Link Search Menu Expand Document

A Review of Studies on Estradiol Levels and Testosterone Suppression with High-Dose Transdermal Estradiol Gel and Ointment

By Aly W. | First published June 30, 2019 | Last modified December 22, 2020

Abstract / TL;DR

Studies were conducted in the 1970s and 1980s on high-dose transdermal estradiol gel and ointment in men with prostate cancer. Estradiol at 3 to 20 mg/day appeared to give 84 to 473 pg/mL estradiol and 105 to 260 ng/dL testosterone in these men. However, the findings of the studies are only so useful due to various unfortunate limitations. In any case, transdermal estradiol gel and related forms as a method of high-dose estradiol monotherapy don’t seem to be well-suited for achieving the high levels of estradiol that are needed for robust testosterone suppression with this approach. Transdermal estradiol patches or genital application may be more useful options on the other hand.


In the late 1970s and early 1980s, high-dose transdermal estradiol gel/ointment was assessed by French researchers in three clinical studies for the treatment of prostate cancer in men (Steg & Benoit, 1979; Steg et al., 1979; Steg et al., 1980; Steg et al., 1983; Steg & Benoit, 1983; Benoit, 1985). The dosages used in the studies ranged from 3 to 20 mg/day. Since there’s not a lot of available information on estradiol levels or testosterone suppression with high-dose transdermal estradiol gel and similar formulations, I opted to do an article reviewing the relevant details from the studies. These studies appear to be the only clinical studies of high-dose transdermal estradiol gel/ointment for prostate cancer that have been published. This is in contrast to high-dose transdermal estradiol patches and parenteral estradiol ester injections, which have been much more widely used.

Data and Excerpts

A Note on Estradiol Delivery Rates

The authors use “γ” (“gamma”) as a unit in their papers. This is a now-deprecated non-SI unit of mass equal to 1 μg (Wiki). That is, “1 γ” means the same thing as “1 μg”. The researchers provided values using this unit as an estimate of the amount of estradiol actually absorbed into the body after taking into account the “cutaneous absorption coefficient” (i.e., transdermal bioavailability) of estradiol. However, when compared to established μg/day values for commercial transdermal estradiol patches and the estradiol levels associated with these patches, the values in the papers are very different from what would be expected. For this reason and the fact that the estimates for estradiol patches are more likely to be accurate, the delivery rates reported by the authors probably should be given little credence. It was important to include them in this article nonetheless however because the researchers altered their transdermal estradiol formulations between studies, resulting in inconsistency between the estradiol mg doses in gel/ointment and their estimated μg/day delivery values.

Study 1: Estradiol Gel 3 mg/day versus 6 mg/day

The following three papers pertain to the first study:

  1. Steg, A., & Benoit, G. (1979). Percutaneous 17 β-estradiol in treatment of cancer of prostate. Urology, 14(4), 373–375. [DOI:10.1016/0090-4295(79)90083-9] [PDF]
  2. Steg, A., Benoit, G., Limouzin-Lamotte, A., Mahoudeau, J., Caillens, M., & Raichvarg, D. (1979). Cancer de la prostate: effets métaboliques des bêta-estradiol par voie percutanée. [Cancer of the prostate: metabolic effect of percutaneous beta-estradiol. / Prostatic carcinoma: metabolic effect of percutaneous beta-estradiol.] La Nouvelle Presse Medicale, 8(46), 3801–3802. [Pascal and Francis] [PubMed] [PDF]
  3. Steg, A., Benoit, G., Limouzin-Lamotte, A., Mahoudeau, J., Caillens, M., & Raichvarg, D. (1980). Cancer de la prostate: Effets métaboliques du bêta-estradiol par voie percutanée. [Cancer of the prostate: metabolic effects of percutaneously administered beta-estradiol.] Revue Médicale de la Suisse Romande, 100(11), 895–897. [Pascal and Francis] [PubMed] [PDF]

In the study, 21 men with prostate cancer were divided into two groups and treated for 1 month with:

  1. 3 mg/day (“300 μg/day”) estradiol in 5 g/day gel applied to the abdomen in one daily dose
  2. 6 mg/day (“600 μg/day”) estradiol in 10 g/day gel applied to the abdomen in two daily divided doses

Hormone levels in the study were as follows:

Estradiol levelsTestosterone levels
Pre-treatment23.2 ± 10.8 pg/mLPre-treatment341 ± 180 ng/dL
E2 gel 3 mg/day84.2 ± 54.3 pg/mLE2 gel 3 mg/day188 ± 125 ng/dL
E2 gel 6 mg/day184.7 ± 98.46 pg/mLE2 gel 6 mg/day105 ± 113 ng/dL
Estrone levelsDHT levels
Pre-treatment37.6 ± 16.2 pg/mLPre-treatment36 ± 27 ng/dL
E2 gel 3 mg/day86.7 ± 42.5 pg/mLE2 gel 3 mg/day23 ± 14 ng/dL
E2 gel 6 mg/day150.9 ± 101.4 pg/mLE2 gel 6 mg/day12 ± 12 ng/dL
LH levelsFSH levels
Pre-treatment6.21 ± 6.2 mIU/mLPre-treatment7.56 ± 4.9 mIU/mL
E2 gel 3 mg/day4.15 ± 3.3 mIU/mLE2 gel 3 mg/day6.44 ± 5.0 mIU/mL
E2 gel 6 mg/day3.41 ± 3.2 mIU/mLE2 gel 6 mg/day4.13 ± 5.0 mIU/mL

The researchers stated that the testosterone levels resulting with 6 mg/day estradiol gel were similar to those with 3 mg/day oral diethylstilbestrol. However, this is not quite accurate as 3 mg/day oral diethylstilbestrol consistently suppresses testosterone levels to the castrate range (≤50 ng/dL) (Wiki).

Study 2: Estradiol Ointment 10 mg/day versus Diethylstilbestrol Oral 3 mg/day

The following paper pertains to the second study:

  • Steg, A., Benoit, G., Maisonneuve, P., Tallet, F., Nahoul, K., Sultan, Y., Raichwarg, D., & Limouzin-Lamotte, M. A. (1983). Étude comparative du diéthylstilboestrol et du 17 bêta-oestradiol per-cutané dans le traitement du cancer de la prostate. [A comparative study of percutaneous 17 beta-estradiol and diethylstilbestrol in the treatment of prostatic cancer.] Annales d’Urologie, 17(4), 197–202. [Google Scholar] [PDF]

In the study, 56 men with prostate cancer were divided into two groups and treated for 3 months with:

  1. 10 mg/day (“600 μg/day”) estradiol in “ointment” applied to the abdomen in two daily divided doses
  2. 3 mg/day oral diethylstilbestrol in three daily divided doses

Hormone levels in the study were as follows:

Estradiol levelsTestosterone levels
Pre-treatment (E2 group)30 pg/mLPre-treatment (E2 group)450 ng/dL
E2 ointment 10 mg/day107 ± 81 pg/mLE2 ointment 10 mg/day180 ± 160 ng/dL
Pre-treatment (DES group)26 pg/mLPre-treatment (DES group)420 ± 130 ng/dL
DES oral 3 mg/day19 pg/mLDES oral 3 mg/day51 ± 0.9 ng/dL
LH levelsFSH levels
Pre-treatment (E2 group)3.2 mIU/mLPre-treatment (E2 group)4.7 mIU/mL
E2 ointment 10 mg/day2.0 ± 0.9 mIU/mLE2 ointment 10 mg/day1.7 mIU/mL
Pre-treatment (DES group)4.0 ± 3.2 mIU/mLPre-treatment (DES group)5.8 mIU/mL
DES oral 3 mg/day1.6 ± 1.2 mIU/mLDES oral 3 mg/day1.7 mIU/mL

(Some SDs (±) are missing above because not all were provided. The SDs above were actually reported in Study 3.)

Some noteworthy translated excerpts from the paper:

This study shows that the clinical effects were more dramatic in the DES group, with a sharper drop in the free plasma testosterone level, than in the E2 group.

E2 has been administered too weakly because it is too low in the commercially used ointment and because application to the thick abdominal skin has a lower absorption coefficient than a thinner skin such as forearms for example. It is possible that a dose equivalent to 15 or 20 mg of gel, applied on the forearms, may allow to lower the testosterone below [100 ng/dL]. The marketing of a more concentrated ointment is therefore desirable, as well as the use of a more efficient absorption site.

The search for new estrogens and new routes of administration is therefore necessary: E2, a natural female hormone, induces no thromboembolic events. It seems to us, therefore, that a larger dose must be proposed to be as effective as 3 mg DES. A gel twice more concentrated is therefore currently under study [(see Study 3 below)].

It should be noted that the transdermal estradiol formulation in this study is different from that used in Study 1. Specifically, “6 mg/day estradiol gel” was stated to deliver “600 μg/day estradiol” in Study 1, whereas “10 mg/day estradiol ointment” was stated to deliver “600 μg/day estradiol” in Study 2. These discrepancies might have relation to the unexpectedly lower estradiol levels and testosterone suppression in Study 2 compared to Study 1.

Study 3: Estradiol Ointment 20 mg/day versus Bilateral Orchiectomy

The following paper pertains to the third study:

  • Steg, A., & Benoit, G. (1983). Étude comparative de fortes doses d’oestradiol-17 β administrées par voie per-cutanée et de l’orchidectomie bilatérale dans le traitement du cancer de la prostate. [Prostatic carcinoma. Bilateral orchiectomy versus percutaneous administration of large doses of 17 β-estradiol. A comparative study. / Comparative study of percutaneous administration of large doses of 17β œstradiol and bilateral orchidectomy in prostatic carcinoma.] Annales d’Urologie, 17(5), 286–288. [Pascal and Francis] [PDF]

In the study, 38 men with prostate cancer were divided into two groups and treated for 1 month with:

  1. 20 mg/day (“1200 μg/day”) estradiol in an ointment applied to the skin in four daily divided doses
  2. Bilateral orchiectomy (i.e., surgical removal of the testes)

Hormone levels in the study were as follows:

ConditionEstradiol levelsTestosterone levelsLH levels
Pre-treatment (E2 group)52 ± 30 pg/mL460 ± 230 ng/dL2.6 ± 0.9 mIU/mL
E2 ointment 20 mg/day473 ± 375 pg/mL260 ± 160 ng/dL1.7 ± 0.8 mIU/mL
Pre-treatment (orchi group)Not reported500 ± 160 ng/dL5.5 ± 6 mIU/mL
Bilateral orchiectomyNot reported20 ± 10 ng/dL19 ± 8 mIU/mL

Some noteworthy translated excerpts from the paper:

The purpose of this study was to determine whether hormonal efficacy [in the treatment of prostate cancer] was enhanced by the administration of large doses of E2. […] The study shows that, whereas orchiectomy lowers plasma testosterone levels dramatically, E2, at [the double dose used in this study compared to our previous study] (1200 [μg] versus 600 [μg]), surprisingly does not further lower it at all, but plasma E2 is substantially increased.

We have found that the administration of E2 causes significant biological changes; E2 is well-absorbed, testosterone is lowered, and LH and FSH are also lowered. But when comparing these effects with those obtained by castration, and with those also observed after administration of 3 mg of diethylstilbestrol (in our previous study), the results are much less brilliant. Indeed: the plasma testosterone is moderately lowered by E2 whereas the fall is much more pronounced with DES and is spectacular with castration; pituitary suppression is comparable with E2 and DES and of course orchiectomy increases LH; and if we compare the results obtained before and after treatment, it is clear that if before treatment the figures are comparable, after treatment the fall is pronounced only with DES and especially with orchiectomy.

The objective of this study was to determine whether doubling the dose of E2 enhances its antiandrogenic effect. [With both a lower dose and a higher dose], E2 is very well-absorbed. [With both doses], the pituitary suppression is identical, but on the other hand, not only is plasma testosterone not lower with [the higher dose], but it seems to be even [higher than with the lower dose]. This difference is also found clinically since with [the higher dose], 10% of patients were improved while with [the lower dose], 30% improvement was obtained.

How can this result be explained? Perhaps this is due to the contrary effects of E2 on androgens. Indeed: E2 on the one hand suppresses pituitary activity, which leads to a fall of testicular testosterone secretion, and at the same time E2 decreases 5α-reductase activity in the prostate; on the other hand, E2 causes an increase in SHBG and competes with free testosterone binding to SHBG and in this way increases the proportion of active free testosterone. In addition, testosterone also increases the level of cytoplasmic androgen receptors.

This work shows that E2 is well-absorbed percutaneously and causes practically no cardiovascular events. On the other hand, it appears that the antiandrogenic effect is significantly less by this route than that obtained by diethylstilbestrol and castration, and that in any case, we cannot improve the efficacy by increasing the dose of E2.

The finding that the 20 mg/day estradiol ointment resulted in less suppression of testosterone levels than the 10 mg/day estradiol ointment despite markedly higher estradiol levels is surprising. However, I don’t agree with the researchers’ interpretation and explanation of the result. Other clinical research of high-dose estradiol for prostate cancer (e.g., transdermal patches, polyestradiol phosphate injections, estradiol undecylate injections) extensively contradicts the finding. Rather, I think that something may have been awry with the researchers’ testosterone assay, or perhaps some other methodological issue was responsible. Studies have consistently shown that estradiol levels of 200 to 500 pg/mL and above suppress testosterone levels by 90% and greater (to 50 ng/dL and below) (Wiki; Wiki-Graphs).

The concentration and dosage of estradiol ointment in this study were double those of Study 2, yet the estradiol levels measured in this study were almost 5-fold higher than those in Study 2. It’s possible that the higher concentration of ointment used may have resulted in disproportionately greater absorption, as one study found that the smaller the application area of estradiol gel (and hence higher the post-application concentration), the greater the resulting estradiol levels (Wiki-Graph). On the other hand, it’s possible that something was off with their assays (c.f. testosterone above). Lastly, in contrast to the previous two papers, the authors don’t state where the ointment was applied in this study. It was probably the abdomen similarly, but this isn’t certain; they might have used one or more other locations with greater skin permeability (as they discussed the possibility of doing in Study 1). You would expect them to have mentioned this change though, which perhaps makes the possibility unlikely.

Review Article

A review on hormone therapy for prostate cancer was also published by one of the researchers:

  • Benoit, G. (1985). Que Penser du Traitement Hormonal du Cancer de la Prostate / Hormonothérapie du Cancer de la Prostate. [Thoughts on the Hormonal Treatment of Prostate Cancer / Hormone Therapy of Prostate Cancer.] Gazette Médicale, 92(5), 33–39. [PDF] [Translation]

In the paper, she briefly mentioned the studies on high-dose transdermal estradiol gel/ointment and provided some additional information pertaining to these studies. Here are the relevant translated excerpts from the paper:

In a study in Cochin (16), the efficacy of diethylstilbestrol was 50% of objective responses, but at a cost of 23% of thromboembolic events. These thromboembolic events are similar to the accidents that occur in women taking oral estrogen–progestogens. These are oral synthetic estrogens. In a series of diethylstilbestrol-treated prostate cancers, Steg (16) has shown that this treatment induces an increase in triglycerides and a decrease in coagulation factor VIII. Abbou (1) found, on his behalf, an increase in circulating immune complexes in patients with thrombosis under synthetic estrogen–progestogen therapy.

It therefore seems to us no longer possible to prescribe diethylstilbestrol as first intention. To reduce these cardiovascular events, most authors have tried to use natural estrogens administered intramuscularly, subcutaneously, or percutaneously. Cochin’s experience of using a natural estrogen (17-beta-estradiol) percutaneously has shown that this treatment may be effective, but that it is very difficult to administer: one in two patients is unable to apply the treatment correctly and regularly, which greatly reduces the use of such treatment (15).

As an indication, we report several studies made successively in Cochin (16, 17) with diethylstilbestrol, a natural estrogen (17-beta-estradiol), testicular pulpectomy, and an LHRH analogue administered subcutaneously.

The results confirm that diethylstilbestrol causes thromboembolic events. These different estrogenic treatments promote gynecomastia. Medical castration by LHRH analogues or surgical pulpectomy causes hot flushes. All these treatments reach, in a variable way, libido and sexual potency. This last effect is complex: it is known that castration, in young men, after puberty, does not regularly cause impotence, as shown by the history of castrates. However, it is necessary to warn the patient that each of these treatments can reach his manhood.

Synthetic estrogens, especially diethylstilbestrol, should no longer be used as first-line agents in the treatment of metastatic prostate cancer because of their thromboembolic risk. Natural estrogens do not seem efficient enough to be used regularly. Complete castration performs excessive mutilation. In our opinion, only testicular pulpectomy, LHRH analogues, and antiandrogens remain in competition.

The review also contains the following adapted table, with data from their studies:

 DESE2 gel/ointmentOrchiectomyGnRH analogue
Testosterone levels50 ng/dL100 ng/dL20 ng/dL50 ng/dL
50% partial remission50%30%33%41%
Cardiovascular events23%0%0%0%

The gonadotropin-releasing hormone (GnRH) analogue data were from other studies and papers by the same researchers (Steg et al., 1984; Steg et al., 1985). These studies did not assess transdermal estradiol and so are not otherwise discussed in the present article.

The authors’ comments in this paper raise the additional question of whether inconsistent and suboptimal exposure to estradiol gel/ointment may have influenced the findings of their studies.

Summary and Discussion

In the reviewed clinical studies of high-dose transdermal estradiol gel/ointment in men with prostate cancer, the following results were obtained in terms of changes in measured hormone levels:

Estradiol doseEstradiol levelsTestosterone levels
Before treatment23–52 pg/mL341–460 ng/dL
3 mg/day gel (“300 μg/day”)84 pg/mL188 ng/dL
6 mg/day gel (“600 μg/day”)185 pg/mL105 ng/dL
10 mg/day ointment (“600 μg/day”)107 pg/mL180 ng/dL
20 mg/day ointment (“1200 μg/day”)473 pg/mL260 ng/dL

Unfortunately, the data reported in these studies is not as useful as might have been hoped. This is due to inconsistencies with the transdermal estradiol formulations and doses used, reported issues with patient compliance in terms of consistent and correct administration, inconsistencies with the measured estradiol and testosterone levels, and omission of certain important details (e.g., ointment concentrations, gel/ointment compositions and differences, how the estimated estradiol delivery rates were determined, etc.). Hence, there may be only so much that be taken away from the research. With that said however, these studies do contribute to an impression that transdermal estradiol gel is relatively weak in strength even at high therapeutic doses. Other studies of high-dose transdermal estradiol gel/cream (e.g., 3 to 8 mg/day) have consistently reported relatively low estradiol levels similarly (~40 to 220 pg/mL; see the table here).

Based on these studies, transdermal estradiol gel doesn’t seem to be a very effective or affordable way of achieving the estradiol levels that are needed for adequate suppression of testosterone levels in transfeminine people. Or at least not as a high-dose monotherapy. In contrast to transdermal estradiol gel, transdermal estradiol patches can consistently achieve much higher estradiol levels and testosterone suppression in comparison (Wiki). About 100 pg/mL estradiol per 100 μg/day estradiol patch is achieved on average (Wiki). This makes estradiol patches generally a more favorable form of estradiol for transdermal administration than gel and related forms.

In any case, genital application of transdermal estradiol gel and other transdermal estradiol formulations can allow for dramatically higher estradiol levels (e.g., 5- to 8-fold) than application to conventional transdermal skin sites (e.g., forearm, abdomen) (Aly W., 2019; Aly W., 2019). This may provide these transdermal estradiol formulations with greater therapeutic potential.