Abstract
Clinical use of hormone therapy in postmenopausal women has drastically changed over the past 5 years. The latest data relate to breast cancer risk among users of estradiol, a formulation more commonly in use in Europe than the USA [1]. Prior to publication of the Women's Health Initiative (WHI) results in 2002 [2], hormone therapy was used to prevent and treat the chronic conditions associated with menopause. In addition to short-term vasomotor symptoms, hormone therapy was used in the treatment of osteoporosis and prevention of coronary artery disease, dementia and colon cancer. A series of studies from the WHI determined the risks and benefits of estrogen, with or without progestin, on healthy postmenopausal women. The study of combined hormone therapy found overall harm, with excess cardiovascular deaths, breast cancer and thromboembolic events [2,3]. In the year immediately following the publication of the WHI data, annual hormone therapy prescriptions in the USA decreased by over 50% [4]. The US FDA and Preventive Services Task Force have since issued recommendations against the routine use of combined hormonal therapy for these chronic conditions [5].
More recently, women and their physicians have turned to other treatment options for the sequelae of menopause. Bisphosphonates with calcium and vitamin D are the primary treatment for osteoporosis. Aggressive risk-factor modification, statins and other cardiac medications are the mainstays of primary and secondary cardiovascular disease prevention. Alternative treatments for vasomotor symptoms, however, are less obvious. At present, a unified approach to women with menopausal symptoms is absent.
Approximately 80–90% of all women going through menopause will experience disruptive symptoms and 20–40% will seek medical attention as a result [6]. The most bothersome symptoms are hot flashes, night sweats, vaginal dryness and sleep disturbance [7]. Hot flashes improve within a few months in approximately 30–50% of cases and resolve within 4–5 years in 85–90% of cases [8,9]. However, 10–15% of women continue to have hot flashes many years after menopause [8]. Estrogen replacement is effective in the treatment of these symptoms, with significant improvements in symptom frequency and severity [10]. Since publication of the WHI, the number of studies of nonhormonal treatments has increased [11]. However, many of these trials have been small, of poor quality, and of short duration. Convincing evidence that behavioral, herbal or natural remedies significantly improve hot flashes is lacking [7]. Results of multiple trials looking at the widely used herb black cohosh and soy have been conflicting [12]. One recent, well-designed, randomized, controlled trial of black cohosh and dietary soy again failed to show a reduction in menopausal symptoms [13]. At present, few safe and effective options are available for the treatment of debilitating vasomotor symptoms of menopause.
Given the lack of alternatives, there has been growing interest in the careful re-examination of the evidence of harm associated with hormone therapy. The possibility exists of finding the right dose, preparation and patient subgroups for which risks would be acceptable. In women for whom endometrial cancer is not a concern because of previous hysterectomy, the progestin component of hormone therapy is unnecessary, and estrogen-alone (unopposed) preparations have traditionally been used. There is abundant evidence that the risks of cardiovascular disease and breast cancer are different for combined estrogen–progestin versus unopposed estrogen alone. The increased incidence of coronary heart disease and breast cancer events was not seen in the estrogen-only arm of the WHI [14,15]. In this group, the relative risk of breast cancer was found to be 0.77 (95% confidence interval [CI]: 0.59–1.01) with almost 7 years of follow-up, although this finding has been contradicted in numerous other studies [16]. Given the controversies, the US Preventive Service Task Force concluded in 2005 that it could not assess the effects of unopposed preparations [5]. Therefore, in this group of postmenopausal women, it has been suggested that short-duration use of estrogen for vasomotor symptoms is safe [17–19]. With long-term use, breast cancer remains the primary concern.
Results
In an observational study published in the December issue of Obstetrics & Gynecology, Lyytinen and colleagues investigated the effect of estrogen-only therapy on the risk of breast cancer in Finnish postmenopausal women [1]. A total of 110,980 women aged 50 years or older who used estradiol- or estriol-based hormone therapy for at least 6 months were identified from the Finnish national registry from 1994–2001. Hormone use was classified into oral or transdermal estradiol, oral estriol or vaginal estrogens. Women were followed to the end of 2002 or death for the end point of breast cancer development.
At the end of the study period, 2171 cases of breast cancer were identified among the study population. The risk of breast cancer was not increased with systemic or transdermal estradiol use of less than 5 years, with nonsignificant standardized incidence ratios (SIRs) of 0.93 for estradiol, 0.98 for estriol and 0.67 for vaginal estrogens. The risk only increased significantly for hormone use of greater than 5 years: the SIR was approximately 1.4 for all three preparations.
The study also examined whether average daily estradiol dose affects the risk of breast cancer. For users of greater than 5 years duration, the highest daily dose (oral dose >1.9 mg/day) was associated with an increased risk (SIR: 1.49; 95% CI: 1.25–1.75). No significant trend towards a dose response was seen for low-, medium-, and high-dose users for greater than or less than 5 years. For transdermal estradiol users, no elevated risk was found for use of less than 5 years. However, use for 5 years or more was associated with nonsignificantly elevated risk. The SIR was 1.6 for low dose, 1.32 for medium dose and 1.44 for high dose. Again, no significant trend for a dose response was observed.
Significance
This study adds to the body of literature on the risk of estrogen-only HRT and the risk of breast cancer. It is unique in providing evidence specifically for estradiol-based estrogen in a relatively homogenous population. The results are most applicable to postmenopausal women in many European countries where estradiol-based estrogen preparations are more commonly used. Since the only estrogen studied in randomized, controlled trials is conjugated equine estrogen (CEE), one must rely on observational studies such as this one to provide information on other estrogen preparations for women with a history of hysterectomy.
The findings by Lyytinen and colleagues are consistent with past epidemiological data, in that significantly increased risk of breast cancer is only seen after 5 years of estrogen use. The SIR from this study was 1.34 (95% CI: 1.16–1.54) for use of 5–10 years of estradiol, resulting in two to three cases of breast cancer per 1000 women over 10 years of follow up. The UK Million Women Study found that risks of current estrogen use were significant: odds ratio 1.05 (95% CI: 0.69–1.59) and odds ratio 1.34 (95% CI: 1.24–1.44) for users of less than and greater than 5 years, respectively [20]. Our own meta-analysis of observational studies found that the summary risk of current use was 1.16 (95% CI: 1.02–1.32) for less than 5 years use and 1.20 (95% CI: 1.06–1.37) for more than 5 years use [16].
The major limitations of this study are inherent to the observational design. Given the multiple factors that affect the risk of breast cancer, the results can potentially be biased. The registry looked at all postmenopausal women aged over 50 years. Only approximately 50% of those enrolled were aged 50–59 years, and their vasomotor symptomatology was unknown. Confounding by indication – wherein a woman takes a particular medication because of her perceived risks, cannot be estimated or controlled for.
This study also attempts to quantify the influence of daily estrogen dose and the route of administration on breast cancer risk. Past studies did not show different risks associated with transdermal or vaginal estrogens [20,21]. The UK Million Women Study and a collaborative reanalysis did not find that lower daily estrogen dosage conferred a lower risk of breast cancer [20,22].
What does this study mean for women seeking safe and effective treatments for vasomotor symptoms of menopause? Data support that oral and transdermal estrogens have similar efficacy in treating hot flashes [9]. Symptom relief has been demonstrated with low doses of estrogen [23,24]. Compared with higher doses whose effects are reached by 4 weeks, relief with lower-dose estrogens is not maximal until 8–12 weeks. However, lower rates of uterine bleeding and breast tenderness were observed with low-dose estrogens [25]. Therefore, physicians must guide patients in reaching the balance of decreasing vasomotor symptoms and living with the risks of estrogens.
Current evidence suggests that the risk of breast cancer for users of less than 5 years remains low and is similar to the risks of moderate alcohol ingestion, lack of regular exercise and postmenopausal obesity [26]. The FDA and the American College of Obstetrics and Gynecologists endorse the use of postmenopausal hormone therapy at the lowest dose and for the shortest possible time for treatment of moderate-to-severe hot flashes [27,28]. Furthermore, lower rates of baseline diseases among younger women would lead to a smaller absolute increase in risk than in older women. In the subgroup of women aged 50–59 years in the WHI with hot flashes at baseline, no significant increased risk for coronary artery disease or stroke was observed [29].
Future perspective
The risk of breast cancer is small for short-term use of estrogen. However, little is known regarding long-term risks. Large clinical trials of different types of low-dose estrogen could adequately document long-term safety [12], but would likely be prohibitively expensive and hampered by large losses to follow-up and crossovers.
In the meantime, physicians and patients need to be educated on the availability of low-dose estrogens and the minimal increased risks associated with their use. National surveys showed that many women have continued the use of high-dose estrogen, even after publication of the WHI and American College of Obstetricians and Gynecologists guidelines [4,25,30]. In 2003, 0.625 mg oral CEE still constituted 50% of estrogens prescribed in the USA [4]. For a woman burdened by vasomotor symptoms, what constitutes ‘acceptable risk’ is a personal decision to be based on known risks of alternatives.
Executive summary
Most women experience disruptive menopausal symptoms, but fewer than half seek medical attention for relief.
Despite significant interest in nonhormonal therapies for the treatment of these symptoms, existing evidence does not provide any convincing data on safe and effective treatments.
Short-term use of oral or transdermal estradiol in women with a history of hysterectomy confers minimal risk of breast cancer.
Significantly increased risk of breast cancer was only observed after 5 or more years of estradiol use.
Definitions of ‘acceptable’ risk vary for each woman burdened by vasomotor symptoms.