Longitudinal C-Peptide End Point Could Strengthen T1D Trial Design: Samuel Shangwu Wu, PhD

A novel end point—time to minimal residual C-peptide—could offer a more sensitive and interpretable alternative to the stimulated C-peptide area under the curve (AUC) that has long anchored efficacy assessments in recent-onset type 1 diabetes (T1D) immunotherapy trials, according to research presented at the American Diabetes Association 2026 Scientific Sessions.

The conventional approach captures a single C-peptide measurement window, typically at 12 to 24 months, leaving the longitudinal trajectory of beta-cell decline largely unused. Samuel Shangwu Wu, PhD, professor and deputy director of the Health Informatics Institute at the University of South Florida, whose work was selected as a Presidents' Abstract, argued that this single-timepoint framework fails to reflect how beta-cell loss actually unfolds across the course of a trial.

"We more efficiently utilize the data captured in the trial," Wu explained, noting that the time-to-event model incorporates measurements from baseline through 3, 6, 9, 12, and up to 30 months. Beyond statistical efficiency, he emphasized the clinical communication advantage: a time-to-event framing is more intuitive for both clinicians and patients than an AUC difference.

The threshold defining "minimal residual" C-peptide was anchored at 0.3 nmol/L, drawn from a 2021 publication demonstrating that patients above vs below this level show meaningful differences in glycemic outcomes. Importantly, no significant clinical distinction was observed between patients falling below 0.3 nmol/L and those below the assay detection limit of 0.003 nmol/L—supporting 0.3 as a clinically relevant inflection point.

Sensitivity analyses further validated the approach: Modest adjustments to the threshold did not materially alter treatment comparisons between active and placebo arms, suggesting the end point is robust rather than threshold-dependent.

The findings position time to minimal residual C-peptide as a candidate primary end point for future T1D immunotherapy trials, with potential to improve statistical power and clinical relevance simultaneously.