Effect of endogenous and exogenous hormones on breast cancer: epidemiology

Epidemiological studies of risk factors for breast cancer have revealed many associations which have been extensively reviewed. The associations with menstrual and reproductive factors are of special relevance in the context of this paper. The main features of these associations are briefly summarised with the conclusion that endogenous hormones play a role in the aetiology of breast cancer. If endogenous oestrogens are important, effects on risk might also be expected from administration of oestrogens for therapeutic reasons. Oestrogens (either alone or with progestogens) have been given to women on a large scale in three particular circumstances--(a) during pregnancy to try to prevent abortion and late pregnancy toxaemia; (b) during the childbearing years to prevent pregnancy and (c) around the time of the menopause and subsequently to relieve menopausal symptoms and to try to prevent the development of osteoporosis. The available epidemiological evidence is reviewed in the paper with the following conclusions: a) Administration of hormones during pregnancy. There is reasonable evidence that silboestrol acts as a weak carcinogen with respect to the breast when given during pregnancy. As yet there is no evidence that trans-placental exposure to stilboestrol increases breast cancer risk in offspring. b) Administration of hormones to prevent pregnancy. The 1996 report of the Collaborative Group on Hormonal Factors in Breast Cancer has provided state of the art information about the risks associated with the pill. Current pill users experience a 24% increase in the risk of breast cancer while taking the preparations; after cessation the increase in risk steadily diminishes to become undetectable ten years after last use. c) Administration of hormones around the time of the menopause and subsequently. The report of the Collaborative Group which will provide a state of the art overview is awaited. The results of the review presented in this paper are as follows. i) Most of the available data come from North America. The oestrogen used has mostly been conjugated equine oestrogens while the progestogen has usually been medroxy progesterone acetate. ii) Any effect of HRT on breast cancer risk must be fairly small in terms of relative risk, but breast cancer is common in postmenopausal women; small increases in relative risk are therefore important in terms of absolute risk. iii) Although the data are variable, there is probably a modest increase in risk of breast cancer with increasing duration of HRT use. Any such risk may be concentrated in current users. iv) There is little to pick between the findings for oestrogen alone and those for combined therapy. Data concerning the latter are, however, still sparse, especially with respect to long term use. v) Any effect of HRT on breast cancer risk may be greater for carcinoma in situ, possibly due to detection bias in HRT users. This observation links with earlier work suggesting more favourably staged invasive cancers in HRT users than in non-users. vi) Further work is required on the administration of HRT to women with inactive breast cancer and on the possible adverse effects of HRT use on the interpretation of the results of mammographic screening.