*Purpose: HLA genes are in linkage disequilibrium with each other. Sequence defined HLA targets (eplets) are associated with alleles and may be shared within and across HLA loci. This results in complex relatedness of eplets in the context of particular donor:recipient mismatches. To model interactions between eplet mismatches, we studied the Scientific Registry of Transplant Recipients.

*Methods: The dataset included 118,382 pairs of donors and unsensitized US first kidney transplant recipients (2000 to 2015). Allele-level HLA genotypes (A, B, C, DRB1 and DQB1) were imputed using an algorithm from the National Marrow Donor Program. HLA genotypes were converted to epitypes consisting of 449 potential eplets (223 Class I: 72 AbVer, 151 Non-AbVer, 226 Class II: 72 AbVer, 154 Non-AbVer). Eplet mismatches were defined as eplets present in the donor but absent from the recipient’s epitype. We applied Weighted Gene Correlation Network Analysis (WGCNA) to identify eplet mismatch profiles. To verify the observed eplet mismatch profiles, we studied an independent dataset of 48,384 pairs of sensitized transplant recipients and their donors and confirmed the consistency of eplet mismatch profiles.

*Results: A total of 67 profiles were identified. Each profile involved 2 to 12 eplets. Both AbVer and non-AbVer eplet mismatches were represented. The observed eplet profiles segregated by class such that only eplets from the same class (I or II) formed a profile. Of 449 potential eplet mismatches, 239 (109 class I: 40 AbVer and 69 non-AbVer, 101 class II: 42 Abv and 59 non-AbVer) were singleton mismatches and did not form any profile with other eplets.

*Conclusions: Our study suggests that consideration of cumulative eplet mismatch load may be insufficient to inform strategies to mitigate immune-mediated injury. Profiles of interacting eplets should also be considered to enhance compatibility at the time of organ allocation.

« Back to 2020 American Transplant Congress