Efficacy and Safety of Pubertal Induction Using 17β-Estradiol in Transgirls

Abstract

Context:

Puberty suppression using gonadotropin-releasing hormone agonists, followed by induction of the desired sex characteristics using sex steroids, has been recommended by the current guidelines as the treatment of choice for gender dysphoric adolescents, although little evidence is available.

Aim:

To evaluate the efficacy and safety of estrogen treatment for pubertal induction in transgirls (female-identifying adolescents assigned male at birth).

Methods:

Twenty-eight adolescents treated with oral estrogen for ≥1 year were included. The Tanner stage, anthropometry, laboratory parameters, bone age, and body composition were evaluated.

Results:

Breast development started within 3 months in 83% of adolescents, and after 3 years, 86% had Tanner breast stage 4 to 5. The hip circumference increased and the waist/hip ratio decreased. The median serum estradiol was 100 pmol/L (range, 24 to 380) at the standard adult dose of 2 mg of 17β-estradiol. The adult height standard deviation score was +1.9 (for females). The body mass index standard deviation score, lean body mass percentage, fat percentage, and blood pressure did not change. No abnormalities of creatinine or liver enzymes were detected, and the hematocrit and hemoglobin A1c did not change. One individual developed hyperprolactinemia during high-dose ethinylestradiol treatment to limit growth.

Conclusions:

Pubertal induction using estradiol is effective; however, an adult dose of 2 mg does not always result in appropriate serum estradiol levels. Monitoring renal function, liver enzymes, hematocrit, and hemoglobin A1c during pubertal induction with estradiol is not necessary. Further studies are needed to establish effective and safe methods to limit growth.

Increasing numbers of adolescents with gender dysphoria, characterized by an incongruence between the experienced gender and the sex assigned at birth, are seeking medical treatment. This treatment can consist of puberty suppression, followed by masculinizing or feminizing hormone treatment and gender-affirming surgery as outlined in the Endocrine Society clinical practice guideline “Endocrine Treatment of Transsexual Persons” and standards of care issued by the World Professional Association for Transgender Health (available at: www.wpath.org) (1). Such treatment has been shown to improve psychological functioning in several domains and result in well-being comparable to that of age-matched controls (2). Data on the efficacy and safety of puberty suppression using gonadotropin-releasing hormone analog (GnRHa) treatment were recently published (3). However, few data are available on the outcome of estrogen treatment in transgirls (female-identifying adolescents assigned male at birth). Breast development during estrogen treatment has been studied in adults; however, those findings might not apply to adolescents treated with GnRHa and exposed to testosterone for a much shorter period (4). No consensus has been reached on the optimal estrogen dose; the doses used in adolescents are generally lower than those used in adults. The Endocrine Society guideline, which is currently being revised, has recommended rather extensive safety monitoring, during pubertal induction, including renal and liver function, prolactin, hematocrit, and glucose metabolism (1). However, it is unclear how frequently abnormalities in these laboratory parameters occur.

Therefore, we sought to describe how effective 17β-estradiol (estradiol) treatment is in inducing breast development and female body habitus and to determine whether adequate serum estradiol levels are reached with the dose currently recommended for adolescents. In addition, we describe the effects on growth and body composition and the results from safety monitoring.

Materials and Methods

Subjects and protocol

Gender dysphoric adolescents seen at the Centre of Expertise on Gender Dysphoria at the VU University Medical Centre in Amsterdam from 1998 to 2009 were invited to participate in a study on brain development, brain functioning, growth, and metabolic aspects of their treatment. Gender dysphoria was diagnosed in these adolescents in accordance with the existing guidelines (1), fulfilling the “Diagnostic and Statistical Manual, 4th ed., text revision” criteria for gender identity disorder (5). These adolescents had experienced lifelong extreme gender dysphoria, were psychologically stable, and were living in a supportive environment. The design of the study was observational and prospective. Treatment initially consisted of intramuscular injections of triptorelin (Decapeptyl-CR; Ferring Pharmaceuticals, Copenhagen, Denmark) 3.75 mg every 4 weeks. For transgirls, oral estradiol treatment was added at ~16 years of age. For those who had started triptorelin before age 16 years, the starting dose of estradiol was 5 µg/kg/d, once daily, and was increased by 5 µg/kg/d every 6 months until an adult dose of 2 mg once daily was reached. For those who had started triptorelin at age ≥16 years (and had completed endogenous puberty), the starting dose was 1 mg, which was increased to 2 mg after 6 months. Individuals were seen at 3-month intervals. In a few individuals, a different type or dose of estrogen was used to limit growth; two individuals were treated with 200 µg ethinylestradiol and four with 6 mg estradiol. Only those who had been treated with estrogen for ≥12 months (n = 28) were included in the present study. None discontinued estrogen treatment.

Physical examination

The Tanner stage was determined by the same examiners at each visit and included breast development and testicular volume using a Prader orchidometer (6, 7). Blood pressure was measured using a Dinamap automated device (Critikon Corp., Tampa, FL). Weight and height were measured using an electronic scale and wall-mounted stadiometer (Seca, Hanover, MD) to the nearest 0.1 kg and 0.1 cm, respectively. The waist circumference (smallest abdominal circumference) and hip circumference (largest hip circumference) were measured to the nearest 0.1 cm. The standard deviation scores (SDSs) for height, sitting height, sitting height/height ratio, hip circumference, waist circumference, and waist/hip ratio were calculated using Dutch reference data (8), and the body mass index (BMI) SDS was calculated using reference data from Cole et al. (9). Changes in the SDS were studied to assess whether the changes in anthropometric parameters in transgirls were different from the changes in these parameters in adolescents from the reference population as they grew older.

Laboratory investigations

Every 6 months, blood was drawn for measurement of luteinizing hormone (LH), follicle-stimulating hormone (FSH), testosterone, estradiol, prolactin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), γ-glutamyl transferase, and creatinine. The following assays were used: an immunometric assay for LH and FSH (Delfia; PerkinElmer, Wallac Oy, Finland; lower limit of quantification, 0.3 U/L for LH and 0.5 U/L for FSH) and a radioimmunoassay for estradiol (Diasorin, Saluggia, Italy; lower limit of quantitation, 18 pmol/L). The laboratory participates in external quality control and meets the ISO (International Organization for Standardization) 15189 criteria.

Radiological investigations

Once yearly, a radiograph of the left hand was used to determine bone age according to Greulich and Pyle (10) using male reference photographs. Fat mass, fat percentage, and lean body mass percentage were measured using dual energy X-ray absorptiometry with a Hologic QDR 4500 scanner (Holologic Inc., Waltham, MA).

Ethical approval

The local medical ethics committee granted medical ethical approval, and all participants and their parents or guardians provided informed consent. The study was placed in the International Standard Randomized Controlled Trial Number register and ascribed the registration number ISRCTN 81574253 (available at: http://www.controlled-trials.com/isrctn/).

Statistical analysis

The statistical package used was SPSS, version 22 (SPSS Inc., Chicago, IL). Changes in various (laboratory) parameters between two time points were compared in the individuals with complete data using a paired samples t test, if normally distributed, and a Wilcoxon signed rank test, if not normally distributed. The results were considered statistically significant at P < 0.05.

Results

Twenty-eight transgirls started estrogen treatment at a median age of 16.0 years (range, 13.9 to 18.9) after treatment with triptorelin alone for a median of 24.8 months (range, 6.4 to 51.6). In five individuals, estrogen treatment was started before the age of 15.5 years because of tall stature; all 5 subjects were >180 cm tall at the start of estrogen treatment.

Pubertal development

Breast development had started within 3 months in 15 of 18 individuals for whom data were available and within 1 year in 27 of 27 of subjects. The median Tanner breast stage was 3 (range, 2 to 5) after 1 year, 4 (range, 2 to 5) after 2 years, and 5 (range, 2 to 5) after 3 years. After 3 years, one individual had Tanner stage 2, 1 had stage 3, 3 had stage 4, and 9 had stage 5 breast development. The individuals with Tanner stage 2 and 3 after 3 years had taken the longest to reach the adult dose of 2 mg (30 and 27 months, respectively).

The waist circumference did not change substantially, but the waist circumference SDS decreased from 0.67 SDS to 0.22 SDS (n = 13; P = 0.07; female reference values) during the 3 years of estrogen treatment (Table 1). The hip circumference increased during the first 2 years of estrogen treatment from 93.5 ± 8.2 cm to 97.5 ± 9.0 cm (n = 21; P = 0.003), but the hip circumference SDS did not change. The waist/hip ratio decreased from 0.79 ± 0.04 to 0.75 ± 0.05 (P < 0.001), and the waist/hip ratio SDS also decreased from 0.48 to −0.04 SDS (n = 13; P = 0.002; female references) during the 3 years of estrogen treatment.

Table 1.

Changes in Anthropometric Measures, Bone Age, Body Composition, and Blood Pressure During 3 Years of Estrogen Treatment

Variable	Treatment Duration (y)
	0	1	2	3
Subjects (n)	28	28	21	16
Age (y)	16.0 (13.9–18.9)
Tanner breast stage	1	3 (2–5)	4 (2–5)	5 (2–5)
Testicular volume (mL)	8 (3–25)	8 (3–23)a	10 (3–18)	6.5 (4–11)
Height (cm)	178.0 ± 6.9	181.3 ± 7.1b	181.8 ± 8.4b	180.0 ± 9.4c
Height SDS
Female	1.48 ± 1.11	1.85 ± 1.14b	1.84 ± 1.32b	1.53 ± 1.50a
Male	−0.08 ± 1.15	0.05 ± 1.12a	−0.12 ± 1.24	−0.49 ± 1.36
BMI (kg/m2)	20.8 ± 3.0	21.0 ± 3.3	21.3 ± 3.7	21.5 ± 3.3a
BMI SDS
Female	0.16 ± 1.21	0.05 ± 1.28	0.02 ± 1.35	−0.00 ± 1.35
Male	0.34 ± 1.32	0.12 ± 1.40	−0.03 ± 1.44	−0.14 ± 1.47
WC (cm)	73.9 ± 7.3	73.1 ± 9.0	72.8 ± 8.1	73.7 ± 9.5
WC SDS
Female	0.72 ± 0.89	0.42 ± 1.10a	0.27 ± 1.04	0.22 ± 1.29
Male	0.07 ± 1.10	−0.42 ± 1.41c	−0.75 ± 1.38c	−0.90 ± 1.70c
HC (cm)	93.9 ± 7.8	95.1 ± 8.8	97.5 ± 9.0c	97.4 ± 7.9a
HC SDS
Female	0.38 ± 0.96	0.37 ± 1.07	0.51 ± 1.09	0.42 ± 0.98
Male	0.60 ± 1.16	0.54 ± 1.30	0.70 ± 1.34	0.64 ± 1.26
WHR	0.79 ± 0.04	0.76 ± 0.05b	0.75 ± 0.05b	0.75 ± 0.06b
WHR SDS
Female	0.49 ± 0.68	0.18 ± 0.85c	−0.16 ± 0.86b	−0.04 ± 1.01c
Male	−0.74 ± 0.86	−1.19 ± 1.09c	−1.71 ± 1.06b	−1.48 ± 1.29b
SH/H	0.509 ± 0.013	0.509 ± 0.014	0.513 ± 0.016	0.518 ± 0.158
SH/H SDS
Female	−0.91 ± 0.82	−0.98 ± 0.92	−0.75 ± 1.01	−0.68 ± 0.99
Male	0.15 ± 0.89	−0.27 ± 0.94	−0.07 ± 1.06	0.21 ± 1.12
Bone age (y)	14.3 (13–18)	15.5 (13.5–18)	17 (15–19)	18 (16–19)
Fat percentage	26.0 ± 7.0	24.0 ± 8.7	24.0 ± 9.3	25.9 ± 8.1
Lean body mass percentage	71.0 ± 6.5	72.8 ± 8.2	72.7 ± 8.9	70.9 ± 7.6
Systolic BP (mm Hg)	119 ± 16	121 ± 14	122 ± 15	125 ± 16
Diastolic BP (mm Hg)	69 ± 9	68 ± 9	69 ± 9	66 ± 9

Data presented as mean ± standard deviation or median (range).

Abbreviations: BP, blood pressure; HC, hip circumference; SH, sitting height; SH/H, sitting height/height ratio; WC, waist circumference; WHR, waist/hip ratio.

^a

P < 0.05 compared with baseline.

^b

P < 0.001 compared with baseline.

^c

P < 0.01 compared with baseline.

The testicular volume slightly decreased during the first year of treatment (P = 0.02) and did not significantly change thereafter (Table 1).

Growth and bone maturation

In those with bone age ≤15 years, the median height increase was 3.9 cm (range, 0.9 to 7.3; n = 19) in the first year, 2.4 cm (range 0.6 to 5.8; n = 12) in the second year, and 0.5 cm (range 0 to 2.3; n = 8) in the third year of estrogen treatment [Fig. 1(a)]. Bone age advanced by a median of 1 year (range, 0 to 2) during the first year of estrogen treatment, 1.5 years (range, 0.5 to 3) during the second year, and 1 year (range, 0.5 to 1.7) during the third year [Fig. 1(b)]. Seven individuals had a bone age >15 years at the start of estrogen treatment and grew little (maximum, 3.1 cm in 4 years) or not at all [Fig. 1(c)]. During the study, 25 individuals reached a bone age >15 years, indicating they were near or at their adult height. They had reached an average height of 182.6 ± 7.9 cm, which is −0.17 SDS for adult Dutch males and +1.9 SDS for adult Dutch females. The height gain of these adolescents was related to their bone age at the start of estrogen treatment [Fig. 1(c)]. The sitting height/height ratio and the sitting height/height SDS did not significantly change.

Figure 1.

Growth and bone maturation during pubertal induction in transgirls. (a) Growth during pubertal induction in adolescents with bone age of ≤15 years at the start of treatment. Data shown from transgirls treated with a gradually increasing dose of estradiol (E2; n = 15; solid lines) and from two adolescents treated with 6 mg of estradiol (dotted lines) and two treated with 200 µg ethinylestradiol (EE; dashed lines) to reduce growth. (b) Bone age progression during pubertal induction (assessed according to Greulich and Pyle [10] using male references; n = 28). (c) Height gain from the start of treatment until adult height (last height measurement was considered adult height if bone age >15 years had been reached) vs bone age at the start of treatment (n = 23).

High-dose estrogen treatment to limit growth

Six individuals used high doses of estrogen to limit growth; two were treated with 200 µg/d ethinylestradiol (from 4 and 19 months, respectively, of estrogen treatment) and four with 6 mg estradiol (from 16, 16, 19, and 27 months, respectively). Growth data were available for four individuals during the period of high-dose estrogen treatment (Fig. 1). One transgirl, with a bone age of 14 years at the start of treatment, who had used 200 µg of ethinylestradiol from the fourth month, grew only 1.9 cm during 5 years (Fig. 1). Another transgirl with a bone age of 13 years had grown 7.8 cm during the first 1.5 years of normal dose estradiol treatment, and her bone age had only advanced to 13.5 years after 1 year. After 1.5 years, she was switched to 200 µg/d ethinylestradiol, after which bone maturation accelerated (15.3 years after 2 years and 17 years after 3 years), and growth velocity decreased (2 cm during the next 1.5 years; Fig. 1). Two transgirls with a bone age of 13.5 and 14 years at baseline had grown 6.6 cm and 7.3 cm, respectively, during the first year of estradiol treatment. In contrast, the bone age had advanced by 0.5 year in both. They were treated with 6 mg/d estradiol from 19 and 16 months and grew 5.8 cm and 4.3 cm during the second year, with the bone age advancing by 1 year during the second year of estrogen treatment. The growth velocity had decreased to 2.3 cm/year in the transgirl from whom data for the third year were available (Fig. 1).

BMI, body composition, and blood pressure

The BMI and fat mass did not change during 2 years of estrogen treatment but were greater than at baseline after 3 years (BMI, 21.5 ± 3.2 kg/m² vs 20.1 ± 2.4 kg/m², n = 13, P = 0.01; fat mass, 20.5 ± 9.1 kg vs 17.2 ± 8.1 kg, n = 10, P = 0.007). No changes were observed in the BMI SDS, lean body mass percentage, or fat percentage (Table 1). Neither systolic nor diastolic blood pressure changed during treatment.

Gonadotropin and estradiol levels

Gonadotropin levels were suppressed in all but one individual, who was noncompliant with GnRHa. After gonadectomy, when GnRHa treatment was stopped, the gonadotropin levels increased. The levels after gonadectomy were available from six individuals, with a median LH of 21.5 U/L (range, 1.1 to 37) and FSH of 26 U/L (range, 1.4 to 58). They were greater than the normal range in five of the six subjects. As expected, the serum estradiol levels increased during the first 2 years as the estradiol dose was increased (Table 2, Fig. 2). The adult dose of 2 mg was used for a median treatment duration of 2 years (range, 3 to 30 months) and resulted in a median serum estradiol of 100 pmol/L (range, 24 to 380). The serum estradiol levels were greater in those treated with 6 mg (to reduce growth) than in those treated with the standard adult dose of 2 mg (Fig. 2).

Table 2.

Changes in Laboratory Parameters During 3 Years of Estrogen Treatment

Parameter	Estrogen Treatment Duration (y)
	0	1	2	3
LH (U/L)	0.4 (<0.3–6.4)	<0.3 (<0.3–5.2)	<0.3 (<0.3–6.5)	<0.3 (<0.3–37)
FSH (U/L)	<0.5 (<0.5–3.8)	<0.5 (<0.5–3.4)	<0.5 (<0.5–9.6)	<0.5 (< 0.5–58)
E2 (pmol/L)	28 (<18–95)	61 (32–169)	118 (34–413)	98 (29–199)
ALP (U/L)	195 ± 67	178 ± 94	123 ± 48	92 ± 36
Creatinine (µmol/L)	70 (46–89)	72 (56–96)	73 (44–86)	65 (41–90)

Data presented as median (range) or mean ± standard deviation.

Abbreviation: E2, estradiol (excluding two individuals who used ethinylestradiol).

Figure 2.

Estradiol levels during pubertal induction in transgirls. E2, gradually increasing dose of estradiol; EE, ethinylestradiol.

Other laboratory investigations

The median prolactin levels did not change during the 3 years of estrogen treatment. However, one subject developed hyperprolactinemia and galactorrhea during treatment with 200 µg of ethinylestradiol. Prolactin had increased to just greater than the upper limit of normal for females after 3 months and slowly continued to increase to 3.2 U/L after 42 months (normal, <0.6 U/L). No pituitary lesion was seen on magnetic resonance imaging. She was successfully treated with dostinex and was switched from ethinylestradiol to oral estradiol 2 mg/d.

No changes in hemoglobin or hematocrit were observed. Hemoglobin A1c (HbA1c) also did not change during treatment, and no elevation of liver enzymes (AST, ALT, or γ-glutamyl transferase) was encountered. The mean ALP did not change during the first year of treatment but did decrease during the second and third year (Table 2). Creatinine slightly decreased during 2 years of treatment (Table 2). One individual had a creatinine level just greater than the normal range after 1 year but this spontaneously normalized.

Discussion

Our findings have shown that estrogen treatment effectively induces breast development and female fat distribution in transgirls. Breast development started within 3 months of treatment in most adolescents, just as was observed in adults (11). Tanner breast stage 4 or 5, both seen in adult cisgender females, had been reached by 86% of adolescents after 3 years of treatment. It is difficult to directly compare our results to those from adults, because few adult studies have used the Tanner stage as an outcome measure. In a group of 17 transwomen who had used estrogen for a median of 59 months, breast development was at Tanner stage 2 to 3 in all those without breast implants (12). This might suggest that breast development is more easily induced in adolescents who have not undergone full male puberty, than in adults, who have. However, in the adult study, different forms and doses of estrogen were used, which could have influenced the outcome. Also, the estrogens were combined with progestins, rather than a GnRHa, which might have resulted in less optimal suppression of testosterone and might also explain the poorer breast development.

Estrogen treatment also induced a female pattern of fat distribution, as evident from an increase in hip circumference and decrease in the waist/hip ratio, such as was observed in adult transwomen (11). After 3 years of estrogen treatment, adolescents had a waist/hip ratio similar to the female reference population (0 SDS).

Although treatment effectively induced breast development and female fat distribution, it is not clear whether the estradiol doses used were optimal. The median serum estradiol level during treatment with 2 mg estradiol (100 pmol/L) was lower than the median level of 116 pmol/L reported in healthy midpubertal girls (Tanner breast stage 3 to 4), although the measurements were not directly comparable owing to the use of different assays (13). A Belgian study aimed for serum estradiol levels in the normal female range, defined as 55 to 200 ng/L (202 to 734 pmol/L) but found levels <200 pmol/L in 63% of adult transwomen treated with various estrogen preparations; 38% had used 2 mg oral estradiol (14). The greater percentage (88%) of adolescents with serum estradiol <200 pmol/L in our study might indicate that the other preparations used in the adult study (mainly transdermal estradiol) resulted in higher serum levels than 2 mg of oral estradiol. Alternatively, noncompliance might have caused lower estradiol levels in the present study. This might also explain some of the variation in serum estradiol levels observed within individuals taking a stable dose. The female reference range for serum estradiol is wide, and it is unclear whether having levels at the lower or higher end of the range for a prolonged period is associated with an increased risk of complications such osteoporosis or thromboembolic events. A British study reported a narrower target range for serum estradiol corresponding to the upper follicular range, 400 to 600 pmol/L (15). However, because of the different assays used, it is difficult to define a universal optimal range.

In addition to the estradiol concentrations, the measurement of LH might be useful in monitoring treatment. Serum LH levels in the normal range might be a measure of adequate estrogen dosing (16). In the present study, most samples were taken during GnRHa treatment; however, postgonadectomy gonadotropin levels were available from six individuals, with estradiol levels of 45 to 137 pmol/L. LH was elevated in five of them. The individual with normal LH had undergone gonadectomy only 4 months before the last LH measurement; thus, suppression from the GnRHa might not have been completely eliminated at the time of the measurement. Similarly, elevated LH levels have been found in adult transwomen taking 2 mg of oral estradiol (14). Thus, this dose might be insufficient for some, which could have a negative effect on bone health. An inverse relationship has been found between serum LH concentrations and bone mineral density (17). In addition, decreased bone mineral density Z-scores have been found at 22 years of age in transwomen treated with GnRHa and estradiol from adolescence (18). The investigators suggested this might have resulted from the relatively low estrogen dosage (18).

The Endocrine Society guideline (1) recommends 2 to 6 mg of oral estradiol for treatment of adult transwomen, with monitoring of serum levels, which should not exceed 200 pg/mL (734 pmol/L). It seems appropriate to apply these same recommendations to adolescents at the end of pubertal induction and to increase the estradiol dose if the serum estradiol levels are low, despite good compliance with a dose of 2 mg, especially in the presence of a high serum LH concentration and low bone mineral density. High doses of estrogen might increase the risk of venous thromboembolic complications, which have been reported in 0% to 6% of adult transwomen (15, 19–21). However, the use of conjugated estrogens, ethinylestradiol and cyproterone acetate, in addition to risk factors, such as smoking, might largely explain the high risk reported in some of these adult studies (15, 19, 20, 22). Discouraging smoking and providing adequate instructions for situations of immobilization are therefore essential. Transdermal estradiol might also provide a safe alternative to oral estradiol and has been successfully used to induce puberty in hypogonadal girls although patches are only available in adult doses and need to be cut to size to achieve appropriate dosing for pubertal induction (23).

Many adolescents had advanced bone age at baseline and grew only slightly. In those with a bone age of 13 to 15 years, the height gain was more variable. Which factors determine the difference in height gain between individuals with a similar bone age at the start of treatment is unclear. Differences in the estrogen dosing schedule or bone age at the last measurement (reflecting remaining growth potential) did not seem to explain this variation. Only a small number of adolescents received high-dose estrogen treatment to limit growth. This treatment was effective in the individual who had used 200 µg of ethinylestradiol from the third month and grew only 1.9 cm despite a bone age of 14 years at the start. In the other three individuals, it was not obvious whether the treatment had reduced their growth. This might have been because they had initially been treated with the standard dose and had only switched to the high dose later on. Low doses of estrogen are thought to stimulate growth, and high doses are thought to accelerate bone maturation. Starting high-dose estrogen treatment at a more advanced bone age would likely have made the treatment less effective, which was shown in constitutionally tall adolescents treated with high-dose sex steroids (24). Perhaps, 6 mg of estradiol is also less effective than 200 µg of ethinylestradiol to reduce growth. However, because of the increased risk of thromboembolic complications, the use of ethinylestradiol is no longer recommended for transgender individuals (1). Future studies are needed to investigate whether high-dose estradiol is effective and safe and to determine the optimal dose to limit growth in adolescents who wish to prevent extremely tall stature.

No change in BMI SDS, fat percentage, or lean body mass percentage was observed. This finding is in contrast to those from adult studies, in which an increase in fat mass and decrease in lean body mass was found (11). This might be explained by the GnRHa treatment that adolescents received before the estrogen treatment. GnRHa treatment was previously shown to result in a decrease in lean body mass percentage and an increase in fat percentage, likely caused by the decrease in testosterone levels (3). It is reassuring that estrogen treatment did not cause a further decrease in the lean body mass percentage or an increase in the fat percentage.

One adolescent treated with high-dose ethinylestradiol developed hyperprolactinemia with galactorrhea. Hyperprolactinemia has previously been observed in 0% to 4% of transwomen using estrogen (11, 25). It is unclear whether those treated with lower doses of estradiol are also at risk. None of the adolescents treated with 2 mg of estradiol developed hyperprolactinemia, and the median prolactin level did not change. In adults, contrasting findings have been reported. One study of adults using GnRHa and 6 mg of estradiol found stable prolactin levels (25); however, a study of adults using cyproterone acetate and 4 mg of estradiol observed an increase in the median prolactin level (11). This difference might have resulted from the use of cyproterone acetate, which was also found to cause an increase in prolactin when used without estrogen treatment (11). Until more data become available, it seems justified to monitor prolactin levels, as recommended by the Endocrine Society (1).

We did not observe the decrease in hematocrit that was reported in adult transwomen after 1 year of estrogen treatment (11). Previous GnRHa treatment might have caused the hematocrit to decline before estrogen treatment was started in adolescents. The creatinine levels did decrease slightly during estrogen treatment, which has also been described in adult transwomen (11). The cause is not clear, because the decrease in lean body mass found in adults was not observed in adolescents. We did not observe elevations of AST or ALT. Transient elevations of liver enzymes exceeding twice the upper limit of normal were observed in adult transwomen, but this might have resulted from the use of cyproterone acetate, in addition to estradiol, because they were not observed in adults using GnRHa rather than cyproterone acetate (11, 25). The decrease in ALP we observed during the second and third year of estrogen treatment was most likely related to reduced growth velocity. No changes in HbA1c were noted. On the basis of these findings, it does not seem necessary to monitor the hematocrit, creatinine, liver enzymes, or HbA1c during pubertal induction with estradiol.

In conclusion, we have shown that pubertal induction with a gradually increasing dosing schedule of estradiol effectively induces breast development and female fat distribution in transgender adolescents. However, an adult dose of 2 mg does not always result in appropriate serum estradiol levels. In particular, in the presence of elevated LH and low bone mineral density, a greater dose should be considered. No adverse effects were observed during estradiol treatment, and we believe that monitoring of renal function, liver enzymes, hematocrit, and HbA1c during pubertal induction is not necessary. After completion of pubertal induction, the adult guidelines for monitoring can be followed (1). Further studies are needed to establish effective and safe methods to limit growth.

Abbreviations:

 
• ALP

  alkaline phosphatase

 
• ALT

  alanine aminotransferase

 
• AST

  aspartate aminotransferase

 
• BMI

  body mass index

 
• FSH

  follicle-stimulating hormone

 
• GnRHa

  gonadotropin-releasing hormone analog

 
• HbA1c

  hemoglobin A1c

 
• LH

  luteinizing hormone

 
• SDS

  standard deviation score.

Acknowledgments

We thank all the adolescents who participated in our study.

This work was supported by an unrestricted educational grant from Ferring BV, Hoofddorp, The Netherlands.

Disclosure Summary: S.E.H. has received lecture fees from Pfizer and Ferring.

Author notes

Address all correspondence and requests for reprints to: Sabine E. Hannema, MD, PhD, Department of Pediatrics, Leiden University Medical Centre, Postbus 9600, 2300 RC Leiden, The Netherlands. E-mail: [email protected].

†

Deceased 13 February 2014.