We know that female reproductive hormones cause breast cancer and that, among postmenopausal women, much evidence points to estrogen as a major determinant of breast cancer risk. The benefits of tamoxifen(Drug information on tamoxifen) (Nolvadex) and oophorectomy correlate lower estrogen levels with a greater likelihood of survival. Nevertheless, debate continues regarding the safety of prescribing hormones in women after treatment for breast cancer. Despite calls for randomized trials among breast cancer survivors, data are not yet available from such clinical studies for us to use as the basis for making informed decisions. In this paper, I will review the existing evidence on the long-term use of hormones by women diagnosed with breast cancer and caution against this use. Alternative approaches to the prevention of chronic disease are recommended.
Female reproductive events have long been related to the risk of breast cancer. Early menarche and late menopause are associated with increased risk. The contribution of late menopause is substantial, and suggests that longer exposure to cyclic hormonal patterns increases the risk of developing breast cancer through the remainder of life. This increased risk is perhaps a consequence of greater accumulated DNA damage resulting from more frequent cell division stimulated by levels of circulating hormones during the menstrual cycle.
Much evidence now supports the concept of cell proliferation as the underlying process by which DNA damage (genetic error) accumulates and risk of breast cancer increases. In premenopausal women, the proliferative activity of the breast is greatest during the luteal phase of the menstrual cycle.[2,3] Progesterone(Drug information on progesterone) may therefore be the major mitogen in normal breast epithelium. After menopause, the rate of breast cell proliferation drops considerably. Pike et al estimate that the use of unopposed estrogen after menopause elevates the annual rate of breast cancer risk to 2.1% above that for women not using hormones. These authors also predict that adding progestins to the estrogen regimen further increases the rate of breast cell proliferation and, thus, the rate of breast cancer.
Role of Endogenous Hormones
Endogenous female hormones increase the risk of developing breast cancer. The positive correlation of obesity to hormone levels in postmenopausal women[6-8] reflects the biological function of peripheral fat cells, which convert androgens to estrogens(Drug information on estrogens). Compared with heavier postmenopausal women, lean women have both lower estrogen levels and a lower age-specific incidence of postmenopausal breast cancer.[9,10]
Few studies have prospectively assessed the risk of breast cancer according to blood levels of hormones. Results of studies that have been conducted have varied, perhaps because of the difficulty of measuring hormone levels in postmenopausal women. In a follow-up study that identified some 130 cases of breast cancer, Toniolo and colleagues showed a positive relationship with endogenous estradiol(Drug information on estradiol) levels. Follow-up of the prospective Study of Osteoporotic Fractures confirmed the strong relationship of breast cancer to estrone(Drug information on estrone) levels in 97 women who were not using postmenopausal hormones. Comparing women in the extreme quartiles of estrone levels, Cauley and associates observed a relative risk of 3.2 (95% confidence interval [CI], 1.4 to 7.0).
Several studies have explored bone density in relation to the risk of breast cancer. Low bone density reflects lower levels of estrogen.[14,15] The Study of Osteoporotic Fractures indicated that bone density in the wrist, hip, and spine is strongly related to breast cancer risk. Likewise, data from the Framingham Study showed bone density (assessed by uniform review of hand x-rays) to be positively related to breast cancer risk. The magnitude of the relative risk in these studies is in the range of 3 to 4 for the extreme quartiles of bone density in the general population. These data support results from previous studies of fractures[18,19] indicating that high bone mass is a risk factor for breast cancer.
Thus, bone density, as a marker of postmenopausal estrogen exposure (and perhaps of longer exposure), is directly related to the risk of breast cancer. This finding supports the biological role of estrogen in breast cancer etiology.
Role of Exogenous Hormones
Numerous epidemiologic studies have addressed the relationship between duration of hormone use and risk of breast cancer. These studies have largely underestimated the adverse effect of hormone use in general because women who take hormones are initially at lower risk of developing breast cancer than those who do not. Low estrogen levels are related to increased frequency of hot flashes and other symptoms of menopause (the traditional indication for use of estrogen). Low bone density is related to low estrogen levels (a more recent indication for the use of hormones after menopause), and statistical analysis has not completely factored in the confounding effects of age at menopause and type of menopause. Despite these limitations, a meta-analysis of the published studies shows that longer duration of use actually increases the risk of developing breast cancer.
Use of hormone replacement therapy, or exogenous hormones, also increases the risk of breast cancer. Traditionally, postmenopausal hormones have been used to relieve menopausal symptoms. More recently, their use has been advocated to reduce the risk of coronary heart disease and osteoporosis. Mounting evidence suggests that women with menopausal symptoms have lower estrogen levels at menopause than women without symptoms.[21-24] Moreover, lean women are more likely to report hot flashes at menopause than are heavier women. In addition, bone density is lower in women who use postmenopausal hormones than in those who do not. Together, this evidence indicates that women who use postmenopausal hormones are, on average, at lower risk for breast cancer than those who do not.
In addition to these biological correlates of hormone therapy, women who reach menopause at an earlier age are more likely to use postmenopausal hormones and for longer durations. Because early menopause is associated with a substantially lower risk of breast cancer, users of hormones, again, are, on average, at lower risk than nonusers of the same age. Data from the Nurses' Health Study relating use of hormones to risk of breast cancer support this view. Short-term use of postmenopausal hormones (ie, for less than 2 years) is related to a lower relative risk of breast cancer than no use at all in women of the same age with the same type (natural or surgical) of menopause. For women who previously took postmenopausal hormones for 1 to 23 months, the adjusted relative risk is 0.92 (95% CI, 0.79 to 1.08). Biologically, it is unlikely that such short-term use can result in a prolonged reduced risk of breast cancer after cessation of therapy. Rather, these women are probably already at lower risk prior to the initiation of use. A similar reduced risk in women who took hormones for short intervals has been reported by other studies.
Therefore, in epidemiologic studies comparing users with nonusers, the adverse effects of postmenopausal hormones will be underestimated because the majority of these studies do not take into account measures of bone density or menopausal symptoms as markers of hormone levels. Few studies have included a collection of blood samples for prospective assessment of the underlying levels of circulating hormones. Thus, the underlying level of hormones prior to the addition of exogenous hormones cannot be assessed or controlled for. Although typical analyses of hormones and risk of breast cancer factor in age at menopause, age is usually categorized into broad ranges, which can seriously confound the conclusions. Thus, women who use hormones for longer durations will have had earlier menopause and physiologically will be at lower risk of breast cancer than women on hormone therapy for shorter durations.
A positive correlation between the duration of hormone use and the risk of breast cancer has been reported by many investigators conducting meta-analyses of epidemiologic studies.[28-31] As noted above, a crucial clinical question relates to when the risk becomes elevated. An extensive systematic review of the evidence was conducted by Steinberg et al. Analyzing US case-control studies published from 1977 to 1991, they found a positive relationship between duration of use and the risk of breast cancer in 11 of 12 studies with community controls and in 4 of 9 studies with hospital-based controls. A rising risk with longer duration of use was reported in all four of the prospective studies. These investigators focused on the type of relationship between use of postmenopausal hormones and risk.
The authors subsequently combined the data in a formal meta-analysis and estimated that the risk of breast cancer increased by 1.8% for each year of use in case-control studies with community controls and by 3.6% per year in prospective studies. When a variable intercept was allowed for each study (with the assumption that all women were not necessarily at equal risk of developing breast cancer before they began taking estrogen), the results became somewhat stronger for the case-control studies, with a 3.7% increase in risk per year of use of postmenopausal estrogen.
Since the review by Steinberg and associates, additional studies have been published.[26,33-35] Although they do not uniformly support the relationship between duration and risk, they also do not rule it out. Like earlier studies, they all use broad ranges of age at menopause when they control for this important predictor of risk. Case-control studies have less consistently documented the relation between duration of use and breast cancer risk; thus higher levels of in-study variation or heterogeneity in results are reported by Steinberg et al from investigations with a case-control design. One reason for this could be the possiblity that hormone users are more health conscious and therefore more likely to participate in health-related studies. An overrepresentation of such hormone users among controls would negatively bias the relationship between hormones and breast cancer.
In summary, the combined epidemiologic data show a significant increase in the risk of breast cancer with longer use of postmenopausal hormones. The estimates from epidemiologic studies have, for the most part, been underestimates because of incomplete control for predictors of hormone use, age at menopause, and underlying hormone levels. Growing evidence supports a positive relationship between blood levels of estrogens and breast cancer risk, although the magnitude of this association varies among studies.