Samuel Shangwu Wu, PhD, is a professor and deputy director of the Health Informatics Institute at the University of South Florida.

Samuel Shangwu Wu, PhD

Time to Minimum Residual C-Peptide As a Predictor of Treatment Efficacy in T1D

Time to minimum residual C-peptide (TMRCP) is a clinically interpretable and statistically efficient endpoint for future studies in 

American Diabetes Association (ADA) Scientific Sessions 2026

 in New Orleans, Louisiana, by Samuel Shangwu Wu, PhD, professor and deputy director of the Health Informatics Institute at the University of South Florida, these data reflect a potentially more reliable endpoint for patients with T1D enrolled in clinical trials, utilizing all available longitudinal data to provide a stable and reliable efficacy measure.

Existing methods of measuring beta-cell function have historically proven difficult due to the challenges inherent in direct tissue sampling. Alternative methods, including insulin or C-peptide measurements, are similarly difficult due to changing levels of insulin resistance or cellular exhaustion.

“We feel that this time is more interpretable, and the patients will be easier to understand,” Wu told 

 in an exclusive interview. “And then we estimated that time with, of course, some extrapolation, and then we used these to compare whether their treatment is effective or not beyond the single time point in the measurement.”

TMRCP is defined as the time until C-peptide drops to a prespecified value, which represents the functional loss of beta-cell secretory capacity. Wu and colleagues estimated TMRCP by utilizing a nonlinear mixed-effects model and applying it to log-transformed mean area under curve (MAUC) C-peptide values collected via mixed-meal tolerance testing. The team states that this approach utilizes all available longitudinal measurements while accounting for repeated observations within individuals. Additionally, the strategy can manage variable visit schedules or incomplete follow-up from a given patient.

Wu and colleagues applied this estimation framework to 6 randomized, placebo-controlled, double-blind trials, the results of which were collected via TrialNet. Each trial investigated patients with recent-onset T1D, totaling 597 participants. Via Wilcoxon rank-sum tests on TMRCP estimates, the team highlighted the same 3 main treatments as effective as the original analyses.

Median TMRCP for the 3 effective treatments compared to placebo were 128 vs 84 days for Rituximab (

 = .03), and 110 vs 53 days for anti-thymocyte globulin (

 = .027). Differences were <4 days for the 3 non-effective treatments, including canakinumab, mycophenolate mofetil and daclizumab combined therapy, and glutamic acid decarboxylase.

Ultimately, the team concluded that, unlike fixed-time endpoints, TMRCP measures the loss of residual insulin secretion while providing patient-specific estimates to support robust treatment comparisons.

“For this TMRCP in the estimate, we used something like a random effect – in theory, we borrow the information from other patients to estimate each patient’s decline,” Wu said. “There is lots of variation between patients in terms of the rate of decline, but we borrow that information to estimate their slope and make it a little less variable.”

Editors’ Note: Wu reports disclosures with Vertex Pharmaceuticals and United HealthCare Services.

Wu S, Ding A, Krischer J, et al. Time to Minimal Residual C-Peptide as an Efficacy Measure in Recent-Onset Type 1 Diabetes Immunotherapy Trials. Abstract presented at the American Diabetes Association (ADA) Scientific Sessions 2026, New Orleans, LA. June 5-8, 2026.

Hannon TS, Kahn SE, Utzschneider KM, et al. Review of methods for measuring β-cell function: Design considerations from the Restoring Insulin Secretion (RISE) Consortium. 

Videos

Samuel Shangwu Wu, PhD, discusses his recent work evaluating TMRCP as an effective endpoint for future T1D trials.

Samuel Shangwu Wu, PhD

Articles