Abstract P6-14-18: Updated Survival Analysis of a Phase III Study (EMBRACE) of Eribulin Mesylate Versus Treatment of Physician's Choice in Subjects with Locally Recurrent or Metastatic Breast Cancer Previously Treated with an Anthracycline and a Taxane

Abstract

Abstract Background: Eribulin mesylate (E7389) is a non-taxane microtubule dynamics inhibitor with a novel mode of action. EMBRACE was the first trial to compare overall survival (OS) of this new chemotherapeutic agent to real-life treatment choices (treatment of physician's choice; TPC) in heavily pretreated subjects with advanced breast cancer. TPC was any monotherapy (cytotoxic, hormonal, biologic) or supportive care only. It was previously reported that the study met its primary endpoint with a significant improvement in OS by a median of 2.5 months with eribulin vs. TPC. That primary analysis was based on 422 of 762 events that occurred by May 12, 2009; here we present an updated survival analysis requested by the FDA of the pivotal Phase 3 study through March 3, 2010. Methods: This report is an updated survival analysis based on 589 of 762 (77.3%) events that occurred by March 3, 2010. Data were censored for subjects who were still alive, lost to follow-up, or withdrew consent on or before the date of data cut-off (March 3, 2010). A treatment arm comparison of OS was performed (with geographic region, HER2/neu status and prior capecitabine use as stratification factors for randomization) using a stratified log-rank test. Hazard ratios (HR) and 95% confidence intervals (CI) were calculated to assess the magnitude of the treatment benefit. Results: Seven hundred sixty two women were randomized in the study; 508 to the eribulin arm and 254 to the TPC arm. 386 (76.0%) events occurred in the eribulin-treated arm while 203 (79.9%) events were recorded in the TPC-treated group. Nine (1. 8%) eribulin and 5 (2.0%) TPC patients were lost to follow-up or withdrew consent. OS was significantly longer in the eribulin treatment arm as compared with the TPC treated arm. (p=0.014; HR 0.805; 95% CI 0.677, 0.958). Median OS was 403 days (13.2 months) compared with 321 days (10.5 months) for TPC. A sensitivity analysis in the following 3 populations showed that the HR was similar to the primary (intent to treat) analysis and the treatment difference was statistically different: population of subjects treated (p = 0.016; HR 0.806; 95% CI 0.676, 0.960), at the cut-off of 572 (75%) deaths (p=0.008; HR 0.787; 95% CI 0.660, 0.939) and with no stratification terms (p=0.024; HR 0.822; 95% CI 0.694, 0.975). Based on the Kaplan-Meier analysis, eribulin-treated subjects had a 1-yr and 2-yr survival rate estimate of 54.5 and 21.9% compared with 42.8 and 19.2% for TPC, respectively. Conclusions: The OS advantage associated with eribulin treatment reported in the primary analysis was maintained throughout the observation period. The results of this OS-updated analysis support the initial OS analysis of the pivotal trial EMBRACE, demonstrating the superiority of eribulin over TPC in heavily pretreated women with advanced breast cancer. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P6-14-18.