RESUMO
OBJECTIVES: Predictive factors of response to eribulin are lacking. We aimed to investigate the activity and safety of eribulin in a real-world population of metastatic breast cancer (MBC) patients and to identify possible predictive factors of progression-free survival (PFS) and objective response. METHODS: We retrospectively analyzed 71 eribulin-treated MBC patients. Best response rate, PFS, and adverse events (AEs) were evaluated. The impact of different clinical-pathological factors on PFS was evaluated using the Cox proportional hazards model. Predictive factors of response were identified by discriminant function analysis (DFA). RESULTS: Median PFS was 3.75 months (95% CI, 2.39-4.48); 12 patients (16.90%) achieved partial response (PR), 27 (38.03%) stable disease. The most common AEs were fatigue (25.83%), neutropenia (16.56%), and peripheral neuropathy (13.91%). A worse performance status (p = 0.025) and a higher number of metastatic organ sites (p = 0.011) were associated with a worse PFS under eribulin. Overall, in the DFA-predictive model, neutrophil-to-lymphocyte ratio at baseline, estrogen receptor, Ki67, histology, and age were predictive of PR with 100% accuracy. CONCLUSIONS: Activity and safety profiles of eribulin were consistent with literature data. Performance status and number of metastatic sites were predictive factors of PFS. DFA could be a promising tool to discriminate responses to eribulin among MBC patients.