A group sequential design and sample size estimation for an immunotherapy trial with a delayed treatment effect.

Area of Science:

• Clinical Trials
• Biostatistics
• Immunotherapy

Background:

• Delayed treatment effects are common in immunotherapy trials, violating proportional hazards assumptions of standard statistical tests.
• Traditional log-rank test-based designs may lack statistical power and efficiency when dealing with delayed treatment effects.

Purpose of the Study:

• To propose a novel group sequential design for immunotherapy trials that accounts for delayed treatment effects.
• To enhance statistical power and optimize sample size determination in the presence of delayed treatment effects.

Main Methods:

• A group sequential design utilizing a piecewise weighted log-rank test for early efficacy termination.
• Defining information time based on events occurring after a specified delay time.
• Employing a one-dimensional search algorithm with analytical estimation and simulation-based power functions for sample size determination.

Main Results:

• The proposed design effectively handles delayed treatment effects, improving statistical power.
• The search algorithm accurately determines the maximum sample size, demonstrating advantages in sample size savings.
• Simulation results validated the accuracy of the sample size determination and the design's efficiency.

Conclusions:

• The developed group sequential design is a powerful and efficient tool for immunotherapy trials with delayed treatment effects.
• This approach offers a statistically sound alternative to traditional methods, leading to optimized sample size and improved trial outcomes.