#2822 Can KDPI Estimate Long-Term Graft Survival in a Cohort of Portuguese Patients?

Abstract Background and Aims The Kidney Donor Profile Index (KDPI) was introduced in the United States in 2014 to support clinicians' decision-making regarding the acceptance or decline of potential donors. In Portugal, there is no valid stratification score to assist physicians in accepting or declining an organ offer. Our study aims to characterize a cohort of Portuguese kidney transplant recipients and donors and evaluate the correlation between KDPI and long-term graft survival. The results are interpreted considering the nuances of the Portuguese allocation system, which favors an age-matched donor-recipient allocation. Method We conducted an observational single-center retrospective cohort study, which included patients who received a deceased donor kidney transplant at our unit between January 2013 and December 2015. We excluded patients for whom the donor KDPI index was unavailable for retrieval. Besides the KDPI index, we collected data regarding the receptor's demographic, clinical, and immunological characteristics, as well as graft and recipient survival. The primary outcome was death-censored survival, and follow-up ended on the 31st of December 2023. The secondary outcome was survival with a functioning graft. Survival estimates were obtained using Kaplan-Meier curves. We performed univariate and multivariate survival analysis with log-rank tests and Cox proportional-hazards regression models. Results We selected 364 deceased-donor kidney recipients and 212 kidney donors. The mean KDPI was 60% and 53.3% of donors were classified as expanded criteria. The kidney recipients were predominantly male (68.4%), with a mean age of 50.16 years and an average of 4 HLA mismatches. Median follow-up was 102 months. Considering recipient characteristics according to pre-defined KDPI strata (≤ 20%, 21-85%, > 85%), recipients from higher KDPI were significantly older, with a mean age of 40.6, 48.8, and 61.1 years, respectively (p < 0.001), and had a higher average body mass index (p = 0.041). There were no significant differences regarding the number of HLA mismatches (p = 0.624) or the proportion of hypersensitized patients (p = 0.253) between classes. During follow-up, 16.8% patients presented graft failure and 9.9% died with a functioning graft. Cumulative death-censored graft survival at 5, 8, and 10 years was 87.7%, 87.1%, and 84.4% respectively. The cumulative percentage of patients living with a functioning graft at 5, 8, and 10 years was 83.6%, 76.7%, and 68.6%. Death-censored graft survival was lower in recipients of donors with higher KDPI (Fig. 1), although it did not reach statistical significance (p-value = 0.166). The number of HLA mismatches (p = 0.592) or the recipient's age (p = 0.513) did not significantly impact death-censored graft survival, whereas delayed graft function (DGF) was the variable with the strongest association in univariate analysis (p < 0.001). In a multivariate Cox regression model, KDPI (HR 1.02, p = 0.005) and DGF (HR 3.2, p < 0.001) correlated with graft loss, while older recipients (HR 0.971, p = 0.019) seemed to have a protective effect over death-censored survival. Patient survival with a functioning graft was significantly associated with KDPI strata (p = 0.008; Fig. 2), the recipient's age (p = 0.006), and DGF (p < 0.001) in univariate analysis. Considering KDPI kidneys ≥ 90%, 10-year cumulative death-censored survival was 67.9%. Conclusion In our sample, despite not reaching statistical significance, there was a trend for improved death-censored graft survival in patients with lower KDPI, which further emulates the results of other European cohorts. Nevertheless, the 10-year cumulative death-censored survival of higher KDPI kidneys (≥ 90%) was remarkable, showing that long-term acceptable outcomes can be achieved. The impact of KDPI on patient survival with a functioning graft is probably overestimated, given the bias introduced by the Portuguese matching system, which rewards an “old-for-old” donor-recipient allocation.