The value of a pre-diagnostic surrogate endpoint for T1D prevention trials is based on detection of efficacy and association with T1D. Thus, among multiple antibody+, dysglycemic individuals with baseline and 6-month OGTTs, we assessed metabolic endpoints for: 1) detecting a delay in metabolic decline from baseline to 6 months between placebo and teplizumab groups in the TrialNet teplizumab T1D prevention trial (TN 10), and 2) predicting T1D by metabolic change from baseline to 6 months in the TrialNet Pathway to Prevention study (TN01). The Table shows that the 6-month change in C-peptide (Cpep) measures (AUC and 30-0 minute) differed significantly between placebo and teplizumab groups, but not the change in glucose AUC. However, prediction of T1D, indicated by receiver operating curve AUC (ROCAUC), tended to be more accurate for 6-month change of glucose AUC than for Cpep changes. Although detection of a teplizumab effect by Cpep measures was comparable to that of combined measures (Index60 and Cpep AUC/glucose AUC), the latter were more accurate predictors.
In conclusion, among metabolic endpoints there was discordance between detection of a teplizumab effect and prediction accuracy of T1D. However, such discordance suggests that metabolic endpoints provide more insight into understanding effects of teplizumab and other preventive treatments than the T1D endpoint alone.
E. K. Sims: None. D. D. Cuthbertson: None. L. M. Jacobsen: None. K. C. Herold: Consultant; Self; Provention Bio, Inc., Viela Bio, Consultant; Spouse/Partner; Provention Bio, Inc. J. Sosenko: None.