🤖 AI Summary
This work proposes a Bayesian optimization framework to address key limitations in clinical trial design, including poorly calibrated priors, inefficient sequential allocation, and decisions that inadequately prioritize patient benefit. The approach leverages historical data to inform prior distributions, employs sufficient statistics and backward induction to transform the infinite-horizon stopping problem into a finite decision table, and replaces conventional error-rate control with a joint efficacy–toxicity utility model to maximize patient benefit. Integrating Thompson sampling, Bayesian inference, and utility theory, the method demonstrates robust performance in real-world applications—including ECMO trials, the CALGB 49907 breast cancer study, and platform trials—thereby providing theoretical support for the upcoming 2026 FDA guidance on Bayesian clinical trial design.
📝 Abstract
We examine three landmark clinical trials -- ECMO, CALGB~49907, and I-SPY~2 -- through a unified Bayesian framework connecting prior specification, sequential adaptation, and decision-theoretic optimisation. For ECMO, the posterior probability of treatment superiority is robust across the range of priors examined. For CALGB, predictive probability monitoring stopped enrolment at 633 instead of 1800 patients. For I-SPY~2, adaptive enrichment graduated nine of 23 arms to Phase~III. These case studies motivate a methodological contribution: exact backward induction for two-arm binary trials, where Beta-Binomial conjugacy yields closed-form transitions on the integer lattice of success counts with no quadrature. A P\'olya-Gamma augmentation bridges this to covariate-adjusted logistic regression. Simulation reveals a fundamental tension: the optimal Bayesian design reduces expected sample sizes to 14--26 per arm (versus 42--100 for alternatives) but with substantially lower power. A calibrated variant embedding the declaration threshold in the terminal utility improves power while maintaining sample-size savings; varying the per-stage cost traces a power frontier for selecting the preferred operating point, with suitability highest in patient-sparing contexts such as rare diseases and paediatrics. The P\'olya-Gamma Laplace approximation is validated against exact calculations (mean absolute error below 0.01). We discuss implications for the 2026 FDA draft guidance on Bayesian methodology.