Dr. Steel's suggestion that hormone therapy be started at menopause and be stopped in 5 to 10 years if new studies show a significantly increased risk for breast cancer is problematic. First, it assumes that it is ethical to prescribe hormone therapy before proving that it does more good than harm. Good data on the effect of long-term hormone therapy on breast cancer risk may be obtained from the Women's Health Initiative Clinical Trial, which is now in the planning stages, but this will take at least 15 years. Beginning hormone therapy later in life might merit further investigation. Beginning hormone therapy at age 60 or 65 and continuing until age 75 or 80 would provide it for a period of time longer than that shown to be beneficial in relation to heart disease and osteoporosis in most observational studies. It would also provide hormone therapy during the ages at which women are at highest risk for osteoporotic fractures and heart disease and would minimize the duration of therapy and possibly the risk of endometrial and breast cancer. Because most available studies were done in perimenopausal women, adequate data do not exist to allow pooled risk estimates for women who only took hormones late in life. Finally, the American College of Physicians is committed to updating the guidelines periodically.

Drs. Thacker and Booher question our recommendations for endometrial biopsy [1, 2] in women taking cyclic estrogen plus progestin therapy. Because women taking combination estrogen plus progestin therapy are expected to have some bleeding, it is difficult to decide when uterine bleeding should be evaluated with biopsy, especially because the incidence of endometrial hyperplasia and cancer in women receiving combination therapy is low. Conflicting data make it difficult to correlate bleeding pattern with endometrial abnormalities [3, 4]. The American College of Obstetricians and Gynecologists recommends that women receiving cyclic hormone therapy have an endometrial biopsy if "bleeding occurs ... before day 6 of progestin therapy" [5].

Regarding the use of cyclic estrogen plus progestin therapy, there is no good evidence that either cyclic or combined continuous estrogen plus progestin therapy has the same benefit for coronary disease or risk for breast cancer as unopposed estrogen. Both regimens, however, have been shown to prevent endometrial hyperplasia. Thus, given equally poor data on outcomes, if combination hormone therapy is used, a woman should be allowed to choose the bleeding pattern that she prefers (predictable cyclic bleeding with the cyclic regimen or unpredictable temporary bleeding with the continuous regimen). Finally, we agree that quality-of-life issues are important to women and urge physicians to discuss thoroughly the potential effects of hormone therapy.