Conundrum: A negative crossmatch in a patient with high level donor specific antibody

• Published in: Human immunology 2015-10, Vol.76, p.57-57
• Main Authors: Rewinski, Michael J, Alberghini, Tod, Bow, Laurine, Devivo, Joseph, Doyle, Pamela, Michalski, Patricia, Ortiz-Rivera, Julio
• Format: Article
• Language: English
• Subjects: Allergy and Immunology
• ISSN: 0198-8859; 1879-1166
• DOI: 10.1016/j.humimm.2015.07.082

With the advent of sensitive luminex single antigen bead (SAB) antibody testing, virtual crossmatches have become reliable ways to predict and eliminate incompatible donor-recipient pairs. In our laboratory, luminex SAB testing (One Lambda) has repeatedly shown that patient serum containing antibody levels with normalized median fluorescence intensity (MFI) values greater than 5000 will yield a positive flow crossmatch almost 95% of the time. We present a patient (EP) with high level Donor Specific Antibody (DSA) to DQ7 who had a negative crossmatch with her DQ7 positive donor. EP and her living donor (AH) did not match for any HLA antigens. Most of the crossmatching in our lab is performed prior to HLA typing, so the initial negative crossmatch between this pair was not noteworthy. However, following low resolution SSP typing by RT-PCR (Linkage Biosciences), donor AH typed phenotypically for DQ7. Since EP was known to have DQ7 antibodies by both luminex ID and SAB (MFI = 8241) assays, it was predicted that this crossmatch would be positive. The flow crossmatch between EP and AH, however, was B and T cell negative, with a B cell delta channel shift (DCS) of only 13. Higher resolution typing by SSP and SSO (One Lambda) was then performed to confirm and elucidate the class II, specifically DQB, typing. Interestingly, AH was shown to be DQB1∗03:03, DQB1∗03:19 which is phenotypically DQ9, DQ7 or DQ9, – (presumably no serological equivalent or weakly expressed antigen). The SAB assay has five DQ7 beads, all of which contain antigens isolated from DQB1∗03:01 cells, but with different DQA associations. With this information, it was not surprising that this pair’s crossmatch was negative. Additional surrogate crossmatches were then performed with EP sera against known DQB1∗03:01 and DQB1∗03:19 typed cells. Two cells expressing DQB1∗03:01 had positive B cell crossmatches with DCS of 106 and 120, while two cells which typed as DQB1∗03:19 had negative B cell crossmatches (DCS = 44 and 9). This particular donor-recipient pair emphasizes that virtual crossmatches, while valuable, cannot be considered accurate for every donor and recipient. Further studies need to be performed to determine the clinical significance of a negative crossmatch in the presence of DSA to closely related, at the allelic level, HLA antigens.