Association of kidney transplantation with mortality on hemodialysis after graft failure

Background Although a substantial number of patients return to dialysis after kidney transplant failure, it remains controversial whether transplant-failure patients have a higher mortality risk than transplant-naïve patients on dialysis who have never undergone kidney transplantation. We compared outcomes of transplant-failure and transplant-naïve patients on hemodialysis. Methods Data from the Japanese National Dialysis Registry (2012–2013) were analyzed, including 220,438 prevalent hemodialysis patients. Multivariable Cox models were used to compare all-cause, cardiovascular, and infection-related mortality during 1-year follow-up between transplant-failure and transplant-naïve patients. Multiple imputation and propensity score matching were utilized as sensitivity analyses. Results During 209,377 patient-years of follow-up, 18,648 all-cause deaths (8.5% of all patients), 7700 cardiovascular deaths (41% of all-cause deaths), and 3806 infection-related deaths (20% of all-cause deaths) were observed. Adjusted hazard ratios [95% confidence intervals] for all-cause, cardiovascular, and infection-related deaths among transplant-failure patients were 0.81 [0.59–1.11], 0.54 [0.30–0.98], and 1.54 [0.92–2.59], respectively. Sensitivity analyses using multiple imputation and propensity score matching yielded similar results. Conclusions This Japanese cohort study suggested that a cardiovascular mortality risk of transplant-failure patients could be significantly lower than that of transplant-naïve patients, while there might be a trend toward a higher infection-related mortality risk in transplant-failure patients. However, this retrospective, single-country study can introduce an immortal time bias in transplant-failure patients, and limit the external validity. Further prospective studies are warranted to improve the comparability of outcomes between transplant-failure and transplant-naïve patients, and to examine worldwide the generalizability of the potential cardiovascular benefit of kidney transplantation even after returning to dialysis. Graphic abstract