Oocyte cryopreservation with in vitro maturation for fertility preservation in girls at risk for ovarian insufficiency

Abstract

Purpose

To assess the feasibility and outcomes of oocyte cryopreservation with in vitro maturation (IVM) in post-pubertal girls undergoing fertility preservation (FP) for primary ovarian insufficiency (POI) risk.

Methods

Ovarian stimulation was performed with an antagonist protocol or progesterone priming. Ultrasound monitoring was performed transabdominally. Oocytes were retrieved transvaginally under IV sedation. Immature oocytes were subjected to IVM for up to 36 h. All MII oocytes were vitrified. The main outcome measure was the total number of mature oocytes cryopreserved. The secondary outcome was the increase in the mature oocyte yield after IVM.

Results

Indications for FP included mosaic Turner syndrome (mTS; n = 10), malignancy (n = 3), and POI risk (n = 2). The mean ± SD age, antral follicle count (AFC), and AMH levels were 14.2 ± 1.4 years, 8 ± 5.2 and 1.3 ± 1.3 ng/mL. In girls with mTS, the ovarian reserve was low for age (AFC 7.4 ± 4.7 and AMH 1.4 ± 1.6 ng/mL). Oocyte cryopreservation was possible in all girls with a range of 1–27 mature oocytes obtained, even in those who were previously exposed to chemotherapy or with low ovarian reserve, and no surgical complications were encountered. After IVM, the median mature oocyte yield increased significantly from 7.5 to 10.5 (p = 0.001).

Conclusions

Oocyte cryopreservation appears to be feasible and safe in girls as young as 12 years of age at risk for POI The utility of IVM increases the yield of cryopreserved mature oocytes. Prior exposure to chemotherapy or low ovarian reserve should not be an automatic reason to exclude these girls from FP consideration.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10815-023-02932-7.

Introduction

Fertility preservation (FP) is relatively rarely performed in children, compared to adults [1]. Among the most common indications for FP in children are planned gonadotoxic treatments, oophorectomy, and being at risk of primary ovarian insufficiency (POI) due to endocrine or genetic conditions such as Turner syndrome (TS) [2, 3]. TS is the most prevalent sex chromosome abnormality in females, affecting 1 in 2500 newborn girls [3]. It is associated with accelerated follicular loss and ovarian insufficiency [3] even in cases of 45,X mosaicism [4]. Spontaneous pregnancies have only been reported in about 5% of young adults with typically mosaic karyotypes (mTS) [5].

Performing FP in girls clinically predisposed to POI is recommended to reduce the future impact of gonadal insufficiency and infertility on the quality of life [6]. Accordingly, international guidelines state that appropriate multidisciplinary counseling on FP options including ovarian tissue and oocyte cryopreservation should be provided to these patients and their parents or guardians as early as possible [7]. Cryopreservation of ovarian tissue for future transplantation [8] is the only FP option in pre-pubertal girls [9, 10] with or without retrieval of immature oocytes followed by in vitro maturation (IVM) and vitrification [11, 12]. While ovarian tissue cryopreservation has previously been utilized for FP in girls with TS [13–15], its success potential is currently uncertain as most of these girls have diminished ovarian reserve at the time of tissue cryopreservation, and the initial ischemia following the transplantation may cause the loss of up to two-thirds of the primordial follicle reserve [16–20]. Furthermore, there have been no reports of pregnancy from thawed and transplanted ovarian tissue in TS.

Alternatively, cryopreservation of mature oocytes may have the advantages of requiring a less invasive intervention, being more widely available, and preserving fertilizable MII oocytes as the final product rather than primordial follicles. In addition, girls with TS have been found to have morphologically and chromosomally normal oocytes [21], and the first live birth after FP using vitrification of oocytes in a woman with mTS has recently been published [22]. Because of these reasons, we consider oocyte cryopreservation as the primary approach in post-pubertal girls, especially with TS.

Even though ASRM Practice Committees currently recommends that oocyte cryopreservation should be offered to select post-pubertal girls [23], ovarian tissue cryopreservation seems to be the more common method of fertility preservation in post-pubertal girls [24]. As a result, utility of ovarian stimulation and oocyte cryopreservation remains scarce in post-pubertal girls < 18 years of age [13], and the publications are generally limited to case reports and small case series [25–27]. Some hypothesized that the ovarian response to gonadotropins might be altered in adolescent girls due to the HPO axis immaturity [13, 28]. Other studies suggested that oocyte maturity rates are lower among adolescents with TS undergoing oocyte cryopreservation [27, 28]. These prior observations highlighted a need to improve oocyte maturity after ovarian stimulation in the pediatric-adolescent group of patients, including the utility of IVM. IVM is an approach that has been utilized in adult infertility [29] and breast cancer patients undergoing FP [30]. However, there is a paucity of data on IVM in the pediatric-adolescent population [31].

Given this background, we sought to evaluate the feasibility and outcomes of oocyte cryopreservation as well as the utility of IVM in post-pubertal girls who are at risk for POI due to mTS, exposure to cancer treatments, or idiopathic reasons.

Methods

We screened our clinical database for all post-pubertal girls (aged < 18 years) who were at risk for POI and evaluated for FP management between 2009 and 2022 by the senior author (KO) and one of his trainees (VT), referred by their pediatric endocrinologists or oncologists. Among them, all girls who underwent ovarian stimulation for oocyte cryopreservation were included in the study cohort.

In each case, preprocedural counseling and evaluation were performed with parental involvement to assess the patient’s physical and psychosocial development, degree of risk for POI, and to discuss the ovarian stimulation process. Parental consent was obtained for all procedures. Baseline ovarian reserve was assessed by serum AMH measurements and antral follicular counts (AFC) via transabdominal pelvic ultrasound prior to ovarian stimulation. Ovarian stimulation was performed with recombinant follicle-stimulation hormone (rFSH) (Follistim, Organon, West Orange, NJ or Gonal-F, Serono, Rockville, MD) and human menopausal gonadotropin (hMG) (Menopur, Ferring, Parsippany, NJ), or when available, recombinant luteinizing hormone (rLH) (Luveris, Merck, Whitehouse Station, NJ). Gonadotropin doses were individualized based on the serum AMH, antral follicle counts, and clinical experience [32]. In two cases where the cost was a concern, progesterone-priming was utilized to prevent a premature LH surge; otherwise, a GnRH antagonist (Ganirelix, Organon, West Orange, NJ) was added when at least one follicle reached a diameter of 14 mm. The cycle monitoring was performed by testing serum estradiol (E2) and LH levels and by transabdominal ultrasound exams. We measured E2 levels only until they reached 200 pg/ml, then we added ultrasound monitoring. Final oocyte maturation was triggered using recombinant human chorionic gonadotropin (rhCG) (Ovidrel, EMD Serono, Rockland, MA, USA) and/or leuprolide acetate (Lupron, Ferring Pharmaceuticals, Parsippany, NJ, USA), when at least 2 follicles reached the minimal diameter of 17 mm. Trigger type was decided based on the potential risk of hyperstimulation (in which case hCG might have been avoided, if baseline LH levels were not suppressed and likelihood of failure to respond to GnRHa trigger-only was low). Oocyte retrieval was performed transvaginally under intravenous sedation. Retrievals were performed by the senior author (KO, cases 1–7, 11–15) or one of his trainees (VT, cases 8–10).

All retrieved oocytes with their cumulus cells were extended on the glass slides as per protocol, oocyte stages were classified without removing the cumulus cells, and the culture time was set for all oocytes. MII oocytes were cryopreserved with the Cryotech Vitrification Kit (Reprolife, Tokyo, Japan) within 2 h, and all immature oocytes were incubated in open culture for IVM (Sage In vitro Maturation Media Kit, Cooper Surgical, Inc., Trumbull, CT). Cumulus cells were removed from MI oocytes after 4–6 h and put in IVM media, while GV oocytes remain in IVM with their cumulus cells. Per our protocols, all MI oocytes remained in IVM culture media for at least 2 h and up to 24 h after the retrieval, while GV oocytes were cultured for at least 18 h and for a maximum of 36 h. If there was no maturation, the oocytes were discarded at the end of the 36-h IVM period.

The main outcome measure was the total number of oocytes cryopreserved. The secondary outcome measure was the increase in the MII oocyte yield after IVM. Only the first cycle of each patient was included in the statistical analysis. Participant demographics, clinical characteristics, and outcomes were summarized using descriptive statistics. Quantitative variables were reported as means and range or standard deviation (SD). For IVM, the data were reported as median values and interquartile ranges (IQR) due to non-normal distribution. For the same reason, to compare the number of MII oocytes before and after IVM, Wilcoxon test was used with the level of significance set at p = 0.05. The retrospective review was approved by the Institutional Review Board at Yale University School of Medicine (Yale IRB Protocol ID 2000030279).

Results

Initial population

We identified 40 post-pubertal girls under the age of 18 (age range 11–18), who were evaluated for fertility preservation management in our center. All subjects in this study were referred by their pediatric endocrinologists. Of those, 14 girls had initial counseling but never returned for ovarian reserve testing or follow-up, and 6 girls were found to be already in POI and were recommended future oocyte donation. Out of the remaining 20 patients eligible for FP, 3 girls with TS (of which two were with mosaic karyotype) did not proceed with any FP procedures either due to having age-appropriate ovarian reserve, feeling anxiety towards the procedure, or cost concerns. Of the remaining 17, two underwent ovarian tissue cryopreservation (suspected acute autoimmune POI and risk of recurrent torsion of the contralateral ovary after unilateral oophorectomy). The remaining 15 girls underwent 19 ovarian stimulation cycles for oocyte cryopreservation. Five of these cases with 6 cycles had been reported in a previous publication [26].

Study population

There were 19 oocyte cryopreservation cycles from 15 girls aged between 12 and 17 years. These included 10 girls with mTS (45,X mosaicism ranging from 14 to 90%), and 5 others with germ cell tumor, acute lymphoblastic leukemia, non-Hodgkin lymphoma, family history of POI, and POI. The baseline characteristics of the study population are detailed in Table 1. The mean ± SD E2 was 40.1 ± 17.4 pg/mL, FSH 8.3 ± 6.5 IU/L, LH 7.4 ± 3.2 IU/L, AMH 1.3 ± 1.3 ng/mL, and AFC 8 ± 5.2. In patients with TS (cases 1–10), the mean values were E2 36.7 ± 18.4 pg/mL, FSH 9.4 ± 7.8 IU/L, LH 7.8 ± 3.2 IU/L, AMH 1.4 ± 1.6 ng/mL, and AFC 7.4 ± 4.7.

Table 1.

Patient population

Case	Age at FP (years)	Diagnosis	45,X (%)	AMH (ng/mL)	FSH (IU/L)	LH (IU/L)	AFC	BaseE2 (pg/mL)
1	13	45,X/47,XXX	90	1.6	5.7	3.9	6	15.1
2	14	45,X/46,XX	45	0.9	5.3	9.5	12	65.2
3	13	45,X/46,XX	20	0.8	5.6	5.3	6	33.5
4	15	45,X/46,XX	14	5.6	5.8	7.7	8	43.8
5	14	45,X/46,XX	17	2.1	5.4	13.1	7	37.6
6	13	45,X/46,XX	30	1	4.9	2.6	16	38
7	13	45,X/46,XX	20	1.8	7.9	7	12	36
8	16	45,X/46,XX	30	0.1	19	11	3	-
9	12	45,X/46,XXdelXq24	14	0.4	6	9	2	56
10	13	45,X/46,XX	30	0.2	28	9	2	5
11	15	Germ cell tumor	-	1.6	5.6	9.2	11	66
12	14	ALL	-	1.3	7.8	8.1	5	28.2
13	16	NHL	-	1	3	4	18	54
14	15	Familial POI	-	1.8	6.7	8.9	11	33.8
15	17	POI	-	0.05	7	2	1	49

Case 11 was previously treated with fertility-sparing surgery with staging, which included laparoscopic left salpingo-ophorectomy. We performed oocyte cryopreservation prior to the initiation of adjuvant systemic chemotherapy. Case 12 had received multi-agent chemotherapy with cyclophosphamide + methotrexate + dexamethasone + vincristine + pegaspargase + thioguanine + ARA-C at the age of 5 years (9 years prior to OC). Her AMH level was 0.8 at age 12. The AMH value represents the measurement performed before her first stimulation cycle at age 14. A second cycle was performed at age 16. Case 13 received multi-agent chemotherapy with doxorubicin + 6-mercaptopurine + methotrexate + prednisone + vincristine at the age of 11 years (5 years prior to oocyte cryopreservation)

Ovarian stimulation outcomes

Ovarian stimulation protocols and outcomes are summarized in Table 2. In the seven girls with mTS without diminished ovarian reserve (cases 1–7), the mean stimulation length was of 10.6 ± 0.8 days, reaching a peak E2 of 2,083.7 ± 552.1 pg/mL. A mean number of 13.7 ± 3.6 oocytes were retrieved. Of those, 7.9 ± 2.1 were at MII stage and 5.9 ± 3.8 were at GV or MI stage with maturity rate of 60.2 ± 19.8%. Of the immature oocytes, a mean of 1.7 ± 2.1 (range 0–5) was successfully in vitro matured with an IVM success rate of 20.7 ± 25.6%. A mean of 9.6 ± 2.3 MII oocytes was cryopreserved after IVM, representing a rise in the MII oocyte yield by 24.7 ± 29.8% per patient.

Table 2.

Ovarian stimulation protocols and outcomes of oocyte cryopreservation cycles

Case	Protocol	Total stim dose	Stim days	Peak E2 (pg/mL)	Trigger med	Oocytes retrieved	Cryo
1	Ant	2475 IU rFSH + 150 IU rLH	11	1548	3300 IU hCG	19	10
2	Ant	1800 IU rFSH + 450 IU hMG	10	2275	1 mg Leupr	11	8
	Ant	3750 IU rFSH + 2100 IU hMG	14	2029	1 mg Leupr	7	4
3	Ant	2025 IU rFSH + 75 IU rLH	10	1613	250 mcg rhCG	16	12
4	Ant	850 IU rFSH + 150 IU hMG	10	3000	3 mg Leupr	13	12
5	Ant	3150 IU hMG	12	1778	1 mg Leupr + 3250 IU hCG	14	11
6	Ant	1775 IU hMG	10	2592	3 mg Leupr + 2500 IU hCG	15	8
	Ant	1605 IU hMG	11	2150	3 mg Leupr + 3300 IU hCG	22	8
	Ant	3075 IU hMG	13	1813.6	3 mg Leupr + 4925 IU hCG	15	8
7	Ant	1575 IU hMG	11	1780	6500 IU hCG	8	6
8	Ant	1950 IU rFSH	13	252	250 mcg rhCG	1	1
9	PP	1500 IU hM	10	282	250 mcg rhCG	1	1
10	PP	1650 IU hMG	11	514	250 mcg rhCG	2	2
11	Ant	1837.5 IU rFSH + 300 IU rLH	11	1004	1 mg Leupr	8	4
12	Ant	1550 IU rFSH + 225 IU rLH	12	2279	1 mg Leupr	21	11
	Ant	2125 IU rFSH + 700 IU hMG	12	1384	3 mg Leupr	20	13
13	Ant	3450 IU rFSH + 900 IU hMG	13	2380	2 mg Leupr	17	15
14	Ant	1387.5 IU rFSH	9	3200	500 IU hCG	30	27
15	Ant	900 IU hMG	3	100.5	2 mg Leupr + 4000 IU hCG	1	1

Case 2 and 12 underwent two cycles, and case 6 underwent three cycles of oocyte cryopreservation. Case 7 also had a previous cycle at another center, resulting in 6 MII oocytes cryopreserved. Progesterone priming instead of antagonists was used in cases 9 and 10

Ant antagonist, PP progesterone priming, Peak E2 E2 level on the day of the trigger decision, Cryo total number of MII oocytes cryopreserved

Of the three girls with mTS whose ovarian reserve was severely diminished (AMH 0.1–0.4 ng/mL, cases 8–10), two had progesterone priming to lower the risk of a premature LH surge as the cost of antagonists was of concern. In these girls, the mean length of ovarian stimulation was 11.3 ± 1.5 days, with a peak E2 level of 349.3 ± 143.4 pg/mL. A mean of 1.3 ± 0.6 MII oocytes was retrieved; IVM was not needed as all retrieved oocytes were mature.

In cases 11–13, where the indication was the increased risk of POI due to previous exposure to cancer treatments, a standard gonadotropin dose-antagonist protocol was used for a mean of 12 ± 1 days, reaching a peak E2 level of 1887.7 ± 766.9 pg/mL. A mean number of 16.5 ± 5.9 oocytes were retrieved; 10 ± 4.2 were at MII stage and 6.5 ± 3.3 were immature (maturity rate of 59.8 ± 13.5%), from which 0.8 ± 0.9 oocytes were successfully matured in vitro and a final mean number of 10.8 ± 4.8 MII oocytes were cryopreserved after IVM.

Case 14 was evaluated because of suspected galactosemia, irregular periods, hot flashes, and a strong family history of POI. A standard gonadotropin dose-antagonist cycle for 9 days resulted in a peak E2 of 3200 pg/mL, and 30 oocytes were retrieved. Of those, 27 were mature and cryopreserved, while the remaining 3 were atretic and discarded.

Case 15 had already been diagnosed with POI. She had a previous ovarian stimulation cycle elsewhere that was canceled due to a lack of response. In a subsequent evaluation, the patient was found to have a dominant follicle of 17-mm size. An abbreviated ovarian stimulation was performed with an antagonist protocol for 3 days, where E2 levels peaked at 100.5 pg/mL. An MI oocyte was retrieved, which was successfully in vitro matured and cryopreserved.

Overall IVM outcomes

The in vitro maturation (IVM) results are summarized in Table 3. In 8 out of 12 cases, IVM resulted in the cryopreservation of additional MII oocytes. In those cases, 1–5 additional MII oocytes were cryopreserved after IVM, significantly increasing the median mature oocyte yields from 7.5 to 10.5 (p = 0.001). This represented a statistically and clinically significant improvement in mature oocyte yield, had the GV oocytes been discarded as is the case in routine practice. The mean IVM success rate, referred as the proportion of immature oocytes that reached MII status after in vitro culture where IVM was indicated (cases 1–7 and 11–15), was 27.4 ± 31.8%.

Table 3.

In vitro maturation outcomes

	a	b	b/a × 100	c	d	b + d
Case	Oocytes retrieved	N of MII pre-IVM	Maturity %	Immature (placed in IVM)	N of in vitro matured	Total MII Cryo post-IVM
1	19	9	47.4	10	1	10
2	11	8	72.7	3	0	8
	7	4	57.1	3	0	4
3	16	7	43.8	9	5	12
4	13	12	92.3	1	0	12
5	14	7	50	7	4	11
6	15	6	40	9	2	8
	22	7	31.8	15	1	8
	15	6	40	9	2	8
7	8	6	75	2	0	6
11	8	4	50	4	0	4
12	21	10	48	11	1	11
	20	13	65	7	0	13
13	17	13	76	4	2	15
14	30	26	87	3	1	27
15	1	0	0	1	1	1

Case 2 and 12 underwent 2 cycles of OC. Case 6 underwent 3 cycles of OC. Case 7 also had a previous cycle at another center, resulting in 6 MII oocytes cryopreserved. Progesterone priming instead of antagonists was used in cases 9 and 10

Complications and tolerability

In all cases, ovarian stimulation, monitoring, and transvaginal oocyte retrievals were tolerated well. There were no surgical complications during the stimulation or post-operatively. However, one girl (case 6) had an anxiety attack prior to her 3rd retrieval, which was managed via reassurance and family involvement. All patients were discharged within 1 h of the retrieval. Similar pain management to adults was utilized post-operatively.

Discussion

In this cohort of post-pubertal girls who were at risk for or with POI, we assessed the feasibility and outcomes of oocyte cryopreservation with IVM. We found that ovarian stimulation for oocyte cryopreservation is well tolerated and safe in this population of girls 12–17 years of age. We showed that, in most cases, a number of mature oocytes that can have > 26% per IVF cycle live birth probability, based on past data in young adults [33], could be obtained. We also found that the utility of IVM increased the number of MII oocytes cryopreserved. IVM is not routinely practiced in IVF laboratories, and immature oocytes are commonly discarded. However, in young adult fertility preservation patients, we showed that IVM increases the mature oocyte yield by 40% compared to the initial yield of mature oocytes without the IVM [30]. There is no data on oocyte vitrification success rates in children; however, a large study showed that per-cycle IVF success rate with vitrified oocytes is 41.8% in women < 35 years of age [34]. In that study, the addition of each mature oocyte from three to five mature oocytes more than doubled the live birth rates, implying a 3.6% increase per additional oocyte. The same gain was 4.4% per additional oocyte when increasing mature yield from five to eight and 5.5% when increasing it from eight to ten mature oocytes. However, another study showed that the per-cycle IVF success rate with IVM’ed oocytes was 26% in a similar age group [33]. Extrapolating from these two studies, we estimate that each IVM’ed oocyte will increase the live birth rate by 1.8–3.3% when three or more mature oocytes were retrieved initially. To our knowledge, this is the first report on ovarian stimulation for oocyte cryopreservation with IVM in post-pubertal girls < 18 years of age, with an age range including the youngest children ever reported.

In the 7 girls with mTS who did not have severely diminished ovarian reserve, the yield of mature oocytes was above the commonly utilized low response criteria of < 3 oocytes [35]. Although the number of oocytes retrieved was lower than expected for age in some cases (cases 2, 7, and 11), our results suggest that a number of oocytes that are associated with live birth rates of > 26% [33, 34] can be obtained and cryopreserved from adolescents with sex chromosome disorders, and these outcomes are better than that reported previously [27]. We noted that the oocyte maturity rates (72.1% in all first cycles in mTS and 64% in the standard dose gonadotropin cycles) were slightly lower than expected for the general infertility population (usually > 75%) [36]. However, our maturity rates were similar to the recently reported in TS by others [4].

Girls with TS are more likely to develop ovarian follicles when presenting with low FSH and high AMH levels, and/or history of spontaneous puberty and menarche [28, 37]. The severity and speed of oocyte depletion among TS girls vary after birth, and they often present with extremely low AMH levels before reaching adolescence [38]. It has been previously shown that there is a strong correlation between AMH levels and the mosaic karyotype [39]. In fact, only 40–50% of TS girls will undergo puberty and about 10% will experience menarche, almost exclusively in cases with a mosaic karyotype [40]. Although TS mosaicism correlates with ovarian function and ovarian reserve markers, and it has been related to the chance of spontaneous pregnancy [5, 28], the specific karyotype does not reliably predict the response to ovarian stimulation and the maturity rates of oocytes retrieved [40]. In our cohort, even those without a normal 46XX cell line (cases 1 and 9) yielded mature oocytes (Tables 1 and 2). However, our study did not have the power to correlate the level of mosaicism with the yield of mature oocytes.

In girls with cancer, it has been suggested that age, pre-treatment AMH levels, and the existence of previous gonadotoxic insult should be considered before fertility preservation, as those are the main predictors of response to ovarian stimulation [41]. We presented three cancer patients, one with a germ cell tumor (case 11) and two with hematologic malignancies (cases 12 and 13). Interestingly, although cases 12 and 13 underwent multi-agent chemotherapy regimens 9 and 5 years before oocyte cryopreservation, respectively, we were able to cryopreserve 11 and 15 MII oocytes from them. Case 12 received cyclophosphamide and case 13 received doxorubicin, which have both been shown to damage the primordial follicle reserve [42–46]. These cases reinforce the point that previous gonadotoxic treatments should not disqualify young girls from fertility preservation.

Although oocyte cryopreservation is not recommended for 3–6 months after the completion of gonadotoxic chemotherapy due to the induction of DNA double-strand breaks in oocytes [47–49], no increased fetal malformation or genetic abnormalities have been reported compared to healthy controls when oocyte retrieval is performed 6 months beyond the completion of chemotherapy [50, 51]. It has been shown that the teratogenic effects on the offspring observed in animals after acute cyclophosphamide exposure dissipate after a time period representing growth from primordial follicle to antral follicle stages. In humans, the equivalent of that time period is 3–6 months [52]. Based on our earlier studies, this is likely because of the ongoing DNA repair in oocytes, and those which fail to repair their DNA are eliminated via apoptosis and other check mechanisms [44, 45]. We showed that chemotherapy-induced DNA damage activates the ATM-mediated DNA repair pathway, and oocytes with sufficient DNA repair ability may survive [44, 53]. On the other hand, those primordial follicle oocytes with less efficient DNA repair mechanisms may be lost because of severe DNA damage triggering apoptotic death pathways [47–53]. As a result, primordial follicles that survive chemotherapy insult do not appear to have impaired DNA integrity and will be able to replenish the antral follicle population, which can be safely stimulated within 3–6 months of the exposure [49].

For post-pubertal girls who are eligible for hormonal stimulation and transvaginal oocyte retrieval, careful counseling of the patient and parents is needed to ensure compliance. GnRH antagonist protocols are currently preferred in oocyte cryopreservation cycles for FP, because of the shorter preparation time and as they can be utilized with the random start approach in time-sensitive cases [54]. However, because HPO axis may be immature in young girls [13, 28], ovarian stimulation protocols may need to be modified. First, either the addition of rLH or an LH-containing preparation may be preferable to better drive the follicle growth. Second, an hCG trigger rather than a GnRHa trigger or a combination of both may be preferred [26]. A GnRHa trigger alone may not be effective if serum LH is over-suppressed [55]. As we have done in cases 9 and 10, progesterone-priming may be used when there is a concern with the cost of injectable medications. But whether progesterone priming is equally effective as GnRH antagonists in suppressing spontaneous LH surge remains to be determined.

Finally, monitoring can be easily performed by transabdominal ultrasound scans in post-pubertal girls, and transvaginal retrieval under sedation is feasible in most cases [26].

Pre-pubertal girls and who are < 12 years of age and not yet suitable for ovarian stimulation can be managed according to our algorithm for FP that we previously reported [13] (Suppl. Figure 1). According to that algorithm, we follow pre-pubertal girls with serial serum AMH assessments [13]. Though individualized, we consider ovarian function exhausted if AMH < 0.1 ng/ml, no antral follicles are seen, and FSH > 30 IU/L. We consider ovarian reserve severely diminished if AMH < 1.1 ng/ml, AFC < 7, and FSH 12–30 IU/L. If there is no acute risk to ovarian reserve (such as imminent chemotherapy), we only consider ovarian tissue freezing if the AMH declines < 2 ng/mL pre-pubertally (2 SD below normal) during the periodic monitoring. If the AMH levels remain steadily above this level, we wait for the girls to reach puberty and attain sufficient sexual and psychosocial maturity, so that ovarian stimulation for oocyte cryopreservation can become an option.

In conclusion, ovarian stimulation for oocyte cryopreservation appears to be feasible and safe in young girls with various medical conditions that predispose them to POI. A tailored approach that combines an antagonist cycle with LH supplementation, close monitoring, hCG or dual triggering, and transvaginal retrievals results in oocyte cryopreservation without complications. The utility of IVM increases the yield of mature oocytes. However, to determine the full feasibility of ovarian stimulation and oocyte cryopreservation approaches in post-pubertal girls, and to assess the pregnancy outcomes, longer-term and larger studies will be needed. Since FP by oocyte cryopreservation has only recently been utilized in children, and these children are remote from adulthood, the pregnancy outcomes will not be available soon. In the meantime, our data should provide guidance to those who are planning oocyte cryopreservation in young girls with mTS and other medical conditions associated with risk for POI.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

K.O. conceived the idea, directed the study, and wrote the manuscript; S.G. collected data and wrote the manuscript. V.T. collected data and wrote portions of the manuscript. S.G. and K.O. had full access to all data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Funding

K.O. is funded by NIH HD R01HD053112. Foundation Alfonso Martín Escudero (Madrid, Spain) provided grant support for S.G. research fellowship position.

Declarations

Competing interests

The authors declare no competing interests.