Real-world 1-year survival analysis of patients with metastatic breast cancer with liver or lung metastasis treated with eribulin, gemcitabine or capecitabine

Abstract

Background Pooled analyses from 2 phase III trials of patients (pts) with metastatic breast cancer (mBC) treated with eribulin (E) versus other chemotherapy showed improved overall survival (OS) favoring E. Subsequent subgroup analyses, including pts with visceral metastases (VM), also showed improved OS when treated with E. Current 3rd line mBC chemotherapy options include E, gemcitabine (G) and capecitabine (C), among others. This study evaluated 1-year (yr) OS in a real world setting among mBC pts with VM treated with E, G, or C in a US cancer center network. Methods This retrospective study investigated pts with mBC with biopsy proven VM (liver and/or lung), treated with E, G, or C in 3rd line chemotherapy from 1/1/2012 to 11/1/2018. Electronic health records and cancer registry data were used to detect differences in OS (number of mons from start of 3rd line until death). Univariate and multivariate analyses of 1-yr survival between E, G, and C were conducted using Kaplan Meier (KM) and Cox proportional hazard (CPH) models respectively. Results Of 1,828 mBC pts, 1,417 had VM, and 443 (27.5% triple negative, 13.7% HER2+ and 51.9% ER+/PR+) received 3rd line therapy with E=229 (51.7%), G=134 (30.2%) or C=80(18.1%). On average, E, G, and C pts did not differ by age, ethnicity, BMI or ECOG score (categorized as good [0-1] and poor [>2]), but E pts had higher mean income compared to G or C. 1-yr KM survival analyses included 370 pts (n=195 E, 117G, and 58 C) with at least 1-yr follow up. 150 patients (45% E, 33% G, and 41% C) were alive at the end of 1-yr. Survival curves between E, G, and C differed significantly (Logrank p<0.05; Wilcoxon p<0.05). Censored median survival of E pts was 9.75 mons compared to 6.95 for G (p<0.05) and 7.48 for C pts, respectively. CPH analysis adjusted for age, ethnicity, ECOG score, BMI, and adverse events showed significant survival benefit for E pts compared to G (HR=0.71, p<0.05), but no significant differences between E vs C or C vs G pts. Conclusions This real-world 1-yr survival analysis of pts with mBC and VM demonstrated significantly better OS for pts receiving E as compared to G, and a trend favoring OS for patients receiving E in comparison to C. Legal entity responsible for the study Eisai Inc. Funding Eisai Inc. Disclosure S.M. Kazmi: Speaker Bureau / Expert testimony: Eisai, Takeda, Lilly, Merck; Advisory / Consultancy: Merck, Takeda. E. Wang: Non-remunerated activity/ies, Eisai employee at the time of research: Eisai. R.S. Hauser: Research grant / Funding (institution): Eisai. P.A. Kaufman: Advisory / Consultancy, Research grant / Funding (institution), research grant support, and consulting fees: Eisai. All other authors have declared no conflicts of interest.