Randomized phase II clinical trials.

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Cancer Research

Background:

• Phase II trials traditionally use single-arm designs with historical controls.
• Historical controls present bias due to small sample sizes, differing populations, or varied assessment methods.
• Randomized Phase II trials offer a prospective approach to mitigate bias.

Purpose of the Study:

• To propose a two-stage randomized Phase II trial design.
• To utilize exact statistical methods, specifically Fisher's exact test, for unbiased comparisons.
• To enhance the reliability of experimental therapy evaluations.

Main Methods:

• Development of a two-stage randomized Phase II trial framework.
• Application of Fisher's exact test for hypothesis testing without needing control arm response probability.
• Numerical simulations to evaluate performance.

Main Results:

• The proposed method effectively controls Type I error rates.
• High statistical power is maintained in the trial designs.
• The approach demonstrates accurate error control and robust power.

Conclusions:

• The proposed two-stage randomized Phase II trial design is a statistically sound and unbiased alternative to traditional methods.
• Fisher's exact test provides a robust foundation for these trials.
• Design optimization using Simon's concepts can further improve trial efficiency.