LAUSR

Clinically tolerable, permissive HLA-DPB1 mismatches defined by the T-cell epitope (TCE) model improve the selection of unrelated donors in allogeneic hematopoietic cell transplantation (HCT). Mechanistically, overlapping immunopeptidomes in structurally close HLA-DP allotypes with similar peptide-binding motifs are fundamental for permissiveness in vitro, but their relevance in vivo is still unknown. Here, we hypothesized that a similarity measure reflecting the peptide-binding region of HLA-DPB1 alleles could constitute a proxy for immunopeptidome overlap and hence predict permissive mismatches in the clinical setting. To test this, we used multidimensional scaling techniques based on 28 polymorphic amino acid positions, resulting in the stratification of HLA-DPB1 alleles from the heterogeneous TCE group 3 (TCE3) into a subgroup of four frequent, structurally related, and less immunogenic "core" TCE3 alleles compared to the remaining "non-core" alleles. In a CIBMTR cohort of 5140 10/10-matched patients transplanted for AML, ALL, or MDS from 2008-2017, the risks of aGVHD II-IV increased progressively from "core" TCE3 permissive (N=930; HR 1.12 [0.98-1.28]; p=0.1012) to "non-core" TCE3-permissive (N=1286; HR 1.24 [1.06-1.46]; p= 0.0082), and non-permissive mismatches (N=2023; HR 1.32 [1.16-1.50]; p<.0001) compared to allele-matched patients (N=785). "Core" TCE3-permissive pairs (HR 0.78 [0.68-0.88]; p=0.0002), but not "non-core" TCE3-permissive pairs (HR 0.95 [0.83-1.09]; p=0.4578) showed significantly lower risks of TRM when compared to non-permissive pairs. Our results suggest that frequent mismatches between structurally similar "core" HLA-DPB1 alleles are the main drivers of permissiveness after HCT, and provide evidence for a role of immunopeptidome differences between mismatched HLA-DPB1 alleles in the clinical outcome of HCT.