Eplet Matching in Pediatric Heart Transplantation

Abstract

<h3>Purpose</h3> To compare antigen to molecular-level HLA matching in regards to post-heart transplant outcomes in a pediatric population. <h3>Methods</h3> Retrospective single centre cohort study including all heart transplant patients aged < 18 years between 2013-2020. HLA typing, antigen mismatch, and eplet mismatch analysis were performed on all donor - recipient pairs. The primary endpoint was graft loss (death/ re-transplantation) and the secondary enpoint was the development of allograft rejection. Multivariable Cox regression analysis was used to examine associations between antigen or eplet mismatching and adverse outcomes. <h3>Results</h3> 77 patients were included, 31 (40%) males, median age at transplant 4.3 years [range 0.05 - 18]. Median number of HLA class I and II donor-recipient eplet mismatches was 10 (range 1-22) and 11 (range 1-23), respectively. Median number of donor-recipient antigen mismatches was 5 (range 1-6) for HLA class I and 4 (range 0-6) for HLA class II. In multivariable analysis, recipients with a number of HLA Class II DPB eplet mismatches above the median for this cohort had an increased risk of graft loss (HR 8.14 [95% CI: 1.26 - 49], p = 0.02). Neither HLA class I nor class II eplet mismatching was significantly associated with rejection. Similarly, HLA class I and II antigen mismatching was not significantly associated with graft loss or rejection. A logistic regression analysis was performed, with the primary outcome as the dependent variable. After iteration, the most accurate prediction was obtained with prior heart surgery, prior ECMO, sensitization against HLA class II antigens, and number of HLA Class II DPB eplet mismatches as independent variables. The model accounted for 35% - 67% of variance in the primary outcome, with 95% of outcomes correctly predicted. <h3>Conclusion</h3> In our cohort of paediatric heart transplant patients, the number of HLA Class II DPB donor-recipient eplet mismatches was independently associated with graft loss. Molecular-level HLA mismatching may aid in identifying recipients at increased risk of long-term graft loss but further study is required.