M-DODII: Bayesian dose optimization design for randomized phase II study with multiple indications.

Area of Science:

• Oncology
• Clinical Trial Design
• Biostatistics

Background:

• Oncology drug development is shifting towards novel agents like molecularly targeted therapies (MTA) and immunotherapies.
• Traditional dose optimization methods may not suit these novel therapies, necessitating new approaches.
• Current methodologies often fail to address dose optimization and indication selection concurrently in early-phase trials.

Purpose of the Study:

• To propose a novel Bayesian Dose Optimization Design for Randomized Phase II trials with Multiple Indications (M-DODII).
• To integrate Bayesian continuous monitoring and a Bayesian pick-the-winner strategy for simultaneous dose and indication selection.
• To evaluate the operating characteristics of M-DODII in simulation studies.

Main Methods:

• The M-DODII design utilizes both efficacy and toxicity endpoints.
• It employs Bayesian continuous monitoring and a Bayesian pick-the-winner approach.
• Simulation studies were conducted to assess performance against other adaptive designs.

Main Results:

• M-DODII demonstrated favorable operating characteristics with controlled selection error.
• The proposed design showed a lower probability of selecting suboptimal doses.
• M-DODII achieved a higher probability of selecting the optimal dose and reduced the total sample size.

Conclusions:

• M-DODII offers an effective solution for simultaneous dose optimization and indication selection in early-phase oncology trials.
• The design improves upon existing adaptive methods by enhancing dose selection accuracy and efficiency.
• This Bayesian approach is crucial for advancing the development of novel cancer therapies.