The Women's Health Initiative (WHI) was designed to evaluate the benefits and risks of strategies that potentially could reduce the incidence of heart disease, breast and colorectal cancers, and fractures in postmenopausal women. One of its trials examined the effects of hormone replacement therapy in women with an intact uterus.
Postmenopausal women between 50 and 79 years of age were included if they had an intact uterus; only 7.7 percent of women in the study had a history of cardiovascular disease; and in women who had been taking postmenopausal hormones, a three-month washout period was required. Women were randomized to receive either placebo or combined estrogen and progestin. Patients were contacted by telephone at six weeks and every six months thereafter, and annual clinic visits were required. Dosages could be adjusted somewhat for symptoms such as breast tenderness or vaginal bleeding. Women who developed breast cancer, endometrial pathology, deep venous thrombosis, pulmonary embolism, malignant melanoma, meningioma, or hypertriglyceridemia were excluded, as were women whose doctor prescribed estrogen, testosterone, or selective estrogen-receptor modulators for them. The main outcome measured was coronary heart disease (CHD). The main adverse effect tracked was invasive breast cancer. Other outcomes, such as osteoporotic fracture, also were followed.
There were 16,608 participants (8,506 who received study medication and 8,102 who received placebo). The study began in 1997 and was prematurely halted in May 2002. Initially, an early adverse effect was seen in the cardiovascular outcomes, but none of the preestablished boundaries for ceasing the study was crossed. By spring of 2001, all participants were notified that there were increases in myocardial infarction (MI), stroke, pulmonary embolism, and venous thromboembolism, but that it was not clear whether the risks outweighed the benefits. By May 2002, it had become clear that the incidence of breast cancer exceeded designated boundaries. A global index showed that overall harm was occurring in women taking the treatment medication. At that time, cardiovascular diseases were persisting, although not at a significant level.
Average follow-up was 5.2 years. At baseline, the placebo and treatment groups were not significantly different. However, although overall CHD rates were not high, the rate in women taking the study medication increased by 29 percent (mainly as nonfatal MI), a trend that, had the study continued, ultimately would have led to a poor result for the main outcome. There was a 41 percent increase in stroke in women in the hormone replacement group. Rates of venous thromboembolism were twice as high in the hormone replacement group. There also was a 26 percent increase in invasive breast cancer in the hormone replacement group. On the beneficial side, colorectal cancer rates decreased by 37 percent (starting at year 3), and osteoporotic fractures were significantly decreased, a benefit that increased over time.
The Women's Health Initiative Investigators conclude that use of estrogen and progestin in postmenopausal women with intact uteruses does not provide benefit, other than improvement in colorectal cancer rates and some osteoporosis rates. Specifically, among 10,000 women who take estrogen and progestin for one year, there would be seven more CHD events, eight more strokes, eight more pulmonary embolisms, and eight more invasive breast cancers compared with the control group. There also would be six fewer colorectal cancers and five fewer hip fractures.
Investigators conclude that combined conjugated equine estrogen and medroxyprogesterone acetate should not be given to women for primary prevention of CHD. If a woman currently is taking this combination for prevention of cardiovascular events, strong consideration should be given to discontinuing the treatment.
In an accompanying editorial, Fletcher and Colditz review the strengths of the WHI study. They underscore the absolute risks that can be discussed with patients (described in the paragraph above) and recommend that physicians stop prescribing this combination for long-term use.