Adaptive design for identifying maximum tolerated dose early to accelerate dose-finding trial

Area of Science:

• Clinical trial design
• Pharmacometrics
• Biostatistics

Background:

• Phase I clinical trials are crucial for determining the maximum tolerated dose (MTD).
• Early MTD identification accelerates progression to Phase II trials or expansion cohorts for efficacy confirmation.
• Traditional methods can prolong dose-finding studies.

Purpose of the Study:

• To propose and evaluate a novel adaptive design for early MTD identification.
• To accelerate dose-finding trials by enabling earlier MTD determination.
• To maintain accuracy in MTD selection while reducing study duration.

Main Methods:

• A novel adaptive design using Bayesian analysis and dose-retainment probability for early MTD identification.
• Application of the early identification design to an actual clinical trial.
• Simulation studies to evaluate the performance and efficiency of the proposed design compared to existing methods.

Main Results:

• The actual trial confirmed that MTD could be identified early, leading to a shortened study period.
• Simulation results showed comparable MTD selection accuracy for early identification designs (Keyboard and BOIN) versus non-early identification versions.
• Early identification Keyboard and BOIN designs reduced study duration by approximately 50% compared to model-assisted designs and 20% compared to time-to-event model-assisted designs.

Conclusions:

• The proposed early identification of MTD strategy effectively shortens clinical trial duration.
• This approach maintains the accuracy of MTD selection.
• Accelerated dose-finding trials facilitate faster drug development pathways.