Checkerboard: a Bayesian efficacy and toxicity interval design for phase I/II dose-finding trials

Area of Science:

• Clinical trial design
• Pharmacometrics
• Oncology drug development

Background:

• Conventional phase I trial designs, focused on maximum tolerated dose (MTD), are challenged by targeted therapies and immunotherapies.
• There is a need for adaptive trial designs that integrate both toxicity and efficacy data.

Purpose of the Study:

• To introduce the checkerboard design, a novel model-assisted phase I/II clinical trial design.
• To evaluate the checkerboard design's performance against existing model-based designs.

Main Methods:

• The checkerboard design models the joint distribution of toxicity and efficacy.
• It divides toxicity and efficacy into intervals and adaptively determines optimal doses using interim data.
• The design is extended to accommodate continuous efficacy endpoints.

Main Results:

• The checkerboard design demonstrates competitive performance compared to current model-based phase I/II designs.
• Simulation studies confirm its efficacy in dose selection.
• The design's decision rules can be pretabulated for easier implementation.

Conclusions:

• The checkerboard design provides a simpler and more practical alternative for phase I/II clinical trials.
• It effectively balances toxicity and efficacy, offering an improved approach for dose-finding studies in the era of precision medicine.