Designing cancer immunotherapy trials with delayed treatment effect using maximin efficiency robust statistics.

Area of Science:

• Clinical Trials
• Biostatistics
• Immunotherapy

Background:

• Immunotherapy's indirect mechanism causes delayed treatment effects, violating proportional hazards assumptions.
• Traditional log-rank tests are inefficient for immunotherapy trials with delayed effects.
• Existing methods like the maximin efficiency robust test (MERT) require estimating unknown functions from historical data.

Purpose of the Study:

• To propose a simplified and efficient statistical test for immunotherapy trials with delayed treatment effects.
• To introduce the V0 test as an approximation to the MERT, utilizing the full dataset.

Main Methods:

• The V0 test is proposed, defined as the sum of the log-rank test on the full dataset and the log-rank test on data beyond the lag time.
• Sample size formula for the V0 test was derived.
• Simulations were conducted to compare the V0 test with existing methods.

Main Results:

• The V0 test efficiently uses all trial data.
• It demonstrates higher efficiency than the log-rank test when a delay exists.
• Minimal efficiency loss occurs when no delay is present.

Conclusions:

• The V0 test is a robust and efficient statistical method for designing cancer immunotherapy trials with delayed treatment effects.
• It offers a practical alternative to more complex methods, balancing efficiency and simplicity.
• The V0 test can be effectively illustrated using real-world cancer immunotherapy trial data.