Escalation strategies for combination therapy Phase I trials

Area of Science:

• Clinical Pharmacology
• Biostatistics
• Drug Development

Background:

• Phase I clinical trials identify maximum tolerated doses (MTDs) for single agents.
• Combination therapy trials present challenges due to multiple potential MTDs and complex dose-escalation strategies.
• Existing Bayesian adaptive models for dose-escalation in combination therapies require further investigation.

Purpose of the Study:

• To investigate the properties of two Bayesian adaptive models for combination therapy dose-escalation.
• To assess the impact of different escalation strategies on MTD identification.
• To compare the efficiency of various dose-escalation approaches in identifying multiple MTDs.

Main Methods:

• Simulation studies were conducted to evaluate operating characteristics of Bayesian adaptive models.
• Two existing Bayesian adaptive models were applied with different escalation strategies.
• The number of identified MTDs and efficiency of each strategy were assessed.

Main Results:

• 'Non-diagonal' escalation strategies, restricting simultaneous dose increases, were found to be inefficient.
• Restricted strategies identified fewer MTDs, often escalating one agent while keeping the other fixed.
• Bayesian D-optimality designs allowed more varied dose space exploration and identified more MTDs.

Conclusions:

• For Phase I combination trials, identifying multiple MTDs for Phase II experimentation is beneficial.
• Careful consideration of the dose-escalation strategy and model is crucial for effective MTD identification.
• Bayesian D-optimality designs show promise for efficiently identifying multiple MTDs in combination therapy trials.