Abstract:
|
Adaptive design has become more popular for novel drug development and more accepted by regulatory agencies in recent years. A Bayesian adaptive phase 2a hypertension study design example is presented and discussed, where multiple doses of a novel drugs are evaluated against placebo control. Placebo control arms is augmented using historical data. The primary endpoint is change from baseline at trough in seated Systolic Blood Pressure (SBP) at end of treatment phase (6 weeks with daily dosing).Different analysis models are conducted using Bayesian Mixed Model Repeated Measures (MMRM) analysis, Emax Models and Normal Dynamic Linear Models. Longitudinal Emax Models and longitudinal Normal Dynamic Models are also evaluated. An Interim analysis is planned when approximately 50% of randomized subjects have primary endpoint readout. Interim decision will be based on predictive probability of success calculated at interim. Design calibration for clinical and statistical parameters has been conducted through extensive simulations to explore operating characteristics at both interim and study end under different scenarios.
|