David F. Archer, MD

Endometrial cancer is the most common gynecologic cancer in the developed world, and its incidence is rising—increasing 50% in the past 10 years in Norway alone. Because endometrial hyperplasia is a precursor to cancer, it is vital that we ensure effective treatment for this prevalent problem. Hysterectomy is one option—used most commonly for complex atypical hyperplasia—but oral progestins have become the norm in women who desire to preserve their uterus when surgery is not the best option.

In this randomized trial from Norway, Orbo and colleagues randomly assigned 170 women aged 30 to 70 years to one of three treatment groups:

• placement of an LNG-IUS (Mirena)
• medroxyprogesterone acetate (MPA) 10 mg for 10 days per cycle
• continuous MPA 10 mg.

All women in the trial had low- or medium-risk endometrial hyperplasia.

After 6 months, women in the LNG-IUS arm had a 100% response rate, compared with 96% for the continuous MPA group and 69% for cyclic MPA.

Related article: When should a menopausal woman discontinue hormone therapy? Andrew M. Kaunitz, MD (Cases in Menopause; March 2014)

Histologic interpretation of hyperplasia is highly subjective
There are several problems inherent in a study like this. Although Orbo and colleagues address these problems tangentially, the problems affect the interpretation of results.

For example, the histologic interpretation of endometrial hyperplasia is known to be associated with low interobserver agreement.¹ Clinical trials that use endometrial safety as an outcome require two primary pathologists to review the histology, with a third pathologist standing by in case of disagreement.

In the current study, two pathologists in the same department independently reviewed the histology. Orbo and colleagues used World Health Organization criteria for hyperplasia. However, as an adjunct, they also used a D-evaluation morphometric assessment.² When I put in a casual call to local gynecologic pathologists, they told me that neither the D-classification nor the immunochemical-detected PTEN protein is used in routine clinical practice to determine the risk of progression.³

Intermittent use of oral MPA is known to be ineffective
The cyclic use of MPA for only 10 days overlooks epidemiology from estrogen-progestin replacement regimens in postmenopausal women. Use of a progestin for fewer than 12 days during estrogen replacement increases the risk of endometrial cancer.⁴ Exogenous progestin must be given for more than 12 days to inhibit hyperplasia and neoplasia. The dose itself is not critical; the duration of administration is.

In the current study, both the LNG-IUS and continuous MPA met this criterion. Local delivery of the progestin with the LNG-IUS allows for a reduction of the delivered dose and mitigates side effects even as it uses a more potent progestin than MPA.^5–8

Assessment of outcomes was questionable
Although Orbo and colleagues suggest that there is no evidence of progression with the LNG-IUS and continuous MPA, they relied on a Pipelle biopsy of the endometrium performed after 6 months of treatment. The clinical settings that led to the hyperplasia in the first place are poorly characterized as either pre- or postmenopausal, and the cause of the hyperplasia is not identified. This approach overlooks such realities as the increased incidence of simple hyperplasia in many perimenopausal women, which appears to regress with further reduction in ovarian estrogen.⁹

The final outcomes for women in the cyclic MPA arm are not provided. Hyperplasia without atypia progresses to carcinoma in 1.6% of cases, but when atypia is present, the progression rate is 23%.¹

What this evidence means for practice
In low-risk women with simple hyperplasia, the use of targeted low-dose progestins—oral or intrauterine—is appealing. While Orbo and colleagues present an interesting study, they do not definitively establish the optimal intervention.
As we enter a cost-conscious phase of medicine in the United States, we may discover that oral generic MPA (given continuously) may be the most cost-effective treatment despite the option of delivering a low-dose progestin via intrauterine device.
David F. Archer, MD

WE WANT TO HEAR FROM YOU!
Share your thoughts on this article or on any topic relevant to ObGyns and women’s health practitioners. Tell us which topics you’d like to see covered in future issues, and what challenges you face in daily practice. We will consider publishing your letter in a future issue. Send your letter to: [email protected] Please include the city and state in which you practice. Stay in touch! Your feedback is important to us!

Endometrial cancer is the most common gynecologic cancer in the developed world, and its incidence is rising—increasing 50% in the past 10 years in Norway alone. Because endometrial hyperplasia is a precursor to cancer, it is vital that we ensure effective treatment for this prevalent problem. Hysterectomy is one option—used most commonly for complex atypical hyperplasia—but oral progestins have become the norm in women who desire to preserve their uterus when surgery is not the best option.

In this randomized trial from Norway, Orbo and colleagues randomly assigned 170 women aged 30 to 70 years to one of three treatment groups:

• placement of an LNG-IUS (Mirena)
• medroxyprogesterone acetate (MPA) 10 mg for 10 days per cycle
• continuous MPA 10 mg.

All women in the trial had low- or medium-risk endometrial hyperplasia.

After 6 months, women in the LNG-IUS arm had a 100% response rate, compared with 96% for the continuous MPA group and 69% for cyclic MPA.

Related article: When should a menopausal woman discontinue hormone therapy? Andrew M. Kaunitz, MD (Cases in Menopause; March 2014)

Histologic interpretation of hyperplasia is highly subjective
There are several problems inherent in a study like this. Although Orbo and colleagues address these problems tangentially, the problems affect the interpretation of results.

For example, the histologic interpretation of endometrial hyperplasia is known to be associated with low interobserver agreement.¹ Clinical trials that use endometrial safety as an outcome require two primary pathologists to review the histology, with a third pathologist standing by in case of disagreement.

In the current study, two pathologists in the same department independently reviewed the histology. Orbo and colleagues used World Health Organization criteria for hyperplasia. However, as an adjunct, they also used a D-evaluation morphometric assessment.² When I put in a casual call to local gynecologic pathologists, they told me that neither the D-classification nor the immunochemical-detected PTEN protein is used in routine clinical practice to determine the risk of progression.³

Intermittent use of oral MPA is known to be ineffective
The cyclic use of MPA for only 10 days overlooks epidemiology from estrogen-progestin replacement regimens in postmenopausal women. Use of a progestin for fewer than 12 days during estrogen replacement increases the risk of endometrial cancer.⁴ Exogenous progestin must be given for more than 12 days to inhibit hyperplasia and neoplasia. The dose itself is not critical; the duration of administration is.

In the current study, both the LNG-IUS and continuous MPA met this criterion. Local delivery of the progestin with the LNG-IUS allows for a reduction of the delivered dose and mitigates side effects even as it uses a more potent progestin than MPA.^5–8

Assessment of outcomes was questionable
Although Orbo and colleagues suggest that there is no evidence of progression with the LNG-IUS and continuous MPA, they relied on a Pipelle biopsy of the endometrium performed after 6 months of treatment. The clinical settings that led to the hyperplasia in the first place are poorly characterized as either pre- or postmenopausal, and the cause of the hyperplasia is not identified. This approach overlooks such realities as the increased incidence of simple hyperplasia in many perimenopausal women, which appears to regress with further reduction in ovarian estrogen.⁹

The final outcomes for women in the cyclic MPA arm are not provided. Hyperplasia without atypia progresses to carcinoma in 1.6% of cases, but when atypia is present, the progression rate is 23%.¹

What this evidence means for practice
In low-risk women with simple hyperplasia, the use of targeted low-dose progestins—oral or intrauterine—is appealing. While Orbo and colleagues present an interesting study, they do not definitively establish the optimal intervention.
As we enter a cost-conscious phase of medicine in the United States, we may discover that oral generic MPA (given continuously) may be the most cost-effective treatment despite the option of delivering a low-dose progestin via intrauterine device.
David F. Archer, MD

WE WANT TO HEAR FROM YOU!
Share your thoughts on this article or on any topic relevant to ObGyns and women’s health practitioners. Tell us which topics you’d like to see covered in future issues, and what challenges you face in daily practice. We will consider publishing your letter in a future issue. Send your letter to: [email protected] Please include the city and state in which you practice. Stay in touch! Your feedback is important to us!

Endometrial cancer is the most common gynecologic cancer in the developed world, and its incidence is rising—increasing 50% in the past 10 years in Norway alone. Because endometrial hyperplasia is a precursor to cancer, it is vital that we ensure effective treatment for this prevalent problem. Hysterectomy is one option—used most commonly for complex atypical hyperplasia—but oral progestins have become the norm in women who desire to preserve their uterus when surgery is not the best option.

In this randomized trial from Norway, Orbo and colleagues randomly assigned 170 women aged 30 to 70 years to one of three treatment groups:

• placement of an LNG-IUS (Mirena)
• medroxyprogesterone acetate (MPA) 10 mg for 10 days per cycle
• continuous MPA 10 mg.

All women in the trial had low- or medium-risk endometrial hyperplasia.

After 6 months, women in the LNG-IUS arm had a 100% response rate, compared with 96% for the continuous MPA group and 69% for cyclic MPA.

Related article: When should a menopausal woman discontinue hormone therapy? Andrew M. Kaunitz, MD (Cases in Menopause; March 2014)

Histologic interpretation of hyperplasia is highly subjective
There are several problems inherent in a study like this. Although Orbo and colleagues address these problems tangentially, the problems affect the interpretation of results.

For example, the histologic interpretation of endometrial hyperplasia is known to be associated with low interobserver agreement.¹ Clinical trials that use endometrial safety as an outcome require two primary pathologists to review the histology, with a third pathologist standing by in case of disagreement.

In the current study, two pathologists in the same department independently reviewed the histology. Orbo and colleagues used World Health Organization criteria for hyperplasia. However, as an adjunct, they also used a D-evaluation morphometric assessment.² When I put in a casual call to local gynecologic pathologists, they told me that neither the D-classification nor the immunochemical-detected PTEN protein is used in routine clinical practice to determine the risk of progression.³

Intermittent use of oral MPA is known to be ineffective
The cyclic use of MPA for only 10 days overlooks epidemiology from estrogen-progestin replacement regimens in postmenopausal women. Use of a progestin for fewer than 12 days during estrogen replacement increases the risk of endometrial cancer.⁴ Exogenous progestin must be given for more than 12 days to inhibit hyperplasia and neoplasia. The dose itself is not critical; the duration of administration is.

In the current study, both the LNG-IUS and continuous MPA met this criterion. Local delivery of the progestin with the LNG-IUS allows for a reduction of the delivered dose and mitigates side effects even as it uses a more potent progestin than MPA.^5–8

Assessment of outcomes was questionable
Although Orbo and colleagues suggest that there is no evidence of progression with the LNG-IUS and continuous MPA, they relied on a Pipelle biopsy of the endometrium performed after 6 months of treatment. The clinical settings that led to the hyperplasia in the first place are poorly characterized as either pre- or postmenopausal, and the cause of the hyperplasia is not identified. This approach overlooks such realities as the increased incidence of simple hyperplasia in many perimenopausal women, which appears to regress with further reduction in ovarian estrogen.⁹

The final outcomes for women in the cyclic MPA arm are not provided. Hyperplasia without atypia progresses to carcinoma in 1.6% of cases, but when atypia is present, the progression rate is 23%.¹

What this evidence means for practice
In low-risk women with simple hyperplasia, the use of targeted low-dose progestins—oral or intrauterine—is appealing. While Orbo and colleagues present an interesting study, they do not definitively establish the optimal intervention.
As we enter a cost-conscious phase of medicine in the United States, we may discover that oral generic MPA (given continuously) may be the most cost-effective treatment despite the option of delivering a low-dose progestin via intrauterine device.
David F. Archer, MD

WE WANT TO HEAR FROM YOU!
Share your thoughts on this article or on any topic relevant to ObGyns and women’s health practitioners. Tell us which topics you’d like to see covered in future issues, and what challenges you face in daily practice. We will consider publishing your letter in a future issue. Send your letter to: [email protected] Please include the city and state in which you practice. Stay in touch! Your feedback is important to us!

Now that the dust is settling from the Women’s Health Initiative (WHI), our patients are again asking reasonable questions about hormone replacement therapy (HRT). I remind them of estrogen’s proven advantages in menopause, as well as its risks. Although most women are generally aware of these risks and benefits, considerable misunderstanding persists. This article reviews what the evidence to date does and does not confirm, particularly regarding breast cancer and coronary heart disease, where most of the uncertainty remains.

HRT stops vaginal atrophy, hot flashes, and bone loss

Three applications form the basis for HRT in postmenopausal women:

1. Hot flashes subside. Hot flashes occur with varying intensity in about 85% of women, and are effectively treated with estrogen, whether given orally, transdermally, or vaginally.^1,2 As long as an appropriate blood level of the hormone is reached, hot flashes diminish.^3-5 This reduction is dose-related.
   
2. Measurable improvements in vaginal atrophy. Estrogen’s efficacy in relieving dryness, itching, burning, and dyspareunia is well demonstrated, regardless of the route of administration.^3,6,7 A fall in vaginal pH from 6.0 to 5.0 after estrogen administration has been documented,⁸ as has the increase in the number of superficial cells of the vagina with exogenous estrogen.⁹
   
3. HRT maintains or increases bone mineral density (BMD). Most estrogen preparations on the US market have been shown to improve BMD.^10-15 “Improvement” means no significant loss, or an increase, in BMD. In the WHI, both vertebral and nonvertebral fractures diminished unequivocally in women using estrogen—alone or with a progestin.^16,17 Other clinical trials also have shown increased BMD, as well as decreased urinary and serum markers of bone turnover.
   

Do new data link progestin to cancer?

Although compelling evidence supports the use of progestational agents in addition to estrogen to prevent endometrial hyperplasia and endometrial cancer,¹⁸ a 2005 report¹⁹ suggests that chronic, long-term use of estrogen with a progestin may increase the risk of endometrial carcinoma. Because this is the only study in which this risk has been found, corroboration is required.

Until then, give progestin at a sufficient dose and duration to inhibit endometrial hyperplasia.^20-25

Effects on heart disease may be age-related

With notable exceptions, the overall conclusion of clinical trials and observational studies to date is that estrogen helps prevent coronary heart disease (CHD).^26-30 This finding was first observed in the late 1980s with evidence that estrogen increases high-density lipoprotein (HDL) cholesterol and reduces total and low-density lipoprotein (LDL) cholesterol.³¹

Some experts argue that these observational trials are biased because many of the women taking estrogen had modified their lifestyles to maintain their weight, control their diet, and exercise regularly.³² Indeed, the randomized, placebo-controlled Heart and Estrogen Replacement Study (HERS) and both arms of the WHI trial found no evidence for a significant increase or decrease in CHD events.^33-35

Time from menopause to HRT may be key

Both the HERS and WHI trials enrolled older women who had entered menopause a few months to several years before starting HRT.³⁶ In addition, the estrogen-progestin arm of the WHI trial lacked sufficient power to detect a significant difference in CHD outcomes.³⁷

The WHI findings contrast those of the large, ongoing, observational Nurses Health Study, which has shown a consistent decrease in CHD incidence in women who began HRT with the onset of menopausal symptoms.^27-30 The most recent data suggest that the interval between menopause and the start of HRT may explain the different findings in randomized, controlled trials and observational studies.³⁸ The WHI data support this theory: CHD was lower in women who began taking HRT within 5 years of menopause, compared with women who initiated HRT more than 5 years afterward.³⁶ In addition, data from the estrogen-only arm of the WHI show fewer CHD events in women younger than 60.³⁴

Several other studies support this hypothesis:

• The surgically postmenopausal cynomolgus macaque had a lower rate of atherosclerotic plaque development when estrogen was given, with or without a progestin.^39,40
  
• In the Rancho Bernardo study, women who had used HRT had less cardiac calcification documented by computed tomography, compared with nonusers.⁴¹
  
• Estrogen has been shown, by measurement of carotid intimal medial thickness, to inhibit atherosclerotic plaque in humans.⁴²
  
• Older women with established atherosclerosis do not undergo any significant change in plaque size with the use of exogenous estrogen.⁴³
  

 

Although these findings support the use of estrogen or estrogen-progestin early after menopause as a way of preventing CHD, further clinical trials are needed.⁴⁴

Stroke risk is small but real

Both arms of the WHI found an increased incidence of stroke in women using hormones, compared with nonusers.^16,36 The exact mechanisms underlying this increased risk are unclear.

The actual attributable risk was an increase of 0.7 cases of stroke per 1,000 women per year over placebo in the estrogen-progestin arm,³⁶ and 1.2 cases per 1,000 in the estrogen-only arm.¹⁶ The relative hazards were 1.31 (95% confidence interval [CI] 1.02–1.68) and 1.30 (95% CI 1.10–1.77), respectively.

Note that women in the estrogen-only arm had a greater incidence of hypertension and diabetes mellitus—known risk factors for stroke—than did women in the estrogen-progestin arm.^16,36

VTE risk is twice as high in HRT users

Postmenopausal women who take estrogen have a higher risk of venous thromboembolism (VTE) than those who do not. This risk translated into a relative hazard of 2.06 (1.57–2.70) in the WHI estrogen-progestin arm, or an attributable risk of 3.6 cases per 1,000 women, compared with 1.8 cases per thousand in the control group.³⁶

The absolute increased risk is 1.8 cases per 1,000 women, or, as expressed in the study itself, 18 cases per 10,000 women per year.

I have deliberately reduced the attributable risk to the number of cases per thousand because I believe this number is more easily understood by the patient and accurately demonstrates the low risk.

In the estrogen-only arm of the WHI, the hazard ratio for VTE was 1.33 (0.99–1.79), or an absolute increased risk of 0.7 cases per thousand—although this finding was not significant. The attributable risk was 2.7 cases per 1,000 women, compared with 2.0 cases per thousand among controls.¹⁶

Like stroke, the risk of VTE may be confounded by other factors besides use of exogenous estrogen.

No cause and effect for HRT and breast cancer

Nothing frightens women as much as breast cancer, and articles focusing on the relationship between breast cancer and HRT have drawn widespread attention. However, despite voluminous literature, the etiology of breast cancer remains elusive—and there is no evidence that either estrogen or progestins cause the disease.^45,46 Rather, there is only an association between the use of estrogen, progestin, and breast cancer. Linking the finding of an increased risk with an implication of causality would be inappropriate.

Breast cancer risk with HRT is not consistently elevated, in studies

In fact, a qualitative review of observational studies from 1975 to 2000 found no significant increase or decrease in the risk of breast cancer with estrogen or estrogen-progestin in 80% of the reports.⁴⁷

Risk factors for breast cancer (TABLE 1) include family history, obesity, late childbirth, and hormone therapy—but obesity and family history have higher relative risks than the use of HRT.⁴⁸

TABLE 1

Relative risk of breast cancer

CHARACTERISTIC	RELATIVE RISK
2 family members with breast cancer	14
1 family member with breast cancer	2.2
Obesity	1.8
Young age at menarche	1.6
Hormone therapy	1.3
>30 years of age at birth of first child	1.3
Menopause	0.7

WHI arms find different risks

In the widely publicized WHI, women in the estrogen-progestin arm had an overall relative hazard for breast cancer of 1.24 (95% CI 1.01–1.54), but there was no increased risk in women who had never before used hormones.³⁶ Women who had previously used hormones for 5 years or more did have an increased risk.³⁶ The incidence of breast cancer in the study population was 3 cases per 1,000 women, and the excess number was 0.7 more cases with the use of estrogen-progestin (TABLE 2).

Conversely, in the estrogen-only arm of the WHI,¹⁶ the relative hazard for breast cancer was 0.77 (95% CI 0.59–1.01), and the reduction in risk was almost statistically significant. There are at least 2 potential explanations for the lower incidence of breast cancer in this arm:

• Without a progestin, estrogen increases breast density only minimally, allowing for easier mammographic interpretation.
  
• Women susceptible to breast cancer because of their previous use of estrogen may not have been present in the at-risk population in sufficient numbers to cause an increase.
  

Neither explanation—separate or combined—fully explains the lowered risk in this population. Each population studied appears to have a different level of risk based on multiple factors that cannot be controlled completely in clinical trials and observational studies.

TABLE 2

Extra cases of breast cancer, by risk factor

RISK FACTOR	BREAST CANCERS DIAGNOSED OVER 20 YEARS FROM AGES 50 TO 70 (PER 1,000)	EXTRA BREAST CANCERS (PER 1,000)
Never used HRT	45	-
>5 years HRT	47	2
>10 years HRT	51	6
>15 years HRT	57	12
Late menopause (age 60)	59	14
Alcohol (2 drinks/day)	72	27
No daily exercise	72	27
Weight gain (>20 kg)	90	45
Reprinted from THE LANCET, Vol. 350: 1047–1059, Collaborative Group on Hormonal Factors in Breast Cancer, Breast cancer and hormone replacement therapy: collaborative reanalysis of data from 51 epidemiological studies of 52,705 women with breast cancer and 108,411 women without breast cancer. Copyright 1997, with permission from Elsevier

 

HRT may promote, rather than induce, breast cancer

The role of hormones in the etiology of breast cancer is difficult to assess. The Million Women Study⁴⁹ found that the elevated risk of breast cancer disappeared within 1 year of stopping HRT. This finding implies that hormones may be a promoter, rather than inducer, of neoplasms in the breast.

Breast cancer may be present in many women, but apparent in few

When autopsies were performed on women in their 40s who had died from other diseases, the incidence of breast cancer was 39%, but the clinical detection rate was only 1% for this population.⁵⁰ This discrepancy suggests that neoplastic cells may be present in the body at any time, but become clinically apparent only under certain conditions.⁵¹

More recent data suggest that undifferentiated stem cells in the breast become dysfunctional and result in cancer.⁵² This theory is supported by the various histologic types of cancer found in the breast.

A weak link

Although it may be compelling to link hormone use with breast cancer, the association is weak and the incidence is lower than in other known relationships such as obesity. At present, the cause of breast neoplasia appears to be multifactorial.

Now that the dust is settling from the Women’s Health Initiative (WHI), our patients are again asking reasonable questions about hormone replacement therapy (HRT). I remind them of estrogen’s proven advantages in menopause, as well as its risks. Although most women are generally aware of these risks and benefits, considerable misunderstanding persists. This article reviews what the evidence to date does and does not confirm, particularly regarding breast cancer and coronary heart disease, where most of the uncertainty remains.

HRT stops vaginal atrophy, hot flashes, and bone loss

Three applications form the basis for HRT in postmenopausal women:

1. Hot flashes subside. Hot flashes occur with varying intensity in about 85% of women, and are effectively treated with estrogen, whether given orally, transdermally, or vaginally.^1,2 As long as an appropriate blood level of the hormone is reached, hot flashes diminish.^3-5 This reduction is dose-related.
   
2. Measurable improvements in vaginal atrophy. Estrogen’s efficacy in relieving dryness, itching, burning, and dyspareunia is well demonstrated, regardless of the route of administration.^3,6,7 A fall in vaginal pH from 6.0 to 5.0 after estrogen administration has been documented,⁸ as has the increase in the number of superficial cells of the vagina with exogenous estrogen.⁹
   
3. HRT maintains or increases bone mineral density (BMD). Most estrogen preparations on the US market have been shown to improve BMD.^10-15 “Improvement” means no significant loss, or an increase, in BMD. In the WHI, both vertebral and nonvertebral fractures diminished unequivocally in women using estrogen—alone or with a progestin.^16,17 Other clinical trials also have shown increased BMD, as well as decreased urinary and serum markers of bone turnover.
   

Do new data link progestin to cancer?

Although compelling evidence supports the use of progestational agents in addition to estrogen to prevent endometrial hyperplasia and endometrial cancer,¹⁸ a 2005 report¹⁹ suggests that chronic, long-term use of estrogen with a progestin may increase the risk of endometrial carcinoma. Because this is the only study in which this risk has been found, corroboration is required.

Until then, give progestin at a sufficient dose and duration to inhibit endometrial hyperplasia.^20-25

Effects on heart disease may be age-related

With notable exceptions, the overall conclusion of clinical trials and observational studies to date is that estrogen helps prevent coronary heart disease (CHD).^26-30 This finding was first observed in the late 1980s with evidence that estrogen increases high-density lipoprotein (HDL) cholesterol and reduces total and low-density lipoprotein (LDL) cholesterol.³¹

Some experts argue that these observational trials are biased because many of the women taking estrogen had modified their lifestyles to maintain their weight, control their diet, and exercise regularly.³² Indeed, the randomized, placebo-controlled Heart and Estrogen Replacement Study (HERS) and both arms of the WHI trial found no evidence for a significant increase or decrease in CHD events.^33-35

Time from menopause to HRT may be key

Both the HERS and WHI trials enrolled older women who had entered menopause a few months to several years before starting HRT.³⁶ In addition, the estrogen-progestin arm of the WHI trial lacked sufficient power to detect a significant difference in CHD outcomes.³⁷

The WHI findings contrast those of the large, ongoing, observational Nurses Health Study, which has shown a consistent decrease in CHD incidence in women who began HRT with the onset of menopausal symptoms.^27-30 The most recent data suggest that the interval between menopause and the start of HRT may explain the different findings in randomized, controlled trials and observational studies.³⁸ The WHI data support this theory: CHD was lower in women who began taking HRT within 5 years of menopause, compared with women who initiated HRT more than 5 years afterward.³⁶ In addition, data from the estrogen-only arm of the WHI show fewer CHD events in women younger than 60.³⁴

Several other studies support this hypothesis:

• The surgically postmenopausal cynomolgus macaque had a lower rate of atherosclerotic plaque development when estrogen was given, with or without a progestin.^39,40
  
• In the Rancho Bernardo study, women who had used HRT had less cardiac calcification documented by computed tomography, compared with nonusers.⁴¹
  
• Estrogen has been shown, by measurement of carotid intimal medial thickness, to inhibit atherosclerotic plaque in humans.⁴²
  
• Older women with established atherosclerosis do not undergo any significant change in plaque size with the use of exogenous estrogen.⁴³
  

 

Although these findings support the use of estrogen or estrogen-progestin early after menopause as a way of preventing CHD, further clinical trials are needed.⁴⁴

Stroke risk is small but real

Both arms of the WHI found an increased incidence of stroke in women using hormones, compared with nonusers.^16,36 The exact mechanisms underlying this increased risk are unclear.

The actual attributable risk was an increase of 0.7 cases of stroke per 1,000 women per year over placebo in the estrogen-progestin arm,³⁶ and 1.2 cases per 1,000 in the estrogen-only arm.¹⁶ The relative hazards were 1.31 (95% confidence interval [CI] 1.02–1.68) and 1.30 (95% CI 1.10–1.77), respectively.

Note that women in the estrogen-only arm had a greater incidence of hypertension and diabetes mellitus—known risk factors for stroke—than did women in the estrogen-progestin arm.^16,36

VTE risk is twice as high in HRT users

Postmenopausal women who take estrogen have a higher risk of venous thromboembolism (VTE) than those who do not. This risk translated into a relative hazard of 2.06 (1.57–2.70) in the WHI estrogen-progestin arm, or an attributable risk of 3.6 cases per 1,000 women, compared with 1.8 cases per thousand in the control group.³⁶

The absolute increased risk is 1.8 cases per 1,000 women, or, as expressed in the study itself, 18 cases per 10,000 women per year.

I have deliberately reduced the attributable risk to the number of cases per thousand because I believe this number is more easily understood by the patient and accurately demonstrates the low risk.

In the estrogen-only arm of the WHI, the hazard ratio for VTE was 1.33 (0.99–1.79), or an absolute increased risk of 0.7 cases per thousand—although this finding was not significant. The attributable risk was 2.7 cases per 1,000 women, compared with 2.0 cases per thousand among controls.¹⁶

Like stroke, the risk of VTE may be confounded by other factors besides use of exogenous estrogen.

No cause and effect for HRT and breast cancer

Nothing frightens women as much as breast cancer, and articles focusing on the relationship between breast cancer and HRT have drawn widespread attention. However, despite voluminous literature, the etiology of breast cancer remains elusive—and there is no evidence that either estrogen or progestins cause the disease.^45,46 Rather, there is only an association between the use of estrogen, progestin, and breast cancer. Linking the finding of an increased risk with an implication of causality would be inappropriate.

Breast cancer risk with HRT is not consistently elevated, in studies

In fact, a qualitative review of observational studies from 1975 to 2000 found no significant increase or decrease in the risk of breast cancer with estrogen or estrogen-progestin in 80% of the reports.⁴⁷

Risk factors for breast cancer (TABLE 1) include family history, obesity, late childbirth, and hormone therapy—but obesity and family history have higher relative risks than the use of HRT.⁴⁸

TABLE 1

Relative risk of breast cancer

CHARACTERISTIC	RELATIVE RISK
2 family members with breast cancer	14
1 family member with breast cancer	2.2
Obesity	1.8
Young age at menarche	1.6
Hormone therapy	1.3
>30 years of age at birth of first child	1.3
Menopause	0.7

WHI arms find different risks

In the widely publicized WHI, women in the estrogen-progestin arm had an overall relative hazard for breast cancer of 1.24 (95% CI 1.01–1.54), but there was no increased risk in women who had never before used hormones.³⁶ Women who had previously used hormones for 5 years or more did have an increased risk.³⁶ The incidence of breast cancer in the study population was 3 cases per 1,000 women, and the excess number was 0.7 more cases with the use of estrogen-progestin (TABLE 2).

Conversely, in the estrogen-only arm of the WHI,¹⁶ the relative hazard for breast cancer was 0.77 (95% CI 0.59–1.01), and the reduction in risk was almost statistically significant. There are at least 2 potential explanations for the lower incidence of breast cancer in this arm:

• Without a progestin, estrogen increases breast density only minimally, allowing for easier mammographic interpretation.
  
• Women susceptible to breast cancer because of their previous use of estrogen may not have been present in the at-risk population in sufficient numbers to cause an increase.
  

Neither explanation—separate or combined—fully explains the lowered risk in this population. Each population studied appears to have a different level of risk based on multiple factors that cannot be controlled completely in clinical trials and observational studies.

TABLE 2

Extra cases of breast cancer, by risk factor

RISK FACTOR	BREAST CANCERS DIAGNOSED OVER 20 YEARS FROM AGES 50 TO 70 (PER 1,000)	EXTRA BREAST CANCERS (PER 1,000)
Never used HRT	45	-
>5 years HRT	47	2
>10 years HRT	51	6
>15 years HRT	57	12
Late menopause (age 60)	59	14
Alcohol (2 drinks/day)	72	27
No daily exercise	72	27
Weight gain (>20 kg)	90	45
Reprinted from THE LANCET, Vol. 350: 1047–1059, Collaborative Group on Hormonal Factors in Breast Cancer, Breast cancer and hormone replacement therapy: collaborative reanalysis of data from 51 epidemiological studies of 52,705 women with breast cancer and 108,411 women without breast cancer. Copyright 1997, with permission from Elsevier

 

HRT may promote, rather than induce, breast cancer

The role of hormones in the etiology of breast cancer is difficult to assess. The Million Women Study⁴⁹ found that the elevated risk of breast cancer disappeared within 1 year of stopping HRT. This finding implies that hormones may be a promoter, rather than inducer, of neoplasms in the breast.

Breast cancer may be present in many women, but apparent in few

When autopsies were performed on women in their 40s who had died from other diseases, the incidence of breast cancer was 39%, but the clinical detection rate was only 1% for this population.⁵⁰ This discrepancy suggests that neoplastic cells may be present in the body at any time, but become clinically apparent only under certain conditions.⁵¹

More recent data suggest that undifferentiated stem cells in the breast become dysfunctional and result in cancer.⁵² This theory is supported by the various histologic types of cancer found in the breast.

A weak link

Although it may be compelling to link hormone use with breast cancer, the association is weak and the incidence is lower than in other known relationships such as obesity. At present, the cause of breast neoplasia appears to be multifactorial.

Now that the dust is settling from the Women’s Health Initiative (WHI), our patients are again asking reasonable questions about hormone replacement therapy (HRT). I remind them of estrogen’s proven advantages in menopause, as well as its risks. Although most women are generally aware of these risks and benefits, considerable misunderstanding persists. This article reviews what the evidence to date does and does not confirm, particularly regarding breast cancer and coronary heart disease, where most of the uncertainty remains.

HRT stops vaginal atrophy, hot flashes, and bone loss

Three applications form the basis for HRT in postmenopausal women:

1. Hot flashes subside. Hot flashes occur with varying intensity in about 85% of women, and are effectively treated with estrogen, whether given orally, transdermally, or vaginally.^1,2 As long as an appropriate blood level of the hormone is reached, hot flashes diminish.^3-5 This reduction is dose-related.
   
2. Measurable improvements in vaginal atrophy. Estrogen’s efficacy in relieving dryness, itching, burning, and dyspareunia is well demonstrated, regardless of the route of administration.^3,6,7 A fall in vaginal pH from 6.0 to 5.0 after estrogen administration has been documented,⁸ as has the increase in the number of superficial cells of the vagina with exogenous estrogen.⁹
   
3. HRT maintains or increases bone mineral density (BMD). Most estrogen preparations on the US market have been shown to improve BMD.^10-15 “Improvement” means no significant loss, or an increase, in BMD. In the WHI, both vertebral and nonvertebral fractures diminished unequivocally in women using estrogen—alone or with a progestin.^16,17 Other clinical trials also have shown increased BMD, as well as decreased urinary and serum markers of bone turnover.
   

Do new data link progestin to cancer?

Although compelling evidence supports the use of progestational agents in addition to estrogen to prevent endometrial hyperplasia and endometrial cancer,¹⁸ a 2005 report¹⁹ suggests that chronic, long-term use of estrogen with a progestin may increase the risk of endometrial carcinoma. Because this is the only study in which this risk has been found, corroboration is required.

Until then, give progestin at a sufficient dose and duration to inhibit endometrial hyperplasia.^20-25

Effects on heart disease may be age-related

With notable exceptions, the overall conclusion of clinical trials and observational studies to date is that estrogen helps prevent coronary heart disease (CHD).^26-30 This finding was first observed in the late 1980s with evidence that estrogen increases high-density lipoprotein (HDL) cholesterol and reduces total and low-density lipoprotein (LDL) cholesterol.³¹

Some experts argue that these observational trials are biased because many of the women taking estrogen had modified their lifestyles to maintain their weight, control their diet, and exercise regularly.³² Indeed, the randomized, placebo-controlled Heart and Estrogen Replacement Study (HERS) and both arms of the WHI trial found no evidence for a significant increase or decrease in CHD events.^33-35

Time from menopause to HRT may be key

Both the HERS and WHI trials enrolled older women who had entered menopause a few months to several years before starting HRT.³⁶ In addition, the estrogen-progestin arm of the WHI trial lacked sufficient power to detect a significant difference in CHD outcomes.³⁷

The WHI findings contrast those of the large, ongoing, observational Nurses Health Study, which has shown a consistent decrease in CHD incidence in women who began HRT with the onset of menopausal symptoms.^27-30 The most recent data suggest that the interval between menopause and the start of HRT may explain the different findings in randomized, controlled trials and observational studies.³⁸ The WHI data support this theory: CHD was lower in women who began taking HRT within 5 years of menopause, compared with women who initiated HRT more than 5 years afterward.³⁶ In addition, data from the estrogen-only arm of the WHI show fewer CHD events in women younger than 60.³⁴

Several other studies support this hypothesis:

• The surgically postmenopausal cynomolgus macaque had a lower rate of atherosclerotic plaque development when estrogen was given, with or without a progestin.^39,40
  
• In the Rancho Bernardo study, women who had used HRT had less cardiac calcification documented by computed tomography, compared with nonusers.⁴¹
  
• Estrogen has been shown, by measurement of carotid intimal medial thickness, to inhibit atherosclerotic plaque in humans.⁴²
  
• Older women with established atherosclerosis do not undergo any significant change in plaque size with the use of exogenous estrogen.⁴³
  

 

Although these findings support the use of estrogen or estrogen-progestin early after menopause as a way of preventing CHD, further clinical trials are needed.⁴⁴

Stroke risk is small but real

Both arms of the WHI found an increased incidence of stroke in women using hormones, compared with nonusers.^16,36 The exact mechanisms underlying this increased risk are unclear.

The actual attributable risk was an increase of 0.7 cases of stroke per 1,000 women per year over placebo in the estrogen-progestin arm,³⁶ and 1.2 cases per 1,000 in the estrogen-only arm.¹⁶ The relative hazards were 1.31 (95% confidence interval [CI] 1.02–1.68) and 1.30 (95% CI 1.10–1.77), respectively.

Note that women in the estrogen-only arm had a greater incidence of hypertension and diabetes mellitus—known risk factors for stroke—than did women in the estrogen-progestin arm.^16,36

VTE risk is twice as high in HRT users

Postmenopausal women who take estrogen have a higher risk of venous thromboembolism (VTE) than those who do not. This risk translated into a relative hazard of 2.06 (1.57–2.70) in the WHI estrogen-progestin arm, or an attributable risk of 3.6 cases per 1,000 women, compared with 1.8 cases per thousand in the control group.³⁶

The absolute increased risk is 1.8 cases per 1,000 women, or, as expressed in the study itself, 18 cases per 10,000 women per year.

I have deliberately reduced the attributable risk to the number of cases per thousand because I believe this number is more easily understood by the patient and accurately demonstrates the low risk.

In the estrogen-only arm of the WHI, the hazard ratio for VTE was 1.33 (0.99–1.79), or an absolute increased risk of 0.7 cases per thousand—although this finding was not significant. The attributable risk was 2.7 cases per 1,000 women, compared with 2.0 cases per thousand among controls.¹⁶

Like stroke, the risk of VTE may be confounded by other factors besides use of exogenous estrogen.

No cause and effect for HRT and breast cancer

Nothing frightens women as much as breast cancer, and articles focusing on the relationship between breast cancer and HRT have drawn widespread attention. However, despite voluminous literature, the etiology of breast cancer remains elusive—and there is no evidence that either estrogen or progestins cause the disease.^45,46 Rather, there is only an association between the use of estrogen, progestin, and breast cancer. Linking the finding of an increased risk with an implication of causality would be inappropriate.

Breast cancer risk with HRT is not consistently elevated, in studies

In fact, a qualitative review of observational studies from 1975 to 2000 found no significant increase or decrease in the risk of breast cancer with estrogen or estrogen-progestin in 80% of the reports.⁴⁷

Risk factors for breast cancer (TABLE 1) include family history, obesity, late childbirth, and hormone therapy—but obesity and family history have higher relative risks than the use of HRT.⁴⁸

TABLE 1

Relative risk of breast cancer

CHARACTERISTIC	RELATIVE RISK
2 family members with breast cancer	14
1 family member with breast cancer	2.2
Obesity	1.8
Young age at menarche	1.6
Hormone therapy	1.3
>30 years of age at birth of first child	1.3
Menopause	0.7

WHI arms find different risks

In the widely publicized WHI, women in the estrogen-progestin arm had an overall relative hazard for breast cancer of 1.24 (95% CI 1.01–1.54), but there was no increased risk in women who had never before used hormones.³⁶ Women who had previously used hormones for 5 years or more did have an increased risk.³⁶ The incidence of breast cancer in the study population was 3 cases per 1,000 women, and the excess number was 0.7 more cases with the use of estrogen-progestin (TABLE 2).

Conversely, in the estrogen-only arm of the WHI,¹⁶ the relative hazard for breast cancer was 0.77 (95% CI 0.59–1.01), and the reduction in risk was almost statistically significant. There are at least 2 potential explanations for the lower incidence of breast cancer in this arm:

• Without a progestin, estrogen increases breast density only minimally, allowing for easier mammographic interpretation.
  
• Women susceptible to breast cancer because of their previous use of estrogen may not have been present in the at-risk population in sufficient numbers to cause an increase.
  

Neither explanation—separate or combined—fully explains the lowered risk in this population. Each population studied appears to have a different level of risk based on multiple factors that cannot be controlled completely in clinical trials and observational studies.

TABLE 2

Extra cases of breast cancer, by risk factor

RISK FACTOR	BREAST CANCERS DIAGNOSED OVER 20 YEARS FROM AGES 50 TO 70 (PER 1,000)	EXTRA BREAST CANCERS (PER 1,000)
Never used HRT	45	-
>5 years HRT	47	2
>10 years HRT	51	6
>15 years HRT	57	12
Late menopause (age 60)	59	14
Alcohol (2 drinks/day)	72	27
No daily exercise	72	27
Weight gain (>20 kg)	90	45
Reprinted from THE LANCET, Vol. 350: 1047–1059, Collaborative Group on Hormonal Factors in Breast Cancer, Breast cancer and hormone replacement therapy: collaborative reanalysis of data from 51 epidemiological studies of 52,705 women with breast cancer and 108,411 women without breast cancer. Copyright 1997, with permission from Elsevier

 

HRT may promote, rather than induce, breast cancer

The role of hormones in the etiology of breast cancer is difficult to assess. The Million Women Study⁴⁹ found that the elevated risk of breast cancer disappeared within 1 year of stopping HRT. This finding implies that hormones may be a promoter, rather than inducer, of neoplasms in the breast.

Breast cancer may be present in many women, but apparent in few

When autopsies were performed on women in their 40s who had died from other diseases, the incidence of breast cancer was 39%, but the clinical detection rate was only 1% for this population.⁵⁰ This discrepancy suggests that neoplastic cells may be present in the body at any time, but become clinically apparent only under certain conditions.⁵¹

More recent data suggest that undifferentiated stem cells in the breast become dysfunctional and result in cancer.⁵² This theory is supported by the various histologic types of cancer found in the breast.

A weak link

Although it may be compelling to link hormone use with breast cancer, the association is weak and the incidence is lower than in other known relationships such as obesity. At present, the cause of breast neoplasia appears to be multifactorial.

Although the Women’s Health Initiative¹ discouraged many menopausal women from using oral estrogen, it failed to address the risks of treatment with other dosages or forms of estrogen and progestin.

Nor has any other trial of similar size and complexity taken up the issue. However, numerous smaller studies have been published.

This article summarizes findings on percutaneous delivery of estradiol gel (EstroGel), the most recently FDA-approved estrogen option for treatment of menopausal symptoms.

It describes the overall safety of estradiol gel, as well as its effectson:

• menopausal symptoms,
  
• bone,
  
• metabolism, and
  
• endometrium.
  

(In this article, “percutaneous delivery” refers to estradiol gel applied to the skin, and “transdermal estradiol” indicates delivery via a transdermal reservoir or matrix system, otherwise known as “the patch.” I have used an arbitrary definition to distinguish the gel from other methods of delivering estradiol across the skin. For example, Estrasorb is a liposomal formulation that is applied to the skin of the thigh. Although it is a percutaneous estradiol similar to the gel, this article focuses only on the latter option.)

Easy to apply, few skin reactions

Percutaneous estradiol gel formulations have been available for almost 30 years in Europe, where they are utilized by a majority of women on hormone therapy. In the United States, the hydroalcoholic gel is packaged in a pump that delivers 64 standardized 1.25-g doses, which contain 0.75 mg of 17ß-estradiol. Once it is applied, the gel is absorbed into an intradermal reservoir (FIGURE) and dries in 2 to 5 minutes, leaving no residue.

Patient selection. Estradiol gel is well-suited for patients who are concerned about the risks of oral estrogen (as portrayed in the mainstream press following the Women’s Health Initiative) and want to avoid that route of administration, as well as women who dislike or have difficulty swallowing pills. Percutaneous administration also is appropriate for physically active women who may have adhesion problems or skin irritation with the transdermal patch, or those who have had reactions to local adhesives in the past. The patient should be motivated to apply the gel on a daily basis.

Indications are moderate to severe vasomotor symptoms in menopausal women, and moderate to severe symptoms of vulvar and vaginal atrophy, although topical vaginal products should also be considered for the latter indication.

Contraindications are undiagnosed abnormal vaginal bleeding; history of breast cancer, other estrogen-dependent malignancy, stroke, heart attack, or liver disease or dysfunction; active thrombophlebitis or thromboembolic disorders (or a history of these); and known or suspected pregnancy.

Common side effects include headache, breast pain, irregular vaginal bleeding or spotting, stomach cramps or bloating, nausea and vomiting, and hair loss.

• Skin reactions are infrequent, but should be taken into account when discussing per-cutaneous or transdermal delivery of any drug. However, estradiol gel appears to cause fewer skin reactions than the patch. In a study over more than 5 years, 0 of 157 women treated with percutaneous estradiol reported skin irritation.² Other comparisons found similar outcomes.^3,4
  

After the gel dries, other lotions or per-fumes can be applied to the site, if desired, and the woman can cover the site with clothing.

Onset of action is rapid, as it is with the transdermal patch.⁵ Dose variability is minimized when the gel is applied at the same time every day to a large area of skin, preferably the arm, although all application sites appear to produce similar results: abdomen, shoulders, arms, and inner thigh.⁶ The gel should not be applied to the breast or vagina.

Diminished effect with skin washing. In 1 trial, site washing 1/2 hour after application significantly decreased bioavailability and time to reach peak plasma concentrations.⁷ For this reason, the gel should be not applied before a bath, shower, or sauna.

Dosing options. The initial dose is 1.25 mg of gel, which is 1 pump of the bottle. The gel is collected in the palm of 1 hand and applied to the skin of the opposite arm from the wrist to the shoulder. The dose can be titrated by adding a second pump of the gel and applying it to the opposite arm. The dose can be lowered by using less than a full depression of the pump.

 

Stable, physiologic estrogen levels

Estradiol gel produces relatively stable serum estradiol levels, and therapeutic estradiol levels similar to those seen with other formulations, routes of administration, and dosages.

Percutaneous administration produces serum estrone to estradiol ratios close to 1, in contrast to higher ratios (5:1) with oral administration.^8-11 The lower ratio approximates levels during the menstrual cycle of premenopausal women. (TABLE 1 gives estradiol and estrone levels from different studies following administration of estradiol as a gel, oral formulation, and transdermal patch.)

The percutaneous route also allows for delivery directly to the systemic circulation, avoiding gastrointestinal and first-pass hepatic metabolism and elimination. In contrast, oral micronized estradiol causes large fluctuations in serum estradiol and estrone levels due to absorption and metabolism.⁸

More stable serum levels than with the patch. One study¹² comparing percutaneous and transdermal estradiol found similar interindividual variability but less stable serum levels in women using the transdermal system. A separate study¹³ also reported greater fluctuation of serum estradiol levels in women using a transdermal system than in those using the gel.

TABLE 1

Serum estradiol and estrone levels for various routes of estradiol

AUTHOR	THERAPY	ESTRADIOL LEVEL (PG/ML)	ESTRONE LEVEL (PG/ML)
Scott et al8	E2gel 3 mg/day	102.9±39.9	120.0±50.5
	E2gel 1.5 mg/day	68.1±27.4	90.6±45.7
	Transdermal estradiol 50 μg/day	41.1±13.5	45.0±15.9
	Oral micronized estradiol 2 mg/day	114.0±65.2	575.2±279.9
Palacios et al10	E2gel 1.5 mg/day	75.7±1.0	58.5±2.9
	Oral conjugated estrogen 0.625 mg/day	39.6±4.6	126.0±7.6
Basdevant et al11	E2gel 3 mg/day	221.0±35.0	146.0±19.0
	Oral micronized estradiol 2 mg/day	121.0±27.0	811.0±370.0
Archer15	E2gel 0.75 mg/day	33.5*	49.0*
	E2gel 1.5 mg/day	65.0*	58.0*
	Placebo	5.0*	23.0*
*Median value
E2gel = percutaneous estradiol gel

Studies compared relief of menopausal symptoms

Estradiol gel effectively relieved meno-pausal symptoms in randomized, double-blind studies, open label comparisons, and observational trials in postmenopausal women, with and without the addition of various progestins.

Symptom relief in comparison with baseline values is statistically significant as early as 2 weeks after initiating treatment. Relief of up to 2 years’ duration has been reported.^3,14

Several studies have compared symptom relief achieved with estradiol gel, oral estrogen, transdermal delivery systems, and placebo^3,9,14-18; findings are summarized in TABLE 2.

Same efficacy when progestogen is added. The following studies, and other studies,¹⁴ demonstrated that estradiol gel relieves menopausal symptoms whether it is administered alone or in combination with a progestogen, with efficacy similar to other estrogen formulations:

Climacteric symptoms decreased to the same extent when estradiol gel was combined with a levonorgestrel-releasing intrauterine device, oral micronized progesterone, or vaginal micronized progesterone.¹⁹

A study²⁰ that added lynestrenol decreased the frequency of hot flushes and night sweats more than in women using estradiol gel alone. However, negative mood symptoms were more pronounced in the progestin-treated group.

Estradiol gel 1 mg/day in combination with monthly or quarterly oral medroxyprog-esterone acetate reduced the severity of hot flushes, sweating, and vaginal dryness, according to a 12-month trial.²¹

Symptoms decreased the same whether a levonorgestrel-releasing IUD or oral or vaginal progesterone was added.

Bone mineral density maintained or increased

Several randomized, controlled trials have documented the effects of estradiol gel on bone mineral density (BMD) and various markers of bone metabolism. In these studies, BMD remained steady^22,23 or increased^10,24,25 following treatment, and estradiol gel remained effective for up to 4 years.²⁵ Estradiol gel maintained or increased BMD with or without addition of progestins.^22,23,26

These investigations involved measurement of BMD at the lumbar spine, forearms, or hip, as well as biologic markers of bone turnover such as urinary hydroxyproline/creatinine ratio, serum alkaline phosphatase, and serum osteocalcin.

Serum estradiol and skeletal uptake of a bone-seeking agent also were determined. Estradiol gel regimens ranged from 0.75 mg/day to 3 mg/day, and populations included both surgical and natural menopausal subjects in several countries.

Effects comparable to oral estrogen. Compared with oral conjugated estrogen, which increased BMD at the lumbar spine by 4.3% (±3.2%), estradiol gel produced increases of 5.6% (±2.9%) at 24 months and 4.7% (±3.2%) at 36 months.¹⁰

 

Minimum level of protection achieved. Following a comparison of oral and percutaneous estradiol, Reginster et al²⁷ suggested that a minimum estradiol level of 60 pg/mL is necessary to prevent postmenopausal bone loss. Mean serum estradiol levels in women receiving 1.5 mg/day estradiol gel were 75.7 pg/mL and 78.4 pg/mL at 24 and 36 months, respectively, in a study by Palacios et al,¹⁰ and 85.8 pg/mL in a study by Devogelaer and colleagues.²⁴ In these trials, BMD increased, and it remained steady in other investigations.^28,29

More recent trials suggest that lower serum estradiol levels secondary to smaller estrogen doses have the capacity to maintain BMD.^30,31 A 1.9% mean increase of BMD at the lumbar spine was reported in women with a mean serum concentration of 17 pg/mL in a 2-year study³⁰ of a transdermal estradiol delivery system. The percutaneous route does not appear to limit these beneficial effects.

How are metabolic factors affected?

In general, oral estrogens produce beneficial changes in lipid metabolism, particularly higher levels of high-density lipoprotein (HDL) cholesterol. However, they also elevate triglyceride and glucose levels. How this plays out clinically is unclear. Both the Women’s Health Initiative and the Heart and Estrogen/Progestin Replacement Study (HERS) found no cardioprotective effects of oral estrogen despite increases in HDL and decreases in low-density lipoprotein (LDL) levels.^1,32

Oral versus percutaneous estradiol. No long-term studies of this magnitude have investigated the effect of percutaneous estradiol on coronary artery disease, although numerous clinical trials have shown that the route of estrogen delivery affects many of the metabolic variables used to estimate the risk of negative cardiac outcomes. It remains to be determined whether percutaneous estradiol affects these metabolic factors in ways that influence morbidity and mortality.

In 1 trial,³³ oral conjugated estrogen led to the following significant changes in lipids:

• increase in very low density lipoprotein (VLDL) cholesterol,
  
• decrease in LDL cholesterol (but not the LDL apoprotein B),
  
• increase in HDL and apoprotein A1, and
  
• significant increase in HDL2 cholesterol.
  

In contrast, percutaneous administration increased only HDL2 and the triglyceride and cholesterol content of the whole HDL fraction.

In a separate study,⁹ oral conjugated estrogen had a 2.5-fold increase in serum angiotensin, and percutaneous estradiol gel, no effect.

Same effects when progestin is added. Beneficial effects were not diminished when oral micronized progesterone was added to percutaneous estradiol.³⁴ Estradiol gel significantly reduced total serum cholesterol and LDL cholesterol in the first year of treatment, compared with placebo.

Coagulation effects. Percutaneous estradiol has fewer negative effects on coagulation factors than its oral counterpart. One study³⁵ investigating combination therapy with micronized progesterone compared the effects of percutaneous estradiol and oral estradiol valerate. The group receiving percutaneous estradiol gel/micronized progesterone had no significant changes in plasminogen activator inhibitor, tissue-type plasminogen concentration, and global fibrinolytic capacity. The other group had a significant decrease in mean tissue-type plasminogen concentration and plasminogen activator inhibitor activity and a significant rise in global fibrinolytic capacity.

The oral estradiol—but not the percutaneous formulation—significantly increased the mean value of prothrombin activation peptide and decreased mean antithrombin activity compared with no treatment.

Poor glycemic control may increase the risk of cardiovascular disease, but oral and percutaneous estradiol appear to have similar glycemic effects. A comparative trial³⁶ of oral estradiol valerate 2 mg/day and percutaneous estradiol 1 mg/day found no differences in glycosylated hemoglobin A1c levels (declined in both groups) or in fasting and 2-hour post-prandial blood glucose levels (constant in both groups) or insulin sensitivity.

Treatment duration may also influence how percutaneous estradiol affects metabolic factors. An open-label longitudinal prospective study³⁷ of 30 women receiving estradiol gel 1.5 mg/day for 6 months found a significant decrease in lipoprotein (a), apoprotein A-I, apoprotein B, HDL cholesterol, and HDL3 cholesterol. At 1 year, however, these changes were not significant.

No association with venous thromboembolism. Transdermal estradiol administered as a patch or percutaneous gel had no effect on the risk of venous thromboembolism in a multicenter case-control investigation.³⁸

In contrast, a recent retrospective study found a risk of venous thromboembolism that was at least 4 times greater with oral estrogen than with transdermal estradiol.³⁸

Use a progestogen to prevent endometrial hyperplasia

Like other forms of estrogen, percutaneous estradiol stimulates the endometrium. For this reason, women who have an intact uterus should use a progestogen in an adequate dose to prevent hyperplastic changes.³⁹ Although numerous regimens appear to be effective, the optimal route, dose, and duration of progestin in women using percutaneous estrogen remain to be determined.

Percutaneous estradiol versus other routes of administration. Endometrial thickness and amenorrhea rates over 6 months were not statistically different among 54 women treated with estradiol gel 1.5 mg/day, transdermal estradiol 50 μg/day, or oral estradiol valerate 2 mg/day—all in combination with nomegestrol acetate 2.5 mg/day.⁴⁰ The overall rates of no bleeding or spotting over 6 cycles of treatment were 78% in the percutaneous estradiol group, 48% in the transdermal group, and 60% in the oral group.

 

Although the percutaneous estradiol group had a lower overall incidence of no bleeding or spotting than the other groups, the difference was not significant. Nor were there significant differences in the variation of endometrial thickness from baseline: A mean increase of 1.5 mm (±0.4 mm) was reported in the percutaneous group, compared with 1.5 mm (±0.7 mm) and 1.7 mm (±0.6 mm) in the transdermal and oral estrogen groups, respectively.

Estrogenic and progestogenic effects were similar for transdermal and percutaneous estradiol (with dydrogesterone 10 mg/day for days 1 to 12) after a baseline atropic endometrium was identified.³

Estradiol gel produces stable, physiologic serum estradiol levels and has a serum estrone to estradiol ratio close to 1.

The most effective progestogen dosage and duration are unknown, although many regimens have been studied:

• Percutaneous estradiol 1.5 mg/day for the first 24 days of the month in combination with nomegestrol acetate 5 mg/day for days 11 to 24: Over 6 months, researchers found a secretory pattern in the majority of women and no evidence of hyperplasia.¹⁸
  
• Percutaneous estradiol 3 mg/day for 3 of every 4 weeks in combination with nomegestrol acetate 5 mg/day for 10 days: No reduction in estrogenic endometrial effects.^41,42
  
• Estradiol gel 3 mg/day for 3 of every 4 weeks in combination with 200 mg or 300 mg oral micronized progesterone for the last 10 days of treatment: Dose was too low or treatment too short to produce a complete secretory transformation of the endometrium.⁴³ However, the Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial⁴⁴ found no increased occurrence of hyperplasia in women using micronized progesterone 200 mg for 12 days of each cycle for more than 3 years in combination with oral conjugated estrogen 0.625 mg.
  
• Percutaneous estradiol 1.5 mg/day combined with micronized progesterone 100 mg daily (orally or vaginally) for the first 25 days of the month: Fully inhibited mitoses and induced amenorrhea in most of the women studied, with amenorrhea rates of 93.3% at 3 months and 91.6% at 6 months.⁴⁵
  
• Estradiol gel 1.5 mg/day with 100 mg vaginal micronized progesterone for 21 days per cycle: Stable endometrial thickness and no endometrial hyperplasia over 12 months.⁴⁶ However, breakthrough bleeding was reported in up to 30% of subjects, principally in the second 3 months of treatment.
  
• Percutaneous estradiol 1 mg/day for 3 months with oral medroxyprogesterone acetate 20 mg/day for the last 14 days, followed by a 7-day free interval: No reports of endometrial hyperplasia or suspect changes at 12 and 24 months.¹⁷
  
• Estradiol gel 2 mg/day for 21 days with 10 mg oral medroxyprogesterone acetate for the last 14 days, followed by a 7-day free interval: No endometrial hyperplasia or suspect changes at 12 and 24 months.¹⁷
  
• Estradiol gel 1 mg/day with medroxyprogesterone acetate 10 mg on days 1-12 every month or every 3 months: Endometrial hyperplasia was found in 1 woman (0.3%) in the group receiving the progestogen every 3 months. Endometrial histology did not differ between women taking medroxyprogesterone monthly and those taking it every 3 months.²¹
  
• Estradiol gel 3 mg/day with oral micronized progesterone 200 mg for 12 days of each cycle: Regular withdrawal bleeding in 70% of women.³⁴
  
• Percutaneous estradiol 1.5 mg/day in combination with a levonorgestrel-releasing intrauterine device: 80% of women were amenorrheic at 1 year, with a mean endometrial thickness 3 mm; at 5 years, 100% of women had epithelial atrophy.⁴⁷
  
• Estradiol gel 1.5 mg for 21 days with 200 mg oral progesterone for 14 days (126 women); 3 mg percutaneous estradiol for 21 days with 300 mg oral progesterone for 10 days (23 women); 1.5 mg estradiol gel with 300 mg oral progesterone (3 women); or 3 mg estradiol gel for 28 days with 200 mg progesterone for 14 days (5 women): No evidence of hyperplasia after 5 years of treatment.²
  

Dr. Archer is a consultant to Solvay, and has had grant support from Solvay for clinical trials.

Although the Women’s Health Initiative¹ discouraged many menopausal women from using oral estrogen, it failed to address the risks of treatment with other dosages or forms of estrogen and progestin.

Nor has any other trial of similar size and complexity taken up the issue. However, numerous smaller studies have been published.

This article summarizes findings on percutaneous delivery of estradiol gel (EstroGel), the most recently FDA-approved estrogen option for treatment of menopausal symptoms.

It describes the overall safety of estradiol gel, as well as its effectson:

• menopausal symptoms,
  
• bone,
  
• metabolism, and
  
• endometrium.
  

(In this article, “percutaneous delivery” refers to estradiol gel applied to the skin, and “transdermal estradiol” indicates delivery via a transdermal reservoir or matrix system, otherwise known as “the patch.” I have used an arbitrary definition to distinguish the gel from other methods of delivering estradiol across the skin. For example, Estrasorb is a liposomal formulation that is applied to the skin of the thigh. Although it is a percutaneous estradiol similar to the gel, this article focuses only on the latter option.)

Easy to apply, few skin reactions

Percutaneous estradiol gel formulations have been available for almost 30 years in Europe, where they are utilized by a majority of women on hormone therapy. In the United States, the hydroalcoholic gel is packaged in a pump that delivers 64 standardized 1.25-g doses, which contain 0.75 mg of 17ß-estradiol. Once it is applied, the gel is absorbed into an intradermal reservoir (FIGURE) and dries in 2 to 5 minutes, leaving no residue.

Patient selection. Estradiol gel is well-suited for patients who are concerned about the risks of oral estrogen (as portrayed in the mainstream press following the Women’s Health Initiative) and want to avoid that route of administration, as well as women who dislike or have difficulty swallowing pills. Percutaneous administration also is appropriate for physically active women who may have adhesion problems or skin irritation with the transdermal patch, or those who have had reactions to local adhesives in the past. The patient should be motivated to apply the gel on a daily basis.

Indications are moderate to severe vasomotor symptoms in menopausal women, and moderate to severe symptoms of vulvar and vaginal atrophy, although topical vaginal products should also be considered for the latter indication.

Contraindications are undiagnosed abnormal vaginal bleeding; history of breast cancer, other estrogen-dependent malignancy, stroke, heart attack, or liver disease or dysfunction; active thrombophlebitis or thromboembolic disorders (or a history of these); and known or suspected pregnancy.

Common side effects include headache, breast pain, irregular vaginal bleeding or spotting, stomach cramps or bloating, nausea and vomiting, and hair loss.

• Skin reactions are infrequent, but should be taken into account when discussing per-cutaneous or transdermal delivery of any drug. However, estradiol gel appears to cause fewer skin reactions than the patch. In a study over more than 5 years, 0 of 157 women treated with percutaneous estradiol reported skin irritation.² Other comparisons found similar outcomes.^3,4
  

After the gel dries, other lotions or per-fumes can be applied to the site, if desired, and the woman can cover the site with clothing.

Onset of action is rapid, as it is with the transdermal patch.⁵ Dose variability is minimized when the gel is applied at the same time every day to a large area of skin, preferably the arm, although all application sites appear to produce similar results: abdomen, shoulders, arms, and inner thigh.⁶ The gel should not be applied to the breast or vagina.

Diminished effect with skin washing. In 1 trial, site washing 1/2 hour after application significantly decreased bioavailability and time to reach peak plasma concentrations.⁷ For this reason, the gel should be not applied before a bath, shower, or sauna.

Dosing options. The initial dose is 1.25 mg of gel, which is 1 pump of the bottle. The gel is collected in the palm of 1 hand and applied to the skin of the opposite arm from the wrist to the shoulder. The dose can be titrated by adding a second pump of the gel and applying it to the opposite arm. The dose can be lowered by using less than a full depression of the pump.

 

Stable, physiologic estrogen levels

Estradiol gel produces relatively stable serum estradiol levels, and therapeutic estradiol levels similar to those seen with other formulations, routes of administration, and dosages.

Percutaneous administration produces serum estrone to estradiol ratios close to 1, in contrast to higher ratios (5:1) with oral administration.^8-11 The lower ratio approximates levels during the menstrual cycle of premenopausal women. (TABLE 1 gives estradiol and estrone levels from different studies following administration of estradiol as a gel, oral formulation, and transdermal patch.)

The percutaneous route also allows for delivery directly to the systemic circulation, avoiding gastrointestinal and first-pass hepatic metabolism and elimination. In contrast, oral micronized estradiol causes large fluctuations in serum estradiol and estrone levels due to absorption and metabolism.⁸

More stable serum levels than with the patch. One study¹² comparing percutaneous and transdermal estradiol found similar interindividual variability but less stable serum levels in women using the transdermal system. A separate study¹³ also reported greater fluctuation of serum estradiol levels in women using a transdermal system than in those using the gel.

TABLE 1

Serum estradiol and estrone levels for various routes of estradiol

AUTHOR	THERAPY	ESTRADIOL LEVEL (PG/ML)	ESTRONE LEVEL (PG/ML)
Scott et al8	E2gel 3 mg/day	102.9±39.9	120.0±50.5
	E2gel 1.5 mg/day	68.1±27.4	90.6±45.7
	Transdermal estradiol 50 μg/day	41.1±13.5	45.0±15.9
	Oral micronized estradiol 2 mg/day	114.0±65.2	575.2±279.9
Palacios et al10	E2gel 1.5 mg/day	75.7±1.0	58.5±2.9
	Oral conjugated estrogen 0.625 mg/day	39.6±4.6	126.0±7.6
Basdevant et al11	E2gel 3 mg/day	221.0±35.0	146.0±19.0
	Oral micronized estradiol 2 mg/day	121.0±27.0	811.0±370.0
Archer15	E2gel 0.75 mg/day	33.5*	49.0*
	E2gel 1.5 mg/day	65.0*	58.0*
	Placebo	5.0*	23.0*
*Median value
E2gel = percutaneous estradiol gel

Studies compared relief of menopausal symptoms

Estradiol gel effectively relieved meno-pausal symptoms in randomized, double-blind studies, open label comparisons, and observational trials in postmenopausal women, with and without the addition of various progestins.

Symptom relief in comparison with baseline values is statistically significant as early as 2 weeks after initiating treatment. Relief of up to 2 years’ duration has been reported.^3,14

Several studies have compared symptom relief achieved with estradiol gel, oral estrogen, transdermal delivery systems, and placebo^3,9,14-18; findings are summarized in TABLE 2.

Same efficacy when progestogen is added. The following studies, and other studies,¹⁴ demonstrated that estradiol gel relieves menopausal symptoms whether it is administered alone or in combination with a progestogen, with efficacy similar to other estrogen formulations:

Climacteric symptoms decreased to the same extent when estradiol gel was combined with a levonorgestrel-releasing intrauterine device, oral micronized progesterone, or vaginal micronized progesterone.¹⁹

A study²⁰ that added lynestrenol decreased the frequency of hot flushes and night sweats more than in women using estradiol gel alone. However, negative mood symptoms were more pronounced in the progestin-treated group.

Estradiol gel 1 mg/day in combination with monthly or quarterly oral medroxyprog-esterone acetate reduced the severity of hot flushes, sweating, and vaginal dryness, according to a 12-month trial.²¹

Symptoms decreased the same whether a levonorgestrel-releasing IUD or oral or vaginal progesterone was added.

Bone mineral density maintained or increased

Several randomized, controlled trials have documented the effects of estradiol gel on bone mineral density (BMD) and various markers of bone metabolism. In these studies, BMD remained steady^22,23 or increased^10,24,25 following treatment, and estradiol gel remained effective for up to 4 years.²⁵ Estradiol gel maintained or increased BMD with or without addition of progestins.^22,23,26

These investigations involved measurement of BMD at the lumbar spine, forearms, or hip, as well as biologic markers of bone turnover such as urinary hydroxyproline/creatinine ratio, serum alkaline phosphatase, and serum osteocalcin.

Serum estradiol and skeletal uptake of a bone-seeking agent also were determined. Estradiol gel regimens ranged from 0.75 mg/day to 3 mg/day, and populations included both surgical and natural menopausal subjects in several countries.

Effects comparable to oral estrogen. Compared with oral conjugated estrogen, which increased BMD at the lumbar spine by 4.3% (±3.2%), estradiol gel produced increases of 5.6% (±2.9%) at 24 months and 4.7% (±3.2%) at 36 months.¹⁰

 

Minimum level of protection achieved. Following a comparison of oral and percutaneous estradiol, Reginster et al²⁷ suggested that a minimum estradiol level of 60 pg/mL is necessary to prevent postmenopausal bone loss. Mean serum estradiol levels in women receiving 1.5 mg/day estradiol gel were 75.7 pg/mL and 78.4 pg/mL at 24 and 36 months, respectively, in a study by Palacios et al,¹⁰ and 85.8 pg/mL in a study by Devogelaer and colleagues.²⁴ In these trials, BMD increased, and it remained steady in other investigations.^28,29

More recent trials suggest that lower serum estradiol levels secondary to smaller estrogen doses have the capacity to maintain BMD.^30,31 A 1.9% mean increase of BMD at the lumbar spine was reported in women with a mean serum concentration of 17 pg/mL in a 2-year study³⁰ of a transdermal estradiol delivery system. The percutaneous route does not appear to limit these beneficial effects.

How are metabolic factors affected?

In general, oral estrogens produce beneficial changes in lipid metabolism, particularly higher levels of high-density lipoprotein (HDL) cholesterol. However, they also elevate triglyceride and glucose levels. How this plays out clinically is unclear. Both the Women’s Health Initiative and the Heart and Estrogen/Progestin Replacement Study (HERS) found no cardioprotective effects of oral estrogen despite increases in HDL and decreases in low-density lipoprotein (LDL) levels.^1,32

Oral versus percutaneous estradiol. No long-term studies of this magnitude have investigated the effect of percutaneous estradiol on coronary artery disease, although numerous clinical trials have shown that the route of estrogen delivery affects many of the metabolic variables used to estimate the risk of negative cardiac outcomes. It remains to be determined whether percutaneous estradiol affects these metabolic factors in ways that influence morbidity and mortality.

In 1 trial,³³ oral conjugated estrogen led to the following significant changes in lipids:

• increase in very low density lipoprotein (VLDL) cholesterol,
  
• decrease in LDL cholesterol (but not the LDL apoprotein B),
  
• increase in HDL and apoprotein A1, and
  
• significant increase in HDL2 cholesterol.
  

In contrast, percutaneous administration increased only HDL2 and the triglyceride and cholesterol content of the whole HDL fraction.

In a separate study,⁹ oral conjugated estrogen had a 2.5-fold increase in serum angiotensin, and percutaneous estradiol gel, no effect.

Same effects when progestin is added. Beneficial effects were not diminished when oral micronized progesterone was added to percutaneous estradiol.³⁴ Estradiol gel significantly reduced total serum cholesterol and LDL cholesterol in the first year of treatment, compared with placebo.

Coagulation effects. Percutaneous estradiol has fewer negative effects on coagulation factors than its oral counterpart. One study³⁵ investigating combination therapy with micronized progesterone compared the effects of percutaneous estradiol and oral estradiol valerate. The group receiving percutaneous estradiol gel/micronized progesterone had no significant changes in plasminogen activator inhibitor, tissue-type plasminogen concentration, and global fibrinolytic capacity. The other group had a significant decrease in mean tissue-type plasminogen concentration and plasminogen activator inhibitor activity and a significant rise in global fibrinolytic capacity.

The oral estradiol—but not the percutaneous formulation—significantly increased the mean value of prothrombin activation peptide and decreased mean antithrombin activity compared with no treatment.

Poor glycemic control may increase the risk of cardiovascular disease, but oral and percutaneous estradiol appear to have similar glycemic effects. A comparative trial³⁶ of oral estradiol valerate 2 mg/day and percutaneous estradiol 1 mg/day found no differences in glycosylated hemoglobin A1c levels (declined in both groups) or in fasting and 2-hour post-prandial blood glucose levels (constant in both groups) or insulin sensitivity.

Treatment duration may also influence how percutaneous estradiol affects metabolic factors. An open-label longitudinal prospective study³⁷ of 30 women receiving estradiol gel 1.5 mg/day for 6 months found a significant decrease in lipoprotein (a), apoprotein A-I, apoprotein B, HDL cholesterol, and HDL3 cholesterol. At 1 year, however, these changes were not significant.

No association with venous thromboembolism. Transdermal estradiol administered as a patch or percutaneous gel had no effect on the risk of venous thromboembolism in a multicenter case-control investigation.³⁸

In contrast, a recent retrospective study found a risk of venous thromboembolism that was at least 4 times greater with oral estrogen than with transdermal estradiol.³⁸

Use a progestogen to prevent endometrial hyperplasia

Like other forms of estrogen, percutaneous estradiol stimulates the endometrium. For this reason, women who have an intact uterus should use a progestogen in an adequate dose to prevent hyperplastic changes.³⁹ Although numerous regimens appear to be effective, the optimal route, dose, and duration of progestin in women using percutaneous estrogen remain to be determined.

Percutaneous estradiol versus other routes of administration. Endometrial thickness and amenorrhea rates over 6 months were not statistically different among 54 women treated with estradiol gel 1.5 mg/day, transdermal estradiol 50 μg/day, or oral estradiol valerate 2 mg/day—all in combination with nomegestrol acetate 2.5 mg/day.⁴⁰ The overall rates of no bleeding or spotting over 6 cycles of treatment were 78% in the percutaneous estradiol group, 48% in the transdermal group, and 60% in the oral group.

 

Although the percutaneous estradiol group had a lower overall incidence of no bleeding or spotting than the other groups, the difference was not significant. Nor were there significant differences in the variation of endometrial thickness from baseline: A mean increase of 1.5 mm (±0.4 mm) was reported in the percutaneous group, compared with 1.5 mm (±0.7 mm) and 1.7 mm (±0.6 mm) in the transdermal and oral estrogen groups, respectively.

Estrogenic and progestogenic effects were similar for transdermal and percutaneous estradiol (with dydrogesterone 10 mg/day for days 1 to 12) after a baseline atropic endometrium was identified.³

Estradiol gel produces stable, physiologic serum estradiol levels and has a serum estrone to estradiol ratio close to 1.

The most effective progestogen dosage and duration are unknown, although many regimens have been studied:

• Percutaneous estradiol 1.5 mg/day for the first 24 days of the month in combination with nomegestrol acetate 5 mg/day for days 11 to 24: Over 6 months, researchers found a secretory pattern in the majority of women and no evidence of hyperplasia.¹⁸
  
• Percutaneous estradiol 3 mg/day for 3 of every 4 weeks in combination with nomegestrol acetate 5 mg/day for 10 days: No reduction in estrogenic endometrial effects.^41,42
  
• Estradiol gel 3 mg/day for 3 of every 4 weeks in combination with 200 mg or 300 mg oral micronized progesterone for the last 10 days of treatment: Dose was too low or treatment too short to produce a complete secretory transformation of the endometrium.⁴³ However, the Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial⁴⁴ found no increased occurrence of hyperplasia in women using micronized progesterone 200 mg for 12 days of each cycle for more than 3 years in combination with oral conjugated estrogen 0.625 mg.
  
• Percutaneous estradiol 1.5 mg/day combined with micronized progesterone 100 mg daily (orally or vaginally) for the first 25 days of the month: Fully inhibited mitoses and induced amenorrhea in most of the women studied, with amenorrhea rates of 93.3% at 3 months and 91.6% at 6 months.⁴⁵
  
• Estradiol gel 1.5 mg/day with 100 mg vaginal micronized progesterone for 21 days per cycle: Stable endometrial thickness and no endometrial hyperplasia over 12 months.⁴⁶ However, breakthrough bleeding was reported in up to 30% of subjects, principally in the second 3 months of treatment.
  
• Percutaneous estradiol 1 mg/day for 3 months with oral medroxyprogesterone acetate 20 mg/day for the last 14 days, followed by a 7-day free interval: No reports of endometrial hyperplasia or suspect changes at 12 and 24 months.¹⁷
  
• Estradiol gel 2 mg/day for 21 days with 10 mg oral medroxyprogesterone acetate for the last 14 days, followed by a 7-day free interval: No endometrial hyperplasia or suspect changes at 12 and 24 months.¹⁷
  
• Estradiol gel 1 mg/day with medroxyprogesterone acetate 10 mg on days 1-12 every month or every 3 months: Endometrial hyperplasia was found in 1 woman (0.3%) in the group receiving the progestogen every 3 months. Endometrial histology did not differ between women taking medroxyprogesterone monthly and those taking it every 3 months.²¹
  
• Estradiol gel 3 mg/day with oral micronized progesterone 200 mg for 12 days of each cycle: Regular withdrawal bleeding in 70% of women.³⁴
  
• Percutaneous estradiol 1.5 mg/day in combination with a levonorgestrel-releasing intrauterine device: 80% of women were amenorrheic at 1 year, with a mean endometrial thickness 3 mm; at 5 years, 100% of women had epithelial atrophy.⁴⁷
  
• Estradiol gel 1.5 mg for 21 days with 200 mg oral progesterone for 14 days (126 women); 3 mg percutaneous estradiol for 21 days with 300 mg oral progesterone for 10 days (23 women); 1.5 mg estradiol gel with 300 mg oral progesterone (3 women); or 3 mg estradiol gel for 28 days with 200 mg progesterone for 14 days (5 women): No evidence of hyperplasia after 5 years of treatment.²
  

Dr. Archer is a consultant to Solvay, and has had grant support from Solvay for clinical trials.

Although the Women’s Health Initiative¹ discouraged many menopausal women from using oral estrogen, it failed to address the risks of treatment with other dosages or forms of estrogen and progestin.

Nor has any other trial of similar size and complexity taken up the issue. However, numerous smaller studies have been published.

This article summarizes findings on percutaneous delivery of estradiol gel (EstroGel), the most recently FDA-approved estrogen option for treatment of menopausal symptoms.

It describes the overall safety of estradiol gel, as well as its effectson:

• menopausal symptoms,
  
• bone,
  
• metabolism, and
  
• endometrium.
  

(In this article, “percutaneous delivery” refers to estradiol gel applied to the skin, and “transdermal estradiol” indicates delivery via a transdermal reservoir or matrix system, otherwise known as “the patch.” I have used an arbitrary definition to distinguish the gel from other methods of delivering estradiol across the skin. For example, Estrasorb is a liposomal formulation that is applied to the skin of the thigh. Although it is a percutaneous estradiol similar to the gel, this article focuses only on the latter option.)

Easy to apply, few skin reactions

Percutaneous estradiol gel formulations have been available for almost 30 years in Europe, where they are utilized by a majority of women on hormone therapy. In the United States, the hydroalcoholic gel is packaged in a pump that delivers 64 standardized 1.25-g doses, which contain 0.75 mg of 17ß-estradiol. Once it is applied, the gel is absorbed into an intradermal reservoir (FIGURE) and dries in 2 to 5 minutes, leaving no residue.

Patient selection. Estradiol gel is well-suited for patients who are concerned about the risks of oral estrogen (as portrayed in the mainstream press following the Women’s Health Initiative) and want to avoid that route of administration, as well as women who dislike or have difficulty swallowing pills. Percutaneous administration also is appropriate for physically active women who may have adhesion problems or skin irritation with the transdermal patch, or those who have had reactions to local adhesives in the past. The patient should be motivated to apply the gel on a daily basis.

Indications are moderate to severe vasomotor symptoms in menopausal women, and moderate to severe symptoms of vulvar and vaginal atrophy, although topical vaginal products should also be considered for the latter indication.

Contraindications are undiagnosed abnormal vaginal bleeding; history of breast cancer, other estrogen-dependent malignancy, stroke, heart attack, or liver disease or dysfunction; active thrombophlebitis or thromboembolic disorders (or a history of these); and known or suspected pregnancy.

Common side effects include headache, breast pain, irregular vaginal bleeding or spotting, stomach cramps or bloating, nausea and vomiting, and hair loss.

• Skin reactions are infrequent, but should be taken into account when discussing per-cutaneous or transdermal delivery of any drug. However, estradiol gel appears to cause fewer skin reactions than the patch. In a study over more than 5 years, 0 of 157 women treated with percutaneous estradiol reported skin irritation.² Other comparisons found similar outcomes.^3,4
  

After the gel dries, other lotions or per-fumes can be applied to the site, if desired, and the woman can cover the site with clothing.

Onset of action is rapid, as it is with the transdermal patch.⁵ Dose variability is minimized when the gel is applied at the same time every day to a large area of skin, preferably the arm, although all application sites appear to produce similar results: abdomen, shoulders, arms, and inner thigh.⁶ The gel should not be applied to the breast or vagina.

Diminished effect with skin washing. In 1 trial, site washing 1/2 hour after application significantly decreased bioavailability and time to reach peak plasma concentrations.⁷ For this reason, the gel should be not applied before a bath, shower, or sauna.

Dosing options. The initial dose is 1.25 mg of gel, which is 1 pump of the bottle. The gel is collected in the palm of 1 hand and applied to the skin of the opposite arm from the wrist to the shoulder. The dose can be titrated by adding a second pump of the gel and applying it to the opposite arm. The dose can be lowered by using less than a full depression of the pump.

 

Stable, physiologic estrogen levels

Estradiol gel produces relatively stable serum estradiol levels, and therapeutic estradiol levels similar to those seen with other formulations, routes of administration, and dosages.

Percutaneous administration produces serum estrone to estradiol ratios close to 1, in contrast to higher ratios (5:1) with oral administration.^8-11 The lower ratio approximates levels during the menstrual cycle of premenopausal women. (TABLE 1 gives estradiol and estrone levels from different studies following administration of estradiol as a gel, oral formulation, and transdermal patch.)

The percutaneous route also allows for delivery directly to the systemic circulation, avoiding gastrointestinal and first-pass hepatic metabolism and elimination. In contrast, oral micronized estradiol causes large fluctuations in serum estradiol and estrone levels due to absorption and metabolism.⁸

More stable serum levels than with the patch. One study¹² comparing percutaneous and transdermal estradiol found similar interindividual variability but less stable serum levels in women using the transdermal system. A separate study¹³ also reported greater fluctuation of serum estradiol levels in women using a transdermal system than in those using the gel.

TABLE 1

Serum estradiol and estrone levels for various routes of estradiol

AUTHOR	THERAPY	ESTRADIOL LEVEL (PG/ML)	ESTRONE LEVEL (PG/ML)
Scott et al8	E2gel 3 mg/day	102.9±39.9	120.0±50.5
	E2gel 1.5 mg/day	68.1±27.4	90.6±45.7
	Transdermal estradiol 50 μg/day	41.1±13.5	45.0±15.9
	Oral micronized estradiol 2 mg/day	114.0±65.2	575.2±279.9
Palacios et al10	E2gel 1.5 mg/day	75.7±1.0	58.5±2.9
	Oral conjugated estrogen 0.625 mg/day	39.6±4.6	126.0±7.6
Basdevant et al11	E2gel 3 mg/day	221.0±35.0	146.0±19.0
	Oral micronized estradiol 2 mg/day	121.0±27.0	811.0±370.0
Archer15	E2gel 0.75 mg/day	33.5*	49.0*
	E2gel 1.5 mg/day	65.0*	58.0*
	Placebo	5.0*	23.0*
*Median value
E2gel = percutaneous estradiol gel

Studies compared relief of menopausal symptoms

Estradiol gel effectively relieved meno-pausal symptoms in randomized, double-blind studies, open label comparisons, and observational trials in postmenopausal women, with and without the addition of various progestins.

Symptom relief in comparison with baseline values is statistically significant as early as 2 weeks after initiating treatment. Relief of up to 2 years’ duration has been reported.^3,14

Several studies have compared symptom relief achieved with estradiol gel, oral estrogen, transdermal delivery systems, and placebo^3,9,14-18; findings are summarized in TABLE 2.

Same efficacy when progestogen is added. The following studies, and other studies,¹⁴ demonstrated that estradiol gel relieves menopausal symptoms whether it is administered alone or in combination with a progestogen, with efficacy similar to other estrogen formulations:

Climacteric symptoms decreased to the same extent when estradiol gel was combined with a levonorgestrel-releasing intrauterine device, oral micronized progesterone, or vaginal micronized progesterone.¹⁹

A study²⁰ that added lynestrenol decreased the frequency of hot flushes and night sweats more than in women using estradiol gel alone. However, negative mood symptoms were more pronounced in the progestin-treated group.

Estradiol gel 1 mg/day in combination with monthly or quarterly oral medroxyprog-esterone acetate reduced the severity of hot flushes, sweating, and vaginal dryness, according to a 12-month trial.²¹

Symptoms decreased the same whether a levonorgestrel-releasing IUD or oral or vaginal progesterone was added.

Bone mineral density maintained or increased

Several randomized, controlled trials have documented the effects of estradiol gel on bone mineral density (BMD) and various markers of bone metabolism. In these studies, BMD remained steady^22,23 or increased^10,24,25 following treatment, and estradiol gel remained effective for up to 4 years.²⁵ Estradiol gel maintained or increased BMD with or without addition of progestins.^22,23,26

These investigations involved measurement of BMD at the lumbar spine, forearms, or hip, as well as biologic markers of bone turnover such as urinary hydroxyproline/creatinine ratio, serum alkaline phosphatase, and serum osteocalcin.

Serum estradiol and skeletal uptake of a bone-seeking agent also were determined. Estradiol gel regimens ranged from 0.75 mg/day to 3 mg/day, and populations included both surgical and natural menopausal subjects in several countries.

Effects comparable to oral estrogen. Compared with oral conjugated estrogen, which increased BMD at the lumbar spine by 4.3% (±3.2%), estradiol gel produced increases of 5.6% (±2.9%) at 24 months and 4.7% (±3.2%) at 36 months.¹⁰

 

Minimum level of protection achieved. Following a comparison of oral and percutaneous estradiol, Reginster et al²⁷ suggested that a minimum estradiol level of 60 pg/mL is necessary to prevent postmenopausal bone loss. Mean serum estradiol levels in women receiving 1.5 mg/day estradiol gel were 75.7 pg/mL and 78.4 pg/mL at 24 and 36 months, respectively, in a study by Palacios et al,¹⁰ and 85.8 pg/mL in a study by Devogelaer and colleagues.²⁴ In these trials, BMD increased, and it remained steady in other investigations.^28,29

More recent trials suggest that lower serum estradiol levels secondary to smaller estrogen doses have the capacity to maintain BMD.^30,31 A 1.9% mean increase of BMD at the lumbar spine was reported in women with a mean serum concentration of 17 pg/mL in a 2-year study³⁰ of a transdermal estradiol delivery system. The percutaneous route does not appear to limit these beneficial effects.

How are metabolic factors affected?

In general, oral estrogens produce beneficial changes in lipid metabolism, particularly higher levels of high-density lipoprotein (HDL) cholesterol. However, they also elevate triglyceride and glucose levels. How this plays out clinically is unclear. Both the Women’s Health Initiative and the Heart and Estrogen/Progestin Replacement Study (HERS) found no cardioprotective effects of oral estrogen despite increases in HDL and decreases in low-density lipoprotein (LDL) levels.^1,32

Oral versus percutaneous estradiol. No long-term studies of this magnitude have investigated the effect of percutaneous estradiol on coronary artery disease, although numerous clinical trials have shown that the route of estrogen delivery affects many of the metabolic variables used to estimate the risk of negative cardiac outcomes. It remains to be determined whether percutaneous estradiol affects these metabolic factors in ways that influence morbidity and mortality.

In 1 trial,³³ oral conjugated estrogen led to the following significant changes in lipids:

• increase in very low density lipoprotein (VLDL) cholesterol,
  
• decrease in LDL cholesterol (but not the LDL apoprotein B),
  
• increase in HDL and apoprotein A1, and
  
• significant increase in HDL2 cholesterol.
  

In contrast, percutaneous administration increased only HDL2 and the triglyceride and cholesterol content of the whole HDL fraction.

In a separate study,⁹ oral conjugated estrogen had a 2.5-fold increase in serum angiotensin, and percutaneous estradiol gel, no effect.

Same effects when progestin is added. Beneficial effects were not diminished when oral micronized progesterone was added to percutaneous estradiol.³⁴ Estradiol gel significantly reduced total serum cholesterol and LDL cholesterol in the first year of treatment, compared with placebo.

Coagulation effects. Percutaneous estradiol has fewer negative effects on coagulation factors than its oral counterpart. One study³⁵ investigating combination therapy with micronized progesterone compared the effects of percutaneous estradiol and oral estradiol valerate. The group receiving percutaneous estradiol gel/micronized progesterone had no significant changes in plasminogen activator inhibitor, tissue-type plasminogen concentration, and global fibrinolytic capacity. The other group had a significant decrease in mean tissue-type plasminogen concentration and plasminogen activator inhibitor activity and a significant rise in global fibrinolytic capacity.

The oral estradiol—but not the percutaneous formulation—significantly increased the mean value of prothrombin activation peptide and decreased mean antithrombin activity compared with no treatment.

Poor glycemic control may increase the risk of cardiovascular disease, but oral and percutaneous estradiol appear to have similar glycemic effects. A comparative trial³⁶ of oral estradiol valerate 2 mg/day and percutaneous estradiol 1 mg/day found no differences in glycosylated hemoglobin A1c levels (declined in both groups) or in fasting and 2-hour post-prandial blood glucose levels (constant in both groups) or insulin sensitivity.

Treatment duration may also influence how percutaneous estradiol affects metabolic factors. An open-label longitudinal prospective study³⁷ of 30 women receiving estradiol gel 1.5 mg/day for 6 months found a significant decrease in lipoprotein (a), apoprotein A-I, apoprotein B, HDL cholesterol, and HDL3 cholesterol. At 1 year, however, these changes were not significant.

No association with venous thromboembolism. Transdermal estradiol administered as a patch or percutaneous gel had no effect on the risk of venous thromboembolism in a multicenter case-control investigation.³⁸

In contrast, a recent retrospective study found a risk of venous thromboembolism that was at least 4 times greater with oral estrogen than with transdermal estradiol.³⁸

Use a progestogen to prevent endometrial hyperplasia

Like other forms of estrogen, percutaneous estradiol stimulates the endometrium. For this reason, women who have an intact uterus should use a progestogen in an adequate dose to prevent hyperplastic changes.³⁹ Although numerous regimens appear to be effective, the optimal route, dose, and duration of progestin in women using percutaneous estrogen remain to be determined.

Percutaneous estradiol versus other routes of administration. Endometrial thickness and amenorrhea rates over 6 months were not statistically different among 54 women treated with estradiol gel 1.5 mg/day, transdermal estradiol 50 μg/day, or oral estradiol valerate 2 mg/day—all in combination with nomegestrol acetate 2.5 mg/day.⁴⁰ The overall rates of no bleeding or spotting over 6 cycles of treatment were 78% in the percutaneous estradiol group, 48% in the transdermal group, and 60% in the oral group.

 

Although the percutaneous estradiol group had a lower overall incidence of no bleeding or spotting than the other groups, the difference was not significant. Nor were there significant differences in the variation of endometrial thickness from baseline: A mean increase of 1.5 mm (±0.4 mm) was reported in the percutaneous group, compared with 1.5 mm (±0.7 mm) and 1.7 mm (±0.6 mm) in the transdermal and oral estrogen groups, respectively.

Estrogenic and progestogenic effects were similar for transdermal and percutaneous estradiol (with dydrogesterone 10 mg/day for days 1 to 12) after a baseline atropic endometrium was identified.³

Estradiol gel produces stable, physiologic serum estradiol levels and has a serum estrone to estradiol ratio close to 1.

The most effective progestogen dosage and duration are unknown, although many regimens have been studied:

• Percutaneous estradiol 1.5 mg/day for the first 24 days of the month in combination with nomegestrol acetate 5 mg/day for days 11 to 24: Over 6 months, researchers found a secretory pattern in the majority of women and no evidence of hyperplasia.¹⁸
  
• Percutaneous estradiol 3 mg/day for 3 of every 4 weeks in combination with nomegestrol acetate 5 mg/day for 10 days: No reduction in estrogenic endometrial effects.^41,42
  
• Estradiol gel 3 mg/day for 3 of every 4 weeks in combination with 200 mg or 300 mg oral micronized progesterone for the last 10 days of treatment: Dose was too low or treatment too short to produce a complete secretory transformation of the endometrium.⁴³ However, the Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial⁴⁴ found no increased occurrence of hyperplasia in women using micronized progesterone 200 mg for 12 days of each cycle for more than 3 years in combination with oral conjugated estrogen 0.625 mg.
  
• Percutaneous estradiol 1.5 mg/day combined with micronized progesterone 100 mg daily (orally or vaginally) for the first 25 days of the month: Fully inhibited mitoses and induced amenorrhea in most of the women studied, with amenorrhea rates of 93.3% at 3 months and 91.6% at 6 months.⁴⁵
  
• Estradiol gel 1.5 mg/day with 100 mg vaginal micronized progesterone for 21 days per cycle: Stable endometrial thickness and no endometrial hyperplasia over 12 months.⁴⁶ However, breakthrough bleeding was reported in up to 30% of subjects, principally in the second 3 months of treatment.
  
• Percutaneous estradiol 1 mg/day for 3 months with oral medroxyprogesterone acetate 20 mg/day for the last 14 days, followed by a 7-day free interval: No reports of endometrial hyperplasia or suspect changes at 12 and 24 months.¹⁷
  
• Estradiol gel 2 mg/day for 21 days with 10 mg oral medroxyprogesterone acetate for the last 14 days, followed by a 7-day free interval: No endometrial hyperplasia or suspect changes at 12 and 24 months.¹⁷
  
• Estradiol gel 1 mg/day with medroxyprogesterone acetate 10 mg on days 1-12 every month or every 3 months: Endometrial hyperplasia was found in 1 woman (0.3%) in the group receiving the progestogen every 3 months. Endometrial histology did not differ between women taking medroxyprogesterone monthly and those taking it every 3 months.²¹
  
• Estradiol gel 3 mg/day with oral micronized progesterone 200 mg for 12 days of each cycle: Regular withdrawal bleeding in 70% of women.³⁴
  
• Percutaneous estradiol 1.5 mg/day in combination with a levonorgestrel-releasing intrauterine device: 80% of women were amenorrheic at 1 year, with a mean endometrial thickness 3 mm; at 5 years, 100% of women had epithelial atrophy.⁴⁷
  
• Estradiol gel 1.5 mg for 21 days with 200 mg oral progesterone for 14 days (126 women); 3 mg percutaneous estradiol for 21 days with 300 mg oral progesterone for 10 days (23 women); 1.5 mg estradiol gel with 300 mg oral progesterone (3 women); or 3 mg estradiol gel for 28 days with 200 mg progesterone for 14 days (5 women): No evidence of hyperplasia after 5 years of treatment.²
  

Dr. Archer is a consultant to Solvay, and has had grant support from Solvay for clinical trials.