Confirmatory adaptive group sequential designs for single-arm phase II studies with multiple time-to-event endpoints

Area of Science:

• Biostatistics
• Clinical Trial Design
• Survival Analysis

Background:

• Current methods for long-term survival in one-armed trials often focus on a single endpoint.
• Existing adaptive designs have limitations in utilizing information from multiple endpoints for interim modifications.

Purpose of the Study:

• To develop adaptive group sequential one-sample tests for multivariate survival distributions using multi-state models.
• To enable data-dependent design modifications based on all time-to-event endpoints.

Main Methods:

• The methodology is applied to tests for joint distributions of progression-free survival (PFS) and overall survival (OS) within an illness-death model.
• It also addresses time to toxicity and time to progression, considering death as a competing event.
• Large sample distributions are derived using a counting process approach, and small sample properties are assessed via simulation.

Main Results:

• The proposed adaptive tests allow for interim design modifications based on all available time-to-event data.
• The approach is illustrated using a multi-state model for non-small cell lung cancer.

Conclusions:

• The developed adaptive procedures enhance the assessment of long-term survival outcomes in one-armed trials by integrating information from multiple endpoints.
• This methodology offers a more robust and flexible framework for clinical trial design and analysis.