Operational post-keratopasty graft tolerance due to differential HLAMatchmaker matching

Purpose: Penetrating keratoplasty can restore vision in corneal blindness. However, immunologic rejection threatens graft survival. Matching donors at swine leukocyte antigen (SLA)-class II convey allo-specific tolerance in a large animal kidney-transplantation model despite mismatches at SLA-class I. The same matching pattern seems to account for the blood transfusion effect in kidney transplantation. Relying on the molecular basis of HLAMatchmaker eplets, we assessed whether this finding would also apply to keratoplasty, and if it would enhance the benefit from matching human leukocyte antigen (HLA)-class I alone. Methods: We retrospectively selected two independent cohorts comprising 586 and 975 penetrating keratoplasties. Our computations revealed a quantitative tolerogenicity factor analogous to the animal model. The number of mismatched HLA-class I eplets functioned as a factor for conventional histocompatibility. In the first cohort, we empirically determined the thresholds with the highest predictive power on graft rejection for both factors, and confirmed those thresholds in the second cohort. We applied Cox proportional hazards regression for these analyses. Results: The thresholds with highest predictive power revealed 220 eplets2 for the tolerance factor and 10 eplets for HLAclass I histocompatibility. The respective hazards ratios were 2.22 (p=0.04) versus 3.63 (p<0.01) in the first cohort and 2.09 (p<0.01) versus 1.51 (p=0.02) in the second, confirmatory cohort. The threshold factors proved to be additive in predicting immune reactions in both cohorts, (hazard ratios 2.66 in cohort 1 versus 1.70; p<0.01 in cohort 2). Conclusions: Operational tolerance may be inducible by balanced matching of HLA-class I and II HLAMatchmaker eplets. Furthermore, such tolerance is additive to histocompatibliity at HLA-class I.