A stochastically curtailed two-arm randomised phase II trial design for binary outcomes

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Oncology Research

Background:

• Phase II oncology trials frequently use single-arm designs due to lower participant requirements compared to randomized controlled trials.
• There is a need for more efficient randomized trial designs in oncology to improve participant allocation and study outcomes.

Purpose of the Study:

• To introduce a novel randomized two-arm binary outcome trial design incorporating stochastic curtailment (SC) and randomized block design.
• To enhance the efficiency of randomized phase II oncology trials, allowing for early stopping.

Main Methods:

• The proposed design integrates stochastic curtailment (SC) for early trial termination and a randomized block design to control interim analyses.
• The new design was compared against existing early stopping designs using a loss function evaluating design characteristics.
• Comparisons were validated with a real-world clinical trial example.

Main Results:

• The proposed randomized design demonstrated superiority over existing early stopping designs across various loss functions.
• The design offers flexibility in the number of interim analyses, potentially increasing practicality.
• While one existing design excels with low anticipated response rates, it is sensitive to null response rate misspecification.

Conclusions:

• The novel randomized design with SC and randomized blocks is recommended for phase II oncology trials over other sequential designs.
• This approach offers a more efficient and practical alternative to traditional single-arm or less flexible randomized designs.
• The design enhances participant efficiency and provides robust decision-making capabilities through controlled interim analyses.