Elevated pretreatment serum activin A and progression-free and overall survival in trastuzumab-treated metastatic breast cancer.

Abstract

38 Background: About half of HER2-positive metastatic breast cancer patients will respond tofirst-line trastuzumab-containing therapy, but most will progress within a year. Trastuzumab resistance remains a vexing clinical problem, and better predictive and prognostic biomarkers are needed. Activin A is a TGF-B superfamily member and regulates cell proliferation, apoptosis, differentiation, and immune response. Methods: Serum activin A was measured using ELISA (R&D Systems, Minneapolis, MN) in 60 metastatic breast cancer patients before starting first-line trastuzumab-containing therapy. Progression-free survival (PFS) and overall survival (OS) were analyzed using the Kaplan-Meier method and Cox modeling with both continuous and dichotomous (median) serum activin A analyses. Results: Pretreatment serum activin A levels averaged 2376 pg/ml, and had a median of 629 pg/ml and 25th and 75th quartile values of 406 and 1791 pg/ml, respectively. Patients who were hormone receptor negative had significantly higher activin A levels (median 1287 vs 450 pg/ml, p=0.002). Higher serum activin A was significant on a continuous basis for predicting reduced PFS to first-line trastuzumab-containing therapy (p<0.003), and for predicting shorter OS (p<0.0001). When analyzed using a dichotomous (median) cutpoint, the elevated serum activin A cohort had a significantly reduced PFS (HR 2.79, p <0. 002) (median 6.6 mo vs. 31.1 mo) and OS (HR 5.24, p <0.0001) (median 19.6 mo vs. median not reached). In multivariate analysis for PFS with other covariates (age, line of therapy, CA 15-3, and hormone receptor status), activin A was the only significant covariate (p=0.021). In multivariate analysis for OS, activin A (p=0.002) and CA 15-3 (p=0.03) remained significant as prognostic factors. Conclusions: Elevated pretreatment serum activin A predicts reduced PFS and overall survival in metastatic breast cancer patients treated with first-line trastuzumab-containing therapy. Activin A deserves further study as an adverse biomarker, and to select patients most likely to respond to activin A-targeted therapy.