autoantibodies and laboratory testing best practices

REVIEW Warm-reactive (immunoglobulin G) autoantibodies and laboratory testing best practices: review of the literature and survey of current practice Alyssa Ziman,1 Claudia Cohn,2 Patricia M. Carey,3 Nancy M. Dunbar,4 Mark K. Fung,5 Andreas Greinacher,6 Simon Stanworth,7 Nancy M. Heddle,8 Meghan Delaney,9,10 and the Biomedical Excellence for Safer Transfusion (BEST) Collaborative

BACKGROUND: Warm-reactive autoantibodies (WAAs) are the most common cause of autoimmune hemolytic anemia (AIHA) and can also be present without clinically significant hemolysis. WAAs complicate immunohematological testing, yet there is no commonly accepted approach to laboratory evaluation and red blood cell (RBC) selection. STUDY DESIGN AND METHODS: We searched PubMed/Cochrane Central for articles that described testing methodology and blood selection for patients with WAAs. We developed a 31-question survey regarding local practice for immunohematology testing and RBC selection in patients with WAAs (with or without AIHA). RESULTS: Eighty-six studies met the inclusion criteria and the aims of this review. Most of the literature was comprised of retrospective studies that often did not correlate laboratory results with clinical findings. Evidence-based protocols to guide testing and RBC selection for transfusion in patients with WAAs are lacking. Individuals representing 54 laboratories completed the survey. The responses indicated that numerous methodologies are used to identify underlying alloantibodies: 75% of respondents use autoadsorption; in patients who have a recent history of transfusion, 76% of respondents use alloadsorption; 58% of respondents perform direct antiglobulin testing (DAT) each time the indirect antiglobulin test is positive; and 48% perform eluate studies at the initial identification of WAAs. Responding laboratories may use phenotyping (98%) or genotyping (80%) at some point in the work-up. Seventyfive percent of respondents provide phenotype-matched or genotype-matched RBCs for transfusion. CONCLUSION: There is wide variability in immunohematology testing and RBC selection practices for patients who have WAAs (with or without AIHA). Future studies are needed to evaluate and compare the effectiveness of different testing algorithms and transfusion strategies.

A

utoimmune hemolytic anemia (AIHA) is an autoimmune disease in which autoantibodies directed toward red blood cell (RBC) antigens lead to accelerated RBC destruction. In the

ABBREVIATIONS: AIHA 5 autoimmune hemolytic anemia; CAD 5 cold agglutinin disease; HBTS(s) 5 hospital-based transfusion services; IRL(s) 5 immunohematology reference laboratory(ies); PCH 5 paroxysmal cold hemoglobinuria; PNH 5 paroxysmal nocturnal hemoglobinuria; SPRCA 5 solid-phase red cell adherence; WAA(s) 5 warm-reactive antibody(ies); XM 5 crossmatch; ZZAP 5 dithiothreitol plus enzyme. From the 1Wing-Kwai and Alice Lee-Tsing Chung Transfusion Service, Department of Pathology and Laboratory Medicine, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, California; the 2Department of Laboratory Medicine and Pathology, University of Minnesota, Minneapolis, Minnesota; the 3Hoxworth Blood Center and Department of Pathology and Laboratory Medicine, University of Cincinnati, Ohio; the 4Departments of Pathology and Medicine, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire; the 5Department of Pathology and Laboratory Medicine, University of Vermont Medical Center, Burlington, Vermont; the 6Department of Immunology and Transfusion Medicine, Ernst-Moritz-Arndt-University, Greifswald, Germany; the 7National Health Service (NHS) Blood and Transplant/ Oxford University Hospitals NHS Trust, John Radcliffe Hospital; and the Radcliffe Department of Medicine, University of Oxford, Oxford, United Kingdom; the 8Faculty of Health Sciences, Department of Medicine, Canadian Blood Services, McMaster University, and Centre for Innovation, Hamilton, Ontario, Canada; and 9Bloodworks NW; and the 10Department of Laboratory Medicine, University of Washington, Seattle, Washington. Address reprint requests to: Alyssa Ziman, MD, Department of Pathology and Laboratory Medicine, David Geffen School of Medicine at UCLA, 757 Westwood Plaza, B403, Los Angeles, CA 90095-1732; e-mail: [email protected]. Volume 57, February 2017 TRANSFUSION 463

ZIMAN ET AL.

majority of patients with AIHA (approximately 80%), the pathologic antibodies are immunoglobulin G (IgG) class and have maximal reactivity at 378C. Autoantibodies with these characteristics are designated as warm-reactive autoantibodies (WAAs).1-4 The incidence of AIHA in adults is 1 to 3 cases per 100,000 per year, which is far less than the reported incidence of WAA detection in antibody identification work-ups (up to 17%), because not all patients with a detectable WAA have clinical symptoms of hemolysis.5-7 In either setting, laboratory evaluations of WAAs are time consuming, expensive, and can be technically challenging. When AIHA is clinically suspected, the detection of circulating WAAs in a patient’s plasma and/or of WAAs that coat circulating RBCs is required to support the diagnosis. In any patient with WAA, the detection of alloantibodies and the selection of appropriate RBC units for transfusion can be problematic, because WAAs may mask the presence of alloantibodies.8,9 Although there are descriptions of serologic approaches for pretransfusion testing, evidence-based protocols to guide the testing practice and selection of RBC units for transfusion in patients who have WAAs (with or without AIHA) are lacking.1,10,11 A comprehensive review of the literature was performed to assess and describe the published practice for immunohematology testing and blood selection in patients with WAAs and to identify knowledge gaps to focus future research programs. In addition, a survey focused on testing methodology and blood selection was administered to collect data regarding current immunohematology practice at immunohematology reference laboratories (IRLs) and hospital-based transfusion services (HBTSs). In this report, the literature review and survey results are compiled to provide practice recommendations supplemented by expert opinion, until comparative evaluations in the form of clinical studies can be undertaken.

tion of WAAs and blood transfusion in patients with WAAs and AIHA. The electronic databases of PubMed (1966 to January 2014) and Cochrane Central (1980 to January 2014) were queried using the terms “test for identifying warm autoantibodies” and “blood transfusion for warm autoantibodies” as search strategies (Fig. 1, Appendix S1, available as supporting information in the online version of this paper). Manual searches of bibliographies were performed to identify additional articles. The search strategy was intentionally broad to minimize the risk of omitting relevant articles. Title and abstract screening was performed by two reviewers (AZ, MD) based on a set of predefined inclusion criteria, which included human (pediatric and adult patients) and English-language publications. Exclusion criteria included animal studies, non-English publications, editorials, reviews, studies that described researchbased testing or testing that was obsolete, as well as specific diseases not associated with WAAs, such as cold agglutinin disease (CAD), paroxysmal cold hemoglobinuria (PCH), paroxysmal nocturnal hemoglobinuria (PNH), drug-induced immune hemolytic anemia, direct antiglobulin test (DAT)-negative hemolytic anemia, Evan’s syndrome, and hemolytic disease of the fetus/newborn (HDFN). Eligible publications were grouped into the following categories: test methodology, alloantibody identification and adsorption, DAT and elution, patient phenotyping/genotyping, RBC selection and transfusion, other testing, general overview and pathophysiology, and case reports and series with less than ten patients. Each category was assigned to an author for confirmation of eligibility/inclusion and assessment of relevance to the practical application in AIHA due to WAAs; the authors evaluated studies for design and summarized key findings (Fig. 1).

Survey

MATERIALS AND METHODS Literature review We performed a comprehensive literature review to identify all publications on laboratory testing for the identificaThe authors are scientific or guest members of the Biomedical Excellence for Safer Transfusion (BEST) Collaborative, an international research organization that works collaboratively to improve transfusion-related services through standardization of analytic techniques, development of new procedures, systematic review of evidence, and execution of clinical and laboratory studies (bestcollaborative.org). Received for publication June 20, 2016; revision received September 1, 2016; and accepted September 10, 2016. doi:10.1111/trf.13903 C 2016 AABB V

TRANSFUSION 2017;57;463–477 464 TRANSFUSION Volume 57, February 2017

To better understand the current state of practice, a 31question survey was designed to collect data about immunohematology testing and RBC selection in patients with WAA (with or without AIHA) (Appendix S2, available as supporting information in the online version of this paper). The survey was administered using SurveyMonkey (http:// www.surveymonkey.com) in September 2013. E-mail invitations were sent specifying that each institution should be represented by only one response in order to avoid duplication; institutions that were invited to participate included members of the Biomedical Excellence for Safer Transfusion (BEST) Collaborative (n 5 50), America’s Blood Centers (ABC) Technical Directors (n 5 68), the College of American Pathologists (CAP) Transfusion Committee (n 5 21), and European Blood Alliance (EBA) physician members (n 5 25). Respondents were asked to categorize their institution as an HBTS, IRL, or other type of laboratory.

AUTOANTIBODIES AND LABORATORY TESTING

RESULTS Literature search In total, 1319 citations were identified by the initial literature search. After a review of the titles according to inclusion/exclusion criteria, 937 citations were excluded. Next, abstracts and full texts were evaluated to determine applicability to study aims; and, after a review of reference lists from the eligible publications to identify additional citations, 86 studies were included and regrouped into the most appropriate categories (Fig. 1). The sections below represent a synopsis of the existing body of literature and do not necessarily represent the opinions of the author group.

Test methodology (primary screening and detection of underlying alloantibodies) Twenty-seven studies focused on testing methodology, including adsorption studies. These studies were predominantly retrospective case reports and case series that concentrated on the technical evaluation of methods, such as tube and/or column agglutination technology (CAT) using various media, with and without enzyme-treated RBCs. The earliest reports of immunohematology testing protocols in the AIHA patient population describe a “hemolytic panel,” which includes a DAT (anti-IgG and anti-serum complement 3 [anti-C3]), eluate with a positive anti-IgG DAT, antibody identification of serum and eluate, use of enzyme-treated and untreated RBC panels at 378C and 48C, and screening for PCH and PNH.12-14 Over the years, there has been an evolving consensus in the published literature that demonstrations of an autoantibody in the eluate and possibly the serum (with 80% demonstrating WAA in both) and of reactivity with autologous RBCs are necessary to support the clinical suspicion for AIHA.1,15,16 When WAA is present, the detection of underlying alloantibodies is paramount to safe transfusion, especially in patients who have AIHA and a history of pregnancy and/or multiple (five or more) transfusion episodes.7,17-26 In the presence of WAAs, reports have demonstrated underlying, clinically significant alloantibodies in 12% to 40% of patient specimens.17,19-21,23,27,28 The most common alloantibody specificities in patients with WAAs include antigens from the Rh (50%) and Kell (20%) antigen systems, with anti-E alloantibodies being most common; these studies did not exclude mimicking autoantibodies, because adsorption/elution studies were not performed.17,18,20,24,29 Approaches used to detect alloantibodies include adsorption and dilution studies. Most authors prefer autoadsorption over alloadsorption due to the risk of adsorbing out alloantibodies against a high prevalence or variant antigen. It is recognized that autoadsorption may not be suitable when there has been a recent pregnancy or transfusion, or when adequate amounts of patient

RBCs are not available.1,27,30-33 Conventional adsorption procedures require pretreatment of adsorbing cells.21 Various techniques (with and without enzyme-treated RBCs and potentiating media) are reported in the literature.30,31,34-36 Adding potentiating media (i.e., polyethylene glycol [PEG], low ionic strength saline [LISS]) to the adsorption process improves efficiency with minimal impact on sensitivity for the detection of underlying alloantibodies.30,34 The use of PEG has the advantage of eliminating the cost and time of pretreating the adsorbing, allogeneic RBCs.35 However, several investigators caution that the presence of PEG can lead to the disappearance/ decreased reaction strength of alloantibodies after adsorption and storage of the adsorbed mixture due to protein precipitation.36-39 When Øyen et al. compared the dilution technique with alloadsorption, they demonstrated that a one-infive dilution was an efficient way to screen for alloantibodies and that no alloantibodies were missed (119 samples).40 However, other investigators have demonstrated that the dilution technique was not as sensitive as the adsorption method, because it failed to detect underlying clinically significant alloantibodies (particularly when these alloantibodies were of lower titer than the WAAs).1,18,34 Given the lack of consensus, authors have surmised that a dilution method may be considered in a patient who requires urgent transfusion, with adsorption studies performed retrospectively.34 The dilution technique, in combination with RBC phenotyping, may be useful for discriminating WAAs with mimicking specificity from alloantibodies.41

DAT and elution studies Forty studies focused on DAT and elution testing. Several authors demonstrated that the traditional tube DAT requires 200 to 500 IgG molecules per RBC for detection and that, in patients with AIHA, the mean IgG load per RBC is 3435 (range 206-20,000). The DAT is positive in 1% to 15% of hospitalized patients and in 1 per 7000 to 14,000 healthy blood donors.42-44 The positive and negative predictive values of positive and negative DATs in determining immune-mediated hemolysis are 83% and 99%, respectively; however, in the absence of hemolysis, the predictive value of a positive DAT is only 1.4%.45,46 Thus, DAT results must be interpreted in conjunction with the clinical picture and laboratory data.47,48 Authors report different DAT techniques, including tube testing, CAT (Gel-DAT), and solid-phase red cell adherence (SPRCA, affinity microcolumn), with Gel-DAT being the most sensitive.49-53 Less commonly used methods include flow cytometry DAT; two-stage, enzyme-linked DAT for the detection of complement; immune-labeling or radioactive labeling; and enzyme-linked immunosorbent assay (ELISA)/enzyme-linked antiglobulin test/enzyme Volume 57, February 2017 TRANSFUSION 465

ZIMAN ET AL.

and and and

and

Fig. 1. Article search and selection results. *Upon review of the “eligible citations,” reviewers could recategorize the citation into a more appropriate category. Evans 5 Evan’s syndrome.

immunoassay techniques used to quantitate the amount of IgG on the RBC.54-64 Several studies examined the value of DAT strength, IgG subclass, and testing with CAT and ELISA/enzymelinked antiglobulin tests. Although the majority of antibodies are IgG1 and IgG3, authors suggest that there may be value in evaluating DAT strength and determining IgG subclass given the correlation with severity of hemolysis, disease outcome, and potential use in monitoring disease activity.65-70 Elution testing in the presence of a positive DAT may be an essential component of a serologic work-up in the setting of a WAA. Johnston and Belota presented an algorithm to determine when to perform elution studies in the setting of WAAs after finding that routine elution studies for positive DATs were nonproductive. The algorithm correlates laboratory, transfusion, and clinical data to identify potential delayed hemolytic 466 TRANSFUSION Volume 57, February 2017

transfusion reactions (HTR) and possible AIHA (Fig. 2).71 Similarly, Nathalang and colleagues advocate for elution studies in patients with positive DATs and a history of transfusion. Furthermore, in patients with a history of recent transfusion, these authors demonstrate that dilution of eluates that exhibit panagglutination could identify underlying alloantibodies that were not detected in the patients’ sera.72 However, concerns about missing clinically significant alloantibodies when testing diluted serum could be applied to the dilution of eluates (particularly when alloantibodies are of low titer).34 Eluate preparation can be accomplished by different methods: xylene, glycine-HCl/ethylenediaminetetraacetic acid (EDTA), acid/glycine, heat, dithiothreitol plus enzyme (ZZAP), and chloroquine diphosphate.73-78 On comparative analysis, the glycine-HCl/EDTA elution method is most efficient in removing antibody from the RBC surface

AUTOANTIBODIES AND LABORATORY TESTING

to characterize the antibody in the eluate.78-80 After elution, tube agglutination, CAT, and SPRCA assays are effective for detecting antibodies in the eluate. CAT methodology offers the added benefits of a smaller specimen volume requirement and less subjectivity with interpretation.49,81 Various studies have investigated the correlation between hemolysis and DAT strength.65-69 Free autoantibodies in the serum and a DAT strength of reactivity 2 correlated with clinical severity of AIHA in some studies, whereas others did not find this correlation.15,29,56,57,62 Other authors have observed that a positive DAT may persist in patients with AIHA who are in remission.82-84 Thus, the value of monitoring patients who have AIHA either with tube tests or ELISAs remains controversial.85

Phenotyping/genotyping Eight publications discussed RBC phenotyping and/or genotyping in patients with AIHA. RBC phenotyping is performed to confirm the identity of suspected alloantibodies, determine the specificity of potential additional antibodies that may be formed, and aid in the selection of adsorbing cells or antigen-matched RBC units for transfusion.86,87 Due to the possibility of variant antigen expression, for example, with the RhD antigen, RBC phenotyping results should be considered in light of the possibility of underlying genetic variation.33 Obtaining an accurate serologic phenotype using hemagglutinationbased methods in this patient population can be challenging when RBCs are coated with IgG or in those who have recently received a transfusion. Methods to facilitate serologic phenotyping under these circumstances include using direct agglutinating antisera, IgG antisera after elution of the autoantibody, reticulocyte-separation by differential centrifugation, or flow cytometry.74-76,79,88 Several techniques have been proposed for autoantibody removal from the surface of circulating RBCs, including the ZZAP treatment, which James and colleagues demonstrated as a reliable means for performing Rh phenotyping, in addition to some of the other methods mentioned above.21,76 However, inactivation of some antigens during the elution process has the potential to limit the laboratory’s ability to determine an accurate phenotype.76 RBC genotyping can serve as a valuable alternative to overcome the limitations of serologic phenotyping.89,90 In multiply transfused patients, Reid and colleagues demonstrated that, by using polymerase chain reaction analysis of DNA, the predicted genotype was consistent with serologic phenotyping.91 Other investigators demonstrated successful posttransfusion increments with antigenmatched RBC units based on RBC genotyping for patients with AIHA in the absence of serologic confirmation of underlying alloantibodies. Furthermore, these authors commented that “RBC genotyping techniques for patients with AIHA allow selecting antigen-matched units without

Positive DAT with Polyspecific Antiglobulin Reagents

Perform DAT with Anti-IgG and Anti-C3 Specific Reagents IgG (+) C3 (+/-)

IgG (-) C3 (-)

Consider eluate if hematology labs abnormal*

IgG (-) C3 (+)

No further testing.

Fig. 2. Testing algorithm for performing eluates on samples from patients WITHOUT recent pregnancy or transfusion history (algorithm adapted from Johnston and Belota71). Note that eluate should be performed in all cases when the patient has a recent pregnancy or transfusion history, regardless of specific reagent direct antiglobulin test (DAT) results. *Abnormal laboratory values included low hemoglobin, elevated reticulocyte count, or elevated total bilirubin. C3 5 serum complement 3; IgG 5 immunoglobuin G.

laborious, costly, lengthy, and sample-consuming adsorption procedures.”92

RBC selection and transfusion Eleven studies described RBC unit selection and crossmatch (XM) requirements in patients with WAA. The selection of RBC units that are phenotypically or genotypically matched to the patient reportedly has the potential to limit alloimmunization and reduce the number or frequency of complex serologic work-ups for patients who require ongoing transfusion support.92,93 Numerous investigators advocate this approach, although the degree of matching varies from Rh and K phenotype matching (partial match) to an extended match, which can also include the Jka, Jkb, Fya, Fyb, and S/s antigens.93-97 El Kenz and colleagues demonstrated in patients with AIHA (n 5 4) that RBC units matched to the patient’s predictive phenotype (based on genotypic analysis) for C, c, E, e, K, Jka, Jkb, Fya, Fyb, M, N, S, and s provided appropriate posttransfusion RBC recovery without reports of transfusion reactions.92 Shirey and coworkers demonstrated similar findings with the provision of prophylactic, antigen-matched RBCs for C, c, E, e, K, Jka, Jkb, Fya, Fyb, S, and s antigens in a cohort of 20 patients, none of whom experienced a transfusion reaction or developed an additional alloantibody.96 In these studies, adsorption studies were abridged if serologic testing was consistent with previous results and matched RBC units for the patient’s phenotype were available.92,96 Although the studies are small and without control populations, the authors find that patients with WAAs can safely undergo transfusion with prophylactic, antigenmatched RBCs. The use of this transfusion strategy potentially simplifies the pretransfusion work-up and Volume 57, February 2017 TRANSFUSION 467

ZIMAN ET AL.

TABLE 1. The use of the DAT and elution studies in patients with WAAs: Survey results Situation

Response

1. For patients with known WAA, a DAT is performed: Every time an IAT is positive If ordered by the treating physician Every 7 days only if there is a history of transfusion (to detect presence of new alloantibody) Yes, every 7 days even if there is no history of transfusion Every 30 days only if there is a history of transfusion (to detect the presence of new alloantibody) Every 30 days even if there is no history of transfusion At the discretion of treating physician or clinical/laboratory scientists If the autocontrol is positive Yes, but none of the above apply 2. For patients with a positive DAT (IgG) and no history of WAA, an elution is performed: Every time the DAT is positive Only with the initial identification of WAA in IAT Only when there is a change in reactivity strength (1 grade) Never Other* 3. For patients with a positive DAT (IgG) and a previous history of WAA, but no diagnosis of AIHA, an elution is performed: Every time the DAT is positive Only with the initial identification of WAA in IAT Only when there is a change in reactivity strength (1 grade) Never Other* 4. For patients with a positive DAT (IgG) and a clinical diagnosis of AIHA, an elution is performed: Every time the DAT is positive Only with the initial identification of WAA in IAT Only when there is a change in reactivity strength ( 1 grade) Never Other* 5. For patients with a positive DAT (C3 only) and laboratory evidence of hemolysis, an elution is performed: No Yes Other*

N 5 52 67% 13% 6% 0% 2% 2% 2% 4% 10% N552 27% 33% 10% 12% 19% N552 27% 10% 25% 17% 21% N552 33% 29% 10% 6% 23% N 5 52 42% 37% 21%

*Respondents’ comment to an answer of “Other” included physician-ordered DATs, recent history of transfusion and pregnancy, prescribed time interval since previous serologic testing, evidence of hemolysis and/or transfusion, or a combination of these variables. C3 5 serum complement 3; WAA 5 warm-reactive autoantibody; DAT 5 direct antiglobulin test; IAT 5 indirect antiglobulin test; AIHA 5 autoimmune hemolytic anemia.

requirement for adsorption studies. The benefits of this practice must be weighed against the urgent clinical need such that the desire to provide matched RBCs units does not lead to a delay in transfusion.98 In addition, this practice requires a large population of genotyped donors and knowledge of the predicted phenotype of these donor units in the current inventory to meet a transfusion service’s goal of providing antigen-matched RBC units.93 At least one report demonstrated a benefit to modifying XM requirements. Lee and coworkers demonstrated that no additional benefit was achieved by performing a serologic (IgG) XM over an immediate spin (IS) XM to confirm ABO compatibility; these authors propose that, in the AIHA patient population, once alloantibodies are excluded, limiting XM testing to an IS XM can decrease turnaround time and cost without compromising patient safety.99

Survey Individuals representing 54 different institutions in the United States (n 5 34), Canada (n 5 3), Europe (n 5 14), Brazil (n 5 1) New Zealand (n 5 1), and Australia (n 5 1) 468 TRANSFUSION Volume 57, February 2017

completed the survey (response rate 5 33% of institutions queried). Respondents had the following affiliations: BEST Collaborative, 52%; ABC, 31%; CAP Transfusion Medicine Resource Committee, 6%; EBA, 11%. They represented the following types of laboratories: HBTS, 39%; IRL, 44%; and other, 17%. Eighty-six of HBTS that responded to this survey stated that the vast majority of testing is performed in the respondent’s laboratory with very little, if any, sent to an IRL, indicating that this cohort of HBTS respondents have the staff and resources necessary to perform and interpret results of complex testing.

Test methodology (primary screening and detection of underlying alloantibodies) Eight survey questions focused on test methodology and detection of underlying alloantibodies. CAT is used by the largest percentage of respondents (43%) as the primary testing methodology; the remaining respondents use standard tube testing (33%) with albumin (2%) or potentiating media, such as PEG (17%) or LISS (15%), and SPRCA (24%). Most sites report more than one methodology available for testing. Responding laboratories use the following techniques to identify underlying alloantibodies in the presence of

AUTOANTIBODIES AND LABORATORY TESTING

TABLE 2. Survey results for the use of RBC phenotyping/genotyping and crossmatching techniques for patients with WAAs Situation

Response

1. For patients with WAA, when performing RBC phenotyping, it is performed: At initial detection of WAA Only with the initial order for RBC units If alloantibody is detected If the patient has not recently received a transfusion To facilitate the selection of cells for adsorption It is never performed Other* Testing is NOT performed in my laboratory; specimen is sent to a reference laboratory 2. For patients with WAA, when performing phenotyping, the following antigens are routinely included: C, c, E, K, Jka, Jkb (in addition to other antigens) Fya, Fyb (in addition to other antigens) S, s (in addition to other antigens) e (in addition to other antigens) M, N (in addition to other antigens) P1 (in addition to other antigens) Lea, Leb (in addition to other antigens) 3. For patients with WAA, when performing RBC genotyping, it is performed: At initial detection of WAA It is rarely or never performed If the patient has recently received a transfusion (within the past 3 months) When phenotyping not possible/unsuccessful When ordered by the physician (clinician or transfusion medicine physician) If alloantibody is detected or identification of alloantibody is too complex Not until an order for RBC transfusion is received To facilitate the selection of cells for adsorption 4. When a WAA is present, the type of sample used in crossmatch: Neat plasma Adsorbed plasma Electronic/computer crossmatch Titrated plasma 5. The technique used for cross-matching when there is a WAA without underlying alloantibodies: Column agglutination test (CAT) Tube—LISS Tube—saline Electronic/computer crossmatch Tube—PEG Tube—albumin Solid-phase red cell adherence (SPRCA)

N 5 50 66% 12% 6% 4% 4% 2% 2% 4% N 5 44 100% 75% 68% 66% 21% 18% 11% 9% N549 20% 20% 18% 14% 12% 10% 4% 2% N546 44% 37% 16% 2% N548 33% 22% 15% 15% 11% 2% 2%

*Respondents’ comment to an answer of “Other” included the need to supply RBC units that were negative for the apparent specificity of WAA. WAA 5 warm-reactive autoantibody; RBC 5 red blood cell; LISS 5 low ionic strength saline; PEG 5 polyethylene glycol.

WAAs (often more than one technique is used): autoadsorption (75%), alloadsorption (differential adsorption; 42%), saline IgG (21%), LISS (8%), 1:3 dilution technique (2%), and 1:5 dilution technique (2%). When autoadsorptions are performed, the following methods are used to treat the patient’s RBCs: ZZAP treatment (dithiothreitol and enzyme; 43%), enzyme treatment (ficin or papain; 26%), untreated in the presence of PEG (19%), untreated (6%), and other (6%). In patients who have a recent history of transfusion, alloadsorption is used by 76% of laboratories that do not send this test to a reference laboratory (with one respondent commenting that they use PEG-adsorption); other methods mentioned in the “other” category include titration, salineIgG, LISS-IgG, and genotyping. Respondents varied in terms of the time period after transfusion that is sufficient for the

performance of autoadsorption versus alloadsorption: the majority (85%) defined this period as 3 months, and others defined it as 13 weeks (2%), 4 months (7%), and 6 months (5%).

DAT and elution studies Nine survey questions focused on DAT and elution studies. Survey respondents were asked about the frequency with which they performed DATs and elution studies (Table 1). Respondents commented that they performed elution studies on positive DATs if the patient has received a transfusion or has been pregnant since the previous studies, every 30 days regardless of transfusion/pregnancy history, per clinician order, and for a clinical suspicion of immune-mediated hemolysis. Volume 57, February 2017 TRANSFUSION 469

ZIMAN ET AL.

Fig. 3. Survey results of antigen-matching policy for RBC unit selection in the following patients with no underlying alloantibodies (n 5 51 responding laboratories).

Respondents were queried regarding the methods used to create an eluate. The majority of respondents (77%) use the acid/glycine method (e.g., EluKit II [Immucor] and ELUClear [Hemo Bioscience]). Other methods include glycine HCl/EDTA (9%), cold acid (4%), the Lui freeze-thaw technique (2%), digitonin acid (2%), chloroform and trichloroethylene (2%), heat (2%) and DiaCidel (2%; Bio-Rad). The primary methods used to detect antibodies in eluates are CAT (43%) and traditional tube with PEG (27%), saline (22%), and LISS (6%); one institution uses tube-saline for strongly reactive DATs versus tube-PEG for weakly reactive DATs. We queried respondents regarding clinician use of DAT results. Of the respondents who were knowledgeable about clinician practice (n 5 25), 64% stated that the clinicians at their institution use the DAT to monitor hemolysis in patients with a WAA during an acute episode of hemolysis (defined as new-onset hemolysis with evidence of pallor and anemia; jaundice with increased indirect bilirubin concentration; increased serum lactate dehydrogenase concentration; reduced/absent serum haptoglobin; increased reticulocyte percentage or 470 TRANSFUSION Volume 57, February 2017

absolute reticulocyte number; the presence of circulating spherocytes on a peripheral blood smear). Furthermore, 39% stated that the DAT is used by clinicians to monitor disease activity during and between episodes of clinical hemolysis and that the frequency of DAT orders varies by clinician.

Phenotyping/genotyping Seven questions focused on phenotyping and genotyping of patient RBCs with the presence of WAAs. Phenotyping and genotyping methods are summarized in Table 2. With regard to genotyping, 20% of respondents never/rarely perform genotyping, and 12% will perform it if ordered by the clinician. The remaining 68% of respondents perform genotyping based on the following indications: initial detection of WAA regardless of request for RBC units (i.e., initial presentation; 20%), with the initial RBC order (4%), with a recent history of transfusion (within the past 3 months; 18%), when phenotyping is not successful/possible (14%), if the detection of an alloantibody or the ability to identify underlying antibodies becomes too complex (10%), or to facilitate the selection of

AUTOANTIBODIES AND LABORATORY TESTING

TABLE 3. Synthesis of literature review and survey: testing and blood selection practices in patients with WAAs

adsorption cells (2%). Policies regarding genotyping date to 2003, and with the majority have been adopted since 2008.

RBC selection and transfusion There were seven survey questions on RBC selection. Patient samples and the techniques used for

crossmatching vary significantly (Table 2). Seventy-five percent of respondents (n 5 51) (78% HBTS, 63% IRL, 100% of “other” institution types) provide phenotypematched or genotype-matched RBC units when a WAA is present and no underlying alloantibodies are detected (Fig. 3); policies vary based on the presence of a diagnosis of AIHA, urgency of transfusion, ability to rule out Volume 57, February 2017 TRANSFUSION 471

ZIMAN ET AL.

TABLE 3. Continued

underlying alloantibodies, results of XM testing, availability of antigen-negative units, and, for some IRLs, according to hospital customer request. The absence or presence of underlying alloantibodies did not impact the overall rate of provision of matched RBC units. Some institutional policies for antigen-matched units have 472 TRANSFUSION Volume 57, February 2017

been in place since the 1970s. Sixty percent of respondents said they provide antigen-negative units when the patient has a WAA that demonstrates specificity, although many factors are taken into consideration, including the clinical severity of hemolysis and the rarity of antigennegative units.

AUTOANTIBODIES AND LABORATORY TESTING

90 80

Number of Publicaons

70 60 50 40 30 20 10 0

Decade Publications Cited

Publications Excluded Following Full Review

Fig. 4. Publications on WAAs (immunohematology testing and RBC selection) are identified for review by decade (n 5 382).

DISCUSSION WAAs can be problematic in the pretransfusion setting given the presence of underlying alloantibodies, the concern for RBC compatibility between patient and donor RBC unit, and the potential for HTRs. Current evidence-based protocols for testing and RBC selection do not exist, and there is a lack of comparative studies with clinical endpoints that address immunohematology laboratory practices in the presence of a WAA. This survey provides an expanded description of current laboratory testing practices for patients with WAAs (with or without AIHA) and shows significant practice variation. These findings, combined with the literature review, which provides a historical viewpoint, highlight the variable practices and methodologies used in laboratories over time. Until robust, clinically oriented research becomes available, immunohematology laboratories may benefit from having access to a pragmatic collation of the literature review and survey results (Table 3). This summary must be considered with respect to the laboratory resources available and the patients’ clinical status regarding the urgency of transfusion. The literature review identified numerous studies focused on immunohematology test methodology. In contrast, only a few studies involved patient phenotyping and/or genotyping and the extent of matching between the RBC unit and the patient (recipient). Interestingly, in our survey cohort, 75% of laboratories provide varying degrees of phenotype/genotype-matched RBC units and have been doing so for decades. This may be due to the preponderance of moderate to high complexity

laboratories that responded to the survey. Although the practice of providing phenotype/genotype-matched RBC units should theoretically provide an increased level of safety by decreasing future alloimmunization and potential for HTRs, as well as decreasing the need for complex serologic work-ups with adsorptions, there is little evidence in the literature to support this practice. There were several challenges with the literature review. Some pertinent references may have been missed, and references after January 2014 are not included. Overall, the available body of literature is decades old; most was published before the 21st century; and, although some methodologies are still common, others are no longer used (Fig. 4). A significant number of publications focused on novel applications that are not readily adaptable to an HBTS or were not within the clinical context of patient care pressures. Many studies were based on small laboratory studies of randomly collected samples or a small patient population without a control population; therefore, no statistical conclusions can be drawn. Furthermore, if the study did not include key details, then it was not possible to include it and compare its methods with those in other published accounts. As of January 2014, there is a paucity of published literature focusing on current topics of interest in transfusion medicine in relation to the WAA patient population: RBC genotyping, genetic variation in blood group antigens, and the practice of personalized medicine (i.e., RBC matching for transfusion). The survey has several limitations. Chief among these is that the number of survey respondents was limited, Volume 57, February 2017 TRANSFUSION 473

ZIMAN ET AL.

only representing the practices at 54 laboratories (predominantly HBTS in academic medical centers and IRLs). As such, the observed testing and RBC selection practices reflect the resources and methodologies of these participating laboratories, which are primarily moderatecomplexity to high-complexity centers and thus may not be reflective of laboratories in other settings. The survey was not designed to assess differences in practice between HBTS and IRLs or to identify differences in practice based on hospital size, the complexity of available testing, or geographic location. The survey also did not specifically assess for practice differences based on the testing platform used (automated, manual gel or solid phase). Future studies are needed to evaluate and compare the effectiveness of different testing algorithms and transfusion strategies, particularly in this era of RBC genotyping, the discovery of genetic variation in blood group antigens, and the practice of personalized medicine (i.e., RBC matching). It would also be advantageous to focus on the feasibility and cost effectiveness of routine phenotyping/genotyping in conjunction with the selection of matched RBC units for transfusion to determine whether this strategy impacts alloimmunization and testing algorithms. In addition, scientific investigations to identify which WAAs will become clinically significant (i.e., will lead to a diagnosis of AIHA), versus those that will not, could guide practice in testing and RBC-selection strategies.

7. Blackall DP. Warm-reactive autoantibodies in pediatric patients: clinical and serologic correlations. J Pediatr Hematol Oncol 2007;29:792-6. 8. Petz LD. A physician’s guide to transfusion in autoimmune haemolytic anaemia. Br J Haematol 2004;124:712-6. 9. Petz LD. Red cell transfusion problems in immunohematologic disease. Annu Rev Med 1982;33:355-61. 10. Mack P, Freedman J. Autoimmune hemolytic anemia: a history. Transfus Med Rev 2000;14:223-33. 11. Reardon JE, Marques MB. Laboratory evaluation and transfusion support of patients with autoimmune hemolytic anemia. Am J Clin Pathol 2006;125 Suppl:S71-7. 12. Bell CA, Zwicker H, Sacks HJ. Autoimmune hemolytic anemia: routine serologic evaluation in a general hospital population. Am J Clin Pathol 1973;60:903-11. 13. Gologan R, Berceanu S. Serologic diagnosis in the differentiation of autoimmune hemolytic anemias. Med Intern 1979; 17:191-6. 14. Sokol RJ, Hewitt S, Stamps BK. Autoimmune haemolysis: an 18-year study of 865 cases referred to a regional transfusion centre. Br Med J (Clin Res Ed) 1981;282:2023-7.  ska B, Lawkowicz W, Go  rska B, et al. Autoimmune hae15. Zupan molytic anaemia in children. Br J Haematol 1976;34:511-20. 16. Issitt PD, Pavone BG, Goldfinger D, et al. Anti-Wrb, and other autoantibodies responsible for positive direct antiglobulin tests in 150 individuals. Br J Haematol 1976;34:5-18. 17. Laine ML, Beattie KM. Frequency of alloantibodies accompanying autoantibodies. Transfusion 1985;25:545-6. 18. Engelfriet CP, Reesink HW, Garratty G, et al. The detection of alloantibodies against red cells in patients with warm-type

ACKNOWLEDGMENT We thank Rikke Ogawa for her assistance with the literature search.

autoimmune haemolytic anaemia. Vox Sang 2000;78:200-7. 19. Morel PABM, Frank BA. A simple method for the detection of alloantibody in the presence of warm autoantibody. Transfusion 1978;18:388. 20. Wallhermfechtel MA, Pohl BA, Chaplin H. Alloimmunization in patients with warm autoantibodies. A retrospective study

CONFLICT OF INTEREST

employing three donor alloabsorptions to aid in antibody detection. Transfusion 1984;24:482-5.

The authors have disclosed no conflicts of interest.

REFERENCES 1. Petz L, Garratty G. Immune hemolytic anemias. Philadelphia (PA): Churchill Livingstone; 2004. 2. Sokol RJ, Booker DJ, Stamps R. The pathology of autoimmune haemolytic anaemia. J Clin Pathol 1992;45: 1047-52. 3. Dacie JV, Worlledge SM. Auto-immune hemolytic anemias. Prog Hematol 1969;6:82-120.

21. James P, Rowe GP, Tozzo GG. Elucidation of alloantibodies in autoimmune haemolytic anaemia. Vox Sang 1988;54:167-71. € fer H, Mueller-Eckhardt C. Red blood cell 22. Salama A, Bergho transfusion in warm-type autoimmune haemolytic anaemia. Lancet 1992;340:1515-7. 23. Issitt PD, Combs MR, Bumgarner DJ. Studies of antibodies in the sera of patients who have made red cell autoantibodies. Transfusion 1996;36:481. 24. So CC, Wong KF, Yu PH, et al. Alloimmunization in Chinese with warm autoimmune haemolytic anaemia—incidence and characteristics. Transfus Med 2000;10:141-3.

4. Packman CH. Hemolytic anemia due to warm autoantibod-

25. Maley M, Bruce DG, Babb RG, et al. The incidence of red cell

ies. Blood Rev 2008;22:17-31. 5. Pirofsky B. Autoimmunization and the autoimmune hemo-

alloantibodies underlying panreactive warm autoantibodies.

lytic anemias. Baltimore (MD): Williams & Wilkins; 1969. 6. Bottiger LE, Westerhold B. Acquired haemolytic anaemia. Acta Med Scand 1973;193:223-6. 474 TRANSFUSION Volume 57, February 2017

Immunohematology 2005;21:122-5. €hne A, et al. Coexistence of autoanti26. Ahrens N, Pruss A, Ka bodies and alloantibodies to red blood cells due to blood transfusion. Transfusion 2007;47:813-6.

AUTOANTIBODIES AND LABORATORY TESTING

27. Laine EP, Leger RM, Arndt PA, et al. In vitro studies of the impact of transfusion on the detection of alloantibodies after autoadsorption. Transfusion 2000;40:1384-7. 28. Haspl ZH, Tomicıc M, Grgicevic´ D. Clinically significant red cell alloantibodies in patients with warm autoimmune hemolytic anemia. Acta Med Croatica 2001;55:149-52. 29. Wheeler CA, Calhoun L, Blackall DP. Warm reactive autoantibodies: clinical and serologic correlations. Am J Clin Pathol 2004;122:680-5. 30. Cheng CK, Wong ML, Lee AW. PEG adsorption of autoantibodies and detection of alloantibodies in warm autoimmune hemolytic anemia. Transfusion 2001;41:13-7. 31. Chiaroni J, Touinssi M, Mazet M, et al. Adsorption of

46. Kaplan HS, Garratty G. The predictive value of the direct antiglobulin test. Diagn Med 1985;8:29-33. 47. Zantek ND, Koepsell SA, Tharp DR, et al. The direct antiglobulin test: a critical step in the evaluation of hemolysis. Am J Hematol 2012;87:707-9. 48. Worlledge SM. The interpretation of a positive direct antiglobulin test. Br J Haematol 1978;39:157-62. 49. Finck RH, Davis RJ, Teng S, et al. Performance of an automated solid-phase red cell adherence system compared with that of a manual gel microcolumn assay for the identification of antibodies eluted from red blood cells. Immunohematology 2011;27:1-5. 50. Dittmar K, Procter JL, Cipolone K, et al. Comparison of

autoantibodies in the presence of LISS to detect alloanti-

DATs using traditional tube agglutination to gel column

bodies underlying warm autoantibodies. Transfusion

and affinity column procedures. Transfusion 2001;41:

2003;43:651. 32. Barron C. Allogeneic red blood cell adsorption for removal of warm autoantibody. Immunohematology 2014;30:153-5. 33. Pham BN, Roussel M, Peyrard T, et al. Anti-D investigations in individuals expressing weak D Type 1 or weak D Type 2: allo- or autoantibodies? Transfusion 2011;51:2679-85. 34. Leger RM, Garratty G. Evaluation of methods for detecting alloantibodies underlying warm autoantibodies. Transfusion 1999;39:11-6. 35. Cid J, Ortın X, Pinacho A, et al. Use of polyethylene glycol for performing autologous adsorptions. Transfusion 2005;45:694-7. 36. Das SS, Chaudhary R. Utility of adsorption techniques in

1258-62. 51. Fabijanska-Mitek J, Namirska-Krzton H, Seyfried H. The value of gel test and ELAT in autoimmune haemolytic anaemia. Clin Lab Haematol 1995;17:311-6. 52. Dutta P, Chatterjee T, Tyagi S, et al. Gel card in the diagnosis of autoimmune haemolytic anemia. Indian J Pathol Microbiol 2005;48:322-4. 53. Nathalang O, Chuansumrit A, Prayoonwiwat W, et al. Comparison between the conventional tube technique and the gel technique in direct antiglobulin tests. Vox Sang 1997;72: 169-71. 54. Asmussen C, Gutensohn K, Wittkopf D, et al. Rapid detection

serological evaluation of warm autoimmune haemolytic

of IgG subclasses on DAT positive RBC membranes by flow

anaemia. Blood Transfus 2009;7:300-4.

cytometry (FC). Beitr Infusionsther Transfusionsmed 1996;

37. Barron CL, Brown M. The use of polyethylene glycol (PEG) to enhance the adsorption of autoantibodies. Immunohematology 1997;13:119-22. 38. Judd WJ, Dake L. PEG adsorption of autoantibodies causes loss of concomitant alloantibody. Immunohematology 2001; 17:82-5. 39. Leger RM, Ciesielski D, Garratty G. Effect of storage on antibody reactivity after adsorption in the presence of polyethylene glycol. Transfusion 1999;39:1272-3.

33:35-9. 55. Bellamy JD, Booker DJ, James NT, et al. Measurement of red blood cell-bound C3b and C3d using an enzymelinked direct antiglobulin test. Immunohematology 1997; 13:123-31. 56. Bodensteiner D, Brown P, Skikne B, et al. The enzyme-linked immunosorbent assay: accurate detection of red blood cell antibodies in autoimmune hemolytic anemia. Am J Clin Pathol 1983;79:182-5.

40. Øyen R, Angeles ML. A simple screening method to evaluate

 ska-Mitek J, Pogłod R, Adamowicz-Salach A, et al. 57. Fabijan

the presence of alloantibodies with concomitant warm auto-

Quantitation of red cell-bound IgG by an enzyme-linked

antibodies. Immunohematology 1995;11:85-7.

antiglobulin test in the patients with warm-type autoimmune

41. Jang MJ, Cho D, Park KU, et al. Autoantibodies with mimicking specificity detected by the dilution technique in

haemolytic anaemia. Clin Lab Haematol 2006;28:241-4. 58. Wang Z, Shi J, Zhou Y, et al. Detection of red blood cell-bound

patients with warm autoantibodies. Ann Lab Med 2013;33:

immunoglobulin G by flow cytometry and its application in

343-8.

the diagnosis of autoimmune hemolytic anemia. Int J Hematol

42. Garratty G. The significance of IgG on the red cell surface. Transfus Med Rev 1987;1:47-57. 43. Domen RE. Warm red blood cell autoantibodies and the direct antiglobulin test revisited. Am J Clin Pathol 2004;122:673-4. 44. Bencomo AA, Diaz M, Alfonso Y, et al. Quantitation of red

2001;73:188-93. 59. Bordin JO, Souza-Pinto JC, Kerbauy J. Measurement of red blood cell antibodies in autoimmune hemolytic anemia. Braz J Med Biol Res 1991;24:895-9. 60. Namirska-Krzton H, Fabijanska-Mitek J, Seyfried H.

cell-bound IgG, IgA, and IgM in patients with autoimmune

Enzyme-linked antiglobulin test for the evaluation of the

hemolytic anemia and blood donors by enzyme-linked

amount of IgG autoantibodies on red blood cells. Clin Lab

immunosorbent assay. Immunohematology 2003;19:47-53. 45. Petz LD. Autoimmune hemolytic anemia. Hum Pathol 1983; 14:251-5.

Haematol 1995;17:221-4. 61. Garratty G. Flow cytometry: its applications to immunohaematology. Baillieres Clin Haematol 1990;3:267-87. Volume 57, February 2017 TRANSFUSION 475

ZIMAN ET AL.

62. Garratty G, Nance SJ. Correlation between in vivo hemolysis and the amount of red cell-bound IgG measured by flow cytometry. Transfusion 1990;30:617-21. 63. Hirano A, Yamada H, Kato K. Quantitation of red-cell-bound IgG in normal and pathologic states by an enzyme immunoassay (EIA) technique. Nagoya J Med Sci 1985;47:5-16. 64. Lin JS, Hao TC, Lyou JY, et al. Clinical application of a flow

80. Rekvig OP, Hannestad K. Acid elution of blood group antibodies from intact erythrocyte. Vox Sang 1977;33: 280-5. 81. Greco VA, Byrne KM, Procter JL, et al. Detection of antibodies in acid eluates with the gel microcolumn assay. Transfusion 2002;42:698-701. 82. Eyster ME, Jenkins DE. Erythrocyte coating substances in

cytometric direct antiglobulin test. Transfusion 2009;49:

patients with positive direct antiglobulin reactions. Correla-

1335-46.

tion of gamma-G globulin and complement coating with

65. Lynen R, Krone O, Legler TJ, et al. A newly developed gel centrifugation test for quantification of RBC-bound IgG antibodies and their subclasses IgG1 and IgG3: comparison with flow cytometry. Transfusion 2002;42:612-8. 66. Das SS, Chaudhary RK. Application of gel technology in the serologic characterization of autoantibody in DATpositive autoimmune diseases. Immunohematology 2007; 23:59-62. 67. Dubarry M, Charron C, Habibi B, et al. Quantitation of immunoglobulin classes and subclasses of autoantibodies bound to red cells in patients with and without hemolysis. Transfusion 1993;33:466-71.  ska-Mitek J, Lopien  ska H, Zupanska B. Gel test appli68. Fabijan cation for IgG subclass detection in auto-immune haemolytic anaemia. Vox Sang 1997;72:233-7. 69. Li Z, Shao Z, Xu Y, et al. Subclasses of warm autoantibody IgG in patients with autoimmune hemolytic anemia and their clinical implications. Chin Med J 1999;112:805-8. 70. Lai M, Leone G, Landolfi R. Autoimmune hemolytic anemia with gel-based immunohematology tests. Am J Clin Pathol 2013;139:457-63. 71. Johnston MF, Belota MK. Determination of need for elution studies for positive direct antiglobulin tests in pretransfusion testing. Am J Clin Pathol 1988;90:58-62. 72. Nathalang O, Sriwanitchrak P, Tubrod J, et al. Antibody elutions in Thai patients with a positive direct antiglobulin test. Blood Transfus 2011;9:306-10. 73. Dumaswala UJ, Sukati H, Greenwalt TJ. The protein composition of red cell eluates. Transfusion 1995;35:33-6. 74. Burin des Roziers N, Squalli S. Removing IgG antibodies from intact red cells: comparison of acid and EDTA, heat, and chloroquine elution methods. Transfusion 1997;37: 497-501. 75. Beaumont AE, Stamps R, Booker DJ, et al. An improved method for removal of red cell-bound immunoglobulin using chloroquine solution. Immunohematology 1994;10:22-4. 76. Mantel W, Holtz G. Characterisation of autoantibodies to erythrocytes in autoimmune haemolytic anaemia by chloroquine. Vox Sang 1976;30:453-63. 77. Branch DR, Petz LD. A new reagent ZZAP having multiple applications in immunohematology. Am J Clin Pathol 1982;78:161-7. 78. Byrne P. Use of a modified acid/EDTA elution technique. Immunohematology 1991;7:46-7. 79. Katharia R, Chaudhary RK. Removal of antibodies from red

underlying diseases, overt hemolysis and response to therapy. Am J Med 1969;46:360-71. 83. Chaplin H. Red cell-bound immunoglobulin as a predictor of severity of hemolysis in patients with autoimmune hemolytic anemia. Transfusion 1990;30:576-8. 84. Gehrs BC, Friedberg RC. Autoimmune hemolytic anemia. Am J Hematol 2002;69:258-71. 85. Goldberg LS, Fudenberg HH. Warm antibody hemolytic anemia; prolonged remission despite persistent positive Coombs test. Vox Sang 1968;15:443-5. 86. Svensson AM, Bushor S, Fung MK. Case report: exacerbation of hemolytic anemia requiring multiple incompatible RBC transfusions. Immunohematology 2004;20:177-83. 87. Liumbruno GM, Tognaccini A, Bonini R, et al. The role of the direct antiglobulin test in pre-transfusion investigations and the approach to selecting blood for transfusion in autoimmune haemolytic anaemia: results of a regional survey. Blood Transfus 2008;6:156-62. 88. Griffin GD, Lippert LE, Dow NS, et al. A flow cytometric method for phenotyping recipient red cells following transfusion. Transfusion 1994;34:233-7. 89. Rios M, Hue-Roye K, Storry JR, et al. Cell typing the sensitized transfusion-dependent patient. Ann Clin Lab Sci 2000; 30:379-86. 90. Lomas-Francis C, DePalma H. DNA-based assays for patient testing: their application, interpretation, and correlation of results. Immunohematology 2008;24:180-90. 91. Reid ME, Rios M, Powell VI, et al. DNA from blood samples can be used to genotype patients who have recently received a transfusion. Transfusion 2000;40:48-53. 92. El Kenz H, Efira A, Le PQ, et al. Transfusion support of autoimmune hemolytic anemia: how could the blood group genotyping help? Transl Res 2014;163:36-42. 93. Westhoff CM. The potential of blood group genotyping for transfusion medicine practice. Immunohematology 2008;24: 190-5. 94. Sokol RJ, Hewitt S, Booker DJ, et al. Patients with red cell autoantibodies: selection of blood for transfusion. Clin Lab Haematol 1988;10:257-64. 95. Cerdas-Quesada C. Specificity of 136 patients’ antibodies to human red blood cells in Dr. Max Peralta J Hospital Blood Bank 2004-February 2009. Transfus Apher Sci 2010;42:105-8. 96. Shirey RS, Boyd JS, Parwani AV, et al. Prophylactic antigen-matched donor blood for patients with warm

cells: Comparison of three elution methods. Asian J Transfus

autoantibodies: an algorithm for transfusion management.

Sci 2013;7:29-32.

Transfusion 2002;42:1435-41.

476 TRANSFUSION Volume 57, February 2017

AUTOANTIBODIES AND LABORATORY TESTING

97. Lau FY, Wong R, Chan NP, et al. Provision of phenotypematched blood units: no need for pre-transfusion antibody

indirect antiglobulin test crossmatch? Transfusion 2007;47: 1290-5.

screening. Haematologica 2001;86:742-8. 98. Ness PM. How do I encourage clinicians to transfuse mismatched blood to patients with autoimmune hemolytic anemia in urgent situations? Transfusion 2006;46: 99.

1859-62. Lee E, Redman M, Burgess G, et al. Do patients with autoantibodies or clinically insignificant alloantibodies require an

SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher’s website: Appendix S1. Literature search strategies Appendix S2. Survey questions

Volume 57, February 2017 TRANSFUSION 477