO was not yet available). The antibodyscreen was positive (3+ with all cells by the IAT). Apanel was tested and all RBCs were reactive (3+);autologous RBCs were nonreactive. Samples weresent to the IRL.

Case #5: A 36-year-old Hispanic woman had ahistory of six pregnancies and four transfusions(associated with two of the deliveries). There was nohistory of HDN. She was admitted with a diagnosisof spontaneous abortion. The antibody screen andpanel were positive (2+ with all cells by IAT);autologous RBCs were nonreactive. Samples weresent to the IRL.

When antibodies to antigens of high frequency arepresent, the resolution can be time- and resource-consuming. The first sign that an antibody to a high-frequency antigen is present is usually uniformreactivity with all reagent RBCs tested; commonly, thisreactivity is only at the antiglobulin phase of testing. Inthe majority of cases, the autologous RBCs arenonreactive,but sometimes they are reactive due to thepresence of recently transfused RBCs (the reaction mayhave a mixed-field appearance) or due to anotherunrelated cause (medication, nonspecific, etc.).Approaches to resolution of the antibody’s specificityinclude testing RBCs that have been treated withdifferent agents, e.g., enzyme or DTT, and looking forenhancement or diminution of reactivity. Additionally,even though the use of a high titer value to steer theinvestigation toward a Knops (KN) system antibody isdownplayed by some experts,one of the authors (STN)still finds it valuable. If the antibody’s titer is equal toor greater than 32, DTT-treated RBCs are nonreactive,and the KN null cell is nonreactive, STN’s laboratoryreports the antibody as a probable KN system antibodyand stops the investigation. This approach wasapproved by the Technical Advisory Committee(customers).

In the evaluation of an antibody to a high-frequency antigen, the IRL’s frozen rare RBC and serumresources are usually utilized to the full extent toidentify the antibody. If the IRL does not have ABO-compatible RBCs negative for all high-frequency



antigens for testing against the patient’s serum, thenthe presence or absence of those antigens on thepatient’s RBCs becomes important. It is helpful toknow the patient’s race when choosing which antigensto test for and it is important to know the patient’stransfusion history when interpreting the RBC typingresults. Although IRLs are continually seeking to freezemore rare sera and RBCs, these are generally notavailable commercially and are obtained throughresource sharing groups like SCARF and an exchangeprogram in the American Red Cross IRLs known asTRANSFERASE. Some of the “more common” high-frequency, antigen-negative RBCs that may be availableon commercial panels are k–, I–, Yt(a–), Js(b–), andKp(b–), especially if a review of recently outdatedpanels is done. If expired panel RBCs are used and anegative reaction is obtained, then additional testingusing in-date RBCs or frozen/thawed RBCs is indicated.

If there is a suspicion that a sample contains anantibody to a high-frequency antigen and underlyingalloantibodies, it is advisable to identify the antibodiesto common antigens first. Test results of RBCs negativefor high-frequency antigens may be unproductive if theRBCs are positive for a common antigen that thepatient also has an antibody against. Underlyingalloantibodies can be identified by first phenotypingthe patient for common antigens, then selecting aphenotype-similar RBC sample for adsorptions. Ifuntreated RBCs are used for the first adsorption, thenan eluate can be prepared from the adsorbing RBCs,and a pure source of the antibody to the high-frequency antigen will be available. The adsorbedserum should be tested back against the adsorbingRBCs for evidence of complete adsorption, asalloantibodies are more variable in their strength (incomparison to most autoantibodies, which haverelatively low titers). The adsorbed serum can be usedfor identification of antibodies against commonantigens that the patient’s RBCs lack, and the eluate

from the first aliquot of adsorbing RBCs can be used totest for the antibody to the high-frequency antigen,once any antibodies to common antigens are identified.

The investigation may be assisted through the useof different media (e.g., enzymes). Frequently, the useof a combination of phenotyping the patient’s RBCs(e.g., for Jsb, S, s, Kpb, I, Jka, Jkb, Lub, H) and testing thepatient’s serum against “null” RBCs (e.g., cord, Oh, orDTT-treated [Ko] RBCs) is valuable. Knowledge of therace of the patient may be helpful early in theinvestigation to choose the more common RBCsnegative for high-frequency antigens known to exist inthat ethnic group (e.g., U–, Sl[a–], Js[b–] RBCs forAfrican Americans). Other clues include reactivity atRT (e.g., anti-H, -PP1P

k) or hemolysis of RBCs (e.g., anti-Vel, -P). See Table 6 for a list of antibodies to high-frequency antigens that are suggested by some of theclues mentioned above. This is not a complete list andthe reader is referred to the Blood Group AntigenFactsBook.21 If such investigation is unproductive,testing of the entire collection of RBCs negative forhigh-frequency antigens or molecular studies are oftenthe solution.

Once the specificity of an antibody to a high-frequency antigen is identified, then the clinicalsignificance must be determined so the hospital canknow what to transfuse. There are excellent texts thatreview the literature on this.21–23 For antibodies that areknown not to be clinically significant, by eitherspecificity or phase of reactivity, transfusion of randomRBCs is recommended. After transfusion, serologicreevaluation of the patient’s sample should beperformed for exclusion of other newly developedalloantibodies. Re-identification of existing antibodiesis optional.

If the antibody is known to be clinically significantand blood availability is usually not a problem (e.g., k–or Kp[b–] RBCs), then a request can be made to thelocal blood supplier and blood can be obtained. If the


Table 6. Some antibodies to high-frequency antigens to consider

Room Decreased African Asian, Hispanic,temperature Hemolysis reactivity in American Caucasian Native American

reactivity seen enzyme patient patient patient

Anti-H Anti-Vel Anti-Ch Anti-U Anti-Kpb Anti-Dib

Anti-I Anti-PP1Pk Anti-Rg Anti-Jsb Anti-Yta

Anti-PP1Pk Anti-P Anti-Inb Anti-Joa

Anti-Vel Anti-Jk3 Anti-JMH Anti-Hy

Anti-Sda Anti-H Anti-EnaTS Anti-Cra

Anti-P Anti-I Anti-EnaFS Anti-Tca

Anti-Ata


blood is not immediately available from the local bloodsupplier, the request can go to an American Rare DonorProgram member facility. The reader is referred to thenext issue of Immunohematology for a discussion ofthe American Rare Donor Program and the ISBT RareDonor Program.

If the antibody is directed against an antigen ofvariable clinical significance, and/or the blood is notimmediately available, then in vitro assays to predict invivo clinical significance should be used. These studiesare valuable; if results are negative, blood may beavailable as liquid random units, thus reducing thelogistics and time needed for transfusion, and the rareunits can be saved for those patients who cannotreceive random units. Patients who should not receiverandom units are those who have had either a positivein vitro test, a transfusion reaction to random units, orboth.

Methods Used to Determine the ClinicalSignificance of Alloantibodies

Before describing methods used to determine theclinical significance of alloantibodies, it is important todiscuss what is meant by the term “clinicallysignificant” in relation to alloantibodies and RBCs.24,25

Everyone would agree that an antibody that causes anobvious, clinical hemolytic transfusion reaction (fever,chills, hemoglobinemia, hemoglobinuria, etc.) is aclinically significant antibody and that antigen-negativeRBCs are needed for transfusion. What about anantibody that does not cause any overt clinicalsymptoms, but is associated only with laboratory signsof hemolysis (increased bilirubin, decreasedhaptoglobin, etc.)? Or an antibody that is notassociated with any clinical or laboratory signs ofhemolysis, but RBCs incompatible with it survive lessthan their normal lifespan? The definition of “clinicalsignificance” may vary depending upon the needs ofthe patient. For a patient with a short-termrequirement for blood, e.g., a surgical patient, it maynot be important for the transfused RBCs to havecompletely normal survival. On the other hand, for apatient with a hematological disorder and long-termrequirements for blood, it would be more importantthat the blood survive as long as possible, to limit thetotal number of transfusions.

The methods used to determine the clinicalsignificance of RBC alloantibodies have been reviewedelsewhere.24,26–32 They include routine serologicmethods (which give information about an antibody’s

specificity and thermal range), IgG subclassingmethods, in vivo survival studies (e.g., 51Cr and the“biological crossmatch”), and in vitro cellular assays.This review will concentrate only on the latter, the invitro cellular assays. These include the antibody-dependent cellular cytotoxicity (ADCC) assay33 (whichhas only been used to assess the clinical significance ofantibodies in the setting of HDFN), the chemi-luminescence test (CLT),34 and the monocyte mono-layer assay (MMA)32,35–37; most of this review will focuson the MMA.

The premise of all in vitro cellular assays is to mimicin the laboratory what occurs in vivo when antigen-positive RBCs are transfused into a patient with thecorresponding antibody. If the antibody-coated RBCsare destroyed, it will be by one of two mechanisms:intravascular or extravascular.20 Intra-vascular RBCdestruction (hemolysis of RBCs directly in thebloodstream) is a rare event due to the action ofcomplement on the RBC membrane.20 Most antibodiesdo not cause RBC destruction via this mechanism, butrather by the extravascular mechanism, which involvesdestruction of RBCs via macrophages in the liver orspleen. Macrophages recognize and attach to RBCs thatare coated with IgG and/or C3 via specific receptors on the macrophage.20 The attached RBCs are thendestroyed in three ways: the RBCs can be completelyengulfed (phagocytized) by the macro-phage, the RBCscan be partially phagocytized and the resultingspherocytes trapped and destroyed in the spleen,or theRBCs can be lysed extracellularly by enzymes secretedby the macrophage.20 In vitro cellular assays attempt to mimic extravascular RBC destruction.

In all in vitro cellular assays, RBCs are sensitizedwith the antibody in question and are then incubatedwith mononuclear cells (usually monocytes, theprecursors of macrophages). The in vitro reactions ofthe monocytes with the antibody-coated RBCs are thenascertained by (1) visual inspection for attachmentand/or phagocytosis of RBCs by monocytes (MMA), (2)measurement of oxygen radicals released by mono-cytes during phagocytosis (CLT), or (3) measurementof direct hemolysis of 51Cr-labeled RBCs (ADCC). TheMMA and CLT assays, which have been used todetermine clinical significance of alloantibodies fortransfusion purposes, have a cutpoint, or value abovewhich the test is considered to be positive. When apositive result is obtained, it is recommended thatantigen-negative blood be transfused in order to avoidan overt transfusion reaction.



A number of variations of the MMA have beenperformed by different investigators over the past 25+years.35–37 Many of the variables involved in the MMAhave been reviewed38; they include the source ofmonocytes,36,39 the RBC sensitization procedure(including whether or not fresh normal serum as asource of complement is added40,41), the “culture”conditions (e.g., a CO2 atmosphere versus ambientair42), and how the results are expressed. The assaycurrently used by the American Red Cross (SouthernCalifornia and Penn-Jersey Regions) is as follows:mononuclear cells from one or more donors areisolated using Ficoll®-Hypaque and incubated inambient air on a glass slide, the nonadherantlymphocytes are suctioned off, the sensitized RBCs(sensitized with and without fresh complementpresent) are added to the monocyte monolayer on theslide and incubated in ambient air; the nonadherantRBCs are removed (via suctioning and washing), andthe slide is stained and examined microscopically forthe percentage of monocytes with RBCs adheringand/or phagocytized.37

When the assay currently used by the AmericanRed Cross was originally set up, a cutpoint of 3 percentmonocyte reactivity was selected based on the resultsof unsensitized RBCs in the assay.37 The SouthernCalifornia Region later used statistical methods anddata from patients who had received incompatibletransfusions to determine that a more appropriatecutpoint for their assay was 5 percent reactivity.25

Using this 5 percent cutpoint, the sensitivity of theMMA is 100 percent (no false negatives) and thespecificity is 61 percent (there are false positives).25

If a result greater than the cutpoint is obtainedwhen testing an antibody, the MMA is considered to bepositive, and to avoid an overt transfusion reaction,antigen-negative blood is recommended as the optimalchoice for transfusion. When a result less than thecutpoint is obtained when testing an antibody, theMMA is considered to be negative; that antibody shouldnot cause acute hemolysis of transfused antigen-positive RBCs at that time. A negative MMA, however,does not guarantee that transfused,sensitized RBCs willhave normal long-term survival. If a patient with anegative MMA is transfused with antigen-positive RBCs,the exposure to more antigen may cause the antibodyto change its characteristics (e.g., strength/titer orsubclass composition). Thus, the MMA must berepeated in these patients prior to transfusion of moreantigen-positive RBCs. There are reports of negative

MMAs becoming positive after transfusion,43,44 as wellas of positive MMAs becoming negative aftertransfusion.45,46 In general, once a positive MMA resultis obtained, further MMAs are not performed.

Case #4 (continued): The antibody detected inthe Caucasian patient’s serum was identified as anti-Ge.The patient received Ge– units for 10 years, untilthe supply of Ge– RBCs in the country was almostdepleted. An MMA was requested and found to benegative (2.8%). The patient received two Ge+ unitsduring dialysis with no untoward reaction noted.The Hb rose as expected and there was no rise inbilirubin. One week after the transfusion, the MMAwas repeated and was still negative (2.7%); theantibody reactivity was 2+. The patient received twomore Ge+ units without an obvious reaction andthere was an expected rise in Hb. The MMA wasrepeated three more times over the next severalmonths. Each time the MMA was negative and thepatient was able to successfully receive Ge+ RBCs.The strength of the anti-Ge weakened; it was onlymicroscopically reactive in the last sample. Thepatient was then started on EPO and required nofurther transfusions.

Case #5 (continued): The antibody in theHispanic woman’s serum was identified as anti-Yta.An MMA was requested and found to be negative(0%). Two months later, during an emergency, thepatient was transfused with two Yt(a+) units ofblood. Her DAT became transiently positive but noclinical symptoms of a reaction were observed. Amonth later another MMA was requested inanticipation of a scheduled hysterectomy. The MMAwas now positive (23%) and the patient received twoYt(a–) units during her surgery.

These two patients demonstrated differentoutcomes after transfusion of incompatible RBCs.Patient #4’s antibody became progressively weakerand the MMAs remained negative. Patient #5’santibody changed its characteristics so that antigen-negative blood became the better choice fortransfusion.

Tables 7 and 8 summarize data from publishedreports on alloantibodies to high-frequency antigensthat may be of variable or unknown clinicalsignificance. In Table 7, MMA results from 29 suchantibodies are compared with the results of 51Crsurvival studies. 51Cr survival study results arecommonly expressed as the percentage ofincompatible RBCs surviving at 1 hour and 24 hours


after transfusion of a small dose of RBCs and/or as thehalf-life of those RBCs (T50Cr). The InternationalCommittee for Standardization in Haematology76

published the following guideline with respect to usingthe 51Cr survival study as a compatibility test: “In casesof urgency or when there is great difficulty in finding

completely compatible red cells,donor red cells may be transfusedwith minimal hazard when,following a test with 0.5 ml of thedonor’s red cells, . . . the red-cellsurvival at 60 minutes is not lessthan 70%.” It is known that anormal 1-hour 51Cr survival resultdoes not mean that the transfusedincompatible RBCs will survivenormally (i.e., the 24-hour survivalor T50Cr may be abnormal). As itcan be difficult to have 51Crsurvival studies performed, invitro assays (e.g., the MMA) weredeveloped to obtain the sameinformation as a 1-hour 51Crsurvival.24 Of the 29 reports inTable 7, only 16 included the 1-hour 51Cr survival data. Results inTable 7 are divided into those

which had 1-hour 51Cr survival data available and thosethat did not (e.g., data were available from other timesposttransfusion or the authors summarized theinterpretation of the 51Cr survival study without givingactual data).

Although most MMA and 51Cr results concurred,there are eight results in Table 7 that appearto be discrepant:one antibody with a negativeMMA had an abnormal 1-hour 51Cr survivalstudy and seven antibodies with positiveMMAs had normal 51Cr survival results (sixhad normal 1-hour results). The one antibodywith the negative MMA and abnormal 1-hour51Cr result was an anti-Ge in an untransfusedmale48; the MMA was negative with nativeserum but was positive when concentratedserum was used for testing. The sixantibodies (three anti-Yta and one each anti-Lu6, -Cra, and -Ata) with positive MMAs andnormal 1-hour 51Cr survival results hadabnormal 51Cr results after the 1-hour result.The other antibody (anti-Lu6) with a positiveMMA and a normal 51Cr result was associatedwith a median survival of 26 days,53 and thusappears to be falsely positive.

Table 8 shows data on 56 antibodieswhere MMAs were performed and thepatients were transfused with incompatibleRBCs (unfortunately, in many cases it is not



Table 7. Data from reports in the literature comparing MMA results of antibodies to high-frequencyantigens with 51Cr survival studies. Results for the 16 antibodies with 1-hour 51Cr survivalstudy results available are listed under the “1-hour 51Cr” heading (shaded columns) as eithernormal (N) or abnormal (Ab). The results for the remaining 13 antibodies without 1-hour 51Crresults are listed under the “51Cr survival” heading.

Antibody MMA = Negative MMA = PositiveSpecificity 1-hour 51Cr 51Cr survival 1-hour 51Cr 51Cr survival(references) # N Ab N Ab N Ab N Ab

Yta 25,35,37,43,47 9 1 0 2 0 3 1 0 2

Ge36,37,44,48,49 5 1 1 0 2 0 0 0 1

Lub 50 1 0 0 0 1 0 0 0 0

Lu651–53 3 0 0 1 0 1 0 1 0

Lu1254 1 1 0 0 0 0 0 0 0

Cra 55,56 2 0 0 0 0 1 0 0 1

Tca 45 1 1 0 0 0 0 0 0 0

Tcab 57 1 0 0 0 0 0 0 0 1

Lan37,58 2 0 0 0 0 0 2 0 0

Ata 59,60 2 0 0 0 0 1 1 0 0

Hya 61 1 1 0 0 0 0 0 0 0

Oka 62 1 0 0 0 0 0 0 0 1

TOTAL 29 5 1 3 3 6 4 1 6

Table 8. Data from reports in the literature comparing MMA results with result oftransfusion (txn) of incompatible RBCs. Results in the “Hemolytic txn, MMApositive reaction” column have been further subdivided into those describedas having (immediate or clinical signs/mild, delayed or laboratory signsonly/no information given).

Antibody MMA = Negative MMA = PositiveSpecificity No txn Hemolytic No txn Hemolytic(references) N reaction txn reaction reaction txn reaction

Yta 24,34,35,37,38,47,63,64 31 22 0 5 4 (1/3/0)

Ge34,36,37,65,66 6 5 0 0 1 (0/0/1)

Cra 67,68 4 2 0 1 1 (0/1/0)

Tca 69 1 0 0 0 1 (0/1/0)

Lub 46 1 0 0 0 1 (0/1/0)

Lu651 1 1 0 0 0

Lu870 1 0 0 0 1(1/0/0)

Hy35 1 0 0 0 1 (0/0/1)

Gy50 1 0 0 0 1 (0/1/0)

Joa 22 1 0 0 1 0

Jra 35,71,72 4 3 0 0 1 (1/0/0)

Lan37 1 1 0 0 0

Dib 73 1 0 0 0 1 (0/1/0)

Dra 74 1 1 0 0 0

Era 75 1 1 0 0 0

TOTAL 56 35 0 7 13 (3/8/2)


known if the MMA was performed before or after thetransfusion). In these cases, information was given asto whether or not there was evidence for a hemolytictransfusion reaction, but often not many details weregiven. The results in Table 8 listed under the“Hemolytic transfusion reaction” heading aresubdivided into those said to have an immediatereaction or clinical signs; those with a mild, delayedreaction or only laboratory signs of a reaction; andthose with no detailed information. Although mostMMA results agreed with the results of transfusion,there appear to be seven discrepant results in Table 8:seven patients with positive MMAs had no signs of atransfusion reaction and the MMA results appear to befalsely positive.

There are other single-case reports in the literature,on patients with antibodies to high-frequency antigenswith variable or unknown significance, where MMAswere performed but the patients were not transfusedwith incompatible blood. One study reported MMAdata on 251 alloantibodies; 192 of these alloantibodieswere directed against high-frequency antigens ofvariable or unknown clinical significance.25 Themajority (68%) of these 192 alloantibodies werepositive in the MMA. Thus, one should presume thatmost alloantibodies to high-frequency antigens havethe potential to be clinically significant, unless provenotherwise by MMA or 51Cr survival studies.

Acknowledgments We thank Margaret Manigly for her preparation of

this manuscript. The authors would also like to expresstheir sincere appreciation and admiration for GeorgeGarratty, PhD, who steadfastly and energeticallymentored and supported us throughout our careers!

References1. Reid ME,Toy PT.Simplified method for recovery of

autologous red blood cells from transfusedpatients. Am J Clin Pathol 1983;79:364-6.

2. Brown DJ. A rapid method for harvestingautologous red cells from patients with hemo-globin S disease. Transfusion 1988;28:21-3.

3. Laine EP, Leger RM,Arndt PA, et al. In vitro studieson the impact of transfusion on the detection ofalloantibodies after autoadsorption. Transfusion2000;40:1384-7.

4. Allogeneic red cell adsorption for removal ofwarm autoantibody. In: Immunohematologymethods and procedures. 1st ed. The AmericanNational Red Cross 1993:47-51.

5. Liew YW, Duncan N. Polyethylene glycol in auto-adsorption of serum for the detection ofalloantibodies (letter).Transfusion 1995;35:713.

6. Chiaroni J,Touinssi M, Mazet M, deMicco P, FerreraV. Adsorption of autoantibodies in the presence ofLISS to detect alloantibodies underlying warmautoantibodies. Transfusion 2003;43:651-5.

7. Barron CL, Brown MB. The use of polyethyleneglycol (PEG) to enhance the adsorption of auto-antibodies. Immunohematology 1997;13:119-22.

8. Judd WJ,Dake L. PEG adsorption of autoantibodiescauses loss of concomitant alloantibody.Immunohematology 2001;17:82-5.

9. Leger RM, Garratty G. Evaluation of methods fordetecting alloantibodies underlying warmautoanti-bodies. Transfusion 1999;39:11-6.

10. Branch DR, Petz LD. Detecting alloantibodies inpatients with autoantibodies (editorial).Transfusion 1999;39:6-10.

11. Cheng C-K,Wong ML, Lee AW. PEG adsorption ofautoantibodies and detection of alloantibodies inwarm autoimmune hemolytic anemia. Transfusion2001;41:13-7.

12. Leger RM,Ciesielski D,Garratty G.Effect of storageon antibody reactivity after adsorption in thepresence of polyethylene glycol (letter).Transfusion 1999;39:1272-3.

13. Storry JR, Mallory D. Misidentification of anti-Veldue to inappropriate use of prewarming andadsorption techniques. Immunohematology 1994;10:83-6.

14. Judd WJ. Controversies in transfusion medicine.Prewarmed techniques:Con. Transfusion 1995;35:271-5.

15. Mallory D. Controversies in transfusion medicine.Prewarmed tests: Pro—why, when, and how—notif. Transfusion 1995;35:268-70.

16. Garratty G. Attack on prewarmed tests—Toomuch hot air (letter)? Transfusion 1996;36:192.

17. Judd WJ, Steiner EA, Knaff P, Davenport RD.Failure to detect potentially significant antibodiesin prewarmed tests (abstract). Transfusion1995;35: (Suppl):68S.

18. Kilsby MA.Comparison of prewarmed techniques:with and without enhancements (abstract).Transfusion 1999;39(Suppl):104S.



19. Leger RM, Garratty G. Weakening or loss ofantibody reactivity after prewarm technique.Transfusion 2003;43:1611-4.

20. Petz LD, Garratty G. Immune hemolytic anemias.2nd ed. Philadelphia: Churchill Livingstone, 2004:133-65,352-6.

21. Reid M, Lomas-Francis C. The blood group antigenfactsbook. 2nd ed. San Diego: Elsevier AcademicPress, 2004.

22. Issitt PD,Anstee DJ. Applied blood group serology.4th ed. Durham: Montgomery 1998.

23. Daniels G. Human blood groups. 2nd ed. Oxford:Blackwell Science, 2002.

24. Garratty G. Evaluating the clinical significance ofblood group alloantibodies that are causingproblems in pretransfusion testing. Vox Sang1998; 74:285-90.

25. Arndt PA, Garratty G. A retrospective analysis of the value of monocyte monolayer assay results forpredicting the clinical significance of blood groupalloantibodies. Transfusion 2004 (in press).

26. Daniels G, Poole J, de Silva M, et al. The clinicalsignificance of blood group antibodies. TransfusMed 2002;12:287-95.

27. Leger RM. In vitro cellular assays and otherapproaches used to predict the clinicalsignificance of red cell alloantibodies: a review.Immunohematology 2002;18:65-70.

28. Reid ME, Øyen R, Marsh WL. Summary of theclinical significance of blood group alloantibodies.Semin Hematol 2000;37:197-216.

29. Nance S. Review: measuring red cell survival anddetermining the clinical significance of red cellantibodies. Immunohematology 1995;11:31-8.

30. Engelfriet CP, Overbeeke MAM, Dooren MC,Ouwehand WH, von dem Borne AEGKr. Bioassaysto determine the clinical significance of red cellalloantibodies based on Fc receptor-induceddestruction of red cells sensitized by IgG.Transfusion 1994;34:617-26.

31. Garratty G. Factors affecting the pathogenicity ofred cell auto- and alloantibodies. In: Nance SJ, ed.Immune destruction of red blood cells. Arlington,VA: American Association of Blood Banks, 1989;109-69.

32. Garratty G. Predicting the clinical significance ofalloantibodies and determining the in vivo survivalof transfused red cells. In: Judd WJ, Barnes A, eds.Clinical and serological aspects of transfusion

reactions: a technical workshop. Washington DC:American Association of Blood Banks, 1982:91-119.

33. Engelfriet CP, Ouwehand WH. ADCC and othercellular bioassays for predicting the clinicalsignificance of red cell alloantibodies. BaillieresClin Haematol 1990;3:321-37.

34. Hadley A, Wilkes A, Poole J, Arndt P, Garratty G.A chemiluminescence test for predicting theoutcome of transfusing incompatible blood.Transfus Med 1999;9:337-42.

35. Schanfield MS, Stevens JO, Bauman D. Thedetection of clinically significant erythrocytealloantibodies using a human mononuclearphagocyte assay.Transfusion 1981;21:571-6.

36. Branch DR, Gallagher MT, Mison AP, Sy Siok HianAL, Petz LD. In vitro determination of red cellalloantibody significance using an assay ofmonocyte-macrophage interaction with sensitizederythrocytes. Br J Haematol 1984;56:19-29.

37. Nance SJ, Arndt P, Garratty G. Predicting theclinical significance of red cell alloantibodiesusing a monocyte monolayer assay. Transfusion1987;27: 449-52.

38. Garratty G. Predicting the clinical significance ofred cell antibodies with in vitro cellular assays.Transfus Med Rev 1990;IV:297-312.

39. Zupanska B, Gronkowska A, Ziemski M.Comparison of activity of peripheral monocytesand splenic macrophages in the monocyte mono-layer assay.Vox Sang 1995;68:241-2.

40. Nance SJ, Arndt PA, Garratty G.The effect of freshnormal serum on monocyte monolayer assayreactivity.Transfusion 1988;28:398-9.

41. Church A, Mallory D, Kavitsky D, Nance S. Impactof complement in the monocyte monolayer assay(abstract).Transfusion 1997;37(Suppl):27S.

42. Branch DR, Gallagher MT. The importance of CO2

in short-term monocyte-macrophage assays(letter).Transfusion 1985;25:399.

43. AuBuchon JP,Brightman A,Anderson HJ,Kim B. Anexample of anti-Yta demonstrating a change in itsclinical significance.Vox Sang 1988;55:171-5.

44. DiNapoli J, Gingras A, Diggs E, Alicea-Tossas E,Kessler L. Survival of Ge+ red cells in a patientwith anti-Ge1,2: data from 51Cr, flow cytometric,IgG subclass, and monocyte erythrophagocytosisassays (abstract). Transfusion 1986;26:545.


.´


45. Anderson G, Gray LS, Mintz PD. Red cell survivalstudies with a patient with anti-Tca. Am J ClinPathol 1991;95:87-90.

46. Nance S, Scandone P, Fassl L, et al. Monocytemonolayer assay (MMA) results are affected bytransfusion of incompatible red cells (abstract).Transfusion 1997;37(Suppl):37S.

47. Kakaiya R, Sheahan E, Julleis J, et al. 51Chromiumstudies with an IgG1 anti-Yta (letter).Immunohematology 1991;7:107.

48. Pearson HA, Richards VL, Wylie BR, et al.Assessment of clinical significance of anti-Ge in anuntransfused man. Transfusion 1991;31:257-9.

49. Issitt PD, Gutgsell NS, Hervis L. Some storedantibodies give unreliable results in the monocytemonolayer assay (letter). Transfusion 1988;28:399-400.

50. Wren MR, Issitt PD. The monocyte monolayerassay and in vivo antibody activity (abstract).Transfusion 1986;26:548.

51. Gibson M, Devenish A, Daniels GL, Contreras M.A transfusion problem in a thalassaemic infantwith anti-Lu6 (abstract). Ann Mtg Br BloodTransfus Soc 1983;27.

52. Issitt PD, Valinsky JE, Marsh WL, DiNapoli J,Gutgsell NS. In vivo red cell destruction by anti-Lu6.Transfusion 1990;30:258-60.

53. Yahalom V, Ellis MH, Poole J, et al. The rare Lu:-6phenotype in Israel and the clinical significance ofanti-Lu6.Transfusion 2002;42:247-50.

54. Shirey RS, Øyen R, Heeb KN, Kickler TS, Ness PM.51Cr radiolabeled survival studies in a patient withanti-Lu12 (abstract). Transfusion 1988;28(Suppl):37S.

55. McSwain B, Robins C. A clinically significant anti-Cra (letter). Transfusion 1988;28:289-90.

56. Leatherbarrow MB, Ellisor SS, Collins PA, et al.Assessing the clinical significance of anti-Cra andanti-M in a chronically transfused sickle cellpatient. Immunohematology 1988;4:71-4.

57. Bell JA, Johnson ST, Moulds M, et al. Clinicalsignificance of anti-Tcab in the second example ofa Tc(a–b–) individual (abstract). Transfusion1989;29 (Suppl):17S.

58. Judd WJ, Oberman HA, Silenieks A, Steiner EA.Clinical significance of anti-Lan (letter).Transfusion 1984;24:181.

59. Ramsey G, Sherman LA, Zimmer AM, et al. Clinicalsignificance of anti-Ata.Vox Sang 1995;69:135-7.

60. Sweeney JD, Holme S, McCall L, et al. At(a–)phenotype: description of a family and reducedsurvival of At(a+) red cells in a proposita with anti-Ata. Transfusion 1995;35:63-7.

61. Baldwin ML, Ness PM, Barrasso C, et al. In vivostudies of the long-term 51Cr red cell survival ofserologically incompatible red cell units.Transfusion 1985;25:34-8.

62. Morel PA, Hamilton HB. Oka: an erythrocytic anti-gen of high frequency.Vox Sang 1979;36:182-5.

63. Gutgsell NS, Issitt LA, Issitt PD. Transfusion ofantigen-positive, serologically-incompatible bloodin immunized patients, based solely on the resultsof a monocyte monolayer assay (MMA) (abstract).Transfusion 1988;28:32S.

64. Eckrich RJ, Mallory DM, Sandler SG. Correlation ofmonocyte monolayer assays and posttransfusionsurvival of Yt(a+) red cells in patients with anti-Yta. Immunohematology 1995;11:81-4.

65. Zupanska B, Brojer E, McIntosh J, Seyfried H,Howell P. Correlation of monocyte-monolayerassay results, number of erythrocyte-bound IgGmolecules, and IgG subclass composition in thestudy of red cell alloantibodies other than D.VoxSang 1990;58:276-80.

66. Hildebrandt M, Hell A, Etzel R, Genth R, Salama A.Determination and successful transfusion of anti-Gerbich-positive red blood cells in a patient witha strongly reactive anti-Gerbich antibody.Infusionsther Transfusionsmed 2000;27:154-6.

67. Chapman RL, Hare V. Successful repeated trans-fusion of Cr(a+) blood to a patient with anti-Cra

(abstract). Transfusion 1992;32:23S.68. Byrne PC, Eckrich RJ, Malamut DC, Mallory DM,

Sandler SG. Use of the monocyte monolayer assay(MMA) to predict the clinical significance of anti-Cra (abstract). Transfusion 1995;35:61S.

69. Kowalski MA, Pierce SR, Edwards RL, et al.Hemolytic transfusion reaction due to anti-Tca.Transfusion 1999;39:948-50.

70. Kobuszewski M, Wallace M, Moulds M, et al.Clinical significance of anti-Lu8 in a patient whoreceived Lu:8 red cells (abstract). Transfusion1988;28(Suppl):37S.

71. Bacon J, Sherrin D,Wright RG. Case report, anti-Jra

(letter).Transfusion 1986;26:543-4.72. Kwon MY, Su L, Arndt PA, Garratty G, Blackall DP.

Clinical significance of anti-Jra: report of two casesand review of the literature. Transfusion 2004;44:197-201.



73. Leger R, Arndt P, Co A, O’Brien L, Garratty G.Clinical significance of an anti-Dib assessed byflow cytometry. Immunohematology 1997;13:93-6.

74. Nakache R,Levene C, Sela R,Kaufman S, Shapira Z.Dra (Cromer-related blood group antigen)-incompatible renal transplantation. Vox Sang1998;74:106-8.

75. Daniels GL, Judd WJ, Moore BPL, et al. A ‘new’ highfrequency antigen Era. Transfusion 1982;22:189-93.

76. International Committee for Standardization inHaematology. Recommended method for radio-isotope red-cell survival studies. Br J Haematol1980;45:659-66.

Sandra Taddie Nance, MS, MT(ASCP)SBB (correspon-ding author), Director, Technical Services, Penn-JerseyRegion and National Reference Laboratory for BloodGroup Serology, American Red Cross Blood Services,700 Spring Garden Street, Philadelphia, PA 19123and Patricia A. Arndt, MS, MT(ASCP)SBB, SeniorResearch Associate, Southern California Region,American Red Cross Blood Services, 1130 SouthVermont Avenue, Los Angeles, CA 90006.

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

Deadlines1st week in January for the March issue

1st week in April for the June issue1st week in July for the September issue

1st week in October for the December issue

Fax these items to Mary H. McGinniss, Managing Editor, (301) 299-7443.


Manuscripts: The editorial staff of Immunohematology welcomes manuscripts pertaining to blood groupserology and education for consideration for publication.We are especially interested in case reports, paperson platelet and white cell serology, scientific articles covering original investigations, and papers on the use ofcomputers 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.Instructions for scientific articles and case reports can be obtained by phoning or faxing a request to Mary H.McGinnniss, Managing Editor, Immunohematology, at (301) 299-7443, or see “Instructions for Authors” inevery issue of Immunohematology or on the Web. Include fax and phone numbers and e-mail addresswith your manuscript.


Introduction“As soon as the blood began to enter into his veins,

he felt the heat along his arm and under his armpits.His pulse rose and soon after we observed a plentifulsweat over his face. His pulse varied extremely at thisinstant and he complained of great pains in his kidneys,and that he was not well in his stomach, and that hewas ready to choke unless given his liberty. He wasmade to lie down and fell asleep, and slept all nightwithout awakening until morning. When he awakenedhe made a great glass full of urine, of a color as black asif it had been mixed with the soot of chimneys.”1

The above description of the signs and symptomsfollowing a blood transfusion were published by JeanDenis, a physician working for King Louis XIV in 1668.We can thank Dr.Denis for providing a vivid and classicdescription of what is now known to be a hemolytictransfusion reaction (HTR). Various in vitro (or, lessfrequently, in vivo) immunohematologic techniqueswhich attempt to prevent HTRs have evolved over theyears. The “goal” is to predict compatibility and trans-fuse RBCs which will survive and function as well asthe recipient’s own RBCs. One is more comfortabletransfusing “compatible” RBCs. Nonetheless, there aretimes when “incompatible” RBCs must be transfused.This paper will review clinical aspects of this decision.

Issues for ConsiderationIssues that should be considered when caring for a

patient requiring incompatible blood are: “Istransfusion really necessary?” “Is the antibody(ies)clinically significant?” and “How should the transfusionbe administered?”

When to transfuse incompatible RBCsIt can be difficult to state with certainty when to

transfuse incompatible RBCs to an anemic patient asthere are many variables which play a role inindividuals’ responses to their Hb (or Hct) levels. Most

physicians still utilize the Hb level as the “trigger” fortransfusion. There is no absolute Hb level below whicha patient will not survive and that makes it difficultwhen considering transfusing incompatible RBCs.

The signs and symptoms experienced by a patientare usually consistent with the rate at which the an-emia developed.2 Weiskopf et al.3 found that acuteisovolemic anemia to a Hb level of < 6 gm/dL resultedin mild, reversible changes in memory reaction time,while < 5 gm/dL impaired immediate and delayedmemory. Cardiovascular condition and age are alsoimportant factors to consider when deciding if (andwhen) transfusion is necessary. Cardiovascularcompensatory mechanisms to anemia, such as increasein heart rate, increase in stroke volume,and decrease inperipheral vascular resistance, are impaired in patientswith cardiovascular diseases and (usually) in theelderly. Carson et al.4 retrospectively analyzed the 30-day mortality of 1958 surgical patients >18 years of agewho refused transfusion for religious reasons. Theycompared outcome in those with and withoutcardiovascular disease (history of angina, myocardialinfarction, congestive heart failure, or peripheralvascular disease). Mortality was 1.3 percent (0.8% to2.0%) in patients with a preoperative Hb of 12 g/dL orgreater and 33.3 percent (18.6% to 51.0%) in patientswith a preoperative Hb < 6 g/dL. A correlation notedbetween serious morbidity and blood loss in the lowpreoperative Hb group was more apparent in patientswith cardiovascular disease than those withoutcardiovascular disease. Hebert et al.5 published amulticenter, randomized, controlled study examiningwhether a restrictive or liberal transfusion strategy hadan effect on mortality in critically ill patients witheuvolia. Patients were randomly assigned to maintaintheir Hb to either between 7.0 and 9.0 g/dL (restrictivetransfusion strategy) or between 10.0 and 12.0 g/dL(liberal transfusion strategy). Interestingly, 30-daymortality rates were similar in the two groups and, infact, were significantly less in the restrictive group in

Review: transfusing incompatibleRBCs—clinical aspectsG. MENY

individuals < 55 years of age and in those with an AcutePhysiology and Chronic Health Evaluation II score of < 20. This benefit was lost, however, in critically illpatients with acute myocardial infarction and unstableangina.

Tachycardia and postural hypotension, which areusually present in acute blood loss, are not often seenwith chronic anemia. A patient with chronic anemiamay have an increased blood volume due to anoverexpanded plasma volume. Rapidly administeringblood to a patient with severe chronic anemia mayprecipitate cardiopulmonary decompensation (orcirculatory overload). Indeed, circulatory overload isan important and underrecognized complication oftransfusion, particularly in the elderly.6 See Popovskyfor a review on this transfusion complication.7

Thus, in summary, while otherwise healthy patientsusually will require a transfusion when their Hbdeclines to < ~ 6 to 7 g/dL, these patients may be ableto tolerate and compensate at these low Hb levelswithout transfusion, particularly if the anemia hasslowly evolved. However, this does not apply to thepatient with cardiovascular disease.

Note that the preceding discussion applies topatients for whom compatible or incompatible RBCsare available for transfusion. Decisions to transfusecompatible blood in urgent situations are frequentlymade with little hesitation. However, transfusingincompatible blood is frequently delayed until thepatient’s life is in serious jeopardy. Transfusion ofincompatible blood must occur where there is anurgent, life-threatening clinical need, and transfusionshould not be withheld because of the serologicfindings.

Is the antibody clinically significant?As one is evaluating the patient’s clinical status and

determining the urgency of the RBC transfusionrequirement, frequently asked simultaneous questionsinclude: “How clinically significant is this RBCantibody?” and “What will happen to the RBCs iftransfused—will they be destroyed intravascularly?extravascularly? how rapidly?” Some antibodies areknown to be clinically significant (i.e., causehemolysis), such as anti-A, -B, -D, and -K. Others areusually not considered clinically significant, such asanti-Lea. M.E.Reid and C.Lomas-Francis have written anexcellent source book on blood group antibodies foruse by transfusionists.8 This book,as well as a literaturesearch (e.g., the National Library of Medicine’s PubMed

at http://www.ncbi.nlm.nih.gov/PubMed) can provideinformation on general “tendencies” of an antibody’sclinical significance. Thus, if time is available, in vitrotesting of the patient’s antibody may be desirable andhelp predict, for example, which RBC units (antigenpositive or antigen negative) should be transfused. Thepaper in this issue by Nance and Arndt providesadditional information on useful laboratory assays, suchas the monocyte monolayer assay (MMA).

In vivo testing is rarely performed, although the 1-hour 51Cr RBC survival study recommended by theInternational Committee for Standardization inHematology is the gold standard for predicting clinicalsignificance of a patient’s antibody.9 A radiolabeled 0.5 mL sample of incompatible RBCs is injected intothe patient and samples are obtained at 10 and 60minutes. If radioactivity in the plasma at 10 and 60minutes is < 3% of the total injected and the RBCsurvival at 60 minutes is at least 70%, then transfusionof the incompatible blood carries minimal risk. Mostfacilities are not equipped to perform this type of invivo testing.

Another in vivo test which may be performed isknown as the “in vivo crossmatch” or “biologicalcrossmatch.” 10 to 50 mL of unlabeled incompatibleRBCs is transfused and a sample is collected from thepatient posttransfusion to check for hemoglobinemia.10

This test only detects intravascular hemolysis.Extravascular RBC destruction cannot be predicted.

The transfusionIf, and when, the decision is made to transfuse,how

can the process be made as free of adverse events aspossible? It is best to always transfuse using leukocyte-reduced blood components. Although the patient’sclinical condition may not warrant their use, leukocyte-reduced blood components can prevent febrilenonhemolytic transfusion reactions, which arecomplications of transfusion and can mimic a HTR. Forexample, one does not want to discontinue a RBCtransfusion to a patient with multiple alloantibodiesand autoantibodies because of fever (due to a febrilenonhemolytic transfusion reaction). Premedicationwith antipyretics may mask the fever of a HTR, butother signs and symptoms of hemolysis will still bepresent. Premedication with an antipyretic orantihistamine (for patients with a history of allergicreactions) may be administered immediately prior totransfusion if administered intravenously. They shouldbe given 30 to 60 minutes prior to the start of the


G. MENY


transfusion if given orally. Some patients may benefitfrom receiving corticosteroids or IVIG (see CaseStudies section of this paper).

If possible, transfuse the patient during a timewhen adequately trained staff is available to monitorthe transfusion. This includes laboratory staff availableto interpret any adverse reactions,should they develop.The most knowledgeable individuals for all aspects of atransfusion are usually available during the day, ratherthan at night.

Systems must be in place to select the unit andissue and match it to its intended recipient. As with anytransfusion, make certain that an informed consentwhich complies with applicable laws is obtained. Vitalsigns, including blood pressure,pulse,and temperature,should be recorded prior to initiating the transfusion.The patient should be kept well hydrated during thetransfusion. Once the transfusion is started, thetransfusionist should remain with and observe thepatient for at least the first 15 minutes. Vital signsshould be documented at that time. The patient shouldbe observed at frequent intervals. Vital signs should berecorded at the end of the transfusion.

Davenport summarized the most frequent clinicalsigns and symptoms observed in 90 cases ofintravascular HTRs and 101 cases of extravascularHTRs (Table 1).11 Signs and symptoms of an acute HTRfrequently become apparent within the first fewminutes of the transfusion, hence the need for thetransfusionist to remain with the patient during thattime. Conscious patients should be educated aboutpossible adverse effects of transfusion and theimportance of their immediately reporting any of thesigns or symptoms of a HTR (acute or delayed). If thepatient is unconscious, vital signs, such as pulse andtemperature, should be checked at regular intervalsthroughout the transfusion and the patient should beobserved for evidence of an acute HTR, includinguncontrollable bleeding (disseminated intravascularcoagulation).

The timing and location of any RBC destruction areimportant to note when transfusing incompatibleblood. Immune-mediated RBC destruction can becategorized by the time of the reaction—acute ordelayed—and by the location of the hemolysis—intravascular or extravascular. Acute hemolysis takesplace within minutes to hours (up to 24 hours aftertransfusion) and delayed hemolysis occurs withinseveral days following transfusion. During intra-vascular reactions, antigen-antibody interactions resultin binding of complement and generation of themembrane attack complex (C5b–9). This causesosmotic lysis of RBCs, resulting in hemoglobinemia and hemoglobinuria. Either IgG or IgM antibodies arecapable of causing intravascular lysis. Intravascularhemolysis, for example, could be observed whentransfusing Vel+ RBCs to a patient with anti-Vel. Duringextravascular reactions, the membrane attack complexis not generated. RBCs become sensitized with IgGantibodies, complement, or both, and are removed bymacrophages within the reticuloendothelial system.12

Extravascular hemolysis, for example, could beobserved when transfusing e+ RBCs to a patient withanti-e.

When a patient has multiple antibodies andtransfusion of incompatible blood is required,incompatible blood is chosen which will survive thelongest and with minimal adverse effects. For example,a patient in need of urgent surgery has “antibody 1”and“antibody 2.” Blood compatible with both cannot beobtained without importing it from out of the countrythrough the American Rare Donor Program and surgerymust be performed. “Antibody 1” frequently destroysRBCs intravascularly, while “antibody 2” frequentlydestroys RBCs extravascularly. Both antibodies aredetectable in the patient’s serum. MMAs show thatboth are clinically significant. If compatible blood canbe found for only one of the antibodies, which shouldbe selected? Blood compatible with “antibody 1”(incompatible with “antibody 2”) should be selectedbecause otherwise intravascular lysis is likely to occur,and it should be avoided.

What laboratory tests can be performed aftertransfusing incompatible blood to assist in the clinicalmanagement of the patient? Tests which can evaluatethe presence of both intravascular and extravascularhemolysis (Table 2) should be performed. Many ofthese tests are the same as those used in the evaluationof a HTR workup when “compatible” blood istransfused. In addition to the Hb and Hct, examination

Review: incompatible transfusion—clinical aspects

Table 1. Signs and symptoms of a HTR at initial presentation11

Intravascular HTR Extravascular HTR(% of Patients) (% of Patients)

Fever, chills, or both 82 50

Pain 19 13

Hypotension, tachycardia, or both 12 —

Nausea, vomiting, or both 12 —

Jaundice — 12

Dyspnea 10 1


G. MENY

of the peripheral blood smear is important.2 Forexample, RBCs may appear in aggregates in coldagglutinin disease. Spherocytes and microspherocytesare usually observed in immune hemolytic anemia.RBC fragments and helmet cells, though, are suggestiveof a microangiopathic hemolytic anemia, not animmune-mediated process.

Other laboratory tests can be used to indicate thepresence of intravascular or extravascular hemolysis.An acute intravascular hemolytic event tends to befollowed by low serum haptoglobin, hemoglobinemia,and hemoglobinuria. Langley et al.13 described twopatients who developed very low serum haptoglobinlevels which persisted for several days after HTRs. Incontrast, there was no consistent change inhaptoglobin levels in 21 patients following compatibleblood transfusion. Haptoglobin is an acute-phasereactant and comorbid conditions such as infectionmay mask the diagnosis by somewhat elevatinghaptoglobin levels.14 However, in classic acuteintravascular hemolysis, the haptoglobin level is usually< 25 mg/dL and may approach 0 mg/dL.

Hemoglobinemia can be detected by inspecting asample of the patient’s plasma. A centrifuged urinesample is examined for hemoglobinuria. Care must betaken to differentiate intact RBCs from freehemoglobin and free myoglobin, if rhabdomyolysis is

suspected.15 Both hemoglobinemia and hemoglobin-uria are rather transient and may be cleared within amatter of hours. The urine should be examined forhemosiderin if it is suspected that the acuteintravascular hemolytic episode took place a few dayspreviously or is chronic in nature. Hemosiderin is iron-containing granules in the urine which are first shed afew days after hemolysis begins. Hemosiderin,however, may be found in any condition which causesdeposition of iron in the renal parenchyma, such ashemochromatosis or multiple transfusions.

Markedly elevated levels of lactate dehydrogenase(LD) are observed with acute intravascular hemolysisdue to RBC lysis. Although rarely tested for, a LD1:LD2flip may be observed in hemolytic anemia if LDisoenzymes are examined. The LD1:LD2 flip is alsoobserved in patients with myocardial infarction.16

An elevated reticulocyte count is indicative of butnot specific for hemolysis. Patients with sickle cell dis-ease have a reticulocytosis as a means of compensatingfor shortened RBC survival. If a reticulocytopenia ispresent and hemolysis is occurring, these patients areat risk for developing life-threatening anemia.17 Theunique challenges of transfusing patients with sicklecell disease are briefly discussed later in this paper.

Extravascular immune-mediated RBC destructionoccurs when immunoglobulin- and/or complement-sensitized RBCs are destroyed by macrophages withinthe reticuloendothelial system. Hemoglobinemia andhemoglobinuria typically do not occur. While serumhaptoglobin levels may decrease, they do not reach thevery low levels seen with intravascular hemolysis.Usually, total serum bilirubin increases only slightly(from the normal < 1.5 mg/dL to < 6 mg/dL) during ahemolytic episode in patients with normal liverfunction.17 An increase in indirect serum bilirubin andurine urobilinogen will also be observed. If the ele-vated total serum bilirubin is composed predominantlyof indirect bilirubin, hemolysis or Gilbert’s disease islikely to be present.14

Case StudiesIn a patient with multiple antibodies or an antibody

to a high-frequency antigen, one may anticipate theneed to transfuse “incompatible” blood. The followingtwo case studies, and a paper in this issue by A.M.Svensson et al., illustrate how a decision is reached.

A 77-year-old woman with a 6.7 g/dL Hb requiredtransfusion (i.e., she was clinically symptomatic), butno compatible blood was available in the United States.

Table 2. A diagnostic approach to hemolysis

Determine CBC, reticulocyte count, haptoglobin, bilirubin (totaland indirect), and urinalysis. Examine peripheral smear.

• Complete blood countDetermine Hb, Hct, and RBC indicesPlatelet count usually normal with intra- and extra-vascularhemolytic anemiaThrombocytopenia with positive DAT: Evans syndrome

• Peripheral smearSpherocytes suggest hemolytic anemiaHelp rule out other causes for anemia (e.g., schistocytes,malignancy, hemoglobinopathy, iron deficiency)

• Serum haptoglobinLow in acute intravascular hemolysisAcute-phase reactant

• Lactate dehydrogenaseElevated in intravascular hemolysisElevated in other damaged tissue or neoplastic cells

• Serum bilirubinElevated total and indirect in extravascular hemolysisSlight increase in total and indirect in intravascular hemolysisHigher levels seen in patients with compromised liver functionor Gilbert’s disease

• UrinalysisHemoglobinuria seen in intravascular hemolysis Rule out hematuria, myoglobinuria, porphyria, which can alsocause “red urine”Hemosiderinuria seen in intravascular hemolysis


Review: incompatible transfusion—clinical aspects

She was group O, D–, with anti-hrB, -E, and -S. MMAswere performed by the American Red Cross NationalReference Laboratory for Blood Group Serology andthe results were interpreted as “clinically significant,”(i.e., these antibodies would result in less than normalRBC survival). Blood was available from anothercountry, but it would take several days to arrive.

How could this patient be managed untilcompatible blood was available? Initially, it was notedthat RBCs negative for her “formed”antibodies, but D+,were available. Two units of D+, hrB–, E–, S– RBCs weretransfused and her Hb increased to 10 g/dL. However,as an anti-D developed, the Hb declined to 5.1 g/dLover the next 6 days. Three units of “incompatible”D–,E–, S–, hrB+ RBCs were then transfused, raising thepatient’s Hb to 8.6 g/dL. These units were destroyedextravascularly, and Hb declined to 6.7 g/dL over a fewdays until compatible blood arrived from South Africa.Two units of group O, D–, hrB–, E–, S– RBCs weretransfused and the Hb increased to 9.6 g/dL, where itremained stable until the patient was discharged.

To summarize this patient’s transfusion manage-ment, multiple alloantibodies were identified. Initially,time was available, which permitted performing an invitro assay (MMA) to assist in predicting the clinicalsignificance of the patient’s antibodies. Her clinicalcondition dictated that transfusion be performed. Thisis important—transfusion should never be withheldfrom a patient with a clinical need based on a serologicincompatibility. Until compatible units could belocated, D+ but otherwise-compatible RBCs weretransfused (which were eventually destroyed by theformation of anti-D), followed by transfusion ofincompatible D–, E–, S–, hrB+ RBCs (also eventuallydestroyed via extravascular mechanisms). Both sets oftransfusions permitted proper clinical management ofthe patient without serious morbidity and mortalitythat could have occurred while awaiting arrival ofcompatible blood.

While the above case illustrates how one maytransfuse against different “incompatibilities” untilcompatible blood arrives, this approach may not workwell in every clinical situation. One such scenario tokeep in mind is the patient with sickle cell disease whois developing a progressively more severe anemia witheach transfusion. The following case illustrates thispoint.

A 21-year-old woman with sickle cell anemia wasadmitted with pneumonia in December. Her medicalhistory is significant for Burkitt’s lymphoma,

successfully treated,at age 15,and pregnancy,at age 19.RBC alloantibodies previously identified included anti-Fya, -S, and -Leb. She received two units of group O, D–,Fya–, S–, Leb– RBCs and was discharged. Several dayslater, she was readmitted with declining Hb (4.5 g/dL).She was transfused with one unit of group O, D–, Fya–,S–, Leb– RBCs, but her Hb declined the next day to 3.1g/dL. Additional antibodies could not be identified asan explanation for the apparent HTR. While one mayfeel the need to continue transfusing this patient, thiscase illustrates many of the features of the “sickle cellhemolytic transfusion reaction,” i.e., manifestation of adelayed HTR, development of a more severe anemiaafter transfusion than was present before, andsubsequent transfusions that further exacerbateanemia that may become life-threatening; both RBCalloantibodies and autoantibodies can be present;serologic studies may not explain the HTR; andadministration of corticosteroids may lead to a gradualimprovement in some patients’ Hb.18–20 This patient’sHb improved to 7.8 g/dL after receiving 60 mg/day ofprednisone without additional transfusion. In a casereport described by Petz et al.,20 their patient withsickle cell anemia and anti-E, -C, -K, -S, -Fya, and -Jkb

received a total of 15 units of RBCs and her Hctdropped from 13.2% to 9.3%. Prednisone (60 mg/day)was started, two units of RBCs were transfused, and theHct increased to 22.4% prior to discharge 1-week later.While this patient responded to steroid medication,consideration can be given to also treating with IVIG to prevent RBC destruction.21

SummaryAs with any transfusion, the risks must be weighed

against the benefits for each individual patient prior totransfusion. Transfusion should not be withheld fromthe patient with an urgent, life-threatening clinicalneed on the basis of serologic test results. If incom-patible blood must be transfused, it is important forclinical care staff and laboratory staff to communicatefrequently to determine which RBC units should beselected for transfusion, when the transfusion will beadministered, and how the patient will be monitored.

References1. Oberman HA. The history of transfusion

medicine. In: Petz LD, Swisher SN, Kleinman S,Spence RK, Strauss RG. Clinical practice oftransfusion medicine. 3rd ed. New York: ChurchillLivingstone, 1996:11-32.

REMEMBER: THE PASSWORD IS “2000” For www.redcross.org/pubs/immuno

Now, as a subscriber, you can enter the password, 2000, and access the back issues. That meanscover to cover! You will receive every article in total, every letter to the editor, every review, every ad, everynotice, and every literature review! All of the other services will continue to be available on the Web site,including sending a letter to the editor, subscribing with a credit card on the secure order site, performing aliterature search, reviewing instructions for authors, and linking to other important sites. Log on now to seethis great service!


G. MENY

2. Hoffman R, Benz EJ Jr, Shattil SJ, et al., eds.Hematology basic principles and practice. 3rd ed.New York: Churchill Livingstone, 2000:367-75.

3. Weiskopf RB, Kramer JH, Viele M, et al. Acutesevere isovolemic anemia impairs cognitivefunction and memory in humans. Anesthesiology2000;92:1646-52.

4. Carson JL, Duff A, Poses RM, et al. Effect ofanaemia and cardiovascular disease on surgicalmortality and morbidity. Lancet 1996;348:1055-60.

5. Hebert PC, Wells G, Blajchman MA, et al. Amulticenter, randomized,controlled clinical trial oftransfusion requirements in critical care. N Engl JMed 1999;340:409-17.

6. Popovsky MA, Audet AM, Andrzejewski C.Transfusion-associated circulatory overload inorthopedic surgery patients: A multi-institutionalstudy. Immunohematology 1996;12:87-9.

7. Popovsky MA. Circulatory overload. In: PopovskyMA, ed. Transfusion reactions. 2nd ed. Bethesda,MD: American Association of Blood Banks, 2001:255-60.

8. Reid ME, Lomas-Francis C. The blood groupantigen factsbook. 2nd ed. San Diego: AcademicPress, 2003.

9. International Committee for Standardization inHematology. Recommended method for radio-isotope red-cell survival studies. Br J Haematol1980;45:659-66.

10. Mollison PL. Survival in vivo as a test for red cellcompatibility. Haematologia 1972;6:139-45.

11. Davenport RD.Hemolytic transfusion reactions. In:Popovsky MA, ed. Transfusion reactions. 2nd ed.Bethesda, MD: American Association of BloodBanks, 2001:1-44.

12. Mechanisms of immune hemolysis. In: Petz LD,Garratty G. Immune hemolytic anemias. 2nd ed.Philadelphia: Churchill Livingstone, 2004:133-65.

13. Langley GR, Owen JA, Padanyi R. The effect ofblood transfusion on serum haptoglobin. Br JHaematol 1962;8:392-9.

14. The diagnosis of hemolytic anemias. In: Petz LD,Garratty G. Immune hemolytic anemias. 2nd ed.Philadelphia: Churchill Livingstone, 2004:33-60.

15. Trainor LD, Solomon HM. Detecting myoglobin-uria: a low-tech analysis. Lab Med 1997;28:569-71.

16. Jacobs DS, Garg U, Oxley DK. Chemistry. In: JacobsDS, Oxley DK, DeMott WR, eds. Laboratory testhandbook.5th ed.Hudson,OH:Lexi-Comp,75-302.

17. Petz LD, Calhoun L, Shulman IA, et al. The sicklecell hemolytic transfusion reaction syndrome.Transfusion 1997;37:382-92.

18. Cullis JO, Win N, Dudley JM, Kaye T. Post-transfusion hyperhaemolysis in a patient withsickle cell disease: use of steroids and intravenousimmunoglobulin to prevent further red celldestruction.Vox Sang 1995;69:355-7.

19. King KE, Shirley RS, Lankiewicz MW, Young-Ramsaran J, Ness PM. Delayed hemolytictransfusion reactions in sickle cell disease:simultaneous destruction of recipients’ red cells.Transfusion 1997;37:376-81.

20. Petz LD,Calhoun L,Shulman IA, Johnson C,HerronRM. The sickle cell hemolytic transfusion reactionsyndrome. Transfusion 1997;37:382-92.

21. Cullis JO, Win N, Dudley JM, Kaye T. Post-transfusion hyperhaemolysis in a patient withsickle cell disease: use of steroids and intravenousimmunoglobulin to prevent further red celldestruction.Vox Sang 1995;69:355-7.

Geralyn Meny, MD, American Red Cross BloodServices, 700 Spring Garden Street, Philadelphia, PA19123


The presence of hemolysis is usually easilyconfirmed by simple tests such as reticulocyte count,review of the peripheral blood film, serum bilirubin(direct and indirect), LDH, and haptoglobin.1 When ithas been established that the patient has a hemolyticanemia, the cause of the hemolysis should be soughtnext. The DAT should be performed on the RBCs ofevery patient in whom the presence of hemolysis hasbeen established, and a positive DAT in a patient withhemolytic anemia does, of course, indicate that themost likely diagnosis is one of the immune hemolyticanemias (IHAs).2 However, there are numerous causesof IHA and the clinical manifestations, prognosis, andtherapy vary so that defining the precise diagnosis is ofconsiderable importance.

Classification of IHAsIn evaluating a patient with IHA, one should first

consider a differential diagnosis as outlined in Table 1.Although a definitive diagnosis rests with serologicstudies, some clinical and routine laboratory pro-cedures are sufficiently distinctive as to stronglysuggest the type of IHA that is present.

Distinctive Clinical Signs and RoutineLaboratory Procedures

Association with exposure to coldA history of acrocyanosis and/or hemoglobinuria

on exposure to cold in an elderly patient with anacquired hemolytic anemia strongly suggests adiagnosis of cold agglutinin syndrome (CAS). However,these manifestations are absent in a majority of patientswith CAS even though they were often emphasized inthe early medical literature.

Although one might assume that paroxysmal coldhemoglobinuria (PCH) is commonly precipitated byexposure to cold, this is only occasionally true. IndeedWolach et al.3 pointed out that the most common formof PCH is the transient type, secondary to infection

(e.g., in childhood); is rarely paroxysmal; is onlyoccasionally clearly precipitated by cold; and is notinvariably expressed as hemoglobinuria, although thelatter finding is very common.4

AutoagglutinationAutoagglutination is a finding that may be noted by

technologists in all sections of the laboratory, not justthose in the blood transfusion or immunohematologylaboratories. Indeed, cold autoagglutinins that reactstrongly at room temperature cause such strikingfindings that they are difficult to ignore. Autoagglu-tination visible to the naked eye occurring at roomtemperature is characteristic of the CAS, but may alsobe noted in about one-third of patients with warmautoimmune hemolytic anemia (WAIHA).1 Although

Review: evaluation of patientswith immune hemolysisLAWRENCE D. PETZ

Table 1. Differential diagnosis of immune hemolytic anemias

Autoimmune hemolytic anemias (AIHA)

Warm-antibody AIHAIdiopathicSecondary (e.g., chronic lymphocytic leukemia, lymphomas,systemic lupus erythematosus)

Cold agglutinin syndromeIdiopathicSecondary

Nonmalignant disorders (e.g., mycoplasma pneumoniaeinfection, infectious mononucleosis, other virus infections)Malignant disorders (e.g., lymphoproliferative disorders)

Combined cold and warm AIHA

Paroxysmal cold hemoglobinuriaIdiopathicSecondary

Viral syndromesSyphilis

Atypical AIHAAIHA with a negative direct antiglobulin testWarm-antibody AIHA caused by IgM or IgA autoantibodies

Drug-induced immune hemolytic anemiaDrug-related antibody identifiableDrug-induced AIHA

Alloantibody-induced immune hemolytic anemiaHemolytic transfusion reactionsHemolytic disease of the fetus and newborn

Note: Tables 1, 2, 4, 5, and 6 were taken from Petz LD, Garraty G. Immune hemolyticanemias. 2nd ed. Philadelphia: Churchhill Livingstone, 2004.1


L.D. PETZ

the autoagglutination caused by cold agglutinins isoften 2+ to 4+, it almost always completely dispersesafter a few minutes of incubation at 37°C, whereas thatcaused by warm autoantibodies is usually muchweaker and will not disperse at 37°C. If the bloodsample has been obtained from a patient known tohave hemolytic anemia, such simple observations offeran important clue to the correct diagnosis.

However,a common error is the over-interpretationof cold agglutination. Many cold antibodies arereactive at room temperature but are clinically benign,albeit somewhat a nuisance in the laboratory. Theserologic criteria for distinguishing clinically benigncold agglutinins from pathologic cold agglutininscapable of causing a CAS are discussed later.

Drug ingestionA temporal history of drug ingestion may suggest

the etiology of the patient’s IHA. A critical aspect ofevaluation of a patient with IHA is the elicitation of ahistory of drug ingestion which, in some instances,mayhave occurred a week or more prior to the onset ofhemolysis or be a single dose given for surgery (e.g.,cefotetan). Knowledge of the drugs that have beenimplicated as a cause of drug-induced IHA is essential.

Many drug-induced IHAs can be distinguished fromautoimmune hemolytic anemia (AIHA) by laboratoryfindings, i.e., the demonstration of a drug-dependentRBC antibody. However, the administration of somedrugs causes hemolytic anemia that is serologicallyindistinguishable from cases of idiopathic WAIHA.Such cases are appropriately termed drug-inducedAIHA, whereas those cases wherein a drug-dependentantibody can be identified are termed drug-inducedIHA.

Alloantibody-induced IHAAlloantibody-induced hemolytic anemias include

HDN and hemolytic transfusion reactions. The clinicalsetting usually strongly suggests these diagnoses.Although autoantibody-induced HDN can occur as aresult of transplacental passage of a mother’s IgG warmautoantibody, this is very unlikely unless the motherhas obvious and quite severe WAIHA.

Also, when hemolysis occurs in the immediateaftermath of a RBC transfusion, the diagnosis of anacute hemolytic transfusion reaction is quite evident.However, distinguishing a delayed hemolytic trans-fusion reaction from AIHA is, on occasion, difficult.5–7

Hemoglobinemia and hemoglobinuriaHemoglobinuria (Hb in the urine) is far less

common than hematuria (RBCs in the urine), and acommon clinical error is the assumption that apatient’s red urine is caused by hematuria. It should beremembered that hemoglobinuria, associated withhemolytic anemia, cannot occur without hemoglobin-emia. If red urine is present without hemoglobinemia,it should be suspected that the cause is hematuria andnot hemolysis. The presence of hemoglobinemia andhemoglobinuria should alert the clinician to a specificgroup of diagnoses and, when considered inassociation with the clinical setting, often makes thespecific diagnosis evident (Table 2). Probably the mostcommon associated diagnoses are hemolytic trans-fusion reactions and severe acute WAIHA, althoughhemoglobinuria may occur in patients with CAS,especially after exposure to cold. Also, drug-inducedIHAs caused by cefotetan and ceftriaxone arecommonly associated with hemoglobinemia andhemoglobinuria.

Hemoglobinuria and hemoglobinemia are muchmore common in children, and both warm and coldAIHA can be the cause. In a child, PCH should bestrongly suspected and a Donath-Landsteiner (DL) testshould be performed whenever an acute severehemolytic anemia occurs with hemoglobinemia andhemoglobinuria. Indeed, hemoglobinuria is a commonpresenting manifestation of PCH4 and thereforebecomes an important diagnostic clue.

Table 2. Causes of hemoglobinuria

Acute HemoglobinuriaIncompatible blood transfusionTransfusion of damaged blood (overheating or freezing, bacterial

contamination, pump-oxygenation)Drugs and chemical agents (immune or nonimmune mechanisms)Paroxysmal cold hemoglobinuriaAcute severe warm-antibody AIHAClostridium perfringens infectionMalaria (blackwater fever)Bartonellosis, babesiosis, leptospirosis, toxoplasmosisPeritoneal hemorrhageSevere hypophosphatemia Snake and spider bitesCold agglutinin syndrome*March hemoglobinuriaMicroangiopathic hemolytic anemiaHypotonic bladder irrigation during prostatic surgeryMistaken intravenous administration of water

Chronic HemoglobinuriaParoxysmal nocturnal hemoglobinuria**Prosthetic cardiovascular materials

*Chronic low grade intravascular hemolysis is common, with acute hemoglobinuriaresulting from exposure to cold**Characteristically associated with intermittent episodes of grossly evidenthemoglobinuria


Review: evaluation of immune hemolysis

RBC morphology and erythrophagocytosisOne easily performed and very valuable test for

determining the specific diagnosis in a patient withhemolytic anemia is examination of the peripheralblood film. RBC morphology often strongly suggests aspecific diagnosis or a limited number of diagnosticpossibilities. Spherocytosis may be a prominent featurein IHAs, especially WAIHA, ABO HDN, hemolytictransfusion reactions, and some instances of drug-induced hemolysis.

Less well appreciated is the fact that RBCadherence and erythrophagocytosis by neutrophils is aprominent finding in PCH, but is seen rarely in otherforms of IHA. Erythrophagocytosis is rarely observedin the peripheral blood film of WAIHA and when it isobserved, monocytes, not neutrophils, are more ofteninvolved.

Association with Mycoplasma pneumoniaeIf a patient with Mycoplasma pneumoniae

infection develops AIHA, the diagnosis of a CAS mustbe strongly suspected, since AIHA in this setting isalmost always caused by cold agglutinins.

Association with underlying disorders (secondary IHAs)

Patients should be evaluated for certain diseaseentities, since approximately half of AIHAs areassociated with an underlying disorder (Table 3).AIHAs are classified as secondary for any of severalreasons. One reason is the association of AIHA with anunderlying disease with a frequency greater than canbe explained by chance alone, as in chronic lympho-cytic leukemia or systemic lupus erythematosus.

Another criterion for categorizing a given case ofAIHA as secondary is the reversal of the hemolyticanemia simultaneously with the correction of theassociated disease. Ovarian tumors are a good ex-ample;well-documented cases of cure of the AIHA aftersurgical removal of the tumor have been reported.8–11

Similarly, AIHA in association with ulcerative colitisalmost invariably goes into remission after colectomy,even when hemolysis is refractory to other therapeuticapproaches.12–14

Still another reason for suspecting a relationshipbetween the occurrence of AIHA and an associateddisease consists of evidence of immunologic aberrationas part of the underlying disorder, especially if theassociated disease is thought to have an autoimmunepathogenesis. In general, the evidence for a

relationship between immunologic disorders, includ-ing immune deficiency states, and AIHA is strong.15–19

Although the pathogenetic basis for the associationbetween cytopenias and congenital immune deficiencyis unclear, defects in T-cell regulation, cytokine defects,abnormal apoptosis, and abnormal production ofimmunoglobulins with autoimmune features arepotential mechanisms.15

Laboratory Diagnosis of IHAsEven in the presence of valuable clinical clues that

may suggest a specific diagnosis, the confirmation ofthe precise diagnosis of the type of IHA presentdepends on the laboratory. The serologic tests to beperformed determine whether the patient’s RBCs arecoated with IgG, complement components, or both.The performance of the DAT supplies such infor-mation. Further tests must be performed to determinethe characteristics of the antibodies in the patient’sserum and in a RBC eluate.

Significance of the DAT in the diagnosis of IHAs The DAT, using polyspecific and monospecific

antiglobulin reagents, provides useful information inthe evaluation of a patient with IHA. However, theresults must always be interpreted in conjunction withclinical and other laboratory data to avoid erroneousconclusions. It must be remembered that a positiveDAT occurs in situations other than IHAs. A positiveDAT does not necessarily indicate the presence of

Table 3. Secondary autoimmune hemolytic anemias

Ovarian tumorsUlcerative colitis

Chronic lymphocytic leukemiaHodgkins disease

Non-Hodgkins lymphomaOther lymphoproliferative disordersSystemic lupus erythematosus (SLE)

Antiphospholipid syndromeCollagen disorders other than SLE

ThymomaCarcinomas

Primary immunodeficiency diseasesAutoimmune lymphoproliferative syndrome (ALPS)

Miscellaneous disorders

AIHA associated with infectious agentsMycoplasma pneumoniae

Infectious mononucleosis (Epstein-Barr Virus)Cytomegalovirus (CMV)

Human immunodeficiency virus (HIV)Varicella (chickenpox)

RubellaParvovirus B19

HepatitisMalaria and other blood parasites

Bacterial infections


L.D. PETZ

autoantibody; furthermore, even if autoantibody ispresent, the patient may or may not have a hemolyticanemia. Thus, an independent assessment must bemade to determine the presence or absence ofhemolytic anemia, and the role of the DAT is to aid inthe evaluation of the etiology of hemolysis whenpresent. The results of DATs in patients with varioustypes of IHAs are indicated in Table 4.

AIHA associated with warm IgM autoantibodieswithout the presence of IgG occurs on rare occasions.However, even with potent antisera, RBC-bound IgM isdifficult to detect with the antiglobulin test.1,20,21

Fortunately, IgM antibodies that cause IHA character-istically fix complement, which is much more readilydetected. IgA antibodies only infrequently play a rolein RBC sensitization, and in such cases other immuneglobulins and/or complement components are almostalways, although not invariably, found on the cellsurface as well. The clinical and hematologic featuresof WAIHA associated only with IgA autoantibodies arevery similar to AIHA associated with warm IgGautoantibodies.

The AIHA serum screen testThe “AIHA screen” is an elaboration of the routine

antibody screen performed for pretransfusion testing.In summary, the patient’s serum is tested against a poolof untreated and enzyme (e.g., ficin)-treated allogeneic

group O RBCs and autologous RBCs at 20°C (roomtemperature can be used) and 37°C (all stages strictlyat 37°C). A duplicate set of tubes is used with addedfresh complement (pooled normal sera) at optimal pH(6.5 to 6.8) for lysis. After incubation, the tubes areinspected for hemolysis, agglutination, and sensitiz-ation (antiglobulin test using polyspecific antiglobulinserum).

The results of this screen usually indicate whetherone is dealing with a cold autoagglutinin, a “warm”autoantibody, or possibly a combination of both. Otherpoints of interest are whether there is a hemolysinpresent and whether it is a “cold”or “warm”hemolysin.If agglutination occurs at 20°C, a cold agglutinin titerand thermal amplitude should be performed. Resultsof the AIHA serum screen in WAIHA and in CAS areindicated in Tables 5 and 6, respectively.

With the results of the DAT, AIHA screen, andpossibly a cold agglutinin titer/thermal amplitude, onegenerally has a good idea of whether the diagnosis isWAIHA or CAS. Sometimes the patient’s history andresults of the DAT/AIHA screen will lead to furthertests, such as the DL test for PCH, or detection of drug-dependent antibodies.

Characteristic serology of WAIHAAutoantibodies causing WAIHA are usually IgG but

can be IgM or IgA. These proteins can be presenttogether; it is rare for only IgM or IgA to be the causeof AIHA. Autoantibodies are usually only found in theserum when all autoantigens are saturated, apparentlybecause the patient’s RBCs are adsorbing warmautoantibodies continuously. Only about 60 percent ofpatients’ sera will react with saline-suspended RBCs,

Table 4. Typical DAT results in patients with immune hemolytic anemias

Causes of hemolytic anemia Anti-IgG Anti-C3*

Warm-antibody AIHA 67% + +20% + 013% 0 +

Warm-antibody AIHA associated with systemic lupus erythematosus + +

Cold agglutinin syndrome (100%)* 0 +

Paroxysmal cold hemoglobinuria (100%) 0 +

Drug-induced immune hemolytic anemiasDrug-dependent antibodiesPenicillin and first-generation cephalosporins + (+)†

Second- and third-generation cephalosporins + +Associated with “Immune Complex Mechanism” + +

Drug-independent antibodies + (+)

Drug-induced nonimmunologic adsorption of proteins* + (+)

Hemolytic disease of fetus/newborn + 0

Hemolytic transfusion reactions + (+)

*IgG +C3 were both detected on the RBCs of one patient with CAS because thepatient was on methyldopa therapy and made IgG autoantibodies†(+) = sometimes positive

Table 5. Results of serum screening in 244 patients with warm-antibodyAIHA

Acidified serum Untreated plus acidified

Results serum complement

Serologic reactions (% Positive reactions) (% Positive reactions)

20°C Untreated RBCsLysis 0.4 0.8Agglutination 34.8 34.8

20°C Enzyme-treated RBCsLysis 1.6 2.5Agglutination 78.6 78.6

37°C Untreated RBCsLysis 0.4 0.4Agglutination 4.9 4.9Indirect anti-globulin test 57.4 57.4




but a higher percentage will react in the presence ofpotentiators (e.g., PEG) or enzyme-treated RBCs.

Some sera contain warm hemolysins, which reactoptimally at 37°C but may react at 20°C. The hemolysisis enhanced by adding fresh complement at pH 6.5 to6.8 to the patients’ sera. von dem Borne et al.22 showedthat such hemolysins were usually IgM autoantibodies;they were maximally reactive at a pH of 6.5 and 7 of 11(64%) reacted optimally at 30°C; 4 of 11 reactedoptimally at 37°C. About one-third of sera frompatients with WAIHA contain IgM cold autoagglutininsthat can react quite strongly at 20°C (or roomtemperature), but have a normal cold agglutinin titer at4°C and do not react at 30°C. These may be naturallyoccurring cold antibodies that become boosted (e.g.,raised thermal amplitude) during the pathogenicautoimmune response.

A diagnosis can usually be reached on the basis ofthe serologic tests described thus far. In spite of theseemingly complicated nature of the foregoing, theusual or “typical” essential diagnostic tests that lead toa reasonably confident diagnosis of WAIHA may be verysimply summarized as follows: (a) the presence of anacquired hemolytic anemia, (b) a positive DAT, and (c)an unexpected antibody in the serum and eluate thatreacts optimally at 37°C. The antibody usually reactswith all normal erythrocytes but, in some cases, it canreadily be shown to react preferentially with antigenson the patient’s own RBCs.

Characterization of antibodies in CAS The mere presence of cold autoagglutinins is not

diagnostic of CAS, and when they are encountered, thetask is to determine whether the patient has clinically

insignificant, albeit abnormal, cold agglutinins; hasWAIHA with an associated but probably insignificantelevation of cold agglutinin titer and/or thermalamplitude; or has CAS. A rather common diagnosticerror is over-diagnosis of CAS in patients who havebenign cold antibodies and pathogenic warmautoantibodies. Also, in rare patients, characteristicfindings of both WAIHA and CAS occur simultaneously.

A diagnosis of CAS must be considered in allpatients with acquired hemolytic anemia who have apositive DAT using anti-C3 and a negative DAT usinganti-IgG. A practical initial serum screening procedureis to test the ability of the patient’s serum to agglutinatesaline-suspended normal RBCs at 20°C (or roomtemperature) after incubation for 30 to 60 minutes. Ifthis screening test is negative, a CAS is extremelyunlikely; if positive, further studies are necessary todetermine the thermal amplitude of the antibody.

When CAS appears to be a possible diagnosis onthe basis of the preceding evaluation, studies of thethermal range of reactivity of the antibody in saline andalbumin are indicated. It is also convenient to simul-taneously determine possible Ii blood group specificityof the antibody. The titer of the cold agglutinin in CASis invariably highest at 4°C and progressively decreasesat higher temperatures. Of particular note are thereactions at 30°C and 37°C. If the reactions at 30°C arepositive in saline or albumin, the antibody may well beof pathogenic significance, i.e., it may be causing shortRBC survival in vivo. If the reactions at 37°C are alsopositive in the presence of albumin (as is true in about68% of patients) or even when albumin is not present(only about 7% of patients), the antibody will causeproblems in compatibility testing.

Using clinical information, the results of the DAT,and the preceding screening tests, a reasonablyconfident assessment of the presence or absence ofCAS may be made. CAS may be diagnosed if thefollowing are present: (1) clinical evidence of acquiredhemolytic anemia, (2) a positive DAT caused bysensitization with C3, (3) a negative DAT using anti-IgG,(4) the presence of a cold autoagglutinin withreactivity up to at least 30°C in saline or albumin, and(5) a cold agglutinin titer (at 4°C) ≥ 256, except inexceptional cases. An alternative diagnosis must besought for patients who do not satisfy all these criteria.

Patients who have warm and cold autoantibodiesApproximately one-third of WAIHA patients have

cold agglutinins that can react quite strongly at room

Table 6. Results of serum screening in 57 patients with CAS

Acidified serum Untreated plus acidified

Results serum complement

Serologic reactions (% Positive reactions) (% Positive reactions)

20°C Untreated RBCsLysis 2.0 14.3Agglutination 98 98

20°C Enzyme-treated RBCsLysis 24.5 93.8Agglutination 100 100

37°C Untreated RBCsLysis 0 0Agglutination 10.7 10.7Indirect anti-globulin test 5.4 5.4



L.D. PETZ

temperature but have a normal titer (at 4°C) and do notreact at 30°C and 37°C. Such antibodies are probablynot pathogenic. However, some patients with WAIHAhave cold antibodies that react up to 30°C or above andtherefore may be pathogenic. Some of these patientshave IgG and C3 on their RBCs, and their sera containIgG 37°C-reactive antibodies together with high-titer,high-thermal amplitude cold autoagglutinins (i.e., thecombined serology of classical WAIHA and CAS).1,23–28

Other patients have IgG and/or C3 on their RBCs, andtheir sera contain IgG 37°C-reactive antibodiestogether with cold autoagglutinins of normal titer thatreact at 37°C and/or 30°C. There are only threepublished reports that relate to this group, and thereport by Sokol et al.28 does not give any coldagglutinin titers, so some of the patients may belong tothe first group. Shulman et al.27 believe that up to 8percent of WAIHAs may belong to this group.

PCHThe diagnosis of PCH or the exclusion of that

diagnosis in the laboratory is usually considerablyeasier than that of either WAIHA or CAS. The essentiallaboratory test is the DL test. A negative test excludesthe diagnosis of PCH and a positive test is, with rareexceptions (described below), diagnostic of thedisorder.

The autoantibody associated with PCH is termed abiphasic hemolysin, that is to say, it sensitizes RBCs inthe cold but only hemolyzes them when the RBCsreach 37°C. The diagnostic test is the DL test, whereinRBCs are incubated with the patient’s serum at 0°C(e.g.,melting ice) and then moved to 37°C for a furtherincubation. No lysis occurs following the incubation at0°C, and no lysis occurs if the incubation is carried outonly at 37°C. The thermal amplitude of this antibody isusually less than 20°C, that is to say, the antibody willgive a positive DL test only when the initial incubationis < 20°C;stronger results will occur as the temperatureof the initial incubation is lowered. Rare patients havebeen described, in whom the DL test is positive whenthe first incubation phase is as high as 32°C,or their DLantibody sensitized RBCs up to 37°C, as detected bythe IAT.29–31

The autoantibody may sometimes agglutinate RBCsin addition to giving a positive DL test.1 Theagglutination is usually of low titer (< 64) at 4°C and oflow thermal amplitude (< 20°C). The DL antibody isIgG but is usually only detectable by IAT if, followingincubation of the patient’s serum and RBCs at 0°C, the

RBCs are washed with ice-cold saline and ice-coldantiglobulin serum is used. Indeed, Dacie found thatthe IAT was a more sensitive way of demonstratingantibody activity than looking for lysis.32 Suchagglutination tests must be carefully controlledbecause many sera give positive results under thesecircumstances if a broad-spectrum antiglobulin serumis used, due to the presence in human sera of normalincomplete cold antibody. It is important, therefore, tobe sure that monospecific anti-IgG antiglobulin serumis used.

Since PCH is quite rare,one may justifiably questionthe advisability of performing a DL test routinely inpatients with acquired hemolytic anemia. However,one should be liberal with the indications forperformance of the test, since it is simple to performand its inclusion avoids diagnostic errors. The perfor-mance of the test is indicated in any child, any patientwith hemoglobinuria, patients with a history ofhemolysis exacerbated by cold, and in all cases with“atypical” serologic findings. If positive results areobtained in the DL test,determination of the specificityof the autoantibody is indicated. In almost all cases ofPCH the antibody has anti-P specificity. Sokol et al.33

found anti-P specificity in 27 of 30 (90%) patients withPCH; specificity was not clearly defined in the otherthree cases. Nevertheless, rare reports of otherspecificities said to be associated with PCH have beenreported. RBCs necessary for determining anti-Pspecificity are rare but, with the assistance of referencelaboratories, specificity testing can be carried out. Itshould be noted that the DL test must be used todetermine specificity.

Cautions regarding the interpretation of the DL test The DL test is essentially diagnostic for PCH, but

one must be cautious when using serum from patientswith CAS. This is true because about 15 percent of serafrom patients with CAS contain monophasic coldhemolysins that will hemolyze untreated RBCs ataround 20°C. Up to 95 percent of such sera will causedirect lysis of enzyme-treated RBCs at 20°C. During theperformance of the DL test, there is a brief period oftime when cells and serum are at room temperatureafter being moved from the ice bath to a 37°C waterbath. In about 15 percent of patients with CAS, falselypositive (weak) DL tests will occur on this basis. Ifenzyme-treated RBCs are used for the DL test, assuggested by some investigators,4 the falsely positiverate is likely to be higher. Therefore, if enzyme-treated



RBCs are used for the DL test, a control for monophasiclysis set up in parallel should be mandatory.

Performing the DL test in patients withhemoglobinemia

It may be impossible to determine if in vitrohemolysis has occurred in the DL test if the patient’sserum is red because of marked hemoglobinemia. Inthis case, a simple procedure is to perform the coldphase of the DL test using the patient’s serum,but afterincubation at 4°C, carefully replace the patient’s serumwith fresh normal serum before moving to the 37°Cphase of the test. As a control, a similar tube can bekept at 4°C, or a similar test can be performed with P-negative RBCs, if available, and if the specificity of theDL antibody is anti-P. Another simple technical aid is tocompare the size of the RBC buttons followingcentrifugation. If these approaches fail, then a “coldIAT” can be performed.

Differentiating Delayed HemolyticTransfusion Reactions From AutoimmuneHemolytic Anemia

Delayed hemolytic transfusion reactions (DHTRs)are a recognized risk of blood transfusion. The reactionis caused by the reappearance of an antibody,presumably first stimulated by pregnancy or a previoustransfusion. Unlike immediate transfusion reactions,which are usually caused by human error,34–38 delayedreactions are usually not avoidable. Since hemolysis isdelayed in onset (typically 3 to 14 days aftertransfusion), the relationship of hemolytic anemia toprior transfusion may not be suspected, and a diagnosisof AIHA may seem more appropriate. Further,laboratory tests are likely to reveal the presence of apositive DAT and IAT,spherocytosis,and reticulocytosis.If multiple alloantibodies or an alloantibody against ahigh-frequency antigen is formed, the findings may bedifficult to differentiate from AIHA. The diagnosticproblem is compounded by the fact that, in some cases,AIHA may actually develop as a consequence of bloodtransfusion.1 Indeed, therapy with corticosteroids forsuspected AIHA has been instituted5,39 orcontemplated40,41 in some reported cases of DHTRsbefore the correct diagnosis was made.

Diagnostic AidsIf a patient has been transfused within recent

weeks, careful evaluation is needed to distinguish

between a DHTR and AIHA. The following measureswill afford important clues.

Comparison of DAT and IATA simple observation that may yield valuable

information is the comparison of the strength of theDAT and the IAT. In WAIHA, the DAT is almost alwaysstronger than the IAT. It appears that autoantibody islargely adsorbed onto the patient’s RBCs, and onlywhen the RBCs are heavily coated does one find a largeamount of antibody in the serum. In contrast, stronglyreactive alloantibodies may be present in a patient’sserum, but this finding cannot result in a stronglypositive DAT unless large numbers of transfused RBCsof appropriate antigenic type are present. Thus, thepresence of a weakly positive DAT in association witha strongly positive IAT is presumptive evidence for thepresence of an alloantibody. These findings aretherefore highly suggestive of a DHTR. Further, even ifthe DAT is strongly positive at the time of initialevaluation, it may soon become weaker in subsequenttests as a result of the destruction of the transfusedRBCs. This is true even though the DAT may remainweakly positive for several months following a DHTR.In contrast, a rapid diminution in the strength of theDAT would not be expected in AIHA except as a resultof treatment, as with corticosteroids, immuno-suppressive drugs, or splenectomy.

Antibody specificityAn important means of differentiating AIHA from a

DHTR relates to the specificity of the antibody(ies)present in the serum and in a RBC eluate. Someantibodies that commonly cause DHTRs have not beenfound or have been reported only rarely asautoantibodies in AIHA. Examples are anti-K and anti-Fya, which are frequently encountered in publishedcases of DHTRs.6,42–45 Many autoantibodies in warmantibody AIHA demonstrate specificity within the Rhsystem but, even here, a distinction betweenautoantibodies and alloantibodies with Rh specificity isoften possible. Whereas alloantibodies demonstratetruly specific reactions and give clearly negativereactions with cells lacking the appropriate antigen,autoantibodies commonly demonstrate “relativespecificity.” That is,autoantibodies that are described ashaving specificity within the Rh system react morestrongly or to a higher titer against RBCs bearing aparticular Rh antigen, but they will nevertheless reactwith RBCs lacking the antigen. Thus,a truly specific Rh


L.D. PETZ

antibody strongly suggests that it is an alloantibody,whereas an antibody demonstrating “relativespecificity” is characteristic of autoantibody.

A further clue to the differentiation ofautoantibodies and alloantibodies having Rh specificityis the specificity itself. That is, anti-e is a rare cause of aDHTR,6 but it is the most frequently describedautoantibody specificity. In contrast, anti-E is the mostcommon Rh alloantibody responsible for DHTRs,6,44

but it is a relatively unusual autoantibody. Again, ifpretransfusion RBCs are still available, it will bepossible to use them to distinguish alloantibody fromautoantibody with certainty.

If the antibody shows a defined specificity, thepatient’s RBCs should be tested for the relevantantigen. If it is an autoantibody, the patient’s RBCsshould possess the antigen. It is not always easy todetermine the patient’s phenotype, as DAT+ RBCs aredifficult to phenotype and transfused RBCs may bepresent. If transfused RBCs are present, severalmethods are available for determining the phenotypeof only the patient’s RBCs:

1) One method depends on separating out theyounger RBCs (e.g., reticulocytes are presumedto be the patient’s own RBCs).46–50

2) Another method, which may be more reliablethan the reticulocyte method, utilizes flowcytometry.51–53

3) A recent approach is to use DNA typing.54–56

Additional approachesIf the recipient’s DAT was known to be negative

prior to transfusion, or if the recipient’s RBCs are stillavailable for the performance of the DAT and it isdemonstrated to be negative, this can be valuableinformation. An abrupt change in the DAT fromnegative to positive is strong evidence that the patienthas a HTR rather than AIHA.

If it is possible to test the donor units of blood, thisprocedure may be of significance if an antibody havingspecificity that could be that of either an alloantibodyor an autoantibody is detected, such as, for example,anti-E. If by chance, none of the donor units containedthe E antigen, AIHA or an alloantibody-induced HTRcaused by an undetected antibody must be considered.

Thus, with careful serologic testing, it is possible todistinguish a DHTR from AIHA in almost all cases.However, if pretransfusion RBCs are not available (as,unfortunately, is usually the case), and if the patient hasan alloantibody or a mixture of alloantibodies with a

broad range of reactivity, the distinction may bedifficult.

A final note of caution is that the IHA may beneither a DHTR nor AIHA, but instead may be drug-induced IHA. In particular, cephalosporin drugs arefrequently used in association with surgical proceduresthat may require transfusion. In this setting, abruptonset of a positive DAT and hemolysis may bemisinterpreted as a DHTR or the sudden onset of AIHA.

References1. Petz LD, Garratty G. Immune hemolytic anemias.

2nd ed. Philadelphia: Churchill Livingstone, 2004.2. Kaplan HS, Garratty G. Predictive value of direct

antiglobulin test results. Diagn Med 1985;8:29-33.3. Wolach B, Heddle N, Barr RD, Zipursky A, Pai KR,

Blajchman MA. Transient Donath-Landsteiner hae-molytic anaemia. Br J Haematol 1981;48:425-34.

4. Heddle NM. Acute paroxysmal cold hemoglobin-uria. Transfus Med Rev 1989;3:219-29.

5. Croucher BEE, Crookston MC, Crookston JH.Delayed haemolytic transfusion reactions simu-lating auto-immune haemolytic anaemia. Vox Sang1967;12:32.

6. Croucher BEE. Differential diagnosis of delayedhemolytic transfusion reaction. In: Bell CA, editor.A Seminar on Laboratory Management ofHemolysis.Washington, DC: American Associationof Blood Banks, 1979:151-60.

7. Worlledge SM. The interpretation of a positivedirect antiglobulin test. Br J Haematol 1978;39:157-62.

8. Barry KG, Crosby WH. Auto-immune hemolyticanemia arrested by removal of an ovarianteratoma: review of the literature and report of acase.Ann Intern Med 1957;47:1002-7.

9. Dawson MA, Talbert W, Yarbro JW. Hemolyticanemia associated with an ovarian tumor. Am JMed 1971;50:552-6.

10. De Bruyere M, Sokal G, Devoitille JM, Fauchet-Dutrieux MC, De Spa V. Autoimmune haemolyticanaemia and ovarian tumour. Br J Haematol 1971;20:83-94.

11. Payne D, Muss HB, Homesley HD, Jobson VW, BairdFG. Autoimmune hemolytic anemia and ovariandermoid cysts: case report and review of theliterature. Cancer 1981;48:721-4.

12. Shashaty GG, Rath CE, Britt EJ. Autoimmunehemolytic anemia associated with ulcerativecolitis. Am J Hematol 1977;3:199-208.



13. Ramakrishna R, Manoharan A. Auto-immunehaemolytic anaemia in ulcerative colitis. ActaHaematol 1994;91:99-102.

14. Arner O, Brohult J, Engstedt L, Karlson R, SallstromT. Autoimmune hemolytic anemia in ulcerativecolitis, cured by colectomy. Report of a case.ActaMed Scand 1971;189:275-7.

15. Cunningham-Rundles C. Hematologic compli-cations of primary immune deficiencies. BloodRev 2002;16:61-4.

16. Cunningham-Rundles C, Bodian C. Common vari-able immunodeficiency: clinical and immuno-logical features of 248 patients. Clin Immunol1999;92:34-48.

17. Oyefara BI, Kim HC, Danziger RN, Carroll M,Greene JM, Douglas SD. Autoimmune hemolyticanemia in chronic mucocutaneous candidiasis.Clin Diagn Lab Immunol 1994;1:38-43.

18. Cummins L, Spearer W, Stevenson S, et al. Wiskott-Aldrich syndrome:clinical, immunologic,and path-ologic observations. Am J Dis Child 1959;98:579.

19. Pinchas-Hamiel O, Mandel M,Engelberg S,PasswellJH. Immune hemolytic anemia, thrombocytopeniaand liver disease in a patient with DiGeorgesyndrome. Isr J Med Sci 1994;30:530-2.

20. Hsu TC, Rosenfield RE, Burkart P, Wong KY,Kochwa S. Instrumented PVP-augmented anti-globulin tests. II. Evaluation of acquired hemolyticanemia. Vox Sang 1974;26:305-25.

21. Garratty G, Petz LD. An evaluation of commercialantiglobulin sera with particular reference to theiranticomplement properties. Transfusion 1971;11:79-88.

22. Von dem Borne AE, Engelfriet CP, Beckers D, VanDer Kort-Henkes G, Van Der Giessen M, VanLoghem JJ. Autoimmune haemolytic anaemias. II.Warm haemolysins—serological and immuno-chemical investigations and 51Cr studies. Clin ExpImmunol 1969;4:333-43.

23. Sokol RJ, Hewitt S, Stamps BK. Autoimmunehaemolysis: an 18-year study of 865 cases referredto a regional transfusion centre. Br Med J (Clin ResEd) 1981;282:2023-7.

24. Crookston JH. Hemolytic anemia with IgG andIgM autoantibodies and alloantibodies. ArchIntern Med 1975;135:1314.

25. Freedman J, Lim FC, Musclow E, Fernandes B,Rother I. Autoimmune hemolytic anemia withconcurrence of warm and cold red cell autoanti-bodies and a warm hemolysin. Transfusion 1985;25:368-72.

26. Kajii E, Miura Y, Ikemoto S. Characterization ofautoantibodies in mixed-type autoimmune hemo-lytic anemia. Vox Sang 1991;60:45-52.

27. Shulman IA, Branch DR, Nelson JM,Thompson JC,Saxena S, Petz LD. Autoimmune hemolytic anemiawith both cold and warm autoantibodies. JAMA1985;253:1746-8.

28. Sokol RJ, Hewitt S, Stamps BK. Autoimmunehemolysis: mixed warm and cold antibody type.Acta Haematologica 1983;69:266-74.

29. Ries CA, Garratty G, Petz LD, Fudenberg HH.Paroxysmal cold hemoglobinuria: report of a casewith an exceptionally high thermal range Donath-Landsteiner antibody. Blood 1971;38:491-9.

30. Lindgren S, Zimmerman S, Gibbs F, Garratty G. Anunusual Donath-Landsteiner antibody detectableat 37°C by the antiglobulin test. Transfusion 1985;25:142-4.

31. Nordhagen R. Two cases of paroxysmal coldhemo-globinuria with a Donath-Landsteiner anti-body reactive by the indirect antiglobulin testusing anti-IgG. Transfusion 1991;31:190-1.

32. Dacie JV. The haemolytic anaemias. Part II. Auto-immune haemolytic anaemias. 2nd ed. London:J. & A Churchill Ltd., 1962.

33. Sokol RJ, Booker DJ, Stamps R. Erythropoiesis:paroxysmal cold haemoglobinuria: A clinico-pathological study of patients with a positiveDonath-Landsteiner test. Hematol 1999;4:137-64.

34. Myhre BA. Fatalities from blood transfusion. JAMA1980;244:1333-5.

35. Honig CL, Bove JR. Transfusion-associatedfatalities: review of Bureau of Biologics reports1976–1978.Transfusion 1980;20:653-61.

36. Sazama K. Reports of 355 transfusion-associateddeaths: 1976 through 1985. Transfusion 1990;30:583-90.

37. Linden JV, Kaplan HS. Transfusion errors: causesand effects. Transfus Med Rev 1994;8:169-83.

38. Williamson L, Cohen H, Love E, Jones H, Todd A,Soldan K. The Serious Hazards of Transfusion(SHOT) initiative: the UK approach to haemo-vigilance. Vox Sang 2000(78 suppl);2:291-5.

39. Thomson S, Johnstone M. Delayed haemolytictransfusion reaction due to anti-Jka and anti-M.Can J Med Technol 1972;34:159-61.

40. Holland PV, Wallerstein RO. Delayed hemolytictransfusion reaction with acute renal failure. JAMA1968;204:1007-8.


L.D. PETZ

41. Rothman IK, Alter HJ, Strewler GJ. Delayed overthemolytic transfusion reaction due to anti-Uantibody. Transfusion 1976;16:357-60.

42. Pineda AA, Brzica SM Jr., Taswell HF. Hemolytictransfusion reaction. Recent experience in a largeblood bank. Mayo Clin Proc 1978;53:378-90.

43. Mollison PL, Engelfriet CP, Contreras M, eds. Bloodtransfusion in clinical medicine. 10th ed., Oxford:Blackwell Science Ltd., 1997.

44. Pineda AA,Taswell HF,Brzica SM Jr. Delayed hemo-lytic transfusion reaction. An immunologic hazardof blood transfusion. Transfusion 1978;18:1-7.

45. Redman M, Regan F, Contreras M. A prospectivestudy of the incidence of red cell allo-immunisation following transfusion. Vox Sang1996;71:216-20.

46. Branch DR, Hian AL, Carlson F, Maslow WC, PetzLD. Erythrocyte age-fractionation using a Percoll-Renografin density gradient: application toautologous red cell antigen determinations inrecently transfused patients. Am J Clin Pathol1983;80:453-8.

47. Brecher M, ed. Technical manual. 14th ed.Bethesda, MD: American Association of BloodBanks, 2002.

48. Reid ME, Toy PT. Simplified method for recoveryof autologous red blood cells from transfusedpatients. Am J Clin Pathol 1983;79:364-6.

49. Renton PH, Hancock JA. A simple method ofseparating erythrocytes of different ages. Vox Sang1964;9:183.

50. Vengelen-Tyler V, Gonzales B. Reticulocyte richRBCs will give weak reactions with many bloodtyping antisera (abstract). Transfusion 1985;25:476.

51. Garratty G, Arndt PA. Applications of flow cyto-fluorometry to red blood cell immunology.Cytometry 1999;38:259-67.

52. Griffin GD, Lippert LE, Dow NS, Berger TA,Hickman MR,Salata KF. A flow cytometric methodfor phenotyping recipient red cells followingtransfusion. Transfusion 1994;34:233-7.

53. Wagner F. Identification of recipient Rh pheno-type in a chronically transfused child by two-colour immunofluorescence. Transfus Med 1994;4:205-8.

54. Wenk RE, Chiafari PA. DNA typing of recipientblood after massive transfusion. Transfusion 1997;37:1108-10.

55. Reid ME, Rios M, Powell VI, Charles-Pierre D,Malavade V. DNA from blood samples can be usedto genotype patients who have recently received atransfusion. Transfusion 2000;40:48-53.

56. Castilho L, Rios M, Pellegrino J, Rodrigues A, CostaFF. Blood group genotyping for the managementof patients with “warm” antibody-inducedhemolytic anemia (WAIHA) (abstract). Blood2001;98:62a.

Lawrence D. Petz, MD, StemCyte, 400 Rolyn Place,Arcadia, CA 91007; [email protected].


RBC transfusions in a patient with a history of autoimmunehemolytic anemia (AIHA) can represent both a laboratory and aclinical challenge. The development of high-titer low-avidityantibodies and antibodies to high-frequency antigens may furtherimpair the ability to identify compatible donor RBCs. Notinfrequently, incompatible RBCs must be used and the desire toincrease oxygen carrying capacity conflicts with the desire to avoidexacerbating the autoimmune hemolytic process with RBCtransfusions. A 66-year-old Caucasian female with coronary arterydisease and a history of refractory AIHA had recently developedanemia and required multiple RBC transfusions. The patient hadmaintained adequate RBC counts with erythropoietin andprednisone therapy for the previous 16 months. With the recentworsening of her hemolytic anemia, she had developed angina thatwas treated with RBC transfusions in an outpatient setting.However, her angina increased as her RBC counts decreased,leading to hospital admission for further management of herhemolytic anemia and angina. She subsequently required multipleincompatbile RBC transfusions despite increased prednisonetherapy and did not improve until after coronary artery stentplacement and high dose IVIG therapy. This case demonstrates theusefulness of early patient phenotyping in a case of acceleratinghemolytic anemia to aid in donor RBC selection, the value ofcommunicating with clinicians and the patient regarding the use ofleast-incompatible RBCs, and the importance of optimizing thepatient’s clinical condition to avoid ischemia. In addition, itdemonstrates the value of repeated attempts with IVIG treatmentdespite previous refractoriness to this treatment.Immunohematology 2004;20:177–183.

Key Words: autoimmune hemolytic anemia, multipleincompatible transfusions, IVIG therapy

Indications for RBC transfusion of patients withwarm autoimmune hemolytic anemia (AIHA) includethe risk for cardiac or cerebral ischemia. Transfusion ofthese patients may confer a risk of acceleratedhemolysis due to stimulation of the autoimmuneprocess, although this is a controversial notion (seereference 1 for a comprehensive review). However,there is also the risk for intravascular or extravascularhemolytic transfusion reactions due to the presence of

alloantibodies in these often multitransfused patients.These alloantibodies may be very difficult todistinguish from the background of warm autoanti-bodies, and in cases of high-titer, low-avidity (HTLA)antibodies or antibodies to high-frequency antigens,the workup may be referred to a reference laboratorywith access to rare cells and antisera.

It is not possible to define a critical Hb level belowwhich transfusion should be administered during anexacerbation of hemolysis. The individual’s capacity toadjust to anemia depends not only on the level of Hbmeasured, but also on the rate of decline in Hb levelsand the capacity for cardiac compensation. In caseswith a background of cardiovascular disease, clinicalsymptoms and signs of cardiac ischemia guide thethreshold for transfusion on an individual-case basis.The decision to transfuse such a patient thus dependson the clinical situation and extensive communicationwith the clinical staff.

Materials and MethodsABO and Rh grouping, and screening for RBC

antibodies, were performed by standard tubetechniques. Similarly,RBC phenotyping was performedby standard tube techniques with commerciallyavailable typing sera. Antibody screening and panelcells were from multiple sources, including Immucor,Inc., Norcross, Georgia; Gamma Biologicals, Inc.,Houston,Texas; and Ortho-Clinical Diagnostics,Raritan,New Jersey. All samples were tested with at least 11different panel cells plus an autocontrol. DATs wereperformed using polyspecific anti-IgG, C3d (BiocloneOrtho), rabbit anti-IgG (Immucor), and anti-C3b, C3d(Bioclone, Ortho)-specific antisera.

Case report: exacerbation ofhemolytic anemia requiringmultiple incompatible RBCtransfusionsA.M. SVENSSON, S. BUSHOR,AND M.K. FUNG


Eluates were prepared by a rapid acid elutiontechnique (Elu-Kit II, Gamma). Antibody enhancementwas achieved using a PEG-based method (PEP,GTI, Inc.,Waukesha,WI) or treatment of RBCs with freeze-driedpapain (Immucor). PEG autoadsorption was per-formed as previously described.2,3 Alloadsorption ofserum was performed using untreated RBCs. Othermethods performed by the American Red CrossNational Reference Laboratory are as described in thecase report.

Incompatible units were transfused using the invivo crossmatch procedure, in which 30 to 50 mL ofdonor blood was slowly transfused into the patient,then the transfusion was temporarily stopped and asample drawn from the patient and examined forvisible evidence of hemolysis. In the absence ofhemolysis, the remainder of the donor unit wastransfused. A second sample was drawn from thepatient at the end of the transfusion to look for visiblehemolysis. This procedure was repeated with each unittransfused.

Case ReportA 66-year-old Caucasian woman, previously

diagnosed with refractory AIHA, presented to theemergency department with increasingly severehemolysis and angina on exertion, following about 3weeks of increasing transfusion needs on an outpatientbasis. Her Hct had dropped to 22.5% from 33% 4 daysprior to admission. The Hb was 7.5 g/dL and the WBCcount was 22,630/µL, with 88% neutrophils, 4%monocytes, 5% lymphocytes, 1% basophils, 2%metamyelocytes, and 6 nucleated RBCs/100 WBCs. Theplatelet count was 513,000/µL, and the reticulocytecount was > 30%.Total bilirubin was 4.5 mg/dL.

Past historyThe patient was group O, D+ and had refractory

AIHA diagnosed 6 years previously that requiredtreatments including splenectomy, cyclophosphamide,rituximab, cyclosporine, and high-dose IVIG. A bonemarrow biopsy performed at the initial presentation ofdisease showed no evidence of malignancy. For theprevious 16 months she had not required bloodtransfusions and antibody screening tests had beennegative for the previous 2 years. Additional medicalhistory included coronary artery disease requiringbypass graft surgery more than 10 years previously, twominor strokes 2 years previously, hypertension,hyperlipidemia, and four uneventful pregnancies. On

admission her medications included clonidine,diltiazem, prednisone at a dose of 15 mg/day, weeklyerythropoietin injections, dietary iron supplemen-tation, and opioid analgesics for chronic lower backand lower extremity pain. There had been no recentchanges in her medication or any infectious prodromeprior to this most recent exacerbation of her hemolyticanemia.

Recent historyThree weeks prior to admission to the emergency

department, the patient had an office visit forcomplaints of fatigue and mild angina. She hadpreviously experienced chest pain when her Hct hadfallen below 30%. At this time, the Hb was 10.6 g/dLand the Hct was 32.7%. On the basis of her mild anginasymptoms, a RBC transfusion in an outpatient settingwas ordered. The blood bank workup demonstratedstrong microscopic reactivity with all antibodyscreening cells in the antiglobulin phase (but noreactivity at the immediate spin phase). The auto-control was negative. Serum reactivity was positive toa dilution of 1:16 and could not be neutralized withplasma.A screen for cold agglutinins was negative. Theuse of papain-treated RBCs resulted in enhancedreactivity to 1+ with all antibody screening cells. Nodistinct specificity was apparent. An attempt toremove possible autoreactive antibodies by PEGautologous adsorption also generated strong to 2+microscopic reactivity with all antibody screeningcells. The serum also reacted with RBCs negative forRg, Ch, Kpb, Jsb, and U.

The DAT was negative and since the patient hadnot been transfused for more than 1 year, phenotypingwas performed. The patient was negative for Cw, Fya,K,S, and P1. Crossmatching with random RBC unitsshowed strong microscopic reactivity. The interpre-tation of these serologic findings was that theantibodies detected were most consistent withnonneutralizable HTLA antibodies, but the possibilityof antibodies to clinically significant, high-frequencyantigens could not be excluded. Following discussionswith the ordering physician, the patient, and herhusband, the patient received the two least-incompatible K– units by in vivo crossmatch, withoutcomplications.

Before admission to the emergency department,the patient had received transfusions of six least-incompatible K–, Fy(a–) units by in vivo crossmatch onthree occasions without complications, but at

A.M. SVENSSON ET AL.


Hemolytic anemia and multiple incompatible RBC transfusions

increasingly shorter intervals between transfusions.Furthermore, the reactions to antibody screening cellsand crossmatches had increased to macroscopic levels(1+). The DAT and autologous control cells remainednegative. However, RBCs from the patient were notedto weakly autoagglutinate in the saline control. Thiseffect could be removed with warm 37°C salinewashing of the cells, suggesting the development of acold agglutinin. Two allogeneic adsorptions usingantibody screening cells, both with and without PEGenhancement, failed to completely remove reactivitythat persisted against the cells used in the adsorptionprocedure. Antibodies against preservatives in thescreening cells were ruled out by using multiply-washed screening cells. No correlation between thestrength of reactivity and the age of crossmatched RBCunits was identified. An extensive review of her medi-cation list found no drugs reported to inducehemolytic anemia, although the patient’s husband didreport a temporal association with the use of the anti-depressant mirtazapine in the past. However, a searchof the literature showed no reported association.

Extensive serologic workupGiven the increasing strength of reactivity against

allogeneic cells and the shortened RBC survival, therewas a concern that the antibodies detected wereagainst a clinically significant, high-frequency antigen.A more extensive serologic workup was requestedfrom the American Red Cross (ARC) Blood Services—New England Region Reference Laboratory inBurlington,Vermont. Briefly, it was concluded that nospecificity could be identified, but evidence of weakcomplement binding (microscopic) was detectedusing an enhanced DAT (5 minute incubation at 37°C),which suggested the presence of a warm autoantibody.It was further concluded that a nonneutralizable HTLAantibody was present. RBCs lacking the following high-frequency antigens were tested and found to bereactive with the patient’s serum: k, I, Jsb, Kpb,Yta, U, Rg,Ch, Lan,AnWj, Ge2, Ge3, Jra,Vel, PP1Pk, Coa, Joa, Gya, Hy,Yka, Kna, McCa, McCb, Csa, Lub, and JMH. A newspecimen was then sent to the ARC National ReferenceLaboratory for Blood Group Serology in Philadelphia,Pennsylvania, for a more extensive evaluation.

Testing performed by the ARC National ReferenceLaboratory was similarly unable to identify an antibodyspecificity. The sample received by their laboratorywas a later sample. The DAT was 1+ with polyspecificanti-human globulin, negative with monospecific

anti-IgG, and 1+ with monospecific anti-C3d, and asaline control using warm saline 37°C washed RBCswas negative. Weak autoagglutination of the patient’sRBCs could be removed by warm 37°C saline washing.The patient’s serum reacted with all RBCs, includingautologous cells tested by 37°C albumin, IgGantiglobulin test (AGT), and ficin IgG-AGT. RBCstreated with DTT were still reactive. Furthermore,DTTtreatment of the patient’s serum changed reactivityfrom weak reactivity to 3+. The serum was reactive toa dilution of 1:256. Five adsorptions with rr, r′r′, or r″r″RBCs did not remove this reactivity. However, after thefifth adsorption, the remaining reactivity was uniformlyweakly positive against panel cells, suggesting that nostrongly reactive antibodies existed against major RBCantigens. RBCs with the following phenotypes werefound reactive at the albumin-IgG-AGT phase: JMH–,Di(b–), At(a–), Cs(a–), Yk(a–), Jk(a–b–), I–, PP1Pk, –D–,Rhnull, Cde/Cde, cdE/cdE, Lu(a–b–),Yt(a–), Co(a–), Lan–,Vel–, Kp(b–), McLeod, Sc–1, and Jo(a–). Finally, theantibody reactivity was assessed for clinicalsignificance with the monocyte monolayer assay usinggroup O,c–,E–,P1–,S–,K–,Fy (a–);group O,C–,e–,P1–,S–, K–, Fy (a–); and group O, C–, E–, P1–, S–, K–, Fy (a–)RBCs, with and without fresh complement, and usingpooled monocytes from two donors. The reactivity ofthe patient’s serum was 21.5 percent to 34 percentreactive monocytes, consistent with a clinicallysignificant antibody (normal range is 0–3%).

On the present admission to the emergencydepartment, the patient had angina on exertion. Therewere no ischemic EKG changes and Troponin I was < 0.15 ng/mL (normal < 0.15, indeterminate 0.15–1.50,positive > 1.50 ng/mL). The blood bank workup at thatpoint showed increased panel cell reactivity of 1 to 2+.Furthermore, the DAT was 1+ with broad-spectrumantiglobulin reagent, 1+ with IgG monospecific serum,and 1+ with anti-C3 monospecific serum. However, thesaline control was now strongly positive micro-scopically, suggestive of autoagglutination. With warm37°C saline washing of the patient’s RBCs, only thebroad spectrum and anti-C3–specific reactionsremained reactive at 1+. An eluate was performed andfound to have nonspecific reactivity at the antiglobulinphase (weak to 1+ microscopic) for all RBCs tested.The autologous cell control at 4°C was also 1+. Allcrossmatched Cw–, Fy (a–),K–, S– RBC units showed 1+to 2+ reactivity.

At this point the risk for ischemic damage to theheart due to severe anemia was weighed against the



risk of exacerbating the autoimmune processunderlying the hemolytic anemia with the transfusionof incompatible RBCs. Because at this point there wereno clear clinical signs of a developing myocardialinfarction, it was felt that transfusion should be heldoff, if possible, pending further testing results.However, the following night the patient started tocomplain of moderate chest pain at rest. Troponin Ihad increased to 0.29,Hb was 8.5 g/dL,and the Hct was25.6%. The patient subsequently received two 1+crossmatch-incompatible units by in vivo crossmatchwithout complications, following a bolus dose of 60 mgprednisone. She continued on 60 mg of prednisoneper day. On day 3,Troponin I peaked at 0.44. Hb haddecreased to 7.7 g/dL, the Hct was 23%, and thepatient’s chest pain persisted at the same level,necessitating further transfusions. Only 2+ incom-patible units could be found and the patient received aunit of CW–, Fy (a–), K–, S–, P1– RBCs by in vivoprocedure without complications.

On day 4, Hb was 9.3 g/dL and Hct was 27.8%. Itwas decided that the patient would undergo cardiaccatheterization and coronary artery stents were placed,with immediate relief of symptoms. On day 5, Hb haddecreased to 8.0 g/dL, the Hct was 23.9%, and totalbilirubin had increased to 5.8 mg/dL, confirmingincreasing hemolytic activity. Reactivity of cross-matched phenotype matched units (CW–, Fy a(–), K–,S–, P1–) was generally increased to 1+ to 3+; however,two microscopically reactive units were found. Overthe following 5 days, the patient continued to receiveleast-incompatible phenotype-matched units withoutimmediate complications or evidence of intravascularhemolysis. However, there was an increase in the totalbilirubin (7.7 mg/dL on day 6), an increase in thestrength of reactivity against crossmatched units, andincreasing frequency of transfusion requirements (twoRBC units every 2–3 days). Reactivity to crossmatchedunits continued to increase to up to 4+ on day 7.

Following 10 days of high-dose prednisonetreatment without discernible improvement, high doseIVIG thereapy was attempted with a one-time dose of1g/kg body weight. Three days later, Hb had reached10.6 g/dL, Hct was 31.8%, and total bilirubin was 2.1mg/dL. The patient was no longer in need oftransfusions and could be transferred to a rehabilitationunit, where she continued to maintain Hb valuesaround 11 g/dL and Hct well above 30% (range of33%–37%). A bone marrow biopsy was scheduled andthen canceled due to the patient’s dramatic

improvement. The patient’s medication at time oftransfer included prednisone 60 mg/day, erythro-poietin, folic acid, and clopidogrel.

DiscussionDifferentiation between a preexisting warm-

reactive autoantibody and an antibody against a high-frequency antigen appearing after a transfusion isdifficult. However, critical results in this patient’sevaluation were a negative DAT and a negative autolo-gous control. However, with continued transfusion,positive results with gradually increasing strength wereseen in these tests. With these findings, warm-reactiveautoantibodies were considered less likely, but couldnot be completely excluded, and more thought wasgiven to the possibility of a HTLA antibody or anantibody against a high-frequency antigen. A HTLAantibody was initially favored given the weak reactivitythat persisted to dilutions of 1:16. However, concernfor a clinically significant antibody, especially against ahigh-frequency antigen, began to grow as the patientbegan to require more RBC transfusions at shorter timeintervals, and the reactions obtained with panel cellsand crossmatches increased to macroscopic reactivity,considered unusual for a HTLA antibody. Although theDAT and autologous cell control were initially negative,the patient likely had a chronic immune hemolyticanemia that was previously kept under control witherythropoietin and prednisone therapy. Therefore, theconfluence of an autoimmune hemolytic process,HTLA antibody, and possible antibody against a high-frequency antigen rendered the decision to transfusecrossmatch-incompatible RBCs that much moredifficult.

To aid in the selection of RBC units for transfusionof patients with unidentifiable specificities, pheno-typing for common antigens of the Rh, K, Jk, Fy, andMNS systems should be performed on the patient’sRBCs at the first opportunity prior to RBC transfusion.In this case, the patient had not been transfused inmore than a year and presented with a negative DAT.While difficult, even in patients who have beentransfused over the past 3 months, a phenotype can beobtained based on either molecular4–9 or serologictechniques.10,11 With the knowledge of which majorantigens are lacking in the patient, decisions can thenbe made in identifying appropriately phenotype-matched units that would be least likely to generate ahemolytic transfusion reaction and would survivelonger in the patient’s circulation.



When searching for RBC units that are antigennegative, the difficulties increase as multiple antigensare required to be simultaneously absent. The need forusing only partially matched or incompatible unitsarises and choices must be made as to which antigensto ignore in screening for antigen-negative RBC units totransfuse. The degree of immunogenicity of eachantigen must then be taken into account, as well as theseverity of the possible consequences of transfusingeach antigen. Certain combinations of antigen-negativeRBC units are more easily found than others and mustalso be taken into account in a search. Furthermore,results of alloadsorption studies may indicate theabsence of any antibodies against major RBC antigensand help in guiding which antigens could be morecomfortably ignored. In this patient’s situation,alloadsorptions performed at the reference laboratoryfailed to completely adsorb all antibodies but reducedthe reactivity to only microscopic reactivity with allcells tested. The inference from this finding was thatwhile no statement could be made regarding reactivityagainst high-frequency antigens, no strongly reactingantibodies against major RBC antigens were identified.However, alloadsorptions would need to be performedat 3 to 4 day intervals to rule out their emergence if thepatient continued to be transfused with units positivefor antigens that the patient lacked.

In the absence of the ability to identify antibodyspecificities and compatible RBC units by routinelyused serologic tube-based methods, alternativemethods exist for determining if an antibody may beclinically relevant. The monocyte monolayer assayquantifies rosetting or phagocytosis of antibody-sensitized cells by monocytes. In this case it indicatedthat the antibodies in the patient’s serum had clinicalsignificance, i.e., would be likely to cause decreasedsurvival of transfused RBCs. In this particular case, themonocyte monolayer assay might have been moreuseful if donor cell phenotypes were identified thatpredicted increased RBC survival. Other reportedmethods that have been used include 51Cr labeling ofRBCs12 and flow cytometry.13 The results of such assaysmay help evaluate the current immunization status andhelp reveal alloimmunization by comparing thesurvival of autologous RBCs with allogeneic RBCs.However,because an antibody may develop at any timeduring a period of repeated transfusions, the resultscannot be used to exclude the existence of clinicallysignificant alloantibodies in the clinical setting beyondthe point of the next transfusion.

An alternative option that may be available in thefuture is the use of blood substitutes. Hemoglobin-based blood substitutes are currently in phase IIIclinical trials for various indications and have beenused successfully in at least one instance for AIHA.14

The most effective use of hemoglobin-based bloodsubstitutes would be to temporarily stabilize thepatient until difficult-to-find or rare donor blood unitscould be obtained or until more aggressiveimmunosuppressive regimens have had a chance totake effect. In both instances however, there would bea prolonged requirement for the use of bloodsubstitutes and the introduction of free hemoglobininto the patient’s circulation would interfere with thevisual assessment of hemolysis.

The balance between the risks of inadequateoxygenation and subsequent ischemic damage on theone hand, and the risks of transfusing incompatibleRBCs on the other hand,represents a clinical challenge.In situations where there is a critical need to optimizeoxygen delivery by increasing RBC volume but forwhich no compatible units can be found, transfusionby in vivo crossmatch is unavoidable. In theseinstances, such as in this case, close communicationwith the clinicians is essential to give adequate inputinto the decision-making process. In this context itshould be emphasized that the in vivo crossmatch hasa limited value in predicting survival of the transfusedRBCs, since one is only looking for acute gross intra-vascular hemolysis during the transfusion.Therefore, anegative in vivo crossmatch result only indicates thatthe likelihood of acute intravascular hemolysis and itsattendant morbidity and mortality is lessened.Furthermore, in situations where the in vivo cross-match procedure has been repeated multiple timeswith negative results, continuous awareness of theincreased risks of transfusing otherwise incompatibleRBCs is imperative for both the physicians who orderthe transfusions and the nurses who administer them.Finally, it is important to communicate with theclinicians and the patient the exact nature of “least-incompatible blood”15 and its increased risks, to addressany concerns regarding the in vivo crossmatchprocedure, and to further explain why no compatibleblood could be found.

Apart from identifying least-incompatible RBCunits for transfusion, therapeutic options for patientswith AIHA are generally aimed at reducing theclearance of RBCs and reducing the production ofantibodies. Corticosteroids usually constitute the first



line of treatment. Splenectomy is used when thepatient is refractory to treatment with corticosteroids.Immunosuppressive therapy, such as cyclophos-phamide and azathioprine and short-term use ofcyclosporine, is sometimes successful in inducingremission in these otherwise refractory patients. Theuse of rituximab in AIHA, a monoclonal antibody thatcauses specific B-cell depletion by targeting the B-cellCD20 antigen,has been reported.15–18 For patients whoare critically anemic and who are refractory totransfusion, plasma exchange may be considered, but itis highly inefficient at the removal of IgG antibodies,and better suited for patients with hemolysis due toIgM antibodies. 19

The immunomodulatory action of IVIG therapy,although not well understood, has been usedsuccessfully for a myriad of autoimmune disorders.20

Although unresponsive to this therapy in the past, ourpatient responded promptly to administration of IVIG.There are substantial differences in IVIG, based ondifferences in purification and chemical stabilizationby manufacturers as well as lot-to-lot variations due tovariation in the donor pool, even with the samemanufacturing process. Such differences have beennoted in the use of IVIG as an immunomodulatorytherapy in solid organ transplantation21 and mayexplain the different responses to IVIG in this patient.

In conclusion, this case illustrates the managementof accelerating AIHA with critical levels of Hb inassociation with threatened cardiac ischemia andpossible alloantibodies. It demonstrates the usefulnessof performing phenotyping in these patients at firstopportunity and the necessity to thoroughly work upthe antibody reactivity in the presence of a negativeautologous cell control. Furthermore, it exemplifiesthe importance of optimization of clinical conditions,in this case performing cardiac catheterization andstent placement to maximize oxygen delivery to theheart. In addition, it shows that IVIG may be useful andshould be attempted in these patients even if priortreatments with IVIG have been unsuccessful.

AcknowledgmentsWe thank the laboratory staff of the Fletcher Allen

Health Care Blood Bank; American Red Cross (ARC)National Reference Laboratory for Blood GroupSerology in Philadelphia, Pennsylvania; and the ARCBlood Services-New England Region ReferenceLaboratory in Burlington,Vermont, for their invaluableassistance.

References1. Issitt PD, Anstee DJ. “Warm” antibody-induced

hemolytic anemia (WAIHA). In: Applied bloodgroup serology. 4th ed. Durham, NC: ScientificPublications, 1999: 939-93.

2. Liew YW, Duncan N. Polyethylene glycol inautoadsorption of serum for detection of allo-antibodies (letter). Transfusion 1995;35:713.

3. Leger RM, Garratty G. Evaluation of methods fordetecting alloantibodies underlying warm auto-antibodies. Transfusion 1999;39:11-6.

4. Reid ME, Rios M, Powell VI, et al. DNA from bloodsamples can be used to genotype patients whohave recently received a transfusion. Transfusion2000;40:48-53.

5. Wenk RE, Chiafari FA. DNA typing of recipientblood after massive transfusion. Transfusion 1997;37:1108-10.

6. Legler TJ, Eber SW, Lakomek M, et al. Applicationof RHD and RHCE genotyping for correct bloodgroup determination in chronically transfusedpatients. Transfusion 1999;39:852-5.

7. Rozman P, Dov T, Gassner C. Differentiation ofautologous ABO, RHD, RHCE, KEL, JK, and FYblood group genotypes by analysis of peripheralblood samples of patients who have recentlyreceived multiple transfusions. Transfusion 2000;40:936-42.

8. Kroll H,Carl B,Santoso S,et al.Workshop report onthe genotyping of blood cell alloantigens. TransfusMed 2001;11:211-9.

9. Reid ME, Lomas-Francis C. Molecular approachesto blood group identification. Curr Opin Hematol2002;9:152-9.

10. Reid ME,Toy P. Simplified method for recovery ofautologous red blood cells from transfusedpatients. Am J Clin Pathol 1983;79:364-6.

11. Branch DR, Hian AL, Carlson F, et al. Erythrocyteage-fractionation using a Percoll-Renografindensity gradient:application to autologous red cellantigen determinations in recently transfusedpatients. Am J Clin Pathol 1983;80:453-8.

12. Baldwin ML, Barrasso C, Ness PM, Garratty G. Aclinically significant erythrocyte antibody detect-able only by 51Cr survival studies. Transfusion1983;23:40-4.

13. Zeiler T, Müller JT, Hasse C, et al. Flow cytometricdetermination of RBC survival in autoimmunehemolytic anemia. Transfusion 2001;41:493-8.



14. Mullon J, Giacoppe G, Clagett C, et al. Transfusionsof polymerized bovine hemoglobin in a patientwith severe autoimmune hemolytic anemia. NEngl J Med 2000;342:1638-43.

15. Petz LD. Least incompatible units for transfusionin autoimmune hemolytic anemia: should weeliminate this meaningless term? A commentaryfor clinicians and transfusion medicineprofessionals.Transfusion 2003;43:1503-7.

16. Ahrens N, Kingreen D, Seltsam A, Salama A.Treatment of refractory autoimmune haemolyticanaemia with anti-CD20 (rituximab). Br JHaematol 2001;114:244-5.

17. Quartier P, Brethon B, Philippet P, et al. Treatmentof childhood autoimmune haemolytic anaemiawith rituximab. Lancet 2001;358:1511-3.

18. Morselli M, Luppi M, Potenza L, et al. Mixed warmand cold autoimmune hemolytic anemia:complete recovery after 2 courses of rituximabtreatment. Blood 2002;99:3478-9.

19. Silberstein LE, Berkman EM. Plasma exchange inautoimmune hemolytic anemia (AIHA). J ClinApheresis 1983;1:238-42.

20. Knezevic-Maramica I, Kruskall MS. Intravenousimmune globulins: an update for clinicians.Transfusion 2003;43:1460-80.

21. Wassmuth R, Hauser IA, Schuler K, et al.Differential inhibitory effects of intravenousimmunoglobulin preparations on HLA-alloanti-bodies in vitro. Transplantation 2001;71;1436-42.

Annika M. Svensson, MD, PhD, Sharon Bushor,MT(ASCP), and Mark K. Fung, MD, PhD (correspon-ding author), Fletcher Allen Health Care/Universityof Vermont Department of Pathology, 111 ColchesterAvenue, Burlington, VT 05401.


Delayed hemolytic transfusion reactions (DHTRs) usually occurbetween 3 and 14 days posttransfusion as a result of a secondaryimmune response, with a drop in Hb level, fever, jaundice, orhemoglobinuria. DHTRs caused by a primary immune response areparticularly rare events, and only a few reports have been known.In this report, we describe an unusual case of a DHTR caused byanti-Fyb in a 42-year-old man, who had no prior history oftransfusion. Although it seems to be a rare phenomenon, wesuggest that DHTRs by a primary immune response may beconsidered even in the case of the patient who had typicalevidence of hemolysis but who had no previous transfusion history.Immunohematology 2004;20:184–186.

Key Words: delayed hemolytic transfusion reactions,anti-Fyb, primary immune response

The most important Duffy antigens in routineblood bank serology are Fya and Fyb.1 Because Duffyantigens are only moderate immunogens, anti-Fya

occurs with one-third the frequency of anti-K. Inaddition, anti-Fyb antibodies have an incidence rate of 1to 20 compared to anti-Fya.2 However, these alloanti-bodies are usually IgG; react best at the antiglobulinphase; and are known as clinically significant,unexpected antibodies, related to acute hemolytictransfusion reactions (HTRs) and delayed hemolytictransfusion reactions (DHTRs).1,3 IgM Duffy antibodiesusually occur as a primary immune response. IgGDuffy antibodies form via a secondary immuneresponse after another exposure to Duffy antigens.4,5

DHTRs usually become apparent between 3 and 14days posttransfusion and are generally attributed to ananamnestic immune response. HTRs caused by aprimary immune response are extremely rare events,and only a few reports have been published.6–9

Case ReportA 42-year-old man suffering from multiple fractures

was admitted to the emergency room of Pusan NationalUniversity Hospital on January 15, 2003, after fallingwhile he was on duty at a construction site. He wasscheduled for emergency surgery since an L-Spine MRIrevealed an unstable bursting fracture (L1) with cordcompression. Right distal radius fracture and rightcalcaneal fracture were also present. His RBCs typed asgroup O,D+. Antibody screening was negative. Duringsurgery, the patient received three units of packedRBCs and two units of FFP. Postsurgical bleedingcaused the patient’s Hb level to continuously fall. Overthe following 7 days, he received an additional 18 unitsof saline crossmatch-compatible packed RBCs until theHb level stabilized at 11.5g/dL. The prothrombin timewas mildly increased to 18.7 seconds (reference range11–14.1). On the 11th hospital day, jaundice and anunexpected fall in Hb (9.9 g/dL) occurred. Antibodyscreening was performed again. Anti-Fyb was identifiedand confirmed by gel test using LISS/Coombs andNaCl/Enzyme cards (DiaMed AG, Cressier Sur Morat,Switzerland). The patient’s RBCs were negative in theDAT. The patient had no previous history of transfusionnor history of any particular medical or surgicalillnesses, except that he had received minor surgery onone of his right toes 5 years previously. The laboratoryfindings on the 11th hospital day were as follows: totalbilirubin 2.15 mg/dL (reference range 0.3–1.3 mg/dL),direct bilirubin 0.87 mg/dL (reference range 0.05–0.40mg/dL), LDH 1146 IU/L (reference range 218–472IU/L), C3 121 mg/dL (reference range 50–90 mg/dL),

Delayed hemolytic transfusionreaction due to anti-Fyb caused by a primary immune response:a case study and a review of theliteratureH.H. KIM,T.S. PARK, S.H. OH, C.L. CHANG, E.Y. LEE,AND H.C. SON


Primary response: anti-Fyb and delayed transfusion reaction

C4 23.5 mg/dL (reference range 10–40 mg/dL), and C-reactive protein (CRP) 5.17 mg/dL (reference range0–0.5 mg/dL). The haptoglobin was below detectablelevels (< 10.0 mg/dL), but the level of alpha-1-acidglycoprotein was 171.0 mg/dL (reference range 45–98mg/dL). The patient was diagnosed with a DHTRcaused by anti-Fyb. Three compatible units of Fy(b–)packed RBCs were issued after crossmatching usingthe antiglobulin method with LISS/Coombs card(DiaMed AG). No further hemolytic reactions oc-curred, and the unexpected antibody, present in hisserum for 46 days after antibody detection, could nolonger be detected.

Materials and MethodsFor detection of RBC antibodies, column

agglutination methods were used. The DAT wasperformed on the patient’s RBCs from an EDTA-anticoagulated sample using the polyspecific LISS/Coombs card (DiaMed AG). Duffy phenotypes weredetermined on RBCs from EDTA-anticoagulatedsamples using ID-Antigen profile III (DiaMed AG). TheRBCs used as screening and identification panels werefrom commercially prepared reagent panel cells(DiaMed AG). Antibody screening and identificationtests were performed with LISS/Coombs and NaCl/Enzyme cards (DiaMed AG) at 37°C for 15 minutes.DTT (Sigma Chemical Co., St. Louis,MO) treatment wasalso performed to determine the immunoglobulin classof the alloantibody in the patient’s serum.

ResultsThe serum of the patient was reactive with all

Fy(b+) RBCs and negative with all Fy(b–) RBCs. Anti-Fyb was not detectable in the enzyme phase. Thepatient’s RBCs were Fy(a+b–) and were negative in theDAT when anti-Fyb was detected. DTT treatment of thepatient’s serum did not alter reactions, suggesting thatthe anti-Fyb was an IgG alloantibody.

DiscussionA study of 11 cases related to DHTRs by a primary

immune response in burned children, who had neverbeen transfused, was reported by Bacon and co-workers.6 All 11 patients had a negative DAT and IATbefore transfusion. In this study, anti-K and anti-E werethe most frequently identified antibodies. The authorsalso reported DHTRs due to anti-S, -C, and -Jka,respectively. Another case of DHTR due to anti-C by a

primary immune response was reported by Patten etal.7 Solanki et al.8 described three patients with notransfusion history who had DHTRs caused by aprimary immune response. Recently, a DHTR due to aprimary immune response that stimulated anti-Jka and -K in a 24-week-old female was reported.9 However,noknown report has yet been published on a DHTR by aprimary immune response involving Fyb, an antigenwhich has relatively low immunogenecity.

Haptoglobin depletion is usually the most sensitivelaboratory indicator of hemolysis. On the other hand,haptoglobin synthesis is increased in the presence ofacute inflammatory processes. In our patient, weobserved an undetectable haptoglobin level but highalpha-1-acid glycoprotein and CRP, so the possibility ofextravascular hemolysis plus an acute phase responsemust be considered. Unfortunately, we could not testthe urine or test for plasma Hb because of the presenceof hematuria caused by trauma and because a plasmaspecimen was not available.

ConclusionOn the basis of clinical and laboratory findings, as

well as the patient’s past history, DHTR by a primaryimmune response was strongly suspected,even thoughDHTRs caused by a primary immune response areextraordinarily rare events. It is thought that such anatypical hemolytic reaction must be related to thetransfusion of the large amount of packed RBCs withina short period of time. Thus, although it is a very rarephenomenon, we suggest that DHTR by a primaryimmune response should be considered, even in thecase of the patient who had typical evidence ofhemolysis but no previous transfusion history.

AcknowledgmentWe thank Dr. S. Gerald Sandler for many insightful

comments and his recommendation, and MIRR SciTechCorp. for technical assistance.

References1. Harmening DM. Modern blood banking and trans-

fusion practices. 4th ed. Maryland: F.A. Davis,1999:161-99.

2. Marsh WL, Ehrich CC. The Duffy blood groupsystem: a review of recent developments.Infusionsther Klin Ernahr 1975;2:280-9.

3. Brecher ME. Technical manual. 14th ed. Bethesda,MD: American Association of Blood Banks, 2002:315-59.


H.H. KIM ET AL.

4. Mollison PL, Engelfriet CP, Contreras M. Bloodtransfusion in clinical medicine. 10th ed. Oxford:Blackwell Science, 1997:81-2.

5. Anderson KC, Ness PM. Scientific basis oftransfusion medicine. 1st ed. Philadelphia:Saunders, 1994:507-8.

6. Bacon N, Patten E, Vincent J. Primary immuneresponse to blood group antigens in burnedchildren. Immunohematology 1991;7:8-11.

7. Patten E, Reddi CR, Riglin H, Edwards J. Delayedhemolytic transfusion reaction caused by aprimary immune response. Transfusion 1982;22:248-50.

8. Solanki D, McCurdy PR. Delayed hemolytictransfusion reactions. An often-missed entity.JAMA 1978;239:729-31.

9. Albiero AL, Novaretti MC, Llacer PE, Chamone DA.Early primary immune response against erythro-cytes: a case report. Transfus Med 2003;13:93-7.

Hyung Hoi Kim, MD, PhD (corresponding author),Tae Sung Park, MD, Seung Hwan Oh, MD, Chulhun L.Chang, MD, PhD, Eun Yup Lee, MD, PhD, and HanChul Son, MD, PhD, Department of LaboratoryMedicine, College of Medicine, Pusan NationalUniversity, 1-10 Ami-dong, Seo-gu, Busan 602-739,Korea.


A 24-year old female, gravida III, para III, delivered a full-term infantby cesarean section. A maternal blood sample at the time ofadmission showed antibody in her serum that had apparent anti-especificity and that her RBCs were e+. Further studies determinedthat the antibody was anti-hrS. Cord RBCs had a negative DAT anda normal Hb level. There was no clinical evidence for increasedhemolysis in the infant. We describe an hrS+ infant with noevidence of HDN due to anti-hrS. Immunohematology2004;20:187–189.

Key Words: anti-hrS, variant e antigen, HDN due toanti-hrS

The e antigen is present in about 98 percent of theCaucasian population. The hrS antigen, a variant formof the e antigen, is a high-frequency Rh antigen presenton all cells except for e– RBCs and rare e+ RBCs thatlack hrS. Alloanti-e–like antibodies may be made bythose individuals with e+ RBCs lacking this variantantigen. The hrS antigen was first reported in 1960 byShapiro, in a Bantu patient whose serum reacted withall cells bearing E or e.1 Shapiro estimated that about 6percent of Bantu Rh haplotypes were Dce or dceencoding no hrS antigen. Following adsorption of thepatient’s serum with R2R2 RBCs, the serum no longerreacted with E+e– RBCs but still reacted with most e+RBCs.2 The antigen, defined by the antibody reactingonly with e+ RBCs (except for a rare e antigen detectedin black South Africans) was named hrS and the broaderspecificity reacting with E+e+ RBCs was named anti-Hr(later, anti-Rh18). Thus, anti-hrS should be suspected inpatients with apparent alloanti-e specificity whoseRBCs are positive for the e antigen. This antibody hasbeen described as analogous to the anti-D made by D+individuals.

Some anecdotal reports suggest that antibodies toportions of the e antigen may be clinically significant.2

According to Reid and Lomas-Francis,3 the availabledata for the clinical significance of this alloantibody inHDN are limited. We describe an hrS+ infant with noevidence of HDN due to anti-hrS and review some ofthe available information concerning the heterogeneity

of the antibody response in individuals whose RBCsexpress a variant e antigen or lack some epitopes of thee antigen.

Case ReportA 24-year-old African American female, gravida III,

para III, presented for elective cesarean section inOctober 2003. Maternal history included two previouscesarean sections with the last delivery in February2000, at which time an anti-e was detected in thematernal serum. The second child was not affectedwith HDN. No information was available for the firstpregnancy and delivery. The father’s ethnic origin isunknown and he was not available for testing.

Results

Maternal sampleA maternal blood sample obtained before the

cesarean section was tested during the presentadmission. The RBCs were A+, C–, E+, c+, e+. A DATwas negative. Serum tested in a routine LISS panel(Immucor-Gamma, Houston, TX) showed no reactionsat immediate spin, a weak reaction with eight of elevencells at 37°C, and micro to weak reactions with sevenof eleven cells at the antihuman globulin (AHG) phase.A routine PEG panel (Immucor-Gamma) revealedapparent anti-e specificity with 1+ to 2+ reactions atthe AHG phase. An autocontrol run with both panelswas negative. Serum was also tested against a seven-member select-cell PEG panel consisting of e– RBCs.The e– PEG panel was negative.

Because the patient’s RBCs were e+, the possibilitythat the maternal antibody might be anti-hrS and notanti-e was considered. To confirm this possibility, theserum was tested against two RBC samples that weree+, hrS–. This testing showed negative reactions withboth cells. In addition, positive reactions were seenwith the maternal serum against two RBC samples thatwere e+ and hrB–, ruling out the presence of anti-hrB in

Maternal alloanti-hrS—an absenceof HDNR. KAKAIYA, J. CSERI, B. JOCHUM, L. GILLARD,AND S. SILBERMAN


R. KAKAIYA ET AL.

the maternal serum. At the time of these tests, an anti-hrS was not available to confirm the patient’s RBCs ashrS–. The titer of the maternal serum versus rr RBCswithout the addition of enhancement media was asfollows: undiluted,2+;1:2,1+;1:4 and 1:8,microscopicreactions.

Child’s sampleA cesarean section was performed and a healthy

baby girl was delivered. The cord blood was O+ with anegative DAT. Antibody screen on the cord serum wasnegative. At birth, the child’s Hb level was 18.8 gm/dLand the Hct was 56.5%. The total bilirubin levels on thefirst day of life ranged between 2.5 and 4.0 mg/dL.Clinically and serologically, the baby did not show signsof hemolysis; she was discharged on the third day.

At 2 months of age, the child’s blood sample wasobtained and tested. The child tested positive for e andhrS antigens. (At this time an anti-hrS from a patient wasavailable to test the child’s RBCs). A complete Rhphenotype was determined with the following results:D+, C–, c+, E+, e+, and hrS+.

DiscussionIn our case, the maternal antibody in the serum

appeared to have anti-e specificity on routine testingusing LISS and PEG panels. The reactions were weakwith a LISS panel, but stronger with a PEG panel. Thematernal RBCs were e+. Because an apparent alloanti-e was present in the serum of a patient who was e+,wesuspected the possibility that the patient may have analloantibody to the e-variant hrS. Therefore, we under-took further testing using RBCs that were e+, hrS–.Using these rare RBCs, we were able to confirm thepresence of anti-hrS in the maternal serum, suggestingthat the patient’s RBCs were hrS–, a variant form of thee antigen.

Testing of the infant revealed an absence ofserologic or clinical signs of hemolysis, including anegative DAT on the cord blood sample. At 2 monthsof age, the child’s RBCs were tested and were shown tobe e+ hrS+. These findings show that the maternal anti-hrS failed to cause HDN.

Unpublished observations suggest that someantibodies produced in people with a e-variantphenotype can be significant in vivo and sometimeseffect clearance of small labeled aliquots ofincompatible RBCs. However, recent data from Noizat-Pirenne et al.4 show that anti-Rh18 (containing anti-hrS)caused fatal transfusion reactions in two patients after

transfusions of incompatible RBCs. Therefore,transfusion in patients with these antibodies should beapproached with caution. According to Moores, theimmune response of hrS– people is highly variable.5

The results of the cord blood DAT in her report variedfrom microscopic positive to 4+ due to anti-Rh18.Seven affected infants received successful exchangetransfusions, although one died later. In addition, threeinfants suffered mild hemolytic disease, and sevenother infants either were miscarried or were deliveredelsewhere. Grobbelaar and Moores6 described an anti-hrS which would now be considered an anti-Rh18 inthe serum of a Bantu woman whose newborn sufferedfrom mild HDN. In our case, the maternal antibody wasanti-hrS (not anti-Rh18) and was not clinicallysignificant. This was our patient’s third pregnancy andat least the last two of her infants were unaffected. Theantibody tested only weakly with LISS and was strongerwith PEG. Moreover, this antibody did not react withRBCs that are both E+ and e+, the child’s phenotype.The reasons for the absence of HDN in our case areunclear but could be related to the nature or strengthof maternal alloantibody and/or a decreased antigendensity on fetal RBCs.

Noizat-Pirenne et al.7 determined that severalgenetic events give rise to the absence of hrS in Blackindividuals. Specifically, she determined that loss ofcertain amino acid residues in a transmembranedomain might alter the protein configuration and theimmune response to the antigen.

The ability to identify alleles responsible forvariable expression of the e antigen likely will havepractical application. For instance, fetal DNA obtainedfrom maternal plasma during pregnancy can beexamined to determine if the allele of interest wastransmitted to the fetus.

References1. Shapiro M. Serology and genetics of a new blood

factor: hrS. J Forensic Med 1960;7:96-7.2. Issitt PD, Anstee DJ. The antigen hrS. In: Applied

blood group serology. 4th ed. Durham, NC:Montgomery Scientific Publications, 1998:368-72.

3. Reid ME, Lomas-Francis C. hrS antigen. In: Theblood group antigen factsbook. New York:Academic Press, 1997:119-20.

4. Noizat-Pirenne F, Lee K, Le Pennec PY, et al. RareRHCE phenotypes in black individuals of Afro-Caribbean origin: identification and transfusionsafety. Blood 2002;100:4223-31.


Alloanti-hrS—absence of HDN

5. Moores P. Rh18 and hrS blood groups andantibodies. Vox Sang 1994;66:225-30.

6. Grobbelaar BG, Moores P. The third example ofanti-hrS. Transfusion 1963;3:103-4.

7. Noizat-Pirenne F, Mouro I, Le Pennec PY, et al. Twonew alleles of the RHCE gene in Black individuals:the Rhce allele ceMO and the RhcE allele cEMI. BrJ Haematol 2001;113:672-9.

Ram Kakaiya, MD, Medical Director, LifeSourceBlood Services, 1205 North Milwaukee Avenue,Glenview, IL 60025; Jill Cseri, MT(ASCP)SBB,LifeSource Blood Services, Beth Jochum,MT(ASCP)SBB, LifeSource Blood Services, and LaurieGillard, MT(ASCP)SBB, and Simone Silberman, MD,Loyola University Medical Center, Maywood, IL.

Attention SBB and BB Students: You are eligible for a free 1-year subscription to Immunohematology. Askyour education supervisor to submit the name and complete address for each student and the inclusive datesof the training period to Immunohematology, P.O. Box 40325, Philadelphia, PA 19106.

Attention: State Blood Bank Meeting OrganizersIf you are planning a state meeting and would like copies of Immunohematology for distribution, pleasecontact Mary McGinniss, Managing Editor, 4 months in advance, by phone or fax at (301) 299-7443.

C O M M U N I C A T I O N S


Letter to the Editors

HAMA (Human Anti-Mouse Antibodies) do not Cause False Positive Results in PAKPLUS

Regretfully, we wish to inform readers that a recentarticle which appeared in Immunohematology (Falsereactivity in GTI PAKPLUS ELISA kits due to thepresence of anti-mouse antibody in patient’s samples,MF Leach, JP AuBuchon, Volume 19, No.4, p 112.)contained several significant errors.

In Figures 1 and 2 the authors provide illustrationsof how a human anti-mouse antibody (HAMA) cancause both false positive and false negative results intesting done with the GTI kits. On the one hand, theyclaim that a false positive result can be caused bybinding of a HAMA to the mouse monoclonal antibodyused to immobilize platelet antigens in the well. Onthe other hand, they suggest that a false negative resultcan be caused by binding of a HAMA to the secondaryantibody-enzyme conjugate, thereby preventing it fromrecognizing a patient antibody bound to the targetantigen (resulting in “neutralization” of the secondaryantibody-enzyme conjugate).

We wish to point out that these suggestions areflawed for the following reasons:(1) The PAKPLUS kit is not a simultaneous sandwich

immunoassay (i.e., the sample is not simul-taneously incubated with both the capture andthe detection antibody reagents). As stated in thePAKPLUS direction insert, there is a wash stepbetween the addition of the patient sample andthe addition of the secondary antibody-enzymeconjugate. Thus, any unbound antibody, (includ-ing HAMA) would be removed from the test wellprior to the addition of the secondary antibody-enzyme conjugate. Therefore, a false negativeresult due to HAMA antibodies could not occurin the way described by the authors in Figure 2.

(2) The authors claim to have identified HAMAantibodies that cause false positive results witheach of the five glycoproteins immobilized in thewells. However, as described in the packageinsert, only two glycoproteins in PAKPLUS arecaptured in the wells by the use of immobilizedmonoclonal antibodies. The other three

glycoproteins are affinity purified and immobil-ized directly. Therefore, the suggestion that aHAMA can cause false positive results in theGPIb/IX, GPIV, and HLA wells is untenable.

(3) Lastly, false positive results due to HAMA are veryunlikely because the specimen diluent providedwith the GTI PAKPLUS kit already containsmouse immunoglobulin. The amount ofimmunoglobulin present in the diluent wasoptimized to prevent interference from HAMA.The use of this diluent in PAKPLUS has beentested using well-characterized samples con-taining HAMA. None was found to give apositive result in the PAKPLUS kit. This includesHAMA induced by treatment with therapeuticreagents containing mouse monoclonal anti-bodies and those naturally occurring in humanpopulations.

We would also like to add that, althoughcommercial ELISA assays are available for confirmingthe presence of HAMA antibodies, none of the samplesin this study were tested by the authors to confirm theHAMA specificity.

The authors recommended that mouse IgG shouldbe added routinely to all patient samples prior totesting. We strongly disagree with them on thispoint. Users are generally not aware of all of thecomplexities that underlie a commercial assay.Manipulating the components without a full under-standing of these complexities can easily lead toerroneous results. Finally, the mouse IgG used by theauthors was not knowingly provided by GTI for thetype of study done by the authors.

Leigh Ann Tidey, MS, MT(ASCP)SBBDirector of Operations

QA ManagerGTI

Suzette Chance, PhDDirector of Product Development

GTI


C O M M U N I C A T I O N S C O N T ’ D

Marilyn Clarke, PhDProduction/Process Scientist

GTI

RH Aster, MDSenior Investigator

Blood Research InstituteThe Blood Center of Southeast Wisconsin

Milwaukee WI

The above letter was sent to Leach and AuBuchon.They offered the following reply.

We appreciate the interest of Tidey et al.1 in ourrecent studies of the effect of heterophile antibodies inenzyme immunoassays for platelet alloantibodies.2 Wewere frankly surprised at many of their commentsinasmuch as it was a staffmember of their companythat first led us to consider the presence of thisphenomenon as an explanation for some anomalousresults we had experienced and provided (on June 9,2000) mouse immunoglobulin and instructions for itsuse to allow us to investigate this problem.

The authors of the letter contend that the washingstep before the addition of the conjugate precludes aheterophile antibody from remaining in the well.However, the capability of heterophile antibodies tobind with sufficient avidity such that they remain inthe test system to be recognized by the secondaryantibody-antigen conjugate has been seen in manyother test systems. Thus, when we encounteredunexplained reactivity in samples that were notexpected to contain platelet alloantibodies, and whenGTI’s suggestion to add (additional) mouse immuno-globulin removed this reactivity, we felt it was notunreasonable to conclude that heterophile antibodieshad been the cause. The authors of the letter alsostated that the antigens in some wells are immobilizedsuch that an anti-mouse antibody would have no target.While we are not privy to the details of the manu-facture of the kit and did not directly analyze samplesfor the presence of anti-mouse antibodies, ourobservation that the addition of murine immuno-globulin blocks otherwise inexplicable activity isprima facie evidence of the presence of anti-mouseantibody reactivity and an obvious and simple solution

to the problem we encountered. We thank GTI forsuggesting this approach to us.

The presence of mouse immunoglobulin in thespecimen diluent of this kit as routinely supplied bythe manufacturer undoubtedly helps reduce thefrequency of this problem. This amount apparentlymay not be sufficient to neutralize anti-mousereactivity in some samples. We attempted to identifyresults that might make the user of the test kitsuspicious for the unexpected effect of the heterophileantibodies that eluded neutralization in order toprompt the use of additional mouse immunoglobulinto obtain an accurate result. As this and many otherlaboratories depend on these test results to detect andcharacterize alloantibodies in order to provideeffective platelet hemotherapy, we hope that othersmay find this supplemental technique useful in certainsituations.

Researchers in our laboratory have enjoyedsuccessful collaborations with many manufacturers indeveloping and validating a wide variety of techniquesand equipment. We have found open communicationsessential in these efforts to improve our collaborativeservice to patients, and our communications to GTI in investigating the situation we were encounteringwere no different. The letter’s implication of nefariousconduct is unwarranted and unsubstantiated. Weappreciate GTI’s provision of mouse IgG to investigatethis problem after their staff had alerted us to itspossibility, and we offered GTI review of an abstract ofthe work3 (on April 9, 2002) at the same time that werequested additional mouse immunoglobulin tocontinue our investigations. We regret that themanufacturer did not pursue this opportunity.

Enzyme immunoassays remain an important toolfor the detection of platelet alloantibodies. Asadditional labs report their experiences with usingthese techniques in “real life” situations, furtherknowledge may accrue that facilitates improvement oftheir useful attributes.


References1. Tidey LA, Chance S, Clarke M, Aster RH. HAMA

(human anti-mouse antibodies) do not cause falsepositive results in PAKPLUS. Immunohematology2004;20:

2. False reactivity in GTI Pak Plus® ELISA kits due tothe presence of anti-mouse antibody in patients’samples. MF Leach, JP AuBuchon. Immuno-hematology 2003;19:112-7.

3. Leach MF, AuBuchon JP. False positive results inGTI-Pak Plus ELISA kits due to the presence ofanti-mouse antibodies in patient samples.Transfusion 2002;42:64S.

Miriam Fogg Leach, MS, MT(ASCP)SBBJames P. AuBuchon, MDDepartment of PathologyDartmouth-Hitchcock Medical CenterOne Medical Center DriveLebanon, NH 03756


Notice to Readers: Immunohematology, Journalof Blood Group Serology and Education, is printed on acid-free paper.


In February 2004, I was informed by the NationalLibrary of Medicine (NLM) that its Literature SectionTechnical Review Committee had completed afavorable review of Immunohematology and that theNLM would begin to cite and index the journal inIndex Medicus and MEDLINE. The NLM’s decision is akey milestone in the history of Immunohematology. Itmeans that articles in Immunohematology—past,present, and future—will be listed on electronic data-bases, making them readily accessible to researchers,authors, and readers worldwide. Preparing the appli-cation for the NLM’s review of Immunohematologywas an informative and inspiring experience. I wouldlike to share some of my observations of thatexperience with you.

The NLM’s Technical Review Committee haddeclined three prior applications for Immuno-hematology to be listed in Index Medicus andMEDLINE. In 2003, I approached Delores Mallory,Editor-in-Chief, and volunteered to initiate a fourthattempt. I was convinced that Immunohematology, aswell as its distinguished contributors, deserved thisrecognition and status. I drafted an extensively refer-enced appeal to the NLM’s Technical ReviewCommittee, basing arguments on the following threeobservations:(1) Immunohematology serves a unique niche among

scientific publications. It is the only journalworldwide that limits its focus to blood groupserology. No other journal, including Transfusion,Blood, or Vox Sanguinis, targets a readership ofblood group serologists and immunohematologiststo provide up-to-date practical information.

(2) Immunohematology’s editors and editorial boardare the international leaders in blood groupserology. I listed the members of the editorialboard and, one by one, cited examples of eachmember’s scientific contributions and expertise inthe field of blood group immunohematology.

(3) I provided an article-by-article critique of the threemost recent issues of Immunohematology,explaining to the reviewers (I assumed none were“blood bankers”) what each article contributed toour discipline. I identified each of the authors and,I must admit, the list of contributors was trulydistinguished and impressive.

Compiling this information to illustrate Immuno-hematology’s high standards and quality was both aninspiring and a sentimental experience. I recall, asthough it were only yesterday, that Delores Malloryarrived at the National Headquarters of the AmericanRed Cross to direct our newly relocated NationalReference Laboratories in Bethesda, MD. Herresponsibilities included coordination of the NationalReference Laboratory Committee, which published theRed Cell Free Press. Delores had a vision for the RedCell Free Press. She believed that there was a need andan opportunity to dress up the newsletter and establisha quarterly journal to be named Immunohematology,Journal of Blood Group Serology and Education. In1986, she recruited Mary McGinniss, who had recentlyretired from a distinguished career at the NationalInstitutes of Health, to assist in scientific editing.Within a few years, Mary was appointed ManagingEditor of Immunohematology. Volume 1, number 1 ofImmunohematology appeared in September 1984.Delores announced the change and her vision in a page1 editorial, as follows:

“The title of our newsletter—Red Cell Free Press—has been put to rest and in its place is a new name—Immunohematology—and a new direction. Therewill be more articles relevant to blood groupserology and education . . .”

Contributors to this first issue included Rebecca H.Buckley, MD; Herbert A. Perkins, MD; Peter Issitt, PhD;Kay Beattie, MT(ASCP)SBB;Tabbie Bolk, MT(ASCP)SBB;


Letter to the Editor-in-Chief

Immunohematology to be listed in Index Medicus and MEDLINE

S. Gerald Sandler, MD



Judy Robinson; Mary N. Crawford, MD; Denise A.Valko,MS, MT(ASCP)SBB; Joan Barker, MT(ASCP)SBB; RogerCollins; and Dorothy C. Malamut, SBB(ASCP). Thejournal’s standards were set high from the beginningand these high standards have been maintained for 80quarterly issues during the past 20 years. Happy 20thbirthday, Immunohematology! Those of us who have

had an opportunity to share in your progress are proudof your achievements and progress.

S. Gerald Sandler, MDAssociate Medical Editor, Immunohematology

Professor of Medicine and PathologyGeorgetown University Medical Center

Washington, DC

IMPORTANT NOTICE ABOUT MANUSCRIPTSFOR

IMMUNOHEMATOLOGY

Mail all manuscripts (original and 2 copies) to Mary H. McGinniss, Managing Editor, 10262 Arizona Circle,Bethesda, MD 20817. In addition, please e-mail a copy to Marge Manigly at [email protected]



Letters From the Editor-in-Chief

Ortho dedication

The final 20th anniversary issue

14 years! Thank you Ortho-Clinical Diagnostics forthe financial support needed to publish the Septemberissue of Immunohematology for the 14th year in arow! It demonstrates once again the commitmentOrtho-Clinical Diagnostics has to the support ofeducation and to the advancment of scientificknowledge and original ideas.

Please let your Ortho-Clinical Diagnosticsrepresentatives know how much you appreciate theirsupport of Immunohematology. It means a great dealto the publication of this journal! Thank you!

The final 20th anniversary issue is a grand finalethat you should not miss! Ralph R.Vassallo, MD, SeniorMedical Director of the American Red Cross, Penn-Jersey Region, is the final guest editor and he has a four-star, two-thumbs-up issue to finish the year. There willbe five invited reviews: The Function of Blood Groupsby Jill Storry, PhD, FIBMS; Immunohematologic Aspectsof Transplantation by Jeffrey McCullough, MD, et al.;IgA Anaphylactic Transfusion Reactions: I. LaboratoryDiagnosis, Incidence, and Supply of IgA-DeficientProducts by Ralph R.Vassallo, MD; IgA AnaphylacticTransfusion Reactions: II. Clinical Diagnosis andManagement by S. Gerald Sandler, MD; and TheAmerican Rare Donor Program by Ann Church,CynthiaFlickinger, and Tammy Petrone. In addition, there willbe three interesting and informative original papers.This will be an issue for you to read and reread!

In addition, a letter that Dr. Tibor Greenwaldgraciously sent to me describes some of his memoriesof the early days of the reference laboratory atAmerican Red Cross National Headquarters and thestart of two rare donor programs in the United States.The first program was the Rare Donor File of theAmerican Association of Blood Banks and the second

was the Rare Donor Registry of the American RedCross. Dr Greenwald started both programs. The twoprograms were combined into the American RareDonor Program which is now housed in the Penn-Jersey American Red Cross. One of the invited reviewarticles in the December issue is a report of theAmerican Rare Donor Program by Ann Church and co-authors.

Dr Greenwald was honored this year by theAmerican Association of Blood Banks with the KarlLandsteiner Memorial Award, the highest award givenby the association. Congratulations, Dr. Greenwald.Well done and thank you for your contributions totransfusion medicine and the American Rare DonorProgram.

Finally, the December issue will contain somereprints from early issues of the Red Cell Free Press,including a poem by a very famous blood banker. Itmay be a good thing he concentrated on immuno-hematology! He might be very thin now if he had keptto poetry!

Delores MalloryEditor-in-Chief


Elution of Anti-SA patient whose serum contained multiple allo

antibodies including anti-S was transfused with 2 unitsof blood that lacked all appropriate antigens except S.The direct antiglobulin test post transfusion wasweakly positive with broad spectrum and anti-IgG anti-globulin serum but negative with anti-complementreagent. The anti-S was IgG in nature, as determined bymonospecific (anti-IgG and anti-IgM) antiglobulin sera.

An ether eluate prepared from the patient’s posttransfusion cell sample possessed no blood groupantibody activity; however a heat eluate containedclear cut anti-S specificity. Has anyone else experiencedifficulty in recovering cell-bound anti-S by the etherelution method? We are interested in your findings.

M. ReidReference LaboratoryCentral California Regional Red CrossBlood ProgramSan Jose, California

Reprinted as published in the American Red CrossReference Laboratory Newletter. No. 1., October 15,1976.

S P E C I A L S E C T I O N

Hemolytic Transfusion ReactionA patient with hereditary hemorrhagic telangi-

ectasia was admitted to a local hospital because of G.I.bleeding. The patient had a ten year history ofhemoglobinuria post transfusion. Our pretransfusionsamples were negative by standard and enzymetechniques. The patient received 3 units of packedcells uneventfully. Seven days later he experiencedhemoglobinuria. Serum samples tested within the nextweek demonstrated weak unidentified albumin-antiglobulin reactivity. In an idle conversation with atechnologist from another Blood Center in the area, Ilearned that a trypsin-only anti-hr”(e) was demon-strated in the patient’s serum 3 years previously. Thereport had gone to the clinician who neglected to passthat information on to the Blood Center or transfusionservice. The patient had not been transfused since thattime.

The patient eventually developed an identifiableanti-Chido. The serum was tested in numerous labora-tories and the anti-hr”(e) could not be detected bystandard techniques which included many differentenzyme techniques. The anti-hr”(e) has been detectedusing the microtiter technique and using theautoanalyzer. The patient has received approximately40 units of hr”(e) negative, Chido positive blood in thelast three years without any hemoglobinuria or othersigns of a transfusion reaction.

S. EllisorReference Laboratory


Notice to Readers: All articles published,including communications and book reviews,reflect the opinions of the authors and do notnecessarily reflect the official policy of theAmerican Red Cross.


S P E C I A L S E C T I O N C O N T ’ D

Controls for Enzyme PremodificationThe following controls have proven useful in our

laboratory for demonstrating adequate enzymepremodification of red cells.

1. Using an antibody/antigen system known to beenhanced by enzyme treatment. A weakly reactiveantibody, with no saline reactivity, is selected andtested against cells possessing the weakestavailable form of the appropriate antigen.For example:anti-c and heterozygous – positive cellsanti-D and R1r cellsanti-Lea and Le(a+b-) cellsanti-Leb and Le(a-b+) cells

2. Using an antibody/antigen system known to bedenatured by enzyme treatment. A stronglyreacting antibody is selected and tested againstcells possessing the strongest available form of theappropriate antigen.For example:anti-M and MM cellsanti-Fya and Fy(a+b-) cells

3. Using a substance that demonstrates reduction inred cell sialic acid levels.For example:Polybrene – untreated cells aggregate

Enzyme premodified cells donot aggregate

Soybean extract – untreated cells do not aggregateEnzyme premodified cellsaggregate

We do not use milk flocculation or X-ray paperdigestion methods for quality control of enzymeactivity since they only measure proteolytic activity.

M. ReidReference LaboratoryCentral California Regional Red CrossBlood ProgramSan Jose, California


IMMUNOHEMATOLOGY IS ON THE WEB!

www.redcross.org/pubs/immuno

Password “2000”

For more information or to send an e-mail message “To the editor”

[email protected]


Monoclonal antibodies available at no cost. TheLaboratory of Immunochemistry at the New YorkBlood Center has developed a wide range ofmonoclonal antibodies (both murine and humanized)that are useful for screening for antigen–negativedonors and for typing patients’ RBCs with a positiveDAT. Monoclonal antibodies available include anti-M,-Fya, -Fyb, -K, -k, -Kpa, -Jsb, -Dob, -Wrb, and –Rh17. For acomplete list of available monoclonal antibodies,pleasesee our Web site at www.nybloodcenter.org/framesets/FS-4C7.htm. Most of those antibodies aremurine IgG and, thus, require the use of anti-mouse IgGfor detection, i.e, anti-K, -k, and -Kpa. Some are directlyagglutinating (anti-M, -Wrb, and -Rh17), and a few havebeen humanized into the IgM isoform and are directlyagglutinating (anti-Jsb and -Fya). The monoclonalantibodies are available at no charge to anyone whorequests them. Contact: Marion Reid ([email protected]) or Gregory Halverson ([email protected]), New York Blood Center, 310 East67th Street, New York, NY 10021.

HEMATOLOGÍA HABANA′′ 2005–First Announcement. The 5th National Congress and the 7th Latin AmericanMeeting in Hematology, Immunology, and TransfusionMedicine will present a scientific program at theInternational Conference Center, Havana, Cuba, May16–20, 2005. A preliminary program lists malignanthemopathies, disorders of RBC membranes,immunotherapy, histocompatibility, immunohema-tology, hemolytic disease of the newborn, regenerativemedicine, and blood components as some of thetopics. For more information contact: Prof. José M.Ballester, President, Organizing Committee,Hematology Habana′ 2005,Apartado 8070,Ciudad de laHabana, CP 10800, Cuba, e-mail: [email protected];Web site: www.loseventos.cu/hematologia2005.

A N N O U N C E M E N T S

Annual Symposium. The National Institutes ofHealth, Department of Transfusion Medicine, will holdtheir 24th annual symposium, Immunohematologyand Blood Transfusion, on September 23, 2004. Thesymposium is co-hosted by the Greater Chesapeakeand Potomac Region of the American Red Cross and isfree of charge. Advance registration is encouraged. Formore information and registration, Contact: KarenByrne, NIH/CC/DTM, Bldg. 10/Rm. 1C711, 10 CenterDrive, MSC 1184, Bethesda, MD 20892-1184; e-mail:[email protected]; or visit our Web site:www.cc.nih.gov/dtm>education.


A N N O U N C E M E N T S C O N T ’ D

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 2004 and

will last for one year. A part-time option lasting three years is also available. There may also be

opportunities to continue studies for PhD or MD following MSc. The syllabus is organised jointly by The

Bristol Institute for Transfusion Sciences and the University of Bristol, Department of Transplantation

Sciences. It includes:

• Scientific principles underlying transfusion and transplantation

• Clinical applications of these principles

• Practical techniques in transfusion and transplantation

• Principles of study design and biostatistics

• An original research project

Applications 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.bloodnet.nbs.nhs.uk/ibgrl/MSc/MScHome.htm

For further details and application forms please contact:

Professor Ben Bradley

University of Bristol, Department of Transplantation Sciences

Southmead Hospital, Westbury-on-Trym, Bristol BS10 5NB, England

FAX +44 1179 595 342, TELEPHONE +44 1779 595 455, E-MAIL: [email protected]


NATIONAL REFERENCE LABORATORY FORBLOOD GROUP SEROLOGY

Immunohematology ReferenceLaboratory

AABB,ARC, New York State, and CLIA licensed(215) 451-4901—24-hr. phone number

(215) 451-2538—Fax

American Rare Donor Program(215) 451-4900—24-hr. phone number

(215) 451-2538—[email protected]

Immunohematology(215) 451-4902—Phone, business hours

(215) 451-2538—[email protected]

Quality Control of Cryoprecipitated-AHF(215) 451-4903—Phone, business hours

(215) 451-2538—Fax

Granulocyte Antibody Detection and Typing

• Specializing in granulocyte antibody detectionand granulocyte antigen typing

• Patients with granulocytopenia can be classifiedthrough the following tests for proper therapyand monitoring:—Granulocyte agglutination (GA)—Granulocyte immunofluorescence (GIF)—Monoclonal Antibody Immobilization of

Granulocyte Antigens (MAIGA)

For information regarding services, call Gail Eiberat: (651) 291-6797, e-mail: [email protected],

or write to:Neutrophil Serology Reference Laboratory

American Red CrossSt. Paul Regional Blood Services

100 South Robert StreetSt. Paul, MN 55107

CLIA LICENSED

National Platelet Serology ReferenceLaboratory

Diagnostic testing for:

• Neonatal alloimmune thrombocytopenia (NAIT)

• Posttransfusion purpura (PTP)

• Refractoriness to platelet transfusion

• Heparin-induced thrombocytopenia (HIT)

• Alloimmune idiopathic thrombocytopenia

purpura (AITP)

Medical consultation available

Test methods:

• GTI systems tests

—detection of glycoprotein-specific platelet

antibodies

—detection of heparin-induced antibodies (PF4

ELISA)

• Platelet suspension immunofluorescence test

(PSIFT)

• Solid phase red cell adherence (SPRCA) assay

• Monoclonal antibody immobilization of platelet

antigens (MAIPA)

For information, e-mail: [email protected]

or call:

Maryann Keashen-Schnell

(215) 451-4041 office

(215) 451-4205 laboratory

Sandra Nance

(215) 451-4362

Scott Murphy, MD

(215) 451-4877

American Red Cross Blood ServicesMusser Blood Center

700 Spring Garden StreetPhiladelphia, PA 19123-3594

CLIA LICENSED

A D V E R T I S E M E N T S


A D V E R T I S E M E N T S C O N T ’ D

IgA/Anti-IgA Testing

IgA and anti-IgA testing is available to do the following:• Monitor known IgA-deficient patients• Investigate anaphylactic reactions• Confirm IgA-deficient donors

Our ELISA assay for IgA detects antigen to 0.05 mg/dL.

For information on charges and sample requirements, call (215) 451-4351, e-mail:

[email protected],or write to:

American Red Cross Blood ServicesMusser Blood Center

700 Spring Garden StreetPhiladelphia, PA 19123-3594

ATTN: Ann Church

CLIA LICENSED

Reference and Consultation Services

Red cell antibody identification and problem resolution

HLA-A, B, C, and DR typing

HLA-disease association typing

Paternity testing/DNA

For information regarding our services, contact

Zahra Mehdizadehkashi at (503) 280-0210, or

write to:

Pacific Northwest Regional Blood Services

ATTENTION: Tissue Typing Laboratory

American Red Cross

3131 North Vancouver

Portland, OR 97227

CLIA LICENSED, ASHI ACCREDITED


SCIENTIFIC ARTICLES, REVIEWS, AND CASE REPORTSBefore submitting a manuscript, consult current issues of

Immunohematology for style. Type the manuscript on white bondpaper (8.5" × 11") and double-space throughout. Number the pagesconsecutively in the upper right-hand corner, beginning with thetitle page. Each component of the manuscript must start on a newpage in the following order:1. Title page2. Abstract3. Text4. Acknowledgments5. References6. Author information7. Tables—see 6 under Preparation8. Figures—see 7 under Preparation

Preparation of manuscripts1. Title page

A.Full title of manuscript with only first letter of first word capitalized (bold title)

B. Initials and last name of each author (no degrees; all CAPS),e.g., M.T. JONES and J.H. BROWN

C.Running title of ≤ 40 characters, including spacesD.3 to 10 key words

2. AbstractA.One paragraph, no longer than 300 wordsB. Purpose, methods, findings, and conclusions of study

3. Key words—list under abstract4. Text (serial pages)

Most manuscripts can usually, but not necessarily, be divided intosections (as described below). Results of surveys and reviewpapers are examples that may need individualized sections.A. Introduction

Purpose and rationale for study, including pertinent back-ground references.

B. Case Report (if study calls for one)Clinical and/or hematologic data and background serology.

C.Materials and MethodsSelection and number of subjects, samples, items, etc. studiedand description of appropriate controls, procedures, methods,equipment, reagents, etc. Equipment and reagents should beidentified in parentheses by model or lot and manufacturer’sname, city, and state. Do not use patients’ names or hospitalnumbers.

D.ResultsPresentation of concise and sequential results, referring to perti-nent tables and/or figures, if applicable.

E. DiscussionImplications and limitations of the study, links to other studies;if appropriate, link conclusions to purpose of study as stated inintroduction.

5. AcknowledgmentsAcknowledge those who have made substantial contributions tothe study, including secretarial assistance; list any grants.

6. ReferencesA. In text, use superscript, arabic numbers.B. Number references consecutively in the order they occur in

the text.C.Use inclusive pages of cited references, e.g., 1431–7.D.Refer to current issues of Immunohematology for style.

7. TablesA.Head each with a brief title, capitalize first letter of first word

(e.g., Table 1. Results of ...), and use no punctuation at the endof the title.

B. Use short headings for each column needed and capitalize firstletter of first word. Omit vertical lines.

C.Place explanations in footnotes (sequence: *, †, ‡, §, ¶, **, ††).8. Figures

A.Figures can be submitted either by e-mail or as photographs (5″ × 7″ glossy).

B. Place caption for a figure on a separate page (e.g.,Fig.1.Resultsof ...), ending with a period. If figure is submitted as a glossy,place first author’s name and figure number on back of eachglossy submitted.

C.When plotting points on a figure, use the following symbols ifpossible: ll l ss s nn n.

9. Author informationA.List first name, middle initial, last name, highest academic

degree, position held, institution and department, and complete address (including zip code) for all authors. Listcountry when applicable.

SCIENTIFIC ARTICLES AND CASE REPORTS SUBMITTED AS LETTERS TO THE EDITOR

Preparation1. Heading—To the Editor:2. Under heading—title with first letter capitalized.3. Text—write in letter format (paragraphs).4. Author(s)—type flush right; for first author: name, degree,

institution, address (including city, state, ZIP code, and country);for other authors: name, degree, institution, city, and state.

5. References—limited to ten.6. One table and/or figure allowed.

ImmunohematologyJOURNAL OF BLOOD GROUP SEROLOGY AND EDUCATION

Instructions for Authors

Send all submissions (original and two copies) to:

Mary H. McGinniss, Managing Editor, Immunohematology,

10262 Arizona Circle, Bethesda, MD 20817 and e-mail your

manuscript to Marge Manigly at [email protected]


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 EducatorsSupervisors of Reference Laboratories Research Scientists Consumer Safety OfficersQuality 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

Fact #1: In recent years, the average SBB exam pass rate is only 38%.Fact #2: In recent years, greater than 73% of people 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 BankTechnology Program

Becoming a Specialist in Blood Banking (SBB)

Contact the following programs for more information:PROGRAM CONTACT NAME CONTACT INFORMATION

Walter Reed Army Medical Center William Turcan 202-782-6210;[email protected]

Transfusion Medicine Center at Florida Blood Services Marjorie Doty 727-568-5433 x 1514; [email protected]

Univ. of Illinois at Chicago Veronica Lewis 312-996-6721; [email protected]

Medical Center of Louisiana Karen Kirkley 504-903-2466; [email protected]

NIH Clinical Center Department of Transfusion Medicine Karen Byrne 301-496-8335; [email protected]

Johns Hopkins Hospital Christine Beritela 410-955-6580; [email protected]

ARC-Central OH Region, OSU Medical Center Joanne Kosanke 614-253-2740 x 2270; [email protected]

Hoxworth Blood Center/Univ. of Cincinnati Medical Center Catherine Beiting 513-558-1275; [email protected]

Gulf Coast School of Blood Bank Technology Clare Wong 713-791-6201; [email protected]

Univ. of Texas SW Medical Center Barbara Laird-Fryer 214-648-1785; [email protected]

Univ. of Texas Medical Branch at Galveston Janet Vincent 409-772-4866; [email protected]

Univ. of Texas Health Science Center at San Antonio Bonnie Fodermaier SBB Program: 210-358-2807,Linda Smith [email protected]

MS Program: 210-567-8869; [email protected]

Blood Center of Southeastern Wisconsin Lynne LeMense 414-937-6403; [email protected]

Additional information can be found by visiting the following Web sites: www.ascp.org, www.caahep.org, and www.aabb.org

Immunohematology - American Red Cross...mefenamic acid mefloquine melphalan mephenytoin methoin...

Documents

Transcript of Immunohematology - American Red Cross...mefenamic acid mefloquine melphalan mephenytoin methoin...