FOXC1 and SOX10 in Estrogen Receptor-Low Positive/HER2-Negative Breast Cancer: Potential Biomarkers for the Basal-like Phenotype Prediction.

ABSTRACT

CONTEXT.­ Breast cancer with low (1%-10%) estrogen receptor (ER) expression (ER-low positive) constitutes a small portion of invasive breast cancers, and the treatment strategy for these tumors remains debatable. OBJECTIVE.­ To characterize the features and outcomes of ER-low positive patients, and clarify the clinical significance of FOXC1 and SOX10 expression in ER-low positive/HER2-negative tumors. DESIGN.­ Among 9082 patients diagnosed with primary invasive breast cancer, the clinicopathologic features of those with ER-low positive breast cancer were characterized. FOXC1 and SOX10 mRNA levels were analyzed in ER-low positive/HER2-negative cases from public data sets. The expression of FOXC1 and SOX10 in ER-low positive/HER2-negative tumors was evaluated by immunohistochemistry. RESULTS.­ The clinicopathologic study of ER-low positive tumors indicated more aggressive characteristics compared with those tumors with ER >10%, while they had more overlapping features with ER-negative tumors irrespective of the HER2 status. The intrinsic molecular subtype of ER-low positive cases with high FOXC1 and SOX10 mRNA expression was more likely to be nonluminal. Among the ER-low positive/HER2-negative tumors, 56.67% (51 of 90) and 36.67% (33 of 90) were positive for FOXC1 and SOX10, respectively, which was significantly positively correlated with CK5/6 expression. In addition, the survival analysis demonstrated no significant difference between patients who received and who did not receive endocrine therapy. CONCLUSIONS.­ ER-low positive breast cancers biologically overlap more with ER-negative tumors. ER-low positive/HER2-negative cases demonstrate a high rate of FOXC1 or SOX10 expression, and these cases might be better categorized as a basal-like phenotype/subtype. FOXC1 and SOX10 testing may be used for the intrinsic phenotype prediction for ER-low positive/HER2-negative patients.