SRS187 - Reducing sample sizes in pancreas transplant trials: the role of one-year HbA1c as a continuous outcome and surrogate endpoint

Abstract Background One-year graft survival is a widely used outcome measure in pancreas transplantation research, but its declining event rate makes powering clinical trials increasingly challenging. Our recent analysis of the UNOS registry identified one-year HbA1c as the strongest independent predictor of long-term graft survival. Building on this, we evaluated whether one-year HbA1c can serve as a surrogate endpoint to reduce sample size requirements for future trials. Methods We conducted a retrospective cohort study of simultaneous pancreas-kidney (SPK) transplants reported to UNOS (2017–2024). Cox regression models were used to confirm the association between one-year HbA1c and graft survival. Simulations compared sample sizes required to detect clinically meaningful differences using graft loss (binary outcome) versus HbA1c (continuous outcome) as trial endpoints. Results Among 2917 SPK recipients with functioning grafts at one year, HbA1c was the strongest predictor of graft survival. In trial simulations, detecting a 0.5% reduction in mean one-year HbA1c (for example 7.0 to 6.5%) required 65 participants per group. By contrast, detecting a reduction in one-year graft loss from 12 to 9% required 1631 participants per group. Across a range of effect sizes, HbA1c consistently enabled sample size reductions of >90% compared with graft loss outcomes. Conclusions HbA1c at one year is not only the most powerful predictor of long-term pancreas graft survival but also a feasible and efficient surrogate endpoint for clinical trials. Adoption of HbA1c as a primary outcome measure could substantially reduce sample size requirements, accelerate evaluation of new interventions, and improve trial feasibility in pancreas transplantation.