Summary Immunohistochemical assays have been employed to study the expression of ER, PgR, EGFR and Ki67 immunostaining in normal breast tissue (n = 76). The expression of ER and PgR was highly variable in both pre and postmenopausal women and was characterised by large numbers of apparently negative cells. This was most evident for ER-ICA staining in tissues removed from premenopausal women. PgR levels were highest in the ducts of premenopausal women, while EGFR expression was elevated in both ducts and lobules. Ki67 expression was observed in < 10% of all normal cells and was suppressed by the menopause in lobular tissue. (Fentiman, 1989;Powles et al., 1989). The concept is largely based on the epidemiological observations that an early age of natural menopause or early oophorectomy for reasons other than breast cancer, substantially reduces the incidence of the disease (Pike et al., 1989). Currently tamoxifen is the most likely candidate for such a prophylactic regime, since the antioestrogen has not only proven effectiveness in both primary (SBCT report, 1987; Nato report, 1990) and advanced (Patterson et al., 1981;Furr & Jordan, 1984) breast cancer, but also shows a low incidence of side-effects. Moreover, in primary breast cancer patients treated with tamoxifen as an adjuvant to surgery it has now been observed that there is a reduction in the development of contralateral breast cancer, suggesting that the drug is indeed preventing the development of the disease (Nato report, 1990). Unfortunately, one of the major concerns about tamoxifen is that little is known of its effects on normal breast tissue. This is of particular concern since animal experiments have demonstrated that tamoxifen can show oestrogen-like properties (Furr & Jordan, 1984) and is capable of promoting full ductal development in the rat mammary gland (Nicholson et al., 1988 Ki67 immunostaining was performed using methods previously described (Bouzubar et al., 1989). The immunohistochemical detection of EGFR was undertaken using a previously unpublished procedure. Briefly, cryostat sections (5 Lm) were mounted on slides coated with a tissue adhesive and air dried for at least 2 h and stored at -70'C prior to assay. Sections were fixed in acetone/chloroform (1:1) at 4'C for 10 min then washed in Tris buffered saline (10 mM Tris, pH 7.4; TBS) before incubation with a normal goat serum (diluted 1:10 with TBS) for 10 min. Excess serum was removed and the slides were incubated for a further 60 min with the primary antibody (1 Lg ml final concentration, Amersham, UK) in 10% normal goat serum and 5% normal human serum in TBS). The slides were washed three times in TBS and reincubated for 30 min with rabbit anti-mouse peroxidase conjugate (diluted 1:50 in 10% normal goat serum and 5% normal human serum in TBS) followed by 2 washes in TBS. Sections were immersed for 6 min in a chromogen substrate bath containing DAB (150 mg) and imidazole (150 mg) in 300 ml TBS, to which had been added 99 il of 30% (w/v) hydrogen peroxide. The reaction was stopp...
