A phase I/II seamless dose escalation/expansion with adaptive randomization scheme (SEARS).

Area of Science:

• Clinical trial design
• Pharmacometrics
• Drug development

Background:

• Traditional drug development separates Phase I (dose finding) and Phase II (dose expansion) sequentially.
• This sequential approach is administratively burdensome and time-consuming, with limited information sharing between phases.

Purpose of the Study:

• Propose a seamless dose escalation/expansion with adaptive randomization scheme (SEARS).
• Integrate Phase I dose escalation (toxicity-based) with Phase II dose expansion and comparison (efficacy-based) into a single, continuous design.
• Enable simultaneous dose escalation, graduation, and adaptive randomization for efficient trial progression.

Main Methods:

• SEARS integrates three components: modified toxicity probability interval (mTPI) for Phase I dose escalation.
• Promising doses are graduated to Phase II for adaptive randomization based on efficacy.
• Phase I dose escalation, dose graduation, and Phase II adaptive randomization proceed concurrently.

Main Results:

• SEARS demonstrated superior performance compared to two other designs.
• The SEARS design reduces sample sizes without compromising statistical power.
• An R program and demo slides for SEARS are available online.

Conclusions:

• The SEARS design assumes binary efficacy and toxicity responses are measurable within the same timeframe.
• This assumption is often met in practice through the use of surrogate efficacy markers.