A new statistical method for dose-finding based on efficacy and toxicity in early phase clinical trials

Area of Science:

• Clinical Trials
• Biostatistics
• Pharmacology

Background:

• Traditional Phase I trials focus on toxicity, while Phase II trials assess efficacy, creating a gap in early-stage drug development.
• Existing methods often separate dose-finding from efficacy evaluation, leading to suboptimal treatment selection.

Purpose of the Study:

• To present a novel statistical strategy for dose-finding in single-arm clinical trials that integrates both patient response (efficacy) and toxicity.
• To describe a phase I/II hybrid approach that adaptively selects the optimal dose based on predefined safety and efficacy standards.

Main Methods:

• The strategy employs a mathematical model where response and toxicity probabilities are interdependent functions of dose.
• It utilizes fixed standards for minimum response and maximum toxicity to guide adaptive dose selection for successive patient cohorts.
• The method allows for early stopping if no dose appears both safe and efficacious.

Main Results:

• The described method successfully selects a best acceptable dose for future patients.
• It provides estimates for response and toxicity probabilities at the chosen dose.
• The application to a trial for chemo-refractory AML/MDS patients demonstrates its practical utility.

Conclusions:

• This hybrid phase I/II strategy offers a more comprehensive approach to dose-finding by balancing efficacy and toxicity.
• The adaptive methodology enhances clinical trial efficiency and patient safety.
• The approach is flexible and applicable to various therapeutic areas, including rare response rate targets.