Female fertility preservation: 25 years of progress, expanding indications and future prospects

Abstract

BACKGROUND

There has been an unprecedented progress in the field of fertility preservation (FP) beginning in the late 1990s. Specifically, technological innovations, refinements in the protocols, and a deeper understanding of reproductive physiology have collectively contributed the increased success and utilization of FP methods.

OBJECTIVE AND RATIONALE

The objectives of this review are: (i) to identify the most recent and significant advances in FP, and (ii) based on evidence, to provide a comprehensive and up-to-date source of contemporary FP management approaches to guide clinicians in critical decision-making. In addition to cancer treatments, the indications for FP have expanded to include various systemic conditions such as haematological, metabolic, genetic, and immunological disorders, as well as gonadal surgery and a wish to delay childbearing. Due to the introduction of random start ovarian stimulation protocols and use of anti-oestrogen agents along with ovarian stimulation drugs, coupled with increased success with oocyte cryopreservation, improvements in ovarian tissue cryopreservation and refinements of transplantation techniques, women can now benefit from various FP options through an individualized approach.

SEARCH METHODS

We searched for peer-reviewed articles in PubMed, Embase, and Cochrane Library databases containing the key words: FP, ovarian ageing, chemotherapy, radiotherapy, embryo cryopreservation, oocyte cryopreservation, ovarian tissue cryopreservation, and in vitro follicle growth, in the English-language literature from inception to May 2025.

OUTCOMES

Cryopreservation of embryos have long been performed successfully in the field of ART. With the advent and widespread of use vitrification, the experimental tag was removed and oocyte cryopreservation was defined as a standard technique of FP. The applicability, success, and safety of random start ovarian stimulation protocols have been demonstrated in many studies including meta-analyses. Improvements in ovarian tissue cryopreservation outcomes have been reported with robotic surgery, use of neovascularizing extracellular matrix, and adjuvant pharmacotherapy. The use of GnRH analogues along with chemotherapy has been trialled as a way of avoiding the need for FP. Although the rate of premature ovarian insufficiency was reported to be lower in some patient populations treated this way, no improvements in live birth rates have been demonstrated. Among the emerging and future options are the use of ovarian tissue freezing and pharmacological approaches to delay menopause and reproductive ageing, non-suppressive gonadoprotective pharmacotherapy, in vitro gametogenesis and in vitro purging of cancer cells from ovarian tissue for cryopreservation. Animal studies have reported success with in vitro follicle growth, and progress is being made with human ovarian tissue.

WIDER IMPLICATIONS

The evolution of FP techniques has profound implications for clinical practice, not only for individuals facing fertility-compromising treatments or conditions, but also for the potential deferral of reproductive ageing. The advent of in vitro primordial follicle growth and gametogenesis may further revolutionize the landscape of reproductive medicine and FP.

REGISTRATION NUMBER

N/A.

Graphical Abstract

Advances in embryo, oocyte, and ovarian tissue cryopreservation, along with expanding indications and novel technologies, now enable individualised fertility preservation with improved outcomes.

Introduction

In an era marked by delayed family building, coupled with the advancements in cryobiology and surgical techniques, fertility preservation (FP) has gained increased attention both in cancer care and in the wider community. Advancements in cancer screening protocols, improvements of therapeutic interventions, development of novel anticancer drugs, and refinements of surgical techniques have all significantly enhanced the survival rates and life expectancy for females undergoing gonadotoxic treatments. The scope of FP has broadened to include not only patients with cancer and haematological diseases undergoing gonadotoxic treatments but also individuals with genetic, metabolic, and immunological disorders that may lead to primary ovarian insufficiency (POI).

Methods

This narrative review was informed by a structured literature search strategy designed to identify the most relevant and high-quality evidence in the field of female FP. We searched PubMed, Embase, and the Cochrane Library from database inception to June 2025 using predefined keywords, including ‘fertility preservation’, ‘ovarian ageing’, ‘chemotherapy’, ‘radiotherapy’, ‘embryo cryopreservation’, ‘oocyte cryopreservation’, ‘ovarian tissue cryopreservation’, ‘ovarian tissue transplantation’, and ‘in vitro follicle growth’. Articles were screened for eligibility based on clinical relevance, methodological quality, and date of publication. Preference was given to systematic reviews, randomized controlled trials, large cohort studies, and key mechanistic investigations. The final selection was based on expert consensus within the author team, ensuring a comprehensive and evidence-based synthesis of the current literature.

A brief overview of ovarian physiology and ovarian ageing

In a human ovary, the ovarian reserve is made up of primordial follicles (PFs) enclosing an oocyte arrested at the prophase of first meiotic division (Oktay et al., 1997). Approximately 99% of follicles undergo atresia during follicular growth and development, mainly through granulosa cell apoptosis. Although regulation of PF growth and maintenance is not fully understood, many signalling pathways, genes, enzymes, and other crucial molecules, as well as autocrine and paracrine factors, have been implicated (Rooda et al., 2025; Yano Maher et al., 2025). Because there is no neo-oogenesis in post-natal ovaries, recent research has focused on in vitro gametogenesis using pluripotent stem cells (PSCs). Some studies have demonstrated the feasibility of differentiating PSCs into germ cells in both humans and mice resulting in the formation of primordial germ cell-like (PGC) cells (Wu et al., 2023). Notably, only a limited number of studies have successfully generated granulosa cells (Lan et al., 2013). Various mechanisms have been implicated to take part in the process of ovarian ageing, such as impaired DNA double-strand break (DSB) repair mechanisms, mitochondrial dysfunction, telomerase shortening, disruption of cohesion proteins, oxidative stress, and chronic inflammation (Beverley et al., 2021; Smits et al., 2023; Suzuki et al., 2024).

Specifically, the proper functioning of BRCA genes and ataxia-telangiectasia-mutation (ATM) mediated DNA DSB repair may play a crucial role in maintaining ovarian reserve (Turan and Oktay, 2020). Due to the decline in the DNA repair function from their single healthy allele in oocytes with age, women with germline BRCA1 mutations appear to have an accelerated loss of ovarian reserve, associated with accumulation of DNA DSBs in the oocytes leading to reduced PF pool (Oktay et al., 2015; Lin et al., 2017; Turan et al., 2021). In addition, presence of BRCA mutations was found to be the most significant determinant of the persistence of post-chemotherapy amenorrhea in women with breast cancer (Oktay et al., 2023). While there were a few contradictory reports on this topic (Prokurotaite et al., 2023; Dellino et al., 2024), they have methodological weaknesses which we previously reviewed (Turan and Oktay, 2020). In human and mouse oocytes, our laboratory has shown that the expression of BRCA1 and other key ATM-mediated key DNA DSB repair genes decline with age, resulting in the accumulation of DNA DSBs in PF oocytes, which in turn triggers apoptotic follicle death (Oktay et al., 2015). We have also shown that ATM knockdown in mouse oocytes significantly inhibits or blocks the progression of meiosis in vitro, and retards or reduces embryo cleavage after parthenogenesis (Suzuki et al., 2024). Aneuploidy rates were also increased in ATM knockdown group compared to controls. Furthermore, ATM knockdown increases the sensitivity of the oocytes to a genotoxic active metabolite of cyclophosphamide, with increased formation of DNA DSBs, reduced survival, and earlier apoptotic death compared to controls. The available evidence suggested that the BRCA-related ATM-mediated DNA repair pathway is a possible candidate that regulates oocyte ageing, and the age-related decline of this pathway likely impairs oocyte health (Turan and Oktay, 2020). In another study combining single-cell RNA sequencing and spatial transcriptomics to systematically characterize human ovarian ageing, Wu et al. (2024) described FOXP1 as a regulator of ovarian ageing, as it declines with age and inhibiting CDKN1A transcription.

Some researchers have surmised that, with ageing, accumulation of mitochondrial DNA mutations and deletions result in declining mitochondrial function in oocytes. However, there has not been clear evidence in support of this hypothesis in human. Non-mitochondria-dependent energy mechanisms have also been shown to be operative in developed oocytes (Li et al., 2011; Oktay et al., 2015). Using immunofluorescence imaging on human ovarian tissue, Smits et al. (2023) demonstrated that oxidative damage by protein and lipid peroxidation starts already at the PF stage. Age-related changes in polar metabolites have implied a decrease in mitochondrial function, as demonstrated by NAD, purine, and pyrimidine depletion. On the contrary, glycolysis substrates and glutamine accumulate with age.

It has been hypothesized that ageing oocytes have undergone more cell cycles during oogenesis and have therefore accumulated more oxidative damage through the reproductive lifespan (Robinson et al., 2024). Telomeres can be shortened by reactive oxygen species or other environmental toxins leading to oocyte meiotic arrest. The female germline has minimal telomerase activity to repair the telomeres after cessation of the mitotic multiplications. It has been suggested that telomere disorders are associated with POI, embryo developmental problems, and infertility (Wang et al., 2018). An inverse correlation between telomere length and aneuploidy rates, as well as steroidogenic genes expression with increasing telomerase reverse transcriptase expression, has also been demonstrated (Longo et al., 2024). However, given the scant telomerase activity in post-mitotic germ cells, and the fact that a clear shortening of telomeres has not been shown in mature oocytes, the role of telomere shortening in oocyte ageing is not proven.

Chromosome cohesion decreases with advancing maternal age. In a human study, the inter-kinetochore distance between sister chromatids was increased significantly with maternal age, indicating weakened cohesion function (Duncan et al., 2012). Moreover, these authors observed unpaired sister chromatids from females of advanced age. Interestingly, ATM regulates cohesion subunit SMC1 and we have hypothesized, based on our recent rodent study, that the age-related decline in ATM function may cause aneuploidy in human oocytes and embryos through dysregulation of chromosome cohesion function (Suzuki et al., 2024).

Overview of mechanisms of reproductive organ damage from chemo- and radiotherapy

The long-term effects of chemotherapy and radiation primarily include massive follicular damage and stromal fibrosis leading to POI and infertility and, in some cases, uterine dysfunction.

Chemotherapy

The type, dosage, and number of courses of chemotherapy as well as pre-existing ovarian reserve are critical factors in determining a patient’s likelihood of having POI post-chemotherapy (Meirow et al., 1999). Given the fact that ovulation may occur despite a significant loss of the follicular pool, continuation of regular menstruation post-chemotherapy does not prove that the ovaries have escaped the damage (Meirow et al., 1999).

Chemotherapy exposure is associated with marked follicle loss, stromal fibrosis, and disordered extracellular matrix (ECM) composition (Shai et al., 2021; Gonzalez-Molina et al., 2022). Various mechanisms have been put forward to explain the chemotherapy-related ovarian damage, including the induction of DNA DSBs, stromal-microvascular damage, and enhanced follicle activation resulting in follicle burn-out (Bedoschi et al., 2016; Szymanska et al., 2020; Titus et al., 2021). However, our recent work with human ovarian tissue has not shown growth activation and ‘burn-out’ as a consequence of chemotherapy exposure. Single follicle RNA-sequencing from human ovarian xenografts showed that chemotherapy exposure activates pro-apoptotic pathways and does not result in follicle activation (Titus et al., 2021). Moreover, pathway analysis showed that the overall state favours the suppression of PF activation. Following cyclophosphamide exposure, no change occurred in the expressions of Akt rpS6, Foxo3a, and anti-apoptotic Bcl2 in PF oocytes by qRT-PCR and immunostaining. DNA damage and apoptosis were increased by γH2AX and active-caspase-3 staining in the PFs, whereas there was no change in the ratio of growing follicles. Following cyclophosphamide exposure, a significant decrease in the expression of anti-apoptotic pro-Akt PECAM1, IKBKE, and ANGPT1, and reduced activation of PI3K/PTEN/Akt occurred. Moreover, DNA damage, and apoptosis increased in the PFs, despite no change occurring in the ratio of growing follicles 12 h after chemotherapy. These findings suggest that the underlying mechanism causing acute follicle loss by cyclophosphamide occurs through DNA DSB-activated massive apoptosis, rather than growth activation of PF oocytes (Titus et al., 2021). Because oocytes in developing follicles may incur DNA damage during chemotherapy and still survive, they are likely to result in abnormal conceptions (Pydyn and Ataya, 1991; Meirow et al., 2001). Therefore, it is recommended that conception should not be attempted for at least 3–6 months after the exposure, to allow for the development of new antral follicle cohorts from the surviving PF pool (Meirow et al., 2001).

ATM-mediated DNA repair pathway appears to play a role in the defence against chemotherapy-induced DNA damage in human PFs (Oktay et al., 2022b; Erden and Oktay, 2025). There is evidence that some PFs can repair chemotherapy-induced DNA damage and survive. It is probable that PFs that are endowed with better DNA repair mechanisms survive chemotherapy, explaining the partial ovarian reserve loss in many instances. In an in vitro mouse oocyte BRCA knockdown bioassay, BRCA deficiency resulted in increased oocyte susceptibility to doxorubicin (Oktay et al., 2020).

Chemotherapeutic agents can be classified based on their risk of causing PF loss as very low or no risk, high risk, and unknown risk (Sonmezer and Oktay, 2004) (Table 1). Quiescent PFs are not damaged by anti-metabolite cancer drugs. In contrast, alkylating agents result in crosslinking and DNA DSBs and thus pose the greatest risk on the PF reserve through the massive activation of apoptotic pathways through these DNA DSBs (Bedoschi et al., 2016; Szymanska et al., 2020; Titus et al., 2021). The damage appears to be dose-dependent; a recent human study demonstrated that there was no difference in follicle density between patients exposed to chemotherapy with a low-dose alkylating agent <3400 mg/m² and the control group (Houeis et al., 2025). However, since the ovarian tissue cryopreservation (OTC) was performed less than 3 months after the completion of chemotherapy in that study, the long-term impact of chemotherapy exposure on ovarian follicle density could not be determined.

Table 1.

Gonadotoxic classification of chemotherapeutic agents.

Drug	Class (action)
Strongly probable ovarian damage
Nitrogen mustard	Mechlorethamine (alkylating agent)
L-phenylalanine mustard	Mechlorethamine (alkylating agent)
Chlorambucil	Chloroethylamine (alkylating agent)
Cyclophosphamide	Chloroethylamine (alkylating agent)
Melphalan	Mechlorethamine (alkylating agent)
Busulfan	Alkylalkane sulfonate (alkylating agent)
Procarbazine	Substituted hydrazine
Dacarbazine	Alkylating agent
Doxorubicin	Anthracydine
Probable ovarian damage
Cis-platinum	Heavy metal
Carmustine	Nitrosourea (alkylating agent)
Lomustine	Nitrosourea (alkylating agent)
Daunorubicin	Anthracydine
Low probability of ovarian damage
Methotrexate	Antimetabolite
Fluorouracil (5-FU)	Antimetabolite
6-mercaptopurine	Antimetabolite
Vincristine	Vinca alkaloid
Mitomycin	Antibiotic (alkylating agent)
Bleomycin	Peptide
Vinblastine	Vinca alkaloid
Cytosine arabinoside (Ara-C)	Antimetabolite
VP-16 (etoposide)	Podophyllotoxin
Imatinib	Tyrosine kinase inhibitor
Unknown probability of ovarian damage/limited evidence
VM-26	Podophyllotoxin
Vindesin	Vinca alkaloid
Ribociclib	Cyclin-dependent kinase inhibitor
Palbociclib	Cyclin-dependent kinase inhibitor
Pembrolizumab	Immune checkpoint inhibitor
Ipilimumab	Immune checkpoint inhibitor

Cisplatin is another non-cell cycle specific agent producing crosslinks with the purine bases and interfering with DNA repair mechanisms (Dasari and Tchounwou, 2014). Alternatively, doxorubicin primarily acts through intercalating into DNA and disrupting the nuclear enzyme topoisomerase type II function. Another mode of action for doxorubicin is the generation of free radicals leading to damage in cellular membranes, DNA, and proteins, as well as to the disruption of mitochondrial function (Soleimani et al., 2011a; Kciuk et al., 2023). Doxorubicin also causes significant microvascular damage in the human ovary.

There are also numerous targeted novel cancer therapeutics and their impact on reproduction remains to be determined (see Table 1). It is likely that these targeted ‘smart’ drugs will have limited systemic side effects, but for some, ovary-specific damage cannot yet be ruled out.

Radiotherapy

Ionized radiation is a well-established cause of gonadal and uterine damage (Howell and Shalet, 1998), particularly when directed to the pelvis and lower abdomen. In addition to the patient’s age and ovarian reserve, the extent of gonadal damage from ionized radiation depends on the mode (whether whole body irradiation, external pelvic radiation or brachytherapy is implemented), and total dosage of radiation, with fractionation increasing the risk of final damage (Andersen et al., 1977; Kim et al., 2021).

Granulosa cells and oocytes are particularly susceptible to radiation-induced gonadotoxicity (Wallace et al., 2003). Evidence from studies indicates that a single dose of 1 Gy irradiation can cause significant ovarian damage (Reiser et al., 2022). Using a mathematical model to calculate the radiosensitivity of the human oocyte, the lethal dose to 50% (LD50) of human oocytes was demonstrated to be <2 Gy in healthy women (Wallace et al., 2003). Women under 40 years of age typically require around 20 Gy for irreversible damage, while older women may need only about 6 Gy (Lushbaugh and Casarett, 1976). PFs generally exhibit greater resistance to radiation compared to larger maturing follicles. Radiotherapy not only directly damages germ cells but also affects the ovarian stroma, leading to vascular damage, atrophy, and eventually fibrosis (Stroud et al., 2009).

In a recent mouse study, it was demonstrated that radiation exposure with an LD50 below 50 mGy induced a massive loss of PFs (Puy et al., 2021). Although growing follicles survived doses up to 8 Gy, their long-term fertility was compromised following exposure to a 2 Gy dose. This difference in radiosensitivity was not due to a different amount of radiation-induced DNA damage, and checkpoint kinase 2 was activated in all oocytes. Surviving oocytes were able to complete folliculogenesis and could be fertilized. This allowed irradiated females to produce a single litter albeit with a high rate of foetal abortion, which emphasizes the abnormality of irradiated oocytes. Although successful full-term pregnancies can occur in patients who have received high-dose total body or pelvic irradiation, these patients may still have increased risks of obstetric complications such as early pregnancy loss, premature labour, and low birthweight due to impaired uterine growth and blood flow (Critchley et al., 2002). The available evidence suggests that uterine radiation doses of <4 Gy do not appear to impair uterine function, however, a TBI dose of 12 Gy is associated with significantly decreased pregnancy rates and increased pregnancy complications.

Overview of the expanding indications for fertility preservation

FP requires an individualized approach, as detailed in Fig. 1. The indications for and use of FP have substantially expanded over time due to greater availability of various FP options, improved cryopreservation success, multidisciplinary approaches through integration of other medical disciplines, and advancements in cancer treatments (Table 2).

Figure 1.

Algorithmic approach to female and male fertility preservation. Experimental or non-established options are shown in dotted lines. When appropriate, ovarian tissue cryopreservation can also be combined with oocyte or embryo cryopreservation to increase chances of treatment success. *Ovarian stimulation is not recommended in prepubertal girls; tissue should be cryopreserved for potential future procedures such as the in vitro follicle growth. ^#In most patients, random start protocols allow ovarian stimulation to be initiated at any day of the menstrual cycle.

Table 2.

Common and expanding and indications for fertility preservation.

Childhood cancers

Hodgkin lymphoma

Non-Hodgkin lymphoma

Leukaemia*

Neuroblastoma*

Ewing sarcoma

Wilms tumour

Genital rhabdomyosarcoma*

Pelvic osteosarcoma

Burkitt lymphoma*

Adult cancers

Haematological malignancies*

Breast cancer**

Cancer of the cervix

Squamous cell carcinoma

Adeno-/adenosquamous carcinoma*

Colon Cancer

Any disease requiring pelvic-abdominal radiation

Other cancers with a moderate to high risk of ovarian insufficiency post-treatment

Autoimmune, hematologic, and metabolic diseases

Sickle cell disease

Thalasemia

Aplastic anaemia

Systemic lupus erythematosus

Behçet’s syndrome

Rheumatoid arthritis

Steroid-resistant glomerulonephritis

Inflammatory bowel disease

Progressive systemic sclerosis

Juvenile idiopathic arthritis

Multiple sclerosis

Pemphigus vulgaris

Autoimmune thrombocytopenia

Benign ovarian disease

Benign ovarian masses requiring surgery with risk of ovarian reserve reduction

Patients receiving pelvic radiation ***

Prophylactic oophorectomy

Pathogenic BRCA and related gene variants

Any malignant or benign disease requiring preconditioning chemotherapy for hematopoietic stem cell transplantation

Most common genetic and metabolic conditions leading to premature ovarian insufficiency

Mosaic Turner syndrome

Galactosemia

Other indications

Planned fertility preservation

Transgender care

^*If ovarian tissue is cryopreserved, ovarian metastases risk should be considered.

^**Occult ovarian metastasis should be ruled out in women with BRCA and related germline pathogenic variants.

^***Rectal cancer, solid organ tumours presenting in the pelvis, Ewing’s sarcoma, osteosarcoma, retroperitoneal sarcoma, tumours of the spinal cord, idiopathic bone diseases treated with radiation therapy.

Malignant diseases

Breast cancer

Breast cancer is the most common malignancy in reproductive-aged women (Siegel et al., 2024). Many of these women are subjected to multiagent chemotherapy involving a highly gonadotoxic agent cyclophosphamide. Since breast cancer is generally hormone sensitive, safer stimulation protocols have been defined and therefore OTC is rarely performed in breast cancer settings (see below). However, for patients with pathogenic BRCA germline mutations, risk-reducing bilateral salpingectomy can concomitantly be performed with OTC (Gaba et al., 2022). However, some centres do not offer this option because of the increased risk of ovarian cancer development which may occur after the tissue is later transplanted. On the other hand, for women with BRCA and related mutations, it is our opinion that OTC can be performed at an age when the risk of ovarian cancer is low. We hypothesize that when the ovarian tissue is transplanted, it will carry the risk of ovarian cancer associated with the age at cryopreservation, not at the transplantation. Thereafter, the graft can be removed as soon as a sufficient number of embryos are obtained via IVF and cryopreserved. In addition, future developments of PF growth techniques may enable us to use these tissues without a need to transplant them back to the patient.

Haematological malignancies

Preconditioning regimens that are implemented before haemopoietic stem cell transplantation (HSCT) are highly gonadotoxic, causing immediate POI in at least 90% of the patients (Couto-Silva et al., 2001). In an observational study, the likelihood of occurrence of menarche in girls, continuing ovarian function, and incidence of POI were assessed in a follow-up study of 178 patients who were diagnosed with leukaemia and underwent HSCT (Chabut et al., 2023). In that study, >65% of patients who underwent HSCT before the age of 4.8 years had spontaneous menarche at the age of puberty, and around 50% did not have POI. On the other hand, >85% of those undergoing HSCT after the age of 10.9 years did not have spontaneous menarche and required puberty induction. These results can be explained by the larger ovarian reserve as well as the enhanced DNA repair capacity of PFs in prepubertal children. In a recent study, the most common indications for OTC in girls younger than 15 years were leukaemia, myeloproliferative diseases, and myelodysplastic syndromes (Armstrong et al., 2018). Despite the contradictory findings, the presence of a haematological malignancy may be associated with decreased ovarian reserve even in younger patients compared with healthy controls (Hussein et al., 2021; Katzir et al., 2024). This may be due to underlying compromises in systemic health, possibly affecting follicle growth dynamics.

Gynaecological cancers

Nearly 75% of reproductive-age patients diagnosed to have gynaecological cancer express a desire for future childbearing (La Rosa et al., 2020). In selected cases of early-stage gynaecological cancers, the uterus and one ovary can be preserved by fertility-sparing surgery (FSS) without compromising overall survival (Schuurman et al., 2021).

Germ cell tumours of the ovary, which are most frequently seen during reproductive years and in young girls can be successfully treated with FSS. Platinum-based multiagent chemotherapy provides a high probability of a cure at all stages of germ cell ovarian tumours with 10-year progression-free survival and overall survival rates of 80–90% (Uccello et al., 2020). The post-treatment POI rates are reported to be between 3.4% and 5%, with cumulative pregnancy rates reaching 85% (Bergamini et al., 2023). In a systematic review including 15 studies on borderline ovarian tumours (BOT), recurrence rates of 11.6% were found in patients undergoing FSS compared to 3.7% in those who underwent definitive surgery (Bercow et al., 2021). Moreover, no difference was demonstrated in the 5-year disease-free survival (DFS) rates between those undergoing FSS and standard surgery. The cumulative spontaneous pregnancy rate is around 50–60% after FSS in BOTs (Daraï et al., 2013). A recent retrospective study enrolling 275 patients suggested that ART may increase the recurrence rates in borderline tumours, but the overall survival was not impaired (Gao et al., 2024). For epithelial ovarian cancers (EOC), a recent meta-analysis encompassing 2906 patients found no difference between those undergoing FSS for stage I EOC and radical surgery in terms of DFS and recurrence rates (Guan et al., 2024). However, overall survival was found to be reduced in the FSS group. Of note, in about 15% of the BOT or early-stage EOC patients undergoing FSS, there may be a residual disease in the preserved ovary (Cacciottola et al., 2024).

In early-stage cervix cancer, FSS includes conization, simple trachelectomy, vaginal radical trachelectomy, abdominal radical trachelectomy, and laparoscopic radical trachelectomy. In many well-planned studies, pregnancy outcomes were demonstrated to be superior, with similar cancer recurrence rates, following fertility-preserving surgical approaches compared to conventional radical surgery (Nezhat et al., 2020; Vesztergom et al., 2024). High-dose oral progestin therapy alone or in combination with intrauterine progestational agents constitutes the core conservative treatment in atypical endometrial hyperplasia or stage I endometrial cancer with superficial myometrial invasion, with complete remission rates of around 70%. However, all patients should be counselled regarding the risk of cancer progression during conservative treatment and the necessity of close surveillance (Jang et al., 2024).

In a recent study assessing the utilization of OTC and ovarian tissue transplantation (OTT), it was found that 7.5% and 9.6% of patients had gynaecological cancers, respectively (Erden et al., 2024a). The return rate for OTT after gynaecological cancers (6.0%) was not different from other indications. If an ovarian stimulation cycle is scheduled, a letrozole-IVF protocol is typically recommended in patients with endometrial cancer (Azim and Oktay, 2007). Bleeding from a fragile cervix, and the risk of tumour spillage from ovarian tumours, are potential complications during an oocyte retrieval (Ingold et al., 2023).

Non-cancer systemic disorders including haematological, immunological, and genetic conditions

Haematological and immunological conditions

Radiotherapy, chemotherapy or HSCT has long been implemented to treat various haematological disorders and autoimmune diseases for patients who are refractory to conventional therapy (Condorelli and Demeestere, 2019). One study revealed that one-third of girls under the age of 15 years who have undergone OTC have done so for non-cancer indications (Armstrong et al., 2018).

Repeated blood transfusions in haematological disorders can lead to iron deposition in various organs including the ovaries. Beyond ovarian iron deposition, repeated transfusions can disrupt the functioning of the hypothalamic-pituitary-ovarian axis, resulting in hypogonadism, impaired embryo implantation, altered endometrial receptivity and, ultimately, infertility (Zhang et al., 2024). Sickle cell disease (SCD) is associated with delayed menarche and decreased pregnancy rates (Carvalho et al., 2017). Some studies with small sample sizes have shown lower anti-Müllerian hormone (AMH) levels in SCD patients treated with hydroxyurea (Elchuri et al., 2015; Silva et al., 2024). However, AMH was commonly measured while patients were receiving hydroxyurea or shortly after and it is likely that lower AMH levels reflected destruction of AMH-producing developing follicles. In fact, in a recent study assessment of digitized slides of ovarian tissue from girls and young women with SCD, who underwent OTC, demonstrated similar PF density between the hydroxyurea and non-hydroxyurea groups (Diesch-Furlanetto et al., 2024).

Systemic lupus erythematosus (SLE) is another common disorder that primarily affects women of reproductive age and patients are advised to delay pregnancy until their disease is in remission. The treatment includes antineoplastic drugs such as cyclophosphamide and methotrexate in resistant cases. In a recent meta-analysis of 13 studies including 1017 patients with SLE, serum AMH, and antral follicle counts (AFCs) were found to be significantly lower compared to healthy subjects. In subgroup analyses according to age of disease onset, adult-onset SLE patients had significantly lower AMH and AFC levels, whereas those with juvenile onset SLE showed no difference in serum AMH and FSH levels, although there was a difference in AFCs (Han et al., 2024). Nevertheless, some studies demonstrated a reduction of ovarian reserve only in those treated sequentially with conventional disease-modifying antirheumatic drugs and cyclophosphamide, as well as in those with severe disease (Di Mario et al., 2019). Moreover, good-quality blastocyst formation, implantation, and cumulative live birth rates (LBR) were demonstrated to be lower in SLE patients undergoing IVF, compared with the general infertile population (Mao et al., 2023). Similar to that seen in females with SLE, studies demonstrated reduced serum AMH levels in rheumatoid arthritis patients, who are commonly treated with methotrexate (Zhang et al., 2022). Likewise, lower AMH levels, poor IVF outcomes, and lower pregnancy rates were reported in patients with Sjogren syndrome (Mao et al., 2023). However, this is likely due to the antimetabolite effect of methotrexate on developing follicles (de Araujo et al., 2014), not the PF reserve. Hence, the true ovarian reserve is unlikely to be impacted by methotrexate treatment.

Genetic and metabolic conditions

There are various genetic and metabolic conditions where patients could benefit from FP due to accelerated loss of ovarian reserve. These include mosaic Turner syndrome, Fragile X syndrome, and classical galactosemia (Dunlop et al., 2023; Le Poulennec et al., 2024). Recently, a clinical pregnancy was reported following OTT in a patient with Turner Syndrome (Dunlop et al., 2023). In a retrospective study including 48 patients and 71 cycles, low basal FSH and high AMH concentrations, and a higher percentage of 46XX cells in the karyotype, were positively associated with the number of cryopreserved oocytes in patients with Turner syndrome (Brouillet et al., 2023). Another study reported that >80% of 89 girls with classical galactosemia had low to undetectable AMH levels (Badik et al., 2011).

Endometriosis

The presence of endometrioma itself may compromise ovarian reserve by mechanical stress on ovarian cortex, impairing granulosa cell functions, inflammatory-mediated damage, apoptosis, and reduced mitochondrial energy metabolism (Sanchez et al., 2017). In an age and calendar time-adjusted model, laparoscopically confirmed endometriosis was associated with a 50% greater risk for early menopause (Thombre Kulkarni et al., 2022). While the evidence is not robust, some studies with small sample sizes and short follow-up periods have demonstrated lower and faster declining serum AMH levels and lower AFCs in patients with endometriosis compared with healthy subjects (Nieweglowska et al., 2015; Kasapoglu et al., 2018). Some have reported lower AMH levels only in those with bilateral endometriomas (Nieweglowska et al., 2015).

An important aspect is the higher utilization rates of frozen oocytes in patients with endometriosis, reaching up to 50%, compared to 3.1–8.7% in those having their oocytes cryopreserved for malignant indications (Cobo et al., 2020; Wnuk et al., 2023). Age appears as the most significant factor to define ovarian responses and cumulative LBR in endometriosis patients undergoing oocyte cryopreservation (OC) (Cobo et al., 2020, 2021a). In an observational study, the cumulative LBR using approximately 20 oocytes was 95.4% in patients aged <35 years, versus 79.6% in women aged >35 years (Cobo et al., 2021a). Cobo et al. (2020) demonstrated that the number of frozen oocytes per cycle was higher in patients who did not undergo surgery compared to the unilateral or bilateral surgery groups; 6.2, 5.0, and 4.5, respectively. Collectively, it is crucial to counsel all endometriosis patients before a scheduled surgery, especially in young patients, since the oocyte yield is reduced by more than 50% in those undergoing ovarian surgery compared to patients with endometrioma in situ (Elizur et al., 2023). Despite this, in a European survey study including 58 fertility centres, only about half of the centres (51.7%) provided fertility counselling, especially for patients with severe endometriosis (Sänger et al., 2023).

To increase FP success, multiple cycles can also be performed. In addition to OC or embryo cryopreservation (EC), despite the lack of limited long-term data, OTC may also be considered in selected patients during the endometriosis surgery or when substantial resection of ovarian tissue or an oophorectomy is inevitable (Harzif et al., 2022). A live birth has already been reported following transplantation of frozen-thawed OTC in a woman treated by chemotherapy and subsequent endometriosis surgery (Dietl et al., 2022). The most recent ESHRE guideline recommends counselling women with severe endometriosis regarding all available FP options (Becker et al., 2022).

Planned (elective) oocyte cryopreservation

When oocyte donation or adoption is not a viable option, elective OC can be planned ideally before the age of 40 years (Hirsch et al., 2024). Low awareness, associated costs, and possible impact of social stigmata still stand as the greatest barriers against pursuing planned OC (Stevenson et al., 2021). Recent studies have demonstrated that even medical students, general practitioners, and other physicians are lacking in knowledge and awareness of age-related decline in fertility, ovarian reserve testing, and elective OC (Slater et al., 2022; Hartman et al., 2024). This underlines a strong need for providing medical professionals with the most up-to-date information. On the public side, a survey conducted in the UK revealed that 10.4% of the patients believed a single cycle was adequate to collect an acceptable number oocytes for cryopreservation (Kasaven et al., 2023). Strikingly, 11.0% believed OC may pose significant health risks and impact future fertility. Less than half agreed that the lack of knowledge on OC for age-related decline in fertility impacted the pursuit of this FP method. Among the patients who underwent planned OC, more than 90% reported no regrets about their decision to undergo oocyte freezing (Jones et al., 2020).

Current methods of fertility preservation

Embryo cryopreservation

EC is a well-established and widely utilized option for FP with a cumulative LBR around 56–65%. Overall, large amounts of data demonstrate that vitrification is superior to slow-freezing in terms of clinical outcomes, survival rates for oocytes, cleavage-stage embryos, and blastocysts (Rienzi et al., 2017). In the context of FP, complications related to ovarian stimulation and oocyte retrieval, such as ovarian hyperstimulation syndrome (OHSS), bleeding, thromboembolism, and infection carry greater risks in cancer patients compared to healthy infertile subjects. Some studies have reported a high risk of thrombosis in cancer patients undergoing OS, with 4/127 patients experiencing a venous thromboembolic event within 6 months of ovarian stimulation (Melo et al., 2022). To prevent the risk of thrombosis, prophylactic low molecular heparin can be administered in selected high-risk women experiencing OHSS with serum oestradiol >3600 pg/ml and oocyte yields >20 (Chen et al., 1997; Cavagna et al., 2018; Tsampras et al., 2025).

Ovarian stimulation in hormone-sensitive tumours

Since most breast and endometrial cancers are oestrogen sensitive, supraphysiological serum oestradiol can possibly induce tumour growth. In a recent retrospective study, follow-up data of 168 breast cancer patients with BRCA1/2 mutations who subsequently conceived either using ART (n = 22) or naturally (n = 146) were reported (Condorelli et al., 2021). Over a median follow-up of 3.4 years from pregnancy in the ART group and 5.0 years in the non-ART group, no patients died in the ART group, compared with 10 patients (6.9%) in the non-ART group. In another cohort study, over a median follow-up was 7.8 years, there was no significant difference in disease-free survival between patients with (n = 659) or without a pregnancy (n = 4073) after breast cancer (Lambertini et al., 2024). In another recent study by the same group, among 543 BRCA1/2 carriers with a pregnancy after breast cancer, ART use did not increase the risk of disease-free survival events in 107 women conceived using ART compared to 436 women who conceived spontaneously (Magaton et al., 2025).

In breast cancer, adjuvant chemotherapy is usually initiated approximately 6 weeks after the surgery providing adequate time to perform ovarian stimulation for OC or EC. In 2005, we first defined specific protocols using letrozole and tamoxifen as a safer option for ovarian stimulation in breast cancer. In that prospective study, embryo yield was higher and peak oestradiol levels were lower in the letrozole-IVF group compared with the tamoxifen-IVF group (Oktay et al., 2005). However, later experience and RCTs showed that the protocols are equally effective (Letourneau et al., 2021; Balkenende et al., 2022). In breast cancer patients undergoing ovarian stimulation with the letrozole-FSH, a GnRHa ovulation trigger was demonstrated to increase the number of cryopreserved embryos and oocytes compared with an hCG trigger, while decreasing the post-trigger oestradiol exposure and OHSS risk (Oktay et al., 2010b; Iorio et al., 2021). Most studies have demonstrated decreased ovarian reserve and potentially reduced embryo quality in women with pathogenic BRCA mutations (Finch et al., 2013; Oktay et al., 2014; Prokurotaite et al., 2023). In an individual patient-level data (IPD) meta-analysis using five data sets on 828 evaluable women, after the adjustments, those with BRCA1/2 mutations had significantly lower AMH levels compared with controls (23% lower; 95% CI, 4–38; P = 0.02). When the adjusted analysis was limited to affected women, the difference persisted (25% lower). The serum AMH levels were 33% lower in women with BRCA1 but not BRCA2 mutations compared with controls. This large IPD set from multiple international centres strongly supports that young women with BRCA pathogenic variants, particularly those affected with BRCA1 mutations, have lower serum AMH levels compared with controls (Turan et al., 2021).

With the letrozole-FSH protocol, IVM rates of germinal vesicle (GV) oocytes collected during stimulated cycles were reported to be higher than IVM rates of immature oocytes collected in stimulated cycles without using letrozole (72% vs 26.7–45.1%) (Oktay et al., 2010a). This has been hypothesized to be due to accumulating intraovarian androgen precursors enhancing follicle growth. In patients who have sufficient time before chemotherapy, consecutive letrozole-gonadotropin cycles can also be performed to increase the embryo yield without compromising overall survival (Turan et al., 2013). Even though some recent studies have reported use of progestin-primed ovarian stimulation in breast cancer, given that many breast tumours are progesterone receptor positive and that progesterone is a stimulant of breast oncogenesis, this approach may create safety concerns (Huang et al., 2022; Zhao et al., 2023).

In terms of the impact on survival, studies with up to 10 years follow-up did not demonstrate an increase in recurrence or compromised survival in patients undergoing ovarian stimulation with letrozole-FSH or tamoxifen-FSH compared to controls (Azim et al., 2008; Benvenuti et al., 2024). In another study enrolling a total of 337 women with invasive breast cancer, the hazard ratio for recurrence was 0.77 after ovarian stimulation, without a compromised survival compared with controls (Kim et al., 2016). The mean follow-ups in the FP and control groups were 5.0 and 6.9 years, respectively. In the same study, neither BRCA gene mutation nor oestrogen receptor status had an effect on survival. In another multicentre retrospective study, 7 patients in the letrozole ovarian stimulation (n = 41) and 4 patients in the conventional ovarian stimulation (n = 56) groups relapsed, respectively, whereas one patient died in each group after a median follow-up of 4 years (Goldrat et al., 2022). Of note, tumour size and HER2 status were less favourable in the letrozole group. In a recently published retrospective observational cohort study with 213 patients aged 18–43 years, 5-year disease-free survival was 80% for the FP group (n = 74) and 86% for controls (n = 141) (Shapira et al., 2025).

Ovarian stimulation with letrozole and levonorgestrel-releasing untra-uterine system (LNG-IUS) in situ and progestin primed ovarian stimulation have been suggested as safe options in women with atypical endometrial hyperplasia or endometrial cancer (Chen et al., 2021; Gallo et al., 2024). Gallo et al. (2024) reported no difference in the number of oocytes retrieved and metaphase II (MII) oocytes between patients undergoing ovarian stimulation with letrozole and LNG-IUS in situ (n = 15) and control patients (n = 30). Another study reported lower recurrence rates in patients with LNG-IUD in situ during ovarian stimulation compared with the control group (12.1% vs 35.5%) (Yin et al., 2023). Vaugon et al. (2021) reported 2-year recurrence rates as 37.7% and 55.7% in the IVF and non-IVF groups, respectively.

Oocyte cryopreservation

With the advent of vitrification, pregnancy rates have significantly improved with OC (Oktay et al., 2006), and it has become an established FP approach (Cobo et al., 2010; Rienzi et al., 2010). Additionally, every GV oocyte in these stimulated cycles can then be subjected to IVM, which can increase the mature oocyte yield significantly (Oktay et al., 2010a). Alternatively, IVM can be performed without ovarian stimulation (Nogueira et al., 2023). Improved IVM methods such as the capa-IVF have been developed to enhance the oocyte maturation process, and pregnancies have already been reported with this approach (Le et al., 2024).

In terms of OC success, a recent meta-analysis on elective FP reported oocyte survival and clinical pregnancy rates with frozen-thawed oocytes as 81.4% and 34.2% (Kirubarajan et al., 2024). In some studies, worse reproductive outcomes were reported in cancer patients (Cobo et al., 2021b). In a large study including 5289 patients in the elective FP group and 1073 patients in the cancer group, the oocyte survival rate (91.4% vs 81.2%), clinical pregnancy rate (65.9% vs 42.8%), and cumulative LBR (68.8% vs 42.1%) were higher in the age-matched elective FP patients compared with the cancer patients (Cobo et al., 2018). Conversely, other studies have reported similar oocyte survival (80.9% vs 75.4%) and pregnancy rates between cancer and non-cancer patients (31.5% vs 31%) (Porcu et al., 2022). In a recent systematic review and meta-regression analysis on planned OC, comprising 10 studies and 8750 women, the mean age at the time OC and at the time of return to use oocytes were 37.2 and 38.1, respectively (Hirsch et al., 2024). The mean number of frozen oocytes was 12.6, the post-thaw survival rate was 78.5%, and the LBR per patient was 28%. Notably, the LBR in women who underwent OC at age ≤35 years was 52%, whereas it was 19% for those who underwent OC at age ≥40 years (Cobo et al., 2020).

Some concerns were expressed about the epigenetic modifications which can potentially be induced by the cryopreservation process that includes changes in DNA and histone methylation and alterations in microRNA and mRNA expression. In this regard, a recent systematic review, analysing a total of 4159 babies with up to 6 years of follow-up reported no increased risk of adverse neonatal outcomes following OC (Da Luz et al., 2022). In the context of the impact of cryostorage time on success rates, healthy livebirths were reported for up to 13 years of storage for elective OC (Azambuja et al., 2023). Since all cellular functions are arrested with cryopreservation in subzero temperatures, no deterioration is expected to occur with prolonged cryostorage. There is only a negligible theoretical risk due to cosmic background radiation, which can potentially induce DNA DSBs.

GV oocytes are presumed to be more resistant to cryodamage due to lower cell volumes and the lack of a metaphase spindle. However, several studies have demonstrated cryo-related damage in cryopreserved immature oocytes such as meiotic spindle alterations, reduction in cortical granules, increased vacuoles, mitochondrial derangements, and partial condensation of chromosomes (Shahedi et al., 2013). Only a few pregnancies have been reported using frozen-thawed human immature oocytes thus far (Tucker et al., 1996; Wu et al., 2001).

Another critical issue to discuss with patients is the low return rates (Ter Welle-Butalid et al., 2024). Following OC, between 2.5% and 10% of the individuals were reported to return to utilize their oocytes after up to 18 years of follow-up (Hirsch et al., 2024; Kirubarajan et al., 2024). A recent study reported a significantly higher return rate in poor responders compared with normal responders (4.0 vs 2.3%) (Fouks et al., 2024). Achieving a pregnancy without the need for frozen oocytes or preferring not to have a child were among the most common reasons for not using cryopreserved oocytes.

Regarding oocyte retrieval before cryopreservation, in some selected patient groups who are ineligible for vaginal access, transabdominal oocyte retrieval using a vaginal ultrasound probe may alternatively be practiced as a safe and effective method in experienced hands.

Random start ovarian stimulation protocols

We have developed the first protocols in the emergency setting where ovarian stimulation can be started at any time during the cycle, and have coined the term random start ovarian stimulation (RSCOS) in 2011 (Fig. 2) (Sönmezer et al., 2011). Physiological principles of performing RSCOS are based on two different theories: (i) the continuous recruitment theory, and (ii) the follicular wave theory (de Mello Bianchi et al., 2010; Baerwald et al., 2012). The continuous recruitment theory proposes that follicles develop and regress continuously throughout the menstrual period. On the other hand, the follicular wave theory suggests development of two or three major follicle waves in a menstrual cycle. Our own unpublished experiences and histological observations support the continuous recruitment theory.

Figure 2.

Random start controlled ovarian stimulation protocols. If LH levels are suppressed, hMG can be added for LH replacement. In some instances, low-dose HCG can be added to the GnRH analogue (GnRHa) trigger. Late follicular phase is defined as cycle day 5 to ovulation and/or by the presence of a dominant follicle >10 mm. The ate follicular random start ovarian stimulation (RSCOS) protocol is similar to the early follicular letrozole protocol except that the GnRH antagonist (GnRH-ant) is started sooner.

In the most recent meta-analysis assessing 11 studies with 688 patients in the RSCOS arm and 1076 patients in the conventional start ovarian stimulation (CSCOS), we demonstrated that the cumulative gonadotrophin consumption was higher, and the duration of ovarian stimulation was longer in RSCOS compared with CSCOS (Sönmezer et al., 2023). The mean number of oocytes collected, MII oocytes, MII/AFC ratio, fertilization rate, and number of embryos available for freezing were all comparable between the groups. Likewise, similar clinical pregnancy and LBRs were reported in patients undergoing donor cycles with RSCOS versus CSCOS (Guerrero et al., 2024). However, possible physiological mechanisms of the prolonged duration of stimulation and increased gonadotropin use include low endogenous FSH levels suppressed by high serum oestradiol levels both in the follicular and the luteal phase, low LH levels due to luteal progesterone suppression, increased antagonist use to prevent follicular asynchrony, and molecular and hormonal differences in the intraovarian microenvironment (Nippoldt et al., 1989). Cycle outcomes were also not different when late follicular versus luteal phase start RSCOS were compared (von Wolff et al., 2016; Guerrero et al., 2024). However, some studies reported a trend towards higher mean numbers of embryos with luteal start compared to late follicular start protocols (9.4 ± 4.2 vs 6.9 ± 2.7, respectively) (Turan et al., 2023). Another study reported similar obstetric and perinatal outcomes between patients undergoing luteal phase and late follicular phase ovarian stimulation following euploid blastocyst transfer (Vaiarelli et al., 2020). In a retrospective study, the embryo euploidy rate was found to be 36.2% per patient in the RSCOS group which was comparable to the 45% aneuploidy rate from similar-aged women (Turan et al., 2023). Gonadotropin receptor expression status and enzymes involved in steroid synthesis were also not found to be different in HCG-exposed granulosa cells obtained from RSCOS versus CSCOS cycles in cancer patients (Esmaeilian et al., 2023; Galati et al., 2023).

Safety and effectiveness of ovarian stimulation and oocyte retrieval in pre- and post-pubertal children

Pre-pubertal children

OTC is generally the only option for FP in prepubertal girls, with two reported live births from cryopreserved premenarchal tissue (Demeestere et al., 2015; Matthews et al., 2018). It is not an established clinical practice to perform ovarian stimulation and oocyte retrieval during the pre/peripubertal period. Data regarding the applicability, effectiveness, and safety of ovarian stimulation and oocyte retrieval in paediatric patient populations and young adolescents are scarce and are mostly limited to individual case reports or small sample size studies (Azem et al., 2020; Martel et al., 2022). In prepubertal children, the definitive impact of an oestradiol rise following ovarian stimulation on pubertal transition is not known. Three premenarchal girls were reported in a systemic review, with successful cryopreservation of 2, 6, and 18 mature oocytes (Slonim et al., 2023). However, no data was given on breast development and accelerated bone growth. Another recent study reported neither a detrimental effect on linear growth nor an accelerated breast development in three prepubertal patients following ovarian stimulation (Meltem et al., 2024).

The American Society of Reproductive Medicine (ASRM) and American Society of Clinical Oncology (ASCO) state that OTC is currently the only way to cryopreserve gametes in prepubertal girls (Oktay et al., 2018; Fertility Preservation in Patients Undergoing Gonadotoxic Therapy or Gonadectomy: A Committee Opinion, 2019).

Post-pubertal children

Evidence is more abundant on the safety and utility of ovarian stimulation and OC in the young adolescent population. In a recent systematic review comprising of 23 studies and 468 children with a median age of 15.2, 470/488 cycles (96.3%) ended up with successful OC (median 10 oocytes) (Slonim et al., 2023). While some studies evaluating the effect of age on the oocyte yield have shown no significant differences, others indicated better oocyte yields with increasing age. For instance, patients aged 12 years or younger had a median yield of 4.5 oocytes (range 0–6), those aged 13–15 years had a median yield of 9.5 oocytes (range 0–22), and those aged 16–18 years had a median yield of 14 oocytes (range 0–35) (Slonim et al., 2023). IVM may also be successfully implemented in paediatric patient populations to increase oocyte yield (Meltem et al., 2024). In a recent retrospective study, AFC but not AMH was found to be a predictor of stimulation outcomes in pre-peripubertal age groups (Gayete-Lafuente et al., 2023). The complication rate is <1%, and one pregnancy has been reported from a female who underwent ovarian stimulation at the age of 17 years (Kim and Hong, 2011). In the largest study, that included 449 patients (306 were ≤18 years), a significantly increased risk of OHSS was observed in those younger than 20 years of age (0.9%) compared to older women (0.4%) (Hipp et al., 2019). Three patients <20 years (0.6%) were either hospitalized or developed an infection. Mild to moderate OHSS was reported in two different patients requiring hospitalization, both of which were triggered with HCG. Notably, aneuploidy rates have been suggested to be higher in very young females due to increased risk of whole-chromosome nondisjunction (Gruhn et al., 2019).

Ovarian tissue cryopreservation and transplantation

Current successes and limitations

As opposed to OC and EC, a significant advantage of OTC is its prompt implementation. OTC has some additional advantageous over OC and EC; not only does it allow freezing a large reserve of PF oocytes at once, but it also allows for spontaneous conception and resumption of ovarian function (Ozkavukcu et al., 2013). The first successful ovarian transplantation with cryopreserved tissue was performed in 1999 with restoration of ovarian endocrine function (Oktay and Karlikaya, 2000). Twenty years later, ASRM removed OTC for medical indications from the experimental category in 2019 (Fertility Preservation in Patients Undergoing Gonadotoxic Therapy or Gonadectomy: A Committee Opinion, 2019). The European Society for Human Reproduction and Embryology (ESHRE) recommended that OTC be used for patients undergoing moderate/high-risk gonadotoxic treatment where OC or EC is not feasible, or for patient preference (Anderson et al., 2020).

The Edinburgh criteria, which is a population-based validation criterion proposed to select women who will benefit from OTC in Scotland, has been highly debated (Wallace et al., 2014). It excludes patients previously exposed to chemotherapy, although numerous livebirths have already occurred in ovarian transplant recipients with ovarian tissues cryopreserved after exposure to gonadotoxic chemotherapy. Likewise, having a child should not preclude patients from utilizing OTC, as these patients may desire a larger family or wish to use ovarian tissue to restore ovarian endocrine function. Other guidelines support OTC until the age of 42 years (Backhus et al., 2007). Not only the age, but also the ovarian reserve should also guide case selection. A limited number of pregnancies have occurred following the auto-transplantation of ovarian tissues cryopreserved after the age of 35 years, whereas no pregnancy has been reported in women whose ovarian tissue was cryopreserved after the age of 40 years. Resumption of endocrine function was reported in women up to the age of 42 years (Gellert et al., 2018; Khattak et al., 2022; Marin et al., 2024). In the light of these data, we suggest performing OTC before the age of 38 years for the purpose of FP, and before 40–41 years for endocrine functions. The need for OTC has recently been debated in women with breast cancer due to reassuringly high pregnancy rates in young breast cancer patients despite the exposure to gonadotoxic chemotherapy (Macklon and De Vos, 2024; Andersen et al., 2025). In our centre, we rarely perform OTC for women with breast cancer as in nearly all cases, there is sufficient time for ovarian stimulation.

Ovarian tissue can also be transported to centralized facilities where expertise in cryopreservation exists, thereby increasing the availability of ovarian tissue freezing for patients in remote locations. We recently demonstrated that the transport of human ovarian tissue in specialized media for up to 21 hours does not alter PF integrity and does not induce aptotic death (Erden, 2025).

There are two main approaches to OTT: the orthotopic transplantation approach and the heterotopic transplant approach (Fig. 3) (Oktay and Oktem, 2010; Gook et al., 2021). Success rates with frozen-thawed OTT using IVF were relatively poor in the initial reports, with pregnancy rate per cycle of 6.9%, and LBR per cycle of 2.8% (Greve et al., 2012b). In a recent meta-analysis of 11 studies, the natural LBR and LBR with IVF following OTT were reported to be 33% per woman transplanted and 19% per IVF cycle, respectively (Fraison et al., 2023). In the same report, the restoration rate of ovarian function was 100%, while the ‘no ovarian function’ outcome was calculated as 6%. However, it is not clear whether the analysis was restricted to those who had the OT after proven menopause, and how ovarian function restoration was defined, and whether repeat transplantations were excluded. In another meta-analysis, when stricter criteria were applied to include only those with at least 12 months follow-up and clear POI criteria for transplantation, and with the exclusion of repeat transplantations, the endocrine function restoration rate was about 80% worldwide.

Figure 3.

Techniques for ovarian tissue transplantation. (A, B) Ovarian cortical pieces are strung with a 5-0 delayed absorbable suture and mounted onto a polycellulose (Surgicel) frame. This construct is then wedged retroperitoneally in the ovarian fossa with the aid of anchor sutures. This approach was described in the report of the first successful cryopreserved ovarian tissue transplantation (Oktay and Karlikaya, 2000). (C) A variation of the approach is the deposit of cortical pieces in a retroperitoneal pocket in the pelvic side wall which we do not recommend as ovarian revascularization may be suboptimal. (D) Cortical pieces mounted onto a Surgicel frame, ready to transplant. (E) Grafting of ovarian cortical pieces onto denuded menopausal ovary. (F) Insertion of ovarian cortical pieces under bluntly dissected cortical cavities. (G, H) Suturing of ovarian cortical pieces onto an extracellular matrix scaffold (ECMS) followed by the anastomosis of that construct onto the bivalved menopausal ovary with robotic surgery. When the recipient ovary is atrophic, the transplant is extended onto mesosalpinx. (I) Ovarian cortical pieces constructed on an ECMS. (J) Heterotopic transplantation of ovarian cortical pieces to the brachioradialis fascia in the forearm or (K) between the rectus abdominus sheets. (L, M) Laparoscopic heterotopic preperitoneal transplantation to the anterior abdominal wall. (N) Robotic preperitoneal ovarian transplantation with the ECMS approach to the abdominal wall, following delineation of the recipient area with blue dye. This delineation is performed to localize the graft at a location suitable for future transabdominal oocyte retrievals.

In a Danish registry study including 40 women with 53 ovarian transplantations, the endocrine function resumption rate was reported as 94% (Colmorn et al., 2022). Furthermore, grafts continued to function in 50% of the patients at 2 years follow-up after OTT. The LBR per women was reported as 41%, but pregnancy rates were higher with natural conception than after ART. The pregnancy loss rates following ART and natural conception were 43% and 20%, respectively. In an individual patient data meta-analysis including 20 studies with 568 women, the pregnancy rate was 37%, LBR was 28%, and miscarriage rates were 37% following frozen-thawed OTT (Khattak et al., 2022). The median time to return of serum FSH to <25 IU/l was 19 weeks. The pregnancy rates for frozen and fresh transplants were 37% and 52%, respectively.

The graft longevity usually ranges from 1 to 7 years following successful frozen-thawed OTT, but >10 years of graft function was also reported (Colmorn et al., 2022). In our recent study, the mean graft longevity was longer following ovarian transplantation with robotic surgery using a neovascularizing human ECM scaffold when compared to longevity from meta-analytic data (43.2 months vs 29.4 months, respectively) (Oktay et al., 2022a). Notably, that study used stricter criteria to include only studies with recipients who had clearly defined ovarian insufficiency or menopause (a minimum of 12-month amenorrhea) or who have (Erden et al., 2024b) undergone bilateral oophorectomy before OTT, and have had at least 1 year of follow up. Repeat transplant recipients were also excluded. A 25% pregnancy rate was reported from the meta-analytic data. For pregnancy outcomes, among the seven patients, one recipient did not attempt to conceive and two needed a surrogate. Of the four who attempted pregnancy with IVF or naturally, all four conceived and delivered a total of seven healthy children. Another patient is undergoing embryo transfer with a gestational surrogate and one decided to use donor oocytes with a surrogate and deferred any further oocyte retrievals from her graft (Pacheco and Oktay, 2017). In a recent systematic review, birthweights and perinatal complication rates were similar following OTT compared with the general pregnant population, with the exception of hypertensive disorders of pregnancy. However, the increase in hypertensive disorders is likely due to underlying medical conditions in these patients.

Success rates were reported to diminish with repeated transplantations and increasing age at the time of OTC (Colmorn et al., 2022; Lotz et al., 2022). Another important parameter to assess OTT success is the experience of the centre. Lotz et al. (2022) demonstrated that pregnancy rates were higher in centres having ≥10 total number of transplantations (35.1%) compared to centres with <10 transplantations (25.4%). The utilization rates for cryopreserved ovarian tissue have been reported to be between 3.9% and 8.7% (Jadoul et al., 2017; Hoekman et al., 2020; Erden et al., 2024b). This low utilization rate is due to natural conception, continued disease or death, conception via the utility of frozen oocytes or embryos (when combined forms of FP were used), and fear of disease recurrence via transplanted tissues in haematological malignancies (Erden et al., 2024b). However, given that many females have tissues cryopreserved at very young ages, the seemingly low rate of utilization could also be due to limited length of follow up and is likely to increase with time. Despite the seemingly low utilization rates, the majority of patients felt that OTC had a positive impact on their quality of life and well-being during their treatment (Ozkavukcu et al., 2013; Leflon et al., 2022).

Approach to the ovarian transplant patient

Preoperative assessment before OTT requires a detailed evaluation (Fig. 4) (Oktay and Turan, 2022). In a recent meta-analysis, it was reported that on average the transplanted tissue percentages in the first and second transplantations were 34% and 30% of one ovary, respectively (Pacheco and Oktay, 2017). Undoubtedly, as the post-transplant follicle survival rates improve, amount of tissue to be harvested and later transplanted will be less. Prior to the transplant, we place the patient on unopposed transdermal oestrogen for 4–6 weeks to enhance recipient site tissue vascularization (Oktay et al., 2022a). Likewise, we administer low-dose aspirin 10 days before the surgery to prevent microclotting and better perfusion to the graft.

Figure 4.

Preoperative evaluation and preparation for ovarian tissue transplant surgery. Preop, preoperative; ASA, aspirin. From: Oktay K (Ed), Principles of Ovarian Tissue Cryopreservation and Transplantation, First Edition, Elsevier, June 2022 (283).

While orthotopic ovarian transplants have resulted in superior quality oocytes and embryos, the restoration rate and longevity of endocrine function appear to be similar to those with the heterotopic approaches (Oktay and Marin, 2024). Thus, orthotopic ovarian transplants should be preferred when the primary aim is restoring fertility. However, when the sole goal is endocrine restoration, or when orthotopic transplantation is not feasible, heterotopic transplants may be preferred especially since they are less invasive and can easily be performed under intravenous sedation. We have developed an algorithmic approach to the ovarian transplantation site selection (Fig. 5) (Oktay and Turan, 2022). One systematic review assessed the impact of ovarian cortical tissue sizes and dimensions on OTT outcome (Diaz et al., 2022). The mean time from OTT to ovarian hormone restoration was not different among the groups with strips, squares or fragments: 3.88, 3.56, and 3 months, respectively.

Figure 5.

An algorithm for ovarian tissue transplant site selection. OTT, ovarian tissue transplantation. From: Oktay K (Ed), Principles of Ovarian Tissue Cryopreservation and Transplantation, First Edition, Elsevier, June 2022 (283).

Novel strategies to improve ovarian tissue cryopreservation and transplantation success

The most significant rate-limiting step after OTC is post-transplant ischaemia leading up to the loss of nearly two-thirds of the total PF pool. While there are losses due to freezing and thawing, minimal impact was shown on the oocyte transcriptome after the transplantation of slow-frozen/thawed human ovarian cortex (Machlin et al., 2025). Following transplantation of ovarian cortical strips, graft survival relies on the natural process of neovascularization from the graft and recipient site, which is completed in about 10 days (Oktay et al., 2001). Therefore, it is plausible that the transplant techniques that allow contact of both sides of the cortical pieces with recipient surfaces may yield greater follicle survival (i.e. use of extracellular matrices, see below).

Of the pharmacological agents tested for enhancing ovarian graft revascularization, sphingosine-1-phosphate (S1P) and its analogues may be closest to enter clinical application (Soleimani et al., 2011a). Sphingolipids, mainly S1P and ceramide-1-phosphate, are important lipid mediators with various critical actions both inside and outside the cells, such as tissue growth, cell survival, migration, and angiogenesis. In a xenograft study, the impact of S1P was assessed in ovarian grafts to enhance angiogenesis and to improve follicle survival (Soleimani et al., 2011b). It was found that S1P treatment increased vascular density, reduced ischaemic reperfusion injury, enhanced stromal cell populations, and resulted in a significant decrease in the percentage of apoptotic follicles compared to the vehicle-treated control. A synthetic analogue of S1P, Fingolimod is an FDA-approved drug that is used in the treatment of multiple sclerosis. Studies are undergoing in our laboratory to test the clinical effectiveness of this agent in ovarian graft revascularization.

There are also some positive reports on the effectiveness of gonadotropin stimulation to increase transplant success and prevent massive follicle loss, possibly through modifying VEGF and bFGF expressions (von Schönfeldt et al., 2012). However, we do not typically utilize FSH injections to improve graft revascularization since most ovarian transplant recipients already have elevated FSH levels due to ovarian insufficiency. In our clinical practice, we recommend continuous transdermal oestradiol for 4–6 weeks before ovarian grafting, followed by cyclical micronized progesterone following OTT until the resumption of ovarian function. Oestrogen use may help improve reproductive tissue quality at the transplant sites and promote angiogenesis (Rubanyi et al., 2002). In addition, we also use low-dose aspirin (80–100 mg/day) prior to OTT to prevent micro-clotting after the grafting (Oktay et al., 2022a). In addition, we use ECM scaffolds with potential ability to enhance neovascularization. Some researchers have proposed using stem cells or platelet-derived factors to improve angiogenesis, however, definitive benefits of this technology need to be clarified in future studies (Manavella et al., 2018; Chung et al., 2025).

ECM components regulate and modify PF responses to growth factors and modulate their activation, and also regulate follicle development through the stimulation of integrin receptors which activate integrin signalling pathways (Oktay et al., 2000, 2008; Oktay and Oktay, 2004; Oktem et al., 2011; Hu et al., 2022). Based on this, in a human ovarian xenograft model, we first demonstrated that ovarian cortex and Alloderm integrated well without any pathological changes, and the ECM membrane was not ovotoxic (Oktay et al., 2016). Our group then reported a successful robot-assisted ovarian transplantation using an ECM scaffold to facilitate ovarian reconstruction, handling, and revascularization, which resulted in multiple live births (Oktay et al., 2016). The procedure consists of three steps: (i) suturing the cortical pieces onto an ECM scaffold under a surgical microscope (ii) bivalving and preparation of the contralateral menopausal ovary as the recipient site, and (iii) anastomosis of the reconstructed graft to the bivalved contralateral ovary. In a further meta-analytic comparison study, the mean graft longevity was longer than the meta-analytic mean even after matching age at cryopreservation, 43.2 ± 23.6 and 9.4 ± 22.7 months, respectively (Oktay et al., 2022a). The decellularized ECM-based tissue scaffolds may also constitute a suitable niche for ex vivo culture and survival of isolated ovarian cells (Pennarossa et al., 2021).

In a mouse model, the benefit of human embryonic stem cell-derived mesenchymal progenitor cells that are incorporated to an ECM scaffold to support ovarian grafts was studied. It was found that the scaffold improved vascularization, reduced fibrosis, and normalized oestrogen and AMH production. This was associated with an increased the number of offspring compared to controls (Kim et al., 2025).

Vitrification versus slow freezing of ovarian tissue

Currently, conventional slow freezing is the established method for OTC (Isachenko et al., 2009), although recent research has demonstrated promising outcomes with vitrification (Sugishita et al., 2021). While there have been around 10 livebirths after the transplantation of vitrified ovarian cortex, the remaining live births following OTT all have been achieved with slow frozen tissue (Sänger et al., 2024). Some animal studies have demonstrated a lower percentage of intact follicles with vitrification compared to slow freezing, whereas others have documented no change in follicle density, cell proliferation, and DNA damage in ovarian stroma (Luizari Stábile et al., 2024). Using caspase-3 staining, one study reported higher apoptotic follicle death with vitrification compared to slow freezing in canine ovaries (Luizari Stábile et al., 2024). A human study did not find significant difference in various steroid gene expression patterns between the two techniques (Jaeger et al., 2023). Moreover, vitrification had no effect on the histological quality of the follicles at any stage of development compared to fresh tissue. To further improve vitrification success, various efforts have been made with limited success including encapsulation of ovaries in 1% alginate hydrogel before immersion in liquid nitrogen, changing equilibration times and temperature, antioxidant treatments, and addition of synthetic polymers to a conventional cryoprotectant solution (El Cury-Silva et al., 2024).

Safety of ovarian tissue cryopreservation and transplantation in bloodborne malignancies

The safety of performing OTT in patients with leukaemia has been debated due to the concern with reintroduction of malignant cells with the transplanted tissue (Dolmans et al., 2010; Greve et al., 2012a; Jahnukainen et al., 2013). In acute leukaemia, malignant cells are inherently present in all tissues through circulation including the ovaries (Jahnukainen et al., 2013). Most patients with acute leukaemia are initially treated with low or non-gonadototoxic induction/consolidation treatments. They are typically referred for FP when they are scheduled to receive highly gonadotoxic chemotherapy before HSCT. To reduce the risk of the presence of residual leukaemia cells in ovarian tissue, a course of chemotherapy may be allowed before the harvesting. One study demonstrated that after the first chemotherapy cycle, follicular density was not reduced although TUNEL staining revealed increased apoptosis with no difference according to the pubertal status (Devos et al., 2023). Moreover, a significantly higher percentage of gH2AX-positive follicles was observed in chemotherapy-exposed group compared to non-exposed controls, indicating DNA damage. Others demonstrated that the success of OTC was not affected by whether or not the patient received chemotherapy prior to cryopreservation, with 3-year cumulative pregnancy rates of 45% versus 57% in post-chemotherapy group versus pre-chemotherapy group (Poirot et al., 2019).

The results from animal studies are conflicting, with some signifying a possible risk of leukemic cell reimplantation, especially when ovarian tissue is found to be positive for leukaemia by PCR (Dolmans et al., 2010; Greve et al., 2012a). The degree of tumour load is critical. The few remaining leukemic cells may not be sufficient to induce a relapse of the disease. Moreover, since the leukaemia effect of the graft in severe combined immune-deficient animal models is not known, cancer cells may spread more readily, and thus the findings from xenograft experiments may not be directly translatable to humans.

The reported risks of disease relapse after OTT have been relatively low (3.9–4.2%). These relapse rates are not higher than expected and thus unlikely to be related to the ovarian grafting. Soares et al. recently reported a new live birth in an acute myeloid leukaemia survivor following frozen-thawed OTT. Her tissue was frozen when she was in complete remission. The patient remained disease-free for 3 years post-transplantation (Soares et al., 2025). In a recent case series study, we demonstrated no increased risk of leukaemia relapse over a mean follow-up of 51 months in six patients with acute leukaemia (Sönmezer et al., 2024). The lack of relapse may be ascribed to the following factors: (i), we performed OTC only in patients who received consolidation chemotherapy which has the potential to sterilize leukaemia cells within the ovary; (ii) before OTT a sample of ovarian cortical piece was carefully screened using histology, immune histochemistry, and molecular marker evaluation where appropriate to exclude leukemic infiltration; and (iii) it is unlikely that a low level of tumour burden can cause leukaemia relapse in a patient engrafted with a new bone marrow having different immunologic properties.

It has been suggested that minimal residual disease (MRD) status in the bone marrow or ovarian cortex can be used as a marker to aid the decision to perform OTC. In a mouse study, Li et al. (2024) demonstrated that mice receiving ovarian transplants showed a significantly higher incidence of death and emaciation when the patients were MRD-positive compared to the MRD-negative group. Supporting this, other studies showed that when complete remission was achieved in the bone marrow, the cryopreserved tissues were also negative for leukaemia markers (Greve et al., 2012a; Jahnukainen et al., 2013). An alternative approach to reducing the risk of leukemic cell reimplantation is ex vivo elimination of leukemic cells from the tissue. Some researchers have demonstrated promising results with ex vivo pharmacologic inhibition of YAP/TAZ oncoproteins by verteporfin, whereas others have demonstrated a favourable effect by invoking mitotic catastrophe via inhibition of Aurora B/C kinases or with a photodynamic therapy approach using tumour infiltration mimicking models (Mulder et al., 2019; Moghassemi et al., 2023).

Ovarian tissue cryopreservation and transplantation to defer reproductive ageing

With increased success of OTC for medical indications, there has been an increased interest in its utility to preserve fertility and delay menopause in healthy women. The first successful OTT with cryopreserved tissue was already performed with the primary goal of restoring ovarian endocrine function in a case where medical hormone replacement was not effective (Oktay and Karlikaya, 2000). It has thus been suggested that OTC can be used as natural means of hormone replacement (Kristensen and Andersen, 2018; Oktay et al., 2021).

In our recent study, a stochastic model of PF wastage was developed to determine the years of delay in menopause by ovarian transplantation near menopause (https://www.fertilitypreservation.org/contents/probability-calculator/nopauze-calculator) (Johnson et al., 2024). This model considered: (i) the age at ovarian tissue harvest (21–40 years), (ii) the amount of ovarian cortex harvested, (iii) transplantation of harvested tissues in single versus multiple procedures (fractionation), and (iv) post-transplant follicle survival (40% [conservative] vs 80% [improved] vs 100% [ideal or hypothetical]). According to this model, the delay in menopause for a woman with a median ovarian reserve who cryopreserves 25% of her ovarian cortex at the age of 25 years, would be approximately 11.8 years with 40% follicle survival rate, and extends to 15.5 years if the survival is 80%. Moreover, the model predicted that transplanting tissues in fractions rather than all at once extends the benefit. This model took into account the potential loss of ovarian longevity by cortical excision. Given the lack of redundancy in ovarian follicles, even the removal of one ovary prepubertally is associated with menopause 7 years earlier or menopause 1–3 years earlier with removal from most adults (Rosendahl et al., 2017). Despite the encouraging results from stochastic models, the success of OTC in extending reproductive life span and delaying menopause is still being clinically tested. A pilot study demonstrated that heterotopic transplantation of ovarian tissue (cryopreserved at the mean age of 28.2 years) in menopausal patients at the mean age of 47.4 years resulted in the successful restoration of ovarian function that continued for at least 6 months (Petrikovsky and Zharov, 2020).

Studies on women who naturally have late menopause (at >55 years of age and compromising 11% of all women experiencing menopause) have reduced health risks associated with hormone deficiency and live for longer, although the latter could be due to better DNA repair ability which is a shared advantage for both oocyte and somatic cell longevity (Oktay et al., 2015). Given its less invasive nature, subcutaneous heterotopic ovarian transplantation is suggested as a more appropriate technique for OTT to delay menopause (Oktay et al., 2021). Additionally, heterotopic OTT excludes the risk of a possible unintended pregnancy.

Combined approaches for fertility preservation

If there is adequate time, patients undergoing OTC can undergo ovarian stimulation first for OC or EC. It may be more practical to let ovaries involute before ovarian tissue harvesting to have a more even cortex to work with, and in the case of total oophorectomy, easier laparoscopic extraction through access ports. Some argue that ovarian excision following ovarian stimulation could render the ovary excessively hyperaemic, but we suggest performing ovarian stimulation in the first place to not reduce the oocyte yield after ovarian stimulation (Hourvitz et al., 2015). While waiting for menses may allow ovaries to reduce in size and make tissue harvesting easier, the time interval between the two procedures can be shortened if cancer treatments are needed to be initiated sooner. In a cohort study with a small sample size in the study group (n = 16), MII oocyte yield (8.3 vs 8.1) and the good-quality embryo rate (4.2 vs 4.4) were comparable between patients undergoing OTC followed by ovarian stimulation compared to an age-matched patients undergoing IVF (Dolmans et al., 2014). Total gonadotropin dose was higher in the study group (2440 IU) versus the control group (1681 IU). However, it is not clear how much of the ovarian cortex was removed following ovarian stimulation. In another retrospective cohort-controlled study, Puy et al. (2023) demonstrated no detrimental effect of performing OTC on the day of oocyte retrieval on ovarian follicle density or cell integrity. Nevertheless, ovarian stimulation increased haemorrhagic suffusion and oedema. The apoptotic oocyte rate and the number of small blood vessels were not statistically different between the ovarian cortex samples of those undergoing ovarian stimulation versus those who were not. In another small retrospective study with 12 patients, ovarian stimulation was started 1–3 days after ovarian tissue harvesting (Huober-Zeeb et al., 2011). In the study group, one half of an ovary was removed and outcomes were compared to 28 age-matched 28 cancer patients with a similar diagnosis who only underwent ovarian stimulation. The mean number of retrieved oocytes was similar between the groups (study group; 12.1 vs control group; 13.1). Despite the fact that the proportion of MII oocytes per total aspirated oocytes was significantly lower in the excision group (65.5% vs 73.8), fertilization rates did not differ (75% and 60% in the study and control group, respectively). However, these results should be interpreted with caution as the sample size is small, and the data are not adjusted for age.

Another combined approach is to extract oocytes from the harvested ovarian tissue or filter from spent media and then freeze them after IVM. Segers et al. (2024) investigated the outcomes of OC following IVM of immature oocytes retrieved ex vivo in 77 patients undergoing OTC. The mean oocyte maturation rate was 39% with successful OC in 64 patients (6.7 oocytes per patient). ICSI was performed in 13 patients, with resultant vitrification of a mean of 2 embryos per patient. In two patients, oocytes were thawed with survival rates of 86% and 60%, resulting in one healthy live birth. This complementary strategy may especially be of value when one ovary is already involved with cancer at the time of oophorectomy. Healthy live births were also reported from frozen-thawed IVM oocytes collected from removed ovarian tissue ex vivo in patients with ovarian tumours (Prasath et al., 2014). This approach also allows OC from prepubertal girls undergoing OTC (Sugishita and Suzuki, 2022). Collectively, maturation rates of oocytes retrieved ex vivo were reported between 31% and 57% (Segers et al., 2020; Karavani et al., 2024). Of note, recent chemotherapy exposure will have a negative effect on the number and normality of the ex vivo collected oocytes, whereas harvesting of large numbers of ovarian fragments and ovulatory status during peripubertal period are associated with a larger oocyte harvest (Abir et al., 2016). Immature oocytes recovered from very young pre-menarche girls may have lower IVM success compared to older prepubertal and post-pubertal girls (Fouks et al., 2020).

Ovarian and uterine transposition

For transposition, one or both ovaries are laparoscopically repositioned out of the radiation field. There have been various techniques to transpose ovaries including medial transposition, lateral transposition, transposition to the kidney poles, and tunnelling the ovary through the peritoneum (Sonmezer and Oktay, 2004; Liu et al., 2016; Christianson and Oktay, 2019).

Ovarian transposition yields successful results with sustained ovarian function in 16–94% of the patients (Sonmezer and Oktay, 2004; Gubbala et al., 2014). In one study, the rate of long-term ovarian function at 5 years was 60.3% in cervical cancer patients who underwent ovarian transposition (n = 27), whereas it was 0% in controls (n = 29) (Hoekman et al., 2018). Ovarian transposition can also be combined with OTC (Martin et al., 2007; Elizur et al., 2009). When patients are cured of their disease and ready to conceive, the transposed ovary can be repositioned back into the pelvis. Without the repositioning, oocytes can still be retrieved transabdominally from the transposed ovary (McLaren and Bates, 2012). Pregnancy rates were reported between 66% and 78% either naturally or using IVF, in retrospective studies following ovarian transposition (Morice et al., 1998; Terenziani et al., 2009). Of note, many of pregnancies (26/32) occurred while the ovaries were still in the transposed location.

Another strategy, the uterine displacement, appears as a novel but unestablished method to preserve fertility in selected patients with cervical and vaginal cancers, with one live birth reported already (Moretti-Marques et al., 2024).

Fertility preservation in transgender individuals

There has been an increasing demand to pursue FP by transgender individuals, likely due to increased public awareness on the utility and success of FP technologies, and changing legislation through various countries (Mattelin et al., 2022; Asseler et al., 2024).

In a survey study encompassing transgender and non-binary people, while very few participants had undergone FP (7%), 95% indicated that FP should be offered to all transgender and non-binary people (Riggs and Bartholomaeus, 2018). An increasing number of transgender women are placed on GnRH agonists prepubertally to prevent gender dysphoria. Although GnRH analogues act in a reversible fashion, some concerns exist with long-term effect on spermatogenesis when they are started before puberty, such as testicular atrophy, poor semen quality, and spermatogenesis arrest (Barnard et al., 2019; Marins et al., 2024). In a recent study, four of the six transgender females who received previous GnRH analogues produced azoospermic samples or samples with poor sperm quality (Ralph et al., 2025). It is reasonable to counsel transgender individuals before undergoing gender-affirming hormone treatments and surgeries. For transgender women for whom masturbation is not practical, viable spermatozoa can be obtained with TESE (Stolk et al., 2024). In another study, the quality of semen collected through masturbation or retrieved by surgery in transgender adolescent females was analysed. Semen parameters collected through masturbation varied but were generally abnormal, being normospermic in only 43.7% of the samples. (Ralph et al., 2025). In the surgical sperm retrieval group, electroejaculation was successful in 4 of 21 of the subjects and the rest underwent TESE. In 16 of 21 subjects, an average of 5 vials were stored. Another study found no significant difference in semen parameters in transgender adolescent patients diagnosed with cancer compared with that of cis men (Dilday et al., 2022). Behavioural factors may also affect sperm parameters in transgender women, such as wearing tight undergarments and extensive tucking (de Nie et al., 2022). In prepubertal transgender women, testicular tissue cryopreservation is still an investigational approach.

In transgender men, OTC can be performed simultaneously with gender-affirmation surgery. While some studies have reported no change in histology and tissue quality by prior testosterone treatment, others have reported inferior quality and low developmental capacity of in vitro matured oocytes from ovarian tissues of transgender men following testosterone treatment (De Roo et al., 2017; Christodoulaki et al., 2023). Normal spindle structure analysis, chromosomal alignment, and cortical follicle distribution were also reported following testosterone use (Lierman et al., 2017). Conversely, Bailie et al. (2023) demonstrated reduced follicle growth activation and increased DNA damage in cultured ovarian tissue taken from transgender men who were put on long-term testosterone treatment. Currently, no data is available on the future use, as well as the feasibility and effectiveness of transplantation of frozen-thawed ovarian tissue for the purpose of FP in transgender men.

Outcomes of ovarian stimulation were assessed in a systematic review including 23 studies with 468 transgender men aged ≤18 years (Slonim et al., 2023). Of the 488 cycles completed, 96.3% successfully resulted in cryopreserved mature oocytes (median 10 oocytes). One pregnancy was reported from a transgender male who underwent ovarian stimulation at the age of 17 years. The optimal time for ovarian stimulation in transgender man is reported either 3 months after cessation of testosterone or upon the resumption of menses (Ghofranian et al., 2023). To avoid gender dysphoria with increasing E2 levels, a letrozole-gonadotropin protocol can be used for ovarian stimulation (personal experience, K.H.O.).

Pharmacological approaches to decrease chemotherapy-induced ovarian damage

In two different meta-analyses, the rate of POI was found lower in patients exposed to GnRHa with chemotherapy compared to a chemotherapy-only group, however, multiple types of cancers and chemotherapy regimens were included (Bedaiwy et al., 2011; Del Mastro et al., 2014). Notably, in one of the meta-analyses, the beneficial effect of ovarian protection was limited to only breast cancer patients in the subgroup analysis (Del Mastro et al., 2014). In the most recent meta-analysis including 15 RCTs on breast cancer, GnRHa cotreatment significantly increased pregnancy rates, improved menstrual recovery rates, and decreased the rate of amenorrhea 1–2 years after chemotherapy (Yuan et al., 2023). Conversely, in a different, larger meta-analysis of 10 trials, no improved function was found in patients receiving GnRHa in combination with chemotherapy as opposed to those chemotherapy-only group (68% vs 60%) (Elgindy et al., 2015). Further analysis using ovarian reserve markers, such as FSH, AFC or AMH reported no benefit with GnRH agonist co-treatment (Elgindy et al., 2015). In a Cochrane meta-analysis of 12 RCTs involving 1369 patients, GnRH agonist was found to be effective in terms resumption and maintenance of menstrual functions, however, no beneficial effect was demonstrated for protection of fertility (Chen et al., 2019). In the GnRH agonist group, 326 of 447 participants (72.9%) had menstruation recovery or maintenance as opposed to the control group, in which 276 of 422 patients (65.4%) had menstruation recovery or maintenance. On the other hand, the incidence of pregnancy was 9% (32/356) in the GnRH agonist, whereas it was 6.3% (22/347) in the control group, with no difference between the groups. Another recent large population-based cohort study enrolling 24 922 patients diagnosed with cancer at ages 15–45 years, GnRHa co-treatment was not associated with higher rates of post-cancer childbirth (Rodriguez-Wallberg et al., 2024).

There has not been a double blind, placebo-controlled, randomized study to test ovarian suppression in FP, and the mixed results seen with breast cancer patients could be due to lower toxicity of drugs and positive recollection bias due to lack of blinding and placebo controlling. Studies demonstrating GnRHa cotreatment as effective are flawed due to the inclusion of patients with various disease indications, heterogeneity in chemotherapy regimens, the lack of age-matched subgroup analysis, and the non-uniformity in the definition of POI. Moreover, evidence shows that GnRH analogues are ineffective in patients with other types of malignancies especially when high-dose gonadotoxic therapy is administered, as well as in men (Waxman et al., 1987; Benor and Decherney, 2024).

Some studies have surmised a possible protective effect of novel tumour-targeted therapies including mTOR, proteasome and glycogen synthase kinase 3 inhibitors, and AS101 (Vallet et al., 2021). In mouse studies, inhibition of PI3K/Akt/mTOR pathway with an mTOR inhibitor rapamycin presumably prevented PF activation induced by cyclophosphamide (Zhou et al., 2017). Some non-anticancer molecules were reported to minimize chemotherapy-associated ovarian damage possibly by targeting apoptosis or PF activation pathways, such as sphingosine-1-phosphate, ceramide phosphate, AMH, granulocyte colony-stimulating factor alone or in combination with stem cell factor, and strong antioxidants resveratrol and dexrazoxane. Among these, sphingosine-1-phosphate is the only one with human ovarian xenograft evidence (Soleimani et al., 2011b) and potential clinical promise as it has an FDA-approved synthetic analogue.

In vitro growth of primordial follicles and in vitro gametogenesis

In vitro interventions can be combined with cryopreservation, such as in the case of in vitro activation of follicles, where ovarian stimulation alone is insufficient (Suzuki et al., 2015; Zhai et al., 2016). In vitro-grown follicles have successfully led to births only in mouse models, providing a proof of concept (Eppig and O’Brien, 1996), but there is only one published report of successful oocyte meiotic maturation from PFs in human ovarian tissue with a multi-step culture strategy up until now (McLaughlin et al., 2018).

Derangements in gap junctions and the intercellular communication between the oocyte and granulosa cells were reported during in vitro follicle culture (Telfer et al., 2023). Some argue that there is a dynamic oxygen transition from relative hypoxia in PFs compared to a greater oxygen tension in preantral follicles, and exposure of follicles to inappropriate oxygen concentrations can lead to poor follicle development and ineffective IVM (Malo et al., 2024). The follicular growth is also affected by the mechanical flexibility in the ovary. While in vitro PF growth excludes hormonal stimulation and eliminates the potential risk of malignant cell contamination, the procedure will require an extended laboratory work load. Currently, only the final stages of oocyte maturation could be recapitulated with clinical success (Telfer and Andersen, 2021). Significant challenges include optimization of culture conditions to maintain the structural integrity of growing follicles in tissue, establishment of an optimal method for viable follicle isolation, and finding the appropriate hormonal support for each stage of growth. However, in the Oktay laboratory, ongoing research has yielded hundreds of oocytes being retrieved from previously cryopreserved ovarian cortical pieces of 5 × 10 mm size, and efforts are underway for achieving fertilization (unpublished preliminary data). Therefore, IVM of human PFs has a potential to succeed, revolutionizing both the FP and the IVF field.

Efforts are also underway for complete in vitro gametogenesis where germ cells can be generated from pluripotent stem cells. Bone morphogenetic proteins (BMP) play a critical role including specification, migration, and maintenance of PGCs (Paulini and Melo, 2011). Differentiation of human PGC-like cells derived from pluripotent stem cells (hPGCLC) entails offsetting of MAPK pathway and maintenance of DNA methyltransferase activities (Theunissen et al., 2014). hPGCLC deficient in TET1, an active DNA demethylase abundant in human germ cells, fail to fully activate genes vital for spermatogenesis and oogenesis, and promoters of these genes remain methylated (Murase et al., 2024). hPGCLCs cultured with BMP2 propagate steadily with reduced levels of MAPK signalling, possibly promoting replication-coupled, passive genome-wide DNA demethylation to differentiate into mitotic pro-spermatogonia or oogonia-like cells.

Once embryonic or induced PSCs are collected, the first step is to transform these cells into PGC-like cells and induce PGC-like cells to differentiate into gametes (Bhartiya et al., 2017). However, the scarcity of stem cells poses the main hurdle for this technique. Research efforts are focused on optimizing the process of differentiating stem cells into gametes, with most studies still in the preclinical phase. To address the scarcity issue, very small embryonic-like stem cells and mesenchymal stem cells (MSCs) are being investigated for their potential in gamete production (Bhartiya, 2017). Another hurdle to overcome is to enhance stem cell engraftment and prevent rejection. To overcome this, some experts suggested human umbilical cord MSC spheroid treatments which have resulted in better treatment effects by creating resistance to autophagy, apoptosis, and oxidative damage in granulosa cells (Dai et al., 2024). Another critical issue to consider related with embryonic stem cells is the potential for tumour formation (Hyun, 2010).

Ethical aspects of fertility preservation

FP techniques such as oocyte, embryo, and OTC are associated with ethical concerns, particularly in the context of autonomy, consent, and access. One central ethical issue is ensuring informed consent, especially in vulnerable populations such as paediatric or adolescent children undergoing gonadotoxic treatments. For EC, additional ethical complexities arise concerning the disposition of unused embryos, particularly in cases of divorce or death. OC, now widely accepted for both medical and elective reasons, invites debates about equitable access and potential socioeconomic disparities, as the procedure remains costly and may not be covered by insurance (Table 3). The use of OTT to defer menopause includes some health concerns related to extended natural oestrogen production. One of the common concerns is the small increased breast cancer risk as the incidence is generally higher in women with late menopause However, women who experience menopause after age >55 have numerous health benefits such as the reduced risk of Alzheimer’s disease, osteoporosis, depression and vascular diseases, and longer lives. Whether this is the benefit of late menopause or late menopause is a surrogate marker of overall better health is currently unknown. Across all methods, considerations of future use, reproductive autonomy, and the implications of delayed childbearing require ongoing ethical scrutiny to balance technological possibilities with patient-centred care.

Table 3.

Success rates and cost-effectiveness of established fertility preservation techniques.

Technique	Success rates (%)	Advantages	Disadvantages	Limitations	Return rates	Cost
Embryo cryopreservation	9–75% (Ter Welle-Butalid et al., 2024)*	Established, widely utilized	Minimum of 10–14 days delay for OS	Limited number of embryos available	9–63% (Ter Welle-Butalid et al., 2024)	$13 350–23 000
Oocyte cryopreservation	33–50% (Ter Welle-Butalid et al., 2024)*	Established, widely utilized	Minimum of 10–14 days delay for OS	Limited number of oocytes available	2.6–21.5% (Wnuk et al., 2023)	$11 000–17 000
Ovarian tissue cryopreservation and transplantation	72–97% (Khattak et al., 2022; Oktay et al., 2022a)** 25–37% (Khattak et al., 2022; Oktay et al., 2022a)***	Resumption of endocrine function and the ability to conceive without ART Quick implementation Cryopreservation of large oocyte reserve	Not widely available Ischaemic primordial follicle loss after transplantation	Requires minimally-invasive surgical procedure with general anaesthesia	3.9–8.7% (Khattak et al., 2022; Oktay et al., 2022a)	$10 300–20 000
Ovarian transposition	16–90% (Sonmezer and Oktay, 2004)****	Preserves ovarian function. Can be performed at the time of primary cancer surgery	Not utilized if chemo is given, repositioning may be required when pregnancy is planned	May fail due to scattered radiation, vascular kinking or remigration of the ovaries	N/A	$7500–24 000

^*Cumulative live birth rates.

^**Resumption of ovarian function.

^***Pregnancy rate per transplanted patient.

^****Continuing ovarian function following the procedure.

NA, not available; OS, ovarian stimulation.

Concluding remarks and future perspectives

The FP field has experienced significant advances within the last two decades. Once seen as a luxury for cancer patients, it is now an integral part of cancer care. FP indications have expanded from cancer to non-cancer medical conditions, and eventually to healthy women who wish to extend their reproductive potential. Improvements in our understanding of both the chemotherapy-induced and physiological ageing pave the way for future pharmacological interventions that can be both gonadoprotective against cytotoxic agents and preventative against reproductive ageing. Customized ovarian stimulation protocols have made oocyte and embryo freezing both safer and more adaptable to fit with patient treatments. Ovarian tissue freezing and transplantation success is increasing and its potential role for menopause delay is being discussed. For those where OTT is not deemed safe due to cancer cell involvement, in vitro growth approaches are being developed. It appears that, based on the progress in our research laboratory, in vitro PF growth is nearing clinical success. The future is likely to present us with more non-invasive options to preserve fertility. As more options become available to safeguard future fertility, we must strive to achieve equality and ensure that all segments of the population have access to FP care.

Data availability

No new data were generated or analysed in support of this work. All data are available via published manuscripts cited in this article.

Authors’ roles

M.S. and K.H.O. prepared the first outline of the article, tables and the figures. K.G.S. and M.S. assisted in abstract screening and wrote the first draft of the manuscript. K.H.O. provided intellectual input and guidance, supervised the review process, and reviewed and extensively revised the manuscript, tables and figures. All authors meet the ICMJE criteria for authorship and all authors reviewed and approved the final version for publication.

Funding

Grant support: Eunice Kennedy Shriver National Institute of Child Health and Human Development (grant number R01HD053112 to K.H.O.).

Conflict of interest

All authors declare no conflicts of interest.