Abstract
Previous work in a small, unselected series showed that up to 83% of breast carcinomas overexpress ECM1 by immunohistochemistry (IHC) and that tumors with lymph node metastases are more likely to be ECM1-positive. We sought to further evaluate ECM1 expression and its effect on prognosis in an unselected cohort of patients with breast cancer.
ECM1 expression was examined by IHC in 134 women diagnosed with invasive breast cancer between 1986 and 1989 and correlated with clinical parameters and outcomes, including disease-free survival (DFS), disease-specific survival (DSS), and overall survival (OS) using Cox proportional hazards regression.
During follow-up, 83 of 134 (66%) patients died. The median follow-up was 211 (range, 183-245) months for surviving patients. Based on a previously described cutoff of 10% staining, 47% of breast cancers were ECM1-positive. ECM1-positive tumors were associated with increasing patient age (P = 0.01). In multivariate analyses, while controlling for age, ER status, tumor grade, stage, and treatment, ECM1 expression emerged as a significant predictor of DSS (hazard ratios, 4.16 (P = 0.009) and 11.6 (P = 0.01) at 10 and 15 years, respectively) and DFS (hazard ratio, 3.08 (P = 0.03) at 15 years) with ECM1 overexpression predicting poorer survival.
ECM1 was overexpressed in approximately half of invasive breast carcinomas and is an important prognostic marker, particularly for predicting poorer DSS, with its predictive value increasing with time from diagnosis. Further work is needed to confirm these findings and determine whether ECM1 expression is predictive of response to specific therapy.