The development of the triphenylethylene derivatives has been a major advance in the treatment of breast carcinoma, due to their low toxicity. Currently there is controversy not only about their mode of action but in three main areas of clinical practice: (A) adjuvant therapy in premenopausal and oestrogen receptor-negative cases; (B) primary therapy in the elderly; (C) the prevention of breast cancer.

The United States' National Institute of Health consensus of 1985 that adjuvant tamoxifen be confined to postmenopausal women with positive nodes and ER positive tumours has been refuted by randomised clinical trials. The evidence that ER negative tumours also respond to tamoxifen has added to laboratory evidence that tamoxifen acts other than purely by oestrogen antagonism at the ER.

The use of tamoxifen as sole treatment in primary breast cancer in the elderly is being evaluated in randomised clinical trials. Two that have been reported so far have flaws which make their contradictory conclusions difficult to assess.

Recent results from the CRC trial suggest that tamoxifen has more benefit than simply preventing cancer deaths and clarification of the effects of triphenylethylenes on bone metabolism and the cardiovascular system may strengthen the case for a trial of prophylactic tamoxifen in those at risk of developing breast carcinoma.

The use of tamoxifen in the treatment of both metastatic and primary breast cancer is reviewed. In metastatic disease, tamoxifen slows progression in at least one-third of patients and is now the preferred first systemic therapy. Results from the larger trials of adjuvant tamoxifen are reviewed in relation to total dose and receptor status. It is suggested that variations in total dose may account for the variable results reported. Although benefit is not confined to those with tumours which are oestrogen-receptor (ER)-positive, the greatest chance of increasing disease-free survival would seem to be in those with high ER levels. Duration of tamoxifen and its effect within defined ER categories, with and without chemotherapy, are areas requiring further study by direct comparison.

Tamoxifen, a non-steroidal antioestrogen, is the first line endocrine therapy for breast cancer. Laboratory studies during the past decade have provided valuable clues to understand the mode of action of tamoxifen so that the drug can be used to its best advantage in the clinic. Animal models (carcinogen-induced rat mammary carcinoma models, spontaneous mouse mammary tumours, and athymic mice inoculated with hormone-dependent human breast cancer cell lines) all demonstrate that tamoxifen is a tumouristatic agent and long-term or indefinite therapy is required to prevent the appearance of tumours. However, when treatment is stopped, tumours appear spontaneously or can be encouraged to re-appear with oestrogen therapy. These data clearly support the view that adjuvant clinical trials with tamoxifen should employ an indefinite treatment policy or at least until the time of treatment failure. Several clinical trials organisations in Great Britain (CRC and Scottish trial) have recruited a significant number of premenopausal patients. However, unlike postmenopausal patients, long-term tamoxifen therapy of premenopausal women often causes an increase in circulating oestrogen. Since tamoxifen has been shown to be a competitive inhibitor of oestrogen action, strategies to reduce ovarian steroidogenesis (oophorectomy or Zodalex®) may provide an optimal environment to sustain the long-term effectiveness of adjuvant tamoxifen therapy.

Recently described 7α-alkylamide analogues of oestradiol, exemplified by ICI 164,384, have the pharmacological characteristics of pure antioestrogens. ICI 164,384 is completely free of oestrogenic activity in rodents and blocks the trophic actions of exogenous and endogenous oestradiol and of partial agonist antiooestrogens like tamoxifen. In addition, ICI 164,384 is peripherally selective, producing complete involution of the uterus without affecting LH secretion in intact female rats.

In MCF-7 oestrogen-responsive, human breast cancer cells ICI 164,384 is a more potent and effective inhibitor of cell growth than tamoxifen. This difference in efficacy is reflected by a greater reduction in the proportion of cells continuing DNA synthesis in ICI 164,384-treated cultures than in tamoxifen-treated cells. This observation implies that a pure antioestrogen may provide more effective therapy of breast cancer but this will have to be demonstrated in clinical studies.

The breast can be deprived of oestrogen by blockade of the synthesis of oestradiol or of its precursors. In postmenopausal women, oestrogen synthesis occurs almost exclusively in extraglandular tissues. Thus, liver, muscle and fat, as well as other tissues, convert precursor androgens such as androstenedione and testosterone of adrenal origin to oestrone and oestradiol through the enzyme, aromatase. Breast cancer tissue and its surrounding stroma also contain aromatase which could potentially convert a biologically important amount of androgen to oestrogen. Sulphatases are present in multiple tissues including breast and act to reform free oestrogens from their conjugated derivatives. Each of the steps outlined (i.e. precursor androgen formation, aromatase, and sulphatase), serve as potential sites of blockade of oestrogen synthesis. Aromatase can be inhibited by either suicide or competitive inhibitors. In our studies, an imidazole inhibitor of aromatase which acts competitively (i.e. CGS 16949A) is 1000-fold more potent than aminoglutethimide and ten-fold more potent than 4-hydroxyandrostenedione in vitro. In sixteen postmenopausal women, doses 500-fold less than standard doses of aminoglutethimide inhibited plasma and urinary oestrogens measured by radioimmunoassay or GLC mass spectrometry to the same extent as with aminoglutethimide. The use of sulphatase inhibitors, another means of oestrogen deprivation, may potentially be feasible. Sulphatase inhibitors are active in vitro but currently lack potency and specificity for the sulphatase in breast cancer tissue itself.

In premenopausal women, the oestrogens are directly aromatised from androgens in ovarian tissues through the stimulatory influence of LH and FSH. Suppression of these gonatrophins with or without additional aromatase inhibitors provides a means of inhibiting oestrogen in premenopausal women. Aromatase inhibition, alone, will probably not be effective because of the reflex rise in LH and FSH observed. Our studies indicate effective suppression of oestradiol in premenopausal women with the GnRH superagonist analogue, Leuprolide. However, combined use of this superagonist analogue plus an aromatase inhibitor would have the advantage of blocking synthesis of oestrogen, both in ovary and in extraglandular sites.

The effects of LH-RH agonists in advanced breast cancer patients are reviewed and compared to the efficacy of surgical ablation of the ovaries and X-ray induced menopause in the treatment of this disease. In both pre- and postmenopausal women, LH-RH agonists produce pituitary gland desensitisation and a fall in concentrations of LH and FSH. In premenopausal patients plasma progesterone and oestradiol levels fall to the castrate or postmenopausal range within three to four weeks. Tumour remissions have been observed in approximately 30% of premenopausal women (50% ER-positive) and approximately 10% of postmenopausal patients. The mechanism of action of LH-RH agonists is discussed and their current application to combined endocrine therapy, early breast cancer and mastalgia briefly outlined.

The results of a phase I/II study in advanced breast cancer in postmenopausal women of the aromatase inhibitors 4-hydroxyandrostenedione (40HA), miconazole and CGS16949A are described. 40HA, a steroidal compound which is an irreversible inhibitor of aromatase, has been administered to 131 patients by weekly (500 mg) or fortnightly (250 mg) i.m. injections, and daily (500 mg) by mouth. The overall response rate (complete regression (CR) and partial regression (PR)) was 33%; all three dose schedules had similar clinical efficacy. Oestradiol levels were suppressed to a mean of 37–45% of pretreatment values with oral and with weekly i.m. 40HA; suppression was marginally suboptimal with 250 mg i.m. fortnightly. 40HA was well tolerated, especially when administered by fortnightly injections; this dose schedule is to be preferred to weekly injections.

Significant oestradiol suppression and clinical responses did not occur with the imidazole antifungal, miconazole, which was administered to twenty-two patients at doses of 500–1000 mg daily. The non-steroidal aromatase inhibitor, CGS16949A, induced comparable oestradiol suppression to 40HA at doses of 0.6–4 mg daily; 0.6 mg daily appears to be suboptimal. Five of thirty-one patients treated had objective responses, and disease stabilised for at least six months in nine patients. Some suppression of aldosterone occurred with doses of 2 and 4 mg daily, but clinical toxicity of the compound was minor.

Inhibitors of the aromatase enzyme represent a class of therapeutic agents which potently inhibit oestrogen biosynthesis in vivo. This inhibition of oestrogen biosynthesis is well established as effective therapy in the treatment of oestrogen-dependent breast cancer. CGS 16949A [4-(5,6,7,8-tetrahydroimidazo-[l,5-a]pyridin-5-yl)-benzonitrile hydrochloride] is a non-steroidal imidazole derivative which is a potent competitive aromatase inhibitor in vitro. At a maximally effective concentration, it selectively inhibits aromatase and does not affect glucocorticoid production from the adrenal in vitro.

In vivo in the rat, CGS 16949A effectively reduces ovarian oestrogen content and potently inhibits an aromatase-mediated androgen-induced uterine hypertrophy. Oral treatment of adult, cyclic female rats with CGS 16949A disrupts cyclicity, inhibits ovulation, reduces uterine weight and suppresses serum oestradiol, all expected sequelae of oestrogen deprivation. At maximally effective doses, there is no evidence of adrenal hypertrophy, indicating that adrenal steroidogenesis is unaffected. In the DMBA-induced mammary carcinoma model in the rat, CGS 16949A caused almost complete regression of palpable tumours and significantly suppressed the appearance of new tumours at a maximally effective oral dose. Thus, CGS 16949A is a potent and selective inhibitor of the aromatase enzyme. In the rat, it is very efficacious in inhibiting oestrogen biosynthesis and in suppressing the growth of DMBA-induced mammary tumours.

In some ways, this splendid conference marks the end of a road and I doubt if another will be held under the same title. Not that there is not much fundamental work still to be done on the mechanism of oestrogen action and its effects on normal and diseased breast tissues, but its general role in relation to breast cancer would appear now to be clear. Oestrogens are necessary for the development of the normal human breast, for the initiation and promotion of breast cancer, and may even act as a mitogen in established disease. But their role can be regarded as largely permissive. It is not clear that excess oestrogens are in any way causative of breast cancer in humans.