Persistent Müllerian Duct Syndrome

A 46,XY difference of sex development in which the uterus, fallopian tubes and upper vagina persist in an individual with normally virilised male external genitalia, owing to deficiency of anti-Müllerian hormone (type I, AMH), insensitivity of its type II receptor (type II, AMHR2) or, in a minority, variants in PPP1R12A or a still unknown cause.

Literature and regulatory information updated: 26 October 2026.

0. Initial Reference Table

1. Introduction

1.1. Definition and place among the differences of sex development

Persistent Müllerian duct syndrome (PMDS) is defined by the presence of Müllerian derivatives, that is, uterus, fallopian tubes and upper vagina, in an individual with a 46,XY karyotype and normally virilised male external genitalia [9, 12]. The external phenotype is completely male, with the urethra opening at the tip of the glans and no hypospadias; it is this absence of genital ambiguity that separates PMDS from mixed gonadal dysgenesis and from the other differences of sex development (DSD) in which Müllerian derivatives and undervirilisation coexist [9]. Assignment of male sex at birth raises no doubt, even when one or both testes are impalpable, and the Müllerian derivatives are usually discovered unexpectedly during surgery for cryptorchidism or inguinal hernia [12].

Male differentiation of the genital tract depends on two fetal testicular hormones, according to the model established by Alfred Jost in 1953: testosterone, produced by Leydig cells, stabilises the Wolffian ducts and virilises the urogenital sinus and external genitalia; anti-Müllerian hormone (AMH, also called Müllerian inhibiting substance or factor, MIS or MIF), produced by Sertoli cells, causes regression of the Müllerian ducts [9, 18]. Until 7 weeks of gestation the gonads of both sexes are histologically identical and both pairs of ducts are present; the fetal testis produces AMH from the end of the 7th week and the Müllerian ducts have completely disappeared by 10 weeks in the normal male fetus [9, 12, 19]. In PMDS only the AMH arm fails, so that androgenic virilisation is normal and the Müllerian derivatives persist [12].

1.2. Historical milestones

The first case is often attributed to Nilson, who in 1939 described hernia uteri inguinalis in a man, but there are earlier publications, by Jordan (1895) and Webster (1906), and von Seemen observed parental consanguinity in a 1927 case [2, 9]. The typical clinical description recorded in OMIM is that of a man with bilateral cryptorchidism and inguinal hernias, normal male external genitalia and a uterus and fallopian tubes in the inguinal canal discovered at herniorrhaphy, with testicular gonads [2]. Reports of affected brothers (Guell-Gonzalez 1971; Morillo-Cucci and German 1971; Armendares 1973) and parental consanguinity supported the hypothesis of autosomal recessive inheritance with expression limited to males [2].

The molecular phase began with the cloning of the complementary DNA and the AMH gene in 1986 (Cate and colleagues; Picard and colleagues) and with the mapping of the AMH gene to 19p13.3 by Cohen-Haguenauer and colleagues in 1987 [9, 18]. Guerrier and colleagues showed in 1989 that not all cases result from a defect of the AMH gene, because some patients have normal amounts of bioactive testicular AMH, which led to the hypothesis of end-organ resistance [18]. The first pathogenic AMH variant, a nonsense variant in exon 5 called “AMH Bruxelles”, was identified in three brothers from a Moroccan family (Knebelmann and colleagues, 1990–1991) [9, 12, 18]. The complementary DNA and the gene for the AMH type II receptor (AMHR2) were cloned in 1994 (Baarends and colleagues; di Clemente and colleagues) and, in 1995, Imbeaud and colleagues described the first AMHR2 variant in a 3-month-old boy of Pakistani ancestry, whose testicular tissue induced regression of fetal rat Müllerian ducts in co-culture, direct proof that AMH was being produced [2, 9, 20]. In 1996 the same group showed, in 38 families, that the type of genetic defect could be predicted by serum AMH before puberty and that a 27 base-pair (bp) deletion in exon 10 of AMHR2 was the single most frequent cause [21].

The experience accumulated by the Paris group (Josso, Picard and colleagues) between 1990 and 2016, with 157 families, constitutes the reference series for the syndrome; variants in AMH or AMHR2 were identified in 88% of families [9, 22]. In 2020, Hughes and colleagues described loss-of-function variants of PPP1R12A in patients with genitourinary and/or brain malformations, and in 2022 Picard and colleagues identified truncating variants of this gene in five patients with PMDS without a biallelic variant in AMH or AMHR2, the first demonstration of a genetic cause outside the AMH signalling pathway [5, 23].

1.3. Nomenclature and classifications

OMIM brings the two classical monogenic forms together under a single phenotype number (261550, Persistent Mullerian duct syndrome, types I and II): type I results from AMH variants and type II from AMHR2 variants [2, 20]. UniProt designates the same entities PMDS1 and PMDS2 [4]. In the genetic classification of DSD of the European consortium COST Action BM1303 (DSDnet), both belong to the category “46,XY DSD with abnormal AMH secretion or action” [3]. In the aetiopathogenic classification of Y-chromosome DSD based on serum AMH and testosterone, type I corresponds to isolated Sertoli cell dysfunction (low or undetectable AMH, normal testosterone) and type II to failure of the AMH target organ (normal AMH and testosterone); Josso and colleagues place the idiopathic forms with normal AMH, frequently associated with other malformations, among the malformative DSD [12].

The anatomical classification used in surgical practice distinguishes three presentations: the “female” form, with both testes intra-abdominal in an ovarian position, attached to the fallopian tubes; hernia uteri inguinalis or the “male” form, with one descended testis and the uterus and one fallopian tube in the hernial sac, and the contralateral testis undescended; and transverse testicular ectopia, in which both testes lie in the same hemiscrotum or the same hernial sac, with uterus and fallopian tubes [9, 13]. The historical terminology of “internal male pseudohermaphroditism”, still listed as an alternative title in OMIM, has been replaced by DSD nomenclature since the 2006 consensus, revised in 2016 [2, 24].

1.4. Scope of the article

This article covers PMDS type I and type II, PMDS without an identified cause and PPP1R12A-associated PMDS, with genitourinary and/or brain malformation syndrome (GUBS) as a syndromic form and differential. The other 46,XY DSD enter the differential diagnosis: androgen insensitivity, 5-alpha-reductase type 2 deficiency, 46,XY gonadal dysgenesis, campomelic dysplasia, Mayer-Rokitansky-Küster-Hauser syndrome and Müllerian aplasias, and Kallmann syndrome and congenital hypogonadotrophic hypogonadism.

2. Molecular Genetics and Pathophysiology

2.1. Anti-Müllerian hormone and regression of the Müllerian ducts

AMH is a homodimeric glycoprotein of the transforming growth factor β (TGF-β) superfamily, secreted mainly by fetal and prepubertal Sertoli cells and, in smaller amounts, by the granulosa cells of small ovarian follicles [12, 25]. It is synthesised as a 560-amino-acid precursor with a large N-terminal pro-region and a small C-terminal mature domain, which concentrates the homology with the TGF-β family and the bioactivity [4, 9, 18]. Cleavage at arginine 451 generates N-terminal dimers of 110 kDa and C-terminal dimers of 25 kDa, which remain associated in a bioactive non-covalent complex; cleavage is obligatory, whereas dissociation of the complex occurs on receptor binding [9, 12]. Without the N-terminal fragment, the C-terminal domain does not fold correctly and is degraded before secretion, which explains the pathogenicity of variants located in the pro-region [9].

AMH expression begins with the differentiation of the seminiferous cords, at the end of the 7th embryonic week, and remains high until puberty, with a transient perinatal decline [12]. It is triggered by the transcription factor SOX9, which binds to the promoter, and reinforced by SF1 (NR5A1), GATA4 and WT1; it is independent of gonadotrophins at onset and at basal level, follicle-stimulating hormone (FSH) stimulates it through Sertoli cell proliferation and the cyclic AMP/protein kinase A pathway, and at puberty it is inhibited by intratesticular testosterone and the onset of meiosis [12]. In the fetal and neonatal testis androgens do not inhibit AMH because Sertoli cells do not yet express the androgen receptor [12]. In murine models, mutation of the Sox9 binding site in the promoter abolishes the onset of transcription and causes Müllerian persistence, whereas mutation of the Sf1 binding site reduces the amount of transcript without preventing regression [18].

AMH binds to its specific type II receptor, AMHR2, a transmembrane serine/threonine kinase; the complex recruits a type I receptor (ALK2/ACVR1, ALK3/BMPR1A or ALK6/BMPR1B), which phosphorylates the R-SMAD effectors 1, 5 or 8; these associate with SMAD4 and pass into the nucleus [9, 12, 25]. In the mouse, ALK2 and ALK3 act redundantly in Müllerian regression, whereas inactivation of ALK6 does not affect it [12]. The type I receptors and SMADs are shared with the bone morphogenetic proteins (BMP) and are indispensable for early embryonic development, so loss-of-function variants in these components are not compatible with isolated PMDS; the only component of the pathway exclusive to AMH is AMHR2 [5, 9]. Mice with Amhr2 inactivation are internal pseudohermaphrodites, and Amh/Amhr2 double mutants have a phenotype indistinguishable from that of single mutants, which indicates that AMH is the only ligand of AMHR2 [18, 20].

In the tissue, regression depends on cells expressing AMHR2 in the coelomic epithelium and the periductal mesenchyme: under the action of AMH they undergo epithelial–mesenchymal transition and migrate to surround the Müllerian mesoepithelium; after rupture of the basement membrane, the epithelial cells reorient and enter the mesenchymal compartment [5]. Wnt pathways take part in this process: Wnt4 is required for Müllerian duct development, Wnt7a induces Amhr2 expression in the Müllerian mesenchyme and, in mice lacking β-catenin, there is no regression [9]. In normal anatomy, disappearance of the Müllerian ducts frees the testes from their initial pelvic position; when regression fails, the testes remain suspended in the broad ligament, whose attachment to the pelvis is often weak or absent, so that they may descend dragging the uterus and fallopian tubes with them [9].

2.2. The AMH gene and PMDS type I

The AMH gene has 5 exons spanning about 2.8 kb on the short arm of chromosome 19 (19p13.3); the 3' end of the last exon is very GC-rich and encodes the bioactive C-terminal domain [9, 12, 18]. The MANE Select transcript is NM_000479.5 (protein NP_000470.3); the tables of the Paris series use NM_000479.3, with identical c. coordinates for the variants normalised in VariantValidator [7]. Gene–disease validity is classified as Strong in GenCC (Labcorp Genetics and PanelApp Australia) and Supportive in Orphanet, all with autosomal recessive inheritance; ClinGen has no gene–disease validity curation and assigns haploinsufficiency score 30, used for genes with a recessive phenotype [25].

Knowledge of the variant spectrum has accumulated over time. By May 2013, the Paris laboratory had identified 54 different alleles in 65 families, with all exons affected except exon 4 and 63% homozygosity [12]. By 2016–2017 there were 80 families (68 from the Paris group and 12 from other groups) and 64 different pathogenic alleles: 37 missense, 10 nonsense, 1 stop-loss, 9 deletions (including one indel), 2 insertions and 5 splicing [9, 22]. The summary published in 2019 updates the figures to 81 families and 65 variants, concentrated in exons 1, 2 and 5 [9]. The variants are distributed throughout the gene, rarely in exon 4; exons 1 and 2 are, proportionally, the most affected, and in exon 5 the bases encoding the C-terminal domain are affected almost three times more often than those encoding the N-terminal part [9]. Imbeaud and colleagues (1996) had identified exon 1 and the 3' half of exon 5 as the main sites of deleterious variants, including small deletions and missense variants [2]. There are no significant hotspots, but 6 variants occur in four or more families and 13 less frequently [9].

Most AMH variants destabilise the protein, so serum AMH is very low or undetectable, and this also happens with variants that do not affect the C-terminal domain [12, 25]. The expression studies of Belville and colleagues (2004) showed that proteins lacking the C-terminal region are secreted faster, that single substitutions in either domain alter stability or folding and that the addition or removal of cysteines is particularly damaging [9]. Three loss-of-function mechanisms have been described: defective folding of the C-terminal domain with degradation before secretion, truncated protein without a C-terminal domain and hormone that is secreted but unable to bind the receptor [11]. The last mechanism is represented by the variant p.(Gln496His), associated with normal serum AMH and presumed to impair receptor binding [9, 12].

Some alleles illustrate technical limits and unusual mechanisms. A deletion in the promoter, 225 bp upstream of the initiation codon, removes the fourth base of the distal SF1 response element (located at −228 bp and indispensable for human expression, unlike the proximal one at −102 bp) [9]. The variant NM_000479.5:c.3G>T p.(Met1?) affects the initiation codon in a Moroccan patient; the variant c.1683A>T, p.(Ter561CysextTer?), removes the stop codon without another in frame before the polyadenylation site; the recurrent c.35T>G p.(Val12Gly) affects the signal peptide [7, 9]. In a patient homozygous for NM_000479.5:c.941C>G p.(Ala314Gly) and c.1598G>T p.(Gly533Val), with very low serum AMH, glycine 533 lies in the conserved β6 strand of the mature domain [7, 9]. In another patient, AMH messenger RNA obtained by “illegitimate” transcription from uterine tissue showed retention of the first 46 bases of intron 1 and use of a cryptic donor site, demonstrating the effect of an intronic variant [9]. In 6 families only one abnormal allele was identified; the second may have been missed because it lies in deep intronic regions or because of technical difficulty in a very GC-rich gene [9].

The polymorphism published as c.146T>G p.(Ile49Ser) is present in 21% of normal individuals and corresponds to the bovine sequence; Picard and Josso do not expect an effect on bioactivity [9]. The reference sequence NM_000479.5 already contains the G allele (Ser49), so that, relative to the MANE transcript, the same position is described as NM_000479.5:c.146G>T p.(Ser49Ile) [7]. Kevenaar and colleagues reported, however, reduced bioactivity of the 49Ser protein in vitro and, in women with polycystic ovary syndrome, an association with a less severe phenotype, as well as higher follicular-phase oestradiol in women without disease; these data concern ovarian physiology and not PMDS [9, 18].

2.3. The AMHR2 gene and PMDS type II

The AMHR2 gene has 11 exons spanning about 8 kb at 12q13.13; exons 1 to 3 encode the signal peptide and the extracellular domain, exon 4 most of the transmembrane domain and exons 5 to 11 the intracellular serine/threonine kinase domain [9, 20]. The protein has 573 amino acids: 17 of signal sequence, 127 of extracellular domain, 26 of transmembrane domain and 403 of intracellular domain [4, 20]. The MANE Select transcript is NM_020547.3 (NP_065434.1); the Paris series used NM_020547.2, with identical c. coordinates for the normalised variants [7].

By 2013 there were 49 abnormal alleles in 59 families in the Paris laboratory and one family in Boston [12]. In the 2017 update the total was 75 families (68 from the Paris group and others reported in the United States, Brazil, Egypt, Portugal, Turkey and Israel), with 58 different alleles: 35 missense, 11 nonsense, 8 deletions and 4 splicing, distributed across the 11 exons [9, 22]. The 2013 statement that exon 4 was not affected was revised in 2017 with the observation of variants in all exons [9, 12]. The summary published in 2019 gives 79 families and 64 alleles [9]. Two variants affect the initiation codon; the next methionine is at position 76 [9].

The most frequent recurrent variant is the 27 bp deletion in exon 10, NM_020547.3:c.1332_1358del p.(Gly445_Leu453del) (NC_000012.12:g.53430189_53430215del), historically described as “d6331–6357” in a genomic numbering from the transcription start and as p.(Leu444_Glu452del) by Imbeaud and colleagues in 1996 [7, 9, 12]. The variant removes nine amino acids from the αG helix and the preceding loop, at the top of the C lobe of the kinase domain, and molecular modelling predicts a substantial conformational change [9, 20]. In the 1996 series it was present on at least one allele in 10 of 16 families with a receptor variant, homozygous in 4 patients and combined with a missense variant in the others (6 patients according to the OMIM gene entry, 7 according to the syndrome entry), corresponding to about 25% of all patients with PMDS analysed [2, 20, 21]. In 2013 the deletion was present in 21% of all families with PMDS and in 44% of families with a receptor variant; by 2017 it had been observed in 30 patients, 67% of whom were of Northern European origin [9, 12]. ClinVar classifies it as pathogenic, with criteria provided by multiple submitters and no conflicts (VCV000008627, last evaluated April 2026) [26]. The second recurrent variant is c.1219C>T p.(Arg407Ter), historically R407Stop, in exon 9, observed in 6 families [9, 12, 26].

The functional studies of Belville and colleagues (2009) showed that extracellular-domain variants that truncate the protein before the transmembrane domain are not secreted, because the AMHR2 signal sequence is defective and the receptor uses the transmembrane domain to translocate to the endoplasmic reticulum, in the manner of type III membrane proteins [9, 20]. Eight of the ten mutant receptors studied were still expressed at the cell surface [20]. The variant NM_020547.3:c.596del p.(Glu199GlyfsTer10) (formerly described as 1692delA) removes the entire kinase domain; c.1217G>A p.(Arg406Gln), c.1471G>C p.(Asp491His), c.1510C>T p.(Arg504Cys) and c.1511G>A p.(Arg504His) affect residues conserved in all type II receptors of the TGF-β family [7, 9, 20]. Truncated mutant receptors have a dominant-negative effect when overexpressed in cell lines, but in vivo only biallelic carriers are affected, because the mutant and normal alleles are expressed in a 1:1 ratio (Messika-Zeitoun and colleagues, 2001) [9, 20]. Missense variants of the extracellular domain cannot be assessed by three-dimensional modelling, because identity with known structures is below 20% [9].

A patient apparently homozygous for c.532C>T p.(Arg178Ter), with a heterozygous father and a mother carrying only the normal allele, was in fact hemizygous: the mother had a large deletion of all or part of AMHR2 [9]. The first AMHR2 variant described, in the Pakistani patient of 1995, affects the +1 position of the intron 2 donor site (NC_000012.12:g.53424471G>A) and produces two abnormal messenger RNAs, one with in-frame skipping of exon 2 and the other with retention of 12 bp of intron 2 through use of a cryptic donor, with a p.Gly78Asp substitution followed by insertion of four residues [9, 20]. The promoter polymorphism −482A>G is associated with follicular oestradiol and age at menopause in women, with no relationship to PMDS [9, 18].

2.4. Normalised recurrent and founder variants

Table sources: [7, 9, 12, 18, 26, 27]. The correspondence between the historical designation E358X and c.1144G>T derives from the supplementary table of the Paris series, which assigns the Knebelmann family to this allele; the HGVS description was validated in VariantValidator [7, 9, 18].

2.5. PPP1R12A and syndromic PMDS

PPP1R12A (12q21.2–q21.31; 25 exons spanning about 162 kb, with several isoforms through alternative splicing) encodes MYPT1, the 1030-amino-acid regulatory subunit of myosin phosphatase, which with the catalytic subunit PP1c and the small subunit M20/21 forms the heterotrimeric phosphatase that relaxes the myosin II light chain in smooth muscle and in non-muscle tissues [5]. The gene is ubiquitously expressed, including brain, small intestine, endometrium and fallopian tubes, and in the human and murine fetus in the prosencephalic neural folds and the lower urinary tract at critical periods of development [5]. It is intolerant to loss of function in gnomAD (Genome Aggregation Database), with a probability of loss-of-function intolerance (pLI) of 1 and an upper bound of the confidence interval of the observed/expected ratio (LOEUF) of 0.2 [8].

Hughes and colleagues (2020) identified heterozygous de novo variants with a premature stop codon in 12 unrelated individuals with genitourinary and/or brain malformations, distributed across several exons and one intron, without genotype–phenotype correlation [6, 23]. Picard and colleagues (2022) reanalysed 39 patients from the Paris series (samples from 1990 onwards) without biallelic AMH or AMHR2 variants on Sanger sequencing: after confirming the absence of biallelic variants in 24 by targeted next-generation sequencing (NGS), they sequenced the exome and filtered 73 genes of the AMH pathway, DSD, malformations and the fetal testis [5]. Five of the 24 patients (21%) had truncating PPP1R12A variants: c.2253delA p.(Asp752MetfsTer36) in exon 16, c.2741T>A p.(Leu914Ter) in exon 21, c.2298C>A p.(Tyr766Ter) in exon 17, c.2608C>T p.(Gln870Ter) in exon 19 and a complex rearrangement of the first, non-coding, exon, with tandem triplication of two sequences of 527 bp and 236 bp [5]. The article does not state the reference transcript for these variants. Only one patient’s family could be studied, and the variant was de novo [5]. The patient described by Voide and colleagues (2026) has the variant NM_002480.3:c.589_590insT p.(Lys197IlefsTer13) in exon 4 of 25, de novo, absent from gnomAD, ClinVar and LOVD (Leiden Open Variation Database), classified as likely pathogenic (criteria PVS1, PM2 and PM6) [8].

The mechanism by which loss of MYPT1 blocks Müllerian regression has not been demonstrated experimentally. Picard and colleagues propose that migration of epithelial cells into the mesenchyme, dependent on cell motility regulated by the balance between myosin kinases and myosin phosphatase, requires intact phosphatase activity; alternatively, the variants might affect AMH signal transduction [5]. The intestinal atresias are attributed to interruption of the mesenteric blood supply, because loss of MYPT1 increases vasoconstriction [5]. Mice with constitutive inactivation die in the embryonic period, and the available conditional models targeted only smooth muscle [5]. Allelic disorders are the contiguous gene deletion at 12q21 (14 individuals described, without 46,XY DSD) and somatic PPP1R12A variants in colorectal cancer and rhabdomyosarcoma, which are not hereditary [10].

2.6. PMDS without an identified variant

In about 12% of the families in the Paris series and 15% in the 2013 count, no variant was detected in AMH or AMHR2, including the proximal promoters and the intronic sequences studied [9, 12]. All these patients have normal serum AMH; several were born small for gestational age or had other malformations, such as jejunal atresia [12]. Some variants may have escaped older methods or lie in the distal promoter or in unstudied deep introns; the AMHR2 splicing variant described by Hoshiya and colleagues in 2003 would not have been identified by the usual approach [9]. The authors nevertheless consider that these forms constitute a distinct entity, because of the frequency of associated malformations, the absence of transverse testicular ectopia and the low consanguinity (10%, versus 40% in AMH families and 33% in AMHR2 families) [9]. The identification of PPP1R12A in 21% of the reanalysed cases confirmed this hypothesis for some of them [5]. In the mouse, inactivation of Wt1 in the Müllerian duct mesenchyme causes duct retention with unaltered testicular AMH and mesenchymal Amhr2, through reduced expression of Wif1 and Osx and nuclear accumulation of β-catenin; there are no human data linking this mechanism to cases of PMDS [28].

2.7. Inheritance, penetrance and genotype–phenotype correlations

The AMH and AMHR2 forms have autosomal recessive inheritance, with homozygosity or compound heterozygosity in all resolved cases [2, 9, 21]. Consanguinity was recorded in 40% of AMH families, 33% of AMHR2 families and 10% of idiopathic families, and homozygosity in 65% and 57% of patients with AMH and AMHR2 variants, respectively, with variation by region [9]. Heterozygotes with one normal allele are clinically normal; the dominant-negative effect observed on overexpression in vitro does not translate into a phenotype [9, 20]. Expression is limited to males: homozygous sisters and, in one case, the homozygous mother of a proband are normal and fertile [9, 25]. The pedigrees of Sloan and Walsh (1976), with two affected half-brothers born to different fathers, and of Naguib and colleagues (1989), with four affected individuals in a Bedouin family, suggested X-linked transmission, but consanguinity in these communities makes interpretation difficult, and no molecular study has confirmed it [2, 9]. The PPP1R12A-associated form has a different inheritance, autosomal dominant and usually de novo [6, 10].

There is no phenotypic difference between the AMH and AMHR2 forms, apart from serum AMH, which is consistent with the exclusive ligand–receptor relationship [9, 12]. The correlation between genotype and clinical phenotype is incomplete: some biallelic carriers have no Müllerian remnants and are asymptomatic or present only with transverse testicular ectopia, whereas other patients with a typical clinical picture have no variant [9]. In the Chinese series of Tian and colleagues, the older brother of a patient, with the same genotype, had no Müllerian derivatives detectable by laparoscopy or imaging [27]. A 2026 report describes a 34-year-old man with transverse testicular ectopia without macroscopic Müllerian remnants, seminoma and novel compound heterozygous AMHR2 variants [29]. The position of the complex formed by the testes and the Müllerian derivatives may differ between brothers with the same genotype [9].

2.8. Animal models and other contexts of the AMH/AMHR2 pathway

PMDS occurs in other species: the AMHR2 exon 3 variant p.Arg81Ter in the Miniature Schnauzer (published canine protein numbering), a form without an elucidated defect in the Basset Hound and AMH and AMHR2 variants in European bison [9]. In mice, inactivation of Amh or Amhr2 facilitates induced testicular tumours, through suppression of inhibin and WT1/β-catenin imbalance [9]. Outside PMDS, AMHR2 is a therapeutic target in oncology: the antibody murlentamab (GM102) was studied in colorectal and gynaecological cancer in completed trials (NCT03799731 and NCT02978755), unrelated to the syndrome [30].

3. Clinical Spectrum and Manifestations by System

3.1. Anatomical presentations

The position of the testes depends on the mobility of the Müllerian derivatives. When the uterus is fixed in the pelvis, the testes remain in a high, “ovarian” position and the patient presents with bilateral cryptorchidism; when the uterus is mobile, one or both testes descend into the inguinal canal or the scrotum dragging the derivatives with them, and a testis is often found in an inguinal hernia with the uterus and one fallopian tube (hernia uteri inguinalis) [9, 12]. Transverse testicular ectopia, with both testes in the same hemiscrotum and the same hernial sac, accompanied by uterus and fallopian tubes, is the most specific anatomical presentation of the syndrome; testicular fusion has been described [9]. Inguinal hernias are very frequent in all forms and may appear at any time after birth [9]. In Orphanet, cryptorchidism is very frequent (80–99% of patients) and inguinal hernia frequent (30–79%) [1].

Table sources: [9, 11, 27]. The series are not comparable with one another: the Paris series results from referral for molecular study, the Chinese series excludes patients without a molecular diagnosis and the review estimates include cases without genetic study. Transverse testicular ectopia was never observed in idiopathic PMDS in the Paris series, and in the review by Tian and colleagues all cases of PMDS with transverse ectopia had possibly causal variants [9, 27].

Not all transverse testicular ectopia corresponds to PMDS. PMDS is present in about 30% of cases of transverse ectopia (Naouar and colleagues, 2008); in a classification cited by Chua and Samnakay, type 1 (up to 50%) is accompanied only by inguinal hernia, type 2 (up to 30%) by persistent Müllerian structures and type 3 by other anomalies, such as scrotal abnormalities, hypospadias, fused vasa deferentia or horseshoe kidney [12, 13]. Very rarely, transverse ectopia is the only anomaly in biallelic carriers of AMH or AMHR2 variants, without detectable Müllerian derivatives; in the Paris cohort this occurred in 2 patients with AMH variants [9, 12].

3.2. Testes, gubernaculum and excretory ducts

The testis in PMDS is loosely anchored to the bottom of the processus vaginalis by a long, thin gubernaculum, similar to the round ligament; it is abnormally mobile, may be pulled into the contralateral hemiscrotum and has an increased risk of torsion and degeneration [9, 12]. Rarely one or both testes are absent, a situation attributed to torsion favoured by mobility [9]. In children testicular morphology is normal and germ cells are present; in older patients the testes are often hypoplastic, with tubular fibrosis, thickened basement membrane and absent spermatogenesis [1, 9].

The vasa deferentia and testes adhere to the walls of the uterus and vagina [12]. The vas deferens initially runs in the mesosalpinx, reaches the uterine wall and penetrates it to open at the top of the vagina, which corresponds to the prostatic utricle; the spermatic vessels are generally very short [9]. The vas deferens is often narrow, blind-ending or absent, and dissociation between epididymis and testis is common; the Paris series mentions these anomalies in 9 patients with an AMH variant, 13 with an AMHR2 variant and 2 idiopathic patients, and notes that the same anomalies are associated with abdominal cryptorchidism without PMDS [9]. In the Chinese series, the vas deferens was blind-ending and dissociated from the testis in 6 of 12 patients, and biopsies showed testicular microlithiasis in 3 of 12 patients (25%) [27].

3.3. Müllerian derivatives

Development of the uterus is highly variable, even between brothers: it may be comparable in size to a female uterus, but is often much smaller and, when retained in the pelvis, may go unnoticed in an inguinal approach [9, 12]. The Müllerian ducts enter the posterior urethra at the level of the prostate, but the communication is often obliterated [9]. In adults with untreated cryptorchidism or hernia, haematuria may result from hormonal imbalance with ageing, because the testes produce fewer androgens and excess oestrogens (Gricourt and colleagues, 2010); cyclical haematuria led to the diagnosis in a 27-year-old man without apparent hormonal abnormalities (Smith-Harrison and colleagues, 2015) [9]. Retained remnants may cause haematuria, recurrent urinary tract infections, stones and voiding disorders [27]. In the Chinese series, 3 of the 8 patients in whom the uterus was preserved developed cysts of the Müllerian remnants in the scrotum, groin or pelvis, interpreted as hydrocolpos due to congenital vaginal obstruction, the first report of this complication after preservation [27].

3.4. Endocrine function, puberty and sexual function

Leydig cell function is normal, unless there is testicular degeneration; testosterone and gonadotrophins are normal for age and puberty proceeds normally [1, 9, 12]. Sexual function is normal, even when fertility is compromised [1, 11]. Bilateral testicular loss through torsion, degeneration or surgery leads to primary hypogonadism: one of the patients of Picard and colleagues (2022), with transverse ectopia, became azoospermic with hypergonadotrophic testicular failure (FSH 126.4 mIU/mL, testosterone 7.9 nmol/L) after necrosis of both testes [5].

3.5. Fertility

Infertility is the most frequent complication [9]. It results from aplasia of the epididymis and upper vas deferens, poor connection of the testes to the excretory ducts, germ cell degeneration due to prolonged cryptorchidism and ischaemic or structural injury to the vas deferens and testis during excision of the Müllerian remnants [1, 12, 13]. In the Paris experience only one patient with an AMH variant became a father, after testicular sperm extraction (TESE) prompted by oligoasthenozoospermia; in the literature, 19% of reported adults had one or more children and Farag and colleagues (1993) recorded 11% fertile patients in Kuwait and neighbouring populations [9]. All the fertile patients except one had transverse ectopia or hernia uteri inguinalis, that is, at least one scrotal testis; all had children before diagnosis, many developed testicular cancer and paternity was not genetically proven [9]. Fertility is thus rare and possible only with at least one normally descended testis and intact excretory ducts [9, 25]. A 2026 urological review restates the proportion of about one in five men with natural conception reported at the time of diagnosis and adds, among the causes of infertility, extrinsic compression of the ejaculatory duct by the Müllerian structures [31]. In 2013 incontestable proof of natural paternity was lacking [12]. A 2026 case report describes spontaneous conception seven months after conservative bilateral orchidopexy in a 28-year-old man with seven years of primary infertility, without molecular study or genetic proof of paternity in the published abstract [32]. Orphanet mentions paternities achieved by intracytoplasmic sperm injection (ICSI) [1].

3.6. Testicular neoplasms

The incidence of testicular cancer in PMDS has been estimated at 18%, a value considered not higher than that of cryptorchid testes in general [9]. The literature review of the Paris group found some form of malignant testicular degeneration, unilateral or bilateral, in 33% of patients aged 18 years or over; the absolute denominator is not given, the cases are published (with probable publication bias) and belong mainly to an era before early orchidopexy [9, 22]. The values cited in other sources vary: 15% for Josso and colleagues, similar to that of undescended abdominal testes without PMDS; 5–18% in most reports according to Chua and Samnakay; 15–40% for undescended testes in PMDS according to Korkmaz and colleagues; and, for Gagliardi and colleagues, an incidence equal to that of cryptorchid men without PMDS [11–13, 33]. The divergence results from different denominators and inclusion criteria and from data that mix natural history (diagnosis in adulthood, abdominal testes for decades) and patients operated on in childhood. In a reference cancer centre in India, 4 of the 637 patients treated for testicular tumour between 1980 and 2022 had PMDS (3 seminomas and 1 mixed germ cell tumour), all at stage 2B or higher and treated with surgery and chemotherapy, and were disease-free after a mean follow-up of 67 months; the literature review for the same period gathered 49 patients with testicular tumour and PMDS in 44 articles, 59% with a bulky abdominal mass and only 5 (10%) with previously adequately treated cryptorchidism [34]. These figures count patients with a tumour and not patients with PMDS, so they do not provide a risk estimate; they show that the typical presentation of the tumour is that of an adult with untreated cryptorchidism and advanced disease [34].

Seminoma is the most frequent type; choriocarcinoma, mixed germ cell tumours, embryonal carcinoma, gonadoblastoma and yolk sac tumour have also been described [9]. In the Paris cohort there were 3 testicular tumours, all in adults with AMH variants [9]. Germ cell neoplasia in situ (GCNIS) was found in a 17-year-old patient and in a 15-month-old boy with a receptor defect [9]. Early orchidopexy does not eliminate the risk: embryonal carcinoma has been described at 16 years after bilateral orchidopexy at 2 months and a mixed germ cell tumour at 23 years after surgery at 5 years for transverse ectopia, which leads the authors to accept that malposition of the testis is not the only factor [9]. Prognosis depends on histology, and choriocarcinoma and mixed tumours have a poor prognosis [9]. In 2026 a seminoma was published in an ectopic testis of a 34-year-old man with transverse ectopia and AMHR2 variants, and a 30 cm intra-abdominal seminoma in a 29-year-old man, in whom the uterus was discovered at surgery after neoadjuvant chemotherapy [29, 35]. In cryptorchidism in general, a Swedish cohort of about 17,000 operated men showed a relative risk of testicular cancer of 2.2 with orchidopexy before 13 years and of 5.4 after that age, compared with the general population [14].

3.7. Neoplasms of the Müllerian derivatives

Malignant degeneration of the Müllerian derivatives is much less frequent than testicular degeneration [9, 13]. Farikullah and colleagues (2012) found 11 cases of malignant degeneration of male Müllerian ducts, only 3 of them in PMDS, the others being prostatic utricle cysts [9]. Gagliardi and colleagues gathered at least 12 cases among about 300 published patients with PMDS, aged 4 to 68 years, and cite an estimated risk of 8% (Beatty and colleagues, 2013), whereas Chua and Samnakay report 3.1–8.4% in the literature [11, 13]. Haematuria is frequently the presenting sign [9]. Two young men with uterine adenocarcinoma died of metastatic disease, at 14 and 39 years, and another patient had a benign uterine leiomyoma [9]. Tian and colleagues count three patients with a malignant neoplasm of the Müllerian remnants of mucosal origin since 2002 [27].

Source: Table 1 of Gagliardi and colleagues [11]. Cases 1 and 6, with hypospadias, may not meet the strict definition of PMDS adopted by the Paris group, which excludes hypospadias, and the difference in criteria explains the gap between Farikullah’s 3 cases and Gagliardi’s 12 [9, 11].

3.8. Idiopathic PMDS, associated malformations and PPP1R12A-associated PMDS

In the Paris series, patients with associated malformations rarely had AMH or AMHR2 variants. The malformations described include intestinal defects (atresia or lymphangiectasia, Urioste syndrome) in 4 cases, Hirschsprung disease, horseshoe kidney, intellectual disability, lipodystrophy with vitamin D-resistant rickets, polycystic kidney disease, hydronephrosis, deafness and prematurity or low weight for gestational age [9].

The five patients of Picard and colleagues (2022) with PPP1R12A variants had complete external virilisation, without signs of Leydig cell deficiency, and AMH normal for age in the three in whom it was measured (122.80 pmol/L in one patient, without indication of the age at sampling; 201 pmol/L at 3 days in another, 171.36 pmol/L at 7 years in the third) [5]. The authors record prematurity in all (gestational age 34 to 37 weeks). Three had small bowel atresia (type III jejunal with malrotation; type IIIb jejunal with multiple stenoses; jejuno-ileal in a stillborn infant who also had a small cerebellum with numerous heterotopias and mild hypertelorism), one had partial oesophageal atresia and transverse testicular ectopia with the uterus in the hernia, and the fifth had no associated malformations [5]. One patient had severe difficulties in learning to read and write; the others had normal intellectual development [5]. These malformations have never been described with AMH or AMHR2 variants [5].

Genitourinary and/or brain malformation syndrome (GUBS) has a broader spectrum. In the series of Hughes and colleagues, of 12 individuals, there was 46,XY DSD or genitourinary anomalies in 9, developmental delay in 7, midline brain malformations (holoprosencephaly spectrum, up to anencephaly) in 5 and ocular abnormalities in 3 [6, 10]. The 46,XY genitalia range from micropenis, chordee, hypospadias, bifid or high scrotum, urogenital sinus anomalies and normal-appearing female genitalia; the gonads range from cryptorchidism to complete dysgenesis with streak gonads; Müllerian structures may persist (6 of 9 patients) [5, 10]. Other anomalies include omphalocele, jejunal and ileal atresia with aberrant mesenteric blood supply, syndactyly and kyphoscoliosis [10]. Before the report of Voide and colleagues, 16 46,XY cases had been published, with no clear correlation between variant location and severity [8]. Expressivity is variable even with the same genotype: monozygotic 46,XY twins with the same de novo variant NM_002480.3:c.1551-2A>G had, one, Prader III genitalia with bilateral gonadal dysgenesis and female sex assignment and, the other, Prader IV with severe hypospadias, left cryptorchidism, right hernia and functioning testes [36]. The patients of Picard and colleagues and that of Voide and colleagues, with external virilisation and normal AMH, place PPP1R12A-associated PMDS at the less severe end of this spectrum [5, 8].

4. Clinical Recognition and Warning Signs

4.1. When to suspect, by age and context

In the newborn and infant, PMDS presents as cryptorchidism in a boy with normal external genitalia. Unilateral or bilateral cryptorchidism alone is not very suggestive; unilateral cryptorchidism with contralateral inguinal hernia and, above all, transverse testicular ectopia, with a testis crossing the midline on palpation or ultrasound, should raise suspicion [9, 27]. In unilateral cryptorchidism, the contralateral hemiscrotum contains a hernia in addition to the testis [12]. Bilateral impalpable cryptorchidism with a normal penis and the existence of a brother with PMDS are the other classical contexts; when a brother already has the diagnosis, imaging should be planned [9, 11]. Suspicion should precede surgery, to avoid unnecessary biopsies and reoperations and so that dissection spares the vasa deferentia and the testicular blood supply [9].

The phenotypically male newborn with bilateral impalpable testes requires immediate specialist consultation, because the child may be 46,XX with congenital adrenal hyperplasia (CAH) and at risk of a salt-wasting crisis; assessment includes electrolytes, karyotype and a hormonal profile with 17-hydroxyprogesterone, luteinising hormone (LH), FSH, testosterone and androstenedione, in collaboration between paediatric endocrinology and urology [14, 37]. The DSDnet consortium recommends comprehensive steroid analysis in all newborns and infants with DSD, to exclude acute adrenal insufficiency [3].

Intraoperative discovery of a uterus and fallopian tubes, during herniorrhaphy or orchidopexy, is the most common form of diagnosis in children [1, 12]. If Müllerian and Wolffian structures coexist at laparoscopy, other DSD should be considered, such as mixed gonadal dysgenesis, although this is unlikely with a normally formed penis; in case of doubt, initial gonadal biopsies confirm the presence of testicular tissue, and biopsy with karyotype before orchidopexy gives the family time to decide [13].

In adults, PMDS manifests as infertility with a history of cryptorchidism or hernia, as inguinal hernia with an impalpable contralateral testis, as haematuria (sometimes cyclical), as an abdominopelvic mass or as a neoplasm of an undescended testis [9, 35, 38]. The 2026 reports include a 27-year-old man in whom herniorrhaphy revealed a uterus, fallopian tubes and transverse testicular ectopia with an atrophic testis and azoospermia, and a 29-year-old man with a 30 cm intra-abdominal seminoma [35, 38].

In newborns with intestinal or oesophageal atresia, midline brain malformations or developmental delay, the presence of Müllerian derivatives or 46,XY DSD points to PPP1R12A variants [5, 8, 10].

4.2. Warning and emergency signs

4.3. Newborn screening

There is no newborn screening for PMDS. Diagnosis depends on investigation of cryptorchidism and intraoperative detection, which explains the probable underestimation of its frequency [11].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Diagnostic criteria

There are no formal diagnostic criteria published by a scientific society nor a specific PMDS guideline [25]. The operational definition used in the series is the presence of Müllerian derivatives (uterus and fallopian tubes), demonstrated by surgery, laparoscopy or imaging, in a 46,XY individual with normal male external genitalia, without hypospadias, and with testes [9]. The Hangzhou series required normal male external genitalia, a 46,XY karyotype and Müllerian remnants confirmed by surgical exploration [27]. Diagnosis of the type requires biallelic variants in AMH (type I) or AMHR2 (type II) [2, 25].

5.2. Serum anti-Müllerian hormone

Before puberty, Sertoli cells secrete AMH at high concentration, so its measurement reflects the presence of testicular tissue without the need for a stimulation test [12, 25]. In PMDS, very low or undetectable serum AMH in a prepubertal 46,XY boy points to type I, and normal or raised AMH points to type II; in the series of Imbeaud and colleagues (1996), AMH was very low or undetectable in type I and at the upper limit of normal in type II [12, 21]. A specific AMH immunoassay (ELISA, enzyme-linked immunosorbent assay) is useful to guide the molecular study [1]. Testosterone and gonadotrophins are normal for age [12].

Interpretation has limits. In 6 post-pubertal patients in the 1996 series the measurement was no longer informative, because AMH production is normally repressed after puberty; in adults AMH is low in both types and does not distinguish them [11, 12, 21]. Low post-pubertal AMH and, to a lesser extent, neonatal AMH are physiological and should not be interpreted as an AMH variant; in the first 2–3 weeks of life AMH is transiently lower, so it is compared with age-specific references and repeated if in doubt [9, 12]. AMH may fall immediately after surgery [9]. In virilisation defects, such as hypospadias, low AMH reflects testicular dysgenesis and not PMDS [9].

A normal AMH generally excludes an AMH variant, with the known exception of p.(Gln496His), which is thought to affect receptor binding; this hypothesis is considered only after the AMHR2 gene has been completely excluded [9, 12]. In the Chinese series, two patients with AMH variants had raised AMH, interpreted as the result of variants that affect bioactivity without affecting secretion; the values were reported in ng/mL with magnitudes improbable for that unit (up to more than 168), suggesting a unit confusion in the article, and so they are not converted [27]. Circulating AMH in type II is normal for age and generally not increased, unlike in androgen insensitivity [9]. A normal AMH is not specific to AMHR2: it is also normal in idiopathic PMDS and in PPP1R12A-associated PMDS [5, 8, 12].

Serum AMH is undetectable in anorchia and always detectable, between subnormal and normal, with bilateral abdominal testes, which distinguishes bilateral impalpable cryptorchidism from anorchia without a stimulation test [12, 18]. The American Urological Association (AUA) guideline proposes measurement of AMH, with other hormonal tests, in bilateral impalpable testes without CAH, to assess anorchia (optional recommendation, grade C) [37]. Rey and colleagues (1999), in 107 patients with ambiguous genitalia, showed low AMH in 46,XY DSD due to abnormal testicular determination and normal or raised AMH in isolated defects of testosterone secretion [18].

Reference values for male AMH depend on the assay. The following table reproduces the values obtained with the Beckman-Immunotech AMH/MIS ultrasensitive EIA kit (ref. A18893), no longer marketed; no paediatric nomograms have been published for current kits nor is there an international measurement standard, and values measured with different methods are not comparable [9]. To convert pmol/L to ng/mL, divide by 7.14 [9].

Source: data from Grinspon and colleagues (2011), reproduced by Picard and Josso; SD, standard deviation; P, percentile; G1–G5, Tanner genital stages [9].

5.3. Combined interpretation of AMH and testosterone in Y-chromosome DSD

Source: classification by Josso, Rey and Picard [12]. If AMH and testosterone are both below the male range, the cause is testicular dysgenesis; if AMH is in the male range and testosterone low, a specific Leydig cell defect; if both are normal or high, an androgen target-organ defect; with normal male genitalia, undetectable AMH indicates anorchia and low AMH primary or central hypogonadism [12].

5.4. Imaging, laparoscopy and endoscopy

Scrotal and pelvic ultrasound is generally sufficient to suspect PMDS, but magnetic resonance imaging (MRI) may be needed to recognise the structures [9]. In the Chinese series, 8 of 12 patients (66.7%) had a pre-operative diagnosis: 7 by ultrasound showing the testes on the same side or a testis crossing the midline, confirmed by exome, and 1 with bilateral impalpable cryptorchidism diagnosed by exome; 3 were diagnosed by discovery of the uterus during herniorrhaphy or orchidopexy and 1 after surgery, through his brother’s diagnosis [27]. The same authors cite a sensitivity of 54% and a specificity of 50% for pelvic ultrasound in locating Müllerian remnants, and note that the position of the testes is mobile and changes during laparoscopy, so that transverse ectopia is easily confused with unilateral cryptorchidism [27].

Laparoscopy directly demonstrates the uterus and fallopian tubes and is the best means of pre-operative diagnosis, but it establishes Müllerian persistence and not the genetic cause [12, 25]. In the guidelines of the European Association of Urology (EAU) and the European Society for Paediatric Urology (ESPU), laparoscopy is required for the final diagnosis of impalpable gonads and of Müllerian structures in DSD, with gonadal biopsy if indicated, and imaging (ultrasound, MRI) is reserved for selected scenarios, such as identification of Müllerian structures in suspected DSD [14]. The AUA guideline advises against ultrasound or other imaging before referral of cryptorchidism (Standard recommendation, grade B), which does not apply to suspected DSD already referred [37]. Cystourethroscopy and retrograde urethrography may show the verumontanum and a long, narrow prostatic utricle [13, 27].

5.5. Genetic study

The karyotype first determines the DSD subgroup; quantitative fluorescent polymerase chain reaction (QF-PCR), with a result in 1–2 days, and the G-banded karyotype are first line, with screening for sex chromosome mosaicism; array comparative genomic hybridisation and single nucleotide polymorphism arrays do not detect balanced structural rearrangements and may miss mosaicism [3, 19]. The presence of SRY completes confirmation of the male 46,XY background [13, 33]. Copy-number variation (CNV) analysis should be part of first-line testing in DSD, particularly with malformations of other systems, because CNVs explain about one fifth to one third of DSD and intragenic CNVs escape Sanger sequencing [3]. Unless there is a known familial variant, DSDnet prefers a candidate gene panel by high-throughput sequencing or the exome, in parallel with phenotypic and hormonal assessment, reserving the genome for novel genes or suspected oligogenic disease; Sanger sequencing is used for family segregation and difficult regions [3]. The EAU/ESPU guidelines note that NGS and exome are becoming the standard in DSD, with the difficulty of proving causality [14].

When a receptor variant is suspected, particularly in patients of Northern European origin, Picard and Josso recommend as first line PCR targeted at the 27 bp deletion in exon 10 of AMHR2: the normal allele measures 382 bp and the deleted allele 355 bp, which migrates faster on the gel [9]. In the Hangzhou series, an NGS panel of 2742 genes with Sanger confirmation, parental segregation study and CNV analysis (CapCNV/CNVkit) detected a deletion of the entire AMH gene in one patient [27]. Exome reanalysis of unexplained cases identified PPP1R12A; one of the five patients had a complex rearrangement of the first, non-coding, exon of PPP1R12A [5].

The technical limits are concrete. AMH is very GC-rich and difficult to sequence, and in 6 families in the Paris series only one allele was found; variants in the distal promoter and deep introns are not studied by the usual approach; a large maternal AMHR2 deletion mimicked homozygosity in one patient [9]. A negative result in AMH and AMHR2 does not exclude PMDS, because 12–15% of cases have no identified variant [9, 12]. The proportion of 85–88% of families with variants derives from reference series that used mainly Sanger sequencing and is not the clinical sensitivity of a specific test [9, 12]. In the Chinese series, several variants were of uncertain significance (VUS), including patients with two VUS, so the molecular diagnosis was formally incomplete according to the rules of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) [27].

The report should state the gene, the transcript with version (NM_000479.5 for AMH, NM_020547.3 for AMHR2), the HGVS (Human Genome Variation Society) description, zygosity, ACMG/AMP classification and confirmation of phase by parental testing when there are two heterozygous variants; DSDnet recommends reporting pathogenic, likely pathogenic and VUS variants in genes related to the phenotype and registering cases in the international I-DSD registry [3, 7].

5.6. Algorithm

5.7. Prenatal and preimplantation diagnosis

Prenatal diagnosis is possible when the pathogenic variants have been identified in a relative [1]. Fetal genetic sex can be determined from cell-free fetal DNA in maternal plasma (QF-PCR of Y-chromosome/SRY markers) from the 6th–7th week, and the genetic status of a known DSD can be studied in chorionic villi (9th–11th week) or amniocytes (15th–20th week) [3]. In PPP1R12A-associated PMDS, prenatal testing and preimplantation genetic testing are possible once the variant has been identified [10]. The sources consulted on PMDS due to AMH and AMHR2 do not specifically discuss preimplantation testing; technically, it rests on the same premise of known familial variants.

6. Differential Diagnosis

6.1. Comparative table

Sources: [1, 5, 9, 10, 12, 13, 19, 37, 39, 40]. AR, autosomal recessive; AD, autosomal dominant; DHT, dihydrotestosterone; hCG, human chorionic gonadotrophin.

6.2. Differential with the gonadal dysgeneses

The main differential is with the DSD in which Müllerian persistence results from insufficient AMH production due to testicular dysgenesis. In SRY variants, for example, AMH is not produced, and the uterus and vagina persist [39]. About 80% of 46,XY women with gonadal dysgenesis have no known aetiology and frequently have a uterus and streak gonads [19]. In these situations the dysfunction affects both Sertoli and Leydig cells, so AMH and testosterone are both low and the external genitalia are undervirilised or female; in PMDS, virilisation is complete and testosterone normal [1, 9, 12]. The absence of genital ambiguity distinguishes PMDS from mixed gonadal dysgenesis, and mixed dysgenesis is unlikely with a normally formed penis; when laparoscopy shows Müllerian and Wolffian structures and there is doubt about the nature of the gonads, gonadal biopsy confirms testicular tissue [9, 13]. The risk of germ cell tumour is much higher in the dysgeneses (SRY, WT1) than in DSD due to defects of hormone synthesis or action, which makes the distinction decisive for gonadal management [41].

6.3. Differential with PPP1R12A syndrome

GUBS occupies an intermediate position. The cases of Hughes and colleagues are suggestive of gonadal dysgenesis, with streak gonads in two, hypospadias, micropenis or ambiguous genitalia in six and incompletely regressed Müllerian derivatives in six of nine, whereas the five patients of Picard and colleagues had complete external virilisation, without Leydig deficiency and with normal AMH, and met the criteria for PMDS [5]. In practice, PMDS with normal AMH in childhood does not distinguish AMHR2, PPP1R12A and the idiopathic form; intestinal atresias and brain malformations point to PPP1R12A, and in carriers complete neurological and urogenital assessment is recommended [8]. The contiguous gene deletion at 12q21 including PPP1R12A and SYT1 has been described in three patients, one of them a fetus, with brain and urogenital malformations, and should be sought by CNV analysis [5, 10].

6.4. Other situations with altered AMH or cryptorchidism

AMH is low from childhood in trisomy 21, normal in childhood and very low from mid-puberty in Klinefelter syndrome, and low in Prader-Willi syndrome and X-linked adrenal hypoplasia congenita (NR0B1); cryptorchidism occurs in about two thirds of patients with Noonan syndrome [12]. None of these situations is accompanied by Müllerian derivatives, but they affect the interpretation of AMH in a boy with cryptorchidism. In the series of Albers and colleagues (1997), of 33 patients with severe hypospadias, one case was classified as “AMH deficiency”; the presence of severe hypospadias does not correspond to the definition of PMDS and the abstract does not make clear whether this was Müllerian persistence in the context of dysgenesis [42, 43].

6.5. Isolated Müllerian remnants

The enlarged prostatic utricle and midline Müllerian cysts are male Müllerian remnants without a uterus or fallopian tubes; among the 11 cases of malignant degeneration of male Müllerian ducts reviewed by Farikullah and colleagues, 8 were prostatic utricle cysts and only 3 occurred in PMDS [9]. The female counterpart of Müllerian pathology is Müllerian aplasia in the 46,XX woman (Mayer-Rokitansky-Küster-Hauser syndrome).

7. Treatments and Risk Management Strategies

7.1. Objectives and principles

Treatment of PMDS is surgical and supportive; there is no specific pharmacological therapy, no controlled trials and no guideline dedicated to the syndrome, and management rests on expert reviews, case series and guidelines on cryptorchidism and DSD [1, 25]. The objectives are to prevent infertility and cancer [9]. The technical difficulty arises because the male excretory ducts are apposed to or included in the walls of the Müllerian derivatives: en bloc removal of the Müllerian organs automatically injures the vas deferens or the deferential artery, on which the blood supply to the testis depends after division of the short spermatic vessels [9]. For this reason total hysterectomy is not recommended [1].

Chua and Samnakay summarise two principles. First, when Müllerian and Wolffian structures coexist, assess for other DSD and, in case of doubt, confirm testicular tissue by biopsy before orchidopexy. Second, in children, give priority to gonadal preservation and to a successful early orchidopexy; if excision of the Müllerian remnant threatens the deferential vessels or the vas deferens, leave it in situ and monitor it [13]. Orchidopexy in PMDS is not genital surgery for sex assignment: male sex is not in question, and the DSDnet consensus recommendations to defer irreversible genital procedures do not apply to testicular descent [12, 41].

7.2. Age at orchidopexy

The European paediatric urology guidelines (EAU/ESPU, 2026 edition) recommend orchidofuniculolysis and orchidopexy before 12 months and, at the latest, at 18 months (strong recommendation), with treatment started at 6 months of age corrected for gestational age [14]. The AUA guideline recommends surgery within the following year, if there is no spontaneous descent by 6 months of corrected age (Standard, grade B) [37]. Both guidelines recommend endocrinological assessment in bilateral impalpable testes and examination under anaesthesia with diagnostic laparoscopy or surgical exploration to locate the intra-abdominal testis [14, 37]. Picard and Josso recommend early orchidopexy in PMDS, without a guarantee of eliminating the tumour risk [9, 33].

7.3. Surgical technique

With high abdominal testes the cord is often too short, the internal spermatic artery has to be divided and the viability of the testis then depends on the deferential artery; to protect it, Vandersteen and colleagues recommended leaving pedicles of myometrium attached to the vasa deferentia and not dissecting the tubal fimbriae from the testis [9]. The Fowler-Stephens procedures, in one or two stages, divide the testicular vessels proximally while preserving the collaterals via the deferential artery and the cremasteric vessels; the Shehata technique, by traction on the vessels, is an alternative with equivalent rates of atrophy and reascent; the second stage is usually performed 6 months later [14]. In a systematic review cited by the EAU/ESPU, testicular atrophy was 1.83% in primary orchidopexy, 28.1% in one-stage Fowler-Stephens and 8.2% in two-stage Fowler-Stephens [14]. Microvascular testicular autotransplantation is a described alternative (Brandli and colleagues, 2005), and most procedures can be performed laparoscopically or by robotic surgery [9, 11].

Management of the Müllerian remnant is the most debated point. The main reason for intervening on the uterus is mechanical: with both testes abdominal, the uterus blocks descent and orchidopexy is impossible [9]. The options described are partial hysterectomy limited to the fundus and proximal fallopian tubes, simple division of the Müllerian structures in the midline with careful dissection of the walls to free the vasa deferentia (laparoscopic magnification helps), removal of the uterine mucosa to reduce the risk of malignancy and, when descent remains limited by the length of the fallopian tubes even after sacrificing the gonadal vessels, sagittal division of the body of the Müllerian structure to move the halves laterally [9, 12, 13]. Tian and colleagues consider it optimal to divide the uterus in the midline, destroy the mucosal lining and leave a pedicle of myometrium intact, a procedure that frees the testes, protects the vas deferens and the blood supply and reduces the probability of malignancy; this is an opinion based on a series of 12 patients without a comparison group [27]. Gagliardi and colleagues, in a review focused on Müllerian oncological risk, consider removal of the Müllerian derivatives desirable when possible, by laparoscopy or robotic surgery, with information on the risk of injury to the vas deferens and infertility [11]. Other reasons for not always leaving the derivatives in situ are malignant degeneration, discomfort and haematuria in a setting of oestrogen excess [9].

In hernia uteri inguinalis and transverse ectopia, some reports leave the Müllerian structures in situ to preserve the vascular supply and the ducts, whereas others excise them, particularly in adults; in children with hernia uteri inguinalis orchidopexy of the contralateral testis is recommended, and in transverse ectopia the testes can be placed in separate hemiscrota by transseptal orchidopexy [13]. Tian and colleagues recommend bilateral orchidopexy because of the mobility of the uterus [27].

The results of the Hangzhou series illustrate the consequences of the options. All patients had laparoscopic exploration and bilateral orchidopexy; the uterus was preserved in 8 and subtotal hysterectomy was performed in 3, with injury to the right vas deferens in a patient with transverse ectopia and post-operative haemorrhage in another; in the 12th patient, the brother of a proband, no uterus was detected [27]. With a median follow-up of 31 months (7–95), there were complications in 6 patients: bilateral hernia recurrence in one, who was reoperated, and, among the 8 with a preserved uterus, irreducible cryptorchidism or recurrence in 2 (one under surveillance, the other undergoing two further operations) and cysts of the Müllerian remnants in 3 (one stable at 2.3 × 0.8 cm; another that grew to 6.3 × 4.0 cm and regressed to 1.5 × 1.1 cm; a third progressively growing to 6.2 × 1.6 cm over 25 months, monitored without intervention) [27]. In the Australian case of Chua and Samnakay, preservation of the remnant with a small midline apical incision allowed a tension-free two-stage bilateral Fowler-Stephens orchidopexy, with a small, stable remnant and testes of normal volume over 7 years [13].

7.4. Orchidectomy and testicular biopsy

Orchidectomy is necessary when the testis cannot be brought down, because of the risk of cancer, but should be a last resort; there is no justification for castrating children without an attempt at orchidopexy [9, 12]. A report from Pakistan describes en bloc removal of the uterus and scrotal testes in a fertile 35-year-old man with transverse ectopia, presented by the authors of the Paris series as an exceptional mutilation [9]. The EAU/ESPU, by panel consensus, proposes orchidectomy in post-pubertal boys with an intra-abdominal testis and a normal contralateral scrotal testis, and the AUA allows orchidectomy with a normal contralateral testis and very short vessels or vas deferens, a dysmorphic or very hypoplastic testis, or post-pubertal age (clinical principle) [14, 37]. In adults with hernia uteri inguinalis, the intra-abdominal testis is often excised because of the risk of malignancy [13]. The EAU sexual and reproductive health guidelines recommend offering orchidectomy to the adult with unilateral cryptorchidism and normal hormonal function and spermatogenesis (strong) and unilateral or bilateral orchidopexy when there is biochemical hypogonadism or spermatogenic failure, if technically possible (weak) [15].

Testicular biopsy is unnecessary, according to Picard and Josso, when the diagnosis of PMDS is not in doubt, because testicular morphology in children is normal [9]. The DSDnet consensus, for 46,XY DSD with retained gonads in general, suggests considering gonadal biopsy to exclude GCNIS or gonadoblastoma, ideally in late adolescence, for lack of reliable markers or imaging [41]. In cryptorchidism corrected in adulthood, the EAU recommends simultaneous testicular biopsy for GCNIS (strong) and, by consensus, multiple biopsies of the undescended testis [15]. Chua and Samnakay and Tian and colleagues performed biopsy at the first operation to confirm testicular tissue, which shows the variation in practice [13, 27].

7.5. Fertility and assisted reproduction

Ductal anomalies, congenital or iatrogenic, can be bypassed by TESE followed by ICSI when fertility is desired; in the Paris series this succeeded in one patient [9, 12]. The DSDnet consensus proposes micro-TESE if biopsy shows spermatozoa [41]. Exogenous testosterone inhibits gonadotrophins and spermatogenesis and is contraindicated when there is an active wish to have children [15].

7.6. Guideline recommendations and divergences

Sources: [3, 9, 14, 15, 37, 41]. None of these guidelines deals specifically with PMDS; in PMDS descent is mechanically blocked by the Müllerian derivatives, so hormonal treatment for descent has no pathophysiological basis [9, 14]. The 2016 global DSD consensus reaffirms full disclosure and ongoing communication with families and the risk of surgical decisions in childhood that remove germ cells [24].

7.7. Hormonal therapy in cryptorchidism: no indication in PMDS

Human chorionic gonadotrophin (hCG) and GnRH analogues produce descent in about 20% of cryptorchid testes in general, with reascent in about 20% of those that descend, and with less success the higher the testis lies; a recent meta-analysis found no consistent evidence for hormones alone [14]. Short-term adverse effects include scrotal erythema and pigmentation, pubic hair, penile growth and pain from hCG injection, and testicular effects such as interstitial haemorrhage, germ cell apoptosis and lower testicular volume in adulthood have been documented [14]. The AUA advises against hormonal therapy to induce descent, and the EAU/ESPU allows it only, with a weak recommendation and without stating doses, in bilateral cryptorchidism to preserve fertility potential [14, 37]. In PMDS descent is mechanically prevented by the uterus and fallopian tubes, and no source consulted proposes hormonal treatment [9]. In adults with cryptorchidism, the EAU recommends not using hormonal treatment (strong) [15].

7.8. Testosterone replacement in hypogonadism

Testosterone replacement is indicated in PMDS only when there is hypogonadism, for example after bilateral orchidectomy, bilateral torsion or bilateral testicular atrophy and degeneration [15]. Uncorrected cryptorchidism and bilateral anorchia are causes of primary hypogonadism [15]. The EAU considers men with symptomatic hypogonadism and total testosterone below 12 nmol/L, without contraindications, to be candidates [15]. Absolute contraindications are locally advanced or metastatic prostate cancer, male breast cancer, an active wish to have children, haematocrit of 54% or more and uncontrolled congestive heart failure; relative contraindications are an International Prostate Symptom Score (IPSS) above 19, a baseline haematocrit of 48–50% and a family history of venous thromboembolism [15]. Exogenous testosterone suppresses gonadotrophins and spermatogenesis; in primary hypogonadism only testosterone can be considered, unlike secondary hypogonadism, in which fertility is theoretically recoverable with gonadotrophins [15].

The reference doses are those of the EAU guideline (2026 edition); the labels are cited for comparison. Gels and long-acting intramuscular testosterone undecanoate have the best safety profile (level of evidence 1a), and the EAU recommends preferring gels at the start of treatment in high-risk men (weak recommendation) [15].

Sources: Table 3.5 of the EAU sexual and reproductive health guidelines [15] and the labels cited.

In the adolescent who has lost both testes before puberty, androgen replacement is necessary during adolescence for the development of secondary sexual characteristics and must be maintained in adulthood [46]. The sources consulted contain no pubertal induction regimen specific to PMDS; the DSDnet consensus refers hormone replacement in DSD without functioning gonads to the recommendations developed for Turner and Klinefelter syndromes [41]. By extrapolation from the US testosterone enanthate label for delayed puberty, the dose lies between 50 and 200 mg every 2–4 weeks for a limited period (for example, 4–6 months), with low starting doses and gradual increases, adjusted to chronological and bone age, and a hand and wrist radiograph every 6 months to monitor epiphyseal maturation [46]. This is an extrapolation from an analogous indication (delayed puberty) and not a dose studied in PMDS; injectable testosterone undecanoate is not indicated below 18 years in the EU [44]. Pubertal induction regimens follow those of congenital hypogonadotrophic hypogonadism.

Adverse effects and their management include acne and injection-site pain (the most frequent with Nebido®), polycythaemia with increased haematocrit, which requires monitoring and dose adjustment or discontinuation, and pulmonary oil microembolism (POME), with cough, dyspnoea, malaise, hyperhidrosis, chest pain, dizziness, paraesthesia or syncope during or shortly after injection; for this reason, injectable undecanoate is given strictly intramuscularly and very slowly (over more than 2 minutes), with observation of the patient during and immediately after each injection [44]. The US Aveed® label has a boxed warning on POME and anaphylaxis and requires 30 minutes of observation [45]. The enanthate label also mentions venous thromboembolism, increased blood pressure and, with prolonged high doses, peliosis hepatis and hepatic neoplasms, and states that the evidence on major cardiovascular events is inconclusive [46]. The EAU recommends assessing cardiovascular risk factors and excluding a family history of venous thromboembolism before starting (strong), keeping the haematocrit below 54% and performing a digital rectal examination at baseline and at least annually [15]. Nebido® is contraindicated in androgen-dependent carcinoma of the prostate or male breast, in past or current liver tumours and in women, and the SmPC recommends monitoring testosterone, haemoglobin, haematocrit, liver function and lipid profile [44].

7.9. Regulatory status

Sources: [44–48]. The text of the Brazilian package inserts was not consulted; the registration status comes from ANVISA open data [48].

7.10. Available evidence and absence of trials

Sources: [9, 11, 13, 27, 30, 34]. There are no randomised trials or prospective comparisons between preservation and removal of the Müllerian remnant; the series are retrospective, small and selected by referral or molecular diagnosis, and the results of one series cannot be compared indirectly with those of another. A negative search of ClinicalTrials.gov does not exclude unregistered studies [30].

7.11. Developments in the last 6–12 months

There are no regulatory processes, opinions or authorisations relating to PMDS. Outside the 12-month window, the report of monozygotic twins with the same PPP1R12A variant and different genital phenotypes (July 2025) documents the variable expressivity of GUBS [36].

7.12. Special situations

Male identity and sex assignment are not in question; Gagliardi and colleagues recommend reassuring the patient about his sexual identity and explaining that the syndrome does not affect sexual function, although it is often associated with infertility [11, 12]. The DSDnet consensus recommends early and progressive information for the child in age-appropriate language, planned transition to adult care with a specific checklist and multidisciplinary care in adulthood, including contact with support groups [41]. In the perioperative period, AMH may fall immediately after surgery, so it should be measured beforehand [9]. Pregnancy and lactation do not apply to the patient; testosterone is contraindicated in women and should not be used by pregnant or breastfeeding women [44]. Agents to avoid are hormonal therapy for testicular descent and exogenous testosterone in men who wish to be fertile [15, 37]. In PPP1R12A-associated PMDS, GeneReviews recommends considering gonadectomy of dysgenetic gonads, referral of undervirilised 46,XY individuals to urology or gynaecology, endocrinology for pubertal induction and post-pubertal hormonal problems, psychological support or a multidisciplinary DSD clinic, and baseline assessment with total testosterone, dihydrotestosterone, inhibin B, AMH, FSH, LH and ultrasound of the kidneys and the Müllerian structures [10].

8. Reproductive Options and Family Counselling

8.1. Recurrence risks

In PMDS due to AMH or AMHR2, with autosomal recessive inheritance, a couple who are both carriers have, in each pregnancy, a 25% probability of a child with a biallelic genotype; as only 46,XY individuals express the syndrome, each male child has a 25% risk of being affected and the risk of an affected child, without specifying sex, is 12.5% per pregnancy [1]. Daughters with the same genotype are not affected [1, 9]. The children of an affected man are obligate carriers; recurrence in the offspring depends on the carrier status of the partner, which may be high in consanguineous marriages, given the frequency of homozygosity in the families of the Paris series (65% for AMH, 57% for AMHR2) [9]. In PPP1R12A-associated PMDS, which is autosomal dominant, all probands tested had a de novo variant; if the variant is not detected in the parents, the risk to siblings is about 1%, owing to the possibility of germline mosaicism, and the children of an affected individual have a 50% risk [10].

8.2. Cascade testing and carriers

Testing of male siblings, with genetic and imaging study, is recommended, because PMDS may be present without evident cryptorchidism and because surveillance and surgery depend on the diagnosis [11, 25]. Given the value placed on fertility in regions with high endogamy, the authors of the Paris series recommend genetic testing in relatives of patients with identified variants and in future spouses, if marriage is planned [9]. Heterozygous carriers of AMH or AMHR2 variants are clinically normal [9]. Homozygous women are normal and fertile; in female mice lacking Amh there is early depletion of the follicular pool, and early menopause in homozygous women is a hypothesis that could not be verified [9]. Ovarian function appears normal in female relatives of affected individuals [1].

8.3. Prenatal and preimplantation diagnosis and counselling of minors

Prenatal diagnosis is possible when the familial variants have been identified, and preimplantation testing has been described for PPP1R12A once the variant has been identified [1, 10]. Determination of fetal sex from cell-free fetal DNA from the 6th–7th week and study of chorionic villi or amniocytes follow the general principles for DSD [3]. The sources consulted do not describe prenatal ultrasound diagnosis of classical PMDS. In the PPP1R12A-associated form, prenatal ultrasound showed bowel dilatation suggestive of obstruction at 31–32 weeks in two patients, and in the patient of Voide and colleagues the persistent Müllerian duct was seen on abdominal ultrasound on the 5th day of life [5, 8]. In counselling minors, the DSDnet consensus recommends progressive age-appropriate information and involvement of the young person in sensitive decisions [41]. In PMDS, the topics to address with the adolescent are fertility, testicular surveillance and self-examination, and the meaning of the genetic status for future offspring [9, 37].

8.4. Content of counselling

Counselling explains the nature of the syndrome (a man with a residual uterus and fallopian tubes, without ambiguity of sex), the recurrence risk, the need for early surgery and testicular surveillance, the reproductive prognosis and the assisted reproduction options [1, 9, 11]. X-linked inheritance suggested by old pedigrees has not been confirmed by any molecular study and should not guide counselling [2, 9]. When no variants are identified, counselling assumes autosomal recessive inheritance as the most likely hypothesis, without confirming it, and considers the possibility of a de novo PPP1R12A-linked form if there are associated malformations [5, 9].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Frequency and what each figure measures

The incidence of PMDS has not been determined accurately, and Orphanet classifies its prevalence as unknown [1, 9]. The available figures are counts of published cases: about 60 publications predating Medline, a further 200 published between 1964 and 2012 (review by Farikullah and colleagues), 34 new cases in the four years before 2016 (review by Elias-Assad and colleagues), more than 250 published cases according to Tian and colleagues and about 300 since 1939 according to Orphanet and Gagliardi and colleagues [1, 9, 11, 27]. The Paris series, with 157 families between 1990 and 2016, measures the experience of a reference laboratory and not population frequency [9]. The apparent increase in incidence is attributed to early investigation of cryptorchidism, routine laparoscopy for impalpable testes and greater awareness among surgeons, and the true frequency is probably underestimated because diagnosis depends on intraoperative detection [9, 11, 12]. In the HGMD (Human Gene Mutation Database) there were 85 AMH and 84 AMHR2 variants at the time of the review by Tian and colleagues [27].

9.2. Geography, consanguinity and founder effect

The families in the Paris series come from Northern Europe (including Northern France), Southern Europe, Africa (mainly the Maghreb), the greater Middle East (including Turkey, Afghanistan and Pakistan) and Latin America; differences in distribution between AMH and AMHR2 are not significant, and the Northern European predominance reflects recruitment bias [12]. Homozygosity is more frequent in regions with endogamy: 12 of 17 patients from the Maghreb were homozygous [9].

The 27 bp AMHR2 deletion has a Northern European founder effect: 67% of the 30 patients with the deletion were of Northern European origin, versus 36% among the 75 patients with an AMHR2 variant (p<0.05), and the deletion was present in 20 of 27 Northern European patients versus 1 of 11 from Southern Europe [9]. The recorded origins include France, the Netherlands, Germany, Sweden, Switzerland, Denmark, Great Britain, Russia, Portugal, Spain, Argentina, the United States and Réunion [9]. In AMH, the variant c.301G>A p.(Gly101Arg) occurs in four families from the Middle East or Pakistan and c.500A>G p.(Tyr167Cys) in five Northern European families; AMHR2 c.1219C>T p.(Arg407Ter) appears in four families from Mediterranean countries, one from Saudi Arabia and one from Brazil [9]. In China, the Hangzhou cohort (12 patients of unrelated parents) showed a predominance of compound heterozygosity and recurrence, in AMH, of NM_000479.5:c.301G>A in three families and of c.1165G>T p.(Glu389Ter) in two [7, 27].

9.3. United States, Europe, Portugal and Brazil

In the United States isolated cases have been described, such as the infant of New York parents of Italian ancestry with both gonads in the right hernial sac, undetectable AMH and compound heterozygosity for a 14 bp deletion in exon 2 and the nonsense variant p.Arg191Ter of AMH, tabulated in the Paris series as NM_000479.5:c.472_485del p.(Pro158AlafsTer11) and c.571C>T p.(Arg191Ter) (Harbison and colleagues, 1991; Carré-Eusèbe and colleagues, 1992), whose sister with the same genotype was phenotypically normal [2, 9, 18]. In Europe, the Paris series and the review by Gagliardi and colleagues, based on an Italian case, are the main sources; there are no population-based studies [9, 11].

In Portugal, the supplementary tables of the Paris series record two patients homozygous for the 27 bp AMHR2 deletion, one of them published by Rosal-Gonçalves and colleagues (2010), and a patient of Portuguese and Filipino origin with a single identified AMH allele (NM_000479.5:c.343_344del p.(Leu115ThrfsTer58)) [7, 9]. In Brazil, Nishi and colleagues (2012) described three unrelated patients homozygous for the AMH intron 2 acceptor variant, with a possible founder effect; the series also includes Brazilian patients homozygous for AMH NM_000479.5:c.1505G>T p.(Arg502Leu) and for AMHR2 NM_020547.3:c.967G>A p.(Gly323Ser), c.983G>A p.(Gly328Asp) and c.1219C>T p.(Arg407Ter) [7, 9]. The sources consulted contain no national epidemiological data for Portugal or Brazil.

9.4. Special populations

Patients with associated malformations (intestinal atresias, prematurity, brain malformations) constitute a subgroup in which AMH and AMHR2 variants are rare and PPP1R12A is a candidate [5, 9]. Adults diagnosed late concentrate the risk of testicular and Müllerian neoplasm and infertility, and are the group from which most natural history data come [9, 11].

9.5. Reference networks and patient associations

Endo-ERN has a thematic group dedicated to rare conditions of sex development and maturation, with patient representatives (ePAG, European Patient Advocacy Groups) and a public list of reference centres; ERN eUROGEN includes area EA 1.1, complex genital reconstruction in DSD, relevant to paediatric urological surgery [16, 17]. Orphanet lists specialised centres, networks and patient organisations for the entity [1]. The international I-DSD registry collects cases for research [3]. Among the associations, dsdfamilies (United Kingdom) informs families of children with DSD and interACT (USA) advocates for young people with intersex variations or DSD; none is specific to PMDS [49, 50].

10. Illustrative Clinical Case

10.1. Bilateral impalpable cryptorchidism with a preserved Müllerian remnant and 7-year follow-up

Chua and Samnakay (2022) described an Australian boy, a healthy first child born at term, without a family history or consanguinity, referred at 6 months for bilateral impalpable testes [13]. The penis was normal, with a complete foreskin and glanular meatus, and the scrotum was well formed but empty; initial scrotal and pelvic ultrasound showed no testes or pelvic anomaly [13]. Diagnostic laparoscopy at 6 months revealed normal-appearing bilateral intra-abdominal testes and a uterus-like Müllerian structure, with bilateral fallopian tubes and round ligaments; the vasa deferentia were in close association with the tubal structures (figure 1a–b, https://pmc.ncbi.nlm.nih.gov/articles/PMC8977345/figure/fig1/) [13]. Pole-to-pole gonadal biopsies and cystourethroscopy were performed, with no evident channel between the Müllerian structure and the urethra [13].

Histology showed normal bilateral testicular tissue; SRY was present on fluorescence in situ hybridisation and the karyotype was 46,XY [13]. LH, FSH and testosterone were appropriate for age, and AMH was 526 pmol/L (normal below 1 year: 454–944 pmol/L), which led the authors to consider an AMHR2 variant likely, without genetic testing being performed [13]. On the hCG test, testosterone rose from less than 1 nmol/L to 18 nmol/L at 72 hours [13].

Treatment was a two-stage bilateral laparoscopic Fowler-Stephens orchidopexy: in the first stage, division of the limiting gonadal vessels (figure 1c); in the second, the Müllerian structure was left intact so as not to injure the vasa deferentia or the vessels, and a small longitudinal midline apical incision allowed the tubal arms to be separated and both testes to be placed in the scrotum without tension, with intact vasa deferentia and deferential blood supply (figure 1d) [13]. Follow-up included annual pelvic and scrotal ultrasound for 7 years and baseline pelvic MRI at 6 years: the Müllerian remnant remained small and stable (figures 2 and 3, https://pmc.ncbi.nlm.nih.gov/articles/PMC8977345/figure/fig2/ and https://pmc.ncbi.nlm.nih.gov/articles/PMC8977345/figure/fig3/), the testes remained intrascrotal with normal structure and volume for age and there was no epididymitis, urinary tract infection or incontinence [13].

The case shows the most frequent presentation (bilateral cryptorchidism with the uterus fixed in the pelvis), the value of a normal AMH as a pointer to type II and the strategy of preserving the remnant. Two aspects of the management reported differ from recommendations discussed: the absence of a molecular study runs counter to the recommendation to seek a genetic diagnosis in DSD, and gonadal biopsy, which the DSDnet consensus accepts in 46,XY DSD with retained gonads, is considered unnecessary by Picard and Josso when the diagnosis of PMDS is not in doubt [3, 9, 41]. The annual imaging interval reflects the authors’ opinion, without data on the ideal frequency [13].

10.2. Other published cases with molecular data

Korkmaz and colleagues (2017) described a 13-month-old Turkish boy, the son of first cousins, with bilateral cryptorchidism and right transverse testicular ectopia on ultrasound (two structures of 12 × 8 mm and 11 × 7 mm in the inguinal canal), uterus and fallopian tubes at laparoscopy, 46,XY with SRY, AMH above 22 ng/mL (laboratory reference 4.9–264.5), and homozygosity for the novel AMHR2 variant NM_020547.2:c.24G>A p.(Trp8Ter) (= NM_020547.3:c.24G>A in the MANE transcript), with heterozygous parents; the decision of the DSD clinic was orchidopexy, with both testes brought down into the scrotum [33]. This patient corresponds to case R006 of the Paris series [9].

Voide and colleagues (2026) described a 46,XY newborn, the son of non-consanguineous parents of Moroccan and Kurdish origin, small for gestational age (2790 g, −1.3 SD), with a genital tubercle of 2.5 × 0.9 cm, apical meatus, impalpable gonads and a persistent Müllerian duct of 2.5 × 1 cm on ultrasound on day 5 [8]. At 3 years, laparoscopy showed Müllerian derivatives extending to intra-abdominal gonads of testicular appearance, and the exome identified the de novo PPP1R12A variant NM_002480.3:c.589_590insT p.(Lys197IlefsTer13) [8]. AMH (174.4 pmol/L at 34 hours; 342.1 pmol/L at 1.5 months) and inhibin B were normal during minipuberty and fell below the lower limit at 10 months (AMH 149.8 pmol/L; inhibin B 76 pg/mL), with FSH and LH during minipuberty 1.6 and 2.0 times above the upper limit for age [8]. The authors suggest progressive Sertoli cell dysfunction due to intra-abdominal cryptorchidism or progressive dysgenesis, without histological data; neurological examination at 3 years was normal and the family declined brain MRI [8].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

The sources consulted contain no data on mortality or life expectancy; prognosis is determined by fertility and oncological risk [9]. Puberty and sexual function are normal [1, 12]. Fertility is compromised even in treated patients; natural paternity is rare and restricted to patients with at least one scrotal testis and intact ducts, and assisted reproduction (TESE and ICSI) is the most realistic route [1, 9]. For comparison, in cryptorchidism in general, untreated bilateral cryptorchidism is associated with oligospermia in 100% and azoospermia in 75% of men, and successfully treated bilateral cryptorchidism with 75% and 42%; paternity is 62% after bilateral cryptorchidism and 89.5–89.7% after unilateral cryptorchidism [14, 15, 37].

The risk of testicular neoplasm varies, depending on the source, between 5–18% and 33% of adults, the latter value referring to published patients aged 18 years or over and predating the widespread adoption of early orchidopexy [9, 13]. Early orchidopexy reduces, without normalising, the risk, and tumours have been described after surgery in childhood [9, 37]. Early diagnosis and treatment may reduce the risks of degeneration and malignancy associated with long-standing cryptorchidism [1]. The risk of Müllerian neoplasm is lower than the testicular risk (3.1–8.4% in the literature, at least 12 published cases) and manifests mainly as haematuria, with deaths from metastatic adenocarcinoma and adenosarcoma [9, 11, 13]. In PPP1R12A-associated PMDS, prognosis depends mainly on the associated malformations and on neurodevelopment [5, 10].

11.2. Surveillance

There is no validated surveillance protocol or specific guideline, and the proposed intervals result from expert opinion and cryptorchidism guidelines [11, 13, 25].

Systematic gonadal biopsy is a matter of divergence: the DSDnet consensus proposes considering it in 46,XY DSD with retained gonads and the EAU recommends it at adult orchidopexy, whereas Picard and Josso consider it unnecessary in PMDS with a secure diagnosis [9, 15, 41]. During testosterone replacement, the EAU further recommends, in prostate cancer survivors, PSA at 3, 6 and 12 months and annually thereafter [15].

11.3. Transition to adulthood

Transition should be planned, with a checklist, and continue with multidisciplinary adult care, including urology, endocrinology, reproductive medicine and psychology [41]. At this stage the reproductive plan and possible sperm retrieval, teaching of self-examination and continued imaging surveillance of the retained Müllerian remnant are reviewed [13, 37, 41].

12. Summary and Final Considerations

Persistent Müllerian duct syndrome is a rare 46,XY DSD in which a man with normal external genitalia retains a uterus, fallopian tubes and upper vagina owing to lack of anti-Müllerian hormone action between the end of the 7th and the 10th week of gestation. In about 85–88% of families the cause is biallelic variants of AMH (type I, very low or undetectable serum AMH) or AMHR2 (type II, normal or raised AMH), in roughly equal parts, with autosomal recessive inheritance limited to males; the 27 bp deletion in exon 10 of AMHR2, NM_020547.3:c.1332_1358del, is the most frequent variant and has a Northern European founder effect. In the remaining cases, frequently with associated malformations and normal AMH, some result from de novo truncating variants of PPP1R12A, which place PMDS at the less severe end of genitourinary and/or brain malformation syndrome, with autosomal dominant inheritance.

Diagnosis is usually made during surgery for cryptorchidism or hernia, but may be suspected beforehand in the presence of unilateral cryptorchidism with contralateral hernia, transverse testicular ectopia or an affected brother. Serum AMH before puberty, interpreted for age and assay, guides the genetic study, which should include sequencing of AMH and AMHR2 with CNV analysis and, in unexplained cases, exome. A negative result does not exclude the syndrome.

Treatment is surgical: early orchidopexy, ideally before 12 months, with techniques that preserve the vasa deferentia and the deferential artery, and individual management of the Müllerian remnant, without total hysterectomy. The cryptorchidism guidelines do not deal specifically with PMDS, and hormonal therapy for descent has no basis. Testosterone replacement applies only to hypogonadism, with the doses of the European guideline as reference. Infertility is the most frequent complication and assisted reproduction the main route to paternity. The risk of testicular neoplasm, highly variable between sources, and the lower risk of Müllerian neoplasm justify long-term surveillance, with testicular self-examination, periodic imaging of the retained remnant and follow-up by a multidisciplinary team, within an evidence base limited to retrospective series and case reports.

13. References

1. Orphanet. Persistent Müllerian duct syndrome (ORPHA:2856). Reviewer: Picard JY. Last updated: October 2025. Available from: https://www.orpha.net/en/disease/detail/2856
2. Online Mendelian Inheritance in Man (OMIM). #261550 Persistent Mullerian duct syndrome, types I and II; PMDS. Johns Hopkins University. Last edited 06/04/2022. Available from: https://www.omim.org/entry/261550
3. Audí L, Ahmed SF, Krone N, Cools M, McElreavey K, Holterhus PM, et al. Genetics in endocrinology: approaches to molecular genetic diagnosis in the management of differences/disorders of sex development (DSD): position paper of EU COST Action BM 1303 'DSDnet'. Eur J Endocrinol. 2018;179(4):R197-206. doi:10.1530/EJE-18-0256.
4. UniProt Consortium. UniProtKB P03971 (AMH) and Q16671 (AMHR2). Available from: https://www.uniprot.org/uniprotkb/P03971 and https://www.uniprot.org/uniprotkb/Q16671
5. Picard JY, Morin G, Devouassoux-Shisheboran M, Van der Smagt J, Klosowski S, Pienkowski C, et al. Persistent Müllerian duct syndrome associated with genetic defects in the regulatory subunit of myosin phosphatase. Hum Reprod. 2022;37(12):2952-9. doi:10.1093/humrep/deac239.
6. Online Mendelian Inheritance in Man (OMIM). #618820 Genitourinary and/or brain malformation syndrome; GUBS. Johns Hopkins University. Last edited 03/08/2020. Available from: https://www.omim.org/entry/618820
7. 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):61-8. doi:10.1002/humu.23348. Available from: https://variantvalidator.org (version 4.0.1, accessed 2026-10-26 and 2026-11-01).
8. Voide M, Santoni F, Bartoloni L, Meylan-Merlini J, Sanchez O, Hauschild M, et al. Case report: minipuberty hormonal profile in PPP1R12A-related persistent Müllerian duct syndrome. Front Endocrinol (Lausanne). 2026;17:1882699. doi:10.3389/fendo.2026.1882699.
9. Picard JY, Josso N. Persistent Müllerian duct syndrome: an update. Reprod Fertil Dev. 2019;31(7):1240-5. doi:10.1071/RD17501. Author manuscript: HAL hal-02472552 (https://hal.sorbonne-universite.fr/hal-02472552).
10. Alkhunaizi E, Chitayat D. PPP1R12A-related urogenital and/or brain malformation syndrome. 2021 Sep 9 [updated 2024 Mar 7]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK573670/
11. Gagliardi F, Lauro A, De Anna L, Tripodi D, Esposito A, Forte F, et al. The risk of malignant degeneration of Müllerian derivatives in PMDS: a review of the literature. J Clin Med. 2023;12(9):3115. doi:10.3390/jcm12093115.
12. Josso N, Rey RA, Picard JY. Anti-müllerian hormone: a valuable addition to the toolbox of the pediatric endocrinologist. Int J Endocrinol. 2013;2013:674105. doi:10.1155/2013/674105.
13. Chua I, Samnakay N. Persistent Müllerian duct syndrome: understanding the challenges. Case Rep Urol. 2022;2022:2643833. doi:10.1155/2022/2643833.
14. EAU Guidelines on Paediatric Urology. Edn. presented at the EAU Annual Congress London 2026. Arnhem: EAU Guidelines Office; 2026. Available from: https://uroweb.org/guidelines/paediatric-urology
15. EAU Guidelines on Sexual and Reproductive Health. Edn. presented at the EAU Annual Congress London 2026. Arnhem: EAU Guidelines Office; 2026. Available from: https://uroweb.org/guidelines/sexual-and-reproductive-health
16. European Reference Network on Rare Endocrine Conditions (Endo-ERN). Rare sex development and maturation conditions. Available from: https://endo-ern.eu/rare-sex-development-maturation-conditions/
17. European Reference Network for rare urogenital diseases and complex conditions (ERN eUROGEN). Complex genital reconstructions (DSDs) (EA 1.1). Available from: https://eurogen-ern.eu/workstreams-and-expertise/workstream-1/complex-genital-reconstructions-dsds/
18. Online Mendelian Inheritance in Man (OMIM). *600957 Anti-Mullerian hormone; AMH. Johns Hopkins University. Available from: https://www.omim.org/entry/600957
19. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 54–57 (Ambiguous genitalia, including sex reversal).
20. Online Mendelian Inheritance in Man (OMIM). *600956 Anti-Mullerian hormone receptor, type II; AMHR2. Johns Hopkins University. Available from: https://www.omim.org/entry/600956
21. Imbeaud S, Belville C, Messika-Zeitoun L, Rey R, di Clemente N, Josso N, et al. A 27 base-pair deletion of the anti-müllerian type II receptor gene is the most common cause of the persistent müllerian duct syndrome. Hum Mol Genet. 1996;5(9):1269-77. doi:10.1093/hmg/5.9.1269.
22. Picard JY, Cate RL, Racine C, Josso N. The persistent Müllerian duct syndrome: an update based upon a personal experience of 157 cases. Sex Dev. 2017;11(3):109-25. doi:10.1159/000475516.
23. Hughes JJ, Alkhunaizi E, Kruszka P, Pyle LC, Grange DK, Berger SI, et al. Loss-of-function variants in PPP1R12A: from isolated sex reversal to holoprosencephaly spectrum and urogenital malformations. Am J Hum Genet. 2020;106(1):121-8. doi:10.1016/j.ajhg.2019.12.004.
24. Lee PA, Nordenström A, Houk CP, Ahmed SF, Auchus R, Baratz A, et al. Global disorders of sex development update since 2006: perceptions, approach and care. Horm Res Paediatr. 2016;85(3):158-80. doi:10.1159/000442975.
25. GeneGuidance. Entries for the AMH and AMHR2 genes and for the condition persistent Müllerian duct syndrome types I/II (OMIM 261550). Version of 2026-09-26.
26. National Center for Biotechnology Information. ClinVar [database]. VCV000008627, VCV003779441, VCV003583554, VCV002736758. Available from: https://www.ncbi.nlm.nih.gov/clinvar/
27. Tian HJ, Wu DH, Ru W, Wu DW, Tao C, Chen GJ, et al. Surgical management and molecular diagnosis of persistent Müllerian duct syndrome in Chinese patients. Asian J Androl. 2022;24(1):78-84. doi:10.4103/aja202175.
28. Chen M, Qi X, Qin S, Kang J, Cen C, Guo M, et al. Wt1 is required for the regression of Müllerian ducts in male mice by inducing Wif1 and Osx expression. Cell Prolif. 2026:e70264. doi:10.1111/cpr.70264.
29. Ba TDM, Dang QT, Pham VH, Ho VPN, Vo MD, Huynh QH. A novel AMHR2 gene mutation causing transverse testicular ectopia: a rare case report and literature review. Urol Case Rep. 2026;68:103556. doi:10.1016/j.eucr.2026.103556.
30. National Library of Medicine. ClinicalTrials.gov [database], API v2: search “persistent mullerian duct syndrome” and “anti-mullerian hormone receptor” (NCT03799731, NCT02978755, NCT05592730). Accessed 26 October 2026. Available from: https://clinicaltrials.gov/
31. George M, Wong S, Mathur R, Tsampras N, Lewis A, Pearce I, et al. Persistent Müllerian duct syndrome - a rare but important cause of male factor infertility. Nat Rev Urol. 2026;23(6):385-394. doi:10.1038/s41585-025-01108-5. PMID: 41203848.
32. Karimi M, Karimi T. Spontaneous fertility following conservative surgical management of persistent Müllerian duct syndrome: a case report. Cureus. 2026;18(5):e108507. doi:10.7759/cureus.108507.
33. Korkmaz Ö, Özen S, Özcan N, Bayındır P, Şen S, Onay H, et al. Persistent Müllerian duct syndrome with transverse testicular ectopia: a novel anti-Müllerian hormone receptor mutation. J Clin Res Pediatr Endocrinol. 2017;9(2):179-81. doi:10.4274/jcrpe.4058.
34. Philips MR, Menon AR, Kumar GR, Malik K, Chandrasekaran S, Ramaswamy T, et al. Testicular malignancy in persistent Mullerian duct syndrome: experience from an apex cancer center with review of literature. Urol Oncol. 2023;41(5):258.e1-258.e6. doi:10.1016/j.urolonc.2023.02.008. PMID: 37019765.
35. Krimi O, Khessairi N, Bouida MA, Bouhani M, Sassi I, Kacem S, et al. Persistent Mullerian duct syndrome. Urol Case Rep. 2026;67:103494. doi:10.1016/j.eucr.2026.103494.
36. Tian H, Wu D, Yang H, Wu D, Tao C, Wei J, et al. PPP1R12A mutation leads to different genders of twinning: a case report and literature review. Transl Pediatr. 2025;14(7):1708-16. doi:10.21037/tp-2025-166. Erratum: Transl Pediatr. 2026;15(4):164.
37. Kolon TF, Herndon CD, Baker LA, Baskin LS, Baxter CG, Cheng EY, et al. Evaluation and treatment of cryptorchidism: AUA guideline. J Urol. 2014;192(2):337-45. doi:10.1016/j.juro.2014.05.005. Web version: https://www.auanet.org/guidelines-and-quality/guidelines/cryptorchidism-guideline (published 2014; literature review 2018; validity confirmed 2025).
38. Sed ZMD, Filig MOO, Warsame BA, Dahir AA, Dirir SA. Incidental discovery of persistent Müllerian duct syndrome with transverse testicular ectopia during adult inguinal hernia repair: a case report. Urol Case Rep. 2026;67:103524. doi:10.1016/j.eucr.2026.103524.
39. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 344–345 (Androgen insensitivity syndrome).
40. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 620–621 (Wilms tumour).
41. Cools M, Nordenström A, Robeva R, Hall J, Westerveld P, Flück C, et al. Caring for individuals with a difference of sex development (DSD): a Consensus Statement. Nat Rev Endocrinol. 2018;14(7):415-29. doi:10.1038/s41574-018-0010-8.
42. Albers N, Ulrichs C, Glüer S, Hiort O, Sinnecker GH, Mildenberger H, et al. Etiologic classification of severe hypospadias: implications for prognosis and management. J Pediatr. 1997;131(3):386-92. doi:10.1016/s0022-3476(97)80063-7.
43. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 186–187 (Hypospadias).
44. INFARMED. Nebido 1000 mg/4 ml solução injetável (undecanoato de testosterona) — Resumo das Características do Medicamento. Marketing authorisation holder: Grünenthal S.A. Infomed. Available from: https://extranet.infarmed.pt/INFOMED-fo/
45. Endo USA, Inc. AVEED (testosterone undecanoate) injection, for intramuscular use. Prescribing information. NDA022219. Revised 2025-07-15. DailyMed/openFDA: https://api.fda.gov/drug/label.json?search=openfda.brand_name:AVEED
46. Testosterone enanthate injection. Prescribing information. ANDA212659 (Eugia US LLC). Effective 2026-06-12. openFDA.
47. Testosterone cypionate injection, USP. Prescribing information. ANDA091244 (Bryant Ranch Prepack). Effective 2025-11-12. openFDA.
48. Agência Nacional de Vigilância Sanitária (ANVISA). Open data — registered medicines (DADOS_ABERTOS_MEDICAMENTOS.csv). Available from: https://dados.anvisa.gov.br/dados/DADOS_ABERTOS_MEDICAMENTOS.csv
49. dsdfamilies. Available from: https://www.dsdfamilies.org/
50. interACT: Advocates for Intersex Youth. Available from: https://interactadvocates.org/