The University of Texas MD Anderson Cancer Center at Houston
Department of Biostatistics

My research focuses on Bayesian clinical trial design in both early- and late-phase settings and involves methodological development and software implementations.

Early-phase trials evaluate the safety and efficacy of investigational treatments. My research addresses certain complexities of evaluating novel cancer treatments, such as molecularly targeted agents and immunotherapy, which include late-onset toxicity outcomes and non-increasing dose-efficacy probability curves. Examples of existing research include deriving a survival endpoint-based Bayesian optimal interval design that is appropriate for incorporating late-onset toxicity outcomes, and developing a Bayesian random-effects meta-analysis method for identifying the optimal biological dose which allows for a possible plateau in dose-efficacy.

Late-phase trials compare the investigational treatment against the current standard of care. My research involves hybrid designs which incorporate historical controls from real world data, which are beneficial in rare disease and pediatric populations. Examples of existing research include developing a hybrid empirical balancing prior to calibrate the covariate distributions between concurrent and nonconcurrent subjects in platform trials, and developing an extension of the power prior where the discounting weights are computed separately for each external control based on compatibility with the randomized control data.

Active research projects address the practical implementation of Bayesian methods from the regulatory perspective, such as methods for computing the effective sample size of a hybrid control arm and calibrating frequentist operating characteristics of Bayesian designs.

PubMed

Education & Training

PhD - University of North Carolina at Chapel Hill - 2021