63-OR: A Model-Based Clinical Trial Simulation Tool to Optimize Clinical Trial Designs for Type 1 Diabetes Prevention Studies

Abstract

Trials seeking to delay or prevent type 1 diabetes (T1D) continue to face multiple challenges, including both clinical endpoint (s) and cohort selection. While it is clear that individuals who possess islet cell autoantibodies (AAbs) are at elevated risk for developing T1D, it remains unclear as to what additional factors, including metabolic activities and anthropometric data, may contribute to this increased risk. The aim of this project was to inform T1D prevention trial designs by developing a quantitative joint model based-clinical trial simulation (CTS) tool. Individual-level data obtained from the NIH TEDDY and TrialNet TNnatural history studies were used in this tool’s development. To describe T1D disease progression quantitatively, both longitudinal and time-to-event variables were simultaneously quantified as a joint model, with results from a 2-hour oral glucose tolerance test (OGTT) assessed at each visit included as a time-varying longitudinal covariate affecting the T1D diagnosis event. Our study focused on individuals at risk of developing T1D with two or more AAbs. Among the investigated individuals, those with a high BMI and the presence of GADA AAb at baseline had less risk, while those with an elevated HbA1c value, high fasting OGTT value, increased fasting OGTT to 2-hour OGTT ratio, and the presence of the IA2A AAb at baseline had higher risk. The developed model based-CTS tool, where users can specify the trial design, participants’ characteristics and hypothetical drug effect, provides a framework to perform T1D prevention clinical trial simulations with the intention to assist in future trial design decision-making processes. Disclosure J.F. Morales: None. R. Muse: None. J. Podichetty: None. S. David: None. M.A. Atkinson: None. M.J. Haller: Advisory Panel; SAB Biotherapeutics. Consultant; MannKind Corporation, Sanofi. M. Campbell-Thompson: Consultant; TRex Bio. K. Romero: None. I. O'Doherty: None. S. Schmidt: None. S. Kim: None. Funding JDRF (grant #: 2-SRA-2020-903-A-N) ; NIDDK Central Repositories