Two-group time-to-event continual reassessment method using likelihood estimation.

Area of Science:

• Clinical Trial Design
• Biostatistics
• Pharmacometrics

Background:

• Patient heterogeneity, with varying maximum tolerated doses, is common in dose-finding studies.
• Failure to account for heterogeneity can lead to toxic or suboptimal dosing.
• Existing methods like separate trials or arm splitting are often infeasible due to cost and resource limitations.

Purpose of the Study:

• To extend existing dose-finding designs to effectively manage patient heterogeneity within a single trial.
• To propose a novel modification of the time-to-event continual reassessment method (TITE-CRM) for accommodating two distinct patient groups.

Main Methods:

• A modified TITE-CRM employing a two-parameter model and maximum likelihood estimation is proposed.
• The method is designed to share information between patient groups, enhancing efficiency.
• Operating characteristics were evaluated through simulations comparing the proposed design to separate single-group trials.

Main Results:

• The modified TITE-CRM demonstrated effective handling of patient heterogeneity.
• Simulations indicated the proposed method's advantages over conducting separate trials when information can be shared.
• The design maximizes the utility of existing methods within a unified trial structure.

Conclusions:

• The proposed modified TITE-CRM offers a feasible and efficient approach to dose-finding in the presence of patient heterogeneity.
• This method improves upon traditional designs by leveraging shared information across patient groups.
• It enhances patient safety and optimizes dose selection in complex clinical trial settings.