FC 119SURVIVAL BENEFIT OF KIDNEY TRANSPLANTATION COMPARED TO REMAINING ON WAITLIST ACROSS DIFFERENT AGES OF TRANSPLANT CANDIDATES: A RETROSPECTIVE COHORT STUDY USING TARGET TRIAL EMULATION

Abstract

Abstract Background and Aims Kidney transplantation is considered to be the optimal treatment strategy for eligible end stage renal disease patients. However, the body of evidence to underpin the anticipated survival advantage for kidney transplant recipients is weak, as random treatment allocation to either kidney transplantation or remaining on dialysis is not feasible and previously reported results obtained from observational studies did not allow for causal interpretation. The aim of this study is to investigate survival differences of kidney transplantation compared to remaining waitlisted on dialysis across different transplant candidate ages applying causal inference methodology. Method We conducted a retrospective cohort study using the Austrian Dialysis and Transplant Registry. We included all maintenance dialysis patients who were waitlisted for their first kidney transplant between January 2000 and December 2018 and utilized repeated updates on waitlisting status and relevant covariates. To estimate the causal effect of kidney transplantation compared to remaining waitlisted on all-cause survival we applied a sequential Cox approach mimicking a series of target trials, where each trial started at the time of a transplantation. In each of these emulated trials transplanted patients were classified as treated and patients with current active waitlisting status as controls, and the groups were balanced for covariates by inverse probability weighting. Controls who were transplanted at later times were censored but assigned to the treated group in a later target trial of the series. All trials were combined into a single data set and analyzed by a Cox proportional hazards model using inverse probability weighting also to adjust for artificial censoring. Additionally, we evaluated potential effect modifications by age at trial initiation (continuous) and stratified our analyses by time on waitlist before trial initiation (up to 1 year, between 1 and 2 years, and more than 2 years). Results are reported as hazard ratios (HRs), 5-year survival probabilities and restricted mean survival time together with respective bootstrap confidence intervals (CIs). Results The study cohort consisted of 4206 patients, of whom one third were women and the mean age was 52 years. In total, 3399 patients (81%) received a transplant and 1256 patients died. The median time from waitlisting to transplantation was 1.8 years. Overall, patients who received a kidney transplant had a significant survival benefit compared to patients who remained waitlisted (HR 0.36, 95% CI 0.29 to 0.43). Assessing survival across different ages showed a significant benefit for kidney transplantation for patients between 32 and 77 years of age at time of transplantation (e.g. HR at age of 70: 0.43, 95% CI 0.33 to 0.54). For older and younger patients our analysis did not provide definitive conclusions due to limited sample sizes. Transplanted patients had higher predicted survival and longer restricted mean survival time compared to patients remaining waitlisted. For example, within 5 years after engraftment, a transplanted patient 70 years of age at trial initiation had a 0.28 higher survival probability (95% CI 0.20-0.37) and was expected to gain 0.75 years of survival time. Our stratified analyses showed a survival benefit for kidney transplantation regardless of time on waitlist before trial initiation across all ages. Conclusion Our study provides robust evidence based on state-of-the-art causal inference methodology for increased survival after kidney transplantation across different transplant candidate ages and irrespective of time on waiting list.