Description:
We analyzed in a cohort of 68,606 first deceased donor kidney transplantations reported to the Collaborative Transplant Study whether an epitope-based matching of donor-recipient pairs using the Predicted Indirectly ReCognizable HLA Epitopes algorithm (PIRCHE-II) is superior to the currently applied HLA antigen matching. PIRCHE-II scores were calculated based on split antigen HLA-A, -B, -DRB1 typing and adjusted to the 0-6 range of HLA mismatches. PIRCHE-II scores correlated strongly with the number of HLA mismatches (Spearman ρ=0.65, P<0.001). In multivariable analyses both parameters were found to be significant predictors of 5-year death-censored graft loss with high prognostic power (hazard ratio (HR) per adjusted PIRCHE-II score=1.102, per HLA mismatch=1.095; z-value PIRCHE-II: 9.8, HLA: 11.2; P<0.001 for both). When PIRCHE-II scores and HLA mismatches were analyzed simultaneously, their predictive power decreased but remained significant (PIRCHE-II: P=0.002; HLA: P<0.001). Influence of PIRCHE-II was especially strong in presensitized and influence of HLA mismatches in non-sensitized recipients. If the level of HLA-incompatibility was low (0−3 mismatches), PIRCHE-II scores showed a low impact on graft survival (HR=1.031) and PIRCHE-II matching did not have additional significant benefit (P=0.10). However, if the level of HLA-incompatibility was high (4−6 mismatches), PIRCHE-II improved the positive impact of matching compared to applying the traditional HLA matching alone (HR=1.097, P=0.005). Our results suggest the PIRCHE-II score is useful and can be included into kidney allocation algorithms in addition to HLA matching; however, at the resolution level of HLA typing that is currently used for allocation it cannot fully replace traditional HLA matching.