Optimal designs for active controlled dose-finding trials with efficacy-toxicity outcomes.

Area of Science:

• Biostatistics
• Clinical Trial Design
• Pharmacometrics

Background:

• Estimating drug efficacy and toxicity is crucial in clinical trials.
• Dose-finding trials require efficient designs to balance patient safety and therapeutic benefit.
• Bivariate continuous outcomes in dose-finding studies present unique analytical challenges.

Purpose of the Study:

• To derive optimal designs for active controlled dose-finding trials with bivariate continuous outcomes.
• To establish upper bounds on the number of doses and identify conditions for including boundary points in optimal designs.
• To analytically describe minimally supported optimal designs and their independence from outcome correlation.

Main Methods:

• Application of nonlinear regression models to describe bivariate continuous outcomes.
• Derivation of optimal design criteria for efficacy and toxicity estimation.
• Analysis of design space boundaries and correlation effects.

Main Results:

• Optimal designs were derived for estimating efficacy and toxicity in active controlled dose-finding trials.
• Upper bounds for the number of doses were determined, with conditions for boundary point inclusion.
• Minimally supported optimal designs were analytically described and found to be independent of outcome correlation.

Conclusions:

• The derived optimal designs enhance the efficiency of dose-finding trials with bivariate outcomes.
• Understanding dose-response relationships and correlation is key for robust trial design.
• The findings provide a theoretical framework for designing trials that efficiently estimate efficacy and toxicity.