Abstract P5-18-26: Confirmatory overall survival (OS) analysis of CLEOPATRA: a randomized, double-blind, placebo-controlled Phase III study with pertuzumab (P), trastuzumab (T), and docetaxel (D) in patients (pts) with HER2-positive first-line (1L) metastatic breast cancer (MBC)

Abstract

Abstract Background: In CLEOPATRA, 808 pts with HER2-positive 1L MBC were randomized to treatment with placebo (Pla)+T+D or P+T+D. The primary endpoint of independently reviewed progression-free survival was significantly improved with P+T+D vs Pla+T+D (HR = 0.62; P<0.0001; medians, 18.5 vs 12.4 mths) (Baselga et al. NEJM 2012). At the time of our initial report, 43% of OS events planned for the final analysis had occurred; therefore, we have now conducted a second interim OS analysis after longer follow-up. Methods: This interim OS analysis was performed applying the Lan-DeMets α-spending function with the O'Brien-Fleming (OBF) stopping boundary to maintain the overall Type I error at 5%. Based on the number of OS events observed, the OBF boundary for statistical significance at this analysis was P≤0.0138. The log-rank test, stratified by prior treatment status and geographic region, was used to compare OS between arms in the intention-to-treat population. The Kaplan-Meier approach was used to estimate the median OS in both arms; a stratified Cox proportional hazard model was used to estimate HR and 95% CIs. Subgroup analyses of OS were performed for the stratification factors and other key baseline characteristics. Results: At the time of this analysis, median follow-up was 30 mths and 267 deaths (69% of planned events for the final analysis) had occurred. The results showed a statistically significant improvement in OS in favor of P+T+D (HR = 0.66; 95% CI, 0.52–0.84; P = 0.0008). This HR represents a 34% reduction in the risk of death. The analysis achieved statistical significance and is therefore considered the confirmatory OS analysis. The median OS was 37.6 mths in the Pla arm and has not yet been reached in the P arm. The treatment effect was generally consistent in predefined subgroups based on baseline variables and stratification factors, including: prior (neo)adjuvant therapy (HR = 0.66; 95% CI, 0.46–0.94); no prior (neo)adjuvant therapy (HR = 0.66; 95% CI, 0.47–0.93); prior (neo)adjuvant T (HR = 0.68; 95% CI, 0.30–1.55); hormone receptor-negative disease (HR = 0.57; 95% CI, 0.41–0.79); and hormone receptor-positive disease (HR = 0.73; 95% CI, 0.50–1.06). Kaplan-Meier estimates of OS rates show survival benefit with P+T+D at 1, 2, and 3 yrs. The majority of pts received anti-cancer therapy after discontinuation of study treatment (64% Pla arm, 56% P arm). Subsequent therapy with HER2-directed agents (T, lapatinib, T emtansine) was balanced between arms. Causes of death remained unchanged from the first interim OS analysis, with the most common cause being progressive disease. Adverse events leading to death were rare and balanced between arms. Conclusions: Treatment of pts with HER2-positive 1L MBC with P+T+D compared with Pla+T+D was associated with an improvement in OS, which was both statistically significant and clinically meaningful. These results show that combined HER2 blockade and chemotherapy using the P+T+D regimen can be considered a standard of care for pts with HER2-positive MBC in the 1L setting. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P5-18-26.