By Sam | First published April 11, 2020 | Last modified October 5, 2020
In the modern world, the oral and transdermal routes are by far the most common means to administer exogenous estradiol and are widely used as a component of gender-affirming hormone therapy. Current clinical evidence shows no difference in feminising efficacy between these formulations at equivalent doses. Although both are well tolerated in general, the oral route is unphysiological in its metabolism and is associated with a significantly greater incidence of cardiovascular and thromboembolic complications. At low adult replacement doses, transdermal forms do not have these disadvantages and may be superior to their oral counterparts in feminising hormone therapy.
Estrogen replacement is both an important and necessary intervention for many transgender people (Hembree et al., 2017; Cheung et al., 2019). In the past, feminising therapy in this group was done using high dose estrogen monotherapy with parenteral esters of estradiol such as estradiol valerate or estradiol undecylate (Benjamin, 1967; Hamburger, 1969). Non-bioidentical oral estrogens such as conjugated equine estrogens and ethinylestradiol were also widely used (Meyer et al., 1986; Meyer, Walker & Suplee, 1989). However, with the significant progress made in drug development, bioidentical estradiol became widely available as oral and transdermal formulations for gender-affirming hormone therapy.
Today, some transgender individuals prefer to use injectable formulations of estradiol (Geffen et al., 2018). Nonetheless, the oral and transdermal routes of administration are used almost without exception for therapy in Europe and some other parts of the world and are probably most commonplace (Fisher & Maggi, 2015; Hamidi & Davidge-Pitts, 2019; Seal, 2019). Many people receiving or eager to start hormone therapy may be interested to know what data exists regarding differences between oral and transdermal estradiol. As we require long-term therapy with these formulations, a discussion regarding adverse effects between these routes of administration may also be of importance. Although the focus of this review is largely on oral estradiol as compared with transdermal estradiol, a good amount of the discussion specific to transdermal estradiol can likely be extrapolated to other non-oral routes of administration when the doses have similar potency. For instance, estradiol administered by intramuscular or subcutaneous injection is a non-oral route of administration.
Oral estradiol includes pill or tablet formulations, while transdermal estradiol is most commonly available as patches or gels (Kuhl, 2005). Oral estradiol and transdermal gel is usually administered once per day (Rohr, Volko & Schindler 2014). However, doses may be split and taken twice-daily. Theoretically, this would result in more stable estradiol levels, although is obviously less convenient. Estradiol pills may also be administered by the sublingual or buccal routes (Casper & Yen, 1981; Wren et al., 2003). Although some literature exists regarding their use for therapy in feminising hormone therapy, usage of these routes is probably still relatively rare in clinical practice (Jain, Kwan & Forcier 2019). In this review, I have used the term “oral estradiol” to refer only to the swallowing of estradiol tablets. Estradiol patches are applied and worn continuously. Different brands exist and transdermal patches are available for twice-weekly or weekly administration. A 50 μg/day dosage delivered by transdermal patch is considered to have similar potency to a 1 to 2 mg/day dosage of oral estradiol and to a 1.5 mg/day dosage of transdermal gel (Kuhl, 2005; Järvinen, Nykänen & Paasiniemi, 1999). However, as there is massive interindividual variability, these doses likely will not correspond to one another on an individual basis.
Estradiol is secreted by the ovaries into systematic circulation. As a result, the liver is not disproportionately exposed to the effects of the hormone (Gravholt et al., 2017). Transdermal estradiol is effective in mimicking this behaviour. However, orally administered estradiol, owing to passage through the gastrointestinal tract, is associated with disproportionate estrogenic exposure in the liver (Bińkowska, 2014). This behaviour gives rise to a number of differences between oral and transdermal estradiol. One such difference is that about 95% of oral estradiol is metabolised, as a consequence of the first pass effect, into estrone and other clinically weak/insignificant estrogens (Kuhz, Blode & Zimmermann, 1993). The ratio of estrone to estradiol is close to 1:1 in adult women and pubertal girls and with transdermal formulations (Kuhl, 2005; Frederiksen et al., 2020). However, with a dose of oral estradiol, postmenopausal women have been found to have about 5 times the concentration of estrone as estradiol (Kuhl, 2005). In some patients, the concentration of estrone may be 20 times higher than that of estradiol (Kuhnz, Gansau & Mahler, 1993). For this reason, the metabolism of oral estradiol has been described as unphysiologic (Gravholt et al., 2017; Mauras et al., 2019).
A subject of great interest to many transgender people taking feminising hormone therapy seems to be concerning which regimens might be most “effective”. In particular, satisfactory breast development is often sought after. However, to date, no randomised controlled trials assessing the efficacy of different gender-affirming hormone regimens have been conducted (Reisman, Goldstein & Safer, 2019; Iwamoto et al., 2019). Moreover, there are no measures of breast development or other effects of transition that are universally agreed upon by the scientific community. For this reason, it is difficult to directly compare the findings of many studies.
Nonetheless, a number of observational studies have studied and quantified feminisation experienced with hormone therapy. In one prospective study, transgender women (n = 53) were treated with cyproterone acetate plus either oral or transdermal estradiol (Wierckx et al., 2014). It was noted that after 1 year, there was no apparent difference in physical measures such as breast circumference between the oral and transdermal groups. Another longitudinal and multicentre study of transgender women (n = 229) attending clinics in the Netherlands, Belgium and Italy also reported that the increase in breast-chest difference following 1 year of therapy did not differ between those using oral and transdermal formulations (de Blok et al., 2018). As a result of unsatisfactory development, many transgender women seek breast augmentation (Seal, 2016; de Blok et al., 2020). Although breast development itself was not measured, it is interesting to note that one retrospective study found no statistical difference in the rate of augmentation requests between users of different estrogen types (Seal et al., 2012). This suggests that oral estradiol valerate might be no more or less effective than the other estrogens in the study (oral conjugated estrogens and oral ethinylestradiol) in attaining a satisfying amount of breast development. Finally, it has been reported in a large cohort study of transgender women (n = 179) that percent changes in gynoid and android fat, total body fat and total lean body mass were not statistically different between the oral and transdermal estradiol groups if BMI and age were controlled for (Klaver et al., 2018).
Estrogen replacement, being a necessary therapy for the vast majority of individuals with Turner syndrome, has also been studied in adolescent girls (Gravholt et al., 2017; Klein & Phillips, 2019). Girls treated with low dose oral estradiol (n = 56) were described in one study as having “similar” breast development to the normal Dutch population (Bannink et al., 2009). Other studies of puberty induction therapy have found that patients using low doses of transdermal estradiol gel (n = 21) and low dose intramuscular estradiol cypionate (n = 14) also all achieved breast Tanner stage 4 or 5 at final follow-up (Piippo et al., 2004; Rosenfield et al., 2005). A small randomised controlled trial of hypogonadal girls (n = 12) demonstrated that the response to oral and transdermal estradiol at comparable doses was near identical (Shah et al., 2014). All the girls receiving bioidentical estrogens achieved Tanner stage 3 or greater after 18 months of treatment, irrespective of route of administration. Interestingly, a cross sectional study of breast development in women with differences of sex development (DSD), including those with Turner syndrome, reported that breast satisfaction in the sample group was much lower than in women without a DSD (van de Grift & Kreukels, 2019). Some studies have found that breast development, in addition to breast satisfaction, seems to be poorer in Turner syndrome girls than in normal cisgender girls (Guo et al., 2019). Nevertheless, a recent review concluded that all these different regimens seemed to result in similar feminising outcomes (Klein et al., 2018).
In summary, current clinical evidence appears to show no difference in objectively measured outcomes between therapy with different routes of administration when the doses have comparable potency. Rather, when taken together, these findings indicate that the extent of breast development and other feminisation is independent of what route of administration is used (excerpts).
In the past, estrogens in general have been associated with a greater overall incidence of adverse cardiovascular and thromboembolic events (Kuhl, 2005). These events can include deep vein thrombosis and myocardial infarction. Such complications of therapy have been attributed to estrogenic activity in the liver which, at sufficient exposure, causes an increased synthesis of liver proteins such as as sex-hormone binding globulin (von Schoultz et al., 1989; Ockrim, Lalani & Abel, 2006). Synthesis rates of lipids and coagulation factors have also been found to change. However, the type and route of administration of estrogen has been shown to modify risk (Olié, Canonico & Scarabin, 2011; Oliver-Williams et al., 2018).
Synthetic and non-bioidentical estrogens are more resistant to enzymatic metabolism by the liver and have disproportionate estrogenic effects relative to bioidentical estrogens such as estradiol (Kuhl, 2005). Because of this behaviour, they contribute to a much greater synthesis of liver proteins and are associated with a significantly higher risk of venous thromboembolism and other cardiovascular complications (Henriksson & Edhag, 1986; Kuhl, 2005; Lycette et al., 2006). A 2015 retrospective case-control study found that venous thromboembolism was 2 to 5 times more common in young women using combined oral contraceptives containing ethinylestradiol and other synthetic progestins than in non users (Vinogradova, Coupland & Hippisley-Cox, 2015). In 2019, the same authors published another case-control study; this time investigating women receiving hormone therapy at the menopause (Vinogradova, Coupland & Hippisley-Cox, 2019). A key finding was that low doses of oral estradiol (2 mg/day or less) were associated with a slight but significant increase in the incidence of venous thromboembolism, while low transdermal doses (100 μg/day or less) were not. This has also been reported by the ESTHER case-control and E3N cohort studies (Scarabin, 2014). Therefore, a strong advantage of transdermal estradiol over oral estradiol is that the incidence of venous thromboembolism is lower (Files & Kling, 2020). As with the synthetic estrogens, this difference is thought to be attributable to the disproportionate amount of estrogenic exposure in the liver that occurs with oral administration (Olié, Canonico & Scarabin, 2011). Nevertheless, high dose polyestradiol phosphate (160 to 240 mg/month) administered by intramuscular injection has been associated with significantly increased cardiovascular and thromboembolic morbidity and mortality in at least one large study of prostate cancer patients (Mikkola et al., 2005; Mikkola et al., 2007). While the increased incidence of these adverse events is clearly much lower than with oral estradiol, it is much less clear if it may be entirely eliminated by non-oral routes of administration at higher doses (Sam, 2020).
It is difficult to accurately determine the incidence of venous thromboembolism in transgender people receiving hormone therapy because of the diverse range of regimens employed in different geographical regions; which may confer different risks (Goldstein et al., 2019). Moreover, although a number of observational and retrospective studies have reported risk as low or relatively insignificant in our community, most are not adequately powered to accurately report risk (Khan et al., 2019). Based on the available evidence, we can probably safely assume that the incidence is low overall with modern regimens (Getahun et al., 2018; Ott et al., 2010; Pyra et al., 2020). It is particularly of note that these complications are, thankfully, mostly confined to people at higher baseline risk such as elderly individuals or those with inherited mutations that predispose to such toxicity (Silverstein et al., 1998; Bezgin et al., 2016). The absolute risk is likely low for most people. Nonetheless, the association between estrogens and adverse cardiovascular and thromboembolic events is of obvious importance. In the future, I intend on posting a more comprehensive evaluation of these complications with different doses and formulations of estrogens.
In conclusion, oral and transdermal estradiol is metabolised differently. Perhaps most significantly, a dose of oral estadiol is predominantly converted by the liver into estrone and other estrogen metabolites before it enters circulation. By contrast, transdermal estradiol bypasses the liver and the conversion of the medication into these weak estrogens is mostly avoided. On average, a transdermal patch that delivers a 50 μg/day dose is thought to have similar estrogenic potency to a 1 to 2 mg/day dose of oral estradiol and to a 1.5 mg/day dose of transdermal gel.
In spite of these difference, there appears to be no evidence that oral estradiol provides more effective feminisation than transdermal estradiol or vice versa if the doses are similar. Instead, the existing clinical evidence seems to show that the extent of feminising changes such as breast development and fat distribution is independent of the route that estradiol is administered by. Contrariwise, there is a good amount of epidemiological evidence that oral estradiol is associated with a higher incidence of venous thrombosis than is transdermal estradiol at a comparable dose. For this reason, transdermal estradiol at physiological doses is likely safer than oral estradiol in long term gender affirming hormone therapy. I intend to discuss, in more detail, the subject of cardiovascular and thromboembolic toxicity with different doses and types of estrogens in a future review.