Diminished Ovarian Reserve: A Narrative Review of Etiologies and Possible Therapeutic Approaches

Abstract

Diminished ovarian reserve (DOR) occurs unintentionally during treatment or spontaneously. Despite its significant clinical manifestations, such as infertility and early menopause, and its high prevalence, most studies on DOR have focused on premature ovarian insufficiency, and reviews specifically addressing DOR remain scarce. This narrative review aims to provide insight into the diverse etiologies of DOR while discussing promising therapeutic approaches. Iatrogenic DOR can occur during chemotherapy, pelvic radiation, and ovarian surgery. Spontaneous DOR may result from ovarian tumors as well as idiopathic or genetic causes. DOR also inevitably occurs during ovarian fragment transplantation. Stem cell transplantation, in vitro activation, and platelet-rich plasma injection have shown some positive results as therapeutic approaches to DOR; however, more high-quality studies are needed to establish their broader applications in clinical practice.

Graphical Abstract

INTRODUCTION

What is diminished ovarian reserve?

There is no clear and universal definition of diminished ovarian reserve (DOR). Since ovarian reserve is defined as the number of oocytes in the ovary or oocyte quantity [1], DOR usually means a condition in which women have fewer oocytes compared to women in the same age. Clinically, DOR can be characterized by poor fertility outcomes even with assisted reproductive technology (ART) [2]. In this respect, DOR is close to poor ovarian response (POR) and DOR and POR are commonly used interchangeably, but unlike DOR, POR has a clear definition. In the Bologna European Society of Human Reproduction and Embryology consensus [3], POR is defined when women have at least two of three following features: (1) advanced maternal age (≥ 40 years) or any of the risk factors for POR, (2) a previous POR (≤ 3 oocytes with a conventional stimulation protocol), and (3) an abnormal ovarian reserve test (i.e., antral follicular count [AFC] < 5–7 follicles or anti-Müllerian hormone [AMH] < 0.5–1.1 ng/mL). In addition, in the Patient-Oriented Strategies Encompassing IndividualizeD Oocyte Number (POSEIDON) stratification [4], women are classified into 4 groups of POR in ART according to age (35 years), ovarian reserve parameters (AFC = 5 ng/mL and AMH = 1.2 ng/mL), and unexpected poor or suboptimal ovarian response. It means, by definition, age or at least one previous controlled ovarian stimulation is important for a diagnosis of POR in both the Bologna criteria and the POSEIDON stratification, whereas DOR is a more general diagnosis with clinical judgement based on abnormal ovarian reserve testing results. However, besides of ovarian reserve tests such as serum follicle-stimulating hormone (FSH), AMH, AFC, and clomiphene citrate challenge test, age or a prior cycle performance are also used to define DOR in the research [2] and the cut-offs of ovarian reserve markers are arbitrary. This heterogeneity in the definition produces confusion and difficulty in interpreting results from studies. Although there are no definitive criteria to differentiate DOR from premature ovarian insufficiency (POI), a brief comparison between the two conditions is summarized in Table 1.

Table 1. Comparison of DOR and POI based on the international guidelines and reviews.

		DOR	POI
Menstrual cycle		Usually regular	Amenorrhea or irregular
Ovarian reserve tests
	AMH (ng/mL)	< 1.1 or 1.2	Very low
	AFC (n)	< 5–7	Not specified (few or none)
	FSH (IU/L)	> 10	> 25

DOR: diminished ovarian reserve, POI: premature ovarian insufficiency, AMH: anti-Müllerian hormone, AFC: antral follicular count, FSH: follicle-stimulating hormone.

Why do we need to pay attention to diminished ovarian reserve?

Regardless of definition or cause, DOR is associated with a higher risk of a variety of health problems which can reduce quality of life.

Infertility

As DOR implies diminution of normal reproductive potential, it can cause infertility. Pregnancy and live birth rates are expected to be low even with repetitive ARTs in a woman with DOR [5] and indeed many fertility centers use POR criteria to diagnose DOR in clinical practice. Furthermore, DOR is common in infertile women. Therefore, DOR is of great importance in terms of fertility.

Early menopausal change

DOR is not necessarily a matter of pregnancy. At any given age, DOR can be translated into shortened reproductive life. Considering that women reach a menopausal state when having fewer than 1,000 remaining follicles [6], women with DOR are at higher risk of developing POI or early menopause according to their age. The more the number of oocytes decreases, the greater the risk of menopause. It is well known that young women without ovarian function would experience a variety of short-term (vasomotor symptoms) and long-term health problems (osteoporosis, cardiovascular disease, and cognitive dysfunction) [7]. Meanwhile, even DOR in young premenopausal women is significantly associated with low bone mineral density, increased bone turnover, sexual dysfunction, and sleep disturbance [8].

What should we know about diminished ovarian reserve research in humans?

Normal aging is the most common cause of DOR, but DOR is also induced unintentionally by treatment, especially during cancer treatment. Although the occurrence of DOR after treatment in women with normal ovarian function varies according to age and methods of treatment, any women who will receive possibly gonadotoxic treatment have some degrees of risk of developing DOR. Since DOR is an important reason for ART and the success rate does not improve dramatically in spite of recent developments in technologies in assisted reproduction, novel treatment options to overcome or reverse DOR is warranted.

Despite its clinical importance and prevalence, review literature on DOR is still limited, while most reviews focus on POI. This narrative review will present various causes of DOR in human and also will discuss possible therapeutic approaches to DOR. Since ovarian reserve reflects ovarian function at a specific time point, DOR is not a permanent but an intermediate step towards becoming POI. Moreover, DOR and POI can share etiologies and features. Therefore, it is difficult to think of POI and DOR separately and POI also will be addressed along with DOR in some parts of this review.

ETIOLOGIES OF DIMINISHED OVARIAN RESERVE

Iatrogenic diminished ovarian reserve

Chemotherapy

Iatrogenic DOR occurs exclusively after cancer treatment, especially chemotherapy. The ovaries are sensitive to chemotherapy, and alkylating agents are most likely to cause damage on the ovarian reserve [9]. Cyclophosphamide, an alkylating agent, is regarded as one of the most gonadotoxic chemotherapies. It binds alkyl groups to DNA covalently, induces DNA crosslinking, and consequently prevents DNA replication. This drug can damage primordial follicle population directly by inducing granulosa cell apoptosis and follicular atresia [10] or indirectly by accelerating primordial follicle activation [11]. In addition, inflammation and vasculature damage also can contribute to loss of ovarian reserve [12]. Meanwhile, cisplatin, a platinum-based compound, interferes with DNA repair mechanisms and produces apoptotic cell death [13]. It has been reported that this drug has moderate risk of amenorrhea when combined with bleomycin [14], but clinical data regarding ovarian toxicity is still limited in human. Like cyclophosphamide, cisplatin also produces the loss of primordial follicles in both direct and indirect ways. Doxorubicin is another chemotherapeutic agent that can produce ovarian toxicity. It prevents topoisomerase II-DNA complex formation which leads to accumulation of DNA fragments and induction of cell death. It also stimulates the oxidative stress [15].

The likelihood that POI will develop after chemotherapy can vary enormously, and sensitivity to gonadotoxic chemotherapy is age- and dose-dependent. In a meta-analysis of 74 studies in patients with breast cancer, the rate of chemotherapy-induced amenorrhea was increased by age showing that older age of > 40 years old was a strong risk factor for the occurrence of chemotherapy-induced amenorrhea [16]. In addition, it has been suggested that higher doses of chemotherapy correspond with higher rates of amenorrhea, although this correlation may not be consistent and a possible threshold dose for amenorrhea is not clear [17]. Of note, the risk of ovarian failure was assessed as risk of amenorrhea after chemotherapy in most clinical studies and guidelines, since initial work of American Society of Clinical Oncology [9]. However, regular menstruation does not guarantee normal ovarian function, because AMH or AFC can be significantly lower in these women which support some degrees of follicle depletion [18]. In a recent study in young breast cancer patients who underwent ovarian protection using GnRH agonist during chemotherapy [19], 95% women experienced resumption of menstruation at 1 year after doxorubicin/cyclophosphamide-based chemotherapy. However, serum AMH level was reduced by over 70% after treatment, which means the ovarian reserve did not return to the baseline. In this aspect, it can be speculated that some women may have DOR after gonadotoxic chemotherapy irrespective of having resumed menstruation.

Radiation therapy

It is well known that radiation to the female pelvis has detrimental effects on the ovaries. Non-growing follicles or primordial follicles are sensitive to radiation, although quiescent follicles are generally more resistant to radiotherapy that larger maturing follicles [20]. Dividing granulosa cells are the initial target of radiation damage, and cell death presents within hours of radiation [21]. In addition, radiation can produce vascular damage in the stroma leading to atrophy and fibrosis of tissue [22] and it also stimulates oxidative stress [23].

The degree of the damage by radiation differs by dose, field and schedule of radiation as well as age or ovarian reserve at the time of treatment [24]. Among them, dose is the most important determining factor for ovarian damage. High dose of radiation such as total body irradiation and whole abdominal irradiation usually induces POI, even during childhood [25]. LD50 (the dose destroying about 50% of non-growing follicles) is conservatively estimated to be less than 2 Gy [20] and a single oocyte is highly radiosensitive with the Do of 0.12 Gy (reciprocal of the slope on the exponential portion of a survival curve). It can be translated that sterilization is predicted in 5% by 2–3 Gy and 50% by 6–12 Gy in women whose ovaries are exposed to radiation therapy [26]. Meanwhile, age (or ovarian reserve at the initiation of radiation) is also related to the degree of radiation-induced damage. Doses of radiation which would result in ovarian failure were calculated to be 20.3 Gy at birth, 18.4 Gy at age of 10, 16.5 Gy at age of 20, and 14.3 Gy at age of 30, with a probability of 97.5% [27]. In addition, women of < 40 years old need more radiation dose for permanent damage than older women (20 Gy vs. 6 Gy) [28].

There have been several developments in the techniques to decrease the exposure of radiation to the ovaries such as novel beam arrangements, fractionation schedules, highly conformational radiotherapy, intensity modification, and shielding or transposing ovaries during radiotherapy. However, it is not possible to avoid the whole damage from radiation and doses reaching the ovaries remain high during the treatment of cancers [29], and consequently, radiation therapy can induce DOR and POI immediately or subsequently.

Ovarian surgery

Endometriotic cystectomy can cause damage on ovarian reserve by removing primordial follicles adjacent to the cyst [30]. In addition, tearing, bleeding, and coagulation during operations also cause loss of ovarian reserve [31]. Even experienced surgeons using accurate techniques cannot avoid ovarian damage completely [32]. Many studies have demonstrated the detrimental effects of endometriotic cystectomy on ovarian reserve, mainly evaluated by serum AMH level [33,34]. In a recent meta-analysis including 14 studies [35], endometriotic cystectomies are associated with a significant reduction of serum AMH level by 54.2% at the late-postoperative time point (9–18 months).

It is not surprising that excision of a non-endometriotic ovarian cyst such as cystadenomas also can cause ovarian damage. In a prospective study, serum AMH level at postoperative 1 week declined by 33.9% compared with preoperative level, although the amount of decrease was smaller and recovery was faster than endometrioma [36]. In another study, the median serum AMH level was also significantly reduced at 1 month after operation in both endometrioma and non-endometrioma group [37]. Moreover, in a meta-analysis of 367 patients from 10 studies, serum AMH level significantly decreased by 38% after excision of unilateral non-endometriotic ovarian cyst [38]. Although the risk of removing or injuring ovarian tissue is greater for endometrioma and a significant decrease in residual ovarian volume compared with the untreated contralateral ovary was not found after dermoid cystectomies [39], a possibility of ovarian damage should be considered during operations of non-endometriotic ovarian cyst as well as ovarian endometriomas.

Of note, the mean serum AMH level after surgeries was above 1.2 ng/mL in most studies, which means that many women does not meet the criteria for POR based on serum AMH level [3,4]. However, some of them surely have DOR or higher risk of developing DOR, especially when they are at advanced age or already have relatively lower ovarian reserve compared to the same age before surgery.

Spontaneous diminished ovarian reserve

Genetic

As ovarian reserve diminishes naturally with age, all women should experience DOR even without any iatrogenic cause at some point in the reproductive age. Indeed, AMH declines by approximately 5% per year in women in their third decade [40]. Apart from these natural phenomena, it is estimated that 10% of women in the general population might be at risk of early-onset DOR during their early reproductive life [41], but the cause of this “premature” DOR remains largely unknown and no specific cause is identified in most cases. Although various factors such as environmental, psychological, or physical can affect ovarian reserve, genetic defects such as chromosome or gene abnormalities play an important role in the loss of ovarian reserve [42].

Tuner syndrome is the most common sex chromosome abnormality in women. Accelerated loss of germ cells and subsequent early depletion of ovarian reserve in childhood or early adolescence is an important characteristic of Turner syndrome, although the initial migration of primordial cells may not be impaired [43]. Since ovarian reserve would be determined by the proportion of 46,XX cells in the ovary in Turner syndrome [44], patients with mosaicism such as 45,X/46,XX or 45,X/47,XXX more frequently experience spontaneous menstruation [45] and most spontaneous pregnancies (up to 5%) are observed in 45,X/46,XX mosaicism [46]. Nevertheless, patients with Turner syndrome only have a very short period for fertility, more likely in DOR state, until they are becoming POI due to the limited and rapidly decreasing ovarian reserve.

BRCA1/2 are tumor suppressor genes and BRCA mutations are associated with high risks of breast and ovarian cancer. In addition to cancer risk, BRCA mutations are also associated with decreased ovarian reserve or an earlier age of menopause [47]. As BRCA repairs double-strand DNA break and maintains embryogenesis and telomere length [48,49,50], BRCA mutations accumulate DNA damage in the oocytes which triggers apoptosis [51] and BRCA-related ataxia telangiectasia mutated-mediated repair functions a regulator of ovarian aging [52]. In 316 young women (≤ 40 years) with breast cancer, women with any BRCA mutation had a significantly lower median serum AMH level by 32%, showing a negative correlation between mutation and serum AMH level [53]. In a meta-analysis in 250 germline BRCA mutation carriers and 578 controls, women with BRCA mutation had a significantly lower serum AMH levels (23%) after adjustments, and this difference mostly resulted from BRCA1 mutation [54]. Based on these findings, possibilities of shorter fertile period and a tendency to DOR or POI should be considered in reproductive-aged women with BRCA mutations [55].

The fragile X mental retardation 1 (FMR1) gene, a regulator gene in the X chromosome, is associated with DOR as well as neuropsychiatric disorders. The FMR1 gene contains a CGG trinucleotide repeated site at the 5′ untranslated region, and this region may change repeat number during meiosis due to errors in DNA replication, repair, and recombination, and loss of oocytes occurs due to impaired granulosa cell mitosis [56]. Gene expression vary according to the number of CGG repeats, and premutation with the range of 55%200 CGG repeats causes excessive FMR1 RNA transcription and decreased protein translation, and consequently women with premutation experience a significant decrease in ovarian reserve. It has been shown that age of menopause is 5 years earlier than general population [57] and estimated incidence of POI is up to 30% [56] in women with a fragile X premutation. Even though many women with premutation will not develop POI, they probably have DOR or are at higher risk of developing DOR [58].

In addition to these known genetic disorders, various genetic variabilities such as single-nucleotide polymorphisms and non-coding RNA molecules are also associated with DOR or POI [42]. Recently whole-exome sequencing detected 79 heterozygous variants overlapped between DOR and POI, which are likely to be involved in the decline of ovarian function [59]. However, not much is known about genetic causes for spontaneous DOR with early-onset, and more research is needed.

Ovarian endometrioma

Ovarian endometriomas per se can affect ovarian reserve negatively [60]. Focal inflammation by endometrioma induces fibrosis and loss of cortex-specific stroma, and finally, recruitment and atresia of follicles are enhanced (burnout hypothesis) [61]. In addition, it has been proposed that endometrioma contains higher concentrations of free iron, reactive oxygen species, proteolytic enzymes, and inflammatory molecules which cause significant changes affecting the surrounding ovarian tissue negatively [62]. Impaired circulation by the compression of the cyst is also possible [63].

In a meta-analysis of 17 studies evaluating the effects of endometrioma on the serum AMH level, AMH was significantly lower in women with endometrioma than in women with non-endometriotic ovarian cyst or with healthy ovaries [64]. In addition, the longitudinal decline of serum AMH level was faster over 6 months in women with endometrioma compared to controls [65]. Moreover, numbers of oocytes and metaphase II oocytes retrieved were significantly lower in women with endometrioma [66,67], although it may result from technical difficulty during oocyte pick-up procedure. Taken together, current evidence indicates that endometriomas themselves are associated with a decrease in ovarian reserve which can increase the risk of developing DOR.

Diminished ovarian reserve after transplantation of frozen-thawed ovarian tissue

Ovarian tissue cryopreservation and transplantation is a safe and effective fertility preservation option now, but there has been a concern about massive depletion of follicle and becoming DOR after reimplantation of ovarian tissue [68]. Loss of ovarian reserve can occur at any stages of the cryopreservation and transplantation procedures, but it is estimated that the 80% loss of primordial follicles occurs in the post-grafting stage [69]. Ischemia and activation are two main mechanisms which contribute to the massive post-grafting follicle loss. As blood supply is stopped after removal of ovarian tissue, ischemia and hypoxia persist until neovascularization following transplantation which can take up to 10 days to provide sufficient oxygen and nutrients to the graft [70]. During this period, the grafted tissue is exposed to ischemic injury causing follicle loss [71]. In addition, massive activation of the primordial follicle pool such as significant increases in early growing follicle populations and proliferations of granulosa cells in transitional and early growing follicles, occurs shortly after transplantation, and it also contributes to loss of ovarian reserve of the graft [72].

A recent study in 285 European women who underwent transplantation of cryopreserved ovarian tissue [73] reported promising results that 88.7% had resumption of menstruation and about one-fourth gave a birth. However, a 5-year graft survival rate was 55%. In addition, the empty follicle rate was 31% and embryo transfer was possible in only 50% in women undergoing frozen-thawed ovarian tissue transplantation followed by in vitro fertilization. Based on these clinical data as well as physiology, transplantation of frozen-thawed ovarian tissue should be supposed to DOR or condition at higher risk of developing DOR.

POSSIBLE THERAPEUTIC APPROACHES

Currently there is no reliable and established treatment for DOR which can reverse loss of ovarian reserve. Although a variety of protocols and adjuvant therapies for in vitro fertilization are introduced to improve ovarian response to controlled ovarian stimulation and clinical outcomes such as pregnancy and live birth rates in infertile women with DOR [74], these methods cannot reverse DOR, and therefore, cannot be considered as a definite treatment. Following methods are now showing a possibility of being considered as therapeutic approaches in women with DOR.

Stem cell transplantation

Due to its regenerative nature of self-renewal and differentiation into various cells, stem cells are considered as a potential therapeutic option for many diseases and effects of stem cell transplantation also have been researched in POI [75]. When human amnion–derived mesenchymal stem cells were injected into tail vein or directly into ovary in cyclophosphamide and busulfan-induced POI rats [76], AMH increased in both transplantation methods and FSH levels did not differ by the route of administration. In addition, the numbers of follicles at the various stages were significantly increased in both groups compared with the POI group. In another study using menstrual-derived stem cells, the number of healthy follicles substantially increased and FSH decreased in cisplatin-induced mouse POI model [77]. A meta-analyses of 16 pre-clinical studies reported that stem cell-based therapy significantly improved FSH and estradiol levels, ovarian weight, follicle count, and the number of pregnancies in animal models of POI, mostly induced by chemotherapy [78]. However, large-scale and high-quality studies are warranted in the future due to a considerable degree of heterogeneity among the studies. Regarding DOR, infusion of human bone marrow–derived stem cells resulted in higher numbers of follicles, metaphase II oocytes, and embryos in mice DOR models, possibly mediated by promoting ovarian vascularization and follicular growth [79].

In humans, treatment using stem cells derived from several sources also has been tested for this purpose in several studies. Despite some positive results, the number of participants is small and the selection criteria are unclear [80,81].

In vitro activation

The balance between activation (PI3K pathway) and inhibition (Hippo pathway) is important in the early folliculogenesis [82], and disruption of the Hippo pathway by fragmentation of ovarian tissue with (conventional) or without (drug-free) incubation with Akt stimulants has been introduced as in vitro activation technique [83]. In a small study (n = 11), some DOR and POI patients can get pregnant using this technique [84].

Although in vitro activation might increase the number of follicles for ART and improve the pregnancy rate in women with DOR or POI, the pregnancy rate is usually very low based on case series [83]. In addition, ovarian reserve may not restore by conventional in vitro activation, and rather, this technique may have a deleterious effect on the ovarian reserve because a massive follicular activation can compromise the development of growing follicles leading to atresia [85]. Moreover, in vitro culture may trigger some damages such as necrosis or apoptosis on the ovarian tissue [86].

Platelet-rich plasma

Platelet-rich plasma is plasma with high concentrations of platelets which contains various proteins, hormones, growth factors, and cytokines [87]. Considering its regenerative potential, autologous platelet-rich plasma intraovarian injection has been studied for POR or even POI, but only in small case series [88,89]. In a recent pilot study including 120 women with POR, POI, perimenopause and menopause (30 per each group), a significant improvement on AMH and AFC was observed in POR and POI patients and menstruation was recovered in some POI and menopausal patients [89]. A recent meta-analysis of 2,256 women from 38 studies, mostly observational, reported that PRP treatment improved the main fertility parameters such as AMH and AFC, but the incidence of spontaneous pregnancy following PRP treatment was very low (7%) [90]. Further randomized trials with large numbers and a longer follow-up period are warranted.

Ovarian tissue transplantation

Re-implanted ovarian tissue after freezing and thawing could be considered as a new source of ovarian reserve as well as sex hormones [91]. This concept has been tested in two small studies with transplantation of ovarian cortical tissue into the abdominal wall [92,93] but evidence supporting use for this purpose is still scarce.

Recently, duration of ovarian activity after transplantation is getting longer and can be extended over 10 year by repeating reimplantation [94]. The better the ovarian function upon freezing (at younger age), the better the function after transplantation is expected. However, this approach is still conceptual and experimental for women with DOR or POI.

Clinical application

Despite some positive results, research is still limited to draw a clear conclusion on the application of these novel techniques in women with DOR. Most studies have shown the beneficial effects from animals, and reliable human studies are sparse. Therefore, results from uncontrolled case-series should be interpreted with caution, and these methods should be considered experimental until much more evidence can support the effects. Meanwhile, more studies on the underlying mechanisms and risks of each treatment are needed for clinical applications in human.

CONCLUSION

In human, iatrogenic DOR mostly develops during cancer treatment such as chemotherapy or pelvic radiation. As DOR is irreversible once it occurs and it has negative impact on fertility potential and even general health, attention and efforts are required to reduce the risk of occurrence. For cancer therapy-induced DOR, choosing less gonadotoxic treatment modality and using the minimum effective dose is the best strategy. Fertility preservation should be considered when the occurrence of iatrogenic or even spontaneous DOR is expected in young women desiring the future fertility. Meanwhile, reduction of ischemic injury and follicle activation will be beneficial for less prominent loss of ovarian reserve after transplantation of ovarian tissue.

Several possible therapeutic approaches to DOR have been proposed with logical reasons and positive results. However, due to lack and low quality of the available evidence, these results should be interpreted with caution, and further research is warranted to be widely applied in clinical practice.