A standardized metric to enhance clinical trial design and outcome interpretation in type 1 diabetes.

Abstract

The use of a standardized outcome metric enhances clinical trial interpretation and cross-trial comparison. If a disease course is predictable, comparing modeled predictions with outcome data affords the precision and confidence needed to accelerate precision medicine. We demonstrate this approach in type 1 diabetes (T1D) trials aiming to preserve endogenous insulin secretion measured by C-peptide. C-peptide is predictable given an individual's age and baseline value; quantitative response (QR) adjusts for these variables and represents the difference between the observed and predicted outcome. Validated across 13 trials, the QR metric reduces each trial's variance and increases statistical power. As smaller studies are especially subject to random sampling variability, using QR as the outcome introduces alternative interpretations of previous clinical trial results. QR can provide model-based estimates that quantify whether individuals or groups did better or worse than expected. QR also provides a purer metric to associate with biomarker measurements. Using data from more than 1300 participants, we demonstrate the value of QR in advancing disease-modifying therapy in T1D. QR applies to any disease where outcome is predictable by pre-specified baseline covariates, rendering it useful for defining responders to therapy, comparing therapeutic efficacy, and understanding causal pathways in disease.