Impact of an HLA-DPB1 Non-Permissive Mismatch By the Revised T Cell Epitope (TCE)-Core Algorithm on Survival Following Haploidentical Donor Transplantation (HIDT) with Post-Transplant Cyclophosphamide (PTCy)

The presence of an HLA-DPB1 non-permissive mismatch (NPMM) by the TCE-3 model has been associated with improved survival following HIDT utilizing post-transplant cyclophosphamide (PTCy). A revised model (TCE-core) which further separates TCE “group 3” alleles into “core”(C) and “non-core” (NC) alleles has been developed (Arrietta-Bolanos et al. Blood 2022). In this new model, a formerly permissive mismatch (PMM) resulting from “group 3” alleles in both donor and recipient are now considered a C-NPMM if one or more of those alleles is NC (see figure), and the total number of NPMMs by the TCE-core model results from adding these C-NPMMs to conventional NPMM. In order to study the impact of an HLA-DPB1 NPMM by the TCE-core model on outcomes following PTCy-based HIDT, we evaluated 242 consecutive HIDT recipients with ALL, AML or MDS transplanted between 2005 - 2021 at a single institution, (median age 51 [19,80], 38% non-White, 32% DRI high/v. high). We further examined the effect of the vector of non-permissiveness (see figure), hypothesizing that a NPMM in the GVH or bidirectional direction may have a more significant impact on relapse and survival than an HVG-only NPMM. Median follow-up was 62 [23, 199] months. A NPMM was seen in 129 donor-recipient pairs (53%), of which 65 (27%) were either in the GVH-only or bidirectional vector. In univariate analysis, a trend for improved 5-yr OS was seen in the context of a TCE-core NPMM vs. matched/PMM HLA-DPB1 (59% vs. 49%, p=0.06). Restricting TCE-core NPMM to the GVH/bidirectional vector further improved 5-yr OS (64% vs. 49%, p=0.023, see figure). Compared to an HVG-only NPMM, a GVH/bidirectional NPMM was associated with a significantly lower 5-yr risk of relapse (23% vs. 45%, p=0.016). In Cox multivariable analysis, adjusted for disease risk index, regimen intensity and donor age, a GVH/bidirectional mismatch was associated with improved OS (HR 0.62, 95% CI 0.38-0.99, p=0.047), compared with HLA-DPB1 match/PMM. Although not statistically significant, a trend was also seen for DFS (HR 0.69, 95%CI 0.44-1.09, p=0.11) and relapse (HR 0.63, 95%CI 0.35-1.14, p=0.12). Compared with an HVG-only TCE-core NPMM, a GVH/bidirectional NPMM was associated with significantly lower relapse (HR 0.48, 95%CI 0.26-0.89, p=0.020). In summary, the presence of TCE-core GVH/bidirectional NPMM is associated with less relapse and significantly improved OS following PTCy-based HIDT. This association should be confirmed in a large registry analyses and potentially incorporated in future donor selection algorithms.