Integrating dose estimation into a decision-making framework for model-based drug development

Area of Science:

• Pharmacometrics
• Drug Development
• Clinical Trial Design

Background:

• Pharmaceutical go/no-go decisions rely heavily on effect size for commercial targets.
• Accurate dose and dose-response knowledge is crucial, as incorrect dosing risks clinical and commercial failure.
• Existing decision frameworks can be enhanced by integrating efficacy and dose-response estimation accuracy.

Purpose of the Study:

• To develop and illustrate a model-based decision framework for go/no-go decisions in drug development.
• To integrate both efficacy and dose-response estimation accuracy into the decision-making process.
• To characterize development risks more comprehensively by considering dose-response uncertainties.

Main Methods:

• Developed a decision framework using prespecified target and lower reference values for efficacy and dose accuracy.
• Incorporated a set of dose-response models and employed model averaging to address early-stage knowledge gaps.
• Illustrated the framework's operating characteristics through simulation studies.
• Applied the decision approach to a post hoc analysis of Phase 2 data for naloxegol.

Main Results:

• The model-based approach provides a structured method for evaluating drug development risks.
• Integration of efficacy and dose-response accuracy enhances the go/no-go decision process.
• Model averaging effectively handles dose-response uncertainties in early development.

Conclusions:

• The proposed framework offers a more complete characterization of development risk for go/no-go decisions.
• This model-informed approach supports scientific decision-making, potentially increasing efficiency and reducing costs in drug development.
• The naloxegol case study demonstrates the practical application of this enhanced decision framework.