Anti-Müllerian hormone beyond an ovarian reserve marker: the relationship with the physiology and pathology in the life-long follicle development.

AMH levels can decrease due to anticancer drugs at any age ( 49 ). However, AMH production may differ depending on whether anticancer drugs are used before or after the establishment of the follicle cohort. In cases where the follicle cohort has already been established, it is proposed that a cycle begins in which developing follicles, which are more sensitive to anticancer drugs, are damaged, resulting in decreased AMH production. The decrease in AMH thereby enhances the recruitment of primordial follicles, which are similarly damaged by anticancer drugs because of the increase in gonadotropins caused by the lack of negative feedback from the decrease in sex steroids. This mechanism is called the burnout theory ( 35 ). In childhood, on the other hand, a follicular cohort exists but has not yet been established, so it is assumed that the burnout cycle induced by AMH reduction due to developmental follicle damage is less likely to be involved. Several studies in pediatric populations (i.e., under 18 years of age at cancer diagnosis) did not detect a significant difference in post-treatment AMH versus controls despite long-term follow-up periods of 10–30 years in heterogeneous populations of childhood cancer survivors ( 50 – 52 ). This may be due to the difference in the mechanisms mentioned above; however, even in childhood, severe damage to the primordial follicle can result in follicle depletion and cause permanent menopause or diminished ovarian reserve (DOR) after the recovery of the menstrual cycle ( 53 ) (Loss of follicles due to ovarian toxicity in childhood in Figure 1 ). In most cases, AMH levels transiently decline after anticancer therapy. The persistence of undetectable serum AMH levels after treatment is useful as a predictor of continuous AMH deficiency re-elevation and menstrual recovery ( 54 ). On the other hand, the evidence is not sufficient to conclude whether prediction of permanent POI at the end of treatment may be possible by pretreatment AMH levels ( 55 ), although this has been reported to be possible in early-stage breast cancer ( 56 ). Further studies are needed to determine which patients and treatments may or may not be feasible. For endometriosis, it is assumed that a decrease in ovarian reserve due to follicular damage by local inflammation is similar to but weaker than that of anticancer drugs. Kitajima et al. reported that fibrosis of the ovarian cortex due to inflammation during endometriosis may damage follicles, leading to decreased AMH production and increased recruitment of follicles ( 57 ). However, longitudinal data on AMH levels under these conditions have not been elucidated in detail. The effect of surgery is thought to be even more direct owing to the volume reduction of the ovarian cortex ( 58 ). Comparing endometriotic cyst and benign ovarian cyst surgery, Kitajima et al. reported that the decrease in AMH was significant in the former group, with more unintended removal of the ovarian cortex in the explanted specimen ( 59 ). In contrast, Iwase et al. found that the degree of adhesion and surgical details were involved in the difference in AMH recovery after endometriotic cyst surgery. They speculated that blood flow at the time of re-establishment of the follicular cohort was affected ( 60 ). Although there are no reports quantitatively evaluating blood flow to the adnexal area and relating it to AMH variations, the possibility of a relationship between ovarian blood flow through the fallopian mesentery and the follicular cohort cannot be ruled out, based on the results of a meta-analysis evaluating the effect of tubal ablation on AMH levels ( 61 ) (Loss of follicles due to surgery and drugs in Figure 1 ).

Ovarian function can be summarized as 1) the maturation of the oocyte to a state where fertilization is possible and 2) the production of female hormones. These are mainly carried out by the oocyte itself, granulosa cells, and theca cells, all of which are components of the follicle. Therefore, ovarian function is the functional sum of the follicle. It is believed that the main factor in the decline in fertility with age is the qualitative and quantitative decline in oocytes. In other words, the concept of ovarian reserve emerged because it was thought that an index that reflects the quantity and quality of follicles in the ovary could be used to assess its impact on female reproductive function ( 1 ). However, because the qualitative evaluation of individual oocytes leading to pregnancy and live birth remains difficult for oocytes in vivo , the ovarian reserve has been established as a quantitative indicator of the oocytes remaining in the ovary ( 2 ). Research on the usefulness of anti-Müllerian hormone (AMH) as an ovarian reserve test began in 2002, and it was reported that serum AMH levels decline with age in women with normal ovulatory cycles, correlating with antral follicle count and age ( 3 ). Subsequently, several clinical applications have been developed ( 4 ). AMH is produced by follicles at a specific developmental stage and not secreted by primordial follicles; therefore, it does not directly reflect the number of follicles remaining in the ovary. The oocytes present in the ovary form a pyramid called the follicular cohort that decreases in number as the follicles mature. Therefore, AMH is an ovarian reserve marker that indirectly reflects the number of oocytes remaining in the ovary. Thus, AMH levels change, influenced by follicular cohort formation and development. This study examined the association between AMH changes and clinical utility in terms of follicular cohorts during a woman’s lifetime.

There are two periods from embryonic development to menopause, during which AMH production increases in aggregates. The first increase, which initiates AMH production, occurs at the time of primordial follicle emergence, following oocyte nest breakdown. The second increase occurred at the time of follicle cohort establishment during ovulatory cycle formation during puberty, as described earlier. The details of how this period influences the control of follicle development, and thus, ovarian reserve, are not yet known. Hashimoto disease has been reported to be associated with POI. Endocrine abnormalities and impairment of follicular growth due to autoimmunity have been postulated; however, the details of this mechanism are unknown ( 33 ). A meta-analysis of age stratification using AMH level as an indicator reported interesting results. Hasegawa et al. demonstrated that AMH levels were substantially higher in adolescents and young adults with thyroid antibodies, while AMH levels tended to decline in all age groups ( 34 ). Extensive recruitment of primordial follicles may occur under autoimmune conditions and follow the atresia of damaged growing follicles (Increased recruitment and early follicle depletion in Figure 1 ). At this stage, it is difficult to conclude whether over-recruitment leads to the subsequent loss of ovarian reserve in humans. In experimental animals, AMH knockout has been shown to result in increased primordial follicle recruitment, leading to early depletion ( 9 ). The burnout theory with cyclophosphamide, discussed below ( 35 ), is also a similar concept. However, cyclophosphamide can impair follicles at all developmental stages. Although follicle damage caused by autoimmunity has long been a matter of debate ( 36 , 37 ), no studies have clearly demonstrated an association with follicle cohort-regulated AMH, which is an issue for future research.

The next period of change in AMH production related to follicle development occurs approximately 2 years after birth. This period is considered to correspond to mini-puberty. However, changes in females during this period are less pronounced and appear later compared to those in males ( 26 ). Gonadotropins and sex steroids increase during mini-puberty, which corresponds to 6 months in boys and 2 to 4 years in girls. Transient increases in gonadotropins during mini-puberty stimulate follicular growth into antral follicles, followed by an increase in AMH and sex steroids ( 8 ). Although AMH seems to increase until approximately 8 years of age, serum levels of AMH are more stable than those of other reproductive hormones in the prepubertal periods ( 27 ). Unlike boys, the significance of mini-puberty in girls is not clear, and it is not conclusive how changes in the follicular cohort during this period affect the future ovarian reserve.

Several pathological conditions may cause the loss of follicles and lead to primary or secondary amenorrhea, including premature ovarian insufficiency (POI). Follicle formation is also a progressive process of meiosis, during which the loss of oocytes due to chromosomal abnormalities can occur. Apoptosis caused by the failure of chromosomal pairing in meiosis I is believed to be the underlying mechanism (Loss of oocytes due to meiotic abnormalities in Figure 1 .). Turner syndrome (TS) is the most prevalent of these conditions. Women with TS exhibit prenatal or postnatal loss of follicles depending on their karyotypes. While the loss of an X chromosome (45, X) results in streak gonads that exhibit complete loss of follicles, women with mosaicism, including normal chromosomal sets such as 45, X/46, XX, may exhibit spontaneous puberty ( 20 , 21 ). In contrast, independent of karyotype, AMH appeared to be a reliable marker for predicting oocyte cryopreservation success ( 22 ). Moreover, higher AMH levels in TS have been reported to be associated with spontaneous puberty and a lower FSH ( 23 ). Lunding et al. reported that serum AMH levels below a certain threshold may be useful for predicting the absence of puberty and POI ( 24 ). Although longitudinal data regarding serum AMH levels in TS are limited, oocyte cryopreservation should be considered for adolescent TS impending premature ovarian failure regardless of karyotypes ( 22 , 25 ).

Considering the actions of AMH, inhibition of primordial follicle recruitment, and FSH-dependent development, AMH production from the established follicle cohort may maintain dynamic stability in this cohort. Late puberty through the early 20s is the period of the least longitudinal change in AMH levels ( 23 , 28 ). Several ovulation disorders, including PCOS, occurred relatively frequently during this period. Ovulatory failure in PCOS is caused by a hyperandrogenic environment and relatively insufficient FSH action and is characterized by arrested development at the small antral follicle stage (Increased antral follicles due to abnormal follicle development in Figure 1 ). Hence, serum AMH levels in women with PCOS are higher than those in women with normal ovulation. Although meta-analyses have confirmed that women with PCOS of all ages have higher AMH values than age-matched control women ( 38 , 39 ), it is difficult to establish a universal AMH cut-off because AMH levels decline in an age-dependent manner ( 40 ). Relatively fewer age-dependent changes in serum AMH occur between late adolescence and the early twenties when the ovulatory cycle is established and the AMH-producing follicles are in equilibrium. Therefore, it would be helpful to understand pathophysiology based on absolute AMH values. Tsukui et al. reported in a meta-analysis that AMH is useful as an adjunct diagnosis for PCOS in this age group and that its cutoff can be relatively narrowed ( 41 ). In this age group, differentiating PCOS from functional hypothalamic amenorrhea (FHA) has become a clinical issue considering that polycystic ovarian morphology—a PCOS triad—is often found in young FHA women ( 42 ). Although differential diagnosis is not difficult based on clinical findings such as body weight and hyperandrogenism, AMH levels may also be useful. FHA is caused by the impairment of the hypothalamus-pituitary system, resulting in decreased gonadotropin secretion and anovulation due to the absence of gonadotropin-dependent follicular development. In addition, persistently low gonadotropin levels cause a decrease in the number of AMH-producing follicles, in which growth is regulated in a gonadotropin-sensitive manner.

Elevated AMH levels have been reported in several tumors, including granulosa cell tumors (GCTs), wherein tumorigenesis results in the growth of AMH-producing cells. GCTs are the most common subtype of ovarian sex cord-stromal tumors, accounting for 2–5% of all ovarian cancers ( 62 ). GCTs are divided into two subtypes. The juvenile type occurs primarily in children, while the adult type typically occurs in postmenopausal women. The usefulness of AMH as a tumor marker has been reported in numerous studies. In a recent meta-analysis evaluating the performance of AMH in the diagnosis of GCT, the pooled sensitivity was 89%, with a pooled specificity of 93% ( 63 ).

Much is unknown about follicle formation and follicle cohort establishment in the human ovary. Since oocytes are not reproduced after birth, the process of follicle cohort establishment from birth to puberty and sexual maturity may affect the subsequent ovarian reserve ( Figure 1 ). Although follicles in the developmental stage below the small antral follicle are difficult to observe in vivo , follicles at this stage, excluding primordial follicles, producing AMH. Thus, AMH measurement as an ovarian reserve test provides clues to dynamic changes in the follicle cohort through an indirect estimation of the total number of remaining follicles. Longitudinal AMH measurements provide clues regarding the lifetime of the ovary and may provide clues for future interventions for ovarian longevity.

The subsequent period that influences follicular development is puberty, which includes the period of prepubertal follicle formation and the subsequent ovulatory cycles. Although gonadotropin secretion is low during the prepubertal stage, follicles have established their development into the gonadotropin-sensitive stage. AMH is measurable but low at this stage, as shown in a cross-sectional study ( 28 ). The process of establishing a follicular cohort, followed by cyclic menstruation, begins with an increase in gonadotropins during puberty ( 29 ). The decline in AMH, correlating with the transition during the prepubertal-pubertal phase, indicated that follicle development had entered the gonadotropin-dependent phase. Follicles in the AMH-producing stage develop into later stages in which AMH is not produced. However, follicle development at this stage is not accompanied by sufficient FSH stimulation, leading to ovulation. Subsequently, the serum AMH levels transiently decreased. Given that AMH suppresses recruitment from primordial follicles, this transient decrease in AMH levels may encourage increased recruitment to establish a follicular cohort for cyclic ovulation. Central precocious puberty is induced by gonadotropin-stimulated ovarian estrogen secretion. Follicular development should also occur at this time; however, the behavior of AMH levels is not conclusive ( 13 ). In female precocious puberty, serum AMH levels have been reported to be similar to those in healthy controls, which suppress GnRH analogue therapy ( 30 ). Although AMH production declines, estradiol production increases owing to the development of AMH-producing follicles to the upper stage; however, it is not expected to show a constant increase because it does not lead to an oscillation in cyclic follicle development. Jeffery et al. confirmed the AMH model in girls, showing a modest peak around 9 years of age followed by a weak decrease at 10–12 years of age, just before menarche. They also showed the inverse relationship between AMH and FSH before puberty that disappears at the onset of puberty ( 31 ). Although the drive for this change has not been fully established, it may be explained by the different modes of gonadotropin control over the two processes of follicle development, that is, recruitment from the primordial follicle to the growing follicle pool and development from AMH-producing follicles to AMH-nonproducing follicles. Specifically, serum estradiol levels gradually increase during the prepubertal years, possibly corresponding to a transient decrease in AMH ( 32 ).

AI: Conceptualization, Writing – original draft. YH: Resources, Visualization, Writing – review & editing. YT: Resources, Visualization, Writing – review & editing. MK: Resources, Visualization, Writing – review & editing. HH: Resources, Visualization, Writing – review & editing. TN: Resources, Visualization, Writing – review & editing. YK: Supervision, Writing – review & editing.