Assessing the influence of graft loss on 4-year patient survival after simultaneous pancreas-kidney transplantation: Kaplan-Meier versus Competing Risk Analysis model.

Abstract

Graft loss increases the risk of patient death after simultaneous pancreas-kidney (SPK) transplantation. The relative risk of each graft failure is complex due to the influence of several competing events. This retrospective, single-center study compared 4-year patient survival according to the graft status using Kaplan-Meier (KM) and Competing Risk Analysis (CRA). Patient survival was also assessed according to five eras (Era 1: 2001-2003; Era 2: 2004-2006; Era 3: 2007-2009; Era 4: 2010-2012; Era 5: 2012-2015). Between 2000 and 2015, 432 SPK transplants were performed. Using KM, patient survival was 86.5% for patients without graft loss (n=333), 93.4% for patients with pancreas graft loss (n=46), 43.7% for patients with kidney graft loss (n=16), and 25.4% for patients with pancreas and kidney graft loss (n=37). Patient survival was underestimated using KM versus CRA methods in patients with pancreas and kidney graft losses (25.4%vs. 36.2%), respectively. Induction with lymphocyte depleting antibodies was associated with 81% reduced risk (HR.19, 95% CI.38-.98, p=.0048), while delayed kidney function (HR 2.94, 95% CI 1.09-7.95, p=.033) and surgical complications (HR 2.94, 95% CI 1.22-7.08, p=.016) were associated with higher risk of death. Four-year patient survival increased from Era 1 to Era 5 (79%vs. 87.9%, p=.047). In this cohort of patients, kidney graft loss, with or without pancreas graft loss, was associated with higher mortality after SPK transplantation. Compared to CRA, the KM model underestimated survival only among patients with pancreas and kidney graft losses. Patient survival increased over time.