Immunoreactivity of ductal cells with putative myoepithelial markers: A potential pitfall in breast carcinoma.

The identification of an intact layer of myoepithelial cells (MECs) located between epithelial cells and the basal lamina is useful in differentiating benign breast lesions and carcinoma in situ from invasive breast carcinoma. In the present study we used three antibodies considered to be putative markers of MECs (S100 protein, muscle-specific actin [HHF-35], and smooth muscle actin [SMA]) in 100 formalin-fixed, paraffin-embedded histologic sections of breast in an attempt to compare their value in demonstrating MECs in benign breast tissue and breast carcinomas. We concluded that for identifying MECs in benign breast tissue, SMA appears to be the most reliable, followed closely by HHF-35, but S100 is very unreliable for this purpose. In breast carcinoma, all three stains showed variable cross-reactivity with myofibroblasts, being greatest with SMA. A significant number of tumor cells in ductal carcinoma, both intraductal and invasive, stain with these markers and this "cross-reactivity" is extremely high with HHF-35. Thus, immunohistochemistry should be interpreted cautiously in differentiating benign, in situ, and invasive breast neoplasms. The "cross-reactivity" also suggests the possibility of myoepithelial differentiation and/or high actin content of breast tumor cells.