Background: Metastatic breast cancer (mBC) is common in older patients (pts) with treatment often complicated by comorbidities-resulting in increased toxicity from chemotherapy and sub-optimal therapy if treatment is inappropriately modified. The phase III EMBRACE trial showed significant improvement on overall survival (OS) for eribulin mesylate monotherapy compared to common chemotherapy agents used in the 3rd-line mBC setting. We sought to compare the benefits and toxicities of eribulin in older and younger pts. Methods: In this exploratory analysis, OS, progression-free survival (PFS), overall response rate (OR; PR+CR), clinical benefit rate (CBR; PR+CR+SD for >6 mos) and toxicity of eribulin-treated pts were compared in 4 age cohorts. Age was fitted as a categorical covariate in a cox proportional hazard model for OS and PFS, and as an interaction term for the Cochran-Mantel-Haenszel test to assess ORR and CBR. Analyses were stratified by geographic region, HER2/neu status, and prior capecitabine use. A further sensitivity analysis of OS fitted age as a continuous covariate with similar stratification to the primary analysis. Results: A total of 508 pts received eribulin in the EMBRACE trial-mean age (SD), 54.8 (10.3); range 28-85. Median OS trended inversely with age; however, this effect was non-significant when fitted as a categorical (Log Rank Test, p=0.40) or continuous covariate (p=0.10). Examination of secondary endpoints yielded a similar, non-significant trend-PFS (p=0.36), ORR (p=0.25), and CBR (p=0.96). The rate of grade 3/4 adverse events did not differ significantly across the age groups-neutropenia, 36.9%;50.3%; febrile neutropenia, 3.9%;5.0%; and neuropathy, 0.6%,4.5%. Treatment withdrawal rates did not differ between age groups. Conclusions: This analysis, while in a limited number of older pts, suggests survival outcomes with eribulin are independent of age; importantly, toxicity appears no greater in older pts. Age alone should not preclude the consideration of eribulin for older pts.

Abbreviations: REP, response evaluable population.