A Bayesian Dose-finding Design for Drug Combination Trials with Delayed Toxicities.

Area of Science:

• Clinical Pharmacology
• Biostatistics
• Drug Development

Background:

• Drug combination trials present challenges in managing delayed toxicity.
• Accurate dose-finding is crucial for optimizing therapeutic efficacy and safety.
• Existing methods may not adequately address delayed toxicity in combination studies.

Purpose of the Study:

• To propose a novel Bayesian adaptive dose-finding design for drug combination trials.
• To specifically address the challenge of delayed toxicity outcomes.
• To leverage the Finney model for dose-toxicity relationship modeling and prior information incorporation.

Main Methods:

• Utilized a Bayesian adaptive design framework.
• Employed the Finney model to characterize dose-toxicity relationships in drug combinations.
• Incorporated prior dose-toxicity information from single-agent trials via prior elicitation.
• Handled unobserved delayed toxicity outcomes as missing data using Bayesian data augmentation.

Main Results:

• The proposed Bayesian adaptive design demonstrated safety in simulations.
• The design showed high probabilities of selecting the target dose combinations.
• The method effectively managed delayed toxicity under various scenarios.

Conclusions:

• The proposed Bayesian adaptive design is a safe and effective approach for dose-finding in drug combination trials with delayed toxicity.
• The Finney model and Bayesian data augmentation are suitable tools for this complex scenario.
• This design facilitates the efficient identification of optimal drug combinations.