A predictive probability interim design for phase II clinical trials with continuous endpoints.

Area of Science:

• Clinical Trial Design
• Biostatistics
• Pharmacodynamics

Background:

• Molecular targeted therapies offer improved safety profiles compared to traditional cytotoxic treatments.
• Early phase trials increasingly focus on biological activity and target modulation.
• Current interim analysis strategies are insufficient for continuous outcomes in single-arm phase II trials.

Purpose of the Study:

• To develop and evaluate a novel two-stage adaptive trial design for molecular targeted therapies.
• To accommodate continuous endpoints and enable interim monitoring for futility and efficacy.
• To optimize resource allocation and decision-making in early phase drug development.

Main Methods:

• A two-stage adaptive design based on predictive probability was proposed.
• The design assumes continuous endpoints (e.g., tumor size) with a normal distribution and known variance.
• Simulations and a case study were used to assess the design's performance.

Main Results:

• The proposed design effectively allows for interim stopping due to futility or efficacy.
• Using continuous tumor size as an endpoint reduced required sample sizes.
• The design demonstrated robustness across various prior assumptions.

Conclusions:

• The developed statistical design provides a flexible and efficient framework for early-phase trials of molecular targeted therapies.
• This approach supports adaptive decision-making by incorporating interim analyses for continuous outcomes.
• The methodology facilitates prioritizing promising targeted therapies based on robust early-phase data.