Statistical Design of the CAPItello-291 Final Survival Analysis

Erica L. Mayer, MD, MPH, of Dana-Farber Cancer Institute, explains the statistical framework behind the final overall survival analysis of CAPItello-291, the phase 3 trial of capivasertib plus fulvestrant in hormone receptor positive, human epidermal growth factor receptor 2 (HER2)-negative advanced breast cancer. The trial used a hierarchical design with alpha split across dual primary progression-free survival endpoints, both of which were met, before overall survival could be tested first in the PIK3CA/AKT1/PTEN-altered population and then overall. Based on accrual and the number of survival events, the study carried only about 53% power to detect an overall survival difference in the altered population versus 90% power in the overall population, context Dr Mayer says is essential before interpreting the survival curves. She also reviews baseline characteristics, including primary endocrine resistance in about 40% of patients, mostly second-line treatment, prior cyclin-dependent kinase 4/6 inhibitor exposure in about 70%, and well-balanced arms.