Beyond Composite Endpoints Analysis: Semicompeting Risks as an Underutilized Framework for Cancer Research.

Area of Science:

• Clinical Trials
• Biostatistics
• Oncology

Background:

• Composite endpoints (CEPs) are widely used in cancer research, particularly for outcomes like progression-free survival.
• However, CEP analyses present drawbacks when combining terminal events (death) with non-terminal events (progression/recurrence), creating a semicompeting risks scenario.

Purpose of the Study:

• To investigate the semicompeting risks framework as a complementary analysis strategy to address limitations of traditional CEP analyses.
• To evaluate the impact of information loss and the role of death when combining multiple endpoints in cancer clinical trials.

Main Methods:

• Comparison of the illness-death model (semicompeting risks framework) with standard CEP and univariate analyses.
• Utilized data from a Phase III randomized clinical trial (N9741) in metastatic colon cancer (1419 participants).
• Conducted a simulation study to further explore the issues related to semicompeting risks data.

Main Results:

• Ignoring semicompeting risks can lead to misleading conclusions in cancer research.
• Semicompeting risks analyses clearly delineate treatment effects on individual events.
• This framework allows for assessing joint risk and the dependence between event types.

Conclusions:

• The semicompeting risks framework is a valuable, underutilized, and feasible approach for cancer research.
• Analyzing component outcomes within this framework can supplement or replace traditional CEP analyses.
• This method enhances the understanding of treatment efficacy and patient outcomes in oncology studies.