HLA-selected platelets for platelet refractory patients with HLA antibodies: a single-center experience.

Abstract

Platelet refractoriness due to HLA immunization represents a problem in transfusion management of thrombocytopenic hematology patients. Refractory patients can be managed by HLA-selected platelet transfusions, but the optimal matching strategy is debated and how the degree of HLA mismatch influences transfusion outcome is poorly studied. We studied 32 hematology patients who received 142 matched platelet units between 2007 and 2016. Four matching strategies were compared: 1) genomic HLA typing at the two digit level, performed using polymerase chain reaction-sequence-specific oligonucleotide probing; 2) serologic "eplet score" calculated using HLAMatchmaker; 3) cross-matching using lymphocyte cytotoxicity; and 4) matching based on donor-specific antibody (DSA) specificity, determined using Luminex. A 1-hour corrected count increment (CCI) of more than 7.5 × 109 /L was considered a successful response. Selection of platelets with either a complete HLA match or an acceptable HLA mismatch based on genomic typing and DSA information, each predicted 86% successful transfusion responses. For HLA-mismatched transfusions, the eplet score correlated with CCI and the fraction of successful transfusions, but less well compared to DSA matching. Cytotoxic crossmatching was least predictive. For transfusions across one to four DSAs, the antibody reaction strength correlated with the 1-hour CCI, but many transfusions were successful despite the presence of DSA. A complete HLA-A and -B match or an acceptable mismatch based on DSA should guide identification of donors. Still, transfusions across DSAs are often successful, emphasizing that the presence of DSA is necessary but not sufficient for platelet clearance.