Answers are grounded in this paper and cite their sources.
Dr. Kuller is correct that an appropriate comparison of the effects of estrogen replacement therapy (ERT) versus estrogen/progestin hormone replacement therapy (combined hormone replacement therapy [CHRT]) requires comparability of women in terms of their baseline breast cancer risks. Adjustment for age at menopause and for type of menopause, as done in our report (1), may not be sufficient to make such a comparison, if the vast majority of women using one type of therapy (e.g., ERT) have a very different risk factor profile (e.g., have all undergone artificial menopause) from women using the alternative therapy. However, in our study, such adjustments were sufficient. Even when we restrict our analyses to the majority subgroup of women, those with natural menopause, the difference in effect of ERT and CHRT remains virtually unchanged: the odds ratio per 5 years of use of ERT was 1.08 versus 1.22 for CHRT. Adjustment factors were the same as for Table 2 in our original report (1). These results are very comparable to the 1.06 (for ERT) and 1.24 (for CHRT) figures cited in our report for the entire study population. We had insufficient women on long-term CHRT who had undergone hysterectomy and bilateral oophorectomy to adequately compare ERT versus CHRT in this subset of women. The studies by Magnusson et al. (2) and Schairer et al. (3) confirm the need as raised in the “Discussion” section of our report to establish an adequate delivery system for progestins. The sole goal of progestin administration postmenopausally is to protect the endometrium from carcinogenic transformation. The ideal progestin delivery system will provide the necessary dose to the endometrium to provide such protection, while concurrently minimizing or eliminating exposure to other organs, such as the breast, liver, and cardiovascular system negatively affected by progestins. We believe that the results of our report combined with those reported by Magnusson et al. and Schairer et al. have changed the standard by which current and subsequent generation progestins and the methods and regimens for their delivery should be evaluated. The burden of proof should no longer be on epidemiologists and other investigators to demonstrate that such agents increase the risk of breast cancer, rather, it should shift to the proponents of systemic use of such agents to demonstrate that they do not. Although we thank Dr. Kuller for his helpful insights regarding our analysis, we strongly disagree with his opinion that we need results from the Women's Health Initiative to determine which type of hormone replacement causes more breast cancer, ERT or CHRT, which is incidentally, a curious twist on the usual goals for a prevention trial. In fact, the Women's Health Initiative suffers from the precise limitation for which Dr. Kuller criticizes our report; women assigned to ERT and CHRT do not have similar baseline risk of breast cancer, since ERT is assigned only to women who have had hysterectomies and CHRT is assigned exclusively to women who have not. Not only do the epidemiologic data now provide a convincing case that CHRT is substantially more harmful than ERT, highly convincing supportive studies of CHRT versus ERT on cell proliferation indices also exist both in women (4) and in nonhuman primates (5), in addition to the study of mammographic densities cited in our report (6). We believe that it would represent a serious error in judgment with regard to the public health to ignore the clear message from these combined studies.
No takes yet. Share an insight, caveat, or question.
Ross et al. (2000) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: