Hormone replacement therapy and ovarian cancer risk – any news?
T. Riman
DOI: https://doi.org/10.1080/00016340701505341
2007-09-01
Acta Obstetricia Et Gynecologica Scandinavica
Abstract:The publication of data from the Women’s Health Initiative (WHI) randomised controlled trial, where increased risks of breast cancer and cardiovascular events appeared among users of hormone replacement therapy (HRT), set off a change in attitude towards HRT use among both practitioners and health care consumers (1). It is estimated that HRT prescriptions have declined by 40 60% since the first WHI report in 2002 (2,3). The clinical question, to initiate or continue HRT in the management of menopausal symptoms, necessitates a careful risk benefit assessment, which, to date, has only marginally incorporated the risk of ovarian cancer among users of HRT. Two large, well conducted, cohort studies recently provided new information on ovarian cancer risks among HRT users (4,5), which may be added to the decision-making process whether to prescribe HRT or not. Further, particularly the US cohort study examines the effects of estrogen and progestin components of HRT in relation to ovarian cancer risk, a research area that has been incompletely evaluated (6). Between 1996 and 2001, Beral et al. in the UK Million Women Study recruited a cohort of 948,576 postmenopausal women, half of whom were current (30%) or past (20%) users of HRT (4). The results pertaining to HRT and the risk of ovarian cancer were published in the Lancet Online of 19 April. After an average follow-up of 5.3 years for incident disease, and 6.7 years for fatal ovarian cancer, 2,273 incident ovarian cancers were detected, and 1,591 deaths were attributed to the malignancy. Current users of HRT were more likely to develop ovarian cancer, with a relative risk of 1.20 (95% CI: 1.09 1.32) compared with HRT never users, and the incidence of ovarian cancer increased with increasing duration of use. A relative risk of 1.31 (95% CI: 1.12 1.53) was reported among those who had used HRT for ]10 years. Compared with never users of any HRT, current users of unopposed estrogens and estrogen-progestin combinations, who were on the medications for ]5 years, had relative risks for incident ovarian cancer of 1.53 (95% CI: 1.27 1.84) and 1.17 (95% CI: 1.02 1.34), respectively. The relative risks of developing ovarian cancer among current users of estrogen-progestin combinations were 1.14 (95% CI: 0.98 1.32) for sequential regimens, and 1.13 (95% CI: 0.95 1.33) for continuous regimens, when compared with HRT never users. No duration data were presented for the sequential and continuous estrogen-progestin combinations. Over time, women may use different HRT regimens, and the authors did not specify if the analyses were restricted to groups of women who had only used one regimen or if any statistical adjustments had been made to allow for use of 2 or more consecutive types of HRT regimens. No residual risk of incident ovarian cancer was observed among past HRT users (RR: 0.98; 95% CI: 0.88 1.11). In the same study, current users of HRT had a higher relative risk for fatal ovarian cancer of 1.23 (95% CI: 1.09 1.38) than never users, a risk that did not persist among past users (RR: 0.97; 95% CI: 0.84 1.11). Standardised incidence and mortality rates for ovarian cancer per 1,000 women in the study population were 2.2 and 1.3 among HRT never users, respectively, and 2.6 and 1.6 in current HRT users, respectively. The authors concluded that if the differences in rates between never users and current users are due to HRT, the study results imply that over a 5-year period, use of HRT resulted in