Premature Ovarian Insufficiency of Genetic Cause

Approach to premature ovarian insufficiency (POI) and its class of genetic causes: X chromosome abnormalities, the FMR1 premutation (fragile X-associated POI), isolated monogenic forms (OMIM series POF1–26 and ODG1–11: genes of ovarian development, follicular growth, meiosis and DNA repair) and syndromic forms (BPES due to FOXL2*, classic galactosaemia, autoimmune polyendocrinopathy type 1, ovarioleukodystrophies, MCM8/MCM9 syndromes, chromosomal instability syndromes and mitochondrial syndromes).*

Literature and regulatory information updated: 6 November 2026.

0. Initial Reference Table

0.1. Table of the genetic entities of the class

1. Introduction

1.1. A syndrome of endocrinological definition with a predominantly genetic component

Premature ovarian insufficiency (POI) is a syndrome and not a disease: it designates the loss of ovarian activity before the age of 40 years, recognised by the combination of menstrual disturbance with elevated gonadotrophins and low oestradiol, that is, a hypergonadotrophic hypogonadism that the World Health Organization (WHO) classification places in type III anovulation [1, 2]. The median age at natural menopause in Western populations is close to 51 years, with a range considered normal between 45 and 55 years; cessation between 40 and 45 years is termed early menopause and lies outside the scope of POI [1, 12]. Physiological menopause occurs when fewer than about 1000 follicles remain, and twin studies attribute 60–80% of the variability in the age at which it occurs to genetics, or 45–85% according to the review of the 2024 international guideline [1, 11]. This article covers POI as a syndromic class of genetic cause: it describes the common syndrome, the current classifications and the table of all genetic entities, and examines in depth those that are most frequent or carry their own clinical consequences; Turner syndrome, fragile X, the galactosaemias, BPES, Perrault syndrome and Fanconi anaemia are covered only with regard to ovarian function.

1.2. From "early menopause" to premature ovarian insufficiency

The condition has been designated over time by different terms, chiefly "premature ovarian failure" (POF), "early menopause" and "primary ovarian insufficiency" [1]. The 2021 French PNDS recommends abandoning "early menopause", "primary ovarian insufficiency" and "primitive ovarian insufficiency", because of their negative psychological impact when the diagnosis is disclosed, and replacing them with "premature ovarian insufficiency"; it also advises against "46,XX gonadal dysgenesis" in isolated POI, because the term evokes a difference of sex development (DSD) that does not exist in these patients [1]. The word "insufficiency" conveys that the dysfunction is not necessarily permanent: ovarian function may fluctuate for years and spontaneous pregnancies occur [1]. The 2024 ESHRE/ASRM/CRE-WHiRL/IMS guideline adopts the same term, premature ovarian insufficiency, as a good practice point [2]. In genetic databases, the OMIM series "premature ovarian failure" (POF1 to POF26) and "ovarian dysgenesis" (ODG1 to ODG11) are retained for historical reasons and, in the text, correspond to POI with post-pubertal onset or to ovarian dysgenesis with primary amenorrhoea [9, 10]. POI is also distinguished from diminished ovarian reserve, defined by regular cycles in an infertile woman with low anti-Müllerian hormone (AMH), a low antral follicle count and, sometimes, poor response to stimulation; the 2024 guideline does not apply to this latter situation [1, 2].

1.3. Historical and molecular discovery milestones

The first familial descriptions of 46,XX ovarian dysgenesis with streak gonads, normal stature and frequent consanguinity date from 1959 to 1971 (Elliott 1959; Christakos 1969; Simpson 1971, according to OMIM) [10]. The association between blepharophimosis and female infertility was delineated by Zlotogora, Sagi and Cohen in 1983, who separated type I, with ovarian failure and transmitted almost exclusively by men, from type II, without a gonadal phenotype [78]. In 1995, Aittomäki and colleagues mapped a locus on 2p in Finnish families with recessively transmitted hypergonadotrophic ovarian dysgenesis and identified the variant c.566C>T, p.(Ala189Val) in the FSH receptor, the first gene of isolated POI [40]. The FMR1 premutation was established as a cause of POI around 2000, with a frequency of about 20% in carriers versus 1% in the population (Sherman 2000, according to GeneReviews) [15]. Crisponi and colleagues cloned FOXL2 in 2001 as the cause of BPES, with truncated proteins in type I and longer proteins in type II [79]. In 2004, Di Pasquale and colleagues described the first human BMP15 variant in two sisters with ovarian dysgenesis [80]; NOBOX followed in 2007 [81], FIGLA in 2008 [45] and NR5A1 in 2009, in a study of four families with 46,XY DSD and 46,XX POI [48]. A 2015 review estimated that cytogenetics, array comparative genomic hybridisation (array-CGH) and exome sequencing identified a genetic cause in only 20–25% of POI [82, 83]. From 2011 onwards, exome sequencing brought the meiosis and DNA repair genes: PSMC3IP (2011), STAG3 (2014), SYCE1 (2014), HFM1 (2014), MCM9 (2014) and MCM8 (2015) [51–53, 56, 84, 85]. The exome cohorts of 2019 to 2023 changed the scale of the problem: 375 European and North African patients with a yield of 29.3% [13] and 1030 Chinese patients with pathogenic or likely pathogenic variants in 23.5% [14]. In the opposite direction, the analysis of 104,733 women in the UK Biobank showed that 99.9% of heterozygous truncating variants in the described genes were in women with a normal reproductive lifespan, which led to the revision of many dominant models to recessive ones [19].

1.4. Current classifications

The aetiologies of POI are divided into primary (chromosomal, genic and autoimmune) and secondary or iatrogenic (surgical, chemotoxic or radiotoxic and infectious), with most cases remaining without an identified cause after the available investigation [1]. The PNDS groups the genes into four functional categories: ovarian development and early establishment of the follicular pool (for example, NR5A1, also involved in the adrenal gland), DNA repair and meiosis (the most numerous causal family, with a very unfavourable fertility prognosis and, for some genes, tumour susceptibility), follicular growth (FSHR, GDF9, BMP15, favouring residual reserve if the diagnosis is made early) and follicular atresia (few genes) [1]. The 2024 international guideline adds the categories of metabolism and mitochondria (Perrault, galactosaemia), immunity (AIRE) and RNA metabolism (FMR1), and distributes the yield of the European cohorts by pathway: DNA repair and meiosis in 37.4%, follicular growth in 35.4% and metabolism/mitochondria in 19% [11]. Clinically, isolated forms, in which POI is the only manifestation, are distinguished from syndromic forms, in which it is part of a multisystem disease; in the cohort of Heddar and colleagues, POI was the only initial manifestation of a multisystem disease in 8.5% of patients [13]. Presentation is further divided into primary amenorrhoea, with a greater genetic and chromosomal contribution, and secondary amenorrhoea [11, 14].

1.5. Definition of the entities within scope

The article covers: the chromosomal abnormalities that cause POI (X monosomy and mosaics, structural Xq rearrangements, X;autosome translocations, trisomy X), covered from the ovarian perspective; POI associated with the FMR1 premutation (FXPOI, fragile X-associated primary ovarian insufficiency), defined as hypergonadotrophic hypogonadism before the age of 40 years in a carrier of 55 to 200 CGG repeats [15]; the isolated monogenic forms of the POF and ODG series; and the syndromic forms with POI as an obligatory or frequent component (BPES type I, classic galactosaemia, autoimmune polyendocrinopathy type 1, ovarioleukodystrophies, MCM8/MCM9 syndromes, chromosomal instability syndromes, Perrault syndrome, pseudohypoparathyroidism 1A and mitochondrial diseases). Non-genetic iatrogenic and autoimmune causes are described in the differential diagnosis and in prevention, because they determine the sequence of investigation and fertility preservation.

2. Molecular Genetics and Pathophysiology

2.1. The follicular pool and the three mechanisms of POI

After migration of the primordial germ cells to the genital ridge, the oogonia proliferate by mitosis up to a maximum at around the 20th week of gestation and then decline through atresia, leaving about 1–2 million follicles at birth (Baker 1963, according to the Oxford Desk Reference) [12]. POI results from one of three mechanisms, which may coexist: formation of a reduced primordial follicle pool during fetal life, accelerated follicular atresia or a block in follicular maturation at any stage [1]. OMIM illustrates the first two with the Turner variants, in which the follicles form and undergo rapid atresia, and with galactosaemia, in which few follicles form; an ovary insensitive to gonadotrophins, as in inactivation of the FSH receptor, may mimic failure with follicles present [9]. The course of the sequelae depends on the aetiology: surgical POI causes an abrupt fall in oestrogens and androgens, whereas in spontaneous POI oestrogens fluctuate for years and the ovaries continue to produce androgen precursors [1]. Germ cell maturation requires two active X chromosomes, and about 20% of the genes on the inactive X escape inactivation, which explains the sensitivity of the ovary to X haploinsufficiency [11].

2.2. Chromosomal abnormalities

Chromosomal abnormalities account for 10–13% of POI, mostly involving the X chromosome, and are more frequent in primary amenorrhoea (21% abnormal karyotypes) than in secondary amenorrhoea (11%) and at younger ages [1, 11]. In the Finnish registry of 5011 women with POI (1988–2017), the odds ratio (OR) was 275 (95% confidence interval (95% CI) 68.1–1110) for Turner syndrome and 12.7 (4.1–39.1) for other sex chromosome abnormalities [86]. Turner syndrome affects about 25–50/100,000 live-born females (1/2500 female births according to the PNDS); haploinsufficiency causes accelerated loss of primordial oocytes during fetal life and streak gonads, although the ovaries of 45,X fetuses are almost normal early in gestation [1, 11, 36]. In a series of 410 Danish adults, the karyotypes were 45,X in 49%, mosaicism with a structural abnormality of the second X in 23%, 45,X/46,XX in 19% and 46,XX with a structural abnormality of the second X in 9% (Birkebaek 2002, according to the Oxford Desk Reference) [87]. Interpretation of low-level 45,X/46,XX mosaicism is difficult because X loss in lymphocytes increases with age, from 1.5–2.5% of cells in prepubertal girls to about 4.5–5% in women over 75 years (Guttenbach 1995); in one series, the mean percentage of cells with a single X was 5.50 in POI versus 2.42 in age-matched controls (Devi 1998, both according to the Oxford Desk Reference) [12]. The regions Xq13–q21 and Xq23–q27 are considered critical: structural X abnormalities explain 4–12% of POI, and POI occurs in about 50% of X translocations, particularly when the breakpoints fall in these regions, through gene disruption or down-regulation, position effects or meiotic errors [11, 38]. The genes DIAPH2 (POF2A) and POF1B (POF2B), located in the POF2 region of Xq21, are provisional OMIM entries, based on an X;autosome breakpoint in one family and on five homozygous Lebanese sisters, respectively [9, 64, 65]; the mapping of 25 Xq rearrangements and a position effect described in a balanced t(X;1) reinforce the role of chromosomal architecture and not only of individual genes [88, 89]. In trisomy X (1/1000–2000 female births, only about 10% diagnosed), two case-control studies found lower AMH and a series of 269 Italian POI cases found 0.7% 47,XXX karyotypes (five times more than expected), with causality still uncertain [11, 39]. The presence of Y chromosome material implies a gonadal tumour risk of 10–30% (gonadoblastoma, dysgerminoma); 10–12% of women with 45,X/46,XX Turner syndrome have detectable Y material, and gonadoblastoma was found in 4/11 (36.4%) and 6/34 (18%) of women with Turner syndrome and Y material in two series [11].

2.3. FMR1 premutation

The FMR1 gene (Xq27.3; reference transcript NM_002024) contains in its 5′ untranslated region a CGG repeat whose number defines four allele classes, with limits that lack sharp boundaries: normal, about 5–44 repeats; intermediate or "grey zone", 45–54 (about 14% unstable, able to expand to a premutation through maternal transmission, but not directly to a full mutation); premutation, 55–200 (upper limit sometimes up to about 230); and full mutation, more than 200, hypermethylated and silenced [15, 16]. The limits used by the PNDS (fewer than 45 normal; 45–55 grey zone; 56–200 premutation) and by the Oxford Desk Reference (premutation 59–200) differ: an allele of 55 repeats, which is a premutation by GeneReviews limits, falls in the PNDS grey zone; the reference limits are those of GeneReviews, and the ± 2–3 repeat precision of the methods requires, in any case, cautious interpretation of results close to the boundaries [1, 15, 90]. POI occurs with the premutation and not with the full mutation: the mechanism attributed is a toxic gain of function of RNA with expanded repeats and production of the FMRpolyG protein through non-canonical translation, and full mutation carriers have neither increased risk nor diminished reserve (Avraham 2017 and Bennett 2010, according to GeneReviews and ESHRE) [11, 15]. The relationship between size and risk is non-linear and bell-shaped: the ORs for FXPOI were 6.9 for 59–79 CGG, 25.1 for 80–99 CGG and 16.4 for more than 100 CGG [15, 91]; the approximate absolute risks cited in one review were 10%, 32% and 16% for 55–79, 80–100 and more than 100 repeats [92]. In a meta-analysis of 3394 idiopathic POI cases and 8461 controls, the OR of the premutation was 8.13 (95% CI 4.35–15.19) for POI and 14.87 (5.20–42.52) for diminished reserve [11, 93]. Intermediate alleles (45–54) were not associated with POI or early menopause in the British population study [94]. The role of AGG interruptions in transmission stability is established, but their effect on the risk of FXPOI has contradictory results [15, 92]. A family history of early menopause and smoking modify the age at onset: smokers develop POI on average five years earlier [15, 95].

2.4. Follicular growth genes: FSHR, BMP15 and GDF9

The FSH receptor (FSHR, 2p16.3; NM_000145.4) causes ovarian dysgenesis 1 (ODG1) through biallelic loss of function; the Finnish founder variant NM_000145.4:c.566C>T, p.(Ala189Val), in the extracellular binding domain, markedly reduces binding capacity and signal transduction, with apparently normal affinity [40]. In Finnish patients with this variant, ultrasound showed follicles in 6/8 versus 1/11 of other dysgeneses and histology in 9/9 versus 1/4, that is, the ovary has follicles but does not respond to FSH, either endogenous or high-dose recombinant [12, 16]. Monoallelic activating FSHR variants cause familial gestational ovarian hyperstimulation syndrome (#608115, AD), and the polymorphisms rs6165 and rs6166 modulate the response to stimulation without a Mendelian effect [16]. BMP15 (Xp11.22; NM_005448.2) encodes an oocyte-specific factor that stimulates folliculogenesis and granulosa proliferation; the first case was that of two sisters heterozygous for NM_005448.2:c.704A>G, p.(Tyr235Cys), inherited from their unaffected hemizygous father, with a dominant negative effect in vitro [41, 80]. Later evidence reclassified the gene as having a biallelic effect: null homozygosity in two girls with primary amenorrhoea whose heterozygous mothers had a normal menopause [96], compound heterozygosity for two frameshifts in sisters with persistent reserve [97], homozygous p.(Cys320Tyr) in a consanguineous Palestinian family with ovarian dysgenesis, with a 3.8-fold reduction in signalling in vitro and homology with the sterility of homozygous ewes [98] and, in a French cohort of 500 women with POI or diminished reserve, variants in 54 (10.8%), mostly benign or of uncertain significance, with a single loss-of-function homozygote whose heterozygous mother had menarche at 16 and menopause at 53 years [17]. For GDF9 (5q31.1), the same team found variants in 36/1281 (2.8%) women, without an excess of frequent missense variants relative to the population, and one loss-of-function homozygote with a normal heterozygous mother, which favours a recessive model [18]; the first homozygous case, a 1 bp deletion, was described in Brazil [99]. The PNDS stresses that a defect in these follicular growth genes favours persistence of ovarian reserve if the diagnosis is made before atresia destroys it [1].

2.5. Ovarian development genes and transcription factors

NOBOX (7q35) encodes a homeodomain transcription factor expressed in primordial germ cells, oocytes and granulosa cells, which regulates GDF9, BMP15, POU5F1, KITLG and FOXL2; Nobox-/- mice lose all their oocytes in the early postnatal period, whereas heterozygotes have normal primordial follicle formation [100]. The initial estimate of 6.2% of POI due to heterozygous variants [101] was revised in a cohort of 810 patients, in which 25% had at least one NOBOX variant: p.(Arg44Leu) was classified as benign, p.(Asp452Asn) and p.(Arg117Trp) as moderate-risk with very low penetrance, and only p.(Gly91Trp) and p.(Gly152Arg) as pathogenic with moderate impact, with effects dependent on ethnic origin (authors' protein nomenclature) [43]. The most severe and convincing cases are biallelic: two Belgian sisters compound heterozygous for truncating variants, a consanguineous family with homozygous p.(Val350Leu) in the homeodomain and a carrier mother with regular cycles at 43 years, and Brazilian sisters homozygous for c.1489del [44, 100, 102]. OMIM retains POF5 as AD on the basis of the original family of Qin and colleagues [42, 81]; GeneGuidance considers a monoallelic missense variant a risk or uncertain finding and recommends testing sisters before intrafamilial oocyte donation [16]. FIGLA (2p13.3) was described in the heterozygous state and later in the homozygous state in two sisters with primary amenorrhoea, one of them with short stature and delayed bone age; in an Indian series of 219 women, the association of two of seven variants with POI is a signal of association and not of causality [45, 46, 103]. NR5A1 (9q33.3; steroidogenic factor 1, SF1) regulates the hypothalamic-pituitary-steroidogenic axis; in the study by Lourenço and colleagues, the four families with 46,XY DSD and 46,XX POI and two of 25 sporadic POI cases had variants that reduced transactivation, without clinical adrenal insufficiency, and with variable expressivity and penetrance within families [47, 48]. Homozygosity for NR5A1 causes adrenal insufficiency without POI, and there are 46,XX girls with a salt-wasting crisis in the first weeks of life [47]. BNC1 (15q25.2; POF16) was implicated by a 5 bp deletion segregating in a large Chinese family and by missense variants in 4/82 idiopathic POI cases, absent from 332 controls [49]; a 15q25.2 microdeletion including CPEB1 and BNC1 was found in two Estonian patients with primary amenorrhoea [104].

2.6. Meiosis, homologous recombination and DNA repair genes

This is the most numerous gene family and the one with the most unfavourable fertility prognosis, because follicular loss occurs very early [1]. STAG3 (7q22.1) encodes a meiosis-specific subunit of the cohesin ring; Caburet and colleagues identified, in a consanguineous Palestinian family, a homozygous 1 bp deletion with frameshift, and Stag3-/- females are sterile, with fetal oocytes arrested in prophase I and oocyte depletion by the first week of life [51]. In the original family, four of six sisters had primary amenorrhoea diagnosed at 17–20 years and streak gonads, and the youngest developed at 19 years a synchronous bilateral ovarian cancer (gonadoblastoma on the right; dysgerminoma, embryonal carcinoma and choriocarcinoma on the left) [50]. In 2026, a Russian series described the recurrent variant c.2776C>T, p.(Arg926Ter) in six patients from five families with early-onset POI (five with primary amenorrhoea), with a frequency in the local database 18 times higher than in the Genome Aggregation Database (gnomAD) (0.001184 versus 0.000065) [105]. Men with biallelic STAG3 variants have azoospermia with meiotic arrest, accounting for less than 1% of non-obstructive azoospermia [106]. SYCE1 (10q26.3), a component of the synaptonemal complex, causes recessive POI in sisters born to Israeli Arab first-cousin parents who were homozygous for a nonsense variant [52]. HFM1 (1p22.2), a meiotic recombination helicase, was described in compound heterozygous sisters with POI; in men it is associated with non-obstructive azoospermia [16, 53]. MSH4 and MSH5 encode the MutSγ complex, which stabilises Holliday junctions; biallelic MSH4 variants are an established cause of POI and non-obstructive azoospermia, the presentation is mainly secondary amenorrhoea, sometimes diminished reserve, and the later onset has implications for cascade screening and predictive testing of female relatives [28, 54, 55]. PSMC3IP (HOP2) causes ODG3: in a consanguineous Palestinian family, homozygous p.(Glu201del) produced absent puberty, primary amenorrhoea, a hypoplastic uterus and streak gonads; in a Yemeni family, the variant p.(Tyr163Ter) caused POI in four sisters and azoospermia in the brothers; in a compound heterozygous Italian family, the presentation was secondary amenorrhoea at 29 years [56, 57, 107]. The SWS1 complex (SPIDR, ZSWIM7, SWSAP1) is involved in homologous recombination between homologues: SPIDR (ODG9) was described in sisters homozygous for p.(Trp280Ter), ZSWIM7 (ODG10) in families with primary amenorrhoea and absent or incomplete puberty, and in 2025 a French series added SWSAP1 (homozygous deletion with frameshift) and new ZSWIM7 variants in five patients with severe isolated POI, with partial or total loss of recombination activity in mouse embryonic stem cells [58–60]. Genes of the Fanconi pathway and BRCA1/BRCA2 may also cause POI and predispose to tumours, so testing for them should be carried out by a multidisciplinary team experienced in presymptomatic diagnosis [1]. For FANCM, the FinnGen cohort of 500,348 individuals identified 76 homozygotes for c.5101C>T, p.(Gln1701Ter), with a high risk of breast, head and neck, gastrointestinal, gynaecological, haematological, skin and lung cancers, a high recessive risk of ovarian dysfunction and of haematological toxicity after cancer treatment, but without bone marrow failure or physical features of Fanconi anaemia; the authors propose a distinct recessive cancer syndrome [63].

2.7. MCM8 and MCM9: POI, chromosomal instability and tumour susceptibility

MCM8 (20p12.3) and MCM9 (6q22.31) form a hexameric complex that takes part in replication, meiosis, homologous recombination (recruiting BRCA1 and RAD51) and mismatch repair (recruited by hMutS); biallelic variants are loss-of-function [20]. The first families, consanguineous, showed primary amenorrhoea, short stature and defective repair of chromosomal breaks in the lymphocytes and fibroblasts of affected individuals [84, 85], and a 46,XY brother with azoospermia and 3 mL testes in one of the MCM8 families [61, 108]. The 2026 GeneReviews brings together 54 individuals (26 MCM8, 28 MCM9): gonadal insufficiency in 20/22 women with MCM8 and 23/24 with MCM9, recognised between 10 and 30 years of age; germ cell tumours (endodermal sinus tumour, dysgerminoma) in 2/26 and 1/28, between 11 and 15 years; colorectal polyposis in 2/26 and 6/28; colorectal cancer in 0/26 and 6/28, at 30–60 years; and gastric cancer in one individual with MCM9 [20]. In the 100,000 Genomes Project, biallelic MCM9 variants were associated with colonic polyps (OR 6.51; 95% CI 1.24–34.11), rectal polyps (OR 8.40; 1.28–55.35) and gastric cancer (OR 27.03; 2.93–248.5), without equivalent enrichment for MCM8 or in the UK Biobank; the tumours had "clock-like" mutational processes, without a repair-deficiency signature [109]. In a series of 155 women with POI, 3 (2%) had possibly damaging heterozygous MCM8 variants [16, 110]. Heterozygotes have hypogonadism in 29% (14/49, MCM8) and 22% (10/45, MCM9) of described cases, without clear evidence of colorectal cancer risk, and modelling from gnomAD estimates that about 0.02% of the population has biallelic pathogenic variants in one of the two genes [20].

2.8. Molecular basis of the main syndromic forms

BPES is due to heterozygous variants of FOXL2 (3q22.3; NM_023067.4), a single-exon gene with a 14-alanine tract; 87% of variants are in the coding region, truncating variants account for about half, expansions of the tract from 14 to 15–24 alanines for about 24% of intragenic variants (and may escape targeted techniques), whole-gene deletions 8%, regulatory deletions 3% and translocations 2% [21]. The recurrent variant is NM_023067.4:c.672_701dup, p.(Ala225_Ala234dup) [21]. Classically, truncations upstream of the poly-Ala tract are associated with type I with POI and expansions with type II, but the 2026 GeneReviews concludes that there is no consistent genotype-phenotype correlation, that ovarian involvement is variable and age-dependent even with the same variant, and that BPES should be treated as a single entity [21, 111, 112]. The somatic variant NM_023067.4:c.402C>G, p.(Cys134Trp) is present in 90–95% of adult-type granulosa cell tumours and is not hereditary [21]. In classic galactosaemia (GALT, 9p13.3; NM_000155.4), the variant NM_000155.4:c.563A>G, p.(Gln188Arg) accounts for about 70% of alleles in Northern Europeans and homozygosity is associated with POI, whereas NM_000155.4:c.404C>T, p.(Ser135Leu), prevalent in African populations, does not cause POI when treated early [22, 113]. The ovarian mechanism attributed is accumulation of galactose metabolites with oocyte apoptosis, beginning very early: by AMH, most girls have a greatly reduced number of follicles at 2–6 months of age, although histology before 5 years suggests a substantial reserve of primordial follicles [1, 114]. The AIRE gene (21q22.3; NM_000383.4) regulates the expression of autoantigens in the thymus; the variants NM_000383.4:c.769C>T, p.(Arg257Ter) and NM_000383.4:c.967_979del, p.(Leu323SerfsTer51) (the 13 bp deletion, a founder variant in the United Kingdom in 17 of 24 alleles) are the best known, and transmission is recessive for most variants and dominant for some dominant negative missense variants in the PHD1, PHD2 or SAND domains, with milder, later-onset disease [1, 16, 70]. In autoimmune polyendocrinopathy type 1, hypogonadism is associated exclusively with anti-P450scc (CYP11A1) antibodies, with an OR of 12.5 in a series of 90 Nordic patients (Söderbergh 2004, according to OMIM) [70]. The ovarioleukodystrophies result from biallelic variants in the five genes of the initiation factor eIF2B (EIF2B5 in about 66.5% of VWM cases, EIF2B2 16.6%, EIF2B3 7.8%, EIF2B4 7.4%, EIF2B1 1.7%) or in AARS2 [16, 23]. Pseudohypoparathyroidism 1A results from loss-of-function variants in the maternal GNAS allele, with resistance to PTH, TSH and GnRH [1].

2.9. Standardised reference variants

Variants follow the Human Genome Variation Society (HGVS) nomenclature on versioned reference transcripts, preferably MANE Select (Matched Annotation from NCBI and EMBL-EBI), and were validated in VariantValidator [115].

Variants of NOBOX, STAG3 and MCM8/MCM9 cited only in protein notation or on the authors' transcript (for example, STAG3 c.2776C>T, p.(Arg926Ter), or MCM8 c.446C>G, p.(Pro149Arg)) are kept in their published form, because conversion to another transcript was not carried out in a validation tool [85, 105].

2.10. Genetic architecture, penetrance and oligogenism

More than 100 monogenic causes have been described, many with a single case or family, and the 2024 guideline recommends classifying variants by the criteria of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) (only classes 4 and 5 are diagnostic) and using curated virtual panels such as Genomics England PanelApp and PanelApp Australia [11]. Family studies point to a heterogeneous model: the risk of POI in female relatives of affected women in Finland was 4.6 times higher (95% CI 3.3–6.5) [26], 10–15% of patients have an affected first-degree relative and familial forms range from 4% to 31% across series [1]. Panel analyses showed that many patients have rare variants in more than one gene: 7 of 100 patients with mutations at two loci, associated with earlier onset [116], 34/64 with at least two variants on a 295-gene panel in POI with onset between 10 and 25 years [117] and 7.3% of 1030 Chinese patients with variants in several genes [14]. The UK Biobank penetrance study found, among 104,733 women, 2231 (1.14%) with natural menopause before 40 years, limited evidence for any described dominant effect, modest haploinsufficiency effects in TWNK (menopause 1.54 years earlier) and SOHLH2 (3.48 years earlier) and 99.9% (13,699/13,708) of truncating variants in reproductively healthy women [19]. Hence the ESHRE conclusion that oligogenism should not be used for diagnosis [11]. Candidate genes with heterozygous effects continue to emerge, such as HELB, in a Chinese family with POI and early menopause and in a mouse model with a shortened reproductive lifespan, whose clinical relevance has not yet been established [118]. The 2026 Utah data confirm a gradient: chromosomal abnormalities and recessive inheritance predominate in primary amenorrhoea, and heterozygous or multiple variants in DNA repair genes, added to a low polygenic risk for age at menopause, predominate in secondary amenorrhoea [119].

2.11. Allelic disorders and modifiers

Several POI genes have allelic phenotypes outside the ovary: FMR1 (fragile X syndrome with the full mutation; fragile X-associated tremor/ataxia syndrome, FXTAS, in premutation carriers, in 17% of men at 50–59 years and 75% at 80 years, and in 16–20% of women) [11, 90]; NR5A1 (46,XY DSD, complete gonadal dysgenesis, male infertility, 46,XX testicular or ovotesticular DSD due to a recurrent variant) [16]; FOXL2 (granulosa cell tumours due to a somatic variant) [21]; FSHR (familial ovarian hyperstimulation due to activating variants) [16]; STAG3, HFM1, MSH4, SYCE1 and ZSWIM7 (male infertility due to meiotic arrest, with SPGF entries) [16, 106]; MCM9 (polyposis and colorectal and gastric cancer) [20]; ERCC6 (Cockayne syndrome B, recessive, in contrast with dominant POF11) [16]; NUP107 (nephrotic syndrome and Galloway-Mowat syndrome in addition to ODG6) [16]. The MCM8 gene also acts as a quantitative locus for age at natural menopause (MENOQ3) [61], and BMP15 is listed in Orphanet as a susceptibility factor for malignant ovarian teratoma [16].

3. Clinical Spectrum and Manifestations by System

3.1. Common manifestations of the class

In the short term, POI manifests as absent puberty (absence of breast development or arrest of pubertal progression), primary, primary-secondary or secondary amenorrhoea, oligomenorrhoea, hot flushes and night sweats, dyspareunia, reduced libido, sleep and mood disturbances and infertility; in the long term, as bone demineralisation and increased cardiovascular risk [1]. Amenorrhoea may become apparent on stopping a combined oestrogen-progestogen contraceptive that had masked it [1]. POI is not a static condition: in a French cohort of 357 consecutive patients with a normal karyotype, 76% had had normal puberty followed by secondary amenorrhoea, 14% had a family history, 14.3% clinical or biological autoimmunity, 56 women had a fluctuating form, ultrasound suggested follicles in 50% and histology confirmed them in 29%, and 58% had BMD abnormalities [83, 120]. In a Chinese cohort of 955 patients (2006–2015), 85.97% presented with secondary amenorrhoea and 14.03% with primary amenorrhoea; in the secondary forms, ovarian function declined rapidly, the reproductive period was shortened by about 10 years and amenorrhoea set in 1 to 2 years after menstrual irregularities; visible follicles (28.43%) correlated with more favourable fertility indicators [83, 121].

3.2. Natural history without treatment and with hormone therapy

Bone. POI is associated with abnormal bone microarchitecture, reduced BMD and an increased risk of osteoporosis and fracture; many patients have not reached peak bone mass (around the middle of the third decade), and hypo-oestrogenism and hypoandrogenism impair bone formation, so adolescents and young women have the highest risk [1, 2]. In a French cohort of 162 women followed for 7.9 ± 2.8 years, 42.6% interrupted HRT for at least one year, and this interruption was associated with significant loss of femoral BMD; at baseline, 30% had osteopenia and 8% osteoporosis [122]. A systematic review of 16 studies showed that bone mass tends to remain stable with oestrogen and progestogen, that usual doses do not reverse established loss and that higher oestrogen doses and continuous oral contraception appear more favourable [123].

Cardiovascular system. Without HRT, cardiovascular diseases are the main cause of reduced life expectancy; women with POI have endothelial and autonomic dysfunction, an abnormal lipid profile, insulin resistance and metabolic syndrome [1]. Mortality from ischaemic heart disease is 80% higher in women with menopause before 40 years than in those with menopause between 49 and 55 years, particularly in those who never received oestrogens, and menopause before 39 years multiplies by 3 to 5 the risk of death from myocardial infarction at 60 years [1]. Data on stroke are contradictory: one review showed increased risk with a protective role of oestrogen up to the average age of menopause, and one meta-analysis showed no increase [1]. The 2024 guideline considers the increased risk of coronary disease, heart failure and stroke to be established, and associates non-use of HRT with more cardiovascular events and mortality [2]. In the COLA study (mean oestrogen exposure of 19.3 years, mean age 45 years), the longer the period without oestrogens, the greater the cardiovascular risk [1].

Fertility. Infertility affects most patients and is generally permanent, but not always: in a French cohort of 507 patients, the cumulative incidence of resumption of ovarian activity was 23% and that of pregnancy 3% among those who resumed activity, and estimates of spontaneous conception lie between 4% and 10% (4–6% natural pregnancies according to the PNDS), without a reliable predictive marker [1]. Up to 50% of women may have intermittent return of ovarian function, but only 5–10% conceive spontaneously [1]. Adolescents with early POI may have reduced uterine volume; an Italian study showed that, after 5 years of HRT, the uterus does not reach the development of controls, although maturation with prolonged oestrogen is possible [1].

Sexuality, mental health and cognition. Women with POI report less well-being and sexual satisfaction, less frequent intercourse, lower libido and more dyspareunia, the more so the younger the age at diagnosis; a meta-analysis of nine studies (529 spontaneous POI versus 319 controls) showed significantly lower total testosterone [1]. The diagnosis is experienced abruptly, with anger, depression, anxiety and sadness; in an American study of 171 idiopathic POI and 100 women with Turner syndrome, the likelihood of depressive syndrome was high in both groups and greater in the idiopathic group (p < 0.001) [1]. The risk of cognitive impairment and dementia is increased, particularly after bilateral oophorectomy; in genetic POI, the cognitive deficits observed are more likely to be of genetic than hormonal origin [1, 2]. The 2024 guideline adds reduced muscle mass, strength and performance, with a risk of sarcopenia [2].

Autoimmunity, cancer and life expectancy. More than 20% of idiopathic POI cases have an associated autoimmune disease, mainly thyroid disease, but also type 1 diabetes, adrenal insufficiency, rheumatoid arthritis, lupus and, rarely, myasthenia; the presence of anti-21-hydroxylase antibodies implies a 50% risk of potentially fatal adrenal insufficiency [1]. Cancer risk is increased mainly because some POI cases follow chemotherapy, and DNA repair defect aetiologies carry their own risk; HRT in spontaneous POI does not increase breast cancer [1]. In a prospective cohort of 20,555 women, menopause before 40 years was associated with increased all-cause mortality and a life expectancy 2 years shorter than menopause between 50 and 54 years, and mortality was higher after prophylactic bilateral oophorectomy before 45 years [1].

3.3. X chromosome abnormalities

In Turner syndrome, menopause typically precedes menarche; 10–20% of girls have spontaneous puberty and 2–5% spontaneous menstruation, usually transient [12, 36]. In the Danish series of 410 women, 31 had at least one spontaneous pregnancy, and only those with 45,X/46,XX or 46,XX with a structural X abnormality had spontaneous live births; pregnancy is possible when the Xq13–q26 region is spared or a 46,XX cell line is present [87]. Women with structural Xq rearrangements or X;autosome translocations with critical breakpoints present with gonadal dysgenesis, primary amenorrhoea or POI, and men carrying the same translocations are infertile [38]. In trisomy X, POI is more frequent than in controls [39].

3.4. POI associated with the FMR1 premutation

About 20% of premutation carriers have menopause before 40 years, versus 1% of the population (21% in a 2005 series), and the premutation explains about 3.2% of isolated POI and 11% of familial POI [15, 92]. POI may begin very early, as early as 11 years, and present with primary amenorrhoea and delayed puberty, but most present with secondary amenorrhoea [15]. Ovarian function is erratic, with vasomotor symptoms, mood changes and vaginal dryness, and up to 12.6% of women conceive after diagnosis [15]. In an American series of 79 women with FXPOI who were interviewed, the median age at onset was 33 years, 72% had 80 to 100 repeats, the mean interval between symptoms and diagnosis was 1.12 years (longer in the youngest, shorter in those who knew they were carriers), 52% never took HRT, started it years later or stopped it before 45 years, 49% had infertility but 75% had at least one biological child, obstetric outcomes were similar to those of the population, 46% had low BMD or osteoporosis and 19% had never had densitometry [124]. Carriers also have an increased risk of osteoporosis, cardiovascular disease, thyroid disease, anxiety and, later, FXTAS (16–20% of women) [11, 15].

3.5. Isolated monogenic forms

In ODG1 due to FSHR, Finnish patients have primary amenorrhoea with incomplete pubertal development, elevated plasma FSH, very low oestrogen and inhibin B and follicles present, without response to high doses of recombinant FSH [12, 16]; in the follow-up of 26 Finnish women homozygous for p.(Ala189Val), HRT was started late (median 18 years) and 9 of 14 had osteopenia despite it, without an excess of cardiovascular risk factors [16]. In the biallelic forms of BMP15 and NOBOX, the dominant picture is delayed puberty with primary amenorrhoea and streak gonads: the Belgian proband with NOBOX was diagnosed at 17 years without pubertal development, with FSH of 106 IU/L, oestradiol below 12 ng/L, a small uterus and streak ovaries at laparoscopy [100]; Di Pasquale's proband had primary amenorrhoea at 23 years, moderate hirsutism, delayed puberty at 15 and streak ovaries with an underdeveloped uterus at 17 [41]. In the MCM8 and MCM9 syndromes, women present with absent, delayed or incomplete puberty (13/26 and 6/28), primary amenorrhoea, or secondary amenorrhoea with oligomenorrhoea after menarche, an underdeveloped uterus and small or streak ovaries; short stature occurs in 2/26 and 6/28, hypothyroidism in 3/26 with MCM8, and men have azoospermia or severe oligospermia with a Sertoli-cell-only pattern, although the presence of sperm does not exclude the diagnosis [20]. The original STAG3 family illustrates the tumour risk of dysgenetic gonads due to meiotic defects [50]. In the forms due to MSH4 and MSH5, the presentation is mainly later-onset secondary amenorrhoea, sometimes only diminished reserve [28]. In NR5A1, the same variant may cause POI in a woman and 46,XY DSD in a relative, and dysgenetic gonads carry a risk of gonadoblastoma; adrenal insufficiency is rare [16, 48]. In exome studies, primary amenorrhoea has a higher genetic yield than secondary amenorrhoea (28.5% versus 17.8% in 1030 Chinese patients) [14].

3.6. BPES type I

BPES is characterised by blepharophimosis, ptosis, epicanthus inversus and telecanthus, with complete ocular penetrance; ovarian involvement is age-dependent and has reduced penetrance [21]. The ovarian presentation ranges from primary amenorrhoea (a minority) to oligomenorrhoea, secondary amenorrhoea of more than 4–6 months before 40 years and infertility; secondary sexual characteristics and menarche are usually normal, and a hypoplastic uterus, small ovaries, a low antral follicle count (2–10 mm), FSH above 25 IU/L on two occasions, elevated luteinising hormone (LH), oestradiol below 50 pg/mL and, in some patients, hypopituitarism may be found [21]. Many women are initially fertile and later progress to ovarian resistance to gonadotrophins or to true POI; there is at least one case of ovarian failure in early childhood, and the family of Townes and Muechler (1979) had primary ovarian failure in all affected women [66, 112]. Endometrial carcinoma and occasional granulosa cell tumours have been described [112].

3.7. Classic galactosaemia

POI with primary or secondary amenorrhoea and hypergonadotrophic hypogonadism occurs in 75–96% of women with classic galactosaemia by 30 years, unrelated to age at diagnosis or dietary control, and is not prevented by effective treatment [68]. About two-thirds of girls have spontaneous menarche, but most stop having regular periods within the following 5 years; male gonadal function is essentially preserved [114]. In a retrospective survey from 1990 (270 individuals), 81% of 47 girls and women had signs of POI, mean menarche was at 14 years (10–18), 8 of 34 aged over 17 years had primary amenorrhoea (two with streak gonads) and only 5 of 17 aged over 22 years had normal periods [22, 125]. The risk factors are homozygosity for p.(Gln188Arg), mean erythrocyte galactose-1-phosphate above 3.5 mg/dL on treatment and 13CO2 recovery below 5% after 13C-galactose [12, 22, 126]. A minority have become pregnant, including women without residual GALT activity (more than 40% of a small group who tried); metabolites do not rise in pregnancy or the postpartum period and the children are normal [114]. The Duarte variant and S135L homozygosity do not cause POI [22, 114].

3.8. Autoimmune polyendocrinopathy type 1 (APECED)

The classic criterion requires two of three manifestations: chronic mucocutaneous candidiasis, hypoparathyroidism and adrenal insufficiency [70]. In the Finnish series of 68 patients, oral candidiasis was the first manifestation in 60%, hypoparathyroidism occurred in 79%, adrenocortical insufficiency in 72% and gonadal failure in 60% of women aged 13 years or older and in 14% of men aged 16 years or older; the components accumulate over decades (3 to 5 per patient) [16, 127]. The PNDS gives a POI prevalence of about 40% at 15 years and 72% at 40 years in polyendocrinopathy type 1, versus about 10% at 40 years in polyendocrinopathy type 2, which is polygenic and linked to HLA-DR and HLA-DQ alleles of the human leucocyte antigen system [1]. In an American prospective cohort of 35 patients, urticaria, hepatitis, gastritis, intestinal dysfunction, pneumonitis and Sjögren-like syndrome occurred in 40–80%, 80% had a median of three manifestations outside the triad before the classic dyad, and the dyad appeared at a mean of 7.38 years [16]. In Finland, mortality among 91 patients (1971–2018; 29 deaths) had a standardised ratio of 11 (95% CI 7.2–16), with adrenal crisis among the causes of death [16]. In a sister with POI at 33 years and isolated hypoparathyroidism, compound heterozygosity for the 13 bp deletion and a splicing variant was found, which shows that POI may precede the complete picture [70].

3.9. Ovarioleukodystrophies

In leukoencephalopathy with vanishing white matter (VWM), the common form begins between 2 and 6 years of age with progressive ataxia and spasticity, with acute deteriorations after fever, minor head trauma or fright; in the juvenile and adult forms, women may have primary or secondary ovarian failure, sometimes as the first manifestation, before behavioural problems and cognitive decline [23, 128]. Ovarian failure may occur in any form and has been found at autopsy in infantile cases; the association of ovarian failure with eIF2B variants was described in 2003 [23, 129]. The incidence in the Netherlands is 1/80,000 live births [16, 130]. AARS2-related leukodystrophy combines progressive leukodystrophy and ovarian failure [131].

3.10. Other syndromic forms

In Perrault syndrome, ovarian dysfunction ranges from dysgenesis with primary amenorrhoea to POI, associated with bilateral sensorineural deafness, with more than 170 individuals described and at least 12 genes that explain about 80% of cases [24]. In Fanconi anaemia, POI affects up to 77% of women and testicular failure up to 64% of men [20]. Bloom, Werner, ataxia-telangiectasia, Nijmegen and Rothmund-Thomson syndromes appear in the list of syndromes with POI in the 2024 guideline, as do Hutchinson-Gilford progeria, GAPO syndrome, acromesomelic chondrodysplasia due to BMPR1B, proximal symphalangism due to NOG, Woodhouse-Sakati syndrome and 46,XX DSD due to WT1 [11]. Mitochondrial diseases (dominant families with multiple mitochondrial DNA deletions due to POLG, with parkinsonism and POI; progressive external ophthalmoplegia) include ovarian failure among their manifestations [74, 132, 133]. Mild forms of D-bifunctional protein deficiency (HSD17B4) combine juvenile cerebellar ataxia, deafness, peripheral neuropathy and ovarian failure, in a picture overlapping with Perrault syndrome [134]. PMM2-CDG may cause absent pubertal development and hypergonadotrophic ovarian failure [75]; in Mulibrey nanism puberty is normal (median menarche at 12.5 years) but premature ovarian failure follows [76]; in microcephalic primordial dwarfism type II, precocious puberty and ovarian failure are described [135]; in myotonic dystrophy type 1, gonadal atrophy is accompanied by early menopause [77]. In the Finnish registry, 15.9% of women with POI had at least one diagnosis of a genetic disease or congenital malformation, with an OR of 16.5 (95% CI 6.2–43.7) for autosomal monogenic diseases and 24.1 (15.1–38.2) in the group diagnosed between 10 and 14 years [86].

4. Clinical Recognition and Warning Signs

4.1. When to suspect

POI should be considered and excluded in any woman under 40 years with amenorrhoea, irregular cycles or symptoms of oestrogen deficiency (good practice point of the 2024 guideline; the PNDS adds oligomenorrhoea) [1, 2]. Signs of a major decrease in reserve include shortening or lengthening of cycles, menstrual irregularities, abnormal uterine bleeding, vasomotor symptoms, mood changes and, later, atrophic vaginitis, but no longitudinal study has shown that diminished reserve predicts POI and there are no specific signs that herald it [1]. AMH should not be used to predict POI [2]. Amenorrhoea that appears on stopping the pill should be investigated, because hormonal contraception masks POI and lowers FSH [1, 2].

4.2. By age and context

In childhood and adolescence, the presentations that raise suspicion are absent puberty at 12–13 years, arrest of pubertal progression and primary amenorrhoea; gonadal dysgenesis accounts for 48.5% of primary amenorrhoea in a classic series, ahead of Müllerian agenesis (16.2%), gonadotrophin-releasing hormone (GnRH) deficiency (8.3%) and constitutional delay (6.0%) (Timmreck and Reindollar 2003, according to the Oxford Desk Reference) [136]. In galactosaemia, screening is targeted and begins at 12 years if secondary sexual characteristics are insufficient or at 14 years in the absence of regular periods [31]. In Turner syndrome, the 2024 guideline recommends measuring LH, FSH and AMH at 8–9 years and then annually until 11–12 years, to allow timely referral for fertility preservation [32]. In the adult woman, the typical presentation is secondary amenorrhoea or infertility, often with years of delay: in a survey of 220 women, 53% did not know the cause of their POI, 30% had been diagnosed within the first 6 months and 22% waited up to 3 years [1].

4.3. Genetic clues from the history and examination

The history should include a three-generation pedigree with the age at menopause of female relatives, fertility and associated diseases (developmental anomalies, neurological signs, intellectual disability, sensory, cardiovascular, endocrine or metabolic symptoms, tumours and cancers) [1, 11, 12]. The specific clues are: intellectual disability or fragile X in boys in the family, late-onset tremor and ataxia in grandfathers or uncles (FMR1 premutation) [1, 137]; personal or family congenital eyelid malformation or eyelid surgery, particularly with transmission through men (BPES) [1, 112]; brothers with azoospermia or severe oligospermia (meiosis and repair genes) [11]; consanguinity and affected sisters (recessive forms) [20]; sensorineural deafness (Perrault) [24]; autoimmune diseases, oral, nail or genital candidiasis and hypoparathyroidism (APECED) [1]; short stature, short or webbed neck, low hairline, cubitus valgus, shield chest or a heart murmur (Turner) [12]; germ cell tumours before 20 years, 20 or more colorectal polyps, or colorectal or gastric cancer before 50 years (MCM8/MCM9 syndromes) [20]; ataxia, spasticity or cognitive decline (ovarioleukodystrophy) [23]. Ovarian or pelvic surgery, chemotherapy and pelvic radiotherapy should also be recorded, as they allow iatrogenic causes to be excluded before genetic investigation [1].

4.4. Warning and emergency signs

The life-threatening situation is adrenal insufficiency: women with POI and anti-21-hydroxylase antibodies have a 50% risk of developing adrenal insufficiency that is potentially fatal if undiagnosed, and in half of cases POI precedes autoimmune adrenal disease, sometimes by years [1]. Fatigue, weight loss, hypotension, hyperpigmentation, hyponatraemia or hyperkalaemia in a woman with POI require urgent adrenal assessment; in autoimmune polyendocrinopathy type 1, adrenal crisis is among the causes of death [16]. In 46,XX girls with NR5A1 variants, adrenal insufficiency may present as a salt-wasting crisis in the first weeks of life [47]. An adnexal mass in a girl with gonadal dysgenesis, with Y material in the karyotype or with biallelic MCM8/MCM9 or STAG3 variants should raise suspicion of a germ cell tumour [11, 20, 50]. In the pregnancy of a woman with Turner syndrome, chest pain should be treated as possible aortic dissection (about 2% risk of death from rupture or dissection in an American survey of 101 pregnancies by oocyte donation) [87]. In ovarioleukodystrophy, fever, head trauma and fright may precipitate acute neurological deterioration [128].

4.5. Newborn screening

POI has no newborn screening of its own. Classic galactosaemia is detected by newborn screening in most European countries and in the United States (1/48,000 births by screening; 1/16,476 in Ireland), but neonatal diagnosis and early treatment do not prevent POI [16, 68]. FSH and LH measurements during minipuberty (2 weeks to 3 months) may give an early indication of ovarian dysfunction in galactosaemia [22].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Diagnostic criteria and divergences

The 2024 change allowing diagnosis with a single FSH above 25 IU/L is the main change compared with the 2016 guideline; the 2021 PNDS and NICE, whose section 1.7 retains the 2015 text, still require two measurements [1–3]. In practice, the PNDS recognises that an FSH above 25 IU/L in a young woman, outside the preovulatory peak, indicates POI and that confirmation is useful in borderline forms [1].

5.2. Biomarkers and their pitfalls

In the woman or adolescent with amenorrhoea, the initial measurements include hCG (to exclude pregnancy), FSH, oestradiol and prolactin; together with TSH, they allow the most frequent causes of cycle disturbance to be excluded [1, 15]. The FSH threshold was lowered because, in autoimmune POI, FSH is often lower than in idiopathic POI (mean 37 IU/L, range 26–64, frequently between 25 and 35 IU/L) [1]. FSH measured on a combined oestrogen-progestogen or on an antigonadotrophic high-dose progestogen is underestimated, sometimes markedly, and the PNDS recommends stopping treatment and repeating the measurement at least 3 months later [1]. A low oestradiol excludes sampling at the preovulatory peak (150–200 pg/mL per mature follicle) or in the presence of a secreting cyst [1]. AMH reflects the follicular pool, as it is secreted by the granulosa cells, and is collapsed in most POI, but may be normal in autoimmune POI, in which the attack begins in the theca; it is therefore not part of the definition, although a normal AMH points towards an autoimmune aetiology [1]. Pelvic ultrasound is not a diagnostic criterion: about 50% of patients have follicles larger than 2 mm, sometimes preovulatory follicles or corpora lutea, which may be falsely reassuring; many small persistent follicles favour an autoimmune aetiology, uterine size indicates the degree of oestrogen exposure and the transvaginal route is preferred, except in a virgin patient [1]. In the cohort of 357 patients, the negative predictive value of the presence of follicles on ultrasound was 77% [83].

There is no indication for infectious screening [1]. Anti-ovarian antibodies are neither sensitive nor specific and the 2024 guideline recommends against their use (strong) [1, 2].

5.3. Genetic testing strategy

The PNDS sets out the sequence: karyotype as first-line test in all non-iatrogenic POI, after informed consent; FMR1 premutation testing as second line, with discussion of the implications before testing and referral to clinical genetics if positive; next-generation sequencing (NGS) gene panel if the karyotype and FMR1 are normal (in some laboratories simultaneously), in centres able to provide counselling and family studies; and, since 2019, whole-genome sequencing in the Plan France Médecine Génomique 2025 for patients on a "diagnostic odyssey" with suspected genetic disease and a negative panel, after a national multidisciplinary meeting [1, 139]. The 2024 guideline recommends chromosomal analysis and FMR1 premutation testing in all non-iatrogenic POI (both strong, low certainty), and considers that additional genetic testing, for example by NGS, may be offered to all non-iatrogenic POI where available and after counselling (conditional, low certainty); age should not restrict access to genetic testing [2]. Ideally, NGS and FMR1 testing are performed in parallel on a single sample, because the yield of NGS (up to 26% in European sporadic forms and 36.7% in familial forms) is higher than that of FMR1 (1–13%) [11]. A 2021 French position statement recommends a panel or genome in all POI without a diagnosis [11, 140].

The G-banded karyotype is the gold standard for numerical abnormalities, mosaicism, Y material and large structural abnormalities; the Oxford Desk Reference recommends a 30-cell count to screen for mosaicism [11, 12]. Microarray detects small copy number variants (CNV) and refines breakpoints, but does not detect low-level mosaicism or balanced rearrangements; PCR and fluorescence in situ hybridisation (FISH) detect mosaicism; age-related X loss in lymphocytes requires cautious interpretation of low-level mosaicism in older women, with verification in another tissue, such as skin [11]. The FMR1 premutation is sized by PCR, including PCR with primers within the repeat (triplet-primed PCR), and Southern blot confirms size and methylation and allows X inactivation to be estimated; multigene panels, exome and short-read genome sequencing do not detect the premutation, and sequencing and PCR do not detect X deletions in heterozygotes [11, 15, 141]. In the case of FOXL2, alanine tract expansions may escape targeted sequencing, and deletions and regulatory rearrangements require CNV analysis, genome sequencing or optical mapping [21].

5.4. Diagnostic yield

These figures are not comparable with one another: they depend on the proportion of primary amenorrhoea, consanguinity, the number of genes, the classification criterion (an isolated P/LP variant or a diagnosis consistent with the gene's inheritance) and the inclusion of candidate genes. The difference between 19.3% of patients with a P/LP variant and 7.5% of clinically relevant diagnoses in the Belgian study shows that the proportion of variants detected does not equate to the clinical sensitivity of the test [144].

5.5. Limitations of the tests

A negative result does not exclude a genetic cause. The karyotype does not detect microdeletions or very low-level mosaicism; microarray does not detect balanced rearrangements or low-level mosaicism; NGS does not detect the FMR1 premutation, misses many CNVs (CNV analysis is not routine, with very low positivity in the cohorts) and repeat expansions such as those of FOXL2, and does not reliably interpret deep intronic variants [11, 21, 104]. Virtual panels and exome sequencing are not routinely available in most countries [11]. The PNDS stresses that functional studies, indispensable to establish a causal relationship for a missense variant, are rarely performed and only in research laboratories [1]. Heterozygotes for recessive genes and heterozygous missense variants of BMP15, GDF9 and NOBOX should not be reported as diagnostic [17–19, 43].

5.6. Report content, VUS and segregation

The report should state the gene, the versioned transcript (preferably MANE Select), the variant in HGVS nomenclature, zygosity, the ACMG/AMP class and consistency with the gene's mode of inheritance; for FMR1, the number of repeats of each allele, the error range and, when assessed, the AGG interruption pattern and methylation [11, 15]. Biallelic variants of uncertain significance (VUS) in MCM8/MCM9 neither establish nor exclude the diagnosis [20]. Segregation analysis in parents and sisters helps to confirm the phase of variants in recessive genes and to clarify low-penetrance variants, and interpretation should be carried out with geneticists; variant carriers may not develop POI or may develop it later, which requires long-term follow-up [11]. Consent should be written and counselling should follow national standards [11].

5.7. Practical algorithm

5.8. Prenatal and preimplantation diagnosis

Once the familial variant has been identified, prenatal diagnosis by chorionic villus sampling or amniocentesis and preimplantation genetic testing (PGT-M) are possible [20, 92]. In FMR1 premutation carriers, PGT-M allows only non-carrier embryos to be transferred, but it is often difficult because of the limited number of oocytes obtained with stimulation [1, 92]. In fertile Turner syndrome, prenatal diagnosis should be considered because of the increased risk of trisomy 21 and 45,X in the offspring (Tarani 1998, according to the Oxford Desk Reference) [87].

6. Differential Diagnosis

6.1. Comparative table

6.2. Main differential diagnoses

The first step separates hypergonadotrophic hypogonadism (elevated FSH) from conditions with normal or low FSH, which include pregnancy, functional hypothalamic amenorrhoea, hyperprolactinaemia, thyroid dysfunction, congenital hypogonadotrophic hypogonadism and polycystic ovary syndrome; hormonal contraception should be stopped before this distinction is made, because it lowers FSH [1, 2]. In primary amenorrhoea with elevated FSH, the karyotype separates 46,XX POI from Turner syndrome and 46,XY DSD, in which the presence of a uterus with streak gonads is common and the risk of gonadoblastoma requires consideration of gonadectomy [11, 136]. In primary amenorrhoea with normal FSH and normal secondary sexual characteristics, pelvic imaging identifies Müllerian agenesis [136]. LH resistance due to LHCGR and ODG1 due to FSHR mimic POI with follicles present; in the former, LH is disproportionately elevated [12, 16].

Once POI is confirmed, aetiological investigation distinguishes iatrogenic causes, which are clarified by the history, from autoimmune and genetic causes [1]. Autoimmune POI is associated with other autoimmune conditions, normal AMH and many small follicles on ultrasound; anti-21-hydroxylase antibodies are the best available marker and, when positive, identify women with a 50% risk of adrenal insufficiency [1]. Among genetic causes, the karyotype and FMR1 testing identify the most frequent, and the others are recognised by clinical clues: eyelids (BPES), hearing (Perrault), nervous system (ovarioleukodystrophy, mitochondrial diseases), autoimmunity with candidiasis and hypoparathyroidism (APECED), short stature with tumours or polyps (MCM8/MCM9), chromosomal fragility (Fanconi and related syndromes) and a history of galactosaemia [11, 20]. In polyposis syndromes, the absence of gonadal dysfunction argues against MCM9 [20]. Haemochromatosis, which may cause hypogonadism and early menopause, is distinguished by iron studies [145].

7. Treatments and Risk Management Strategies

7.1. Principles and objectives

There is no treatment that has been shown to restore ovarian activity or increase natural conception in POI (strong recommendation of the 2024 guideline, moderate certainty) [2]. The mainstay of treatment is HRT, whose objectives, according to the PNDS, are to develop and maintain breast volume, favour the pubertal growth spurt in prepubertal adolescents, protect bone mass, reduce cardiovascular risk and improve sexual function and quality of life (climacteric syndrome, vaginal dryness, dyspareunia) [1]. The 2024 guideline recommends HRT until the usual age of menopause as primary prevention, with or without symptoms (strong), and, on reaching that age, an individual reassessment according to the rules for usual menopause [2, 11]. The PNDS sets that age at about 51 years [1]. Patients should be clearly informed that there is no evidence of increased breast cancer with HRT at a young age compared with women of the same age without POI, because 56–79% of patients believe the contrary [1, 2]. The diagnosis should be communicated with compassion, with personalised information, time for questions, shared decision-making, continuity of care and access to support groups and mental health services [2].

7.2. Puberty induction

The PNDS recommends inducing puberty with 17β-oestradiol, preferably transdermal, at a low dose from the physiological age of pubertal onset (11–13 years), without delaying beyond 12 years, with progressive increases over 2–3 years; the starting dose in a girl without breast development (stage S1) is about 8–10% of the adult dose, and some teams use 5 µg/kg/day with increases every 6 months; the growth curve and bone age should be assessed beforehand, and low doses should be maintained during the growth spurt; a cyclical progestogen is added after at least 2 years of oestradiol or when bleeding occurs; the contraceptive pill is not indicated for induction [1]. In French practice, fractions of 25 µg transdermal patches are used, as no formulations exist for very small doses [1]. The 2024 guideline recommends low-dose oestradiol from 11 years with gradual increase over 2–3 years (strong, low certainty), a modified regimen in patients diagnosed late without growth concerns (conditional), preferably the transdermal route (conditional, inconclusive evidence), not using the combined oral contraceptive (COC) for induction (strong) and preferring oestradiol to conjugated equine oestrogens and ethinylestradiol [2, 11].

The lowest commercial patches release 25 or 50 µg/day and the equivalence of fractions (1/8, 1/6 or 1/4 of a patch at night or daily, or a whole patch 7–10 days per month) has not been established [11]. The ESHRE induction recommendations adapt those of the 2022 European consensus on puberty induction [11, 146]. Uterine growth was greater with transdermal oestradiol in a study of 12 girls with Turner syndrome, more patients reached a normal adult uterus with 4 mg than with 2 mg of oral oestradiol, and in a standardised transdermal protocol uterine volume remained reduced in 27 of 45 patients, with prior irradiation and the oestradiol dose at introduction of the progestogen as determinants [11]. The 2024 Turner syndrome guideline recommends starting low-dose oestrogen between 11 and 12 years if FSH is elevated on at least two measurements, increasing to the adult dose over 2–4 years, preferably using transdermal oestradiol (ethinylestradiol carries more risks but is better than no treatment), adding cyclical progesterone after the first bleed (usually after 18–24 months), preferably micronised progesterone 200 mg for 10–12 days per month, and aiming for oestradiol concentrations of 100–150 pg/mL [32]. In galactosaemia, the international guideline recommends deciding the age of onset according to FSH, growth and psychosocial maturity, with stepwise low-dose oestrogen followed by cyclical progesterone, and referral to paediatric endocrinology [31]. A non-randomised phase 3 trial of the National Institutes of Health (NIH; NCT06851754), recruiting since July 2025, is evaluating in 185 participants aged 11 to 19 years the effect of transdermal 17β-oestradiol with oral micronised progesterone on lumbar spine BMD at 2 years [147].

7.3. HRT in the adolescent with breast development and in the adult woman: doses

The PNDS recommends an oestrogen dose that reproduces the mean concentration of the cycle, about 100 pg/mL of oestradiol: 50–100 µg/day transdermally, 2 to 3 pumps of gel per day, or 2 to 4 mg/day of oral oestradiol, with progesterone or a progestogen for at least 12 days per month in a woman with a uterus [1]. It adds that patients with POI may need higher doses than women at the usual menopause and that about 100 µg of transdermal 17β-oestradiol achieves physiological levels [1]. The 2024 guideline suggests a daily dose of at least 2 mg of oral oestradiol or 100 µg of transdermal oestradiol, or equivalent, to optimise BMD (conditional, very low certainty), because low postmenopausal doses are insufficient; the target is a serum oestradiol of about 200–400 pmol/L, achievable with 50–100 µg patches, 2–4 doses of gel (or 2–3 mg) or 2–3 sprays; oral oestradiol 2–4 mg gives similar levels, but with supraphysiological oestrone; in women with few symptoms, treatment may start at lower doses and be increased according to tolerance, oestradiol levels and BMD [2, 11].

Vaginal progesterone is used only when progestogenic effects preclude the oral route, 4% vaginal gel for 10 days with 1 mg of oestradiol is insufficient, compounded creams are insufficient and endometrial safety is less assured with micronised progesterone for more than 5 years [11]. With higher oestrogen doses, the progestogen should be increased, and unscheduled bleeding should be evaluated [2]. Absence of bleeding after the progesterone phase requires a pregnancy test, because HRT does not prevent sporadic ovulation [1].

Comparison with labels. Transdermal oestradiol (Vivelle-Dot®, USA) has a United States Food and Drug Administration (FDA) indication for "hypoestrogenism due to hypogonadism, castration or primary ovarian failure", in patches of 0.025, 0.0375, 0.05, 0.075 and 0.1 mg/day applied twice weekly to the lower abdomen or buttocks (not the breasts), starting at 0.0375 mg/day for vasomotor symptoms or 0.025 mg/day for osteoporosis prevention and adjusted according to response; the label includes a boxed warning on endometrial cancer with unopposed oestrogen in a woman with a uterus, stroke and deep vein thrombosis (data from the Women's Health Initiative, WHI, trial in women aged 50 to 79 years), probable dementia and breast cancer, and contraindicates use in undiagnosed genital bleeding, current or previous breast cancer, oestrogen-dependent neoplasia, active or previous venous or arterial thromboembolism, liver disease and thrombophilias [149]. Generic oral oestradiol (equivalent to Estrace®) has the same indication, with a usual starting dose of 1–2 mg/day adjusted to symptoms [150]. Oral micronised progesterone (Prometrium®) is authorised for endometrial protection at 200 mg at bedtime for 12 days per 28-day cycle, and for secondary amenorrhoea at 400 mg for 10 days [151]. Divigel® oestradiol gel (0.25–1.25 g/day) only has a menopausal vasomotor indication [152]. In Portugal, the Summaries of Product Characteristics (SmPC) of Dermestril® (oestradiol, 25, 50 and 100 µg/24 h patches applied twice weekly; 25 µg ineffective for osteoporosis prevention), Lenzetto® (oestradiol 1.53 mg per spray, 1 to 3 sprays per day, maximum 4.59 mg/day), Utrogestan® (micronised progesterone 100 mg; 200 mg/day for 12–14 days per month in HRT; vaginal route for replacement in women without ovarian function in oocyte donation programmes) and Duphaston® (dydrogesterone 10 mg; 10–20 mg/day in the last 12–14 days of the cycle) carry postmenopausal HRT indications; Progeffik®, Cyclogest®, Femoston® 1/5 and 2/10 and Activelle® are also authorised [148, 153–155]. Use in POI before 40 years and for puberty induction therefore lies outside the literal terms of the Portuguese SmPCs, although supported by the PNDS and the international guideline, whereas the American label for oral and transdermal oestradiol includes primary ovarian failure. The reference doses are those of the guidelines; the 3–4 sprays of the 2024 guideline exceed the maximum of 3 sprays in the Lenzetto® SmPC, and doses of 100–200 µg/day transdermally exceed the commercial maximum of 0.1 mg/day per patch [11, 149, 154]. In Brazil, open data from the National Health Surveillance Agency (ANVISA) list with active registration, among others, Estradot®, Systen®, Oestrogel®, Sandrena®, Lenzetto®, Natifa®, Utrogestan®, Duphaston® and Femoston® [156].

7.4. Routes, regimens and contraception

The route and regimen are chosen with the patient; transdermal oestradiol avoids the hepatic first-pass effect on clotting factors and does not increase venous thromboembolic risk, does not alter insulin-like growth factor 1 (IGF-1) and gives more physiological levels [1]. The PNDS describes as options a non-contraceptive tablet of oestradiol with progesterone, separate oestrogen and progesterone tablets, oestrogen gel with progesterone tablets, the COC, and oestrogen gel with a progestogen-releasing intrauterine device; the discontinuous regimen produces withdrawal bleeds, and the continuous regimen avoids them [1]. The 2024 guideline points out that HRT is not contraceptive, that in women with intermittent function who wish to conceive naturally sequential HRT is preferred (without impact on conception), and that compounded "bioidentical" preparations are not recommended [2].

Contraception is an area of divergence. The PNDS considers the COC, outside the usual contraindications, the first choice when contraception is needed, as it provides replacement and contraception at the same time and is more acceptable to younger women, while acknowledging the symptoms in the pill-free week, the supraphysiological steroid doses, the greater thromboembolic risk of the oral route and a single small English study with a less favourable bone effect than HRT; the progestogen-only pill is unsuitable because it does not replace oestrogen [1]. The 2024 guideline accepts the COC if contraception is needed or if it improves adherence, but in a continuous or extended regimen (strong, low certainty) [2, 11]. NICE offers a choice between HRT and combined hormonal contraception and explains that HRT may have a favourable effect on blood pressure compared with the COC [3]. The FMR1 GeneReviews and the galactosaemia guideline warn that combined contraception may fail in women with very high FSH and indicate the intrauterine device or implant as methods with a lower failure rate [15, 31].

Particular situations, according to the PNDS and the 2024 guideline: migraine does not contraindicate HRT (adjust dose, route or regimen if it worsens; transdermal oestrogen in migraine with aura); hypertension is not a contraindication (transdermal oestradiol); a history of venous thromboembolism or thrombophilia requires assessment by a haemostasis specialist and, if possible, the transdermal route, with discussion at a multidisciplinary meeting; obesity favours the transdermal route; fibroids are not a contraindication; in endometriosis with bilateral oophorectomy, HRT treats vasomotor symptoms but may reactivate the disease, so combined HRT is recommended even after hysterectomy [1, 2]. In general, HRT is not recommended after breast cancer (strong), and in BRCA1/BRCA2 carriers without breast cancer it is an option after risk-reducing salpingo-oophorectomy (strong) [2].

7.5. Androgens, non-hormonal and complementary therapies

The PNDS does not recommend androgens, because benefit has only been shown with total absence of libido, mainly after bilateral oophorectomy, and because in France there is no molecule with a dosage and route suitable for women [1]. The 2024 guideline diverges: transdermal testosterone at doses close to physiological premenopausal levels may be considered for hypoactive sexual desire disorder (conditional, low certainty), should be considered in iatrogenic POI with this disorder after exclusion of other causes (strong), has short-term safety without long-term data and should not be offered for muscle benefit (strong) [2]. Dehydroepiandrosterone (DHEA) did not improve ovarian function over 12 months, and acupuncture has no randomised trial showing efficacy or added benefit over HRT [1, 2]. Non-hormonal therapies effective in the usual menopause may be considered when HRT is contraindicated (conditional), but complementary therapies and supplements do not replace HRT (strong) [2]. NICE recommends advice on bone, cardiovascular health and symptoms for women with contraindications to hormonal treatments [3].

7.6. Bone, muscle, cardiovascular system, cognition and sexuality

HRT maintains BMD and, in a randomised trial, improved lumbar and femoral neck BMD after 3 years to values equal to those of controls with regular cycles; it appears to reduce fracture risk when it lasts at least 3 years [1]. Late initiation and non-adherence should be avoided (strong), weight-bearing exercise and a healthy lifestyle added, and calcium and vitamin D supplemented if intake is insufficient or BMD is low (conditional) [2]. Bisphosphonates and selective oestrogen receptor modulators such as raloxifene have no data in POI and should only be considered with an osteoporosis specialist, with caution if pregnancy is desired [1, 2]. POI is associated with reduced muscle mass and strength; resistance exercise is recommended (conditional) and screening for sarcopenia at diagnosis may be considered [2]. From the cardiovascular perspective, HRT is recommended until the usual age of menopause (strong), with at least annual measurement of blood pressure, weight and smoking status, and a lipid profile and diabetes screening at diagnosis; all women with Turner syndrome should be assessed by a cardiologist experienced in congenital heart disease (strong) [2]. HRT until the usual age of menopause is also recommended to reduce the risk of cognitive impairment and dementia (strong, low certainty), even without symptoms, and the risk should be discussed before bilateral oophorectomy before 45 years [2]. Regarding sexuality, systemic HRT improves sexual function only slightly (strong), vaginal oestrogen treats genitourinary symptoms and may be added if they persist (conditional), lubricants and moisturisers are an option, and laser and thermal energy therapies are not recommended [1, 2].

7.7. HRT monitoring

The PNDS recommends an annual clinical examination after starting HRT, with gynaecological examination, breast palpation, weight and blood pressure, and therapeutic education aimed at patients and professionals [1]. The 2024 guideline recommends titrating oestrogen according to symptoms and BMD; serum oestradiol is useful if relief is inadequate, bone protection insufficient or adverse effects occur (assays do not measure ethinylestradiol or oestrone); FSH has no value in monitoring; mammography follows the national programme from 45–50 years, except with risk factors such as BRCA1/BRCA2 or chest irradiation [1, 11]. Adherence is a central problem: after surgical menopause, cumulative use at 36 months was 79.1% (18–29 years), 75.9% (30–34) and 70.2% (35–39); 42.6% of French patients abandoned HRT, through lack of interest or fear of breast cancer, and interruption for more than one year was associated with greater loss of femoral BMD (−52 versus −17 mg/cm²; p = 0.022); in FXPOI, 52% never took HRT, started it late or stopped before 45 years [11, 122, 124].

7.8. Fertility: preservation, natural conception and oocyte donation

Preservation before POI. The PNDS considers eligible for fertility preservation women who are to receive gonadotoxic treatments, those with a familial risk of POI and other at-risk situations; the options are ovarian tissue cryopreservation, ovarian suppression with a GnRH agonist, cryopreservation of mature oocytes and of embryos; hyperstimulation with oocyte or embryo vitrification, as early as possible, is the reference technique after puberty, and in vitro maturation of immature oocytes (24–48 h) is an alternative [1]. Before 38 years, 15 to 20 vitrified oocytes are needed for a 70–80% probability of a live birth, with no additional obstetric or perinatal risk demonstrated after vitrification [1]. An FSH above 20 IU/L predicts little benefit from stimulation; if it persists, natural-cycle retrieval may be attempted in women under 35 years with a normal antral count, without a guarantee of pregnancy [1]. Ovarian tissue cryopreservation, by laparoscopy, is the only technique that can restore both exocrine and endocrine functions, with more than 100 pregnancies described almost exclusively after orthotopic grafting; it is used in prepubertal girls and its success depends on the initial follicular density, which limits its indications in POI [1]. Once POI is established, cryopreservation is no longer possible in most cases, because the follicular pool is exhausted [1, 2]. The 2024 guideline recommends discussing preservation with women at risk and offering it in iatrogenic causes before treatment (conditional) [2].

Genetic forms. In the FMR1 premutation, the PNDS recommends offering, where possible, oocyte self-preservation after stimulation, usually over several cycles, with preimplantation diagnosis possible but difficult [1]. In Turner syndrome, the 2024 guideline recommends controlled ovarian stimulation with oocyte cryopreservation as the first option in post-menarchal girls with adequate psychological maturity, in experienced centres, and annual AMH measurement if preservation is being considered [32]. In galactosaemia, preservation may not be successful and should only be offered to prepubertal girls within a protocol approved by an ethics committee [31]. A 2026 review of the genetic conditions predisposing to POI (Turner syndrome, FMR1 premutation, galactosaemia, rare monogenic diseases and cancer predisposition syndromes) concludes that reserve trajectories are highly variable, that the window for preservation may be brief and unpredictable, and that early referral after genetic diagnosis is justified, with longitudinal monitoring of reserve [157]. Identification of a genetic cause allows oocyte self-preservation to be offered to female relatives who are not yet affected [1]. The BPES GeneReviews notes that there is little evidence for longitudinal monitoring of reserve in young girls [21].

Natural conception. POI greatly reduces natural conception, but in non-surgical POI there may be ovarian activity and spontaneous pregnancy (both strong recommendations), with natural pregnancies in 4–6% and spontaneous conception in 5–10% [1, 2]. Those who do not wish to become pregnant should use contraception (strong) [2]. In FXPOI, there are cases of children with fragile X syndrome conceived after the diagnosis of POI [15].

Oocyte donation. This is the established option and the first to propose for pregnancy: a live birth rate of about 30–40% per embryo transferred and a cumulative pregnancy rate of 70–80% after four cycles [1, 2]. In France, the mean wait for oocytes is 3 years (Agence de la Biomédecine), which justifies informing patients early [1]. Donation by a sister in non-iatrogenic POI implies shared genetic risk and more cycle cancellations (strong, low certainty), and the aetiology of POI should be investigated before donation (strong) [2]; GeneGuidance therefore recommends testing sisters for the familial variant before intrafamilial donation [16]. Pregnancy with donated oocytes is high-risk and should be followed in an appropriate obstetric unit, with disclosure of the origin of the oocytes (strong) [2]. In the DONOR individual-patient data meta-analysis (2747 donation pregnancies, 4699 autologous in vitro fertilisation pregnancies and 33,323 natural pregnancies), the adjusted OR for hypertensive complications was 2.62 (95% CI 2.22–3.10) to 2.85 (2.30–3.54) and that for pre-eclampsia 2.28 (1.88–2.78) to 2.39 (1.94–2.94), independently of age, body mass index, multiple pregnancy and the number of embryos transferred [158]. Adoption and, rarely, embryo donation are alternatives [1].

Pregnancy according to aetiology. Natural pregnancy after idiopathic POI or after most chemotherapies carries no increased obstetric or neonatal risk (strong); pregnancy after uterine irradiation is high-risk (strong); cardiometabolic and thyroid function should be assessed before pregnancy (strong); in women exposed to anthracyclines or cardiac irradiation, cardiological assessment is necessary (strong); some situations carry such risk that donation may be considered inappropriate (strong) [2]. In Turner syndrome, pregnancy carries high obstetric and non-obstetric risk, and full cardiac screening and counselling by maternal-fetal medicine and experienced cardiology are recommended before planning pregnancy, particularly with donation [2, 32]. In FXPOI, pregnancies have no more complications than in the general population [15, 124].

7.9. Specific management by genetic aetiology

The BPES GeneReviews lists the factors to avoid: follow-up without assessment of reserve (reduces options), unsupervised hormone therapy, delay in HRT (osteoporosis and cardiovascular disease, aggravated by smoking and alcohol), chronic glucocorticoids, some anticonvulsants and unsupervised fertility supplements [21]. Oral hormone tablets contain lactose as an excipient; some patients with galactosaemia have used them for years without apparent negative effect (personal observation of an author of the reference chapter) [114].

7.10. Hormone therapy in iatrogenic POI with genetic relevance

The 2024 guideline recommends a personalised approach after gynaecological or breast cancer: HRT recommended after squamous cell carcinoma of the cervix (it does not increase recurrence), personalised after cervical adenocarcinoma, considered after early low-risk endometrial adenocarcinoma and after epithelial ovarian cancer (conditional), and avoided in hormone-dependent tumours (uterine sarcoma, endometrioid carcinoma, clear cell carcinoma of the ovary, granulosa cell tumour and other sex cord-stromal tumours; strong) [2]. Information about POI should precede risk-reducing salpingo-oophorectomy (strong), and personalised HRT or puberty induction are recommended after haematopoietic cell transplantation and other gonadotoxic therapies (strong) [2]. The PNDS summarises gonadotoxicity: alkylating agents (cyclophosphamide, busulfan, melphalan, chlorambucil) at high risk, platinum salts and doxorubicin at intermediate risk, anthracyclines and taxanes at low risk; POI at 6 months in 42% with alkylating agents versus 14% without them; cumulative risk of 60% (95% CI 41–79%) after alkylating chemotherapy versus 3% (1–7%) without alkylating agents in Hodgkin lymphoma; permanent amenorrhoea with cumulative cyclophosphamide of 5 g at 40 years, 9 g at 30 and 20 g at 20; radiotherapy above 10 Gy at high risk; ovarian transposition reducing the risk by about 50% [1, 83]. In a Brazilian phase 2 trial in women aged 35 years or younger with locally advanced cervical cancer (n = 22), subcutaneous grafting of fresh ovarian tissue before pelvic radiotherapy maintained a hormonal profile of preserved function at 6 months in 85% of the intervention group, versus POI in all controls [159].

7.11. Comparative regulatory status

The Brazilian package inserts were not consulted; only active registration was verified.

7.12. Trials and developments in the last 12 months

Trials of cell therapies and PRP are small, heterogeneous and almost all without published results; the improvements in hormonal markers in the PRP meta-analysis did not translate into more pregnancies, and the mixing of patients with low reserve and with POI precludes extrapolation to genetic POI [167]. None of these trials selects patients by genetic aetiology, and none of these interventions is authorised by the FDA or the European Commission for POI. There is no gene therapy for POI: the only registration using that designation is a follow-up study of a cell therapy with autologous mesenchymal stem cells, with no efficacy results published in the sources consulted [166].

In vitro follicle activation. The technique, developed by Hsueh's group, consists of fragmenting the ovarian cortex (which disrupts the Hippo pathway and increases downstream growth factors), treating the fragments with Akt pathway stimulators and regrafting them; in the founding study, it promoted follicular growth in patients with POI, allowed mature oocytes to be obtained and, after in vitro fertilisation, the birth of a healthy child, and a few other births were later described in China [1, 168]. A study of 35 Chinese patients was completed in 2021 [169]. The Heddar cohort suggests that genetic diagnosis, by predicting residual reserve in 60.5% of cases, may come to select candidates for in vitro activation [13]. Other approaches are preclinical or preliminary: stem cells (one clinical pregnancy after transplantation of umbilical cord mesenchymal stem cells on a collagen scaffold), artificial ovaries in rodents and recombinant AMH to limit chemotherapy-induced follicular "burnout" [1].

7.13. Pregnancy and lactation, transition and situations to avoid

HRT is not contraceptive and may coexist with sporadic ovulation; the absence of bleeding after the progestogen phase requires a pregnancy test, and any pregnancy obtained should be followed according to the aetiology of POI and the origin of the oocytes [1, 2]. The transition from paediatric to adult medicine should ensure continuity of HRT without interruption, because interruption for more than one year is associated with loss of BMD [11, 122]. Smoking (smokers with the FMR1 premutation develop POI five years earlier), compounded hormone preparations, unsupervised supplements and chronic glucocorticoids when avoidable should be avoided [2, 15, 21].

7.14. Divergences between recommendations

8. Reproductive Options and Family Counselling

8.1. Principles of counselling

The implications of genetic testing should be discussed before it is performed, and full genetic counselling should be considered (good practice point of the 2024 guideline); consent is written and tests are performed in centres able to provide counselling and family studies, as required by French law [1, 2, 11]. The recognised benefits of genetic diagnosis are psychological (a cause instead of "idiopathic"), prognostic (fertility potential and residual reserve), preventive (screening for comorbidities with the competent teams, for example cancer genetics) and familial (screening of sisters, daughters and brothers, and fertility preservation for female relatives not yet affected) [1, 11]. Female relatives (sisters, daughters) of women with non-iatrogenic POI have an increased risk (strong, low certainty), should be informed of the signs and symptoms and may benefit from reserve testing and family planning, but there are no established methods of prediction or prevention; relatives of women with the FMR1 premutation or another genetic cause should receive counselling and testing [2]. In Finland, the relative risk of POI in female relatives of women with POI was 4.6 (95% CI 3.3–6.5) [26].

8.2. Recurrence risks by aetiology

8.3. Cascade screening and male relatives

Cascade screening of adults is indicated for the FMR1 premutation, because apparently normal brothers may be carriers (normal transmitting males), with risk for their daughters' children and their own risk of FXTAS [90]. In meiosis and DNA repair genes, brothers may have azoospermia or severe oligospermia and should be included in the family study [11, 20]. In the MCM8/MCM9 syndromes, GeneReviews recommends testing asymptomatic siblings to begin tumour surveillance [20]. In the forms due to MSH4/MSH5, the later onset leaves time for predictive testing of female relatives and for fertility preservation [28]. A sister carrying a causal variant may not be a suitable oocyte donor [2, 16].

8.4. Reproductive options

FMR1 premutation carriers who wish to avoid an affected child have three options: in vitro fertilisation with PGT-M, spontaneous conception with prenatal diagnosis by chorionic villus sampling or amniocentesis, or donor oocytes [92]. PGT-M avoids termination of pregnancy and allows only non-carrier embryos to be transferred, but involves costs and difficulties, and is often limited by the number of oocytes [1, 92]. In monogenic forms with a known variant, PGT-M and prenatal diagnosis are possible, but natural conception is infrequent and oocyte donation is the usual route [20]. A diagnosis of FXPOI does not eliminate the possibility of natural conception, and there are cases of children with fragile X conceived after the diagnosis, which justifies discussing contraception or prenatal diagnosis [15].

8.5. Counselling of minors

The 2024 guideline establishes that age should not restrict access to genetic testing in POI [2]. In girls, diagnostic testing of established POI is accepted; predictive testing of asymptomatic minor relatives, particularly for FMR1, raises its own issues and should be discussed with the family, weighing the benefit of possible fertility preservation before follicular depletion [2, 157]. In Turner syndrome, the guideline recommends providing developmentally appropriate information about fertility and counselling the girl and her parents, as early as possible, about family-building options (fertility preservation, fostering, adoption, surrogacy, oocyte or embryo donation, or not having children) [32]. Psychological support is often needed, because options such as oocyte donation may influence how the young woman views future relationships [1].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Frequency of POI

Classic estimates attribute spontaneous POI to 1–2% of women before 40 years, 1/1000 at 30 years and 1/10,000 before 20 years [1]. A meta-analysis of 31 studies (1980–2017) obtained a pooled prevalence of 3.7% (95% CI 3.1–4.3) for POI and 12.2% (10.5–14) for early menopause, higher in countries with a medium and low human development index [25]; the 2024 guideline places the prevalence of non-iatrogenic POI between about 1% in older studies and 3.5% in recent ones [2]. These figures measure different things: surveys of age at menopause (such as the UK Biobank, with 1.14% natural menopause before 40 years among 104,733 women) depend on recall and definition, reimbursement registries capture only diagnosed and treated women, and clinical series exclude those not referred [19, 26]. In Finland, in the HRT reimbursement registry (5011 women, 1988–2017), incidence was highest in the 35–39 year group (73.8/100,000 woman-years in 1993–1997 and 39.9 in 2013–2017), increased since 2007 from 7.0 to 10.0/100,000 woman-years in the 15–19 year group, and cumulative incidence before 40 years was 478/100,000 women [26]. In Sweden, in a registry study of all 1,036,918 women born between 1973 and 1993, 19,253 had POI before 40 years, a total prevalence of 1.9% (95% CI 1.7–2.1), with 1.7% spontaneous and 0.2% iatrogenic forms; 98.8% of cases were identified through the registry of drugs dispensed for climacteric symptoms and only 4.1% by diagnosis code, which may overestimate prevalence [27]. The PNDS, which presents the 19,253 women as the size of the registry, adds the breakdown of 89.3% spontaneous, 10.7% iatrogenic and 2.4% due to bilateral oophorectomy [1, 27]. The incidence of spontaneous forms appears stable and that of iatrogenic forms is increasing with survival after childhood cancer; a 2025 Chinese guideline estimates that iatrogenic forms account for about half of cases and up to 82.2% of cancer survivors under 24 years [1, 171].

9.2. Ethnic origin and founder populations

In the United States, the study by Luborsky and colleagues found a higher prevalence of POI in African American and Hispanic women than in white women, and a lower prevalence in women of Chinese and Japanese ancestry (according to the PNDS) [1]. Prematurity has been proposed as a risk factor [1]. The distribution of genetic causes varies with the population: the FMR1 premutation was found in 2.0% of British POI cases, whereas a meta-analysis of four Asian studies showed no association and frequencies in Chinese women were 0.49–1.6% [11, 94]. The FSHR founder variant c.566C>T, p.(Ala189Val) explains 98% of pathogenic alleles in Finland, with an allele frequency of 0.96% in 1976 Finns (enriched in the North-East; carriers about 1/85), 1/1162 in Swiss individuals and absence in 1094 Danes and 540 Singaporean Chinese [16, 172]. The STAG3 variant c.2776C>T is enriched in the Russian population [105], the 13 bp AIRE deletion is a founder variant in the United Kingdom, and autoimmune polyendocrinopathy type 1 has a high frequency in Finland (4/100,000), in Iranian Jews (1:6500–1:9000) and in Sardinia (R139X carriers 1.7%, about 1/14,400 affected), versus 0.5/100,000 in France [16, 70]. In classic galactosaemia, the incidence is about 1/48,000–1/50,000 in the United States and 1/16,476 in Ireland, very rare in Asians, and the p.(Ser135Leu) variant, without POI on treatment, predominates in African populations (homozygosity in about 1/20,000 African Americans) [16, 22, 114]. VWM has an incidence of 1/80,000 live births in the Netherlands [130]. Consanguineous cohorts (Turkish, North African, Middle Eastern) have a higher genetic yield and a recessive predominance [13, 142], and the absence of recessive cases in an Estonian cohort illustrates population specificity [104].

9.3. Frequency of genetic causes

The FMR1 premutation occurs in about 344/100,000 women in the Hunter meta-analysis (2014) and 677/100,000 in a population sample of 19,996 people (Maenner 2013); FXPOI was estimated at about 1/3560 women in the United States (Hantash 2011, all according to GeneReviews) [15]. The premutation explains 1–5% of sporadic POI and up to 13% of familial POI according to the 2024 guideline, about 5% according to the PNDS and 2% of sporadic and 14% of familial POI according to the Oxford Desk Reference [1, 11, 12]. Chromosomal abnormalities explain 10–13% [11]. In the Chinese cohort of 955 patients, genetic causes accounted for 13.15%, autoimmune causes 12.04% and iatrogenic causes 7.29%, with about 68% idiopathic [121]. BPES has an estimated prevalence of 1/50,000 births, probably underestimated (1–9/100,000 according to the PNDS) [1, 21]. Loss-of-function variants of MCM8 and MCM9 have gnomAD frequencies of about 1.4/100,000 and 2.5/100,000 [20]. Data from Portugal and Brazil in the sources consulted are limited to cases and trials: the first homozygous GDF9 case was described in São Paulo [99], consanguineous Brazilian sisters homozygous for a NOBOX variant [100, 102], a Brazilian cohort with a yield of 12–20% cited by the 2024 guideline [11] and the subcutaneous ovarian grafting trial in São Paulo [159]; the sources consulted do not describe the prevalence of POI in Portugal or Brazil.

9.4. Special populations

Survivors of childhood cancer constitute the most important iatrogenic risk group: among 100 survivors, one in six had POI (Larsen 2003, according to the Oxford Desk Reference), and in the St. Jude cohort (921 survivors, median age 31.7 years) POI was associated with high-dose alkylating agents and ovarian radiotherapy at any dose [12, 83]. Women with the FMR1 premutation have documented underdiagnosis and undertreatment [124]. Adolescents with POI have greater bone risk because they have not reached peak bone mass [1]. Trans and non-binary people registered female at birth are included in the NICE recommendations [3].

9.5. Organisation of care, European networks and associations

In France, POI is managed by the Centre de Référence des Maladies Endocriniennes Rares de la Croissance et du Développement (CRMERCD; paediatric coordination at Hôpital Robert Debré and constituent centres at the Necker, Saint-Antoine and Pitié-Salpêtrière hospitals), in collaboration with the centres for rare gynaecological conditions and for disorders of sex development (DEVGEN), within the FIRENDO network [1, 34, 173]. The PNDS serves as a reference for the attending physician for the care protocol with a request for exemption from patient co-payment (ticket modérateur) as a condition outside the list of long-term diseases (ALD "hors liste") and identifies the drugs used outside the terms of the marketing authorisation [1]. The Plan France Médecine Génomique 2025 provides access to whole-genome sequencing after national multidisciplinary review [139]. The relevant European network is Endo-ERN, which aims to ensure lifelong endocrine care and reduce inequalities of access, and which has published a survey on the genetic diagnosis of DSD and maturation disorders in its centres [33, 174]. Patient associations, for which the PNDS refers readers to Orphanet, include The Daisy Network in the United Kingdom [1, 12, 35].

10. Illustrative Clinical Case

10.1. POI associated with the FMR1 premutation with a five-year diagnostic delay

Fink and colleagues published in 2018 the case of a woman seen at the National Institutes of Health Clinical Center in May 2006, at 20 years of age, because she was "not feeling herself": hot flushes, night sweats, insomnia, crying spells, sadness, loss of interest and a feeling of agitation [92]. At 18 years she had had severe depression that forced her to interrupt her first year at university, and since then she had been on a complex psychotropic regimen [92]. Cascade testing at 4 years had revealed an FMR1 premutation with 100–110 CGG repeats; her older brother had fragile X syndrome diagnosed at 9 years, and her mother and aunt were carriers with "premature menopause" [92]. Menarche occurred at 11 years, she never had regular cycles, took an oral contraceptive from 13 to 16 years for dysmenorrhoea and menorrhagia and, between 16 and 18 years, had only spotting every 3–4 months; between 17 and 18 years night sweats appeared, she was prescribed levothyroxine with normal free T4 and TSH, and at an assessment in another centre Cushing syndrome was considered, because of elevated morning cortisol and urinary free cortisol and an abnormal dexamethasone suppression test [92].

On admission, vital signs were normal, body mass index was 28.3 kg/m² and there were no Cushingoid features; diurnal serum cortisol levels and 24-hour urinary free cortisol were normal [92]. Transvaginal ultrasound showed a 2 mm endometrium, a very small left ovary (1.2 × 1.6 × 1.0 cm) without visible follicles and a right ovary that could not be reliably identified; FSH was 46 IU/L, LH 26 IU/L and oestradiol 23.3 pg/mL, all at menopausal levels; anti-21-hydroxylase, antiperoxidase and antithyroglobulin antibodies were negative, and prolactin, pituitary MRI, TSH and free T4 were normal [92]. The psychiatric consultation concluded that she had a mood disorder secondary to a medical condition, possibly POI [92]. She was discharged with a diagnosis of FXPOI and started HRT with transdermal oestradiol 100 µg/day and oral medroxyprogesterone acetate 10 mg/day for the first 12 days of each month; psychotropic drugs and levothyroxine were tapered and stopped, she began to menstruate regularly and the anxiety and depression resolved [92]. She married in August 2012 and conceived shortly afterwards, without medical intervention and while on HRT; she terminated that pregnancy because prenatal diagnosis showed a full FMR1 mutation in the fetus; she then had two spontaneous pregnancies on HRT, in 2013 and 2016, with the birth of two healthy girls, even though, shortly before the last conception, an in vitro fertilisation clinic had given her a probability of conception below 0.1% and advised oocyte donation [92].

The case illustrates four points of this article: diagnostic delay even with known carrier status, oligoamenorrhoea and elevated FSH (about 5 years without adequate HRT); the psychiatric symptoms and unnecessary treatments that oestrogen deficiency can cause; the possibility of natural conception in FXPOI, which makes it necessary to discuss contraception or prenatal diagnosis; and the risk of expansion to a full mutation in the maternal offspring [15, 92, 124]. The HRT regimen used (transdermal oestradiol 100 µg/day with a progestogen 12 days per month) corresponds to the physiological doses recommended by the 2024 guideline [11]. The review accompanying the case includes figures from other studies: the change in femoral neck and lumbar spine BMD over 3 years in women with spontaneous 46,XX POI on oestrogen and progestogen, with or without testosterone, compared with controls (https://pmc.ncbi.nlm.nih.gov/articles/PMC6278244/figure/F2/), and the changes in serum creatinine, renin activity, angiotensin II and aldosterone with physiological HRT versus oral contraceptives (https://pmc.ncbi.nlm.nih.gov/articles/PMC6278244/figure/F3/) [92].

10.2. Family with POI due to biallelic NOBOX variants

Sassi and colleagues described in 2021 a Belgian patient, born to non-consanguineous parents, who attended a fertility clinic at 35 years for oocyte donation; at 17 years she had been diagnosed with POI because of absent pubertal development and primary amenorrhoea, with FSH of 106 IU/L and oestradiol below 12 ng/L, and laparoscopy had shown a small uterus and streak ovaries [100]. Puberty was induced with HRT; she was 163 cm tall, had a body mass index of 19 kg/m², had no dysmorphic features, and karyotype and array-CGH were normal; her older sister had the same phenotype and her mother had menopause at 51 years [100]. Trio exome sequencing identified compound heterozygosity for NM_001080413.3:c.826C>T, p.(Arg276Ter), inherited from the father, and NM_001080413.3:c.1421del, p.(Gly474AlafsTer76), inherited from the mother, confirmed by Sanger sequencing in the sister; the pedigree is shown in figure 1 of the article [100, 115]. The authors concluded that biallelic null NOBOX variants cause severe POI with primary amenorrhoea and that haploinsufficiency is tolerated, as suggested by the normal menopause of the carrier mother [100]. The case shows the change in the inheritance model of NOBOX, from dominant to recessive, and the importance of a family study before intrafamilial donation.

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

Without HRT, POI reduces life expectancy, mainly through cardiovascular disease (strong): in the prospective cohort of 20,555 women, menopause before 40 years was associated with a life expectancy 2 years shorter than menopause at 50–54 years, and prophylactic bilateral oophorectomy before 45 years with even higher mortality [1, 2]. HRT restores endothelial function and maintains BMD, but a reduction in cardiovascular mortality, although probable, has not been demonstrated in POI [1]. The reproductive prognosis depends on the aetiology: in meiosis and DNA repair genes it is very unfavourable, in follicular growth genes reserve may persist if the diagnosis is made early, and in idiopathic POI ovarian activity resumes in about 23% and natural pregnancy occurs in 4–10% [1]. Some genetic aetiologies determine the vital prognosis in their own right: autoimmune polyendocrinopathy type 1 has a standardised mortality ratio of 11 in Finland, the MCM9 and FANCM syndromes add cancer risk and the ovarioleukodystrophies have a progressive neurological course [16, 20, 23, 63]. Adherence to HRT is the main modifiable factor: discontinuation for more than one year is associated with loss of BMD, and half of women with FXPOI are undertreated [122, 124].

11.2. Surveillance table

The HRT and puberty induction intervals coincide with those of treatment (dose increase every 6–12 months over 2–3 years, progestogen after about 2 years or at the first bleed, HRT until about 51 years) [1, 11].

12. Summary and Final Considerations

Premature ovarian insufficiency is the loss of ovarian activity before 40 years of age, diagnosed by menstrual disturbance of at least 4 months with FSH above 25 IU/L; the 2024 international guideline accepts a single elevated FSH, whereas the French PNDS and NICE retain two measurements [1–3]. Its frequency lies between 1% and 3.7% of women, and the genetic component is the leading identifiable cause, although most cases remain undiagnosed after current investigation [1, 25]. Genetic investigation follows an established order: karyotype with a search for mosaicism and Y material, specific testing for the FMR1 premutation (which panels and exome sequencing do not detect) and, if both are normal, a curated panel or exome in a specialised laboratory, with genome sequencing in negative cases with strong suspicion [1, 2, 11].

The monogenic landscape has changed in recent years. Meiosis and DNA repair genes are the most numerous group, and some (MCM8, MCM9, FANCM, STAG3, Fanconi pathway genes) add a risk of germ cell tumours, polyposis or cancer, with their own surveillance [13, 20, 63]. Genes previously considered dominant, such as BMP15, GDF9 and NOBOX, cause disease mainly when both copies are affected, and the UK Biobank study showed that almost all heterozygous truncating variants in the described genes exist in women with a normal reproductive lifespan; heterozygous missense variants should therefore be treated as of uncertain significance or as risk factors and not as diagnostic [17–19, 43]. The syndromic forms (BPES, galactosaemia, autoimmune polyendocrinopathy type 1, ovarioleukodystrophies, Perrault syndrome, chromosomal instability syndromes) require a search for manifestations in other organs, and autoimmune adrenal insufficiency is the complication not to be missed [1, 11].

Treatment is based on HRT with 17β-oestradiol at physiological doses (at least 100 µg/day transdermally or 2 mg/day orally according to the 2024 guideline; 50–100 µg/day according to the PNDS) and a progestogen for at least 12 days per month, maintained until about 51 years, with gradual puberty induction from 11–12 years in prepubertal girls [1, 11]. In Europe, this use falls outside the terms of the SmPCs, which refer to the usual menopause, whereas the American oestradiol label includes primary ovarian failure [149, 153]. Adherence to HRT, bone, cardiovascular, thyroid and adrenal surveillance, psychological support and information on the possibility of natural pregnancy (4–10%) and on oocyte donation are part of follow-up [1, 2]. Genetic diagnosis makes it possible to predict residual reserve, guide fertility preservation for female relatives not yet affected and test sisters before intrafamilial donation [1, 16, 157]. Therapies intended to restore ovarian function (in vitro follicle activation, platelet-rich plasma, stem cells, immunomodulation in autoimmune POI) remain experimental, with no demonstration in controlled trials of an increase in pregnancies or live births [160, 167, 168].

13. References

1. Centre de Référence des Maladies Endocriniennes Rares de la Croissance et du Développement (CRMERCD), Filière FIRENDO. Protocole National de Diagnostic et de Soins (PNDS): Insuffisance ovarienne prématurée (en dehors du syndrome de Turner). Saint-Denis La Plaine: Haute Autorité de Santé; March 2021 (published 6 May 2021). Available at: https://www.has-sante.fr/jcms/p_3264990
2. Panay N, Anderson RA, Bennie A, Cedars M, Davies M, Ee C, et al. Evidence-based guideline: premature ovarian insufficiency(). Hum Reprod Open. 2024;2024(4):hoae065. doi:10.1093/hropen/hoae065. PMID: 39660328. PMC11631070.
3. National Institute for Health and Care Excellence. Menopause: identification and management (NG23). Recommendations 1.7: diagnosing and managing premature ovarian insufficiency in people under 40. London: NICE; published 12 November 2015, last updated 15 April 2026. Available at: https://www.nice.org.uk/guidance/ng23
4. Orphanet. Primary ovarian failure (ORPHA:619) [Internet]. Paris: INSERM; 2026. Available at: https://www.orpha.net/en/disease/detail/619
5. Orphanet. Rare genetic premature ovarian failure (ORPHA:485382) [Internet]. Paris: INSERM; 2026. Available at: https://www.orpha.net/en/disease/detail/485382
6. Orphanet. Fragile X-associated primary ovarian insufficiency (ORPHA:642691) [Internet]. Paris: INSERM; 2026. Available at: https://www.orpha.net/en/disease/detail/642691
7. Orphanet. 46,XX gonadal dysgenesis (ORPHA:243) [Internet]. Paris: INSERM; 2026. Available at: https://www.orpha.net/en/disease/detail/243
8. Orphanet. 46,XX ovarian dysgenesis-short stature syndrome (ORPHA:444048) [Internet]. Paris: INSERM; 2026. Available at: https://www.orpha.net/en/disease/detail/444048
9. Online Mendelian Inheritance in Man, OMIM®. Premature ovarian failure 1; POF1. MIM #311360 [Internet]. Baltimore (MD): Johns Hopkins University; 2026. Available at: https://www.omim.org/entry/311360
10. Online Mendelian Inheritance in Man, OMIM®. Ovarian dysgenesis 1; ODG1. MIM #233300 [Internet]. Baltimore (MD): Johns Hopkins University; 2026. Available at: https://www.omim.org/entry/233300
11. ESHRE, ASRM, CRE-WHiRL, IMS. Evidence-based guideline: premature ovarian insufficiency. Full guideline. Strombeek-Bever: European Society of Human Reproduction and Embryology; 2024. Available at: https://www.eshre.eu/Guidelines-and-Legal/Guidelines/Management-of-premature-ovarian-insufficiency
12. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Premature ovarian failure (POF). p. 778–779.
13. Heddar A, Ogur C, Da Costa S, Braham I, Billaud-Rist L, Findikli N, et al. Genetic landscape of a large cohort of Primary Ovarian Insufficiency: New genes and pathways and implications for personalized medicine. EBioMedicine. 2022;84:104246. doi:10.1016/j.ebiom.2022.104246. PMID: 36099812. PMC9475279.
14. Ke H, Tang S, Guo T, Hou D, Jiao X, Li S, et al. Landscape of pathogenic mutations in premature ovarian insufficiency. Nat Med. 2023;29(2):483-492. doi:10.1038/s41591-022-02194-3. PMID: 36732629. PMC9941050.
15. Hunter JE, Berry-Kravis E, Hipp H, Todd PK. FMR1 disorders. In: Adam MP, Bick S, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Published 16 June 1998; updated 21 November 2019; revised 16 May 2024. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1384/
16. GeneGuidance. Gene records for BMP15, FIGLA, NOBOX, NR5A1, FOXL2, GDF9, STAG3, SYCE1, MCM8, MCM9, HFM1, PSMC3IP, EIF2B5, FSHR, LHCGR, AIRE, GALT, FMR1, SPIDR, MSH4, MSH5, BNC1, ZSWIM7, NUP107, FANCM, ERCC6, DIAPH2 and POF1B [local database]. Version of 26 September 2026.
17. Hervé B, Verebi C, Bonnier M, Jordan P, Stroe D, Voisin C, et al. From Dominant Assumptions to Recessive Trait: Rethinking BMP15 in Ovarian Dysfunction. Reprod Sci. 2025;32(6):1926-1938. doi:10.1007/s43032-025-01831-9. PMID: 40274730.
18. Jordan P, Verebi C, Hervé B, Perol S, Bernard V, Karila D, et al. Revisiting GDF9 variants in primary ovarian insufficiency: A shift from dominant to recessive pathogenicity?. Gene. 2024;927:148734. doi:10.1016/j.gene.2024.148734. PMID: 38942181.
19. Shekari S, Stankovic S, Gardner EJ, Hawkes G, Kentistou KA, Beaumont RN, et al. Penetrance of pathogenic genetic variants associated with premature ovarian insufficiency. Nat Med. 2023;29(7):1692-1699. doi:10.1038/s41591-023-02405-5. PMID: 37349538.
20. Helderman NC, Ben Yahia S, Rajkovic A, Goldberg Y, Nielsen M. MCM8/MCM9-related syndromes. In: Adam MP, Bick S, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Published 16 June 2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK623157/
21. Matton C, Thomaes A, Cools M, Verdin H, De Baere E. Blepharophimosis, ptosis, and epicanthus inversus syndrome. In: Adam MP, Bick S, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Published 8 July 2004; updated 23 June 2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1441/
22. Berry GT. Classic galactosemia and clinical variant galactosemia. In: Adam MP, Bick S, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Published 4 February 2000; updated 11 March 2021. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1518/
23. van der Knaap MS, Fogli A, Boespflug-Tanguy O, Abbink TEM, Schiffmann R. Childhood ataxia with central nervous system hypomyelination / vanishing white matter. In: Adam MP, Bick S, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Published 20 February 2003; updated 4 April 2019. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1258/
24. Li T, Faridi R, Newman WG, Friedman TB. Perrault syndrome overview. In: Adam MP, Bick S, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Published 25 September 2014; updated 8 May 2025. Available at: https://www.ncbi.nlm.nih.gov/books/NBK242617/
25. Golezar S, Ramezani Tehrani F, Khazaei S, Ebadi A, Keshavarz Z. The global prevalence of primary ovarian insufficiency and early menopause: a meta-analysis. Climacteric. 2019;22(4):403-411. doi:10.1080/13697137.2019.1574738. PMID: 30829083.
26. Silvén H, Savukoski SM, Pesonen P, Pukkala E, Gissler M, Suvanto E, et al. Incidence and familial risk of premature ovarian insufficiency in the Finnish female population. Hum Reprod. 2022;37(5):1030-1036. doi:10.1093/humrep/deac014. PMID: 35134918. PMC9071220.
27. Lagergren K, Hammar M, Nedstrand E, Bladh M, Sydsjö G. The prevalence of primary ovarian insufficiency in Sweden; a national register study. BMC Womens Health. 2018;18(1):175. doi:10.1186/s12905-018-0665-2. PMID: 30359245. PMC6202813.
28. Fergus-Mackie J, Kline BL, Robevska G, van den Bergen J, Bell K, Dulon J, et al. MSH4 and MSH5 variants in premature ovarian insufficiency: A literature review and case study. Maturitas. 2026;211:109034. doi:10.1016/j.maturitas.2026.109034. PMID: 42386447.
29. ESHRE, ASRM, CREWHIRL and IMS Guideline Group on POI; Panay N, Anderson RA, Bennie A, Cedars M, Davies M, et al. Evidence-based guideline: Premature Ovarian Insufficiency. Fertil Steril. 2025;123(2):221-236. doi:10.1016/j.fertnstert.2024.11.007. PMID: 39652037.
30. American College of Obstetricians and Gynecologists. Committee Opinion No. 698: Hormone Therapy in Primary Ovarian Insufficiency. Obstet Gynecol. 2017;129(5):e134-e141. doi:10.1097/AOG.0000000000002044. PMID: 28426619.
31. Welling L, Bernstein LE, Berry GT, Burlina AB, Eyskens F, Gautschi M, et al. International clinical guideline for the management of classical galactosemia: diagnosis, treatment, and follow-up. J Inherit Metab Dis. 2017;40(2):171-176. doi:10.1007/s10545-016-9990-5. PMID: 27858262. PMC5306419.
32. Gravholt CH, Andersen NH, Christin-Maitre S, Davis SM, Duijnhouwer A, Gawlik A, et al. Clinical practice guidelines for the care of girls and women with Turner syndrome. Eur J Endocrinol. 2024;190(6):G53-G151. doi:10.1093/ejendo/lvae050. PMID: 38748847. PMC11759048.
33. European Reference Network on Rare Endocrine Conditions (Endo-ERN) [Internet]. Leiden: Endo-ERN; 2026. Available at: https://endo-ern.eu
34. Filière de santé maladies rares endocriniennes FIRENDO [Internet]. Paris: FIRENDO; 2026. Available at: https://www.firendo.fr
35. The Daisy Network. Premature ovarian insufficiency support group [Internet]. London: The Daisy Network; 2026. Available at: https://www.daisynetwork.org.uk
36. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Turner syndrome (45,X). p. 70–75.
37. Orphanet. Turner syndrome (ORPHA:881) [Internet]. Paris: INSERM; 2026. Available at: https://www.orpha.net/en/disease/detail/881
38. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. X-autosome translocations. p. 700–702.
39. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. XXX and XXXX syndromes. p. 68–69.
40. Aittomäki K, Lucena JL, Pakarinen P, Sistonen P, Tapanainen J, Gromoll J, et al. Mutation in the follicle-stimulating hormone receptor gene causes hereditary hypergonadotropic ovarian failure. Cell. 1995;82(6):959-68. doi:10.1016/0092-8674(95)90275-9. PMID: 7553856.
41. Online Mendelian Inheritance in Man, OMIM®. Ovarian dysgenesis 2; ODG2. MIM #300510 [Internet]. Baltimore (MD): Johns Hopkins University; 2026. Available at: https://www.omim.org/entry/300510
42. Online Mendelian Inheritance in Man, OMIM®. Premature ovarian failure 5; POF5. MIM #611548 [Internet]. Baltimore (MD): Johns Hopkins University; 2026. Available at: https://www.omim.org/entry/611548
43. Jordan P, Verebi C, Perol S, Grotto S, Fouveaut C, Christin-Maitre S, et al. NOBOX gene variants in premature ovarian insufficiency: ethnicity-dependent insights. J Assist Reprod Genet. 2024;41(1):135-146. doi:10.1007/s10815-023-02981-y. PMID: 37921973. PMC10789696.
44. Zidoune H, Mezhoud I, Elzaiat M, Wankanit S, Rjiba K, Salmia L, et al. Primary ovarian insufficiency due to homozygous variants in the homeobox transcription factor NOBOX. Mol Biol Rep. 2025;53(1):150. doi:10.1007/s11033-025-11252-1. PMID: 41324758.
45. Zhao H, Chen ZJ, Qin Y, Shi Y, Wang S, Choi Y, et al. Transcription factor FIGLA is mutated in patients with premature ovarian failure. Am J Hum Genet. 2008;82(6):1342-8. doi:10.1016/j.ajhg.2008.04.018. PMID: 18499083. PMC2427265.
46. Yuan P, He Z, Sun S, Li Y, Wang W, Liang X, et al. Bi-allelic recessive loss-of-function mutations in FIGLA cause premature ovarian insufficiency with short stature. Clin Genet. 2019;95(3):409-414. doi:10.1111/cge.13486. PMID: 30474133.
47. Online Mendelian Inheritance in Man, OMIM®. Premature ovarian failure 7; POF7. MIM #612964 [Internet]. Baltimore (MD): Johns Hopkins University; 2026. Available at: https://www.omim.org/entry/612964
48. Lourenço D, Brauner R, Lin L, De Perdigo A, Weryha G, Muresan M, et al. Mutations in NR5A1 associated with ovarian insufficiency. N Engl J Med. 2009;360(12):1200-10. doi:10.1056/NEJMoa0806228. PMID: 19246354. PMC2778147.
49. Zhang D, Liu Y, Zhang Z, Lv P, Liu Y, Li J, et al. Basonuclin 1 deficiency is a cause of primary ovarian insufficiency. Hum Mol Genet. 2018;27(21):3787-3800. doi:10.1093/hmg/ddy261. PMID: 30010909.
50. Online Mendelian Inheritance in Man, OMIM®. Premature ovarian failure 8; POF8. MIM #615723 [Internet]. Baltimore (MD): Johns Hopkins University; 2026. Available at: https://www.omim.org/entry/615723
51. Caburet S, Arboleda VA, Llano E, Overbeek PA, Barbero JL, Oka K, et al. Mutant cohesin in premature ovarian failure. N Engl J Med. 2014;370(10):943-949. doi:10.1056/NEJMoa1309635. PMID: 24597867. PMC4068824.
52. de Vries L, Behar DM, Smirin-Yosef P, Lagovsky I, Tzur S, Basel-Vanagaite L. Exome sequencing reveals SYCE1 mutation associated with autosomal recessive primary ovarian insufficiency. J Clin Endocrinol Metab. 2014;99(10):E2129-32. doi:10.1210/jc.2014-1268. PMID: 25062452.
53. Wang J, Zhang W, Jiang H, Wu BL, Primary Ovarian Insufficiency Collaboration. Mutations in HFM1 in recessive primary ovarian insufficiency. N Engl J Med. 2014;370(10):972-4. doi:10.1056/NEJMc1310150. PMID: 24597873.
54. Carlosama C, Elzaiat M, Patiño LC, Mateus HE, Veitia RA, Laissue P. A homozygous donor splice-site mutation in the meiotic gene MSH4 causes primary ovarian insufficiency. Hum Mol Genet. 2017;26(16):3161-3166. doi:10.1093/hmg/ddx199. PMID: 28541421.
55. Guo T, Zhao S, Zhao S, Chen M, Li G, Jiao X, et al. Mutations in MSH5 in primary ovarian insufficiency. Hum Mol Genet. 2017;26(8):1452-1457. doi:10.1093/hmg/ddx044. PMID: 28175301. PMC5393145.
56. Zangen D, Kaufman Y, Zeligson S, Perlberg S, Fridman H, Kanaan M, et al. XX ovarian dysgenesis is caused by a PSMC3IP/HOP2 mutation that abolishes coactivation of estrogen-driven transcription. Am J Hum Genet. 2011;89(4):572-9. doi:10.1016/j.ajhg.2011.09.006. PMID: 21963259. PMC3188834.
57. Sirchia F, Giorgio E, Cucinella L, Valente EM, Nappi RE. Biallelic mutations in PSMC3IP are associated with secondary amenorrhea: expanding the spectrum of premature ovarian insufficiency. J Assist Reprod Genet. 2022;39(5):1177-1181. doi:10.1007/s10815-022-02471-7. PMID: 35352317. PMC9107541.
58. Smirin-Yosef P, Zuckerman-Levin N, Tzur S, Granot Y, Cohen L, Sachsenweger J, et al. A Biallelic Mutation in the Homologous Recombination Repair Gene SPIDR Is Associated With Human Gonadal Dysgenesis. J Clin Endocrinol Metab. 2017;102(2):681-688. doi:10.1210/jc.2016-2714. PMID: 27967308.
59. McGlacken-Byrne SM, Le Quesne Stabej P, Del Valle I, Ocaka L, Gagunashvili A, Crespo B, et al. ZSWIM7 Is Associated With Human Female Meiosis and Familial Primary Ovarian Insufficiency. J Clin Endocrinol Metab. 2022;107(1):e254-e263. doi:10.1210/clinem/dgab597. PMID: 34402903. PMC8684494.
60. Lokchine A, Zhang F, Cluzeau L, Le Page L, Belaud-Rotureau MA, Planes M, et al. SWS1-complex in premature ovarian insufficiency: SWSAP1 as a new POI gene. Hum Reprod. 2025;40(11):2189-2201. doi:10.1093/humrep/deaf177. PMID: 40991243.
61. Online Mendelian Inheritance in Man, OMIM®. Premature ovarian failure 10; POF10. MIM #612885 [Internet]. Baltimore (MD): Johns Hopkins University; 2026. Available at: https://www.omim.org/entry/612885
62. Online Mendelian Inheritance in Man, OMIM®. Ovarian dysgenesis 4; ODG4. MIM #616185 [Internet]. Baltimore (MD): Johns Hopkins University; 2026. Available at: https://www.omim.org/entry/616185
63. Nynäs E, Sulkava S, Nurmi AK, Suvanto M, Aittomäki K, FinnGen, et al. Recessive FANCM cancer syndrome with high cancer risks, chemotherapy toxicity, chromosome fragility, and gonadal failure. Genet Med. 2025;27(10):101521. doi:10.1016/j.gim.2025.101521. PMID: 40642874.
64. Bione S, Sala C, Manzini C, Arrigo G, Zuffardi O, Banfi S, et al. A human homologue of the Drosophila melanogaster diaphanous gene is disrupted in a patient with premature ovarian failure: evidence for conserved function in oogenesis and implications for human sterility. Am J Hum Genet. 1998;62(3):533-41. doi:10.1086/301761. PMID: 9497258. PMC1376955.
65. Lacombe A, Lee H, Zahed L, Choucair M, Muller JM, Nelson SF, et al. Disruption of POF1B binding to nonmuscle actin filaments is associated with premature ovarian failure. Am J Hum Genet. 2006;79(1):113-9. doi:10.1086/505406. PMID: 16773570. PMC1474115.
66. Online Mendelian Inheritance in Man, OMIM®. Blepharophimosis, ptosis, and epicanthus inversus; BPES. MIM #110100 [Internet]. Baltimore (MD): Johns Hopkins University; 2026. Available at: https://www.omim.org/entry/110100
67. Orphanet. Blepharophimosis-ptosis-epicanthus inversus syndrome (ORPHA:126) [Internet]. Paris: INSERM; 2026. Available at: https://www.orpha.net/en/disease/detail/126
68. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. Galactosemia. p. 425–434.
69. Orphanet. Classic galactosemia (ORPHA:79239) [Internet]. Paris: INSERM; 2026. Available at: https://www.orpha.net/en/disease/detail/79239
70. Online Mendelian Inheritance in Man, OMIM®. Autoimmune polyendocrine syndrome, type I, with or without reversible metaphyseal dysplasia; APS1. MIM #240300 [Internet]. Baltimore (MD): Johns Hopkins University; 2026. Available at: https://www.omim.org/entry/240300
71. Orphanet. Autoimmune polyendocrinopathy type 1 (ORPHA:3453) [Internet]. Paris: INSERM; 2026. Available at: https://www.orpha.net/en/disease/detail/3453
72. Orphanet. Ovarioleukodystrophy (ORPHA:99853) [Internet]. Paris: INSERM; 2026. Available at: https://www.orpha.net/en/disease/detail/99853
73. Orphanet. Perrault syndrome (ORPHA:2855) [Internet]. Paris: INSERM; 2026. Available at: https://www.orpha.net/en/disease/detail/2855
74. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. The mitochondrial DNA depletion syndromes: mitochondrial DNA polymerase deficiency. p. 415–422.
75. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. PMM2-CDG (congenital disorders of glycosylation type Ia). p. 771–777.
76. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Mulibrey nanism syndrome. p. 120–121.
77. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Myotonic dystrophy (DM1). p. 496–499.
78. Zlotogora J, Sagi M, Cohen T. The blepharophimosis, ptosis, and epicanthus inversus syndrome: delineation of two types. Am J Hum Genet. 1983;35(5):1020-7. PMID: 6613996. PMC1685801.
79. Crisponi L, Deiana M, Loi A, Chiappe F, Uda M, Amati P, et al. The putative forkhead transcription factor FOXL2 is mutated in blepharophimosis/ptosis/epicanthus inversus syndrome. Nat Genet. 2001;27(2):159-66. doi:10.1038/84781. PMID: 11175783.
80. Di Pasquale E, Beck-Peccoz P, Persani L. Hypergonadotropic ovarian failure associated with an inherited mutation of human bone morphogenetic protein-15 (BMP15) gene. Am J Hum Genet. 2004;75(1):106-11. doi:10.1086/422103. PMID: 15136966. PMC1181993.
81. Qin Y, Choi Y, Zhao H, Simpson JL, Chen ZJ, Rajkovic A. NOBOX homeobox mutation causes premature ovarian failure. Am J Hum Genet. 2007;81(3):576-81. doi:10.1086/519496. PMID: 17701902. PMC1950834.
82. Qin Y, Jiao X, Simpson JL, Chen ZJ. Genetics of primary ovarian insufficiency: new developments and opportunities. Hum Reprod Update. 2015;21(6):787-808. doi:10.1093/humupd/dmv036. PMID: 26243799. PMC4594617.
83. CRMERCD, Filière FIRENDO. Argumentaire: PNDS Insuffisance ovarienne prématurée. Saint-Denis La Plaine: Haute Autorité de Santé; 2021. Available at: https://www.has-sante.fr/jcms/p_3264990
84. Wood-Trageser MA, Gurbuz F, Yatsenko SA, Jeffries EP, Kotan LD, Surti U, et al. MCM9 mutations are associated with ovarian failure, short stature, and chromosomal instability. Am J Hum Genet. 2014;95(6):754-62. doi:10.1016/j.ajhg.2014.11.002. PMID: 25480036. PMC4259971.
85. AlAsiri S, Basit S, Wood-Trageser MA, Yatsenko SA, Jeffries EP, Surti U, et al. Exome sequencing reveals MCM8 mutation underlies ovarian failure and chromosomal instability. J Clin Invest. 2015;125(1):258-62. doi:10.1172/JCI78473. PMID: 25437880. PMC4382257.
86. Silvén H, Savukoski SM, Pesonen P, Pukkala E, Ojaniemi M, Gissler M, et al. Association of genetic disorders and congenital malformations with premature ovarian insufficiency: a nationwide register-based study. Hum Reprod. 2023;38(6):1224-1230. doi:10.1093/humrep/dead066. PMID: 37018629. PMC10233236.
87. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Turner syndrome (45,X) and variants. p. 696–699.
88. Rizzolio F, Bione S, Sala C, Goegan M, Gentile M, Gregato G, et al. Chromosomal rearrangements in Xq and premature ovarian failure: mapping of 25 new cases and review of the literature. Hum Reprod. 2006;21(6):1477-83. doi:10.1093/humrep/dei495. PMID: 16497693.
89. Genesio R, Mormile A, Licenziati MR, De Brasi D, Leone G, Balzano S, et al. Short stature and primary ovarian insufficiency possibly due to chromosomal position effect in a balanced X;1 translocation. Mol Cytogenet. 2015;8:50. doi:10.1186/s13039-015-0154-3. PMID: 26175800. PMC4501070.
90. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Fragile X syndrome (FRAX). p. 424–427.
91. Sullivan AK, Marcus M, Epstein MP, Allen EG, Anido AE, Paquin JJ, et al. Association of FMR1 repeat size with ovarian dysfunction. Hum Reprod. 2005;20(2):402-12. doi:10.1093/humrep/deh635. PMID: 15608041.
92. Fink DA, Nelson LM, Pyeritz R, Johnson J, Sherman SL, Cohen Y, et al. Fragile X Associated Primary Ovarian Insufficiency (FXPOI): Case Report and Literature Review. Front Genet. 2018;9:529. doi:10.3389/fgene.2018.00529. PMID: 30542367. PMC6278244.
93. Huang J, Zhang W, Liu Y, Liu Y, Wang J, Jiang H. Association between the FMR1 CGG repeat lengths and the severity of idiopathic primary ovarian insufficiency: a meta analysis. Artif Cells Nanomed Biotechnol. 2019;47(1):3116-3122. doi:10.1080/21691401.2019.1645153. PMID: 31352801.
94. Murray A, Schoemaker MJ, Bennett CE, Ennis S, Macpherson JN, Jones M, et al. Population-based estimates of the prevalence of FMR1 expansion mutations in women with early menopause and primary ovarian insufficiency. Genet Med. 2014;16(1):19-24. doi:10.1038/gim.2013.64. PMID: 23703681. PMC3914024.
95. Allen EG, Sullivan AK, Marcus M, Small C, Dominguez C, Epstein MP, et al. Examination of reproductive aging milestones among women who carry the FMR1 premutation. Hum Reprod. 2007;22(8):2142-52. doi:10.1093/humrep/dem148. PMID: 17588953.
96. Rossetti R, Ferrari I, Bestetti I, Moleri S, Brancati F, Petrone L, et al. Fundamental role of BMP15 in human ovarian folliculogenesis revealed by null and missense mutations associated with primary ovarian insufficiency. Hum Mutat. 2020;41(5):983-997. doi:10.1002/humu.23988. PMID: 31957178.
97. Mayer A, Fouquet B, Pugeat M, Misrahi M. BMP15 "knockout-like" effect in familial premature ovarian insufficiency with persistent ovarian reserve. Clin Genet. 2017;92(2):208-212. doi:10.1111/cge.12970. PMID: 28094433.
98. Cohen A, Rossetti R, Florsheim N, Samson AO, Renbaum P, Carbone E, et al. A Novel Homozygous BMP15 Mutation Causes Ovarian Dysgenesis and Primary Amenorrhea. J Endocr Soc. 2025;9(2):bvae221. doi:10.1210/jendso/bvae221. PMID: 39850788. PMC11756294.
99. França MM, Funari MFA, Nishi MY, Narcizo AM, Domenice S, Costa EMF, et al. Identification of the first homozygous 1-bp deletion in GDF9 gene leading to primary ovarian insufficiency by using targeted massively parallel sequencing. Clin Genet. 2018;93(2):408-411. doi:10.1111/cge.13156. PMID: 29044499.
100. Sassi A, Désir J, Duerinckx S, Soblet J, Van Dooren S, Bonduelle M, et al. Compound heterozygous null mutations of NOBOX in sisters with delayed puberty and primary amenorrhea. Mol Genet Genomic Med. 2021;9(10):e1776. doi:10.1002/mgg3.1776. PMID: 34480423. PMC8580073.
101. Bouilly J, Bachelot A, Broutin I, Touraine P, Binart N. Novel NOBOX loss-of-function mutations account for 6.2% of cases in a large primary ovarian insufficiency cohort. Hum Mutat. 2011;32(10):1108-13. doi:10.1002/humu.21543. PMID: 21837770.
102. Li L, Wang B, Zhang W, Chen B, Luo M, Wang J, et al. A homozygous NOBOX truncating variant causes defective transcriptional activation and leads to primary ovarian insufficiency. Hum Reprod. 2017;32(1):248-255. doi:10.1093/humrep/dew271. PMID: 27836978.
103. Tosh D, Rani HS, Murty US, Deenadayal A, Grover P. Mutational analysis of the FIGLA gene in women with idiopathic premature ovarian failure. Menopause. 2015;22(5):520-6. doi:10.1097/GME.0000000000000340. PMID: 25314148.
104. Valkna A, Kikas T, Jakovlev Ü, Mõttus O, Dutta A, Erlang K, et al. All-in-one exome sequencing approach for genetic testing of unexplained premature ovarian insufficiency. Hum Reprod Open. 2026;2026(3):hoag058. doi:10.1093/hropen/hoag058. PMID: 42465592.
105. Pavlova N, Tsabai P, Kumykova Z, Latypova N, Averkova V, Kochetkova T, et al. STAG3 c.2776C > T, p.Arg926Ter is a recurrent variant in Russian patients with early-onset premature ovarian insufficiency. J Assist Reprod Genet. 2026;43(8):2667-2678. PMID: 42455414.
106. van der Bijl N, Röpke A, Biswas U, Wöste M, Jessberger R, Kliesch S, et al. Mutations in the stromal antigen 3 (STAG3) gene cause male infertility due to meiotic arrest. Hum Reprod. 2019;34(11):2112-2119. doi:10.1093/humrep/dez204. PMID: 31682730.
107. Al-Agha AE, Ahmed IA, Nuebel E, Moriwaki M, Moore B, Peacock KA, et al. Primary Ovarian Insufficiency and Azoospermia in Carriers of a Homozygous PSMC3IP Stop Gain Mutation. J Clin Endocrinol Metab. 2018;103(2):555-563. doi:10.1210/jc.2017-01966. PMID: 29240891. PMC5800840.
108. Tenenbaum-Rakover Y, Weinberg-Shukron A, Renbaum P, Lobel O, Eideh H, Gulsuner S, et al. Minichromosome maintenance complex component 8 (MCM8) gene mutations result in primary gonadal failure. J Med Genet. 2015;52(6):391-9. doi:10.1136/jmedgenet-2014-102921. PMID: 25873734.
109. Helderman NC, Yang T, Palles C, Terlouw D, Mei H, Vorderman RHP, et al. Clinical syndromes linked to biallelic germline variants in MCM8 and MCM9. HGG Adv. 2025;6(4):100480. doi:10.1016/j.xhgg.2025.100480. PMID: 40684266. PMC12361757.
110. Desai S, Wood-Trageser M, Matic J, Chipkin J, Jiang H, Bachelot A, et al. MCM8 and MCM9 Nucleotide Variants in Women With Primary Ovarian Insufficiency. J Clin Endocrinol Metab. 2017;102(2):576-582. doi:10.1210/jc.2016-2565. PMID: 27802094. PMC5413161.
111. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Ptosis, blepharophimosis and eyelid anomalies. p. 292–295.
112. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Blepharophimosis-ptosis-epicanthus inversus syndrome. p. 326–327.
113. Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. Galactosemia. p. 301–302.
114. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 14, section 14.1 Galactose-1-phosphate uridylyltransferase (GALT) deficiency.
115. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-68. doi:10.1002/humu.23348. Accessed 2 and 6 November 2026: https://variantvalidator.org
116. Bouilly J, Beau I, Barraud S, Bernard V, Azibi K, Fagart J, et al. Identification of Multiple Gene Mutations Accounts for a new Genetic Architecture of Primary Ovarian Insufficiency. J Clin Endocrinol Metab. 2016;101(12):4541-4550. doi:10.1210/jc.2016-2152. PMID: 27603904.
117. Rossetti R, Moleri S, Guizzardi F, Gentilini D, Libera L, Marozzi A, et al. Targeted Next-Generation Sequencing Indicates a Frequent Oligogenic Involvement in Primary Ovarian Insufficiency Onset. Front Endocrinol (Lausanne). 2021;12:664645. doi:10.3389/fendo.2021.664645. PMID: 34803902. PMC8600266.
118. Pan Y, Yu Y, Mo J, Ren S, Zhou Z, Yang X, et al. Genetic variants in HELB contribute to premature ovarian insufficiency and early age of natural menopause. JCI Insight. 2025;10(21):e191122. doi:10.1172/jci.insight.191122. PMID: 41212051.
119. Fukutomi S, Hernandez EJ, Allen-Brady K, Marchetti M, Johnstone EB, Verrilli LE, et al. Chromosomal and genetic causes of primary ovarian insufficiency in the Utah population. J Endocr Soc. 2026;10(11):bvag233. doi:10.1210/jendso/bvag233. PMID: 42891080.
120. Bachelot A, Rouxel A, Massin N, Dulon J, Courtillot C, Matuchansky C, et al. Phenotyping and genetic studies of 357 consecutive patients presenting with premature ovarian failure. Eur J Endocrinol. 2009;161(1):179-87. doi:10.1530/EJE-09-0231. PMID: 19411303.
121. Jiao X, Zhang H, Ke H, Zhang J, Cheng L, Liu Y, et al. Premature Ovarian Insufficiency: Phenotypic Characterization Within Different Etiologies. J Clin Endocrinol Metab. 2017;102(7):2281-2290. doi:10.1210/jc.2016-3960. PMID: 28368522.
122. Bachelot A, Nicolas C, Gricourt S, Dulon J, Leban M, Golmard JL, et al. Poor Compliance to Hormone Therapy and Decreased Bone Mineral Density in Women with Premature Ovarian Insufficiency. PLoS One. 2016;11(12):e0164638. doi:10.1371/journal.pone.0164638. PMID: 27906970. PMC5132216.
123. Costa GPO, Ferreira-Filho ES, Simoes RDS, Soares-Junior JM, Baracat EC, Maciel GAR. Impact of hormone therapy on the bone density of women with premature ovarian insufficiency: A systematic review. Maturitas. 2023;167:105-112. doi:10.1016/j.maturitas.2022.09.011. PMID: 36368093.
124. Hipp HS, Charen KH, Spencer JB, Allen EG, Sherman SL. Reproductive and gynecologic care of women with fragile X primary ovarian insufficiency (FXPOI). Menopause. 2016;23(9):993-9. doi:10.1097/GME.0000000000000658. PMID: 27552334. PMC4998843.
125. Waggoner DD, Buist NR, Donnell GN. Long-term prognosis in galactosaemia: results of a survey of 350 cases. J Inherit Metab Dis. 1990;13(6):802-18. doi:10.1007/BF01800204. PMID: 1706789.
126. Guerrero NV, Singh RH, Manatunga A, Berry GT, Steiner RD, Elsas LJ 2nd. Risk factors for premature ovarian failure in females with galactosemia. J Pediatr. 2000;137(6):833-41. doi:10.1067/mpd.2000.109148. PMID: 11113841.
127. Ahonen P, Myllärniemi S, Sipilä I, Perheentupa J. Clinical variation of autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED) in a series of 68 patients. N Engl J Med. 1990;322(26):1829-36. doi:10.1056/NEJM199006283222601. PMID: 2348835.
128. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Leukodystrophy/leukoencephalopathy. p. 208–211.
129. Fogli A, Rodriguez D, Eymard-Pierre E, Bouhour F, Labauge P, Meaney BF, et al. Ovarian failure related to eukaryotic initiation factor 2B mutations. Am J Hum Genet. 2003;72(6):1544-50. doi:10.1086/375404. PMID: 12707859. PMC1180314.
130. Hamilton EMC, van der Lei HDW, Vermeulen G, Gerver JAM, Lourenço CM, Naidu S, et al. Natural History of Vanishing White Matter. Ann Neurol. 2018;84(2):274-288. doi:10.1002/ana.25287. PMID: 30014503. PMC6175238.
131. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 39, sections 39.1 Nucleotide and nucleic acid metabolism and 39.2 tRNA processing disorders.
132. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Mitochondrial DNA diseases. p. 490–495.
133. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 10, Disorders of oxidative phosphorylation, section 10.1 Clinical presentation.
134. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 42, section 42.2 Disorders of peroxisomal fatty acid β-oxidation.
135. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Microcephalic primordial dwarfing syndromes. p. 128–131.
136. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Female infertility and amenorrhoea. p. 722–723.
137. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Intellectual disability. p. 188–193.
138. European Society for Human Reproduction and Embryology (ESHRE) Guideline Group on POI; Webber L, Davies M, Anderson R, Bartlett J, Braat D, et al. ESHRE Guideline: management of women with premature ovarian insufficiency. Hum Reprod. 2016;31(5):926-37. doi:10.1093/humrep/dew027. PMID: 27008889.
139. Plan France Médecine Génomique 2025. Pré-indication: insuffisance ovarienne primitive [Internet]. Paris: Aviesan; 2019. Available at: https://pfmg2025.aviesan.fr/professionnels/pre-indications-et-mise-en-place/insuffisance-ovarienne-primitive/
140. Christin-Maitre S, Givony M, Albarel F, Bachelot A, Bidet M, Blanc JV, et al. Position statement on the diagnosis and management of premature/primary ovarian insufficiency (except Turner Syndrome). Ann Endocrinol (Paris). 2021;82(6):555-571. doi:10.1016/j.ando.2021.09.001. PMID: 34508691.
141. Spector E, Behlmann A, Kronquist K, Rose NC, Lyon E, Reddi HV, et al. Laboratory testing for fragile X, 2021 revision: a technical standard of the American College of Medical Genetics and Genomics (ACMG). Genet Med. 2021;23(5):799-812. doi:10.1038/s41436-021-01115-y. PMID: 33795824.
142. Jolly A, Bayram Y, Turan S, Aycan Z, Tos T, Abali ZY, et al. Exome Sequencing of a Primary Ovarian Insufficiency Cohort Reveals Common Molecular Etiologies for a Spectrum of Disease. J Clin Endocrinol Metab. 2019;104(8):3049-3067. doi:10.1210/jc.2019-00248. PMID: 31042289. PMC6563799.
143. Zhou F, Zhang J, Li Y, Liu S, Wang X. Performance of exome sequencing in premature ovarian insufficiency: a systematic review and meta-analysis. BMC Pregnancy Childbirth. 2026;26(1):714. PMID: 42121113.
144. Sassi A, Peyrassol X, Marangoni M, Gheldof A, Van Dooren S, Alvaro Mercadal B, et al. Exploring genetic variants in premature ovarian insufficiency: a cohort-based investigation of monogenic etiologies. J Endocr Soc. 2026;10(11):bvag215. doi:10.1210/jendso/bvag215. PMID: 42953442.
145. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Haemochromatosis. p. 430–435.
146. Nordenström A, Ahmed SF, van den Akker E, Blair J, Bonomi M, Brachet C, et al. Pubertal induction and transition to adult sex hormone replacement in patients with congenital pituitary or gonadal reproductive hormone deficiency: an Endo-ERN clinical practice guideline. Eur J Endocrinol. 2022;186(6):G9-G49. doi:10.1530/EJE-22-0073. PMID: 35353710. PMC9066594.
147. ClinicalTrials.gov. Hormone replacement therapy in adolescents with premature ovarian insufficiency. NCT06851754. Bethesda (MD): National Library of Medicine; updated 29 October 2026. Available at: https://clinicaltrials.gov/study/NCT06851754
148. INFARMED. Utrogestan 100 mg (progesterona) cápsula mole. Resumo das Características do Medicamento. Lisbon: INFARMED; accessed 2 November 2026. Available at: https://extranet.infarmed.pt/INFOMED-fo/
149. U.S. Food and Drug Administration. Vivelle-Dot (estradiol transdermal system). Prescribing information. NDA 020538. Revised 17 November 2023. Available at: https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=020538
150. U.S. Food and Drug Administration. Estradiol tablets USP (ANDA 040275). Prescribing information. Available at: https://dailymed.nlm.nih.gov/dailymed/search.cfm?query=estradiol+tablets
151. U.S. Food and Drug Administration. Prometrium (progesterone, USP) capsules. Prescribing information. NDA 019781. Revised 23 July 2026. Available at: https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=019781
152. U.S. Food and Drug Administration. Divigel (estradiol gel) 0.1%. Prescribing information. NDA 022038. Revised 27 February 2026. Available at: https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=022038
153. INFARMED. Dermestril 25, 50 e 100 (estradiol) sistema transdérmico. Resumo das Características do Medicamento. Lisbon: INFARMED; accessed 2 November 2026. Available at: https://extranet.infarmed.pt/INFOMED-fo/
154. INFARMED. Lenzetto 1,53 mg/dose (estradiol) solução para pulverização transdérmica. Resumo das Características do Medicamento. Lisbon: INFARMED; accessed 2 November 2026. Available at: https://extranet.infarmed.pt/INFOMED-fo/
155. INFARMED. Duphaston 10 mg (didrogesterona) comprimido revestido por película. Resumo das Características do Medicamento. Lisbon: INFARMED; accessed 2 November 2026. Available at: https://extranet.infarmed.pt/INFOMED-fo/
156. Agência Nacional de Vigilância Sanitária (ANVISA). Open data: registered medicinal products (DADOS_ABERTOS_MEDICAMENTOS.csv), filtered for estradiol, progesterone and dydrogesterone with active registration. Brasília: ANVISA; accessed 6 November 2026. Available at: https://dados.anvisa.gov.br/dados/DADOS_ABERTOS_MEDICAMENTOS.csv
157. Angelini E, Oktay KH. Fertility preservation in patients with genetic conditions predisposing to primary ovarian insufficiency. Fertil Steril. 2026. PMID: 42869368.
158. van Bentem K, van der Hoorn ML, Banker M, de la Calle M, Elenis E, El Demellawy D, et al. The risk for the development of hypertensive complications in oocyte donation pregnancy: a systematic review and individual participant data meta-analysis (DONOR IPD). Hum Reprod Update. 2026;32(4):488-508. doi:10.1093/humupd/dmag006. PMID: 41915703.
159. Bertolazzi M, Genta ML, Lopes A, Fernandes R, Maciel G, Lopez R, et al. Randomized trial of autologous ovarian graft to preserve ovarian function in young women undergoing pelvic radiotherapy for cervical cancer. Eur J Surg Oncol. 2026;52(4):111496. doi:10.1016/j.ejso.2026.111496. PMID: 41759360.
160. ClinicalTrials.gov. Immunomodulatory therapy to restore ovarian function and improve fertility in women with autoimmune premature ovarian insufficiency. NCT07509840. Updated 3 April 2026. Available at: https://clinicaltrials.gov/study/NCT07509840
161. ClinicalTrials.gov. Phase I/IIa trial evaluating intravenous hUC-MSCs injection for safety, tolerability and efficacy in patients with POI. NCT07775391. Updated 20 August 2026. Available at: https://clinicaltrials.gov/study/NCT07775391
162. ClinicalTrials.gov. Finerenone for premature ovarian insufficiency (POI). NCT07815509. Updated 11 September 2026. Available at: https://clinicaltrials.gov/study/NCT07815509
163. ClinicalTrials.gov. Oral GDC-0077 in premature ovarian insufficiency. NCT07904208. Updated 29 October 2026. Available at: https://clinicaltrials.gov/study/NCT07904208
164. ClinicalTrials.gov. Investigating reactivation of ovarian function following autologous PRP intra-ovarian infusion in POI patients. NCT04031456. Updated 8 July 2026. Available at: https://clinicaltrials.gov/study/NCT04031456
165. ClinicalTrials.gov. Does subcutaneous granulocyte colony stimulating factor (G-CSF) improve ovarian reserve in women with premature ovarian insufficiency? NCT06117982. Updated 20 August 2026. Available at: https://clinicaltrials.gov/study/NCT06117982
166. ClinicalTrials.gov. Long term follow up patients with premature ovarian failure ex vivo gene therapy using autologous mesenchymal stem cell and mesenchymal stem cell lyophilisate. NCT04675970. Updated 22 December 2025. Available at: https://clinicaltrials.gov/study/NCT04675970
167. Fang Y, Liu C, He Z, Zheng Y, Zhu L, Ye Z. Efficacy of intraovarian autologous platelet-rich plasma in women with poor ovarian reserve or ovarian insufficiency: a meta-analysis of randomized controlled trials. Front Endocrinol (Lausanne). 2026;17:1827041. PMID: 42205258.
168. Kawamura K, Cheng Y, Suzuki N, Deguchi M, Sato Y, Takae S, et al. Hippo signaling disruption and Akt stimulation of ovarian follicles for infertility treatment. Proc Natl Acad Sci U S A. 2013;110(43):17474-9. doi:10.1073/pnas.1312830110. PMID: 24082083. PMC3808580.
169. ClinicalTrials.gov. In vitro activation of dormant follicles for patients with primary ovarian insufficiency. NCT02322060. Updated 27 May 2026. Available at: https://clinicaltrials.gov/study/NCT02322060
170. Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. Síndrome de Turner. p. 28–31.
171. Ruan X, Xu C, Huang H, Xu B, Yin C, Wu Y, et al. Practice guideline for the treatment and management of iatrogenic premature ovarian insufficiency. Maturitas. 2026;205:108807. doi:10.1016/j.maturitas.2025.108807. PMID: 41447991.
172. Jiang M, Aittomäki K, Nilsson C, Pakarinen P, Iitiä A, Torresani T, et al. The frequency of an inactivating point mutation (566C-->T) of the human follicle-stimulating hormone receptor gene in four populations using allele-specific hybridization and time-resolved fluorometry. J Clin Endocrinol Metab. 1998;83(12):4338-43. doi:10.1210/jcem.83.12.5306. PMID: 9851774.
173. Filière FIRENDO. Synthèse à destination du médecin traitant: Insuffisance ovarienne prématurée. Saint-Denis La Plaine: Haute Autorité de Santé; May 2021. Available at: https://www.has-sante.fr/jcms/p_3264990
174. Persani L, Cools M, Ioakim S, Faisal Ahmed S, Andonova S, Avbelj-Stefanija M, et al. The genetic diagnosis of rare endocrine disorders of sex development and maturation: a survey among Endo-ERN centres. Endocr Connect. 2022;11(12):. doi:10.1530/EC-22-0367. PMID: 36228316. PMC9716404.