A straightforward meta-analysis approach for oncology phase I dose-finding studies

Area of Science:

• Clinical Pharmacology
• Biostatistics
• Oncology Drug Development

Background:

• Phase I clinical studies assess drug safety and dose-toxicity relationships.
• Prior quantitative data from similar drugs is crucial but often underutilized.
• Existing methods may not adequately handle between-study heterogeneity.

Purpose of the Study:

• To propose a novel two-stage approach for estimating maximum tolerated doses (MTDs).
• To leverage quantitative data from previous studies for enhanced MTD estimation.
• To provide a robust methodology for early-phase clinical trial dose selection.

Main Methods:

• Utilized penalized logistic regression for dose-toxicity modeling.
• Employed Bayesian random-effects meta-analysis to account for heterogeneity.
• Implemented the approach using standard R packages for accessibility.

Main Results:

• The proposed two-stage method demonstrated robustness in simulations.
• The approach effectively integrates data from multiple studies.
• Monte Carlo simulations validated the method's performance.

Conclusions:

• The developed method offers a simple and robust way to estimate MTDs.
• This approach enhances the quantitative assessment of drug safety in early trials.
• The methodology is applicable to oncology and other therapeutic areas.