Two-stage screened selection designs for randomized phase II trials with time-to-event endpoints

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Oncology Research

Background:

• Phase II multiarm studies are valuable for exploring new treatments, especially in areas of unmet need like pediatric cancer.
• Existing screened selection designs (SSD) are limited to binary endpoints, necessitating new methods for time-to-event data.
• Faster patient recruitment and more treatment options are advantages of multiarm over traditional two-arm trials.

Purpose of the Study:

• To propose and evaluate novel two-stage screened selection designs (SSD) for randomized phase II trials with time-to-event endpoints.
• To provide methods that ensure a high probability of selecting superior treatment arms.
• To maintain control over the type I error rate when comparing treatments against a common external control.

Main Methods:

• Development of two-stage SSD and its variant for time-to-event data in phase II randomized trials.
• Derivation of sample size calculations tailored for the proposed designs.
• Validation through simulation studies to assess operating characteristics.

Main Results:

• The proposed SSD methods demonstrate a high probability of identifying superior treatment arms.
• Effective control of the type I error rate was confirmed in simulations.
• The design was successfully applied to three real-world clinical trials.

Conclusions:

• The novel two-stage SSD offers an efficient approach for phase II randomized trials with time-to-event outcomes.
• This methodology enhances the selection of promising treatments in areas with unmet needs.
• An accessible R package, frequentistSSD, has been developed for practical application.