There are complex risk-benefit tradeoffs with different transplantation strategies for end-stage liver disease patients on renal support. Using a Markov discrete-time state transition model, we compared survival for this group with 3 strategies: simultaneous liver-kidney (SLK) transplantation, liver transplantation alone (LTA) followed by immediate kidney transplantation if renal function did not recover, and LTA followed by placement on the kidney transplant wait list. Patients were followed for 30 years from the age of 50 years. The probabilities of events were synthesized from population data and clinical trials according to Model for End-Stage Liver Disease (MELD) scores (21-30 and >30) to estimate input parameters. Sensitivity analyses tested the impact of uncertainty on survival. Overall, the highest survival rates were seen with SLK transplantation for both MELD score groups (82.8% for MELD scores of 21-30 and 82.5% for MELD scoresâ€‰>â€‰ 30 at 1 year), albeit at the cost of using kidneys that might not be needed. Liver transplantation followed by kidney transplantation led to higher survival rates (77.3% and 76.4%, respectively, at 1 year) than placement on the kidney transplant wait list (75.1% and 74.3%, respectively, at 1 year). When uncertainty was considered, the results indicated that the waiting time and renal recovery affected conclusions about survival after SLK transplantation and liver transplantation, respectively. The subgroups with the longest durations of pretransplant renal replacement therapy and highest MELD scores had the largest absolute increases in survival with SLK transplantation versus sequential transplantation. In conclusion, the findings demonstrate the inherent tension in choices about the use of available kidneys and suggest that performing liver transplantation and using renal transplantation only for those who fail to recover their native renal function could free up available donor kidneys. These results could inform discussions about transplantation policy.
ASJC Scopus subject areas