Are All Estrogens Created Equal? A Review of Oral vs. Transdermal Therapy

Sources

A review of the available English language literature was undertaken, utilizing existing articles and references culled from reviews and editorials in publications available to the author, as well as a database-directed search as means for detecting literature relating to both effects of estrogen therapy (ET) on the health of menopause-aged women, and the comparison of oral vs. TD administration. Search terms utilized included “transdermal vs. oral hormone therapy,” and “transdermal vs. oral estrogen therapy.” All articles in indexed, English language peer-reviewed publications were considered, and their data were utilized if they contained material referable to the point(s) being addressed and were powered highly enough to provide meaningful data. No attempt was made to judge study validity dependent on methods of data collection or collation. No “combining” of data was undertaken. Articles accessed were from a broad spectrum of investigators across multiple disciplines and journals. Subgroup and secondary analyses were not performed. Areas of comparison include absorption and bioavailability of oral and TD products; metabolic routes and their sequelae; cardiovascular, inflammatory and thrombotic effects; differential effects on lipids, bioavailable testosterone, thyroid proteins, and cortisol; and effects on weight, IR, and other metabolic factors, as well as differential effects on the breast, bones, and uterus.

Study selection

One hundred twenty-five citations were reviewed. Those written in languages other than English, redundant to material already referenced, or with a very small sample size were not utilized. Included were randomized controlled trials (single, multi-centered, blinded, placebo-controlled, cross-over), systematic reviews with and without meta-analysis, randomized primary and secondary prevention trials, cohort studies, prospective comparative studies, and observational (including multi-center and cross-sectional) studies. In consideration of space limitations, the areas of gallstone formation, comparisons involving the oral product tibilone and selective estrogen receptor modulators, continuation issues involving different delivery systems, effects on homocysteine and dimethylarginine reduction, and differential effects on serum amyloid-a, as well as other more peripheral matters were not analyzed for this review.

Definition: TD estradiol products

It is important to note that all TD (and transmucosal) preparations contain the same active ingredient, molecularly identical (or “bioidentical”) E2, synthesized from plant sterols, most commonly soybeans or wild Mexican yam. All TD products are bioidentical (bioidentical hormone replacement therapy [BHRT]), whether U.S. Food and Drug Administration (FDA)-approved pharmaceuticals or individually compounded preparations, although only FDA-approved medications were included in studies accessed. By definition, TD products involve skin application, whether via patches applied once or twice weekly to the lower abdominal area or daily products sprayed or massaged into the skin of the thigh, upper arm, or forearm. An estradiol-containing slow-release vaginal ring is also available; it is speculated that its metabolic effects are similar to transdermal products.

Lipid profile

It is well known that menopause has adverse effects on lipids and lipoprotein serum concentrations, increasing low-density lipoprotein cholesterol (LDL-C) and triglycerides and decreasing high-density lipoprotein cholesterol (HDL-C) concentrations. ^5,18 Studies have shown that oral HT beneficially affects lipid profiles of postmenopausal women, resulting in increases in HDL-C and decreases in LDL-C to a degree greater than transdermals, ¹⁹ while at the same time increasing triglyceride levels, an independent risk factor for stroke. ^20,21 TD HT appears to have a favorable effect on triglyceride levels while maintaining, albeit to a slightly lesser degree, oral HT's beneficial effects on LDL-C and HDL-C. ^{22 –24} Additionally, transdermals have been noted to produce larger LDL-C particles (“pattern A”), more resistant to oxidation. ²⁵ A small study found that postmenopausal women whose triglyceride levels rose after therapy with oral estrogen had significantly reduced levels (−51%) after switching to TD E2 administration. ²⁶ TD, but not oral ET significantly decreased the atherogenic index of plasma compared with baseline, making the differences between the therapies statistically significant. ²⁷ Unknown is whether these differential effects translate into cardiovascular events.

Cardiovascular, inflammatory, and thrombotic effects

Studies in the 1960s suggested that oral estrogen–progestin combinations appeared to increase blood coagulability. ^28,29 In addition to hormone binding globulins, clotting factors are produced by the liver and are stimulated by the high estrogen levels needed to overcome first-pass metabolism. Exogenous oral estrogen appears to affect the activity of platelets and plasma coagulation factors, both key factors in clotting and thrombosis. It has been shown that oral estrogens combined with a progestin leads to an increase in platelet adhesiveness and aggregability and increased plasma fibrinogen and coagulation factor activity, especially factors VII, VIII, and IX, as well as diminishing antithrombin III, an inhibitor of coagulation. Oral contraceptive sex steroid combinations appear to increase venous thromboembolism (VTE) risk by this combination of increased clotting factors and decreased antithrombin activity. It appears that, presumably via hepatic metabolism, these changes apply also to menopausal HT, but whether this applies to estrogen alone, or only oral estrogen–progestin combination, or estrogen combined with progesterone, remains a subject of debate, with conflicting studies in the literature. What roles do the delivery system (oral vs. TD) and the specific progestogen utilized (progestin vs. progesterone) play? A recent study ³⁰ suggests that indeed the progestin MPA utilized in the WHI investigation may play a critical role, noting that after 11 years, women taking CEE alone and starting near to menopause had significantly lower (RR 0.61–0.63) risks of CVD and MI, in contrast with women who took CEE+MPA (RR 1.21–1.23).

Data from the WHI trials suggest that orally administered HT in women 50 to 79 years old was associated with a combined relative risk (RR) of 1.31 for ischemic and hemorrhagic stroke, although the majority (85.2%; RR 1.44) were ischemic strokes; a minority (14.8%; RR 0.82) were hemorrhagic. ⁹ A large Danish observational study ³¹ found no increased risk of myocardial infarction with unopposed estrogen, cyclical combined HT, or tibilone, and a significantly lower risk of MI was found with dermal routes of application (RR 0.63; confidence interval [CI] 0.42–0.93). If the method of taking estrogen was via patch or gel on the skin, the risk of MI was reduced by 38%, and for vaginal application, by 44%, compared with oral unopposed ET. Interestingly, this decrease in MI incidence did not hold with combined ET (oral or TD) and progestin (norethisterone acetate, MPA, or levonorgestrel); progesterone was not utilized in any of the study groups. Another study, ³² in which risk was separated by route of administration, reported similar outcomes for oral and TD, with the multivariate odds ratio for MI was 0.66 and 0.75 for oral and TD routes of administration, respectively.

Cardiovascular findings have been generalized to all patients and HT regimens. However, mounting evidence suggests that many factors influence risk. More recently reported WHI findings and other studies ^{6 –13,15,16,30} have demonstrated that, when started near to menopause (within 6–10 years, in patients younger than 60 years), HT acts to prevent cardiovascular disease (hazard ratio [HR] 0.76, CI 0.5–1.16 ¹⁴ ; HR 0.63, CI 0.36–1.08 ¹¹ ).

Transdermally administered estrogen does not appear to be associated with the same short-term cardiovascular risks as oral HT. ³³ A large population-based cohort study of >78,800 women showed that, compared with women who used HT for <6 months, those exposed for >3 years had a hazard ratio for risk of hospitalization for cardiovascular disease of 0.53 if they utilized transdermals and 1.15 with oral HT. ³⁴ However, when adjusted for income, the overall hazard ratio of 0.65 climbed to 0.94, calling into question confounding factors of many studies in the literature. Other cardiovascular risk factors, such as higher body mass index, compounded the VTE risk with oral HT but did not alter the effects of TD HT. ³⁵ Women with a thrombotic mutation who used oral HT had a >25-fold increased risk of VTE, compared with nonusers and TD HT users, both of which had a fourfold increased risk, ^14,35 findings confirmed by the earlier study of Scarabin and collegues. ³⁶ Canonico and her colleagues ^37,38 in the Estrogen and Thromboembolism Risk (ESTHER) Study Group in France recruited 206 consecutive hospital admissions for first documented episode of idiopathic VTE, paired with 426 controls. The study, with a population averaging 61.6 years of age (average age at menopause, 49 years), found that in postmenopausal women oral estrogen is associated with an increased thromboembolic risk, especially during the first year of treatment, while TD estrogen may be safer in terms of thrombotic risk. The relative risk with adjusted odds ratio for VTE with use of oral estrogens was 4.2; in contrast, the ratio was 0.9 with TD preparations. Other investigators ^39,40 confirm diminished hemostatic activation, thrombosis, and VTE in postmenopausal users of TD ET, when compared with those using oral ET.

In one large retrospective case–control study, oral estrogen was associated with an almost fourfold increased risk of VTE compared to nonuse, whereas transdermals produced no increased risk. ³⁶ TD E2 may also differ from oral estrogen in its ability to lower blood pressure, ^41,42 although no “head-to-head” comparison studies exist.

A 1-year study comparing both TD and oral CEE, both combined with MPA, found little differential effect on blood coagulation factors, ⁴³ again calling into question the role of the progestogen in coagulation effects of HT because most studies of oral medications utilized CEE combined with the progestin MPA, distinct from many TD studies, which eschewed the use of MPA. Another study of the effects of unopposed ET on markers of coagulation and inflammation on postmenopausal women showed an increase in plasminogen and CRP, but little difference on other coagulation markers. ⁴⁴

Scarabin and colleagues ⁴⁵ studied cyclic oral and TD E2, both combined with progesterone. Oral, but not transdermal E2 significantly increased prothrombin activation peptide and decreased mean antithrombin activity compared with no treatment. The oral estrogen group was associated with a significant decrease in both plasminogen concentration and plasminogen activation factor and a significant rise in global fibrinolytic capacity, compared with the TD and control groups, with no significant changes in values of fibrinogen, factors VII and VIII, and other clotting factors between the three groups. They concluded that E2/progesterone replacement therapy may result in increased fibrinolytic potential, whereas unopposed TD E2 appears without substantial effects on hemostasis and emphasized that the increased risk of VTE in users of oral postmenopausal estrogen may alter the route of administration, especially those at high risk of VTE.

Biochemical evidence also appears to support the lower cardiovascular risk of TD estrogen. Although estrogen in general reduces levels of inflammatory markers in the blood, there are some exceptions to this general assertion. Oral estrogen in particular appears to increase inflammatory proteins (namely, CRP and matrix metalloproteinases [MMPs], specifically MMP-9) and suppress their inhibitors, largely a hepatic metabolic effect influenced by the dose and route of administration. ^{46 –56}

Elevated CRP, a strong predictor of adverse cardiovascular events in healthy postmenopausal women, enhances inflammation and promotes tissue damage in acute MI. ⁵⁷ In a cross-over, placebo-controlled study of postmenopausal women, oral conjugated estrogens caused a statistically significant greater than twofold increase in CRP levels, while transdermals had no effect on CRP levels. ⁵⁸ In a recent study, Shifren and her colleagues ⁵⁹ found that TD ET may be preferable to oral ET because it minimizes the hepatic estrogen exposure that increases CRP and levels of coagulation markers and reduces IGF-1. Other investigators have reported similar findings. ^{60 –64} Oral estrogen appears to induce atherosclerotic cardiovascular disease by increasing acute inflammation; TD ET appears to avoid this effect. TD E2 exerts a positive effect on endothelial function similar to oral estrogen; however, TD may be more favorable than oral in terms of vascular disease risk. ⁶⁵ It must be noted, however, that CRP and IGF-1 are surrogate markers and cannot make a determination of true “risk.”

In a presentation at the 20th Annual Meeting of the North American Menopause Society, October 1, 2009 (unpublished data), Greep and her colleagues from University of California at Los Angeles performed a retrospective study based on the UCLA stroke database of 955 women, finding 127 were postmenopausal. The study found that among these women, the use of oral HT was associated with a significant elevation in ischemic stroke, particularly cryptogenic stroke, in patients with a patent foramen ovale (presumably thus allowing clots to directly access arterial circulation). Of the 25 women in the study who were on HT at the time of their stroke, all were on oral therapy; increased stroke risk was not seen in patients using TD E2.

Effects on IGF, IR, and MeS

Several investigators have sought to determine whether oral or TD E2 treatment would be more beneficial for women with IR and MeS, based on changes in inflammation and coagulation markers. In postmenopausal women with obesity and MeS who have increased cardiovascular risk, TD therapy did not affect the ratio of circulating MMP-9 to tissue inhibitor of metalloproteinase (TIMP)-1, whereas the ratio was increased with oral ET. MMPs, notably MMP-9, belong to a family of enzymes that degrade collagen and have a role in promoting plaque rupture. High levels of MMPs are strongly associated with symptomatic cardiovascular disease and are predictive of cardiovascular risk. These metabolic changes in the presence of significant atherosclerosis may have the ability to cause coronary artery plaque instability and may lead to plaque rupture and coronary occlusion. ⁶⁶ Chu and her group ⁶⁶ also found that postmenopausal women with MeS have higher levels of certain coagulation and inflammation markers, facts corroborated by Davis et al. ⁶⁷ who noted differential effects of the route of administration on IGF-1, IGF binding protein-3, and IR.

Sexual effects

As noted earlier, SHBG levels are influenced by the first-pass effect of oral estrogen therapy. This sex hormone–specific plasma transport protein, which is synthesized in the liver, controls hormone activity by binding reversibly but with high affinity to testosterone and E2. Since only the unbound hormone is biologically active, increasing SHBG decreases hormone activity, potentially having a negative effect on energy, mood, sex drive, and sexual responsiveness. ⁶⁸

Oral HT of any type has been shown to increase SHBG, frequently by greater than 100%, although production is dose and individual dependent. ^20,69 However, studies have verified that lower dose TD E2 has little or no effect on SHBG levels. ⁵

In a systematic review on the subject of different types of estrogen therapy in relation to women's sexual function, Nappi and Polatti ⁷⁰ also come to the conclusion that the type, dose, and delivery system matter, and further posit that “…Collectively, these data allow us to reasonably state that TD E2 seems to be the more logical choice to treat peri- and postmenopausal women with female sexual dysfunction, because it avoids a further reduction of circulating androgens without significantly affecting vascular homeostasis.… (pgs 607, 608)” Their focus was on SHBG, which decreases circulating free and bioavailable testosterone. SHBG is variably effected by ET, strongly increased by CEE at a dose of 0.625 mg/d, moderately increased by oral E2 at a dose of 1 mg/d, and minimally increased by TD ET at a dose of 0.05 mg/d, ⁷¹ which was corroborated by a randomized open label cross-over study by Shifren et al. ⁷²

The authors theorize that TD E2, which avoids first-pass metabolism by liver and enzymatic degradation by the gastrointestinal tract, displayed only a mild effect (12% elevation) on SHBG levels, while CEE 0.625 mg/d increased circulating SHBG by 132.2% after 12 weeks. Shifren et al. ⁷² found the amount of free E2 was almost double in women treated with TD E2, while free estrone was significantly increased in women taking CEE. Free testosterone was significantly reduced (−32.7%) in women on CEE, while almost unchanged (+1%) in women on TD E2. They refer to Simon and Snabes's study ⁷³ by stating that TD E2 induces a less pronounced effect on hepatic protein markers of coagulation and fibrinolysis, while oral E2 has more pronounced hyper-coagulant effects on synthesis of CRP and fibrinolytic markers. They quote other epidemiologic data identifying VTE risk differences between oral and TD therapy, ^36,74

Effects on weight

It has been shown that estrogens, both oral and transdermal, appear to reduce midlife weight gain and have a positive effect on weight and body composition. ⁷⁵ Several studies have been performed investigating differences in body composition, comparing orally versus transdermally administered estrogen. Although most of these studies show an improved lean/fat mass ratio for TD estrogen, the evidence is indirect, and there is not universal agreement.

Because of its first-pass effect, oral estrogen causes diminished lipid oxidation resulting in more fat accumulation. ⁷⁶ At the same time, growth hormone and its binding protein are increased, resulting in decreased IGF-1 and reduction in lean mass, with increased fat mass and increases in leptin, changes not found with TD estrogen. ⁷⁷ Increases in leptin have been statistically associated with IR, contributing to more abdominal fat accumulation. ⁷⁸ It has been observed that oral estrogen may cause a worsening of IR in postmenopausal women with obesity and MeS, while TD E2 was neutral. ⁷⁹ Additionally, oral estrogen led to an increased leptin/adiponectin ratio, while leptin was unchanged with TD therapy.

Intuitively as well, any product that increases, even slightly, thyroid binding globulin and cortisol binding globulin might be expected to have the effect of lowering the basal metabolic rate, resulting in a slightly lowered caloric expenditure and weight gain. These data are suggestive, but not overwhelming; there is certainly room for well-designed studies to add to knowledge in this area.

Breast cancer, uterine cancer, and bone density issues

In a study directly comparing the bone-sparing effects of TD and oral therapies compared with placebo, Hillard and colleagues ⁸⁰ found no difference between the two delivery systems. Ettinger and colleagues ⁸¹ have shown that a daily TD dose as low as 0.014 mg E2, average serum concentration (C_ave)=8.5 pg/mL, effectively maintains bone mineral density in postmenopausal women aged 60 or greater. All of the commercially available FDA-approved TD and oral products, at all available dosages, exceed this serum concentration, with the C_ave for lower-dose TD products in the 8.5- to 31-pg range. ⁸²

Although a significantly smaller segment of the female population eventually succumbs to breast carcinoma than to cardiovascular disease, for which ET appears to generously diminish risk (HR 0.63–0.76) if started near to menopause, ^{6 –15,83} the former is of immense concern to our patients. Adequate evidence exists that an individual's development of this disease relates to ongoing levels of estrogen and progestogenic substances circulating through breast tissue, and may be dose dependent, although the data to substantiate this are not robust. No prospective head-to-head reports yet grace the literature regarding this important consideration, although one might posit that TD products, with their generally lower average serum concentrations, may be associated with mildly diminished rates of carcinoma, when compared with average oral doses for which C_ave levels are harder to calculate secondary to individual absorption differences exceeding that of TD products. The C_ave for oral conjugated equine estrogens 0.625 mg has been calculated at ∼87 pg/mL, and for oral E2 1 mg, 50 pg/mL plus estrone levels in the 150–300 mg/mL range. ⁶³ Data are not available for the lower dose formulations of oral products (i.e., conjugated equine estrogen 0.3 and 0.45 mg, oral E2 0.5 mg), but the lower dose TD formulations (e.g., E2 patch 0.025 and 0.0375 mg, aerosolized E2 one to two sprays, E2 gel 0.25–0.5 mg of 0.1% gel) produce average serum concentrations of 10–31 pg/mL, with mid-dose products (E2 patch 0.05 to 0.075 mg, higher dose E2 gels, E2 aerosol three sprays) in the 40–60 pg/mL range. ⁸² Several investigators ^{83 –86} have found similar risks, although one study ⁸⁴ of 73,500 Italian women noted a hazard ratio for ever developing breast cancer to be 1.34 (1.27 with TD and 2.14 with oral), suggesting, at least in this group, that oral therapy may increase risk. Nothing is noted about dosages utilized.

Additionally, increased risk of breast cancer is more likely related to concentrations of estrogen locally in breast tissue rather than the serum levels. Oral products providing higher levels of estrone in breast tissue than TD products may have impact.

Of great importance, whether ET is administered via oral or TD route, is the role of the accompanying progestogen in breast cancer development. Even before the analysis by LaCroix et al. ³⁰ of the ongoing WHI investigation provided evidence of the lack of breast cancer increase in women taking CEE without concomitant MPA, Ross et al. ⁸³ in a landmark study showed a very small increase in breast cancer risk for women consuming ET (CEE) alone, but a very considerable increase in risk for women on estrogen/progestin therapy with MPA as the progestogen. Additionally, they concluded that “… for each incident case of breast cancer, 6 cardiovascular deaths would be prevented.… (p. 331)” These findings have also been confirmed by Stefanick et al. ⁸⁷ in their original analysis of data from the WHI ET (CEE)-only arm of the WHI, finding a relative risk of 0.8 compared with control for women on CEE (no progestin) for 7.1 years.

Clinicians must also consider that individual uterine responsivity to estrogen varies widely and that for uterine protection, even when using a low-dose TD product, continuous or cyclic progestogen treatment remains the standard of care, while at the same time understanding that present data suggest that, certainly if that progestin is MPA, the incidence of breast carcinoma will rise.