HLA Mismatching and Transplant Outcome in the Ptcy Era: A Comprehensive Study By the EBMT Cellular Therapy and Immunobiology Working Party

Introduction HLA compatibility is a mainstay in allogeneic HCT, resulting in a preference for well-matched unrelated donors (UD). Recent interest in improving accessibility for patients lacking a fully matched donor and the introduction of post-transplant cyclophosphamide (PTCy) for GvHD prophylaxis are increasing the use of mismatched (i.e. <10/10) UD. However, the role of HLA mismatching in current HCT practice, including PTCy, remains unclear. Methods We studied outcomes in 17,276 adult patients reported to the EBMT Registry who underwent UD-HCT between 2005-2020. Most patients were treated for AML, ALL, MDS, or MPN (67.5%) using reduced-intensity conditioning (56%) and in vivo T-cell depletion (76%). 23.5% of the transplants had one (9/10; n=3,561) or two (8/10; n=499) high-resolution HLA mismatches (mM). GvHD prophylaxis with calcineurin inhibitors was used in all patients. PTCy was used in 7% (n=924) of the 10/10 and 15% (n=599) of the <10/10 transplants. Multivariable models were constructed to analyze the effect of the number, class, locus, and nature (i.e. high vs low-resolution) of mM in the presence or absence of PTCy. Results OS was significantly lower in transplants across HLA mM (45% [95% CI 40-50%] in 8/10 and 47% [45-49%] in 9/10 vs 52% [51-53%] in 10/10 at 60 months; p<0.001; Figure 1a). In multivariable analysis, the presence of one (HR 1.24 [99% CI 1.15-1.34]; p<0.001) or two (HR 1.29 [1.09-1.54]; p<0.001) mM associated with higher risks of mortality compared to 10/10 transplants. HLA class I (HR 1.31 [1.20-1.42]; p<0.001) but not HLA class II mM (HR 1.07 [0.93-1.22]; p=0.23) were associated with significantly worse OS, even when excluding antigen recognition domain-matched pairs. For class I, HLA-A (HR 1.37 [1.21-1.54]; p<0.001) and HLA-B (HR 1.44 [1.23-1.69]; p<0.001) mM conferred higher risks than HLA-C mM (HR 1.16 [1.01-1.33]; p=0.005), and antigen-level mM associated with worse OS than allelic mM (HR 1.22 [1.01-1.46]; p=0.006). Similar associations were observed for GRFS, RFS, NRM, and aGVHD, but no significant differences were observed for relapse or chronic GvHD. Non-permissive HLA-DPB1 mM increased aGvHD risks in both 9/10 and 10/10 pairs, but NRM only in the latter. The use of PTCy significantly reduced the risks of GvHD and mortality compared to standard prophylaxis. However, the effects of HLA mM were similar regardless of PTCy use (non-significant interaction; p=0.22), with a single mM conferring significantly increased risks of mortality both in the presence (HR 1.38 [1.14-1.68]; p<0.001) and absence (HR 1.23 [1.13-1.33]; p<0.001) of PTCy (Figure 1b). Conclusion In contemporary HCT, HLA disparity is associated with increased risks of mortality, mainly driven by HLA class I. These associations are present even under GvHD prophylaxis with PTCy, highlighting the need to continue defining better tolerated mM to provide the best possible outcome for all patients.