Projecting long-term graft and patient survival after transplantation

Abstract

Objective: In spite of increases in short-term kidney transplant survival rates and reductions in acute rejection rates, increasing long-term graft survival rates remains a major challenge. The objective here was to project long-term graft- and survival-related outcomes occurring among renal transplant recipients based on short-term outcomes including acute rejection and estimated glomerular filtration rates observed in randomized trials.

Methods: We developed a two-phase decision model including a trial phase and a Markov state transition phase to project long-term outcomes over the lifetimes of hypothetical renal graft recipients who survived the trial period with a functioning graft. Health states included functioning graft stratified by level of renal function, failed graft, functioning regraft, and death. Transitions between health states were predicted using statistical models that accounted for renal function, acute rejection, and new-onset diabetes after transplant and for donor and recipient predictors of long-term graft and patient survival. Models were estimated using data from 38,015 renal transplant recipients from the United States Renal Data System. The model was populated with data from a 3-year, randomized phase III trial comparing belatacept to cyclosporine.

Results: The decision model was well calibrated with data from the United States Renal Data System. Long-term extrapolation of Belatacept Evaluation of Nephroprotection and Efficacy as Firstline Immunosuppression Trial was projected to yield a 1.9-year increase in time alive with a functioning graft and a 1.2 life-year increase over a 20-year time horizon.

Conclusions: This is the first long-term follow-up model of renal transplant patients to be based on renal function, acute rejection, and new-onset diabetes. It is a useful tool for undertaking comparative effectiveness and cost-effectiveness studies of immunosuppressive medications.

Keywords: decision model; end stage renal disease; modeling; renal transplant.