Practical recommendations for implementing a Bayesian adaptive phase I design during a pandemic

Area of Science:

• Clinical Trials
• Biostatistics
• Pharmacology

Background:

• Model-based dose-finding designs, like the continual reassessment method, offer superior efficacy compared to traditional 3+3 designs.
• Implementing advanced dose-finding designs requires specialized expertise and significant time investment.

Purpose of the Study:

• To present a practical framework for developing and implementing model-based dose-finding designs.
• To guide the derivation of necessary parameters and decision-making processes for these designs.
• To demonstrate the application of model-based design in a real-world clinical trial setting.

Main Methods:

• A practical approach to parameter derivation and decision-making input for model-based designs is outlined.
• A model-based dose-finding trial was designed and implemented within the AGILE platform trial for COVID-19 treatments.
• The study involved a phase I/II adaptive design for novel therapeutics.

Main Results:

• The practical delivery of the AGILE platform trial is discussed, highlighting information supporting the Safety Review Committee's decisions.
• Key components for a statistical analysis plan in model-based designs are identified.
• Challenges encountered during the study and acceptable adaptations for model-based designs are explored.

Conclusions:

• The study successfully demonstrates the design and delivery of an adaptive, model-based dose-finding trial.
• This work aims to facilitate the broader adoption of advanced dose-finding methodologies in clinical research.