Commonly Used Types of Postmenopausal Estrogen for Treatment of Hot Flashes:  Scientific Review FREE

Estrogen was approved as a hormone supplement in the 1940s to treat menopausal symptoms. During the ensuing years, observational studies indicated additional health benefits for estrogen users such as prevention of chronic diseases like cardiovascular disease and osteoporosis.¹ A national survey conducted in 1995 indicated that 37% of women aged 50 years or older were using estrogen for multiple purposes.²

Estrogen use has decreased since 2002^3- 4 when the results of the Women's Health Initiative (WHI), the first large randomized controlled trial of estrogen for prevention of disease, released its findings contradicting previous beliefs about cardiovascular benefits.⁵ In this study, women using conjugated equine estrogen (CEE) and medroxyprogesterone acetate had significantly increased risks of coronary heart disease events, strokes, and breast cancer than women taking placebo.⁵ The US Food and Drug Administration recently ordered estrogen safety warnings on product labels referring to WHI findings and altered approved indications for its use.^6- 7 Package inserts indicate that treatment of menopausal symptoms remains an indication for estrogen use, although now physicians are advised to use the smallest effective dose for the shortest duration possible.⁷ The National Institutes of Health added steroidal estrogens to its list of known human carcinogens.⁸ The US Preventive Services Task Force⁹ as well as professional organizations¹⁰ updated their recommendations and now advise against using estrogen for prevention of chronic conditions. Recently the estrogen-only treatment group of the WHI was stopped due to increased incidence of strokes (http://www.nhlbi.nih.gov/whi/#estrogen).

Several estrogen preparations are available for symptom management, including oral, transdermal, and topical forms. There is interest in comparing different estrogen agents because of concerns about CEE reported in the WHI study. Differences between agents and routes have been described,¹¹ although it is not known if these differences result in important clinical effects. Treatment with transdermal 17β-estradiol provides higher estradiol levels than corresponding doses of CEE that provide higher levels of estrone and estrone sulfate.¹¹ This difference reflects the hormonal compositions of the different drugs as well as the consequences of the hepatic first-pass metabolism effect with oral use.

Recent trials^1,12 indicate that when estrogen is combined with a progestin or progesterone, the risks of endometrial hypertrophy and endometrial cancer are comparable with placebo. Both agents can be combined into 1 daily pill, although other regimens using separate estrogen and progestin or progesterone pills taken together or distributed cyclically over a month are also used. The effect of progestin or progesterone on other clinical outcomes, such as cardiovascular disease, is not clear.

The purpose of this review was to compare the efficacy and safety of the most commonly used estrogen preparations for reducing menopausal hot flashes. Trials of oral CEE and oral and transdermal 17β-estradiol were focused on because they are commonly used in the United States and our preliminary search of the literature indicated few published trials of other forms. Our preliminary review also indicated that trials of estrogen for treatment of other menopausal symptoms, such as quality-of-life, mood changes, and vaginal atrophy, varied widely in methodological approaches and outcome measures precluding a quantitative analysis of results.

MEDLINE (1966 to July 2003), EMBASE (1980 to July 2003), the Cochrane Database of Systematic Reviews and Cochrane Controlled Trials Registry (2003, issue 1), and reference lists of published articles, including a recently published systematic review listed in the Cochrane database, were searched for trials.^13- 14Citations were obtained from pharmaceutical manufacturers and experts. All citations were imported into an electronic database (EndNote 6.0; Thomson ISI ResearchSoft, Carlsbad, Calif).

The English-language, double-blind, randomized, placebo-controlled trials and systematic evidence reviews of oral CEE and oral and transdermal 17β-estradiol, and treatment of menopausal hot flashes and flushes were included. Included studies were at least 3 months in duration and compared one estrogen preparation with another estrogen or placebo with or without concomitant use of progestin or progesterone administered as cyclic or continuous regimens. Progestin or progesterone preparations were not separately considered.

Study participants included women experiencing menopause who were recruited from health care settings or the general population. When available, data were considered separately for women with natural or surgical menopause (oophorectomy), and for women in perimenopausal or postmenopausal periods. Perimenopausal women were considered as those women transitioning through natural menopause who had irregular menstrual periods within the last 12 months. Postmenopausal women were those women with surgical oophorectomy, or natural menopause and amenorrhea for more than 12 months. Differences based on patient characteristics, such as age, race, comorbidities, and early oophorectomy (<45 years) or premature menopause (<35 years), were also considered. Studies of women with major intercurrent disease were excluded as where those with previous estrogen use within 1 month of commencement of the study due to carry-over effects.

Outcome measures included hot flashes or flushes defined as any otherwise unexplained sensation of flushing or sweating experienced by the woman being studied. Although the term flash indicates a prodromal phase and flush the vasomotor dilation phase, they are combined herein because they were reported inconsistently among the trials. Hot flashes were measured in many ways in the estrogen trials. Most commonly, study participants recorded the number of episodes over a day or week period, and changes indicated treatment responses. Other trials used measures such as percentage of participants experiencing symptoms or severity of symptoms. A cumulative symptom score, the Kupperman Index,¹⁵ was used in some studies to classify the severity and intensity of hot flashes as well as various other menopausal symptoms. However, the use of the score is controversial because it has not been validated. Studies were included if they measured frequency, severity, presence vs absence, or a combination measure of frequency and severity as either primary or secondary outcomes at baseline, 3 months, and/or end of study.

Outcomes were determined by the differences in hot flashes measured at baseline compared with the end of the study. Treatment effects were defined as the differences in outcomes between the estrogen and placebo groups, or second estrogen group for head-to-head comparisons, at the end of the study. For crossover trials, only results from the end of the first phase were used because of the potential carry-over effect.

Adverse effects were also evaluated such as withdrawals from the study, atypical bleeding, endometrial hypertrophy, nausea and vomiting, breast tenderness, headaches, weight changes, dizziness, thrombosis, cardiovascular events, rash and pruritus, cholecystitis, effects on the liver, and other adverse effects, if reported.

From each trial, study design, population characteristics, eligibility criteria, interventions (estrogen type, form, dose and duration, use of progestin or progesterone, cyclic or continuous regimen), comparisons, numbers enrolled and lost to follow-up, method of outcome ascertainment, results for each outcome, and adverse effects were assessed. Intention-to-treat results were recorded if available.

For trials not included in the published Cochrane review,^13- 14 the internal validity (quality) was assessed using predefined criteria based on those developed by the US Preventive Services Task Force and the UK National Health Services Centre.^16- 18 Only trials of good or fair quality were included. All trials included in the Cochrane review were of at least fair quality by these criteria and were not rated in this review. External validity of trials was assessed based on whether the publication adequately described the study population, how similar patients were to the target population in whom the intervention will be applied (menopausal women with hot flash symptoms seeking treatment), and whether the treatment received by the control group was reasonably representative of standard practice. The funding source was also recorded. Overall quality ratings for individual studies were based on ratings of the internal and external validity of each trial.

A meta-analysis was conducted of trials reporting hot flash outcomes to provide a more precise and standard measure of treatment effect. Trials that presented data on frequency of hot flash outcomes after treatment in numerical format and provided standard deviations met criteria for the meta-analysis (Figure 1). DerSimonian-Laird weighted mean differences in mean weekly number of hot flashes were calculated to estimate pooled effects. This assumes a random effect or between-study variation in addition to within-study variation. The calculations were generated by using StatsDirect statistical software version 1.9.14.¹⁹ Funnel plots were constructed and indicated no evidence of publication bias, although they are a crude estimate and were limited by the small numbers of eligible studies.

Four trials compared estrogen preparations head-to-head, including 1 trial of CEE compared with oral 17β-estradiol²⁰ and 3 trials comparing oral CEE with transdermal 17β-estradiol^11,21- 22 (Table 1 and Table 2). Women enrolled in these trials had hot flashes at baseline, and all trials reported improved number, severity of hot flashes, or both for all of the estrogen treatment groups. There were no statistically significant differences in treatment effects in any of the head-to-head estrogen comparisons in any of the trials.

Of 3 trials comparing oral CEE with transdermal 17β-estradiol, 2 were combined in a meta-analysis,^21- 22 and 1 was excluded because data were provided in graphic form.¹¹ The pooled weighted mean difference in hot flashes was not significantly different between 17β-estradiol and CEE treatment groups, thereby favoring neither agent (–0.3; 95% confidence interval [CI], –3.4 to 2.7).

Dose-response trends were demonstrated in trials that tested multiple doses with higher doses corresponding to bigger treatment effects; however, these did not reach statistical significance.^20- 21 In 1 study, patients using CEE at 0.625 mg/d had a reduction of mean daily frequency of hot flashes by 80%, and patients using CEE at 1.25 mg/d had a reduction of 95% (P = .06).²⁰ Too few dose comparisons were conducted between estrogen agents to determine if differences exist at various doses.

Twenty-eight randomized controlled trials comparing CEE or 17β-estradiol with placebo met criteria for this review.^{20- 21,23- 48} Trials were conducted predominantly in the United States or western Europe and recruited participants from general primary care or gynecology practices. Trials enrolled patients with mean age approximately 50 years (range, 25-88 years) and included from 24 to 2763 patients in 1 to 8 comparison groups.

Inclusion criteria varied among studies from most or a percentage of participants with baseline symptoms to a specified level of symptoms, such as "5 or more vasomotor symptoms per day." Trials often enrolled both perimenopausal and postmenopausal women but did not separate them in the analysis-limiting comparisons. Hysterectomy status was clearly reported if inclusion criteria called for women either with or without hysterectomy.³⁷ For trials including both women with and without hysterectomies, data were not separately reported and comparisons could not be made. No trial specifically addressed treatment in women with premature ovarian failure. Reporting of concurrent medications, comorbidities, or other potential confounders was minimal, and inclusion criteria generally focused on healthy symptomatic women.

Different outcomes were reported and lack of standardization limited comparisons. Frequency of hot flashes was the most common measure reported in 19 of 28 trials.^{20- 21,23,25- 28,30,32- 40,45,48} When reported, women in placebo groups also had improvement of symptoms from baseline by as much as 66%.²⁰ Ten trials included estrogen combined with various progestin or progesterone agents.^{23- 24,26- 27,31,35- 36,40,47- 48} Studies that compared groups using estrogen alone with groups using estrogen with progestin or progesterone found no differences in treatment effects.^24,47- 48

Nine of 10 trials of oral 17β-estradiol demonstrated statistically significant improvements in hot flash frequency, severity, or both compared with placebo (Table 1 and Table 2, Table 3, and Table 4).^{20,23- 24,26- 31} The 1 trial that reported no difference between groups was conducted in Chinese women in Hong Kong after oophorectomy.²⁵ Approximately 66% of women in this trial had vasomotor symptoms at baseline and 23% to 35% considered them moderate to severe, a lower level than in some of the other trials. One trial reported that women in early (3-12 months amenorrhea) as well as late menopause (>12 months amenorrhea) had similar benefit.²³ Five trials included concomitant progestin or progesterone use (continuous and cyclic norethindrone acetate, cyclic nomegestrol).^{23- 24,26- 27,31}

All 11 trials of transdermal 17β-estradiol reported statistically significant improvements in hot flash frequency, severity, or both compared with placebo (Table 5 and Table 6).^21,32- 41 Three trials included concomitant progestin or progesterone (cyclic norethindrone acetate, continuous transdermal levonorgestrel).^35- 36,40

All 8 trials of oral CEE reported statistically significant improvements in hot flash frequency, severity, or both compared with placebo (Table 1, Table 7 and Table 8).^20,42- 48 Three trials included treatment groups with concomitant progestin or progesterone use (cyclic and continuous medroxyprogesterone acetate, cyclic micronized progesterone).^42,47- 48 One trial compared 3 doses of CEE alone (0.3, 0.45, and 0.625 mg/d) and reported bigger treatment effects with 0.625 mg than 0.45 mg or 0.3 mg (P<.05).⁴⁸ Differences between estrogen doses were not found in patients provided with CEE (0.3, 0.45, and 0.625 mg/d) and continuous medroxyprogesterone acetate (1.5 or 2.5 mg/d) in this trial.⁴⁸

Of 10 trials of oral 17β-estradiol compared with placebo, 5 met criteria for the meta-analysis.^{23,26- 28,31} The pooled weighted mean difference in hot flashes was –16.8 per week (95% CI, –23.4 to –10.2) compared with placebo (Figure 2). Combining only the 4 trials that included 17β-estradiol and progestin or progesterone did not significantly change results (pooled weighted mean difference, –19.1; 95% CI, –29.6 to –8.6).^{23,26- 27,31} Trials were excluded from analysis because they did not provide data on frequency of hot flashes^24- 25 or did not provide standard deviations.^20,29- 30

Of 11 trials of transdermal 17β-estradiol compared with placebo, 6 met criteria for the meta-analysis.^{21,32,34,37,39- 40} The pooled weighted mean difference in hot flashes for these trials was –22.4 per week (95% CI, –35.9 to –10.4) compared with placebo (Figure 3). Only 1 trial included 17β-estradiol and progestin or progesterone and results were not significantly different from the other studies.⁴⁰ Trials were excluded because they did not provide data on frequency of hot flashes,^32,41 provided data in graphic form,^33,35,38 or did not provide standard deviations.^36,38

Of 8 trials of CEE compared with placebo, 1 met criteria for the meta-analysis.⁴⁶ This trial reported a mean reduction of –19.1 hot flashes per week (95% CI, –33.0 to –5.1) after treatment compared with placebo. The other 7 trials were excluded from analysis because they did not provide data on frequency of hot flashes,^42,45,47 provided data in graphic form,⁴³ or did not provide standard deviations.^{20,43- 44,48}

All but 5 trials^{24- 25,42,47- 48} were less than 1 year in duration and only 3 trials enrolled more than 500 participants.^42,47- 48Studies reported multiple specific adverse effects, including atypical bleeding and endometrial hypertrophy, nausea and vomiting, breast tenderness, headache, weight change, dizziness, venous thromboembolic events, cardiovascular events, rash and pruritus, cholecystitis, liver effects, and other adverse events. These outcomes were reported unevenly across studies and could not be combined in summary statistics.

Head-to-head comparison trials lacked data to determine the relative adverse effects of different estrogens. One trial of CEE and oral 17β-estradiol reported that the incidence of possible drug-related adverse experiences ranged from 20% in placebo, 1-mg/d 17β-estradiol, and 0.625-mg/d CEE groups to 35% in 2-mg/d 17β-estradiol and 1.25-mg/d CEE groups with no statistically significant differences between groups.²⁰ Among trials with placebo groups, comparisons between types of estrogens could not be made with the data provided.

A 4-year trial with 875 patients reported 2 cases of deep vein thrombosis among CEE users,⁴⁷ otherwise no cardiovascular events were reported in the trials. Breast tenderness^{11,20,23- 24,26- 27,33- 34,37- 39,42,47- 49} and vaginal bleeding^{11,21- 24,26,29- 30,33- 34,36,38,41- 42} were the most commonly reported adverse effects among estrogen users in the trials. Two trials each reported 1 case of endometrial cancer in an 17β-estradiol user.^30,34 Bleeding and breast tenderness were more frequent among patients with higher vs lower doses of estrogen regardless of the type of estrogen in some trials.^11,20- 21 Adverse skin reactions were most common among women using transdermal forms of 17β-estradiol or placebo.^{21,32- 35,38- 41,49- 50} Withdrawals from placebo groups due to lack of treatment effect were also reported.^27,43

Trials included in this review used similar methodology and met criteria for at least a fair quality score for internal and external validity. In several studies, it was not apparent whether quality criterion, such as use of intention-to-treat analysis, was met because it was not reported in the publication. The most common problem with the studies was differential loss to follow-up and it was unclear if comparable groups were maintained. Most studies were either funded by industry or the funding source was not reported.

Trials included in this systematic review indicate that CEE and oral and transdermal 17β-estradiol are more effective than placebo in relieving menopausal hot flashes and available evidence does not indicate that one agent is more effective than another. A range of doses is effective, although a dose-response relationship was reported in a limited number of studies. Although data are limited, concomitant use of progestin or progesterone does not influence the effect of estrogen. Available trials do not allow additional comparisons between types of estrogens to determine the effects of cyclic and continuous regimens. There are too few trials of other types of estrogen than CEE and 17β-estradiol to evaluate their relative effectiveness.

These results are consistent with a Cochrane review and meta-analysis of trials of oral estrogens compared with placebo for treating menopausal hot flashes published before 2000.¹⁴ Differences between types of estrogens were not determined in this review, although trials of 17β-estradiol and CEE predominated. Results indicated a 77% reduction in frequency and a significant reduction in severity of symptoms with oral estrogen compared with placebo.

Data from trials evaluated in this review do not allow comparisons of adverse effects because they were reported in incomplete and nonstandardized ways. The most comprehensive data about adverse effects of estrogen are reported in studies designed for purposes other than symptom treatment, such as the WHI.⁵ This trial was designed as a primary prevention trial and enrolled more than 16 000 women with a mean age of 63 years at study entry. After 5 years of continuous administration of 0.625-mg/d CEE and 2.5-mg/d medroxyprogesterone acetate, estrogen users had significantly increased coronary heart disease events,⁵¹ strokes,⁵² deep vein thrombotic events,⁵ and breast cancer⁵³ compared with nonusers. The symptom treatment trials reviewed herein enrolled small numbers of patients for short periods and were inadequately designed to capture the important health outcomes reported by the WHI. One trial reported 2 CEE users with deep vein thrombosis⁴⁷ and 2 trials reported 1 case each of endometrial cancer in 17β-estradiol users^30,34; otherwise, adverse effects included predominantly vaginal bleeding, breast tenderness, and other assorted nuisance symptoms. Although these adverse effects are important to individual women and may result in stopping use of estrogen, they are less serious health outcomes than those reported in the WHI. Available trial data do not prove that serious outcomes will not occur in younger short-term users but it is inconclusive.

Symptom treatment trials have other important limitations. Most trials enrolled white women in the United States or western Europe who were recruited through clinical practices. The few trials conducted in nonwhite women took place in countries where lifestyle factors substantially differ from those in the United States and could potentially influence outcomes. Trials usually included women ranging in age from 40 to 60 years old with a mean age of early 50s. Comparisons of results for these women with women of different age groups, racial or ethnic groups, comorbidities, and risk factors are not possible. No trials considered smokers, women at high risk for ovarian or breast cancer, or other risk factors and comorbidities separately. No trials compared women with early oophorectomy or premature menopause with women undergoing menopause at an older age.

This systematic review and meta-analysis of 32 treatment trials found that the use of CEE and oral and transdermal 17β-estradiol have consistent and comparable effects on treatment of hot flashes in menopausal women with symptoms and may have similar adverse effects. However, many issues remain unresolved by current trial data. Future trials could address these issues by providing a broader demographic sample of women, longer follow-up, larger numbers of patients, and more head-to-head comparisons of estrogens, progestins or progesterones, and other therapies (phytoestrogens, megestrol, clonidine, selective serotonin-reuptake inhibitors). Results of these trials would guide more individualized use of estrogen, including appropriate selection of treatment candidates, monitoring of treatment and adverse effects, and determining when and how to discontinue therapy.