PIRCHE-II Is Related to Graft Failure after Kidney Transplantation

Abstract

Individual HLA mismatches may differentially impact graft survival after kidney transplantation. Therefore, there is a need for a reliable tool to define permissible HLA mismatches in kidney transplantation. We previously demonstrated that donor-derived Predicted Indirectly ReCognizable HLA Epitopes presented by recipient HLA class II (PIRCHE-II) play a role in de novo donor-specific HLA antibodies formation after kidney transplantation. In the present Dutch multi-center study, we evaluated the possible association between PIRCHE-II and kidney graft failure in 2,918 donor-recipient couples that were transplanted between 1995 and 2005. For these donors-recipients couples, PIRCHE-II numbers were related to graft survival in univariate and multivariable analyses. Adjusted for confounders, the natural logarithm of PIRCHE-II was associated with a higher risk for graft failure [hazard ratio (HR): 1.13, 95% CI: 1.04-1.23, p = 0.003]. When analyzing a subgroup of patients who had their first transplantation, the HR of graft failure for ln(PIRCHE-II) was higher compared with the overall cohort (HR: 1.22, 95% CI: 1.10-1.34, p < 0.001). PIRCHE-II demonstrated both early and late effects on graft failure in this subgroup. These data suggest that the PIRCHE-II may impact graft survival after kidney transplantation. Inclusion of PIRCHE-II in donor-selection criteria may eventually lead to an improved kidney graft survival.

Keywords: HLA antigens; HLA matching; PIRCHE-II; graft rejection; kidney transplantation.

Figure 1

Flow chart of enrollment, inclusion, and exclusions for the study.

Figure 2

Modeling of the increased hazard on graft failure for each Predicted Indirectly ReCognizable HLA Epitopes presented by recipient HLA class II (PIRCHE-II). Based on the multivariable analyses, the kidney graft failure risk is plotted as a function of the number of PIRCHE-II for both the overall cohort and for the first-transplantations only group. For example for the overall cohort, a patient with a PIRCHE-II value of 2.718 [ln(2.718) = 1] has a hazard of 1.13 on kidney graft failure, whereas a patient with a PIRCHE-II value of 7.388 [=2.718*2.718; ln(7.388) = 2] has a hazard of 1.26 (an increase of 0.13 compared with a PIRCHE-II value of 2.718) on kidney graft failure. Similar calculations were performed for other PIRCHE-II values and for the first-transplantations only group. The differences in hazard on kidney graft failure between different PIRCHE-II values decrease for higher PIRCHE-II values.

Figure 3

Predicted Indirectly ReCognizable HLA Epitopes presented by recipient HLA class II (PIRCHE-II) affects both early and long-term graft outcome. The hazard on kidney graft failure for ln(PIRCHE-II) at different virtual endpoints after transplantation is displayed in (A). Hazard ratios (HRs) and 95% CI’s were calculated for different virtual endpoints (1–2–3–4–5–6–9–12–18 months and 2–3–4–5–6–7–8–9–10 years after transplantation), as exemplified by the lower panel of (A). HRs were calculated using multivariable Cox proportional hazard models and smooth curve fitting was performed based on these data points. In the upper panel of (A), the solid line represents the HR and the dotted line represents the 95% CI. The kidney graft failure risks for different time spans when excluding subjects who experienced kidney graft failure before different time points are displayed in (B). For these analyses, the graft survival was set to 100% at each analyzed time point (1–2–3–4–5–6–9–12–18 months and 2–3–4–5–6–7 years after transplantation) and the HR were calculated at 10 years after transplantation, as exemplified in the lower panel of (B). The number of patients at risk at each analyzed time span is shown in Table S1 in Supplementary Material. Due to the low number of events and the limited follow-up time around 7 years after transplantation, the later time spans could not be analyzed. HR’s and 95% CI’s were calculated for different time spans using multivariable Cox proportional hazard models and smooth curve fitting was performed based on these data points. In the upper panel of (B), the solid line represents the HR and the dotted line represents the 95% CI. For both (A,B), the individual dots represent the HR’s at different virtual endpoints/time spans. HR’s were calculated using forward stepwise selection considering the covariates implemented in the initial multivariable model [ln(PIRCHE-II), donor age, recipient age, and year of transplantation].