Abstract
PURPOSE: Estrogen receptor-positive (ER(+)) breast cancer (BC) is a heterogeneous disease, and there is an ongoing debate regarding the optimal cut point for clinically relevant ER expression. We used a real-world database to assess the prognostic and predictive values of lower ER expression levels on treatment outcomes with endocrine therapy. METHODS: We used a nationwide electronic health record database. Descriptive statistics were used to evaluate the association between ER expression, tumor characteristics, and treatment patterns among patients with early-stage BC. We used Kaplan-Meier survival curves to estimate recurrence-free survival (RFS) and overall survival (OS). We assessed associations between an alternative ER expression-level cut point and clinical outcomes. RESULTS: Among 4697 patients with early-stage HER2-negative BC, 83 (2.04%) had ER(+)-low BC (ER expression, 1-9.99%) and 36 (0.88%) had ER(+)-intermediate BC (10-19.9%). ER(+)-low tumors were associated with higher tumor grade, larger size, and higher axillary tumor burden than ER(+)-high tumors (≥20% ER expression). African Americans had a higher prevalence of both triple-negative BC (TNBC) and ER(+)-low BC than ER(+)-high BC. Patients with ER(+)-low and ER(+)-intermediate tumors had survival outcomes similar to patients with TNBC and worse survival outcomes than patients with ER(+)-high tumors (P < 0.001). Tumors with <20% ER expression were associated with worse outcomes. CONCLUSION: In our cohort, patients with BCs with ER expression levels <20% had poor clinical outcomes similar to those of patients with TNBC.