Cryopreservation of in vitro matured oocytes in addition to ovarian tissue freezing for fertility preservation in paediatric female cancer patients before and after cancer therapy

Abstract

STUDY QUESTION

Is a protocol that combines in vitro maturation of germinal vesicle-stage oocytes and their vitrification with freezing of cortical ovarian tissue feasible for use in fertility preservation for both chemotherapy-naive paediatric patients as well as patients after initiation of cancer therapy?

SUMMARY ANSWER

Follicle-containing ovarian tissue as well as oocytes that can undergo maturation in vitro can be obtained from paediatric patients (including prepubertal girls) both before and after cancer therapy.

WHAT IS KNOWN ALREADY

Anticancer therapy reduces the number of follicles/oocytes but this effect is less severe in young patients, particularly the paediatric age group. Autotransplantation of ovarian tissue has yielded to date 60 live births, including one from tissue that was cryostored in adolescence. However, it is assumed that autografting cryopreserved-thawed ovarian cortical tissue poses a risk of reseeding the malignancy. Immature oocytes can be collected from very young girls without hormonal stimulation and then matured in vitro and vitrified. We have previously shown that there is no difference in the number of ovarian cortical follicles between paediatric patients before and after chemotherapy.

STUDY DESIGN, SIZE, DURATION

A prospective study was conducted in a cohort of 42 paediatric females with cancer (before and after therapy initiation) who underwent fertility preservation procedures in 2007–2014 at a single tertiary medical centre.

PARTICIPANTS/MATERIALS, SETTING, METHODS

The study group included girls and adolescent females with cancer: 22 before and 20 after chemotherapy. Following partial or complete oophorectomy, immature oocytes were either aspirated manually ex vivo from visible small antral follicles or filtered from spent media. Oocytes were incubated in oocyte maturation medium, and those that matured at 24 or 48 h were vitrified. Ovarian cortical tissue was cut and prepared for slow-gradual cryopreservation. Anti-Mullerian hormone (AMH) levels were measured in serum before and after oophorectomy.

MAIN RESULTS AND ROLE OF CHANCE

Ovarian tissue was successfully collected from 78.7% of the 42 patients. Oocytes were obtained from 20 patients before chemotherapy and 13 after chemotherapy. The youngest patients from whom oocytes were retrieved were aged 2 years (two atretic follicles) and 3 years. Of the 395 oocytes collected, ∼30% were atretic (29.6% in the pre-chemotherapy group, 37% in the post-chemotherapy group). One hundred twenty-one oocytes (31%) were matured in vitro and vitrified: 67.8% from patients before chemotherapy, the rest after chemotherapy. Mature oocytes suitable for vitrification were obtained from 16/20 patients before chemotherapy and from 12/13 patients after chemotherapy (maturation rate, 32 and 26.4%, respectively). There were significant correlations of the number of vitrified oocytes with patient age (more matured oocytes with older age) (P = 0.001) and with pre-oophorectomy AMH levels (P = 0.038 pre-chemotherapy group, P = 0.029 post-chemotherapy group). Oocytes suitable for vitrification were obtained both by manual aspiration of antral follicles (45%) and from rinse solutions after dissection. There were significantly more matured oocytes in the pre-chemotherapy group from aspiration than in the post-chemotherapy group after both aspiration (P < 0.033) and retrieval from rinsing fluids (P < 0.044). The number of pre-antral follicles per histological section did not differ in the pre- versus post-chemotherapy. AMH levels dropped by approximately 50% after ovarian removal in both groups, with a significant correlation between pre- and post-oophorectomy levels (P = 0.002 pre-chemotherapy group, P = 0.001 post-chemotherapy group).

LIMITATIONS, REASONS FOR CAUTION

There were no patients between 5 years and 10 years old in the post-chemotherapy group, which might have affected some results and correlations. Oocytes from patients soon after chemotherapy might be damaged, and caution is advised when using them for fertility-restoration purposes. The viability, development capability and fertilization potential of oocytes from paediatric patients, especially prepubertal and after chemotherapy, are unknown, in particular oocytes recovered from the media after the tissue dissection step.

WIDER IMPLICATIONS OF THE FINDINGS

Although more oocytes were collected and matured from chemotherapy-naïve paediatric patients, ovarian tissue and immature oocytes were also retrieved from young girls in whom cancer therapy has already been initiated. Our centre has established a protocol for potential maximal fertility preservation in paediatric female patients with cancer. Vitrified-in vitro-matured oocytes may serve as an important gamete source in paediatric female patients with cancer because the risk of reseeding the disease is avoided. Further studies are needed on the fertility-restoring potential of oocytes from paediatric and prepubertal patients, especially after exposure to chemotherapy.

STUDY FUNDING/COMPETING INTEREST(S)

The study was conducted as part of the routine procedures for fertility preservation at our IVF unit. No funding outside of the IVF laboratory was received. Funding for the AMH measurements was obtained by a research grant from the Israel Science Foundation (to B.-A.I., ISF 13-1873). None of the authors have competing interests.

TRIAL REGISTRATION NUMBER

N/A.

fertility preservation, paediatric cancer patients, in vitro maturation, oocyte vitrification, atretic oocytes, pre-antral follicles, anti-Mullerian hormone

Introduction

Advances in anticancer treatment have increased survival rates in children and adolescents with cancer (Feigin et al., 2008). Specifically, survival rates of 70% or more have been reported for various forms of leukaemia. It is estimated that one in every 100 adults in their third decade will be a survivor of childhood cancer, and very soon, one in every 715 adults in the general population will be a survivor of childhood or adolescent cancer. As a consequence, the incidence of premature ovarian failure due to chemotherapy is expected to rise (Abir et al., 2008; Feigin et al., 2008; Trudgen and Ayensu-Coker, 2014). Factors affecting the level of infertility are patient age (the younger the patient, the lower the likelihood of severe ovarian failure), drug dose, treatment duration, type of treatment (alkylating agents are considered the most gonadotoxic), and the number of agents used.

Among the currently available options for fertility preservation, especially in young patients, is cryopreservation of ovarian cortical tissue containing primordial follicles followed by autotransplantation (Feigin et al., 2008). This procedure has so far resulted in 60 live births (Donnez and Dolmans, 2015), including a report of ovarian grafting of tissue that was cryostored at adolescence (Demeestere et al., 2015). However, in patients with cancer, this procedure carries a potential risk of reintroducing the malignancy (Feigin et al., 2008; Meirow et al., 2008; Abir et al., 2010, 2014, Chung et al., 2013; Dolmans et al., 2013).

Alongside these findings, IVF laboratories, including our own (Ben-Haroush et al., 2010; Farhi et al., 2011; Ellenbogen et al., 2014), are reporting success with in vitro maturation (IVM) of germinal vesicle (GV)-stage oocytes and oocyte cryopreservation/vitrification (Revel et al., 2009; Ben-Haroush et al., 2010; Farhi et al., 2011; Noyes et al., 2011; Chian et al., 2013). These procedures have become routine in many centres and have resulted in many live births. IVM and oocyte vitrification can be offered together with cryopreservation of ovarian tissue to improve the patients’ fertility-preservation options (Revel et al., 2009; Farhi et al., 2011; Fasano et al., 2011; Michaeli et al., 2012; Chian et al., 2013). Oocytes can be collected by aspiration of small antral follicles (multilayered follicles containing a fluid-filled cavity within the granulosa cells layers, Gougeon, 1996), either in situ in the operating theatre or, if partial or complete oophorectomy is performed, ex situ in a fertility laboratory. Minimal ovarian stimulation may be considered unless contraindicated by the nature of the cancer or the patient's age. In survivors of childhood leukaemia and other malignancies, cryopreserved mature oocytes could provide a safe gamete source for future use because there is no danger of reseeding the cancer (Feigin et al., 2008; Abir et al., 2010, 2014; Dolmans et al., 2013).

Although fertility preservation in cancer patients should ideally be performed before therapy (Abir et al., 2008), it often becomes an option only after treatment has already been initiated. In a previous study from our laboratory, high numbers of viable non-apoptotic pre-antral follicles (follicles preceding the antral stage: primordial, primary and secondary, Gougeon, 1996) were identified in ovaries of patients aged ≤20 years, regardless of cancer therapy initiation. Transmission electron microscopy, however, revealed a deterioration in follicular intracellular quality manifested by an increase in abnormal granulosa cell nuclei and oocyte vascularization.

Anti-Mullerian hormone (AMH) is a glycopeptide produced by granulosa cells from primary stages onwards (La Marca et al., 2010; Lie Fong et al., 2012; Leader and Baker, 2014; Lunsford et al., 2014). Its secretion is reduced in antral follicles measuring ≥8 mm. AMH is considered the most accurate marker of ovarian reserve, also in survivors of childhood/adolescence cancer, including prepubertal girls (Brougham et al., 2012; Lunsford et al., 2014). Chemotherapy has been found to reduce AMH levels in paediatric patients with cancer (Peigne and Decanter, 2014; Dunlop and Anderson, 2015) compared with age-matched controls (Brougham et al., 2012; Lunsford et al., 2014; Peigne and Decanter, 2014).

The combination of immature oocyte aspiration followed by oocyte IVM and vitrification with ovarian cortical tissue cryopreservation has been described for fertility preservation in girls/adolescents (age 0–20 years) including prepubertal girls with cancer before chemotherapy (Revel et al., 2009; Fasano et al., 2011; Segers et al., 2015). However, there are no reports on paediatric patients after chemotherapy. Furthermore, two of these studies (Fasano et al., 2011; Segers et al., 2015) included low numbers of young girls (4 and 7, respectively). The aim of the present study was to describe our experience with this protocol in a larger cohort of paediatric patients with cancer, both before and after chemotherapy, and to determine, for the first time, if the number of maturable oocytes—not only pre-antral follicles—in ovarian tissue differs between these two patient groups (pre- and post-chemotherapy).

Materials and Methods

Patients

The study group consisted of 42 girls aged 2–18 years with cancer or in preparation for stem cell transplantation (sarcoma = 23.8%, leukaemia = 23.8%, lymphoma = 26.2%, other = 26.2%) who underwent fertility preservation at our centre between 2007 and 2014 (Tables I and II).

Table I

Characteristics of paediatric patients and their ovaries before anti-cancer treatment.

Patient information			Oocytes							Ovarian tissue	AMH (ng/ml)
Patient no.	Age (years), biopsy type	Disease	Source Antral follicle aspirated (N)	Source Rinsing Fluids (N)	Total (N)	Atretic (N)	Atretic/total (%)	Frozen (N)	% Frozen/total	Follicles/Section (N)	Before surgery	After surgery
1	2, Cr	AML	2	0	2	2	100	0	0	55	1.23	0.43
2	3, Cl	Brain ATRT	0	0	0	0	–	0	0	41
3	3, Cl	Medulloblastoma	1	1	2	1	50	0	0	0	0.34	0.22
4	5, Cr	AML	10	21	31	10	32.3	6	19.4	20	1.31	0.68
5	7, Cr	Beta thalassemia	2	6	8	3	37.5	2	25	66	4.35	2.82
6	8, Cr	Rhabdomyosarcoma	6	0	6	1	16.7	1	16.7	56
7	9, Cl	Medulloblastoma	6	5	11	3	27.3	4	36.4	23	UD	UD
8	10, C	Ewing sarcoma	16	12	28	7	25	8	28.6	50	7.09	4.18
9	11, Cr	ALL	12	3	15	10	66.7	2	13.3	8	2.03	1.14
10	12, Cr	Brain PNET	7	13	20	10	50	14	70	40	6.01	2.2
11	13, Pl	Ewing sarcoma	4	0	4	0	0	0	0	28
12	13, Cr	Osteosarcoma	1	8	9	1	11.1	5	55.6	30
13	13, Pr	HL	10	11	21	11	52.4	3	14.3	60	3.71	2.18
14	14, Pl	Ewing sarcoma	7	0	7	3	42.9	2	28.6	1
15	15, Pr	Osteosarcoma	4	11	15	5	33.3	3	20	130
16	15, Pr	HL	3	0	3	0	0	2	66.7	2
17	15, Cr	HL	4	0	4	0	0	0	0	4
18	16, Pr	HL	0	0	0	0	–	0	0	10	3.01	1.7
19	16, Pl	Ewing sarcoma	7	0	7	0	0	2	28.6	1
20	17, Pr	HL	18	24	42	3	7.1	21	50	34	1.3	0.7
21	18, Pr	HL	0	11	11	6	54.5	3	27.3	40	UD	UD
22	18, Cr	Lupus Nephritis	1	10	11	0	0	4	36.4	5
Total		22 patients	121	136	257	76		82		704
Mean ± SD	12 ± 5 (Age)		6 ± 6	6 ± 7	12 ± 1	3 ± 4		4 ± 5		32 ± 31	3.0 ± 2.2	1.7 ± 1.4
%			47.1	52.9		29.6	29.6	31.9	31.9

Percentage was calculated from the total number of oocytes retrieved.

Note: N: number; Cl: complete left oophorectomy; Cr: complete right oophorectomy; C: complete unilateral oophorectomy (side not specified); Pl: partial left oophorectomy; Pr: partial right oophorectomy; ALL: acute lymphoblastic leukaemia; AML: acute myeloblastic leukaemia; ATRT: atypical teratoid rhabdoid tumour; HL: Hodgkin lymphoma; PNET: primitive neuroectodermal tumour; UD: undetectable; AMH: anti-Mullerian hormone.

Table II

Characteristics of paediatric patients and their ovaries after anticancer therapy.

Patient information				Oocytes							Ovarian tissue	AMH ng/ml
Patient no.	Age (years), biopsy type	Disease	Prior treatment	Source Antral follicle Aspiration (N)	Source Rinsing Fluids (N)	Total oocytes (N)	Atretic (N)	% Atretic/total	Frozen (N)	% Frozen/total	Follicles/Section (N)	Before surgery	After surgery
23	2, Cl	Neuroblastoma	Alk agents	0	0	0	0	–	0	0	2 atretic	0.21	0.11
24	3, Cl	ALL	Alk agents	2	1	3	1	33.3	0	0	35	UD	UD
25	4, Cr	Wilms tumour	Alk agents	0	0	0	0	–	0	0	120	0.34	0.22
26	11, Cr	Rhabdomyosarcoma	Alk agents	0	0	0	0	–	0	0	0
27	11, Cl	ALL	Alk agents	4	9	13	10	76.9	3	23.1	0
28	11, Cr	Retinoblastoma	Alk agents	0	3	3	1	33.3	1	33.3	0
29	13, Pl	Germ cell tumour	BEP	2	4	6	0	0	3	50	25	UD	UD
30	14, C	Retinoblastoma	VEC	8	23	31	12	38.7	8	25.8	2	6.51
31	14, Cl	AML	Doxo + VP16 + cytarabine	0	0	0	0	–	0	0	136	3.89	2.55
32	15, Cl	AML	Alk agents	0	0	0	0	–	0	0	10
33	16, Pl	AML	Alk agents	1	13	14	5	35.7	3	21.4	10
34	16, Tl	Ewing sarcoma	Alk agents	2	0	2	0	0	1	50	130
35	16, Tl	HL	Alk agents	8	0	8	4	50	2	25	10
36	16, Pr	NHL	ABVD	1	1	2	0	0	1	50	30	2.54	1.9
37	16, Pl	ALL	Alk agents	21	13	34	9	26.5	11	32.4	8
38	16, C	AML	Alk agents	1	13	14	6	42.9	3	21.4	30	1.55	1.04
39	16, Cl	ALL	Alk agents	0	0	0	0	–	0	0	1	UD	UD
40	18, Cl	HL	ABVD	6	0	6	3	50	2	33.3	41
41	18, Cl	HL	ABVD	2	0	2	0	0	1	50	16	0.34	0.22
42	18, C	HL	Alk agents	0	0	0	0	–	0	0	0
Total		20 patients		58	80	138	51		39		606
Mean ± SD	13 ± 5			3 ± 5	3 ± 5	7 ± 10	3 ± 4		2 ± 3		30 ± 44	2.2 ± 2.3	1.0 ± 1.0
%				42.0	58		37	37	28.3	28.3

Percentage was calculated from the total number of oocytes retrieved.

Note: N: number; Cl: complete left oophorectomy; Cr: complete right oophorectomy; C: complete unilateral oophorectomy (side not specified); Pl: partial left oophorectomy; Pr: partial right oophorectomy; ALL: acute lymphoblastic leukaemia; AML: acute myeloblastic leukaemia; HL: Hodgkin lymphoma; NHL: non-Hodgkin lymphoma; UD: undetectable; Alk agents: alkylating agents; BEP: bleomycin, etopside, cisplatin; VEC: vincristine, etoposide, carboplatin; Doxo: doxorubicin; VP16: etopside; ABVD: doxorubicin, bleomycin, vinblastine, dacarbazine.

Twenty-two patients were chemotherapy-naïve (pre-chemotherapy group; Table I) and 20 had been exposed to various chemotherapy protocols either recently or following disease prior to salvage treatment. The ‘post-chemotherapy group’ (Table II) contained no patients aged 5 to 10 years, but there was no significant difference in age range between the two groups (Tables I and II). There was no overlap between the groups (i.e. none of the patients were in both groups) and all patients in the post-chemotherapy group had undergone at least one treatment or course of chemotherapy. Fourteen patients (nine before chemotherapy, five after chemotherapy) had not achieved menarche at the time of the study. Seven patients in the cohort died during the course of the study.

All parents of the minors and patients over 18 years of age signed an informed consent form, as required by the Israel Ministry of Health. The Israel Ministry of Health regulations include IVM and ovarian cryopreservation as routine fertility procedures; therefore, ethics approval was not required for the fertility preservation procedures described in this paper.

Ovarian tissue collection

All patients underwent laparoscopic retrieval of ovarian tissue under general anaesthesia using one 10-mm port positioned at the umbilicus and two 5-mm ports, one at the left lower quadrant and one at the right. Partial or complete unilateral oophorectomy was performed according to the consensus decision of the paediatric oncologist, the surgeon, and the medical staff of the IVF unit, based on ovarian size and the anti-cancer treatment protocol. In general, adolescents with a larger volume of ovarian tissue underwent partial oophorectomy, and prepubertal girls with small ovaries underwent complete oophorectomy. However, we took the gonadotoxic effects of the planned anticancer therapies into consideration. Information on partial or complete oophorectomy for each patient is given in Tables I and II. The average operative time was 40 min. The use of any dissection device that could induce collateral electric or thermal injury to the ovarian tissue was avoided. Ovarian evacuation required an endo-bag. One slice of fresh ovarian tissue was sampled for pathological examination, and was fixed in a 3% formalin solution (Gadot, Netanya, Israel) for histological preparation. It is a routine procedure at our centre to send a small ovarian biopsy sample for pathological examination directly from the operating theatre. The bulk of ovarian tissue was forwarded to the IVF laboratory in Hepes-buffered Oocyte Washing Medium (IVM Kit, SAGE-CooperSurgical Inc., Trumbull, CT, USA) for manual oocyte aspiration within a maximum of 30 min. Our preliminary experience with the laparoscopic technique for ovarian preservation in paediatric patients has been previously published (Feigin et al., 2007, 2008).

Manual oocyte aspiration

Visible antral follicles (at least 2 mm in diameter) were aspirated with a 21-gauge syringe needle connected to a 1-ml syringe (Becton Dickinson Microlance, Madrid, Spain) (Revel et al., 2009; Ben-Haroush et al., 2010). Aspirates were flushed in Hepes-buffered Oocyte Washing Medium (IVM Kit, SAGE/ CooperSurgical) and examined for oocyte-cumulus complexes under a dissecting microscope. After the ovarian cortex was sliced, the spent media from all the dissection dishes were filtered through a 70 μm nylon mesh cell strainer (BD Falcon, Becton Dickinson, Bedford, MA, USA) and the content was thoroughly inspected after the retained cells were resuspended in washing medium.

IVM protocol

Retrieved oocytes were incubated in four-well dishes (Nunc, Roskilde Denmark), up to five oocytes/well, containing oocyte maturation medium (SAGE/ CooperSurgical) supplemented with a final concentration of 75 mIU/ml FSH and 75 mIU/ml LH (prepared from Menogon/Menopur, Ferring, Kiel, Germany) without oil overlay at 37°C in an atmosphere of 5% CO₂ in air (Ben-Haroush et al., 2010; Son and Tan, 2010, Farhi et al., 2011). Denudation was performed 24 h later by brief exposure of the oocytes to hyaluronidase (80 IU/ml) (Irvine scientific, Santa Ana, CA, USA) and gentle pipetting to remove adhering cumulus cells. Immature oocytes were further incubated in IVM-medium (SAGE/CooperSurgical) and re-examined the next day.

Vitrification of mature oocytes

Mature, intact oocytes were cryopreserved using a Vitrification Kit (SAGE/CooperSurgical) and loaded on a cryolock (BioDiseno, Bogota, Columbia) or cryotop (Kitazato, BioPharma Co., Shizuoka, Japan) device for cryostorage in the vapour phase of liquid nitrogen (Fasano et al., 2011).

Cryopreservation of ovarian tissue

Cortical tissue was cut on a warm plate at 37°C into slices measuring 1 cm × 0.5 cm with a thickness of about 1 mm (Feigin et al., 2008). Each sample was placed in a cryotube (Nanc, ThermoFischer Scientific, Copenhagen, Denmark), and frozen slowly in a 1.5 M dimethylsulfoxide (DMSO) (Sigma, St Louis, MO, USA) solution. Before freezing, the samples were kept 45 min on ice for equilibration with the DMSO solution. All samples were frozen slowly at a cooling rate of −2°C/min until −7°C, −0.3°C/min until −30°C and finally −10°C/min until −140°C in a programmable freezer (Kryo 10; series 10/20, Planer Biomed, Sunbury on Thames, UK), and immediately plunged into liquid nitrogen at −196°C.

Histological preparation

The fixed ovarian pieces were dehydrated with increasing concentrations of ethanol (Biolab, Jerusalem, Israel) and xylene or toluene (Biolab) and prepared for paraffin embedding (Abir et al., 2008). Sections were cut at 5 µm thickness and stained with haematoxylin and eosin. The number of pre-antral follicles per section of every patient was counted, as described by us previously (Abir et al., 2008). The number of follicles in every sample was counted in two different section-levels per sample (50 µm between sections, to avoid counting the same follicle twice).

AMH measurement in serum samples

The Ethics Committee of Rabin Medical Center approved the AMH protocol, and informed consent was obtained from the parents of the minors and from patients over 18 years old. Blood samples were collected 24 h before and 24 h after oophorectomy. These time points were strictly maintained in for all patients. The serum was separated and frozen at −80°C until AMH measurement. AMH was measured with an enzyme-linked immunosorbent assay kit (Diagnostic Systems Laboratory, Webster, TX, USA) (Ben-Aharon et al., 2012), with a detection limit of <0.195 ng/ml.

Statistical analysis

Data were statistically analysed by analysis of variance and Pearson correlation coefficients test for the correlation graphs. P-values less than 0.05 were considered significant (StatView 5, and SAS version 9.4, SAS Institute, Cary, NC, USA).

Because of the limited number of individual patients per year of age, the statistical analyses could not be corrected for age factors.

Results

No complications were recorded following ovarian tissue collection. All patients had a benign post-operative course and were discharged the next day. On follow-up, no wound infection or port-side hernias were detected, and the cosmetic results were excellent.

Oocytes (range 2–42) were recovered from 33 of the 42 patients (78.6%; 20/22 before chemotherapy; 13/20 after chemotherapy), including 11 premenarcheal girls (Tables I and II). No immature oocytes were detected in two patients before chemotherapy and seven after chemotherapy. The youngest patients from whom oocytes were retrieved were aged 2 years (two atretic follicles) and 3 years. Moreover, 31 immature oocytes were retrieved from a 5-year-old girl and 28 immature oocytes from a 10-year-old girl. A total of 395 oocytes were collected, 45% by oocyte-aspiration procedures and 55% from left-over rinsing media. There was no statistically significant difference between patients before and after chemotherapy in number of oocytes collected (257 versus 138, respectively) or number of atretic oocytes. Almost 33% of the oocytes were atretic upon collection.

There was a significant positive correlation between total oocytes collected and number of atretic follicles within each group (Fig. 1A) (pre-chemotherapy, r = 0.64, P = 0.0013; post-chemotherapy, r = 0.91, P < 0.0001) and between the two groups (Fig. 1A, r = 0.97, P < 0.0001; increase in atretic oocytes correlated with increase in oocytes collected in both groups). There was a positive significant correlation between total number of oocytes collected and percentage of atretic oocytes and in the post-chemotherapy group (Fig. 1B, r = 0.55, P = 0.015). There was a negative significant correlation between atretic oocytes and patient age in the pre-chemotherapy group (Fig. 1C, r = −0.54, P = 0.013) and between the two patient groups in terms of atretic oocytes and age (reduction in atretic oocytes with increased age) (Fig. 1C, r = −0.83, P = 0.02).

Figure 1

Atretic oocytes identified upon collection for fertility preservation in female paediatric cancer patients before and after chemotherapy. (A) Correlation between total oocytes collected and number of atretic oocytes. *Significant correlation between total oocytes collected and number of atretic oocytes in chemotherapy-naïve patients (P = 0.0013). **Significant correlation between total oocytes collected and number of atretic oocytes in patients after chemotherapy (P < 0.0001). (B) Correlation between total oocytes collected and percentage of atretic oocytes. *Significant correlation between total oocytes collected and percentage atretic oocytes in patients after chemotherapy (P = 0.015). (C) Correlation between age and percentage of atretic oocytes. *Significant correlation between age and percentage atretic oocytes in chemotherapy-naïve patients (P = 0.0013). **Significant correlation between chemotherapy-naïve patients and patients after chemotherapy (P = 0.02).

One hundred and twenty-one oocytes (30.6%) underwent IVM and were vitrified. There was no significant difference in the number of mature oocytes between the pre- and post-chemotherapy groups, although a general trend towards more frozen-matured oocytes was noted in the pre-chemotherapy group (Tables I and II, Fig. 2A and B). There was a positive significant correlation between the pre- and post-chemotherapy groups (more frozen-matured oocytes with age increase) (Fig. 2A, r = 0.87, P = 0.001). In addition, the positive correlation between patient age and percentage of frozen-matured oocytes was significant in the post-chemotherapy group (Fig. 2B, r = 0.46, P = 0.04) and nearly significant in the pre-chemotherapy group (Fig. 2B, r = 0.396, P = 0.068). In both groups, the number of frozen-matured oocytes was significantly positively correlated with pre-surgery (pre-oophorectomy) AMH levels (number of frozen-matured oocytes correlated with increase in pre-surgery AMH levels) (Fig. 2C, pre-chemotherapy group r = 0.69, P = 0.038; post-chemotherapy group r = 0.76, P = 0.029), but not with AMH levels after oophorectomy (Fig. 2D).

Figure 2

Frozen-matured oocytes for fertility preservation in female paediatric cancer patients before and after chemotherapy. (A) Correlation between age and number of frozen-matured oocytes. *Significant correlation between chemotherapy-naïve patients and patients after chemotherapy (P = 0.001). (B) Correlation between age and percentage frozen oocytes/total oocytes collected. *Significant correlation between age and percent frozen-matured oocytes in patients after chemotherapy (P = 0.04). (C) Correlation between pre-surgery anti-Mullerian hormone (AMH) levels and number of frozen-matured oocytes. *Significant correlation between pre-surgery AMH levels and number of frozen-matured oocytes in chemotherapy-naïve patients (P = 0.038). **Significant correlation between pre-surgery AMH levels and number of frozen-matured oocytes in patients after chemotherapy (P = 0.029). (D) Correlation between post-surgery AMH levels and number of frozen-matured oocytes.

The matured oocytes (range: 1–21) were derived from 16/20 patients before chemotherapy (maturation rate, 32%) and 12/13 (oocyte range: 1–11) after chemotherapy (maturation rate, 26.4%). These maturation rates were significantly lower than the 61.7% found in standard IVM treatment cycles for adults in our unit (P < 0.0001) (Ben-Haroush et al., 2010; Farhi et al., 2011). The maturation rate was higher for oocytes from aspirated antral follicles than for oocytes filtered from spent media both at 24 h (25.4 and 17.8%, respectively) and cumulatively by 48 hours (41.5 and 29%, respectively); the difference achieved near significance (P = 0.056). Moreover, the number of matured oocytes retrieved from aspiration procedures was significantly higher in the pre-chemotherapy group at 24 h (35%) than in the post-chemotherapy group at 24 h and 48 h (22.2% for both time points, P < 0.033). It was also significantly higher than the number of oocytes retrieved from the rinsing fluids at 48 h in the post-chemotherapy group (25%, P < 0.044).

Histological analysis yielded no difference in the number of pre-antral follicles per section in the patients before and after chemotherapy. The correlation between age and follicular counts was not statistically significant in either group, but there was a marginally higher follicular count in the youngest patients, regardless of cancer therapy (Fig. 3A). Light microscopy failed to identify any atretic pre-antral follicles before or after chemotherapy. Baseline AMH values were not significantly higher before than after chemotherapy (Tables I and II). They dropped by approximately 40–50% after ovarian removal, either complete or partial, in both groups, and this was not statistically significant. It is noteworthy that among the patients after partial oophorectomy, some did not undergo AMH measurements (patients 11, 14–16, 19), one had an undetectable level of AMH (patient 29) and one showed a drop in AMH of only about 25% (patient 36) (Tables I and II). The correlation between number of cortical follicles and AMH level either before (pre-surgery) (Fig. 3B) or after (post-surgery) oophorectomy (Fig. 3C) was not statistically significant. However, there was a trend in the pre-chemotherapy group of an increase in pre-surgery and post-surgery AMH levels with an increase in the number of cortical follicles.

Figure 3

Cortical follicles identified in histological sections of ovarian tissue collected for fertility preservation in female paediatric cancer patients before and after chemotherapy. (A) Correlation between age and number of cortical follicles. (B) Correlation between pre-surgery AMH levels and number of cortical follicles. (C) Correlation between post-surgery AMH levels and number of cortical follicles.

Although there were no significant correlations between pre-surgery AMH levels and patient age in the pre- and post-chemotherapy group (Fig. 4A), AMH levels increased until about 10 years and then decreased, specifically in the pre-chemotherapy group. There was a positive significant correlation between post-oophorectomy surgery AMH levels and pre-oophorectomy surgery AMH levels in both the pre-chemotherapy group (Fig. 4B, r = 0.91, P = 0.002) and the post-chemotherapy group (Fig. 4B, r = 0.998, P = 0.001).

Figure 4

AMH levels detected before and after ovarian surgery for fertility preservation in female paediatric cancer patients before and after chemotherapy. (A) Correlation between age and pre-surgery AMH levels (ng/ml). (B) Correlation between pre-surgery AMH levels and post-surgery AMH levels. *Significant correlation between pre-surgery AMH levels and post-surgery AMH levels in chemotherapy-naïve patients (P = 0.002). **Significant correlation between pre-surgery AMH levels and post-surgery AMH levels in patients after chemotherapy (P = 0.001).

Discussion

To the best of our knowledge, this is the first report of an approach to fertility preservation using oocyte collection and IVM combined with ovarian tissue cryopreservation in paediatric patients after chemotherapy as well as for chemotherapy-naïve paediatric patients. It is also the largest report on the use of the combined protocol in paediatric patients. Immature oocytes were identified in ovaries of patients before and after chemotherapy; over half were derived from the rinsing fluids. The oocyte maturation rate for oocytes obtained by aspiration in the pre-chemotherapy group was significantly higher than by aspiration or rinsing media in the post-chemotherapy group; overall, the maturation rate for our paediatric patients was significantly lower than in our regular IVM cycles. There were significant correlations between the pre-and post-chemotherapy groups in terms of frozen-matured oocytes and patient age; and between the percentage of frozen-matured oocytes and patient age in the post-chemotherapy group. In addition, the number of frozen-matured oocytes was significantly correlated with pre-surgery AMH levels in both groups. The percentage of atretic oocytes was relatively high on collection (∼30%) in both groups; with significant correlations in both groups between atretic oocytes with the total number of oocytes collected; between both groups in terms of atretic oocytes and total oocytes collected; in the post-chemotherapy group between the percentage of atretic oocytes and the total number of oocytes collected; in the pre-chemotherapy between percentage of atretic oocytes and patient age; and between the two groups in terms of atretic oocytes and patient age. There were no differences in the number of pre-antral follicles per section between patients in both groups. AMH levels dropped by approximately 40–50% in both groups after oophorectomy with a significant correlation between pre- and post-oophorectomy levels in both groups.

AMH levels, representing follicular reserve, are normally related only to age (Guibourdenche et al., 2003; Hagen et al., 2010; Kelsey et al., 2011; Brougham et al., 2012; Lie Fong et al., 2012; Tehrani et al., 2014; Eichuri et al., 2015). They rise from birth to the age of 8 years and then remain stable until a peak in the early-mid twenties, followed by a decline until menopause. These findings are in line with our findings of an increase in AMH levels until 10 years and a decrease thereafter, especially in the chemotherapy-naïve patients, as there were no patients between 5 and 10 years in the post-chemotherapy group. Interestingly, one study reported a decrease in AMH levels in girls (including pre-pubertal) with newly diagnosed cancer, even before anti-cancer therapy, which was attributed to the impairment in their general health (van Dorp et al., 2014). During chemotherapy, studies report a progressive decrease in AMH levels in both pre-pubertal and post-pubertal girls (Brougham et al., 2012), which can recover after chemotherapy only in patients who receive protocols of low/medium gonadotoxicity risk (Lunsford et al., 2014; Dunlop and Anderson, 2015). In the present study, baseline AMH levels were higher in the pre-chemotherapy group than the post-chemotherapy group, but the difference did not reach statistical significance, probably because of the limited numbers of AMH measurements. It is possible that use of a recently introduced hypersensitive assay (detection limit 0.07 pmol/l) will yield more accurate results (also in paediatric patients), with fewer undetectable values (Decanter et al., 2014).

Unilateral oophorectomy/partial oophorectomy (in our study) or cystectomy (wherein healthy ovarian fragments are also removed) (Hwu et al., 2011; Streuli et al., 2012; Litta et al., 2013; Sugita et al., 2013) can reduce in vivo follicular reserve. Accordingly, a decrease in AMH levels was reported after cystectomy, mainly of endometriosis lesions (Hwu et al., 2011; Streuli et al., 2012; Litta et al., 2013; Sugita et al., 2013), sometimes normalizing within 1 year (Sugita et al., 2013). These findings agree with our observation of a decrease in AMH levels after oophorectomy in both pre-chemotherapy and post-chemotherapy groups, and it explains the statistically significant correlation between pre-surgery and post-surgery AMH levels in both patient groups. We also found that the contribution of one ovary to the total AMH measured in serum was about 50%. We presume that the similar decrease in AMH levels observed after both complete or partial oophorectomy can probably be explained by insufficient sensitivity of the assay kit. Again, it is possible that the use of a hypersensitive assay (Decanter et al., 2014) would have yielded more accurate results. On the whole, the number of AMH measurements in our study was limited, and therefore, all conclusions should be made with caution.

In another study that described ovarian tissue collection for cryopreservation in a patient cohort that included paediatric patients (age 0.8–17 years) (Imbert et al., 2014), low AMH levels were identified 5 years after surgery, which was probably related not only to the ovarian surgery but also to the post-surgery chemotherapy.

Although there seemed to be a trend for correlation between AMH levels and cortical pre-antral follicle numbers, the correlation did not reach significance (owing to the limited AMH measurements); a statistically significant correlation was noted between AMH levels and number of frozen-matured oocytes obtained from small antral follicles. This finding can be explained by AMH production in antral follicles (La Marca et al., 2010; Lie Fong et al., 2012; Leader and Baker, 2014; Lunsford et al., 2014). It should be noted that most of the antral follicles identified in our study were smaller than 8 mm, before the reduction in AMH levels occurs (La Marca et al., 2010; Lie Fong et al., 2012; Leader and Baker, 2014; Lunsford et al., 2014).

There are as yet no official policies that establish the amount of ovarian tissue that should be retrieved from paediatric patients. In an earlier study of ovarian tissue collection that also included paediatric patients (Imbert et al., 2014), the authors recommended unilateral oophorectomy for the prepubertal age group. This agrees with our policy to perform complete rather than partial unilateral oophorectomy in prepubertal girls with small ovaries.

The presence of oocytes and antral follicles in young girls, particularly the premenarcheal girls, is probably a consequence of the baseline levels of gonadotrophins during childhood (Gougeon, 1996) that stimulate follicular growth. These levels, however, are probably insufficient to promote growth beyond a certain diameter, such that follicles cannot reach ovulation size. Furthermore, the oocytes in the rinsing fluids may have originated from antral follicles that were not carefully aspirated or follicles that were not properly identified during oocyte aspiration. The lower maturation rate of oocytes derived from the rinsing fluids than oocytes derived by aspiration may be attributable to the relatively long time needed to complete ovarian tissue slicing compared with aspiration. Moreover, although one study reported the presence of pre-antral follicles in follicular aspirates (Wu et al., 1998), it is possible that they were actually cumulus-enclosed germinal vesicle-stage oocytes.

Little is known about the developmental capability of oocytes derived from paediatric patients, especially prepubertal girls, or their fertilisation and embryogenesis potential. The report of a live birth achieved after ovarian grafting of tissue cryostored at adolescence (Demeestere et al., 2015) is encouraging. However, although the patient did not yet menstruate, she was in the early pubertal stages. One study showed an increase in abnormal follicles in ovarian samples from prepubertal girls, with very little development in culture (Anderson et al., 2014), but our experience with ovarian tissue from paediatric patients does not support these findings for pre-antral follicles, either in histological sections or after culture (Abir et al., 2008; Margulis et al., 2009; Kedem et al., 2011; Lerer-Serfaty et al., 2013). Be that as it may, oocytes (aspirated from antral follicles) from paediatric patients showed relatively high rates of atresia in the present study as well as an earlier study (Revel et al., 2009) in a similar age group, and we showed a significant correlation between the percentage of atretic oocytes and patient age in the pre-chemotherapy group (reduction in atretic oocytes with increase in age). Moreover, the overall oocyte maturation rate in girls (even before chemotherapy) reported by us and by others (Revel et al., 2009; Segers et al., 2015) (18–34%) is relatively low compared with the maturation rate in women undergoing IVM in our IVF laboratory (60%) (Ben-Haroush et al., 2010; Farhi et al., 2011) and in other units (40%) (Segers et al., 2015). This suggests that even the in vitro matured oocytes from girls might not be of high quality, possibly because of the insufficient in vivo hormonal milieu and the various atretic processes that take place in antral follicles in the younger age group. In one study only four paediatric girls, all prepubertal, were included; IVM rates were not significantly lower than in older patients (Fasano et al., 2011). However, the authors did not obtain mature oocytes in their two youngest patients (age 9 months and 4 years), in line with our finding of lack of maturable oocytes in patients younger than 4-years either before or after chemotherapy.

Although we found fewer atretic oocytes in the pre-chemotherapy than post-chemotherapy group, the difference was not statistically significant. Nevertheless, there was a significant correlation between percentage of atretic oocytes and number of oocytes collected in patients after chemotherapy, which might have been attributed to the chemotherapy as well as to patient age. The reported rate of atresia of immature oocytes upon collection in regular non-primed IVM cycles is 11.8% (Son et al., 2006). However, some studies, specifically of immature oocytes from infertile women with polycystic ovary syndrome, described a high rate of meiotic abnormalities after IVM (Zhu et al., 2015).

Pregnancies and live births have been reported after IVF of prepubertal calf (Armstrong et al., 1992) and lamb oocytes (Morton, 2008). However, prepubertal oocytes from mice (de Matos et al., 2003), goats (Leoni et al., 2009), sheep (Leoni et al., 2006), cows (Armstrong et al., 1992; Palma et al., 2001) and rhesus monkeys (Zheng et al., 2001) had lower maturation, fertilization and embryogenesis rates as well as a lower cryotolerance (Leoni et al., 2009) than oocytes from adult animals.

It is also unclear if oocytes collected from girls and adolescents after chemotherapy are safe for clinical use. An earlier experimental study investigated the effect of cyclophosphamide in mice across a timeline corresponding to different stages of follicular maturation (Meirow et al., 2001). Conceptions in mice attributed to antral follicles exposed to cyclophosphamide at a late developmental stage were associated with a significantly lower rate of implantation sites than controls, and the pups had a higher malformation rate. These findings suggest that as long an interval as possible should be maintained between oocyte exposure to chemotherapy and oocyte collection.

According to a recent literature review in young women, the IVM procedure is associated with a clinical pregnancy rate of up to 35%, comparable to standard IVF/ICSI cycles (Ellenbogen et al., 2014). At 2-year follow-up, there was no increase in rates of congenital anomalies, or abnormal growth and development rates in the offspring. The success of oocyte vitrification (the cryopreservation method used in the present study) is controversial. One review and meta-analysis found a significantly lower clinical pregnancy rate from vitrified/warmed oocytes than fresh oocytes (Potdar et al., 2014), whereas other reviews and meta-analyses reported no difference (Cobo and Diaz, 2011; Herrero et al., 2011). Several studies reported lower rates of clinical pregnancy and ongoing pregnancy for slow-frozen oocytes than vitrified oocytes (Oktay et al., 2006; Cobo and Diaz, 2011; Herrero et al., 2011). Pregnancies and live births from frozen oocytes have been reported also in cancer survivors (Noyes et al., 2011; Garcia-Velasco et al., 2013). To the best of our knowledge, there are no reports of pregnancy rates after IVM or oocyte freezing conducted during childhood or adolescence.

Similar to our previous report (Abir et al., 2008), we found no significant difference in the number of pre-antral follicles between paediatric patients or patients aged ≤20 years before and after chemotherapy (Abir et al., 2008). This earlier study, however, showed a significant increase in intracellular follicle structure damage after chemotherapy (identified by transmission electron microscopy) in granulosa cells and oocytes (vacuoles).

There is increasing evidence of possible malignancy reseeding from transplanted ovarian tissue, specifically in the many paediatric patients with leukaemia (Abir et al., 2010, 2014; Dolmans et al., 2013). The use of frozen oocytes matured in vitro should avoid this risk and establish an important and safe gamete source for fertility restoration. Therefore, our centre sought to establish a protocol for maximal fertility preservation in paediatric female patients before and after chemotherapy that entails combining maturation and cryopreservation of oocytes with ovarian tissue freezing. If possible, paediatric patients should undergo immature oocyte collection and cryopreservation of ovarian tissue before undergoing any form of chemotherapy. However, if this is not possible, fertility preservation can be conducted after anti-cancer therapy has been initiated. All rinsing fluids should be carefully filtered to identify remaining immature oocytes. To further improve our protocol, immature oocyte collection, if possible, should be conducted with minimal hormonal stimulation in the operating theatre (Farhi et al., 2011), followed by partial or full unilateral oophorectomy. In this manner, oocytes can then be aspirated manually from visible antral follicles even if some were already collected at the operating theatre. In parallel, ovarian cortical slices should be prepared and cryopreserved. However, further reports on the viability and development capability of in vitro matured oocytes from paediatric patients are warranted. Any pregnancies achieved from oocytes that were matured and frozen close to the time of chemotherapy administration warrant careful fetal monitoring, including amniocentesis or PGD and ultrasound examinations.

It is possible that new techniques will be established in the coming years, providing safer options for fertility preservation in paediatric patients with cancer, especially leukaemia. These as yet experimental procedures include IVM of primordial follicles from the frozen-thawed ovarian tissue, followed by IVF of the matured oocytes and embryo transfer (Abir et al., 2006; Telfer et al., 2008) or transplantation of isolated primordial follicles embedded in an ‘artificial alginate hydrogel-based ovary’ (Luyckx et al., 2014; Soares et al., 2015).

Authors' roles

R.A. collected most of the immature oocytes, designed the study and collected the data, conducted most of the statistical analysis and wrote and revised the manuscript; I.B.-A. recruited most of the patients, helped to design the study, performed the AMH measurements and provided the funding by her grant for the AMH assay, wrote certain parts of the manuscript and critically proofed the final manuscript; R.G. collected most of the immature oocytes (especially from the rinsing medium), conducted most of the IVM and vitrification procedures, performed some of the statistical analyses, helped in collecting the data regarding the oocytes, wrote some parts of the manuscript and critically proofed the final manuscript; I.Y. recruited the patients, helped in writing the manuscript and critically proofed the final manuscript; S.A. recruited the patients, helped in writing the manuscript and critically proofed the final manuscript; S.M.S. helped in recruiting the patients, helped in designing the study, critically proofed the final manuscript; A.B-H. recruited the patients, helped in writing the manuscript and critically proofed the final manuscript; E.F. performed the surgery on the patients, wrote certain parts of the manuscript and critically proofed the final manuscript; D.K. performed the surgery on the patients and critically proofed the final manuscript; O.S. helped in designing the study, conducted some of the IVM and vitrification procedures, performed some of the statistical analyses, helped in collecting the data regarding the oocytes, wrote some parts of the manuscript and critically proofed the final manuscript; B.F. recruited some of the patients, assisted in designing the study, helped in writing the manuscript and critically proofed the final manuscript.

Funding

The study did not receive any monetary support other than routine funds of the IVF laboratory. Funding for the AMH measurements was obtained from a research grant from the Israel Science Foundation (to B.-A.I., ISF 13-1873).

Conflict of interest

None declared.

Acknowledgements

The authors are grateful to Gloria Ganzach from the Editorial Board (Beilinson Hospital, Rabin Medical Center) for the English editing. We are also grateful to Inbar Dabah-Aloush, Project Manager of the Computer Unit for assisting us in creating the graphs (Beilinson Hospital, Rabin Medical Center) and to Tzippy Shochat, statistical consultant (Beilinson Hospital, Rabin Medical Centre) for the statistical analysis of the correlation plots.