In recent years, Bayesian methodology is getting more and more popular in the application on the early-development Hematology and Oncology trials (ED Hem-Onc). Mainly because Bayesian framework improves drug development in a statistical meaningful way.
Various methodologies had been originated over years in this space such as Bayesian model-based approaches and Bayesian model-assisted approaches. Applicable areas in ED Hem-Onc include but not limited to: a. Maximum Tolerated Dose (MTD) finding during dose escalation phase. b. Toxicity monitoring in dose expansion phase. c. Go/no-go decision making for advancing to later phase.
How we make the most out of the above implementation? For examples, how we build in a covariate to account for a single-agent via a combination? how we deal with intra-subject dose escalation? or how we utilize the Bayesian tool to guide the subject enrollment? The topic may be broken down to 3 subjects: 1. Landscape of Bayesian Methodology implementation. 2. Challenges encountered during implementation and idea sharing. 3. Software readiness.
I hope that this could be a topic of interest to the Statistical Society especially in Early-Phase Oncology.
|