Intro Header

A journal club style critical review of current publications in evidence-based transgender medical care.

Friday, June 1, 2012

Estrogen replacement in girls with Turner syndrome: What is the optimal estrogen dose and route?

Pharmacokinetic and pharmacodynamic study of various doses oral and transdermal estrogen found that transdermal estradiol 75mcg patches most closely approximated physiologic estrogen levels in agonadal girls with Turner syndrome. (1)

     As is the case in the care of transgender women, the optimal regimen of hormone replacement for girls with Turner syndrome is not well established.  Applying knowledge from Turner syndrome to the cross-sex hormone therapy (csHT) of transwomen is to some degree intuitive; both involve long term estrogen therapy across the lifespan.  Investigators aimed to study the pharmacokinetics and pharmacodynamics of various doses and routes of 17-beta estradiol on 10 girls with Turner syndrome.  Subjects were divided into two groups; one high dose group and one low dose group.  Girls in each group were treated with both transdermal and oral preparations, with a two-week washout period between each regimen (regimens: 75mcg and 37.5mcg patch (Vivelle) changed twice/weekly, 2mg and 0.5mg micronized estradiol tablets (Estrace) once/daily.  Results were compared to integrated mean estrogen levels over one menstrual cycle in a control group of 20 healthy menstruating girls.  Levels of estradiol (e2) and estrone (e1) were measured in all subjects and controls.

     Mean serum estradiol concentrations were closest to those of the control group in the high-dose transdermal group (114 +/- 31pg/ml transdermal, 96 +/- 11pg/ml for controls) for.  Levels in the low dose transdermal and high dose groups were similar.  Levels were lowest in the low dose oral group.  Estrone (e1) concentrations were higher the oral groups, likely because of hepatic and intestinal conversion.  LH/FSH suppression was equal in both high dose groups, but more pronounced in the low dose transdermal vs. oral.  In all cases gonadotropins did not fully normalize (particularly FSH, which responds in part to inhibin, which is absent in hypogonadal states).


     It is unclear if pharmacologic studies in Turner syndrome females are applicable in those with male physiology or genetics.  For example, transwomen’s skin and connective tissue may have different absorption characteristics with respect to transdermal delivery systems.  It is also unclear if targeting physiologic estradiol levels is an appropriate clinical target for medical transition.  While estrone is generally considered to have minimal estrogenic activity, it may serve as an estrogen reservoir and may explain why patients describe feeling “different” on oral vs. transdermal preparations with similar serum levels.  Lastly, measurement of serum estrogens is wrought with difficulties and this study was not adequately powered to draw firm conclusions.


Wednesday, May 30, 2012

Perioperative Hormones: What are Plastic Surgeons Doing?

Survey of the attitudes of British plastic surgeons with respect to menopausal hormone replacement therapy finds a wide range of practices.  (1)

Given the ongoing lack of clear evidence based recommendations regarding the use of perioperative combined oral contraceptives (OCP) or menopausal hormone replacement therapy (HRT) in the context of venous thromboembolism (VTE), the authors distributed a mailed 2004 survey of 285 consultant plastic surgeons in the UK in with respect to their individual practices.  Of the 53% that responded, 50% believed that HRT was a risk factor for VTE but only 20% considered cessation of therapy to their patients.  This is compared with 90% who believed OCP was a VTE risk factor, and 54% considered cessation of this treatment.

     The authors conclude by recommending that barring quality evidence, risks associated with OCP should be extrapolated to HRT, and that all perioperative estrogen of any type should be stopped for 4 weeks before and 2 weeks after any major surgery, any lower limb surgery, or any general anesthesia lasting > 1 hour.


      The authors make several “leaps of faith” in their recommendations.  While they admit that there is little if any evidence linking HRT with perioperative VTE, and also point out that even older (and more thrombogenic) generation OCP preparations from the 1960’s only increase DVT risk from 5/100,000 to 15/100,000 – startling 200% increase to be sure, however an absolute risk still well within reason when operating in the context of informed consent.  It is asserted that anesthesia >1 hr or age >40 raises an individual VTE risk to “intermediate” – which should serve to remind us more about the background morbidity associated with major surgery rather than cause us to attribute all postoperative VTE in transwomen to estrogen treatment (in other words, people get VTE’s postoperatively, estrogen or no estrogen).  

     Oddly, the British National Formulary states that “evidence associating [HRT with] an increase in DVT is questionable”, yet goes on to list thrombophlebitis and thromboembolic disorders as contraindications to usage and (even though they are very different medicines) generalizes recommendations for OCP’s to HRT as well.  In any case, this practice pattern survey with a 53% response rate should serve more to identify the need for evidence-based guidelines rather than serve to inform best practices.


Estrogens in the Perioperative Period: Remarks from the Ansesthesiology Literature

2001 review article is unfortunately one of the more recent publications on the subject.  (1)

     The perioperative use of estrogen cross-sex hormone therapy (csHT) is a topic of controversy in both transgender women as well as in cisgender women taking menopausal HRT or combined oral contraceptives (OCP).  Little data exists in any context with respect to perioperative venous thromboembolic risks (VTE) in those patients taking estrogen of any type and for any reason.  Most guidelines are based on consensus opinion and of varying vintage.

     This review article aimed to review perioperative estrogen therapy from the reference frame of the anasthesiologist.  While the article reviews a number of conditions of interest, the section focusing on perioperative VTE will be discussed here.

     It is first mentioned that studies of VTE in estrogen therapy of any kind suggest relative risk of 2-3 for users vs. non-users.  While these studies are still equivocal and suggest that the risk is greatest in the first year of therapy, the 2-3x increase in risk likely translates into a real world increase of 1-2 cases per 10,000 women per year.  Furthermore, there may be a subset of users at baseline higher risk due to occult prothrombotic states, which have a 6.2% baseline prevalence in the general population, but a 60.2% prevalence in those with a personal or family history of VTE.

     The authors conclude that no evidence-based recommendations can be made with respect to withholding perioperative estrogen therapy, nor with respect to postoperative VTE prophylaxis with heparin.  Recommendations from various British bodies and organizations are presented showing a range of conflicting and equivocal recommendations.  Providers are advised to rely on clinical judgment.   


     This 11 year old article unfortunately represents one of the more recent publications on the issue of perioperative VTE and estrogen therapy; randomized controlled trials are sorely lacking.  In the discussion the author mentions that withholding perioperative OCP may result in an unwanted pregnancy with attendant risks and morbidity; it is recommended that physicians consider these risks when analyzing the pros and cons on an individual basis.  The same thought process should be applied for transwomen, who may experience significant emotional symptoms from hormone withdrawal in what is for many an already extremely stressful event.  In the case of transwomen, it is likely safe to continue csHT in the perioperative period if they have no personal or family risk factors, have been on csHT for > 1 year, wear sequential compression devices postoperatively, and have received informed consent about the minimal absolute risk increase.  Providers should also keep in mind the background rate of VTE in major pelvic surgery without the use of estrogens.


1) Brighouse, D.  Hormone Replacement Therapy and Anaesthesia.  British Journal of Anaesthesia 86(5): 709-16.

Reducing Pain Associated with Pelvic Speculum Exams

Use of gel vs. water resulted in a slight reduction in pain score on pelvic speculum exams in cisgender women without interfering with cervical cytology results. (1)

     Pelvic examinations can be uncomfortable for transgender men for a variety of reasons.  Virginal status or infrequent penetration, vaginal atrophy due to hormone therapy, and psychological factors may impair or even prevent a successful pelvic examination.  Patients may have had past experiences that were painful and may be reluctant to permit necessary pelvic screening or care.

      Cisgender women undergoing pelvic examinations also may experience discomfort.  Historically providers have been hesitant to use lubricating gel when collecting cervical cytology or chlamydia testing specimens out of concern that gels may interfere with test results, and have instead used water for speculum lubrication.  However a growing body of evidence suggests that certain lubricant products that do not contain carbomers do not interfere with such tests.  Investigators aimed to evaluate the pain associated with pelvic speculum examinations using gel lubricant as compared to water.

      This was a single-blind randomized controlled trial of 119 women assigned to either a gel (n=59) or water (n=60) group.  Pain scores were recorded on a 0-10 visual analog scale (VAS) after insertion of the speculum, but before visualization of the cervix.  Study size was determined based on a power calculation to detect a VAS difference of 0.9, which has been shown in the literature to be a threshold of clinical significance.  Menopausal women, pregnant or recently parous women as well as those with pelvic pathology were excluded.  The overwhelming majority of subjects were patients presenting for routine pelvic screening examinations.  A single physician performed all examinations.

      The gel group reported a statistically significant lower VAS score than the water group (1.41 vs. 2.15, difference 0.74).  33.9% in the gel group reported a score of 0, vs. 10% in the water group.  None of the 35 gel or 38 water patients who underwent cytology had unsatisfactory results.


     The authors identified a statistically significant reduction in VAS scores when using gel vs. water lubricant in routine pelvic examinations, with no observed cases of unsatisfactory cytology results.  The findings do not cross the threshold of clinical significance.  However these findings may have applicability in transgender men who may have baseline higher pain scores than the study group seen here; pain scores of 1-2 are remarkably low.  As one’s pain climbs on the pain scale, non-linear differences may be encountered which could cross the threshold of clinical significance.  In the very least clinicians should be reassured by this and other data that using a small amount of gel lubricant for pelvic examinations is reasonable and will not impact cytology results if the gel used does not contain carbomers.  Further study in populations more likely to experience pain on pelvic speculum examination (transgender men, menopausal women) would be useful.


1) Hill DA, Lamvu G. Effect of lubricating gel on patient comfort during vaginal speculum examination: a randomized controlled trial. Obstetrics & Gynecology. 2012;119(2, Part 1):227

Transdermal Estradiol Reduces Risk of Venus Thromboembolism in Menopausal Hormone Replacement Therapy

Matched cohort study looks at risk reduction of using transdermal estrogen vs. oral in menopausal hormone therapy. (1) 

     In this matched cohort study of incidence of VTE in women using menopausal HRT, investigators attempted to quantify the risk reduction for VTE events associated with transdermal relative to oral estrogens.  Claims records from a health insurance database of 30 million people living in the US were reviewed over the period of January 2002 - October 2009 for patients who had received at least two dispensings of a transdermal or oral estradiol and were over age 35.  Users of other estrogens such as vaginal or injections and those with a prior VTE were excluded.  At total of 27,018 women were enrolled in each cohort.  Interestingly while only 22.4% of women in each cohort had undergone hysterectomy, a minority of the women were using a progestagen.  Primary outcome measures were the overall VTE risk as well as hospitalization-related VTE, as measured by incidence rate.

     Mean drug exposure times were 391 days for transdermal estrogen and 401 days for oral estrogens.    A total of 115 VTE events were encountered in the transdermal group vs. 164 in the oral group; a statistically significant difference.  Further analysis revealed more pronounced statistically significant reductions in VTE incidence among hospitalized women and reductions in PE incidence.

     Yet another study highlighting the likely benefits of transdermal HRT with respect to VTE risk.  While this study did not compare incidence to a control population, overall VTE incidence in both cohorts was slightly higher than reported baseline incidences.  This raises the possibility of an overall skewing of the sample population, though it may not affect internal validity of the results.  It is noteworthy that the average times on HRT was just over 1 year in both cohorts;  given that the risk of VTE is highest in the first year of HRT, this may explain such skewing.  Further studies should compare oral vs. transdermal VTE incidences in those on longer term HRT when the overall VTE risk is lower, and specifically in transgender populations.  

     It is possible that there might be a skewing of results toward higher incidences of VTE as the investigators included all oral estrogen forms including conjugated equine estrogens in the oral therapy arm.

Transdermal Estrogen Safe With Respect to Venous Thromboembolism?

Oral estrogen shows increased risk of VTE, but transdermal does not in a French prospective cohort. (1)

     Several studies have suggested that route of estrogen administration can affect risk of venous thromboembolism (VTE) in women using postmenopausal hormone replacement therapy (HRT).  The authors reviewed data from a prospective cohort of French teachers born between years 1925 and 1950 who completed biennial self-administered questionnaires on a range of health issues.  In those who self-reported a thromoboembolic event, follow-up questions as well as a query of the subject’s physician(s) futher populated the data on such events.  Death certificates were reviewed for deceased subjects and revealed 68 additional events (all PE). Thromboembolic events (deep vein thrombosis (DVT) or pulmonary embolism (PE)) were included if they were diagnosed by an imaging procedure.  Exclusion criteria for events included superficial vein thromboses, upper-extremity DVT, central retinal vein obstruction and recurrent events.  Subject exclusion criteria were history of cancer, VTE prior to study period, recurrent VTE, or nonidopathic VTE. 

     A total of 549 incidents of VTE (415 DVT, 134 PE) were identified in 80,308 women who met inclusion criteria out of the total study enrollment of 98,995.  Of those 549 women with a VTE during the study perior, 181 were never users, 66 were past users, and 255 were current users of HRT.  An additional 47 women were using “other” or “unknown” treatments.  Progestagens were used in all but 26 of the women using HRT (micronized progesterone – 47, pregnane deriviatives – 91, norpregnane derivatives – 69, nortestosterone derivatives – 22).

     Mean age at the start of follow-up period was 54, with women followed over an average of 10.1 years.  The hazard ratio for VTE with oral estrogen use was 1.7 (95% CI: 1.1-2.8) and with transdermal estrogens was 1.1 (95% CI: 0.8-1.8).  Use of a norpregnane derivative was independently associated with an increased risk of VTE (HR 1.8, 95% CI: 1.2-2.7), while other progestagens were not. 


     The investigators have obtained results consistent with prior studies (ESTHER) that suggest transdermal estrogens with or without a non-norpregnane progestagen may be safe with respect to VTE.  The current study has a number of limitations; it was a self-reporting survey with a non-trivial exclusion of almost one-fifth of respondents.  While 9.9% of respondents reported tobacco use (a known risk factor for VTE in HRT), the authors did not report prevalence of smokers in each arm of the study.  Since this was simply an observational cohort study, patients at higher risk of VTE based on their individual physician’s clinical impression may have guided the use of HRT type and route or may have prevented the use of HRT all together.  Lastly the study was funded by the manufacturer of a transdermal estrogen product. Nevertheless this study is yet another in a growing series of investigations which suggest that estrogen route and progestagen type can impact VTE risk.  While these results can not be generalized to transgender populations, they are suggestive and are helpful for guiding future research.  

     It is possible that there may be a skewing towards a higher incidence of VTE as the investigators did not separate out users of 17-beta estradiol vs. conjugated equine estrogens (Premarin).



Transgender Men and HIV

Of 59 transmen presenting for HIV screening at public health clinics in San Francisco, none tested positive.  However the sample may not be representative of the population. (1)    

     Transgender men (transmen) may be at risk for HIV infection.  Transmen may have male sexual partners and therefore be identified as men who have sex with men (MSM).  Transmen may also have sex with transgender women (transwomen).  HIV prevalences in San Francisco, USA have been found to be 24% among MSM and 35% among transwomen.  Historically only a few cases of HIV in transmen have been reported in San Francisco.  The authors aimed to identify the prevalence of HIV among those transmen presented to one of 17 public health clinics in San Francisco during 2009.  These sites perform 25% of all HIV tests in the city.

     A total of 64 HIV tests were performed on 59 unique self-identified transmen in 2009 at these sites, representing 0.4% of all tests performed.  Median age was 27, with 79% ages 18-37.  Whites comprised 73% of the sample and 64% identified as gay/lesbian/queer.  With respect to sexual practices in the prior 12 months, 61% had a male sexual partner (51% vaginal receptive, 19% anal insertive, 39% anal receptive), 63% had female partners (55% vaginal or anal sex) and 47% had transgender partners (42% vaginal or anal sex); 44% had 6 or more sexual partners in the prior year.  Unprotected vaginal or anal sex was reported by 63% in the past 12 months, with 10% reporting having had vaginal or anal sex with a known HIV+ person.  None of the subjects tested positive for HIV (98% negative, 2% missing).


     This study demonstrates a number of interesting findings.  First it confirms what is already anecdotally well known; that transmen have a range of sexual partners, practices and orientations.  It also demonstrates an absence of HIV in a population of transmen who are actively engaging in a variety of high risk behaviors.  This absence suggests that transmen are still in a “pre-epidemic” phase with respect to HIV and that there is substantial opportunity to develop HIV prevention measures targeting this population.  It also raises the question if there are factors (anatomical, functional, behavioral) that alter transmale succeptability to HIV infection to a level lower than that of other populations. 

     The study also has a number of limitations.  The study population was primarily white transmen between the ages of 18-35, and only comprised of 25% of all HIV tests performed in San Francisco during the study period.  Barriers may exist for younger transmen or transmen of color to accessing HIV testing at any location.  Older transmen may be more likely to have insurance and a primary care provider and seek testing at a private office.  Some transmen may not self-identify themselves when completing a survey.

     The study demonstrates the complexities and pitfalls of data collection on transgender populations.  The survey used contains numerous flaws and oversights with respect to demographics and sexual practices.  For example, sexual activity with a transgender person is not broken down by transmale/transfemale.  It is unclear if respondents identified transgender partners by their affirmed gender, their transgender status, or both.  It is assumed that by “man” and “woman” the researchers intended to refer to cisgender men and women – such assumptions are counter to the “two-step” recommendation for data collection (Step 1: “What was the sex you were assigned at birth?”; Step 2: “What is your gender identity?”) which allows a more complete and culturally sensitive collection of data.  Anal and vaginal sex with a transgender person or a woman (presumably cisgender) were grouped together and not qualified by insertive or receptive relative to the subject.  Further study with a greater level of detail with respect to demographics and sexual practices as well as a wider catchment of age ranges and ethnicities is essential to identify and manage HIV risks in transgender men.


1) Chen S, McFarland W, Thompson HM, Raymond HF. Transmen in San Francisco: What do We Know from HIV Test Site Data? AIDS and Behavior. 2010 Dec 9;15(3):659–62