Escalation with overdose control for phase I drug-combination trials

Area of Science:

• Oncology Drug Development
• Clinical Trial Design
• Biostatistics

Background:

• Dose finding for combination therapies is crucial in oncology.
• The escalation with overdose control (EWOC) method is established for single-agent trials.
• Existing methods lack specific designs for two-drug combination dose finding.

Purpose of the Study:

• To propose a novel two-dimensional EWOC design for dose finding in two-drug combination oncology trials.
• To adapt the EWOC method for simultaneous dose assessment of two agents.
• To enhance patient safety by controlling the probability of exceeding the maximum tolerated dose (MTD).

Main Methods:

• Developed a two-dimensional EWOC design based on a four-parameter logistic regression model.
• Continuously updated the posterior distribution of the MTD combination.
• Utilized a feasibility bound based on MTD distribution quantiles to control overdose risk.
• Determined dose adjustments (escalation/de-escalation) within a two-drug combination matrix.

Main Results:

• Simulation studies demonstrated the effectiveness of the two-dimensional EWOC design under various scenarios.
• The proposed design successfully identified appropriate dose combinations.
• The feasibility bound effectively reduced the likelihood of assigning doses exceeding the MTD.

Conclusions:

• The two-dimensional EWOC design is a viable and safe model-based approach for dose finding in two-drug combination oncology trials.
• This method provides a structured framework for navigating complex dose-finding landscapes in combination therapy development.
• Further application and validation in real clinical trial settings are warranted.