Background: The role of oestrogen and progesterone receptors in breast cancer has been extensively investigated; however, the androgen receptor has received little attention. We report a flow cytometric method for the measurement of tumour androgen receptor (AR) expression on 43 primary operable breast cancers. AR expression has been correlated with grade, lymph node status, epidermal growth factor receptor (EGFR), and oestrogen receptor alpha (ERα) also measured by flow cytometry. Method: Biotinylated AR antibody AR441 (Dako) was titrated using the LNCaP cell line. Tumour samples were processed through mesh and the resulting suspension were fixed with 2 per cent paraformaldehyde and permeabilized with 0.025 per cent triton prior to incubation with AR antibody, streptavidin PE and 5D3‐FITC. Analysis was performed on cytokeratin‐positive cells by FACScan flow cytometer. Results: Eighty per cent of tumours expressed AR. There was a significant correlation between AR and ERα expression (P‐value = 0.008). No correlation was observed between AR and EGFR expression (P‐value = 0.59). Conclusions: AR appears to be coexpressed with ERα, but not EGFR. The finding of increasing AR expression with increasing grade suggests that AR expression is a more aggressive marker than has hitherto been found by immunohistochemistry. These observations are supported by the previously reported findings of AR expression in 75 per cent of breast cancer metastases, and as the sole steroid receptor in 25 per cent of metastases.

Background: The preliminary results of the IMPACT trial have shown that aromatase inhibitors are the preferred first‐line hormonal treatment for advanced breast cancer. This raises the question as to the significance of circulating testosterone levels in women with breast cancer. Method: We have measured serum testosterone and sex hormone‐binding globulin (SHBG) preoperatively in women presenting with primary operable breast cancer and age‐matched controls. Tumour androgen receptor (AR) expression was measured by flow cytometry on tumour biopsies from 35 patients, and serum testosterone levels were correlated with tumour expression. Results: No correlation was found between serum androgen levels and tumour androgen receptor expression. Conclusions: These results suggest that elevated circulating total testosterone has a role in the aetiology of the breast cancer, presumably by conversion to oestrogens. The finding of a significantly raised free androgen index in postmenopausal controls suggests that ‘active’ testosterone may be protective against the development of breast cancer. This is supported by cell line work, which has shown a growth‐inhibitory affect of androgens on breast cancer proliferation.