The Rise and Fall of Hormone Replacement Therapy
[ed. One of the major issues during the modern period of CVD epidemiology has been the question about use of hormone therapy (HT) in menopause. Three experienced researchers, all women, reviewed this topic in 2005, and, with their permission, we summarize their collaboration (Barrett-Connor, Grady, and Stefanick 2005).]
The prescription of estrogens to relieve menopause-related symptoms was approved by the Food and Drug Administration (FDA) some 60 years ago. As expectations increased in the 1960s, widespread estrogen use was advocated to prevent postmenopausal estrogen deficiency and the “tragedy of the menopause.” In the 1970s, prescriptions decreased, reflecting awareness of an increase in endometrial cancer in hormone users. It subsequently increased again with evidence that the addition of a progestin could prevent estrogen-induced endometrial changes.
In the 1980s there was a renewed emphasis on potential long-term benefits as epidemiological evidence mounted for a reduced risk of osteoporotic fractures and an equally dramatic reduction in coronary heart disease. In 1992, a landmark systematic review and meta-analysis of the results of observational studies of postmenopausal HT described four outcomes plausibly related to HT: heart disease, hip fracture, breast cancer, and uterine cancer. On the basis of summary odds ratios from the observational studies and lifetable models, the authors concluded that the reduced risk of heart disease and hip fracture would outweigh the cancer risks. This favorable risk benefit ratio, based on observational evidence and apparent cardiovascular benefit, increased the recognition that coronary heart disease was the main killer of women and led to the recommendation that hormone replacement therapy be considered for all women.
In the United States, by the mid-1990s, recommendation of HT was counted a criterion for good medical practice, and estrogen became the biggest-selling prescription drug. By 2002, however, results from two large, randomized, placebo-controlled clinical trials, Heart and Estrogen/Progestin Replacement Study (HERS) and the Women’s Health Initiative estrogen plus progestin trial (WHI-EPT), both comparing conjugated equine estrogen (CEE) plus daily medroxyprogesterone acetate (MPA) with placebo, did not show the expected coronary benefit. Estrogen and combined estrogen/progestin therapy prescriptions decreased greatly. In April of 2004, the main results from the Women’s Health Initiative Estrogen Trial (WHI-ET), which compared unopposed CEE with placebo, also reported no reduction in the risk of coronary events.
WHI investigators constructed a global index for eight clinically significant disease outcomes (coronary disease events, stroke, hip fracture, pulmonary embolism, breast, colon, and endometrial cancer, and death from other causes). In the WHI-EPT
analysis, small risks exceeded the small benefits of combined therapy. In WHI-ET, the small benefits of unopposed estrogen equaled the small risks. Other trials with other regimens, routes, and doses are underway, although none is large enough to examine specifically clinical cardiovascular events or cancer.
The investigators concluded that despite estrogen’s multiple biologic effects, clear clinical trial evidence of benefit for systemic HT is presently limited to relief of menopausal vasomotor and vaginal symptoms. HT prevents bone loss and fractures but also increases the risk of other life-threatening events, resulting in no overall benefit. On the basis of WHI-EPT and WHI-ET, the risks for combined or unopposed therapy are an increased risk of stroke that does not decrease with time, an increased risk of VTE that may wane over time, and, in older women, an increased risk of dementia. None of these trials excludes the possibility that other doses, routes, or regimens, or administration at earlier ages, would have a better risk-benefit ratio for the prevention of disease.
Barrett-Connor, E., Grady, D. and Stefanick, M.L. 2005. The rise and fall of menopausal hormone therapy. Annual Review of Public Health, 26: 115–40.