I M M U N O H E M A T O L O G Y Do patients with autoantibodies or clinically insignificant alloantibodies require an indirect antiglobulin test crossmatch? Edmond Lee, Martin Redman, Gordon Burgess, and Nay Win BACKGROUND: Compatibility testing is the standard protocol that identifies suitable blood for patients requir- ing transfusion. If the antibody screen is negative or no clinically significant antibodies are detected, BCSH guidelines and AABB standards allow an immediate- spin crossmatch (IS XM) or even electronic issue. The testing requirement is less clear where autoantibodies or non���clinically significant alloantibodies compromise the indirect antiglobulin test crossmatch (IAT XM). Per- forming an IAT XM will give a mismatched result anyway, delays the supply of blood to the patient, and provides no additional benefit or safety. STUDY DESIGN AND METHODS: From January 2002 to April 2006, the provision of blood for autoimmune hemolytic anemia (AIHA) patients with autoantibodies and no alloantibodies as well as patients with alloanti- bodies that exhibited a ���high-titer, low-avidity��� (HTLA) mode of reactivity was reviewed. RESULTS: A total of 222 AIHA patients (428 samples) with autoantibodies had 1585 units of red cells supplied after IAT XM 1308 (82.5%) were mismatched. In 50 patients (80 samples) with HTLA-like antibodies, 286 units of 328 (87.2%) were mismatched by IAT XM. CONCLUSION: No adverse reactions were reported for the study groups where ���suitable��� blood was provided after a serologically mismatched IAT XM. No additional benefit for these patients can be claimed by performing an IAT XM over an IS XM, as a check of ABO match. The IAT XM is both costly and time-consuming. It is proposed that for these study group patients, a reduc- tion to an IS XM can be applied and can be beneficial. F ree autoantibodies are detected in the majority of patients with autoimmune hemolytic anemia (AIHA) and other conditions associated with a positive direct antiglobulin test (DAT). These autoantibodies usually react with all red cells (RBC) tested of common phenotypes. The overall incidence of RBC alloimmunization with clinically significant alloantibod- ies in patients with warm autoimmune hemolytic anemia (WAIHA) has been reported as approximately 32 per- cent.1,2 Exclusion and identification of underlying alloan- tibodies are important so that antigen-negative blood can be selected. Therefore, in most cases of a WAIHA patient or a positive DAT with free autoantibodies, samples are referred to a reference laboratory for further investiga- tions. Either auto- or alloadsorption studies may be required to identify the presence of alloantibodies. Similarly, patients with ���high-titer, low avidity (HTLA)-like��� antibodies cause difficulty in serologic testing because of the weak reactions they produce in the IAT.3 The term HTLA has been used to categorize the fol- lowing immunoglobulin G (IgG) antibodies: anti-Csa, -Yka, -Kna, -Knb, -McCa to -McCf, -Sla, -Slb, -JMH, -Ch, and -Rg. These antibodies usually exhibit high titers, with low avidity however, not all examples are found with high titers. They do react with nearly all random RBCs tested, although some show racial variation. Samples from this group of patients are also referred to a RBC reference ABBREVIATIONS: AIHA = autoimmune hemolytic anemia EXM = electronic crossmatch HTLA = high-titer, low avidity IS XM = immediate-spin crossmatch NBS = National Blood Service WAIHA = warm autoimmune hemolytic anemia XM(s) = crossmatch(-es). From Red Cell Immunohematology, National Blood Service, London, UK. Address reprint requests to: Edmond Lee, Red Cell Immuno- hematology, National Blood Service, London, UK e-mail: edmond.lee@nbs.nhs.uk. Received for publication October 5, 2006 revision received January 11, 2007, and accepted January 16, 2007. doi: 10.1111/j.1537-2995.2007.01272.x TRANSFUSION 2007 47:1290-1295. 1290 TRANSFUSION Volume 47, July 2007

laboratory to determine the specificity and clinical signifi- cance of the particular antibody and to exclude any other underlying clinically significant antibodies. In England, all blood donations are tested for ABO and D, Rh(C, E, c, and e), and K1. Therefore, for patients with autoantibodies where no alloantibodies are identi- fied, it is relatively easy to select Rh(D, C, E, c, and e)-matched, K:-1 RBCs units for compatibility testing. Alloantibodies with Rh and K1 specificities are most com- monly found in such patients.4 Additional antigen- negative RBC units will be selected if clinically significant alloantibodies are found in the serum samples of these patients. The main challenge to a reference laboratory is to determine whether underlying RBC alloantibodies are present and to provide suitable blood in a timely manner. A total of 704 samples with autoantibodies (n = 325 patients) and 107 samples with HTLA-like antibodies (n = 68 patients) were referred to the Red Cell Immunohe- matology Laboratory, National Blood Service (NBS), Colindale Center, UK, over a 52-month period (January 2002 to April 2006). All of the RBC units crossmatched were found to be mismatched by 37��C IAT technique, but because no underlying alloantibodies were found after extensive investigation, the blood was issued as being ���suitable��� for transfusion. We evaluated the safety of immediate-spin cross- match (IS XM) as an alternative to indirect antiglobulin test crossmatch (IAT XM) in the following two groups of patients where no clinically significant alloantibodies had been found after extensive antibody testing: patients with autoantibodies and patients with HTLA-like antibodies. MATERIALS AND METHODS Materials Data were extracted from compatibility testing records performed by our reference laboratory between January 2002 and April 2006. A summary for crossmatching is shown in Table 1. A total of 1078 XMs were performed for 557 patients, with 4202 units crossmatched and 3801 issued. From the above figures we reviewed data for our two main study groups: AIHA patients with autoantibodies and patients with HTLA-like antibodies, for example, anti- Kna, -Yka, -Chido, and Rogers. Group 1 consisted of 325 patients, with 2645 units crossmatched and 2448 issued, and Group 2 consisted of 68 patients with 439 units cross- matched and 283 issued. Both groups were subdivided into no clinically significant alloantibodies present and significant alloantibodies present, as seen in Table 2. Methods From January 2002 to December 2003, patient samples were grouped for ABO, Rh(D, C, E, c, e), and K1 by micro- plate technique. The tests were incubated at room tem- perature (18-25��C) for 30 minutes and centrifuged at 130 �� g for 2 minutes, and the RBCs were resuspended to allow reading. Anti-A and anti-B reagents were from Sero- logicals Ltd (Livingston, Scotland) anti-D reagent RUM1 was from NBS Reagents (Cambridge, UK) and RS1126 was from Lorne Laboratories (Reading, UK) monoclonal anti-Rh (-C, -E, -c, and -e) were from Biotest (Biotest AG, Dreieich, Germany) and anti-K was from Lorne Laborato- ries. After December 2003, ABO, Rh and K phenotyping was performed by gel column agglutination method involving ABO/D+ reverse grouping (DiaClon, DiaMed AG, Cressier, Switzerland), ABD confirmation, and Rh subgroups + K cards (DiaMed AG), using a 5 percent cell suspension in DiaMed diluent-2 and a 10-minute centrifugation before reading. Antibody identification was performed by standard- tube low-ionic-strength saline (LISS) IAT (LISS from Inverclyde, Scotland). Each patient���s serum or plasma sample was tested against a RBC panel produced by NBS Reagents, to determine antibody specificities. The manual polybrene technique was also employed to determine additional antibody specificities. The 0.05 percent work- ing polybrene, resuspending solution, and low-ionic medium were prepared in house according to Judd.5 Three testing protocols were selectively utilized for the detection of alloantibodies underlying autoantibodies: 1. Dilution technique: patient���s serum or plasma sample diluted in phosphate-buffered saline (PBS) 1 in 3 (one part of serum or plasma and two parts of PBS). Native and diluted serum or plasma was tested by LISS IAT against an antibody identification panel and donor RBCs in the IAT XM. 2. ZZAP6 autoadsorption performed by treating 1 vol of washed and packed patient���s RBCs with 4 vol of ZZAP provided by NBS Reagents (5 mL 0.2 mol/L dithiothreitol + 1 mL of 1% papain + 4 mL PBS) and prepared freshly before use. The patient���s treated cells TABLE 1. Summary of total number of XMs performed: January 2002 to April 2006 XMs Number of XMs Number of patients RBCs units crossmatched Urgency (number of XMs) Compatibility results (number of units) Total Issued Routine Urgent Matched Suitable Mismatched Number 1078 557 4202 3801 673 405 1525 2636 41 Percent 100 90.5 62.4 37.6 36.3 62.7 1 AUTOANTIBODIES OR ALLOANTIBODIES AND IAT CROSSMATCH Volume 47, July 2007 TRANSFUSION 1291