5-Alpha-Reductase Type 2 Deficiency

An autosomal recessive 46,XY difference of sex development caused by biallelic SRD5A2 variants, impairing conversion of testosterone to dihydrotestosterone and producing genital undervirilisation at birth followed by virilisation at puberty; includes 17β-hydroxysteroid dehydrogenase type 3 deficiency as the main differential diagnosis.

Literature and regulatory information updated: 10 October 2026.

0. Initial Reference Table

1. Introduction

1.1. From pseudovaginal hypospadias to enzyme deficiency

In 1961, Nowakowski and Lenz termed “pseudovaginal perineoscrotal hypospadias” a form of atypical sexual differentiation in 46,XY individuals with female-appearing external genitalia, bilateral testes and an internal male urogenital tract, whose ejaculatory ducts ended in a blind vagina [9]. The designation described anatomy without identifying the cause and remained a synonym in the OMIM catalogue (entry 264600) [2]. The mechanism was elucidated in 1974, when two groups simultaneously described deficient conversion of testosterone (T) to dihydrotestosterone (DHT). Imperato-McGinley and colleagues studied, in Science, a large consanguineous Dominican Republic kindred with ambiguous genitalia at birth and marked virilisation at puberty, and demonstrated very low plasma DHT due to deficient 5α-reductase activity and autosomal recessive inheritance [24]. In the same year, Walsh, Wilson and colleagues (Dallas) described in the New England Journal of Medicine two US siblings with “familial incomplete male pseudohermaphroditism type 2” and decreased DHT formation [3]. The original description emphasised three features that continue to define the entity: absent external genital masculinisation in utero, clinical and psychological virilisation at puberty and absence of gynaecomastia [9].

1.2. Untreated natural history and debate on gender identity

In 1979, Imperato-McGinley's group reported that, of 38 affected members of the Dominican kindred, 18 had been unequivocally raised as girls and that, during or after puberty, 17 of those 18 had adopted a male gender identity and 16 a male gender role; the authors interpreted this as evidence that cerebral testosterone exposure in utero, neonatally and at puberty contributes to male identity [25]. In 1980, the same group described a 65-year-old man who had never received medical intervention, with an android habitus, male identity developed at puberty and social expression delayed by social factors; fibroblast enzyme activity showed lower affinity for T and NADPH (reduced nicotinamide adenine dinucleotide phosphate) and thermal instability, suggesting inherited enzymatic heterogeneity [26]. Anthropology added a counterpoint: among the Sambia of Papua New Guinea, Herdt and Davidson described fourteen affected people in 1988, a third social category (“turnim-man”) and gender role changes in a context of social trauma, emphasising sociocultural factors [4]. In 1991, a cluster was described in the Simbari Anga linguistic group, where some villages began to recognise affected individuals in childhood and raise them as boys, using the designation “kwalatmala” [5]. These initial observations gave rise to pharmacological 5α-reductase inhibitors, developed for benign prostatic hyperplasia and androgenetic alopecia [15].

1.3. Molecular discovery

In 1991, Andersson, Russell and colleagues isolated by expression cloning a second 5α-reductase complementary DNA with the properties of the predominant genital tissue isoenzyme, and found a deletion of this gene in two related individuals with the deficiency, confirming at least two human 5α-reductases [27]. The first cloned isoenzyme (from rat liver) was named 5α-reductase type 1 and the second, identified from human prostate and patients, type 2 [9]. In 1992, Thigpen and colleagues characterised SRD5A2 (5 exons, 4 introns), mapped it to 2p23 and identified 18 variants in 23 families of 19 ethnic origins, with a high compound heterozygosity proportion, suggesting a higher carrier frequency than assumed [28]; in the same year they described the molecular basis of the Dominican kindred [29]. In 1996, Cai and colleagues showed that several variants coexisted in that population, contradicting a single common ancestor hypothesis [30].

1.4. Nomenclature and reclassification

The 2005 Chicago Consensus, organised by the Lawson Wilkins Pediatric Endocrine Society (LWPES) and European Society for Paediatric Endocrinology (ESPE) and published in 2006, introduced “disorders of sex development” (DSD) and a karyotype-based classification, abandoning “male pseudohermaphroditism” and “intersex”; 5α-reductase deficiency was classified under “46,XY DSD → disorders of androgen synthesis or action → androgen biosynthesis defect”, alongside 17β-HSD3 deficiency and StAR defects [13, 31]. “Differences of sex development” is now frequently used instead of “disorders”, with the latter remaining acceptable among professionals [12]. Strictly, 5α-reductase type 2 acts in peripheral target tissues rather than testicular T synthesis, although clinical classifications group it among biosynthesis defects [12]. The Oxford Desk Reference titles the corresponding paragraph “5β-reductase deficiency”, a typographical error: the enzyme concerned is 5α-reductase, as the book itself states elsewhere [11, 32].

1.5. Definition of the entities in this article

5α-Reductase type 2 deficiency is defined as a 46,XY DSD caused by biallelic pathogenic SRD5A2 variants, with testes, Wolffian derivatives present, Müllerian derivatives absent and incomplete intrauterine masculinisation of the external genitalia, with a spectrum ranging from female genitalia with a blind vaginal pouch to a male phenotype with pseudovaginal posterior hypospadias and micropenis, or isolated micropenis [1, 9]. Homozygous 46,XX women have a minimal phenotype [33]. 17β-HSD3 deficiency, caused by biallelic HSD17B3 variants, impairs testicular conversion of androstenedione to T; it is addressed in this article as the main differential, with its own subsection, because it shares the presentation (undervirilisation at birth, virilisation at puberty, primary amenorrhoea) and the dilemmas of gender assignment [7, 16]. Androgen insensitivity syndrome and the general approach to DSD enter here only as differential diagnosis and context.

2. Molecular Genetics and Pathophysiology

2.1. Testosterone, dihydrotestosterone and genital differentiation

External genitalia are identical in both sexes until 12 weeks of gestation and, even at 20 weeks, ultrasound determination of fetal sex may be difficult [11]. In the 46,XY fetus, the testis produces anti-Müllerian hormone (AMH), which causes regression of the Müllerian ducts, and T, which stabilises the Wolffian ducts and differentiates them into the epididymis, vas deferens, seminal vesicles and ejaculatory ducts [9, 11]. Male external genitalia develop from 8 weeks in the presence of DHT formed from T by 5α-reductase [32]. DHT has a 2- to 5-fold greater affinity for the androgen receptor (AR) than T and approximately 10-fold greater signalling potency; it is required for prostate growth and for differentiation of the urogenital sinus and genital tubercle into the urethra, penis and scrotum, whereas T, through the AR, differentiates the Wolffian ducts and contributes to male psychosexuality [2, 9]. This division of functions explains the cardinal phenotype: in 5α-reductase type 2 deficiency the Wolffian derivatives form and the Müllerian derivatives regress, but the external genitalia and prostate remain undervirilised [15].

2.2. 5α-Reductase isoenzymes

5α-Reductases stereospecifically and irreversibly reduce the C4–C5 double bond of 3-oxo-Δ4,5 C19 and C21 steroids (T, progesterone, cortisol, aldosterone), with NADPH as cofactor [2, 9]. 5α-Reductase type 1 (259 amino acids, SRD5A1 at 5p15) is expressed in fetal scalp and non-genital skin, 5 to 50 times more in adults than fetuses, and very little in prostate and genital skin; type 2 (254 amino acids, SRD5A2 at 2p23) is highly expressed in fetal prostate and, after birth, prostate, seminal vesicles, epididymis, genital skin and liver [9, 15]. Both genes have 5 exons and 4 introns and approximately 60% identity [9]. SRD5A3 (4q12; 318 amino acids; 19–20% homology) encodes a polyprenol reductase in the dolichol pathway; its biallelic variants cause a congenital glycosylation disorder (SRD5A3-CDG) with ophthalmological and neurological manifestations, without DSD [9]. 5α-Reductase type 1 and direct T action on AR explain pubertal virilisation despite type 2 deficiency [2]. Since hepatic type 2 participates in glucocorticoid and other steroid metabolism, deficiency generally alters the urinary ratio of 5α- to 5β-reduced metabolites, underpinning urinary steroid profiling [5, 15].

2.3. Molecular mechanisms of variants

Variants act through three main mechanisms: altered substrate or cofactor binding, absence of functional isoenzyme (deletions, nonsense variants, splice variants) and production of an unstable isoenzyme [15]. In a three-dimensional model with six α helices, residues Y98, N102, Y107, L167, R171, H231 and Y235 contact NADPH; p.His231Arg mainly affects T binding and p.Arg171Ser greatly reduces NADPH affinity, and the histidine segment 230–232 is highly conserved [9]. In functional studies of 40 variants, one group had no detectable activity (for example p.Tyr136Ter, p.Gly115Asp, p.Pro59Arg, p.Gln126Arg, p.Leu55Gln, p.Arg246Gln and splice variants IVS1-2A>G and IVS4+2T>C), another reduced activity (for example p.Arg145Trp, p.Thr187Met, p.Gly203Ser, p.Arg227Gln, p.Arg246Trp), another altered NADPH binding (p.Arg171Ser, p.Gly183Ser, p.Asn193Ser, p.Gly196Ser, p.Arg246Gln) and another T binding (p.Gly34Arg, p.His231Arg) [9]. IVS4+2T>C, the only splice variant functionally studied, causes exon 4 skipping and a truncated 205-amino-acid protein without activity [9]. Crystal structure analysis in a Chinese Sichuan cohort attributed interference with NADPH binding or protein destabilisation to missense variants, and identified transcript-altering variants [34].

2.4. Variant spectrum and hotspots

An integrative review of the worldwide literature up to 2019, which considered only individuals with a biallelic genotype, described 129 different variants (83 missense, 12 small deletions, 6 splice, 4 nonsense, 20 small indels and 4 large deletions), distributed across all exons but concentrated in exon 1 (33%) and exon 4 (25%), and affecting 76 of the 254 amino acids [9]. Of the 434 cases published at the time, 70% (305/434) were homozygous and 30% (129/434) compound heterozygous; consanguinity increases the proportion of homozygosity [9]. The text of the article names positions 196, 227, 235 and 246 as hotspots (25% of variants), whereas the abstract gives codons 55, 65, 196, 235 and 246, an internal discrepancy within the review itself [9]. The p.Arg246Gln variant has been reported in populations as diverse as Indian, Austrian, Brazilian, Italian, Korean, Pakistani, Dominican, Egyptian, Chinese, Saudi and Mexican, and p.Gly196Ser in Turkey, China, Bulgaria, Italy and the United Kingdom [9]. Other variants appear to be population-specific: p.Pro59Arg (Algeria), p.Asn160Asp (Egypt), c.188_189insTA (India), c.453del (Italy and Turkey), p.Gly183Ser (Brazil) and p.Ala65Pro (Turkey) [9]. In an international series of 55 patients followed in Montpellier, there were 33 different variants, five of them novel, with homozygosity in 69.1%, compound heterozygosity in 25.5% and, in 5.4%, compound heterozygosity only with the “V89L” polymorphism (historical numbering, = NM_000348.4:c.265C>G p.(Leu89Val)); exons 1 (35.8% of alleles) and 4 (21.7%) were the most affected [35]. In 2025, a British series estimated that more than one hundred pathogenic variants had been described and that familial homozygosity accounted for approximately 65% [17].

Table 2.1 standardises the variants most frequently mentioned in this article to MANE Select transcript NM_000348.4 and protein NP_000339.2, with historical designation and ClinVar classification at consultation (10 October 2026) [10, 36].

Table 2.1. Standardised SRD5A2 variants (NM_000348.4; GRCh38)

Some variants cited in the literature without a coding DNA-level description (p.Pro212Arg, frequent in families of Mexican ancestry; p.Gly34Arg, more frequent in Egypt; p.Glu200Lys in the United Kingdom; p.Pro59Arg; p.Asn160Asp) are retained here only in their published protein nomenclature [9, 17, 37]. Frameshift variant nomenclature differs between publications and transcript versions; the 2025 European series use NM_000348.3 [21]. Reports should therefore always state transcript, version and genomic description.

2.5. Non-coding variants, copy-number variation and atypical inheritance

Whole-gene and exon 1–2 deletions have been described; the SRD5A2 genomic region is rich in LINE-1 and Alu elements, and cases with a single identified variant and typical biochemistry suggest mechanisms outside the coding sequence [9]. The Papua New Guinea kindred has a deletion exceeding 20 kb, without enzyme activity, which was the first molecularly described alteration [15]. In 2026, the first large family with homozygous exon 1 deletion was described in Turkey, an example of copy-number variation (CNV) undetectable by simple Sanger sequencing [38]. Also in 2026, a Lübeck study described a 46,XY adult with clinically confirmed deficiency and compound heterozygosity: maternal c.692A>G p.(His231Arg) and paternal c.*66T>G in the 3' untranslated region; RT-PCR in genital skin fibroblasts showed only maternal allele expression, and long-isoform sequencing showed that c.*66T>G creates a splice donor 66 nucleotides after the stop codon, joined to an inverted Alu sequence 15 kb downstream; an exon–exon junction more than 55 nucleotides from the stop codon triggers nonsense-mediated decay of the paternal allele [39]. The authors conclude that variants outside the coding region may be causal and exon-focused panels and exomes may fail to interpret them, explaining some patients with a “missing second variant” [39]. Uniparental disomy has also been described: a patient born to non-consanguineous parents carrying p.Glu197Asp and p.Pro212Arg respectively was homozygous for p.Glu197Asp; two Chinese patients had three variants [9]. With a single heterozygous variant and suggestive biochemistry, possibilities include an undetected second variant (deep intronic, promoter) or, more rarely, a dominant effect [17].

2.6. Genotype–phenotype correlation

External genital severity is usually expressed by the Sinnecker scale (1 to 5, with sublevels, convertible to a 1 to 9 scale) or External Masculinization Score (EMS) [9, 20]. In the worldwide review, mean Sinnecker score was 5.44 ± 2.2 in homozygotes, 5.02 ± 2.1 in compound heterozygotes and 6.89 ± 1.1 in indel carriers (indels versus homozygotes p = 0.02 and versus compound heterozygotes p < 0.001); exon 4 variants, encoding the NADPH-binding domain, were associated with less virilisation (n = 156; F = 10.57; p < 0.001) [9]. Among recurrent homozygous variants, p.Arg246Gln was the most virilising, whereas p.Gly183Ser and p.Gln126Arg were consistently associated with severe undervirilisation; homozygous nonsense variants in Saudi Arabia (p.Arg227Ter, p.Arg103Ter) presented with hypospadias and microphallus [9]. Variants with no residual activity in vitro (such as p.Arg246Gln) are associated with variable virilisation, and intrafamilial variation exists, so the impact of residual activity on phenotype remains controversial [9]. In a Chinese cohort of 130 Han children, p.Arg227Gln homozygotes had mild variable phenotypes, p.Gln6Ter, p.Arg246Gln or p.Gly203Ser homozygotes consistently severe phenotypes, and biallelic nonsense genotypes lower EMS and meatal scores than missense genotypes (p = 0.009 and p = 0.024) [40]. In a Chinese cohort of 103 patients, absence of p.Arg227Gln was associated with female assignment, cryptorchidism, bifid scrotum and lower EMS and AMH [41]. p.Arg227Gln retains approximately 3.2% residual activity and was found in 3.7% of 81 prepubertal Japanese boys with isolated micropenis [18]. In Turkish series, p.Ala65Pro and p.Leu55Gln were more undervirilising than p.Gly196Ser, without statistically significant correlation [42]. Together, series demonstrate no strong correlation beyond exon 4, indels and the hypomorphic nature of p.Arg227Gln [9, 40].

2.7. Polymorphisms, haplotypes and modifiers

The most studied polymorphism is at codon 89. Literature calls it “V89L”, treating valine as the wild-type allele and attributing approximately 30% lower activity in vitro to leucine [9]. Reference transcript NM_000348.4 encodes leucine at this position, so the variant recorded in ClinVar is c.265C>G p.(Leu89Val), classified benign by multiple submitters [10, 36]. This variant does not cause 5α-reductase type 2 deficiency and should not count as a pathogenic allele in genotype interpretation. A meta-analysis of 9 case–control studies (1232 cases and 1616 controls) associated the leucine allele with 46,XY DSD and isolated hypospadias, with a dose effect (allelic OR 2.05, 95% CI 1.28–3.28; LL versus VV OR 3.38, 95% CI 1.51–7.56), corresponding to a low-penetrance risk factor rather than a Mendelian variant [43]. p.Ala49Thr increases activity approximately fivefold in vitro; an unconfirmed prostate cancer association was suggested, and it has conflicting ClinVar classifications; TA repeats (0 to 18) in the 3' untranslated region have an uncertain role [9, 36]. In southern China, a haplotype comprising c.265C>G and c.680G>A (thus Val89 and Gln227 on the same allele) had a frequency of 64.71% in 5α-RD2 patients versus 1.22–2.59% in other DSDs and 1.57% in 2794 controls; homozygotes mainly had microphallus, almost half isolated, and catalytic efficiency was intermediate, constituting a hypomorphic allele with a possible founder effect [44].

2.8. Gene–disease validity and allelic diseases

ClinGen has no gene validity or actionability assertion for SRD5A2; dosage curation (22 August 2012) assigns a haploinsufficiency score of 30, corresponding to “gene associated with an autosomal recessive phenotype” [2]. In GenCC, the disease relationship is classified strong (Labcorp Genetics; PanelApp Australia) and supportive (Orphanet), always with autosomal recessive inheritance; in Genomics England PanelApp the gene is green, with biallelic inheritance, in the “Differences in sex development” and “Fetal anomalies” panels [2]. The allelic spectrum includes mild phenotypes (isolated micropenis or hypospadias), particularly with p.Arg227Gln, and the minimal phenotype of homozygous 46,XX women [33, 45]. Association of the codon 89 polymorphism with isolated hypospadias is susceptibility rather than a Mendelian allelic disease [43].

2.9. 17β-HSD3 deficiency: molecular basis

HSD17B3 encodes 17β-hydroxysteroid dehydrogenase type 3, an NADPH-dependent Leydig cell-specific reductase converting 17-ketosteroids to 17β-hydroxysteroids, mainly androstenedione to T [7]. It is one of five isoenzymes capable of this conversion [21]. Biallelic loss reduces fetal testicular T and consequently DHT available to external genitalia and prostate; Wolffian derivatives may form, suggesting low T concentrations or paracrine secretion suffice for their development [7, 21]. The enzyme does not participate in adrenal steroidogenesis [16]. 9q22 deletions including the gene are also associated with ambiguous genitalia [11]. A 2022 review compiled 70 HSD17B3 variants in 239 patients from 187 families [7].

3. Clinical Spectrum and Manifestations by System

3.1. Genital phenotype at birth

In an affected 46,XY neonate, testes and T production are normal, but lack of DHT leaves external genitalia undervirilised [2]. The typical homozygous presentation includes a clitoral-appearing phallus, bifid scrotum, perineoscrotal hypospadias, a shallow blind vaginal pouch (from incomplete urogenital sinus closure) and rudimentary prostate, with normal epididymis, vas deferens, seminal vesicles and ejaculatory ducts [15]. Testes may lie in the labioscrotal folds, inguinal canals or, more rarely, abdomen [17]. The spectrum ranges from typical female external genitalia to isolated hypospadias or micropenis, with marked variability even among carriers of the same variant [1, 9]. Table 3.1 reproduces the Sinnecker scale used in many series.

Table 3.1. Sinnecker scale for external genitalia (equivalence on the 1–9 scale)

Source: Sinnecker scale (1996), reproduced from Batista and Mendonca [9].

Phenotypic distribution varies with population and recruitment. In the Montpellier international series (55 patients), 49.1% had clitoromegaly, 32.7% microphallus with varying hypospadias severity, 7.3% female external genitalia and 3.6% isolated micropenis [35]. A Turkish multicentre series of 85 patients had external genital abnormalities in 92.9% [42]. At Beijing Children's Hospital (December 2007 to May 2019), among 130 consecutive Han children with median age 1.4 years at first consultation, 113 had hypospadias and 17 a normal meatus with micropenis, isolated or with cryptorchidism; 41.54% had EMS ≤ 3 and 42.31% EMS > 6, and 76.92% had bilaterally descended gonads [40]. In a Birmingham series (United Kingdom, 14 patients), EMS ranged from 1 to 11 (median 6.5), 78% had at least one undescended testis and isolated micropenis was uncommon [17]. In Taiwan, 5 of 6 p.Arg227Gln homozygotes had isolated micropenis [45]. In a Chinese series of 16 patients diagnosed between April 2018 and July 2025, all had ambiguous genitalia at birth, 10 with micropenis and hypospadias and 6 with isolated micropenis [46].

3.2. Childhood

Without neonatal diagnosis, children raised as girls may present with clitoromegaly, inguinal swelling (testis) or vaginal adhesions, and those raised as boys with micropenis, hypospadias or cryptorchidism [47]. In a Turkish series of 29 patients (Adana), none was diagnosed prenatally or neonatally, median age at presentation was 71 months (6–201), consanguinity was present in 89.6% and mean prepubertal phallic length was 1.4 cm; complaints included genital ambiguity (15), family history (4), micropenis (2), masculine behaviour (2), inguinal swelling (2), amenorrhoea (2), vaginal adhesions (1) and short stature (1); a horseshoe kidney was identified [47]. The prepubertal genital phenotype is usually indistinguishable from partial androgen insensitivity (PAIS), particularly without consanguinity or a family history compatible with X-linked inheritance [48].

3.3. Puberty and adulthood without treatment

At puberty, increased T acting directly on AR and 5α-reductase type 1 activity produce virilisation: increased muscle mass, deep voice, libido, penile growth, scrotal rugosity and pigmentation and testicular descent [2, 15]. Facial and body hair remain sparse and temporal hair recession does not occur because these effects depend on DHT; the prostate is impalpable and approximately one-tenth of normal volume on magnetic resonance imaging; final height is similar to unaffected siblings and there is no gynaecomastia, unlike androgen insensitivity [15, 16]. Pubertal penile growth is frequently insufficient [18, 48]. In undiagnosed individuals raised as girls, pubertal presentation combines primary amenorrhoea, absent breast development, deep voice, clitoromegaly and hirsutism [11, 16]. In Papua New Guinea, affected adults had more facial and body hair than Dominicans, although less than unaffected siblings, reflecting familial variability or degree of deficiency [5]. A 65-year-old Dominican man without medical intervention had an android habitus and male identity [26]. In a Turkish series, median pubertal phallic length was 3.75 cm [47]. In the I-DSD registry, men aged 16 years or older had median testicular volume 16 mL (10–20; n = 4), and two had stage B2 gynaecomastia without previous hormone treatment [21].

3.4. Gonadal function, bone and fertility

With gonads in situ, puberty progresses spontaneously [21]. In 8 untreated individuals or individuals assessed before gonadectomy (median age 16 years), median T was 11.1 nmol/L (6.3–19.1); considering 17β-HSD3 and 5α-reductase deficiencies together, T exceeded 10 nmol/L in 5 of 6 individuals aged 16 years or older, but FSH and/or LH exceeded 10 U/L in 4 of 6, justifying surveillance for hypogonadism [21]. In a prospective observational dual-energy X-ray absorptiometry (DXA) study (1989–1999), 16 5α-RD2 patients had lumbar L2–L4 Z-score −0.84 (±1.29; p = 0.02), becoming −0.44 after weight adjustment (p = 0.14), and femoral neck Z-score 0.14, whereas 12 women with complete androgen insensitivity had reduced values (−2.84 and −1.33); the authors concluded that T, even with low DHT, suffices for bone mineral density [49]. Male fertility is reduced but not absent through four mechanisms: uncorrected cryptorchidism, failed spermatogonial transformation into spermatocytes, low semen volume from underdeveloped prostate and vesicles and failure of semen liquefaction because of absent prostate-specific antigen (PSA); ejaculate is scanty and gelatinous, and a short phallus and perineal hypospadias prevent intromission when uncorrected [15]. On biopsy, germ cell numbers may be normal with absent primary spermatocytes, suggesting a role for DHT in their differentiation; histology ranges from normal to absent germ cells [15]. In the I-DSD registry, review of gonadal tissue from a neonate showed normal germ cell numbers with regressive features, and from an 11-year-old child tubules with and without germ cells, Sertoli nodules, testicular angiopathy and separation of the rete testis from the epididymis, contradicting the notion of invariably normal Wolffian derivatives [21].

3.5. Homozygous 46,XX women

Three 46,XX women in the Dominican kindred, homozygous for p.Arg246Trp, had normal internal and external genitalia, reduced body hair, no acne, late menarche, ovulatory cycles, normal sebum (regulated by 5α-reductase type 1) and normal gonadotrophins; all were fertile and two had twins [15, 33]. Biochemically, they had normal to elevated T, low DHT, an elevated T/DHT ratio, low 3α,5α-androstanediol glucuronide and elevated urinary 5β/5α ratios of C19 and C21 steroids [33].

3.6. Gender identity: natural history and cohorts

Table 3.2 summarises cohorts with data on changes in gender role or identity. Numbers are not directly comparable: they differ in period, country, diagnostic timing, proportion of prepubertal gonadectomies, denominator (all affected individuals, only those raised as girls or only those reaching puberty) and what is measured (gender identity, social role or actual change of registration).

Table 3.2. Gender change in 5α-reductase type 2 deficiency: what each cohort measures

When change occurs, it is almost always female-to-male, mainly in childhood and adolescence, even after gonadectomy, and does not systematically depend on virilisation at birth (Sinnecker 5.68 ± 2.07 in those changing versus 5.97 ± 1.97 in those not changing; p = 0.37) [9, 50]. The 2016 gender assignment consensus update confirms a high rate of self-initiated change, adding that risk may be lower after prepubertal gonadectomy and that, in one study, women who did not change role did not always have typically female identities [54]. Social and cultural factors modulate change: in societies where female infertility prevents marriage, the social advantage of a male role influences the decision [4, 13]. Male assignment increased from 26.8% in cases diagnosed before 1999 to 42.8% after 1999 (p < 0.0001), and female-to-male reassignment became less frequent (p = 0.036) [9].

3.7. 17β-HSD3 deficiency: clinical spectrum

46,XY individuals with 17β-HSD3 deficiency have female or atypical external genitalia, no prostate and hypoplastic to normal Wolffian derivatives; most are assigned female and signs between birth and puberty may be mild or absent [7]. In the I-DSD registry, median external genital score was 2 (0.5–6), most gonads were inguinal at diagnosis and all 16 individuals were raised as girls [21]. At puberty, LH rises and marked virilisation occurs, often with gynaecomastia [7, 11]. Presentation is prepubertal (atypical genitalia, inguinal testes) or pubertal (primary amenorrhoea with virilisation) [7]. In a systematic review of 152 genetically confirmed probands, initial male assignment was uncommon (10.5%) and change to a male role occurred in 10.3% of those raised as girls, associated with Asian origin and not reported in patients of European origin [7]. In the I-DSD registry, 1 of 16 (6%) changed to male at 14 years [21]; in Cohen-Kettenis's review, change ranged from 39% to 64% [50]; in São Paulo, 5 of 15 adults assigned female changed to male [51]. Unlike 5α-RD2, germ cell numbers decline with age and spermatogenesis was absent in I-DSD registry samples [21].

4. Clinical Recognition and Warning Signs

4.1. Prenatal period

Discordance between chromosomal sex obtained by chorionic villus sampling, amniocentesis or cell-free fetal DNA and ultrasound or neonatal sex is most likely caused by maternal contamination or sample mix-up; true discordance is rare, and 5α-reductase type 2 deficiency and androgen insensitivity are among causes of a female phenotype with a 46,XY karyotype [11]. Severe hypospadias may be suspected prenatally from ultrasound findings (severe curvature or ambiguous genitalia) combined with a male karyotype [32]. In one published US family case, ultrasound at 20 weeks showed female genitalia and labioscrotal masses were seen at 36 weeks [37].

4.2. Neonates and infants

Specialist assessment is indicated for isolated perineal hypospadias, isolated micropenis, isolated clitoromegaly, familial hypospadias and combinations of abnormalities with EMS below 11 [55]. Warning signs for 5α-reductase type 2 deficiency are: palpable gonads in the labioscrotal folds or inguinal canals of a child with female-appearing genitalia, a blind vaginal pouch with separate urethral opening or a urogenital sinus, bifid scrotum with perineoscrotal hypospadias, micropenis with or without hypospadias and parental consanguinity or an affected sibling [15, 47, 48]. Birth weight should be recorded because fetal growth restriction is associated with hypospadias and 46,XY DSD without an identified monogenic cause [11, 14]. Sex assignment should not occur before specialist team assessment [13]. In a neonate with atypical genitalia, the immediate priority is excluding salt-wasting congenital adrenal hyperplasia (urgent electrolytes, glucose, 17-hydroxyprogesterone), a potentially life-threatening condition not caused by 5α-reductase type 2 deficiency [8, 11].

4.3. Childhood

In older children, suspicion arises with inguinal swelling or a testis found during hernia repair in a girl, progressive clitoromegaly, masculine behaviour, micropenis or proximal hypospadias [47]. In the Birmingham series, presentation ranged from the neonatal period to 13 years and the most common reason was ambiguous genitalia at birth (6 of 14, 43%) [17].

4.4. Adolescence

Pubertal clitoromegaly and hirsutism with primary amenorrhoea are the classical presentation of two 46,XY DSDs, 17β-HSD3 and 5α-reductase type 2 deficiencies, and also occur in SF-1 (NR5A1) deficiency; they are less typical of partial androgen insensitivity, which usually presents at birth [16, 55]. Müllerian structures are absent in all three; 46,XX congenital adrenal hyperplasia and androgen-secreting tumours enter the differential but have a uterus [16]. Gynaecomastia favours 17β-HSD3 deficiency or partial androgen insensitivity [7, 48].

4.5. Emergencies and diagnostic delays

5α-Reductase type 2 deficiency does not cause adrenal insufficiency, but initial assessment of any neonate with DSD should exclude congenital adrenal hyperplasia forms and StAR and P450scc defects that do [8]. Early diagnosis matters because sex assignment may depend on it [16]. Delays are frequent: in the worldwide review, only 29.7% had neonatal diagnosis [9]; in an Italian cohort of 24 patients, approximately 70% had a previous incorrect diagnosis and 8 children underwent gonadectomy before correct diagnosis [52]; at a Turkish centre, mean delay was 5.9 years and 27 consultations, and in the Italian cohort mean delay was 9.1 years [17]. The sources consulted describe no newborn screening programmes targeting this entity.

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Principles and criteria

There are no validated formal clinical criteria for 5α-reductase type 2 deficiency. Diagnosis rests on a 46,XY karyotype, testes, Wolffian derivatives present and Müllerian derivatives absent, undervirilised external genitalia and evidence of deficient T-to-DHT conversion, and is confirmed by biallelic pathogenic or likely pathogenic SRD5A2 variants [8, 16]. The European COST Action BM1303 (DSDnet) consensus establishes that everyone with suspected DSD requires extensive bodily and genital examination and biochemical, genetic and imaging investigation discussed by a multidisciplinary team, that the ultimate goal is molecular diagnosis, not achieved in approximately 50% of 46,XY DSD, and that diagnostic uncertainty should not delay management planning [20]. The 2021 UK guidance adds that, with rapid genetic testing available, SRD5A2 deficiency exemplifies a condition in which molecular genetics is replacing detailed biochemistry as the preferred diagnostic tool [16].

5.2. Initial clinical assessment

History should include a three-generation pedigree with consanguinity, maternal health and medication and pregnancy history; examination measures the phallus, describes labioscrotal folds, seeks gonads in the folds and inguinal regions (if palpable, probably testes or ovotestes), locates the urethral meatus, seeks other abnormalities and records birth weight [11]. Genitalia should be described using EMS; anogenital distance correlates with prenatal androgen exposure but is not standardised [20]. Pelvic and abdominal ultrasound seeks uterus and gonads; genitography or examination under anaesthesia defines the urogenital sinus [2, 11].

5.3. Biochemistry: methods and neonatal physiology

Steroids should be measured by liquid chromatography–tandem mass spectrometry (LC-MS/MS) or immunoassay after organic solvent extraction because direct immunoassays have specificity problems; laboratories should participate in external quality assessment [16, 55]. The DSDnet position recognises that mass spectrometry is superior but unavailable for all hormones or in all European regions, that automated immunoassays cross-react and are unreliable in neonates, and that 5α-reductase type 2 deficiency is diagnosed by serum T/DHT before and after hCG or the ratio of 5α- and 5β-reduced steroids in the urinary profile [56]. In boys, T and DHT may be high at birth, then fall below 1 nmol/L and to undetectable values respectively, rise from day 30 with a peak around day 70 (minipuberty) and return to prepubertal values [55]. Outside minipuberty and puberty, basal androgens are uninformative and hCG stimulation is required [15]. Normal postnatal T, DHT and LH concentrations do not prove they were normal during the critical fetal masculinisation period [14].

5.4. hCG stimulation test

The hCG test is an invasive second-line investigation to be performed late in the diagnostic pathway under regional centre guidance; serum AMH has high predictive value for the T response [16]. Table 5.1 summarises protocols described in the sources.

Table 5.1. hCG stimulation protocols in 46,XY DSD

In very young infants with an active axis, serial samples with elevated T may obviate testing [55]. Prolonged stimulation produces higher T peaks; low stimulated T lowers T/DHT and increases false negatives [17].

5.5. T/DHT ratio: thresholds and performance

Post-hCG T/DHT frequently exceeds 30:1, but several patients with lower ratios have been reported [16]. Proposed thresholds vary: a ratio above 8.5 was considered suggestive and a reason for molecular investigation [15]; in a Turkish series of 85 patients, the stimulated ratio was more sensitive than basal even at puberty, and thresholds with best sensitivity were 8.5 in minipuberty, 10 prepubertally and 17 at puberty [42]; the Birmingham series used 10 [17]. Table 5.2 reproduces the twelve-study review published in 2025, supplemented by later series.

Table 5.2. Stimulated T/DHT performance in 5α-reductase type 2 deficiency series

Sources: the first twelve rows reproduce Table 2 of Balagamage and colleagues, citing original studies [17]; remaining rows derive from [34, 38, 47, 57]. Fan, Bertelloni, Akcay, Veiga-Junior and Maimoun data were also compared with the original publications [35, 40, 48, 52, 58].

This evidence has several limitations. Sensitivity of a stimulated ratio above 10 is below 80% in several series; there are no age- and sex-specific reference values; protocols and assays vary; immunoassays are particularly imprecise outside minipuberty and puberty; the ratio overlaps androgen insensitivity and therefore lacks specificity; and pubertal 5α-reductase type 1 expression lowers the ratio [17, 48, 58]. The ratio is not associated with phenotypic severity or age [17, 40]. In an Indonesian cohort (66 suspected cases and 95 relatives; reference = sequencing), the T/DHT ROC area under the curve was 57.7% (95% CI 43.0–72.4; non-significant) [59]. A below-threshold ratio therefore does not exclude diagnosis, and sequencing should be performed with clinical suspicion [47, 57].

5.6. Urinary steroid profile

GC-MS urinary steroid profiling is qualitative and quantitative, and samples may be frozen for years [55]. It is unsuitable before 3 months because diagnostic 5β- and 5α-reduced metabolite pairs become detectable only from that age; neonatally, fetal adrenal 11-oxocortisone metabolites complicate interpretation [16, 17]. In infants older than 3 to 6 months, deficiency is identified in a urine sample through a reduced 5α:5β ratio of C21 and C19 steroids, possible even after early gonadectomy [16]. Urinary metabolites fluctuate throughout the day, so 24-hour urine should be considered when suspicion is high [16]. Performance studies are small but consistent: in Birmingham, the urinary profile suggested deficiency in all 13 tested patients, including both with T/DHT below 10, with reduced 5α:5β ratios particularly for tetrahydrocortisol and tetrahydrocorticosterone [17]; in Hong Kong (January 2003 to July 2012), 15 tested patients had low 5α/5β ratios and 12 of 12 genotyped patients had two variants, whereas 2 of 4 with DHT measured did not reach threshold [60]; in Hanoi (2017–2021), random urinary 5α-tetrahydrocortisol/tetrahydrocortisol was low in 25 patients aged 6 months to 13 years without overlap with 104 controls, with 100% sensitivity and specificity at threshold 0.19 in that sample [61]; in Indonesia, the etiocholanolone/androsterone ROC area under the curve was 79.7% (95% CI 69.0–90.4; p < 0.001), with proposed threshold ≥ 0.95 [59]. In the Turkish village studied in 1986, tetrahydrocortisol/5α-tetrahydrocortisol was most discriminating (adult homozygotes 35.80 ± 20.10; children 15.48 ± 7.91; heterozygous parents 4.56 ± 1.61; controls 1.07 ± 0.36) [62]. Urinary profiling has not been systematically tested in large samples, but the 2025 British series proposes it as first-line biochemistry because it is more accurate and less invasive than hCG testing [17].

5.7. Genetic diagnosis

The European strategy begins with chromosomal complement (QF-PCR in 1–2 days and karyotype with mosaicism screening, or microarray, also detecting SRY and CNVs), followed in 46,XY individuals by CNV analysis when extragenital malformations exist (CNVs in approximately one-fifth of DSDs) [8, 12]. When the steroid profile points to a gene, targeted sequencing is undertaken (SRD5A2 belongs to this group with CYP21A2, CYP11B1, HSD17B3 and AR); in other cases phenotyping, biochemistry and genetics proceed in parallel, preferring a high-throughput sequencing panel or exome, reserving genome sequencing for novel genes or oligogenic bases, with variant classification following ACMG/AMP (American College of Medical Genetics and Genomics / Association for Molecular Pathology) [8]. In the United Kingdom, DSD testing is in the National Genomic Test Directory (R146.2 panel, including SRD5A2), and interpretation is performed by a diagnostic group with a clinical geneticist, molecular geneticist, clinical biochemist and paediatric endocrinologist [16, 17]. Pretest counselling should address variants of uncertain significance (VUS), exome and genome coverage limitations and unsolicited findings [16]. In December 2025, the Chinese Medical Association's Society of Paediatrics published the first Chinese expert consensus on DSD genetic testing and counselling, with testing indications, report interpretation and counselling [63].

Yield depends on population and method. In an Indian cohort of 147 children with 46,XY DSD (median age 3.8 years), sequential Sanger sequencing of SRD5A2 and AR in 75 suspected cases found pathogenic or likely pathogenic variants in 44, and a 155-gene panel in the remaining 103 found 20 more; overall yield was 46% (68/147), with SRD5A2, AR and NR5A1 most implicated [64]. In Iraqi Kurdistan, microarray and exome established diagnosis in 12 of 21 individuals (57%), 3 with SRD5A2, and all 6 with homozygosity regions exceeding 3% of the genome had a diagnosis [65]. In Turkey, among 51 children clinically diagnosed with androgen insensitivity and stimulated T/DHT below 25, AR variants were identified in 11 and SRD5A2 in 6 (12%), without hormonal differences between groups [58]. In Japan, AR and SRD5A2 amplicon sequencing from dried blood spots produced results in 19 of 22 patients, but a false homozygote due to allelic dropout required Sanger validation [66].

5.8. Test limitations and report content

A negative result or a single variant does not exclude diagnosis. Sanger sequencing and many panels do not detect exon or whole-gene deletions, such as the Turkish exon 1 deletion or Papua New Guinea deletion [15, 38]; exome sequencing may detect exonic CNVs undetected by comparative genomic hybridisation microarray or multiplex ligation-dependent probe amplification (MLPA), and genome sequencing may detect non-coding CNVs [8]. 3' untranslated or deep intronic variants require RNA studies [39]. Allelic dropout may mimic homozygosity [66], and apparent homozygosity requires consideration of a deletion on the other allele or uniparental disomy, with parental segregation confirmation [9, 12]. Proportions of variant types in reviews do not equate to clinical test sensitivity, not established for SRD5A2 in the sources consulted; panel genes vary between laboratories [12]. The report should state gene, versioned transcript (NM_000348.4), DNA- and protein-level HGVS description, genomic coordinate, zygosity, variant phase in compound heterozygosity (confirmed by parental studies) and ACMG/AMP class, and treat c.265C>G p.(Leu89Val) as a benign polymorphism [8, 36]. If molecular diagnosis is not achieved, DNA storage is recommended [12].

5.9. Practical algorithm

Table 5.3. Proposed diagnostic sequence derived from the sources

5.10. Prenatal and preimplantation diagnosis

After familial variants are identified, prenatal diagnosis and preimplantation genetic testing are possible; the ideal counselling time is before pregnancy, and views on using these tests vary among families and professionals [12]. Details are in section 8.

6. Differential Diagnosis

6.1. Comparative table

Table 6.1. Differential diagnosis of 46,XY DSD with testes and undervirilisation

6.2. 17β-HSD3 deficiency

17β-HSD3 deficiency most closely resembles 5α-reductase type 2 deficiency, sharing autosomal recessive inheritance, absent Müllerian derivatives, frequent female assignment and pubertal virilisation with primary amenorrhoea [7, 16]. It differs in absent prostate, frequent pubertal gynaecomastia, a more pronounced LH rise and biochemistry [7, 11]. The enzyme does not affect adrenal steroidogenesis: plasma steroids show elevated androstenedione with low T, particularly after hCG; the British 46,XY DSD table adds increased plasma oestrone, reduced T/androstenedione ratios and oestradiol after hCG and increased FSH and LH [16]. In prepubertal children basal values are uninformative and T/androstenedione should be measured after hCG; basal values are informative in early infancy and adolescence [7]. Under criteria cited in a Brazilian series, 17β-HSD3 deficiency is excluded in infants under 6 months when basal T/androstenedione exceeds 0.8 (100% sensitivity) and prepubertally when post-hCG ratio exceeds 0.8 (90% sensitivity) [48]. A low ratio also occurs in gonadal dysgenesis and its standalone reliability is uncertain; urinary androsterone and etiocholanolone increases may appear only after puberty [16]. Birth prevalence was estimated at 0.68 per 100,000 in the Netherlands [6]. In a cohort of 150 46,XY DSD patients from western India, 17β-HSD3 deficiency was the third cause of non-dysgenetic 46,XY DSD, accounting for 6% (clinical proportion, not prevalence) [7]. Some individuals with a partial androgen insensitivity phenotype have biallelic HSD17B3 variants [14]. In the 2006 consensus, germ cell tumour risk in 17β-HSD deficiency was classified intermediate (28%, 2 studies, 7 patients, recommended action “monitor”) [13]; in the I-DSD registry, 13 of 16 individuals underwent gonadectomy, without germ cell (pre)malignancy in reviewed samples, and germ cell numbers declined with age [21]. These data led the authors to suggest deferring gonadectomy until late adolescence in this entity too, with reversible pubertal suppression when needed (section 7) [21]. Fertility potential in 17β-HSD3 deficiency was considered unknown in the 2006 consensus [13].

6.3. Partial androgen insensitivity

Before puberty, 5α-reductase type 2 deficiency and partial androgen insensitivity (PAIS) are generally indistinguishable, particularly without consanguinity or a family history compatible with X-linked inheritance [48]. At puberty, PAIS involves gynaecomastia, limited genital virilisation and abnormally elevated LH and T, whereas 5α-reductase deficiency involves genital virilisation (not always adequate penile growth), without gynaecomastia and with an absent or hypoplastic prostate [48]. In the Campinas series (2001–2010), consanguinity was more frequent in 5α-reductase deficiency (p = 0.00003), familial recurrence more frequent in PAIS and genital severity greater in 5α-reductase deficiency; radioimmunoassay T/DHT did not distinguish them (28–82 versus 32–72) [48]. In Turkey, hormonal data in children with AR, SRD5A2 or no variant did not differ, and the authors recommend SRD5A2 sequencing in “PAIS” without an AR variant [58]. PAIS should be used only with a confirmed AR variant; AR sequencing identifies a variant in more than 90% of complete forms but only approximately 20% of partial phenotypes [55]. Functional androgen sensitivity assessment (T or DHT applied to the phallus, systemic T, SHBG reduction) has no consensus on androgen, dose, route, timing, duration or response definition [16]. Wolffian structures suggest 5α-reductase deficiency, a partial biosynthesis defect or PAIS; elevated T precursors suggest a biosynthesis defect, and Müllerian derivatives suggest gonadal dysgenesis [14].

6.4. Aldo-keto reductase deficiency and other biosynthesis defects

AKR1C2 and AKR1C4 deficiency impairs the alternative testicular fetal DHT synthesis pathway; it produces varying atypical genitalia, including female genitalia, resembling 5α-reductase type 2 deficiency, but pubertal virilisation is not expected; biochemistry shows DHT deficiency and apparent 17,20-lyase deficiency with normal CYP17A1 and SRD5A2, and the entity's clinical significance is debated [16]. Biosynthesis defects affecting the adrenal gland (StAR, P450scc, 3β-HSD2, P450c17, POR) differ through adrenal insufficiency, hypertension or accumulated precursors [12, 16]. In POR deficiency, common variants p.Ala287Pro and p.Arg457His, craniosynostosis and radiohumeral synostosis and the characteristic urinary profile allow distinction [67, 68].

6.5. Isolated hypospadias and micropenis

In a series of 33 patients with severe hypospadias, a cause was identified in 12, one with partial 5α-reductase deficiency; in a series of 63 children with severe hypospadias, an aetiology was identified in approximately one-third [32]. Boys with moderate or severe hypospadias should undergo hCG stimulation for T and DHT measurement, and AR or SRD5A2 molecular analysis should be considered when endocrine studies are abnormal [32]. In apparently isolated micropenis, particularly in East Asian populations, 5α-reductase type 2 deficiency with p.Arg227Gln should be excluded before treatment, because systemic T fails when its penile effect depends on local DHT conversion [18, 45]. The codon 89 polymorphism is a hypospadias risk factor, not a Mendelian cause [43].

6.6. Pharmacological phenocopy

Finasteride, a 5α-reductase type 2 inhibitor, is contraindicated in women who are or may become pregnant because it may cause male fetal external genital abnormalities; in rats it caused dose-dependent hypospadias in 3.6–100% of male offspring, with lower prostate weight and anogenital distance, and pregnant women should not handle broken or crushed tablets [69]. Dutasteride, inhibiting type 1 and type 2 isoenzymes, is also contraindicated in pregnancy, should not be handled by pregnant women, and treated men should not donate blood until 6 months after the last dose [70]. Maternal exposure history should be elicited when assessing atypical genitalia in a 46,XY neonate.

7. Treatments and Risk Management Strategies

7.1. General principles and organisation of care

Enzyme activity cannot be restored; treatment is hormonal, surgical and psychological, following the general DSD framework without a gene-specific protocol [2]. Assessment and follow-up should occur at a centre with an experienced multidisciplinary team (genetics, paediatric and adult endocrinology, urology or paediatric surgery, gynaecology, psychology, nursing, biochemistry), integrated in clinical networks with audit and national standards [12, 13, 16]. Initial communication with parents is crucial because first impressions persist; families need time and opportunities for repeated discussion, open communication, active participation in decisions and strict confidentiality [12]. The 2016 international update (Global DSD Update), consulted only as an abstract, defines the goal as the best possible quality of life, with ongoing open communication and full disclosure, and emphasises that individual gender development remains impossible to predict with certainty and childhood assignment and surgical decisions altering genital anatomy or removing germ cells carry risk [71].

7.2. Sex and gender assignment

All children should receive a sex assignment, not before expert assessment [12, 13]. Factors to weigh are diagnosis, genital appearance, surgical options, lifelong replacement therapy needs, fertility potential, family opinion and sometimes cultural practices [13]. Specifically in childhood-diagnosed 5α-reductase type 2 deficiency, the 2006 consensus recommends discussing male identity in most cases and documented fertility potential as evidence for assignment [13]. A 2014 clinical synthesis adds that initial male assignment should be considered with understanding of expected pubertal events, and many affected men live in long-term heterosexual relationships [15]. A worldwide review advocates assignment based on molecular diagnosis [9]. US surveys of paediatric endocrinologists and urologists (2003–2004, 2010–2011 and 2019–2020) show that recommending male rearing became increasingly frequent in micropenis and PAIS scenarios, and that in 2019–2020 a minority offered no binary gender assignment [72]. Assignment remains difficult in the most undervirilised phenotypes: in a series of five US patients, consensus was equivocal or favoured male sex in the more virilised, male assignment avoided lifelong hormone therapy and preserved fertility potential, and female assignment was anatomically more consistent in the most undervirilised [37]. Psychological gender identity assessment is mandatory before discussing reassignment when the child is sufficiently developed; gender identity begins forming before 3 years, the rule fixing 18 months as the reassignment limit should be viewed cautiously, and homosexual orientation or cross-gender interests do not themselves indicate incorrect assignment [13]. The 2016 gender assignment update emphasises the lack of evidence-based recommendations with genital staging cut-offs and a tendency to view gender as a continuum [54]. DSDnet distinguishes gender role behaviour from core gender identity, considers distress the criterion for clinical concern and recommends explaining that initial social role decisions may change with the child's self-expression, with gender contentment the goal [20].

7.3. Topical dihydrotestosterone

Topical DHT is used in individuals raised as boys to increase penile length and, according to UK guidance, may assess long-term genital virilisation potential in infants; no guideline gives a dose and no controlled trial has defined a regimen [2, 16]. Systemic T is unnecessary because production is normal and has little penile effect, growth depending on locally formed DHT: in Japanese patients previously treated with testosterone enanthate, gain was 0.2 to 0.6 cm [15, 18].

Product. The only product identified in the sources is ANDRACTIM® (androstanolone, i.e. DHT, 2.5 g per 100 g gel; Besins Healthcare), nationally authorised in France on 23 January 1981, active and marketed (ATC code G03BB02; 30 g and 80 g tubes) [19]. The summary of product characteristics (SmPC) approves it in men for general androgen deficiencies (confirmed organic or functional hypogonadism) and, as local treatment, for lichen sclerosus and male gynaecomastia and vulvar lichen sclerosus in women; it has no paediatric, micropenis or DSD indication or paediatric dose [19]. Use in 5α-reductase type 2 deficiency is therefore off-label. No FDA-approved DHT application or centrally authorised DHT medicine in the European Union was identified; in Japan, the gel is not officially or commercially available [18, 73].

Table 7.1. Topical DHT dosing: label and literature

In the Japanese series, a 10–11-year-old patient treated with 25 mg/day for 8 weeks increased from 2.2 cm (−3.1 SD) to 3.2 cm (−1.9 SD) and, after another 8 weeks, to 5.0 cm (+0.3 SD) [18]. Puberty started at 11–12 years with normal growth, testicular enlargement and pubic hair, but delayed penile growth; a second DHT course at 18, 16 and 12 years did not produce sufficient catch-up, and adults had lengths of 6.0 (−3.4 SD), 6.4 (−3.0 SD) and 7.0 cm (−2.4 SD), T/DHT between 16 and 21 and male identity, some with anxiety about penis size [18]. The authors conclude that DHT effect is age-dependent and postpubertal treatment has limited effect [18]. In the European I-DSD registry, 4 patients received DHT (2 from 0.8 years and 2 at 15 and 17 years, after changing to male gender), without doses stated [21]. An Indian report describes DHT gel response in an affected child, without a dose in the abstract [75].

Adverse effects and monitoring. In the Japanese series, one patient developed local rash and pruritus; there was no effect on bone maturation or lipid profile; serum DHT during treatment was 0.20 to 1.1 ng/mL and LH and FSH were suppressed in all, leading the authors to suggest intermittent administration [18]. In the Chinese trial there were no significant adverse effects other than elevated serum DHT, but 3 children (13%) stopped at 3 months because of parental anxiety about adverse effects [74]. The SmPC reports androgenic hyperactivity (requiring dose reduction), frequent increases in haematocrit, erythrocytes and haemoglobin, monitoring of haemoglobin, haematocrit, liver function and lipids during prolonged treatment, caution in severe cardiac, hepatic or renal failure, ischaemic heart disease, hypertension and thrombophilia, interaction with vitamin K antagonists (increased INR), contraindication in prostate cancer and in women receiving general androgen therapy, positive doping tests and alcohol content (ethanol 2.44 g per 5 g dose; stinging on damaged skin; flammable until dry) [19]. The SmPC states that the medicine should not be used in pregnancy or breastfeeding [19].

7.4. Systemic testosterone and pubertal induction

In 46,XY DSD generally, a short childhood T course is used for micropenis (stretched penile length below −2.5 SD for age) and T is often needed to initiate and maintain puberty [12]. In 5α-reductase type 2 deficiency with gonads in situ, puberty is spontaneous and endogenous T normal or elevated, so exogenous T is unnecessary for puberty and has little penile efficacy [15, 18, 21]. The sources consulted give no T doses for this entity. After gonadectomy, lifelong hormone replacement is necessary and chosen according to gender (oestrogens to induce breasts and female puberty) [12, 13]; the sources consulted do not detail regimens specific to 5α-reductase deficiency.

7.5. Reversible pubertal suppression

In adolescents raised as girls without a stable gender identity, a gonadotrophin-releasing hormone agonist (GnRHa) may prevent virilisation and defer gonadectomy; in the I-DSD registry, one 5α-reductase-deficient patient received GnRHa and one girl was receiving GnRHa with oestradiol from 14 years while awaiting gonadectomy at 18, and in 17β-HSD3 deficiency one patient decided to change to male gender during treatment [21]. This resembles use in adolescent gender dysphoria, but DSDs were not listed as an indication in the 2019 GnRHa update and no long-term data exist in this population; the authors recommend monitoring growth, bone mineral density and blood pressure [21]. When sex of rearing differs from gonadal sex, contrasexual puberty should be monitored, and hormonal suppression with desired hormone replacement may avoid gonadectomy until the age of consent [12]. The sources give no specific drug, dose or duration.

7.6. Surgery

Principles. Surgery should be performed only by DSD-trained surgeons, explaining the full surgical sequence and lifelong consequences and emphasising functional outcome [13]. Many operations are not clinically necessary; deferral should be considered to allow the individual's participation, particularly with female rearing, avoiding irreversible procedures and tissue removal whenever possible; there are no controlled technique trials or consensus on timing [12]. DSDnet recommends deferring sensitive or irreversible procedures until the individual can participate in the decision, recording complications, urinary and sexual function, appearance, quality of life and psychosexual functioning, with validated masculinising surgery instruments (HOPE, HOSE, PPS, SAGAS-M) and no equivalent feminising instruments [20]. In 2026, Endo-ERN's MTG7 thematic group published expert opinion recommending deferral of non-urgent surgery with irreversible effects until the person can consent, stronger family support and a European framework with quality indicators [76].

Masculinising surgery. This includes vaginal pouch resection, urethroplasty, orchidopexy and scrotoplasty, frequently staged [12, 15]. Moderate or severe hypospadias repair is usually staged, generally between 12 months and 2 years or after toilet training at 3–4 years, with complications requiring reoperation in 15–30%; circumcision should be avoided because the foreskin may be needed [32]. The 2006 consensus envisaged standard hypospadias techniques with judicious testosterone and required initial counselling to consider possible adult phalloplasty [13]. Long-term outcomes are heterogeneous: in São Paulo, men changing female-to-male (n = 24, all aetiologies) had sexual parameters similar to those assigned male at birth (n = 20), with more frequent active sexual life (96% versus 75%; p = 0.04) [51]; in an Indian series of 12 adult men after childhood masculinising genitoplasty (2 with 5α-reductase deficiency), all identified as men, 9 urinated standing, 10 (83.3%) were concerned about penis size and 8 (66.7%) had severe erectile dysfunction [77].

Feminising surgery. This includes clitoroplasty, vaginoplasty and urogenital sinus mobilisation with vaginal dilatation [12]. Clitoral surgery may impair orgasmic function and erectile sensitivity, and vaginoplasty carries risks of scarring, repeated revisions and neoplasia [2]. In São Paulo, 46,XY women with atypical genitalia at birth (n = 35) did not differ from women with typical genitalia (n = 60) in sexual parameters after completed surgery; among operated women, 88.6% underwent clitoral reduction, 8% vaginoplasty and 28.6% labioplasty, and 37.1% vaginal dilatation before sexual activity [51].

Legal framework. In Germany, a May 2021 law protects children with DSD against genital surgery without their consent, with emergency exceptions or family court approval supported by an interdisciplinary commission opinion; ten centres assessed 78 patients (55% 46,XY, mainly undervirilisation or proximal hypospadias), with reimbursement and assessment criteria unresolved [78]. In Washington State (USA), hospital policy prohibits parental consent for removing minors' reproductive organs without a court order, except with oncogenic risk or inevitable infertility [37]. Intersex activists have called for a ban on irreversible elective genital surgery without consent, and alternatives to surgery (parental support, psychological adjustment) are underdeveloped [20].

7.7. Gonads and germ cell tumour risk

The 2006 consensus recommended prepubertal gonadectomy in androgen biosynthesis defects reared female, but its own risk table placed 5α-reductase deficiency in the lowest-risk group (0%, action “unresolved”, one study with 3 patients) [13]. Absence of tumours in 3 patients does not demonstrate zero risk. Later data continue to suggest low risk: 14 Turkish patients with bilateral gonadectomy had no malignancy [42]; in the I-DSD registry, 6 gonadectomised patients (median 8 or 10 years, depending on article section) had neither germ cell neoplasia in situ (GCNIS) nor cancer, and neither did 8 with gonads in situ (14–25 years), although seminomas have been described in two deficient patients aged 23 and 49 years [21]. DSDnet notes that risk is much higher in gonadal dysgenesis than hormonal synthesis or action disorders and higher in abdominal than inguinal or scrotal gonads, that neither tumour markers nor imaging reliably detect precursor lesions, and that considering gonadal biopsy in most 46,XY individuals with retained gonads seems prudent, ideally in late adolescence [20]. I-DSD registry authors conclude that data support deferring the gonadectomy decision until late adolescence while careful gender exploration with a specialist psychologist allows the individual to decide [21]. The 2016 gender update acknowledges that patient-initiated change may be less likely after prepubertal gonadectomy, which does not alone justify surgery [54]. Potentially functional gonads may be retained with surveillance [12].

7.8. Psychological support and information

Early support by a psychologist experienced in DSD is recommended, through multiple conversations at the family's pace [55]. The child should be informed early and progressively, with psychological and peer support at all ages [20]. Emotional and non-urgent decisions such as surgery should be deferred until psychological counselling has occurred [20]. Adolescent diagnosis may precipitate severe psychological distress and warrants psychological and psychosexual support referral [79].

7.9. Regulatory status

Table 7.2. Regulatory status of relevant drugs

The sources consulted identify no medicine specifically authorised for 5α-reductase type 2 deficiency or ongoing regulatory processes.

7.10. Clinical trials and therapeutic evidence

Table 7.3. Intervention studies and available therapeutic evidence

There are no randomised or controlled trials in this disease. Available studies are open-label, uncontrolled, with small samples and mixed populations, and do not permit regimen comparisons or estimates of long-term effects or optimal dose [18, 74]. Registered DHT gel trials (0.7% gel in adult men) concern prostate rather than DSD [80]. Negative registry searches do not prove absence of studies in other registries or unpublished studies.

7.11. Developments in the last twelve months

Between October 2025 and October 2026, research focused on diagnosis and care frameworks, without new treatments. Molecular developments described Alu exonisation through a 3' untranslated region variant (human study with RNA analysis) [39], a hypomorphic haplotype with possible founder effect in southern China (case–control study with functional characterisation) [44], predominant homozygous p.Arg227Gln with isolated micropenis in Taiwan (clinical series) [45], the first large family with homozygous exon 1 deletion (clinical series) [38], a Chinese series of 16 patients with a novel variant (clinical series) [46] and a codon 89 polymorphism meta-analysis [43]. In diagnosis, an Indian cohort of 147 children and an Iraqi Kurdistan cohort confirmed SRD5A2 among the most frequent 46,XY DSD genes [64, 65], a Turkish family of four siblings illustrated discordant T/DHT ratios with the same genotype [57], and the Chinese Society of Paediatrics published expert consensus on genetic DSD testing (consensus, December 2025) [63]. In care frameworks, Endo-ERN published expert opinion on genital surgery (February 2026) [76], first data on application of Germany's 2021 law were published [78], and an Indian series of functional and psychosexual outcomes in adult men [77]. None of these works is an authorisation or regulatory opinion.

7.12. Agents to avoid, pregnancy, perioperative care and transition

5α-Reductase inhibitors (finasteride, dutasteride) would worsen the enzymatic block and have no treatment role; pregnant women should not handle them [2, 69, 70]. Systemic T does not replace DHT for penile growth [18]. Androstanolone gel should not be used by pregnant or breastfeeding women [19]. The sources consulted describe no specific anaesthetic risks; absence of adrenal insufficiency distinguishes perioperative care here from other DSDs with adrenal defects [8]. Transition to adult care should follow a predefined plan with early adult specialist involvement and a personal clinical history summary [20]; adults with DSD appear to have more health problems, with morbidity linked to sex steroid deficiency or treatment overrepresented [20].

7.13. Guideline recommendations and divergences

Table 7.4. Recommendations from main guidelines and consensus documents

The main divergence opposes the 2006 recommendation of prepubertal gonadectomy in individuals raised as girls to later European documents prioritising deferral of irreversible decisions, reflecting evolving evidence on low tumour risk and frequency of gender change [13, 20, 21]. None of the sources consulted defines quantitative sex assignment criteria or DHT doses.

8. Reproductive Options and Family Counselling

8.1. Inheritance and recurrence risks

5α-Reductase type 2 deficiency is autosomal recessive, with genital expression limited to a 46,XY karyotype [2, 12]. Proband parents are presumed heterozygous, requiring confirmation by parental molecular testing [12]. Each sibling has a 25% probability of being biallelic, 50% of being a carrier and 25% of being a non-carrier; only biallelic 46,XY siblings have a genital phenotype, while biallelic 46,XX sisters have a subtle phenotype (reduced hair, late menarche) and are fertile [12, 33]. For a carrier couple, the risk of an affected 46,XY child in each pregnancy is the product of biallelic genotype probability (1/4) and 46,XY karyotype probability (1/2), i.e. 1/8. If only one parent is a carrier and paternity is confirmed, a de novo variant or parental mosaicism should be considered; apparent homozygosity requires exclusion of an undetected deletion or uniparental disomy [9, 12]. Parents' siblings have a 50% probability of being carriers, and carriers are usually asymptomatic [12].

8.2. Offspring of affected individuals

A homozygous 46,XX woman obligatorily transmits a pathogenic allele to every child [33]. An affected man with offspring similarly transmits an allele to every child, who will be carriers if the other parent is not a carrier; in a Dominican case, sequencing found no SRD5A2 variants in the partner [15]. Highly consanguineous series (89.6% of Adana patients) illustrate the context in which a related partner is more likely to be a carrier [47].

8.3. Cascade testing and carrier detection

Genotyping is the reference method for cascade testing after familial variants are identified [12]. Biochemistry is imperfect for carrier detection: in the Turkish village studied in 1986, parents had urinary 5β/5α ratios intermediate between affected individuals and controls [62], and in Indonesia etiocholanolone/androsterone ≥ 0.99 identified carriers with an area under the curve of 75.1% [59]. Counselling should be offered to affected or carrier young adults, ideally before pregnancy [12]. The sources consulted do not specifically address carrier testing in asymptomatic minors.

8.4. Prenatal and preimplantation diagnosis

After familial variants are identified, prenatal diagnosis by chorionic villus sampling or amniocentesis and preimplantation genetic testing are possible; views on their use differ among families and professionals [12]. Cell-free fetal DNA sex determination combined with ultrasound may reveal genotype–phenotype discordance, requiring technical causes to be excluded before considering DSD [11]. In the US family described by Byers and colleagues, the second affected sister had female genitalia on 20-week ultrasound and labioscrotal masses at 36 weeks [37].

8.5. Fertility and paternity in affected men

Fertility is markedly reduced in almost all DSDs, but assisted reproduction offers prospects to some; microdissection testicular sperm extraction (micro-TESE) is an option in azoospermic men if sperm are present on biopsy, early tissue cryopreservation is available to a limited number of individuals and in vitro maturation and induced pluripotent stem cell-derived gametes are experimental [20]. Fertility information should be given progressively before adolescence, and gonadal surgery decisions should weigh experimental fertility preservation, germ cell cancer risk and unwanted hormone production, discussing alternatives such as adoption [20]. In 5α-reductase type 2 deficiency, histology does not suggest rapid germ cell decline (unlike 17β-HSD3 deficiency), but spontaneous or assisted paternity is rarely reported, and rete testis and epididymal abnormalities may contribute to infertility; there are no TESE success data in this entity [21]. The first paternity through intrauterine insemination with an affected man's sperm was published in 1997 [81].

A staged approach is proposed: with normal or near-normal concentration and motility, try intrauterine insemination first; with severe oligoasthenoteratozoospermia, low volume and absent liquefaction, use in vitro fertilisation with intracytoplasmic sperm injection (ICSI); with azoospermia, testicular microdissection [15]. Genitalia should be corrected before assisted reproduction, and samples prepared at a centre experienced in semen without prostatic enzymes, with judicious partner gonadotrophin use and psychosocial and genetic support [15]. Two Dominican cases illustrate these options: a p.Arg246Trp homozygous man, with testicular volumes 25 and 20 mL and very viscous semen of 0.2–0.5 mL, had three healthy children through natural-cycle intrauterine insemination after Percoll gradient preparation; another, with 0.1 mL semen and motility below 1%, had twins through ICSI after partner ovarian stimulation with recombinant FSH 200 IU daily and hCG 5000 IU [15]. In São Paulo, two 5α-reductase-deficient men had children after intravaginal insemination [51]. In a cohort of 127 adults with 46,XY DSD at the same centre, 79% of women and 82% of men wanted children, and fear of rejection because of infertility was greater in women (32% versus 16%; p = 0.01) [51].

8.6. Counselling in 17β-HSD3 deficiency

Inheritance is likewise autosomal recessive with expression limited to a 46,XY karyotype [12]. Fertility potential was considered unknown in the 2006 consensus [13]; in the I-DSD registry, germ cell numbers declined with age and spermatogenesis was absent, with implications for still-experimental testicular tissue cryopreservation, although one paternity case has been reported [21].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence and what each number measures

There is no validated population prevalence estimate for 5α-reductase type 2 deficiency: Orphanet records “unknown” point prevalence based on expert opinion [1]. The 2020 worldwide review counted 434 published cases with biallelic genotypes from 44 countries, measuring publication and molecular diagnostic access rather than disease frequency [9]. Between 1966 and 2014, prevalence of women with a 46,XY karyotype was 6.4 per 100,000 female live births (4.1 per 100,000 with androgen insensitivity and 1.5 per 100,000 with gonadal dysgenesis), without a separate 5α-reductase deficiency estimate [8]. For comparison, hypospadias occurred in 6.4 per 1000 male live births in north-eastern France [32], and androgen insensitivity is estimated at 1 per 20,000 live births [79]. For 17β-HSD3 deficiency, Orphanet validates a birth prevalence of 0.68 per 100,000 in the Netherlands, not transferable to populations with frequent consanguinity [6, 7]. Carrier frequency in founder populations was not quantified in the sources consulted. Lack of estimates in the sources does not demonstrate absence of published data.

9.2. Geographical distribution of published cases

Table 9.1. Published cases with biallelic genotypes by country (worldwide review up to 2019)

Source: Table 1 of Batista and Mendonca [9]. The Dominican Republic is cited in the review text but has no table row, and Portugal is absent from countries with published cases up to 2019.

Case concentration in Turkey, the Middle East, India and Pakistan accompanies consanguinity, and the review found no effect of national income on sex assignment, although 86% of cases came from high- or upper-middle-income countries and none from low-income countries [9].

9.3. Founder populations and recurrent variants

Dominican Republic. The large kindred described in 1974 had 23 families and 38 affected individuals, with exon 5 variant p.Arg246Trp; of four additional families, two had the same variant and two p.Gly115Asp or p.Gly183Ser, with similar phenotype and biochemistry, showing several population variants rather than one common ancestor [30]. Experienced Dominican midwives now recognise deficiency at birth [15].

Papua New Guinea. The Simbari Anga linguistic group cluster (Sambia, Eastern Highlands) has a gene deletion exceeding 20 kb [5, 15]. Some villages began recognising affected individuals in childhood and raising them as boys [5].

Turkey. An isolated southern village was the second community described, with 12 individuals with sexual ambiguity and consanguinity [62]. In the 20-centre multicentre series (85 patients), the most frequent variants were p.Ala65Pro (30.6%), p.Leu55Gln (16.5%) and p.Gly196Ser (15.3%) [42]. In Adana, p.Ala65Pro was most common, homozygosity reached 93.1% and consanguinity 89.6% (versus 37% in a European series cited by the authors) [47]. A large family with homozygous exon 1 deletion was described [38].

Brazil. At Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo (HC-FMUSP), 5α-reductase type 2 deficiency was the most frequent aetiology among 127 adults with 46,XY DSD followed (30 patients), ahead of PAIS (24) and partial gonadal dysgenesis (19), with recurrent alleles p.Gly183Ser and p.Gln126Arg and the highest gender change rate among countries in the worldwide review (70%) [9, 51]. In Campinas (Grupo Interdisciplinar de Estudos da Determinação e Diferenciação do Sexo, UNICAMP; 2001–2010), among 58 46,XY patients with ambiguous genitalia and normal post-hCG T, deficiency was identified in 9 patients from 7 families (13.7% of families), and p.Gln126Arg was most frequent [48]. A São Paulo group review reports that adults with male social sex report more satisfactory sexual lives and quality of life than those with female social sex [82].

East Asia. p.Arg227Gln is the most frequent allele in Chinese Han children (39.62% of alleles in a Beijing cohort) and Taiwan (77.8%, 14 of 18 alleles), associated with isolated micropenis or hypospadias and predominantly male rearing [40, 45]. In southern China, the haplotype combining c.265C>G and c.680G>A suggests a founder effect [44]. In Japan, 3.7% of 81 boys with isolated micropenis had deficiency [18]. In Hong Kong and Vietnam, series based diagnosis on urinary profiling [60, 61].

Middle East and South Asia. In Iraq, five of seven unrelated patients were homozygous for c.453del, suggesting a founder effect [83]; in Iraqi Kurdistan, SRD5A2 was among the most frequent genes, and in India it was most implicated in a cohort of 147 children [64, 65]. In Indonesia, a series of 37 patients revealed six novel variants, and limited DHT measurement availability prompted validation of urinary etiocholanolone/androsterone [59, 84]. p.Gly34Arg is more frequent in Egypt and p.Pro212Arg in Mexico [17].

Europe and USA. European series (Montpellier, Pisa, Birmingham, I-DSD registry) are heterogeneous, with many private variants and patients of migrant origin [17, 21, 35, 52]. In the USA, described cases include patients of Mexican (p.Pro212Arg), Caucasian (p.Asn193Ser/p.His231Arg; p.Gln126Arg, a variant described in French and Spanish ancestry) and African American ancestry [37]. The sources consulted include no Portuguese series.

9.4. Special populations and situations

Inequalities in diagnostic access influence diagnostic age and decisions: in Indonesia DHT is poorly accessible because of cost [59], and in Turkey T/DHT was not measured in pubertal patients because insurance did not cover it [47]. In Middle Eastern societies, families may not comply with gender change decisions when early diagnosis fails, and the Adana authors raise the ethical dilemma of athletes with 46,XY DSD hyperandrogenism in female sports categories [47]. Fertility carries particular social weight where female infertility prevents marriage [13].

9.5. Reference networks, registries and patient associations

The European Reference Network on Rare Endocrine Conditions (Endo-ERN) brings together more than 100 reference centres in 27 European Union Member States and Norway; MTG7 (Sex Development & Maturation) covers DSDs, including 5α-reductase type 2 deficiency, and provides centre locations, registries, virtual consultations through CPMS (Clinical Patient Management System), guidelines and patient representatives (ePAGs) [22]. SDMregistries (formerly I-DSD Registry) maintains international I-DSD, I-CAH, I-TS and I-HH registries, with quality improvement and training programmes [23]. DSDnet (COST Action BM1303) and dsd-LIFE were European collaborative projects; the DSDnet website is no longer available [20]. Patient and family associations include dsdfamilies (United Kingdom) [85], interACT, advocating for intersex youth (USA) [86], AIS-DSD Support Group (United Kingdom) [87], Associação Brasileira Intersexo (ABRAI) [88] and Accord Alliance (USA) [12]. No specific Portuguese association was identified in the sources consulted.

10. Illustrative Clinical Case

The case was published by Byers and colleagues in a five-patient series from two US academic centres [37]. The patient was the third daughter of a Mexican couple, after uncomplicated pregnancy and delivery. At birth, genitalia appeared female, with bifid labioscrotal folds of minimal rugosity, no hyperpigmentation or fusion, a very small clitoris, separate urethral opening and posterior introitus; bilateral palpable masses in the labioscrotal folds prompted DSD team referral [37]. Karyotype was 46,XY with SRY present by fluorescence in situ hybridisation, and AR sequencing was normal. Ultrasound suggested a bicornuate uterus and testes with epididymis, raising suspicion of dysgenetic gonads and prompting laparoscopy and gonadectomy recommendations [37]. At laparoscopy at 8 months, no Müllerian structures were present; the vaginal vault measured 2.5 × 3 cm, without a cervix, and labial gonads measured 1 × 0.4 × 0.7 cm and 0.9 × 0.5 × 0.7 cm; biopsy showed normal testes, which were retained [37]. Figure 1 shows external genitalia at 18 months, with testes already retracted into the inguinal canals (Figure 1 of Byers et al., 2017) [37].

At 2.5 years, reassessment by a child psychologist concluded that the child identified fully as a girl. SRD5A2 sequencing revealed homozygosity for p.Pro212Arg, previously described in Mexican-ancestry families, and diagnosis was established at that age [37]. The family strongly preferred female rearing based on genital appearance and expected masculinising reconstruction outcome, and the team supported this decision; parents requested gonadectomy to reinforce the decision, prevent pubertal virilisation and relieve distress, and male fertility potential did not dissuade them [37]. Hospital policy, supported by interpretation of Washington State law, prohibits parental consent to remove a minor's reproductive organs except with oncogenic risk or inevitable infertility, otherwise requiring a court order; without means to obtain one, the family retained the testes, and at 7 years the child identified as a girl and was followed by a psychologist and DSD team before puberty [37].

The case illustrates several points of this article. Ultrasound raised suspicion of dysgenesis and led to a proposal of gonadectomy that the biopsy showed to be unnecessary; the molecular diagnosis emerged only at 2.5 years; and the decision on the gonads was conditioned by legal and not only clinical factors. The decisions described are not routine: the low tumour risk and the frequency of gender change lead current European documents to recommend deferring gonadectomy until the person can participate in the decision, with reversible pubertal suppression if needed (section 7) [20, 21]. The remaining four cases of the series, all gonadectomised between 7 and 19 months and identifying as girls at 3, 7, 9 and 12 years, show that female identity also occurs in this disease [37].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

The sources consulted describe no life-threatening complications or reduced life expectancy; the untreated 65-year-old Dominican patient illustrates somatically benign natural history [26]. Endocrine testicular function allows spontaneous puberty with gonads in situ, but some individuals have elevated gonadotrophins and low T in late puberty or early adulthood, warranting hypogonadism surveillance [21]. Bone mineral density appears normal after weight adjustment [49]. Germ cell tumour risk appears low, at least until early adulthood, but data are sparse (2006 consensus with 3 patients; 14 Turkish gonadectomies; 6 gonadectomies and 8 individuals with gonads in situ in I-DSD), and seminomas have been described at 23 and 49 years [13, 21, 42]. Psychosocial prognosis depends on congruence between assigned gender and gender identity: change rates range from 12.5% to over 90% according to cohort, period and measurement (Table 3.2) [9, 25, 35]. Gender variance was associated with lower self-esteem and more anxiety and depression in DSDs generally [53]. Penile growth frequently remains below normal even with DHT, and penis size concerns and erectile dysfunction are frequent in operated men [18, 77]. Fertility is reduced but possible with assisted reproduction [15, 51].

11.2. Surveillance

Sources define no 5α-reductase type 2 deficiency-specific surveillance intervals; follow-up is by an interdisciplinary team (endocrinology, genetics, gynaecology or urology, psychology), with standardised lifelong somatic assessment [12, 20]. Table 11.1 brings together recommended examinations and sources; when a source gives no interval, the column states this.

Table 11.1. Long-term surveillance

Surveillance implementation is irregular: in the I-DSD registry, only 2 individuals with gonads in situ underwent testicular surveillance [21]. Transition to adult care should be planned (section 7.12) [20].

12. Summary and Final Considerations

5α-Reductase type 2 deficiency is an autosomal recessive 46,XY DSD caused by biallelic SRD5A2 variants, preventing target-tissue conversion of T to DHT. The phenotype reflects the two androgens' division of functions: T forms Wolffian derivatives and produces muscular, vocal and genital virilisation at puberty, while absent DHT leaves external genitalia and prostate undervirilised at birth and facial hair sparse in adulthood. The spectrum ranges from female genitalia with a blind vaginal pouch to isolated micropenis, with hypomorphic variants such as p.Arg227Gln predominant in East Asia and severe variants and founder populations in the Dominican Republic, Papua New Guinea, Turkey, Brazil and Middle East.

Diagnosis now rests mainly on SRD5A2 sequencing, with parallel biochemistry. Stimulated T/DHT sensitivity is below 80% in several series and lacks specificity against partial androgen insensitivity; mass spectrometry urinary steroid profiling, interpretable from 3 months, appears more accurate. A negative genetic result or a single variant does not exclude diagnosis because deletions, non-coding variants and uniparental disomy occur, and c.265C>G p.(Leu89Val), historically “V89L”, is benign and does not count as a causal allele. 17β-HSD3 deficiency is the closest differential and differs in absent prostate, gynaecomastia and low T/androstenedione.

Management is multidisciplinary and has no authorised treatment. Sex assignment should weigh female-to-male gender change observed in highly variable proportions according to cohort, period and culture, and documented fertility potential; the current trend favours male assignment when diagnosis is early and deferral of irreversible surgery and gonadectomy until the person can participate in decisions, considering apparently low tumour risk. Topical DHT, used off-label with doses only from small series, increases penile length mainly before puberty. Paternity is possible with assisted reproduction. Major gaps remain: no prevalence estimates, controlled trials, long-term pubertal suppression data or Portuguese series.

13. Bibliographical References

1. Orphanet. 46,XY difference of sex development due to 5-alpha-reductase 2 deficiency (ORPHA:753) [Internet]. Paris: INSERM; version dated 2 Jul 2026 [accessed 10 Oct 2026]. Available from: https://www.orpha.net/en/disease/detail/753
2. GeneGuidance. SRD5A2 — clinical curation of the gene and condition OMIM 264600. User's database, version dated 26 Sep 2026; primary references indicated in each field.
3. Walsh PC, Madden JD, Harrod MJ, Goldstein JL, MacDonald PC, Wilson JD. Familial incomplete male pseudohermaphroditism, type 2. Decreased dihydrotestosterone formation in pseudovaginal perineoscrotal hypospadias. N Engl J Med. 1974;291(18):944-949. doi:10.1056/NEJM197410312911806
4. Herdt GH, Davidson J. The Sambia "turnim-man": sociocultural and clinical aspects of gender formation in male pseudohermaphrodites with 5-alpha-reductase deficiency in Papua New Guinea. Arch Sex Behav. 1988;17(1):33-56. doi:10.1007/BF01542051
5. Imperato-McGinley J, Miller M, Wilson JD, Peterson RE, Shackleton C, Gajdusek DC. A cluster of male pseudohermaphrodites with 5 alpha-reductase deficiency in Papua New Guinea. Clin Endocrinol (Oxf). 1991;34(4):293-298. doi:10.1111/j.1365-2265.1991.tb03769.x
6. Orphanet. 46,XY difference of sex development due to 17-beta-hydroxysteroid dehydrogenase 3 deficiency (ORPHA:752) [Internet]. Paris: INSERM; prevalence file Orphadata en_product9_prev.xml [accessed 10 Oct 2026]. Available from: https://www.orpha.net/en/disease/detail/752
7. GeneGuidance. HSD17B3 — clinical curation of the gene and condition OMIM 264300. User's database, version dated 26 Sep 2026; primary references indicated in each field.
8. 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-R206. doi:10.1530/EJE-18-0256
9. Batista RL, Mendonca BB. Integrative and analytical review of the 5-alpha-reductase type 2 deficiency worldwide. Appl Clin Genet. 2020;13:83-96. doi:10.2147/TACG.S198178
10. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-68. doi:10.1002/humu.23348
11. Firth HV, Hurst JA. Ambiguous genitalia (including sex reversal). In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 54–57.
12. Mohnach L, Fechner PY, Keegan CE. Nonsyndromic Disorders of Testicular Development Overview. 2008 May 21 [updated 2022 Aug 18]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–. https://www.ncbi.nlm.nih.gov/books/NBK1547/
13. Hughes IA, Houk C, Ahmed SF, Lee PA; LWPES Consensus Group; ESPE Consensus Group. Consensus statement on management of intersex disorders. Arch Dis Child. 2006;91(7):554-563. doi:10.1136/adc.2006.098319
14. Gottlieb B, Trifiro MA. Androgen Insensitivity Syndrome. 1999 Mar 24 [updated 2017 May 11]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–. https://www.ncbi.nlm.nih.gov/books/NBK1429/
15. Kang HJ, Imperato-McGinley J, Zhu YS, Rosenwaks Z. The effect of 5α-reductase-2 deficiency on human fertility. Fertil Steril. 2014;101(2):310-316. doi:10.1016/j.fertnstert.2013.11.128
16. Ahmed SF, Achermann J, Alderson J, Crouch NS, Elford S, Hughes IA, et al. Society for Endocrinology UK Guidance on the initial evaluation of a suspected difference or disorder of sex development (Revised 2021). Clin Endocrinol (Oxf). 2021;95(6):818-840. doi:10.1111/cen.14528
17. Balagamage C, Igbokwe R, Cole T, Baranowski ES, McCarthy L, Chandran H, et al. The diagnostic value of stimulated androgen ratios in 5-alpha reductase type 2 (SRD5A2) deficiency: a case series and review of the literature. J Pediatr Endocrinol Metab. 2025;38(9):922-930. doi:10.1515/jpem-2025-0195
18. Sasaki G, Ishii T, Hori N, Amano N, Homma K, Sato S, et al. Effects of pre- and post-pubertal dihydrotestosterone treatment on penile length in 5α-reductase type 2 deficiency. Endocr J. 2019;66(9):837-842. doi:10.1507/endocrj.EJ19-0111
19. Agence nationale de sécurité du médicament et des produits de santé (ANSM). ANDRACTIM, gel pour application locale (androstanolone 2,5%) — Résumé des caractéristiques du produit. Base de données publique des médicaments, CIS 68001518 [Internet]. Paris: Ministère de la Santé; [accessed 2026-10-10]. Available from: https://base-donnees-publique.medicaments.gouv.fr/affichageDoc.php?specid=68001518&typedoc=R
20. 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):415-429. doi:10.1038/s41574-018-0010-8
21. Boogers LS, Brüggenwirth HT, Wolffenbuttel KP, Hersmus R, Bryce J, Ahmed SF, et al. Gonadal function and pathology in 17beta-HSD 3 and 5alpha-reductase deficiency. Eur J Endocrinol. 2025;192(1):34-45. doi:10.1093/ejendo/lvae154
22. European Reference Network on Rare Endocrine Conditions (Endo-ERN). Sex Development & Maturation (MTG7) [Internet]. [accessed 2026-10-10]. Available from: https://endo-ern.eu/rare-sex-development-maturation-conditions/
23. SDMregistries — International Registries for Rare Conditions Affecting Sex Development & Maturation (I-DSD, I-CAH, I-TS, I-HH) [Internet]. Glasgow: University of Glasgow; [accessed 2026-10-10]. Available from: https://sdmregistries.org/
24. Imperato-McGinley J, Guerrero L, Gautier T, Peterson RE. Steroid 5alpha-reductase deficiency in man: an inherited form of male pseudohermaphroditism. Science. 1974;186(4170):1213-1215. doi:10.1126/science.186.4170.1213
25. Imperato-McGinley J, Peterson RE, Gautier T, Sturla E. Androgens and the evolution of male-gender identity among male pseudohermaphrodites with 5alpha-reductase deficiency. N Engl J Med. 1979;300(22):1233-1237. doi:10.1056/NEJM197905313002201
26. Imperato-McGinley J, Peterson RE, Leshin M, Griffin JE, Cooper G, Draghi S, et al. Steroid 5 alpha-reductase deficiency in a 65-year-old male pseudohermaphrodite: the natural history, ultrastructure of the testes, and evidence for inherited enzyme heterogeneity. J Clin Endocrinol Metab. 1980;50(1):15-22. doi:10.1210/jcem-50-1-15
27. Andersson S, Berman DM, Jenkins EP, Russell DW. Deletion of steroid 5 alpha-reductase 2 gene in male pseudohermaphroditism. Nature. 1991;354(6349):159-161. doi:10.1038/354159a0
28. Thigpen AE, Davis DL, Milatovich A, Mendonca BB, Imperato-McGinley J, Griffin JE, et al. Molecular genetics of steroid 5 alpha-reductase 2 deficiency. J Clin Invest. 1992;90(3):799-809. doi:10.1172/JCI115954
29. Thigpen AE, Davis DL, Gautier T, Imperato-McGinley J, Russell DW. Brief report: the molecular basis of steroid 5 alpha-reductase deficiency in a large Dominican kindred. N Engl J Med. 1992;327(17):1216-1219. doi:10.1056/NEJM199210223271706
30. Cai LQ, Zhu YS, Katz MD, Herrera C, Baéz J, DeFillo-Ricart M, et al. 5 alpha-reductase-2 gene mutations in the Dominican Republic. J Clin Endocrinol Metab. 1996;81(5):1730-1735. doi:10.1210/jcem.81.5.8626825
31. Lee PA, Houk CP, Ahmed SF, Hughes IA; International Consensus Conference on Intersex. Consensus statement on management of intersex disorders. Pediatrics. 2006;118(2):e488-e500. doi:10.1542/peds.2006-0738
32. Firth HV, Hurst JA. Hypospadias. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 186–187.
33. Katz MD, Cai LQ, Zhu YS, Herrera C, DeFillo-Ricart M, Shackleton CH, et al. The biochemical and phenotypic characterization of females homozygous for 5 alpha-reductase-2 deficiency. J Clin Endocrinol Metab. 1995;80(11):3160-3167. doi:10.1210/jcem.80.11.7593420
34. Mao Y, Huang JM, Chen-Zhang YW, Lin H, Zhang YH, Jiang JY, et al. Exploring the clinical implications of novel SRD5A2 variants in 46,XY disorders of sex development. Asian J Androl. 2025;27(2):211-218. doi:10.4103/aja202469
35. Maimoun L, Philibert P, Cammas B, Audran F, Bouchard P, Fenichel P, et al. Phenotypical, biological, and molecular heterogeneity of 5α-reductase deficiency: an extensive international experience of 55 patients. J Clin Endocrinol Metab. 2011;96(2):296-307. doi:10.1210/jc.2010-1024
36. National Center for Biotechnology Information. ClinVar [Internet]. Bethesda (MD): NLM; [accessed 2026-10-10]. Records VCV000459645, VCV000003337, VCV000003351, VCV000003341, VCV000372519, VCV000988403, VCV000003339, VCV000003345, VCV000003340, VCV000436859, VCV000459640, VCV000003346, VCV000097400, VCV000003347. Available from: https://www.ncbi.nlm.nih.gov/clinvar/
37. Byers HM, Mohnach LH, Fechner PY, Chen M, Thomas IH, Ramsdell LA, et al. Unexpected ethical dilemmas in sex assignment in 46,XY DSD due to 5-alpha reductase type 2 deficiency. Am J Med Genet C Semin Med Genet. 2017;175(2):260-267. doi:10.1002/ajmg.c.31560
38. Köprülü Ö, Hazan F, Bal Yüksel E, Erbaş İM, Nalbantoğlu Ö, Korkmaz HA, et al. Clinical and molecular characterization of 5α-reductase deficiency type 2: first report of a family with homozygous deletion in exon 1 of SRD5A2. Arch Pediatr. 2026;33(7):105602. doi:10.1016/j.arcped.2026.105602
39. Werner R, Basina A, Künstner A, Prakash SAC, Kösem K, Kulle A, et al. Exonisation of an Alu element in the 3'-UTR contributes to SRD5A2 deficiency. Sci Rep. 2026;16(1):24030. doi:10.1038/s41598-026-64581-x
40. Fan L, Song Y, Polak M, Li L, Ren X, Zhang B, et al. Clinical characteristics and genotype-phenotype correlations of 130 Chinese children in a high-homogeneity single-center cohort with 5α-reductase 2 deficiency. Mol Genet Genomic Med. 2020;8(10):e1431. doi:10.1002/mgg3.1431
41. Liu Q, Yin X, Li P. Clinical, hormonal, and genetic characteristics of 5α-reductase type 2 deficiency in 103 Chinese patients. Endocr Pract. 2022;28(9):859-866. doi:10.1016/j.eprac.2022.06.002
42. Abacı A, Çatlı G, Kırbıyık Ö, Şahin NM, Abalı ZY, Ünal E, et al. Genotype-phenotype correlation, gonadal malignancy risk, gender preference, and testosterone/dihydrotestosterone ratio in steroid 5-alpha-reductase type 2 deficiency: a multicenter study from Turkey. J Endocrinol Invest. 2019;42(4):453-470. doi:10.1007/s40618-018-0940-y
43. Deeba F, Choudhury P, Sharma J, Jain V, Dhua AK, Yadav DK, et al. Gene-dosage effect of V89L variants in steroid 5α-reductase type 2 and increased risk of 46,XY disorders of sex development: a systematic review and meta-analysis. J Indian Assoc Pediatr Surg. 2026;31(3):334-345. doi:10.4103/jiaps.jiaps_302_25
44. Lei X, Zhou X, Cheng Z, Zhao S, Gui C, Ma Y, et al. A hypomorphic SRD5A2 haplotype with a potential founder effect: composed of common variants in individuals with 5α-reductase type 2 deficiency from South China. Biol Sex Differ. 2026;17(1). doi:10.1186/s13293-026-00847-3
45. Chee SY, Ting CS, Chuang SH, Shih YJ, Chao MC, Hsiao HP, et al. Homozygous p.Arg227Gln mutation with isolated micropenis is the major presentation of 5α-reductase type 2 deficiency in Taiwan. Pediatr Neonatol. 2026 [Epub ahead of print]. doi:10.1016/j.pedneo.2025.12.009
46. Shen L, Fu D, Yang W, Cui Y, Wang H, Li D, et al. Clinical characterization and analysis of the SRD5A2 gene in 16 Chinese patients with 5α-reductase type 2 deficiency: a single-center retrospective study. BMC Pediatr. 2026;26(1):911. doi:10.1186/s12887-026-07012-0
47. Celiloglu C, Turan I, Kotan LD, Cimen AM, Dilek SO, Yuksel B. An evaluation of cases of disorders of sex development related to SRD5A2. Endocrine. 2025;89(2):632-640. doi:10.1007/s12020-025-04252-5
48. Veiga-Junior NN, Medaets PA, Petroli RJ, Calais FL, de Mello MP, Castro CC, et al. Clinical and laboratorial features that may differentiate 46,XY DSD due to partial androgen insensitivity and 5α-reductase type 2 deficiency. Int J Endocrinol. 2012;2012:964876. doi:10.1155/2012/964876
49. Sobel V, Schwartz B, Zhu YS, Cordero JJ, Imperato-McGinley J. Bone mineral density in the complete androgen insensitivity and 5alpha-reductase-2 deficiency syndromes. J Clin Endocrinol Metab. 2006;91(8):3017-3023. doi:10.1210/jc.2005-2809
50. Cohen-Kettenis PT. Gender change in 46,XY persons with 5alpha-reductase-2 deficiency and 17beta-hydroxysteroid dehydrogenase-3 deficiency. Arch Sex Behav. 2005;34(4):399-410. doi:10.1007/s10508-005-4339-4
51. Batista RL, Inácio M, Brito VN, Sircili MHP, Bag MJ, Gomes NL, et al. Sexuality and fertility desire in a large cohort of individuals with 46,XY differences in sex development. Clinics (Sao Paulo). 2023;78:100185. doi:10.1016/j.clinsp.2023.100185
52. Bertelloni S, Baldinotti F, Russo G, Ghirri P, Dati E, Michelucci A, et al. 5α-reductase-2 deficiency: clinical findings, endocrine pitfalls, and genetic features in a large Italian cohort. Sex Dev. 2016;10(1):28-36. doi:10.1159/000445090
53. Kreukels BPC, Köhler B, Nordenström A, Roehle R, Thyen U, Bouvattier C, et al. Gender dysphoria and gender change in disorders of sex development/intersex conditions: results from the dsd-LIFE study. J Sex Med. 2018;15(5):777-785. doi:10.1016/j.jsxm.2018.02.021
54. Meyer-Bahlburg HF, Baratz Dalke K, Berenbaum SA, Cohen-Kettenis PT, Hines M, Schober JM. Gender assignment, reassignment and outcome in disorders of sex development: update of the 2005 consensus conference. Horm Res Paediatr. 2016;85(2):112-118. doi:10.1159/000442386
55. Ahmed SF, Achermann JC, Arlt W, Balen A, Conway G, Edwards Z, et al. Society for Endocrinology UK guidance on the initial evaluation of an infant or an adolescent with a suspected disorder of sex development (Revised 2015). Clin Endocrinol (Oxf). 2016;84(5):771-788. doi:10.1111/cen.12857
56. Kulle A, Krone N, Holterhus PM, Schuler G, Greaves RF, Juul A, et al. Steroid hormone analysis in diagnosis and treatment of DSD: position paper of EU COST Action BM 1303 'DSDnet'. Eur J Endocrinol. 2017;176(5):P1-P9. doi:10.1530/EJE-16-0953
57. Ors A, Bereketoglu MB, Boz A, Akkus G. 5-alpha reductase deficiency - an uncommon disorder of sexual development: a case study of four Turkish patients within a family. Curr Gene Ther. 2026;26(2):1-4. doi:10.2174/0115665232415806250922094852
58. Akcay T, Fernandez-Cancio M, Turan S, Güran T, Audi L, Bereket A. AR and SRD5A2 gene mutations in a series of 51 Turkish 46,XY DSD children with a clinical diagnosis of androgen insensitivity. Andrology. 2014;2(4):572-578. doi:10.1111/j.2047-2927.2014.00215.x
59. Marzuki NS, Idris FP, Kartapradja H, Renata S, Harahap A, Batubara JRL. Accuracy of urinary etiocholanolone/androsterone ratio as alternative to serum testosterone/dihydrotestosterone ratio for diagnosis of 5 alpha-reductase type 2 deficiency patients and carriers in Indonesia. Int J Endocrinol Metab. 2021;19(2):e109510. doi:10.5812/ijem.109510
60. Chan AO, But BW, Lee CY, Lam YY, Ng KL, Tung JY, et al. Diagnosis of 5α-reductase 2 deficiency: is measurement of dihydrotestosterone essential? Clin Chem. 2013;59(5):798-806. doi:10.1373/clinchem.2012.196501
61. Tran TCM, Tran TNA, Le HBN, Nguyen VH, Tran MD, Vu CD, et al. Validation of steroid ratios for random urine by mass spectrometry to detect 5α-reductase deficiency in Vietnamese children. Clin Chem Lab Med. 2022;60(8):1225-1233. doi:10.1515/cclm-2022-0272
62. Akgun S, Ertel NH, Imperato-McGinley J, Sayli BS, Shackleton C. Familial male pseudohermaphroditism due to 5-alpha-reductase deficiency in a Turkish village. Am J Med. 1986;81(2):267-274. doi:10.1016/0002-9343(86)90262-7
63. Subspecialty Group of Endocrinology, Hereditary and Metabolic Diseases, the Society of Pediatrics, Chinese Medical Association; et al. [Expert consensus on genetic testing and genetic counseling for patients with disorders of sex development (2025)]. Zhonghua Er Ke Za Zhi. 2025;63(12):1293-1298. doi:10.3760/cma.j.cn112140-20250618-00529. PMID: 41233124
64. Jain V, Priyadarshini S, Sharma R, Kumar A, Sharma S, Raghupathy P, et al. Clinical characterization and molecular profiling by targeted next-generation sequencing in a large Indian cohort with 46,XY differences in sex development. Clin Endocrinol (Oxf). 2026;104(6):612-627. doi:10.1111/cen.70122
65. Idris FP, Tayeb TH, Robevska G, Sharef KHSH, Jalal DB, van den Bergen J, et al. Molecular analysis of individuals with suspected 46,XY differences of sex development in a homogenous and understudied population. Hum Mol Genet. 2026;35(19):ddag089. doi:10.1093/hmg/ddag089
66. Uehara E, Akiba K, Matsubara K, Fukami M, Tanase-Nakao K. Amplicon-based targeted next-generation sequencing using dried blood spots for 46,XY differences/disorders of sex development: aiming for diagnosis by minimally invasive testing. Clin Pediatr Endocrinol. 2025;34(3):172-179. doi:10.1297/cpe.2024-0081
67. Jones KL, Jones MC, del Campo M. Antley-Bixler syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 596.
68. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Antley-Bixler syndrome. In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 837–838.
69. PROSCAR (finasteride) tablets — prescribing information [Internet]. Silver Spring (MD): U.S. Food and Drug Administration/DailyMed; revision effective 2026-03-18 [accessed 2026-10-10]. Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=6f988153-fc74-4ca4-b29a-111f750c4a4b
70. Dutasteride capsules — prescribing information (generic) [Internet]. DailyMed; revision effective 2026-08-14 [accessed 2026-10-10]. Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=170f01fe-5048-413c-b4cd-8311e1a2728a
71. Lee PA, Nordenström A, Houk CP, Ahmed SF, Auchus R, Baratz A, et al; Global DSD Update Consortium. Global Disorders of Sex Development update since 2006: perceptions, approach and care. Horm Res Paediatr. 2016;85(3):158-180. doi:10.1159/000442975
72. Khorashad BS, Gardner M, Lee PA, Kogan BA, Sandberg DE. Recommendations for 46,XY disorders/differences of sex development across two decades: insights from North American pediatric endocrinologists and urologists. Arch Sex Behav. 2024;53(8):2939-2956. doi:10.1007/s10508-024-02942-1
73. U.S. Food and Drug Administration. Drugs@FDA (openFDA API) — search for dihydrotestosterone/androstanolone/stanolone [Internet]. [accessed 2026-10-10]. Available from: https://api.fda.gov/drug/drugsfda.json
74. Xu D, Lu L, Xi L, Cheng R, Pei Z, Bi Y, et al. Efficacy and safety of percutaneous administration of dihydrotestosterone in children of different genetic backgrounds with micropenis. J Pediatr Endocrinol Metab. 2017;30(12):1285-1291. doi:10.1515/jpem-2016-0400
75. Vupputuri M, Kandepu M, Devireddy HR. 5α-reductase type 2 deficiency: response to dihydrotestosterone gel. Indian J Pediatr. 2014;81(8):821-823. doi:10.1007/s12098-013-1032-9
76. Claahsen-van der Grinten HL, Gravholt CH, White EK, Brösamle M, Breen L, Smyth A, et al; Endo-ERN MTG7. ENDO-ERN expert opinion: addressing the endocrine needs in considering genital surgery in disorders/differences in sex development individuals in Europe. Eur J Endocrinol. 2026;194(2):217-224. doi:10.1093/ejendo/lvag009
77. Singh VK, Sharma S, Bajpai M, Gupta DK, Chaturvedi PK. Traversing from childhood surgery to adult life: long-term functional and psychosexual outcomes in males with differences/disorders of sex development. J Pediatr Urol. 2026 [Epub ahead of print]:106246. doi:10.1016/j.jpurol.2026.106246
78. Neumann U, Binder G, Eckoldt F, Döhnert U, Dunstheimer D, Holterhus PM, et al. German law on protection of children with DSD: first data on care after enactment. Endocr Connect. 2026;15(6):e260194. doi:10.1530/EC-26-0194
79. Firth HV, Hurst JA. Androgen insensitivity syndrome (AIS). In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 344–345.
80. ClinicalTrials.gov [Internet]. Bethesda (MD): U.S. National Library of Medicine. API v2 search for 5-alpha reductase deficiency, SRD5A2, dihydrotestosterone gel, androstanolone, micropenis, 46,XY DSD [accessed 2026-10-10]. Available from: https://clinicaltrials.gov/ (NCT06723938, NCT06692049, NCT05867979, NCT06682078)
81. Katz MD, Kligman I, Cai LQ, Zhu YS, Fratianni CM, Zervoudakis I, et al. Paternity by intrauterine insemination with sperm from a man with 5alpha-reductase-2 deficiency. N Engl J Med. 1997;336(14):994-997. doi:10.1056/NEJM199704033361404
82. Mendonca BB, Batista RL, Domenice S, Costa EM, Arnhold IJ, Russell DW, et al. Steroid 5α-reductase 2 deficiency. J Steroid Biochem Mol Biol. 2016;163:206-211. doi:10.1016/j.jsbmb.2016.05.020
83. Di Marco C, Bulotta AL, Varetti C, Dosa L, Michelucci A, Baldinotti F, et al. Ambiguous external genitalia due to defect of 5-α-reductase in seven Iraqi patients: prevalence of a novel mutation. Gene. 2013;526(2):490-493. doi:10.1016/j.gene.2013.04.070
84. Marzuki NS, Idris FP, Kartapradja HD, Harahap AR, Batubara JRL. Characterising SRD5A2 gene variants in 37 Indonesian patients with 5-alpha-reductase type 2 deficiency. Int J Endocrinol. 2019;2019:7676341. doi:10.1155/2019/7676341
85. dsdfamilies [Internet]. United Kingdom [accessed 10 Oct 2026]. Available from: https://dsdfamilies.org/
86. interACT: Advocates for Intersex Youth [Internet]. USA [accessed 10 Oct 2026]. Available from: https://interactadvocates.org/
87. AIS-DSD Support Group (AISSG) [Internet]. United Kingdom [accessed 10 Oct 2026]. Available from: https://www.aissg.org/
88. Associação Brasileira Intersexo (ABRAI) [Internet]. Brazil [accessed 10 Oct 2026]. Available from: https://abrai.org.br/