Ovarian stimulation outcomes among transgender men compared with fertile cisgender women

Abstract

Purpose

To compare assisted reproductive technology (ART) outcomes among transgender men with those of fertile cisgender women.

Methods

This retrospective cohort study included 12 transgender men, six with no testosterone exposure and six after testosterone treatment, and 12 cisgender women (oocyte donors) who underwent ART in our institution between June 2017 and December 2019. Statistical analyses compared ART data and outcomes between three groups: cisgender women, transgender men without testosterone exposure, and transgender men after testosterone exposure. Comparisons were also made between transgender men with and without testosterone exposure.

Results

The transgender men with no testosterone exposure (23.3 ± 4 years) were significantly younger than the transgender men who had undergone testosterone treatment (30.3 ± 3.8 years; P = 0.012) and the cisgender women (29.1 ± 3.1 years; P = 0.004). The amount of FSH used for ovulation induction (1999 ± 683 mIU/mL) was significantly lower among transgender men without prior testosterone exposure compared with that among cisgender women (3150 ± 487 mIU/mL; P = 0.007). There were no differences in the peak estradiol levels, the number of oocytes retrieved, the number of MII oocytes, and the oocyte maturity rates between the three groups. Five out of six testosterone-treated transgender men underwent embryo cryopreservation, and they all achieved good-quality embryos.

Conclusions

Transgender men have an excellent response to ovulation stimulation even after long-term exposure to testosterone. Oocyte/embryo cryopreservation is, therefore, a feasible and effective way for them to preserve their fertility for future biological parenting.

Introduction

Fertility preservation (FP) with oocyte or/and embryo cryopreservation is a safe and effective option among women at risk of ovarian failure due to medical conditions, medical interventions, or advanced age [1–3]. This technology has also recently been offered to transgender men [4–6]. Transgender men are individuals who identify as male but were assigned female sex at birth [7]. The medical interventions for gender affirmation in transgender men include gender-affirming hormone (GAH, testosterone) and gender-affirming surgery (GAS). Gender-affirming bottom surgery includes hysterectomy and bilateral oophorectomy with or without reconstruction of male genitalia [7].

The fertility potential of transgender men may be affected by exposure to these medical interventions. While bilateral oophorectomy results in permanent sterility, the effect of testosterone on fertility is inconclusive. Testosterone treatment reportedly did not alter the number and the distribution of follicles in the ovary of transgender men [8, 9], nor did it affect the developmental capacity of the follicles upon xenotransplantation [8] or the in vitro maturation potential of the cumulus-oocyte complex obtained during the tissue processing of ovaries procured in these individuals [9]. Normal spindle structures and chromosome alignments were detected before and after vitrification of MII oocytes in transgender men after testosterone treatment [9, 10]. However, several studies on transgender men showed polycystic ovarian syndrome (PCOS)–related functional and morphological changes in their ovaries after testosterone treatment [11–14]. These changes were attributed to a hyperandrogenic environment that exists in both women with PCOS and in testosterone-treated transgender men. Other studies could not confirm PCOS-like changes among transgender men with testosterone exposure [15]. None of these studies [8–15] mention whether participants had previous treatment with GnRH agonist. The anti-Müllerian hormone (AMH) is a well-known ovarian reserve marker whose concentration is high in PCOS and declines as the number of developing follicles diminishes [16]. Its level reportedly decreased in transgender men who were using testosterone [17]. In the latter study [17], the transgender men (mean age 22.4 ± 6.8 years) underwent a combination of pituitary-gonadal suppression with the use of prolong GnRH agonist treatment and then, additional treatment with androgens and an aromatase inhibiter after desensitization had been established.

Therefore, the World Professional Association for Transgender Health (WPATH) [18], the Endocrine Society [19], the American Society for Reproductive Medicine (ASRM) [20], and the European Society of Human Reproductive and Embryology (ESHRE) [21] recommended that transgender persons should be encouraged to consider FP before starting GAH treatment and that they take an obligatory break in treatment of at least 3 months if the hormone therapy had already begun.

Public legitimacy and acceptance of transgender people recently reached unprecedented heights, and transgender individuals have subsequently started to seek treatment at a younger age [22], when reproductive intentions may not yet be clearly defined. Several studies have shown that many transgender individuals want biological children and were interested in preserving fertility [23–25]. Feasible FP options for post-pubertal transgender men include oocyte and embryo cryopreservation [26, 27], both of which require assisted reproductive technology (ART) that includes hormonal ovarian stimulation followed by oocyte retrieval. Embryo cryopreservation requires partner or donor sperm for fertilization. Embryo transfer can be carried out later if the transgender individual is interested in becoming pregnant and the uterus has been preserved, or, alternatively, with the recruitment of a gestational surrogate, thereby enabling transgender men to carry out biological parenting.

Data on ovarian stimulation outcomes of transgender men are limited [28–33]. Three of six papers are case series [28–30], in which the individuals underwent ovulation induction without prior treatments, with the exception of one 15-year-old who had a histrelin implant (a GnRH agonist puberty blocker) from the age of 12 years (the Tanner stage when GnRH was started is not mentioned) and one patient who had already been treated with testosterone. Excellent oocyte yields were observed in all reported cases. Three recent retrospective cohort studies examined the results of ovarian stimulation among transgender men compared with cisgender women [31–33]. None of the participants had been on puberty blockers prior to beginning the process of ovarian stimulation. All three studies [31–33] showed that transgender men have ovarian stimulation outcomes similar to those of matched cisgender women, which seemed to hold true even for those who have already initiated hormonal transition with the use of testosterone in two of the studies [31, 32]. However, the control group in both studies consisted of infertile patients in whom the ovarian reserve and the ovarian stimulation response may be affected by the underlying fertility problem, and they may therefore not be representative of the general healthy population.

The aim of the current study was to examine the ART outcomes of transgender men with and without testosterone treatment in comparison to those of fertile cisgender women. The effect of testosterone on ART outcomes was further examined by comparing transgender men with testosterone treatment to those without testosterone therapy.

Materials and methods

Ethical approval

This study was approved by the ethics committee (Helsinki) of the Tel Aviv Medical Center (#0168-20-TLV).

Study population and participant recruitment

This retrospective study was performed between June 2017 and December 2019 at the IVF Unit, Fertility Institute, in Tel Aviv Sourasky Medical Center, a tertiary university-affiliated medical center. During the study period, 38 transgender men arrived for a fertility consultation, and 12 proceeded with ovarian stimulation. Six of the transgender men seeking ovarian stimulation had not yet initiated the use of testosterone and six had used and discontinued testosterone treatment. In our medical center, egg donors must be residents of the country (it is not possible to receive eggs from abroad), and the number of potential donors is very low. During the study period, twelve cisgender women were planned for oocyte donation, and all of them completed ovarian stimulation. All transgender men and oocyte donors who completed ART between June 2017 and December 2019 were included in the current study. The medical records of 12 transgender men (age range 20–35 years) and 12 fertile cisgender women (age range 23–34 years) were reviewed. All of the transgender men were referred from the Endocrinology Clinic of the medical center after they had been evaluated by a community mental health professional and were diagnosed with gender dysphoria according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition 302.85 (DSM 5) criteria. All of the cisgender women were oocyte donors.

Data collection

All relevant data were collected from the computerized database of the hospital. The data in the electronic charts included the following: clinical details (age, body mass index (BMI), and serum total testosterone level), fertility potential details (follicle-stimulating hormone (FSH) and antral follicle count (AFC)), ART details (ovarian stimulation duration, total FSH dose, and peak serum estradiol), ovarian stimulation outcomes (number of retrieved oocytes, number of MII oocytes, and their maturation rate derived from the number of MII oocytes/number of oocytes aspirated), and the in vitro fertilization (IVF) outcomes (number of 2 pronuclei (2PNs), number of embryos that were frozen, and the day the embryo was frozen). All transgender men who had already initiated androgen therapy stopped testosterone injections for at least 3 months before referral to FP, and menses resumed in all of them after the discontinuation of testosterone. All of the participants had regular menstrual cycles. A regular menstrual cycle was defined when the interval between bleeding periods was in the range of 21–35 days.

Ovarian stimulation, oocyte, and embryo cryopreservation

Prior to FP, all of the participants signed informed assent and consent forms that provided detailed information regarding the known potential side effects of treatment. Controlled ovarian stimulation was carried out by the GnRH antagonist protocol. Follicle follow-up was by transvaginal or transabdominal ultrasound, according to patient choice, and ovum retrieval was by transvaginal access. The stimulation was started with the administration of daily recombinant FSH (rFSH; Gonal F (Serono, Geneva, Switzerland) or Puregon (Organon, Oss, The Netherlands)) from day 3 of the cycle. GnRH antagonist (cetrorelix acetate 0.25 mg, Cetrotide®, Serono or Ganirelix, Orgalutran®, Merck and Co., Inc.) was started when the leading follicle was ≥ 12 mm or the estradiol level was > 450 pg/ml, and it continued until the day of trigger administration. GnRH-α triptorelin 0.2 mg/day (Decapeptyl; Ferring, Kiel, Germany) was administered when at least three follicles achieved a diameter of 18 mm. Ovum pickup was performed 36 h later by transvaginal puncture with the participant under general anesthesia.

Oocyte cryopreservation was performed after the recovered oocytes were vitrified, and embryo cryopreservation was performed following fertilization by conventional IVF or intracytoplasmic sperm injection (ICSI).

Statistical analysis

Data were analyzed with SPSS, version 25.0 (SPSS, Inc., Chicago, IL, USA). The data were summarized as mean ± SD or number of responders (percentage) according to the variables. Comparing continuous variables between groups was done with the t test. The effect of the age on the other variables was tested using Pearson’s correlation. A significant effect of the age was found on FSH total dose with a correlation of 0.487 (P = 0.018). When a significant correlation was found, ANCOVA was used to control age. The effect sizes were calculated using Cohen’s D. A P value of < 0.05 was considered significant.

Results

Clinical characteristics of the study participants

Twelve transgender men preserved fertility during the study period, of whom six were treated with testosterone and six did not initiate testosterone treatment before fertility preservation. Twelve egg donors, who were fertile cisgender women, comprised the control group. The clinical characterizations of the entire cohort are detailed in Table 1. The six transgender men without prior testosterone exposure (23.3 ± 4 years) were significantly younger than the six treated transgender men (30.3 ± 3.8 years; P = 0.012) and the 12 cisgender women (29.1 ± 3.1 years; P = 0.004). There was no significant difference between the mean age of the testosterone-treated transgender men (30.3 ± 3.8 years) and the cisgender women (29.1 ± 3.1 years; P = 0.5). There were no other clinical differences, including the ovarian reserve markers FSH and AFC and serum testosterone levels, between the three groups. Among the subjects with a history of testosterone use, the mean length of testosterone exposure was 77 ± 55.3 months (range 14–144 months) and the mean time of discontinuation of testosterone prior to stimulation was 9.3 ± 5.9 months (range 5–21 months).

Table 1.

Comparison of clinical parameters between cisgender women and transgender men with and without testosterone exposure

	Cisgender women (n = 12)	Transgender men (n = 12)		P value			t			df			Effect sizes#
Characteristic		No T exposure (n = 6)	T exposure (n = 6)	Cis vs. trans with no T exposure	Cis vs. trans with T exposure	T vs no T exposure	Cis vs. trans with no T exposure	Cis vs. trans with T exposure	T vs no T exposure	Cis vs. trans with no T exposure	Cis vs. trans with T exposure	T vs no T exposure	Cis vs. trans with no T exposure	Cis vs. trans with T exposure	T vs no T exposure
Age (year)
Mean (SD)	29.17 (3.12)	23.33 (4.03)	30.33 (3.88)	0.004	0.5	0.012	3.395	− 0.690	− 3.063	16	16	10	1.640	− 0.330	− 1.773
Range	(23–34)	(20–31)	(24–35)
BMI (kg/m2)	24 (3.74)	24.38 (4.81)	22.51 (1.38)	0.858	0.365	0.383	− 0.182	0.931	0.913	16	16	10	− 0.087	0.521	0.518
Testosterone (nmole/L)	1.45 (0.37)	1.55 (0.35)	1.35 (0.43)	0.618	0.658	0.388	−0.514	0.455	0.902	11	11	10	− 0.287	0.254	0.533
FSH (mIU/mL)	7.1 (1.26)	6.81 (1.54)	7.61 (1.47)	0.676	0.449	0.377	0.425	− 0.777	− 0.924	16	16	10	0.205	− 0.376	− 0.534
AFC (total number)	17 (6.8)	20 (6.14)	16 (4.93)	0.364	0.855	0.281	− 0.934	0.186	1.140	16	16	10	− 0.478	0.098	0.651
Testosterone therapy (mo)	–	–		–	–	–
Duration (SD)			77 (55.3)
Range			(14–144)
Time off testosterone (SD)*			9.3 (5.9)
Range			(5–21)

T, total testosterone; BMI, body mass index; FSH, follicle-stimulating hormone; AFC, antral follicle count; SD, standard deviation

*Before initiation of ovarian stimulation

^#The effect sizes were calculated using Cohen’s D. A value of < 0.5 represents a small effect size, a value between 0.5–0.8 represents a medium effect size, and a value of > 0.8 represents a large effect size

Ovarian stimulation outcomes

The ART data and outcomes of the three groups are summarized in Table 2. All transgender men without testosterone exposure, all fertile cisgender women donors, and four out of six transgender men with prior testosterone treatment completed one ovarian stimulation cycle, while two transgender men with testosterone exposure completed two stimulation cycles. The data and results of the first stimulation cycle are shown in Table 2. All six of the transgender men with no testosterone exposure before ovarian stimulation had cryopreserved oocytes, while five of the six transgender men exposed to testosterone had cryopreserved embryos and one underwent oocyte cryopreservation.

Table 2.

Comparison of ovarian stimulation data and outcomes between cisgender women and transgender men with and without testosterone exposure

	Cisgender women (n = 12)	Transgender men (n = 12)		P value			t			df			Effect sizes#
Characteristic		No T exposure (n = 6)	T exposure (n = 6)	Cis vs. trans with no T exposure	Cis vs. trans with T exposure	T vs. no T exposure	Cis vs. trans with no T exposure	Cis vs. trans with T exposure	T vs. no T exposure	Cis vs. trans with no T exposure	Cis vs. trans with T exposure	T vs. no T exposure	Cis vs. trans with no T exposure	Cis vs. trans with T exposure	T vs. no T exposure
Ovarian stimulation duration (days)	11 (1.5)	11 (0.98)	9 (1.16)	0.904	0.095	0.058	− 0.122	1.775	2.138	16	16	10	-0.066	0.920	1.242
FSH total dose (mIU/mL)	3150 (487)	1999 (683)	2862 (664)	0.007*	0.310	0.309**		1.048			16			0.497
Peak E2 (pg/mL)	2262 (954)	2106 (1978)	2058 (974)	0.821	0.677	0.959	0.230	0.424	0.053	16	16	10	0.102	0.211	0.030
Oocytes retrieved (n)	23.75 (9.23)	24.33 (7.14)	22 (9.94)	0.894	0.716	0.651	− 0.135	0.370	0.467	16	16	10	− 0.071	0.183	0.272
MII oocytes (n)	17.67 (5.83)	20.50 (7.39)	^	0.386			− 0.890			16			− 0.421
Maturity rate (%)	76.87 (12.94)	83.25 (9.63)	77.15 (10.96)^^	0.304	0.970	0.379	− 1.062	− 0.039	0.930	16	14	8	− 0.562	− 0.024	0.592

T, testosterone; GT, gonadotropins; FSH, follicle-stimulating hormone; E2, estradiol; SD, standard deviation

*ANCOVA was used to control age; F(1,15) = 9.62, P = 0.007

**ANCOVA was used to control age; F(1,9) = 1.161, P = 0.309

^In testosterone-treated transgender men who underwent embryo cryopreservation, fertilization was accomplished using conventional IVF or ICSI; therefore, MII oocyte data were not available in all cases and that parameter was not included in this subgroup

^^The maturity rate for testosterone-treated transgender men was calculated only when fertilization was performed with ICSI

^#The effect sizes were calculated using Cohen’s D. A value of < 0.5 represents a small effect size, a value between 0.5–0.8 represents a medium effect size, and a value of > 0.8 represents a large effect size

There was no difference in the mean number of FSH-stimulation days between the cisgender women (11 ± 1.5 days) and the transgender men without testosterone exposure (11 ± 0.98 days; P = 0.904), and between the cisgender women (11 ± 1.5 days) and the transgender men with testosterone exposure (9 ± 1.16 days; P = 0.095). Additionally, no difference in the mean number of FSH-stimulation days was observed between the transgender men with testosterone exposure (9 ± 1.16 days) and the transgender men without testosterone exposure (11 ± 0.98 days; P = 0.058). However, the amount of FSH used (1999 ± 683 mIU/mL) was significantly lower within the non-exposed transgender men compared with that within the cisgender women (3150 ± 487 mIU/mL; P = 0.007). There was no significant difference in the amount of FSH used for ovulation induction between the exposed transgender men (2862 ± 664 mIU/mL) and the cisgender women (3150 ± 487 mIU/mL; P = 0.31) as well as between the exposed transgender men (2862 ± 664 mIU/mL) and the non-exposed transgender men (1999 ± 683 mIU/mL; P = 0.309). There were also no differences in peak estradiol levels between the cisgender women, the non-exposed transgender men, and the exposed transgender men (P = 0.821, 0.677, and 0.959, respectively) in the number of oocytes retrieved (P = 0.894, 0.716, and 0.651, respectively), in the number of MII oocytes (P = 0.386), and in the oocyte maturity rates (P = 0.304, 0.970, and 0.379, respectively).

IVF outcomes of the transgender men treated with testosterone

All six transgender men who had been treated with testosterone discontinued the treatment before starting ovarian stimulation. Four of them completed one ovarian stimulation cycle and two completed two stimulation cycles. Five of them underwent embryo cryopreservation and one underwent oocyte cryopreservation. All of the transgender men stated that they do not intend to carry the pregnancy and that the embryos will be transferred to another country for surrogacy, which is prohibited in our country. A summary of IVF outcomes is shown in Table 3.

Table 3.

Ovarian stimulation outcomes for six transgender men who were treated with testosterone

Subject	Testosterone use		Ovarian reserve		Ovarian stimulation			Ovarian stimulation outcomes
	Time on (months)	Time off (months)	FSH	AFC	Length of cycle (days)	Total FSH	Peak E2 (pg/mL)	Oocytes retrieved	2PN	Number of embryos frozen
1a	132	5	7.4	15	12	3600	1131	16	9	7
2b	16	8	6.2	15	9	2700	1428	16	–	–
3c	14	6	8.3	16	10	2250	2598	38	26	16
4d, cycle 1	72	21	9.9	10	9	3600	3525	15	8	4
4d, cycle 2	72	22.5	9.9	10	12	4400	5210	18	13	5
5e, cycle 1	144	7	5.9	25	9	2025	1147	31	12	6
5e, cycle 2	144	12	5.9	25	10	1350	5256	34	18	12
6f	84	9	8	18	10	3000	2524	16	4	1

FSH, follicle-stimulating hormone; AFC, antral follicle count; E2, estradiol; PN, pronuclei

^aConventional fertilization, 7 frozen embryos

^bOocytes cryopreservation, 11 MII oocytes cryopreserved

^cConventional and ICSI fertilization, 16 frozen embryos

^{d, cycle 1}ICSI fertilization, 4 frozen embryos

^{d, cycle 2}ICSI fertilization, 5 frozen embryos

^{e, cycle 1}Conventional fertilization, 6 frozen embryos

^{e, cycle 2}Conventional fertilization, 12 frozen embryos

^fConventional and ICSI fertilization, 1 frozen embryo

Subject 1, a 35-year-old transgender man, presented with his cisgender male partner. He had been treated with testosterone for 11 years and discontinued that treatment 5 months prior to the ovarian stimulation. He had 15 antral follicles at the initiation of his cycle. His ovarian stimulation lasted 12 days during which he received FSH 3600 IU. A peak serum estradiol of 1131 pg/mL was measured on the day of the trigger. His cycle resulted in the retrieval of 16 eggs that were conventionally inseminated with his partner’s sperm. Fertilization resulted in nine 2PNs that developed to seven good-quality blastocysts that were transferred to the USA for surrogacy. The second medicated frozen single embryo transfer resulted in a successful ongoing intrauterine pregnancy.

Subject 2, a 29-year-old transgender man, presented for oocyte cryopreservation. He had been on testosterone for 1 year and 4 months and discontinued it 8 months prior to ovarian stimulation. He had an AFC of 15 at the initiation of his cycle. His cycle lasted for 9 days during which he received a total of FSH 2700 IU. On the day of the trigger, he reached a peak serum estradiol level of 1428 pg/mL. Sixteen oocytes were retrieved of which 11 were MIIs, and all of them were frozen.

Subject 3, a 24-year-old transgender man, presented with his cisgender female partner. He had used testosterone for 1 year and 2 months and discontinued it 6 months prior to embryo cryopreservation. He had an AFC of 16 near the beginning of the cycle. His ovarian stimulation lasted for 10 days during which he received a total of FSH 2250 IU. On the day of the trigger, he attained a peak serum estradiol of 2598 pg/mL. This cycle resulted in the retrieval of 38 oocytes, of which 30 were conventionally inseminated and the remaining eight were fertilized by ICSI with donor sperm. Fertilization resulted in 26 2PNs. Sixteen good-quality blastocysts were developed and frozen.

Subject 4, a 33-year-old transgender man, presented with his transgender woman partner. He had been on testosterone for 6 years and had discontinued it for 1 year and 9 months prior to ovarian stimulation. He had an AFC of 10 immediately prior to stimulation. His cycle lasted 9 days during which he received a total of FSH 3600 IU. On the day of the trigger, he reached a peak estradiol of 3525 pg/mL. Fifteen oocytes were retrieved of which 10 were MIIs. These 10 MIIs were fertilized via ICSI with his partner’s sperm, resulting in eight 2PNs from which four good-quality blastocysts were developed and frozen. One month later, he arrived to undergo an additional IVF cycle. This time, the cycle lasted 12 days during which he received FSH 4400 IU. On the day of the trigger, he reached a peak estradiol level of 5210 pg/mL. Eighteen oocytes were retrieved of which 16 were MIIs. These 16 MIIs were fertilized via ICSI with his partner’s sperm resulting in 13 2PNs from which five good-quality blastocysts were developed and frozen.

Subject 5, a 32-year-old transgender man, was treated with testosterone for 12 years and had stopped for 7 months before starting the embryo cryopreservation process. He had an AFC of 25 before ovarian stimulation. His cycle lasted 9 days during which he received FSH 2025 IU. On the day of the trigger, he reached a peak estradiol level of 1147 pg/mL. Thirty-one oocytes were retrieved and conventionally inseminated with donor sperm resulting in 12 2PNs from which five good-quality day 3 embryos and one blastocyst were frozen. Five months later, he returned for a second IVF cycle. This time, the cycle lasted for 10 days during which he received FSH 1350 IU. On the day of the trigger, he reached a peak estradiol level of 5256 pg/mL. Thirty-four eggs were retrieved and conventionally inseminated with donor sperm, resulting in 18 2PNs from which 10 good-quality day 3 embryos and two blastocysts were frozen.

Subject 6, a 29-year-old transgender man, had been on testosterone for 7 years which he discontinued 9 months prior to ovarian stimulation. At the initiation of his cycle, he had an AFC of 18. His ovarian stimulation cycle lasted 10 days during which he received a total of FSH 3000 IU. On the day of trigger, he reached a peak serum estradiol level of 2524 pg/mL. This cycle resulted in the retrieval of 16 oocytes of which nine were conventionally inseminated and the remaining seven were fertilized by ICSI with donor sperm. Fertilization resulted in four 2PNs that developed to one good-quality day 3 embryo that was frozen.

Discussion

With the increasing awareness of reproductive healthcare in the transgender population, there is growing interest to preserve fertility among transgender men, but there are currently limited published data on FP outcomes from ART in that selective group [28–33]. Two studies [31, 32] compared the ART outcomes of adult transgender men with those of cisgender women undergoing fertility treatments, but no studies have recruited fertile cisgender women to form the control group. This is the first demonstration that FP outcomes from ART among transgender men, even after long-term testosterone treatment, are comparable with those of fertile cisgender women. The control group in the current study consisted of oocyte donors who comprise a highly fertile population. Oocyte and/or embryo cryopreservation is a well-established method of FP in conditions that may cause ovarian failure [1–3], and it has been recently used as a viable method for FP in transgender individuals as well [26, 27]. Our data support the methodology of controlled ovarian hyperstimulation as a feasible means of FP in transgender men before and after testosterone treatment.

The FSH levels and AFC of all participants included in this study were within the normal range [34, 35], and there was no difference in these parameters between the three groups. Since serum FSH and AFC are well-established markers of ovarian reserve [36], it can be concluded that the ovarian reserve of transgender men remains preserved even after exposure to testosterone.

Although the duration of hormonal stimulation did not differ between the three groups, significantly lower total doses of gonadotropins had been used in the stimulation cycles of the transgender men who had not yet initiated testosterone treatment compared with the cisgender group. This is probably because of the significantly younger age of the former group and, consequently, FSH was given with caution for fear of ovarian hyperstimulation.

There were no significant differences in the peak estradiol levels, the number of retrieved oocytes, the number of mature oocytes, and the maturity rate of the oocytes between the three groups. Data on ovarian stimulation outcomes among transgender men are relatively limited. The first published case series [28] included three transgender men who underwent FP before GAH therapy: one cryopreserved oocytes and two cryopreserved oocytes and returned for IVF and embryo transfer. More than 18 oocytes were retrieved in each of the two latter cases, and both had children by means of their cryopreserved oocytes, with the pregnancies carried by their cisgender female partners. A second case series [29] included five adolescent transgender males who underwent oocyte cryopreservation without prior puberty blockers or GAH treatments, and a large number of oocytes was cryopreserved for them all. A recent case series demonstrated good ART outcomes, including the number of retrieved oocytes and the number of MII oocytes and their maturity rate, among three adolescent transgender men and one adult transgender man who underwent FP after varying degrees of exposure to pubertal blockers and/or testosterone [30]. Our group recently published similar results in adolescent transgender males [33]. In that study, we demonstrated that FP outcomes, including the number of retrieved oocytes, the number of mature oocytes, and the maturity rate of the oocytes from ART among adolescent transgender males, are comparable with those of adolescent cisgender females. Two additional studies [31, 32] recently reported their findings on ART outcomes in adult transgender men compared with cisgender women. Leung et al. [31] were the first to show a higher number of oocytes retrieved from the transgender men compared with the cisgender women, a difference not observed in the current study. However, similar to our finding, there was no significant difference in the number of retrieved oocytes when the cisgender women were compared with the transgender men with testosterone exposure. A higher incidence of PCOS was suggested among transgender men before testosterone exposure [37–40] as well as after it [11–14]. Patients with PCOS are typically characterized with high ovarian reserve (i.e., a high AFC value) and a strong response to ovarian stimulation with a high number of retrieved oocytes [41], as also observed by us in our transgender participants. We cannot, however, reach firm conclusions about the incidence of PCOS in our transgender subjects because a full evaluation according to the Rotterdam criteria [42] was not performed. However, Leung et al.’s group [31] also found no difference in the peak estradiol level or percentage of mature oocytes between the two groups. Adeleye et al. [32] observed significantly lower peak estradiol levels in transgender men compared with cisgender women and between testosterone-treated and non-treated transgender men. Those authors also found a significantly lower number of oocytes retrieved in transgender men exposed to testosterone compared with non-exposed transgender men, while we did not. Granulosa cells are the main source of estradiol; therefore, estradiol production, as demonstrated in our study, may indicate normal functioning of the granulosa cells in the early stages of folliculogenesis and may suggest that the follicular development and, potentially, the oocyte quality may not be significantly impacted by prior testosterone exposure. Our findings are compatible with data that suggested that, despite testosterone treatment, the pool of primary oocytes that can be normally matured in vitro is relatively preserved among transgender men [9]. Additionally, the spindle structure is generally preserved when testosterone-exposed oocytes are maturated in vitro immediately or after thawing from cryopreservation [10]. Other ovarian stimulation outcomes of our study, including cycle length, mature oocytes, and maturity rate, were similar between the transgender men and cisgender women in a recent study by Adeleye et al. [32].

All of the subjects who had been exposed to testosterone had good-quality cryopreserved embryos. Two subjects used their partners’ sperm, one from his cisgender male husband and the other from his transgender female spouse. Three patients used sperm donors. One couple, the transgender man and his cisgender male partner, was assisted by a surrogate who is currently in the second trimester of pregnancy. In our cases, fertilization was performed by conventional IVF, ICSI, or a combination of the two. We did not compare the fertilization rate and the number of frozen embryos between transgender men and cisgender women because of the different fertilization methods that had been used, thus precluding meaningful results. Interestingly, the two transgender men (subjects 4 and 5) with longer histories of testosterone use (6 and 12 years, respectively) were the ones who completed multiple cycles of ovarian stimulation and had fewer good-quality blastocysts frozen. No blastocysts were achieved for subject 6 who was exposed to testosterone for 7 years. A similar trend was found in Adeleye et al.’s study [32], in which the patient with the longest testosterone exposure achieved the smallest number of good-quality embryos compared with the transgender men with shorter testosterone exposure. These results suggest that long-term testosterone use may negatively impact the oocyte quality and early embryonic development. However, our current sample size is too small to draw definitive conclusions and further studies are needed to explore these issues.

In the current study, all transgender men with testosterone histories discontinued testosterone for much longer periods before attempting ovarian stimulation (5–21 months) than the recommended 3-month discontinuation period. The retrospective design of this study limits our ability to obtain more details about the reasons for the prolonged period without testosterone. Possible reasons are the time to completion of all tests required for the ART process, obtaining donor sperm, arranging a parenting agreement, financial challenges, and others. Another issue is whether this duration of testosterone withdrawal may be acceptable to transgender men. Previous studies found that one major barrier not to pursue FP in transgender men is the fear that stopping hormone treatment could result in the reversal of many androgen-induced changes (including loss of muscle mass, fat redistribution, and return of menses) and the possible return of female physical characteristics resulting in gender dysphoria [4, 43]. Indeed, Armuand et al. [44] found that discontinuation of testosterone triggered gender incongruence and dysphoria due to these physical changes during FP among transgender men. Again, given the retrospective nature of this study, we were unable to assess the physiological and psychological effects of discontinuation of testosterone therapy on our patients. Additionally, we were not able to assess the physiological and psychological experiences of the patients during the ovarian stimulation process. All of these issues warrant further studies to make the process easier for the patients and increase FP rates among transgender men.

This study has several strengths. (1) Our control group is comprised of egg donors of reproductive age (21–35 years) whose ovarian reserve, including FSH and AFC values, was tested before ovarian stimulation and found to be high (over 20 oocytes on average were retrieved). The control group in two recent studies [31, 32] consisted of women who were receiving fertility treatments, a factor which could have influenced the results. Indeed, Leung et al. [31] mentioned that their control group was problematic because of the infertility factor and noted that a better comparison group would be fertile women, as those in the current work. (2) The subjects in our study consisted solely of adults (above 18 years old), while the participants in the studies of Leung et al. [31] and Adeleye et al. [32] had a broader age range and included individuals younger and older than 18 years. Age has a major effect on fertility potential and therefore, the exclusion of very young people may be significant in such studies.

Several limitations of the present study bear mention. (1) It is retrospective in design. (2) The statistical analyses were performed in small subgroups, which could compromise the strength of the conclusions about differences in ovarian stimulation outcomes between cisgender fertile women and testosterone-exposed and testosterone-unexposed transgender men. Further studies are needed to establish clear-cut conclusions. Furthermore, the small sample size limits the ability to make any conclusive statements about the length of time transgender men should discontinue testosterone prior to ovarian stimulation, if at all. (3) All of the transgender men in the present study, except for one who used a surrogate, underwent oocyte/embryo cryopreservation; therefore, we are unable to report pregnancy outcomes.

In conclusion, this study demonstrates that, even after long-term exposure to testosterone, transgender men have an excellent response to ovulation stimulation similar to age-matched fertile cisgender women. Therefore, oocyte and/or embryo cryopreservation can be considered an effective way for them to preserve their fertility for future biological parenting. In addition, our data indicate that an antagonist-based protocol for ovarian stimulation triggered by a GnRH agonist for oocyte maturation is a feasible means of ART in this population. Further studies are needed to examine whether testosterone treatment in transgender men should be discontinued before ovarian stimulation, and if so for how long. Additionally, more studies are required to explore the pregnancy outcomes in transgender men who underwent FP.

Compliance with ethical standards

This study was approved by the ethics committee (Helsinki) of the Tel Aviv Medical Center (#0168-20-TLV).

Conflict of interest

The authors declare that they have no conflict of interest.

Informed consent

Informed consent was obtained from all individual participants included in the study.