CDH3-Related Disorders (Hypotrichosis with Juvenile Macular Dystrophy and EEM Syndrome)

Autosomal recessive allelic spectrum of P-cadherin: congenital hypotrichosis of the scalp associated with childhood-onset chorioretinal atrophy of the posterior pole, with limb anomalies and ectodermal dysplasia of variable severity in EEM syndrome.

Literature and regulatory information updated: 29 October 2026.

0. Initial Reference Table

The condition has also been called «aniridia / hypotrichosis with retinal dystrophy». None of the sources consulted (Orphanet, OMIM, GeneReviews, clinical series) associates CDH3 with aniridia, which is a disorder of ocular development linked mainly to PAX6; the entities caused by biallelic CDH3 variants are HJMD and EEM syndrome, and the term «retinal dystrophy» corresponds to the macular dystrophy and posterior pole chorioretinal atrophy described in these two diseases [1, 2, 4, 5].

1. Introduction

1.1. From clinical descriptions to the identification of CDH3

The association between a congenital hair anomaly and juvenile-onset macular dystrophy was first described by Wagner in 1935, in two German siblings; OMIM refers to two sisters and the series of Hull and colleagues refers to two brothers, a discrepancy that only reading the original article, in German, could resolve [2, 11]. Isolated reports followed of affected siblings in Japan (Yasakura, 1967), of a sporadic case whose parents came from the same village in Westphalia (Kroll, 1981) and of a brother and sister whose parents were born in the same Portuguese village, in whom the macular dystrophy only became evident after 18 years of age (Souied and colleagues, 1995); occurrence in siblings of related parents or of the same geographical origin suggested autosomal recessive inheritance from an early stage [2].

In parallel, a more extensive picture was described, combining the same macular dystrophy with ectodermal dysplasia and malformations of the hands and feet. Albrectsen and Svendsen reported in 1956 two children of first-cousin parents, in Denmark, with hypotrichosis, syndactyly and retinal degeneration [5, 12]. Ohdo, Hirayama and Terawaki proposed in 1983 the designation EEM syndrome (ectodermal dysplasia, ectrodactyly, macular dystrophy) on the basis of patients from an isolated population on a remote island in Japan, with ectodermal dysplasia, ectrodactyly associated with syndactyly or cleft hand and presumably progressive macular dystrophy; the abstract of the article refers to five patients, and OMIM refers to one family with six cases [5, 21]. There followed a Japanese woman of first-cousin parents (Hayakawa and colleagues, 1979 and 1989), a Brazilian family followed for ten years (Balarin Silva and colleagues, 1999) and two Yemenite Jewish siblings with a less severe form (Senecky and colleagues, 2001) [5, 22–24]. Kuster and colleagues described in 1987 a patient with alopecia, macular degeneration and growth retardation, whom they considered a distinct entity [5].

The molecular basis was identified in 2001 by Sprecher and colleagues, who studied four consanguineous Druze families from northern Israel with eleven affected individuals: homozygosity mapping placed the locus at 16q22.1 and all families shared the homozygous deletion of a guanine in exon 8 of CDH3 (historically designated 981delG), with reduced mRNA and protein in the patients' skin. This was the first demonstration that a cadherin causes a human disease of the hair and retina [6, 25]. In 2002, Indelman and colleagues described the first pathogenic missense variant, p.(Arg503His), in a fifth consanguineous family with four affected individuals, and documented pili torti on light and electron microscopy [26]. In 2005, Kjaer and colleagues showed that EEM syndrome also results from homozygous CDH3 variants in two of the families already published, the Danish family of Albrectsen and Svendsen and the Brazilian family of Balarin Silva, and concluded that CDH3 was the third gene associated with ectrodactyly [5, 27]. In 2008, Shimomura and colleagues showed that CDH3 is a direct target gene of the transcription factor p63, which explained the clinical overlap with the syndromes caused by TP63 [28].

In 2006, Leibu and colleagues, documenting cone and rod dysfunction in 16 patients, proposed the designation «hypotrichosis with cone-rod dystrophy», which was not adopted by the databases [2, 3]. In 2010, the review by Basel-Vanagaite and colleagues, prompted by a new Yemenite Jewish family, concluded that HJMD and EEM syndrome very probably form a continuous phenotypic spectrum and not two distinct syndromes [12]. In 2016, the British series of Hull and colleagues, with ten patients from ten families, systematically characterised retinal imaging and electrophysiology and established that the chorioretinal atrophy is confined to the posterior pole and that sparse scalp hair is universal [11].

1.2. Nomenclature and definition of the entities

OMIM and Orphanet maintain two entries. HJMD (MIM 601553; ORPHA:1573) is defined as an autosomal recessive disease in which scalp hair is lost in the first months of life and progressive macular degeneration, with loss of vision, appears between a few years and two decades later [1, 2]. EEM syndrome (MIM 225280; ORPHA:1897) combines ectodermal dysplasia (hypotrichosis of the scalp, eyebrows and eyelashes, with partial anodontia), ectrodactyly and macular dystrophy, the latter described as central geographic atrophy of the retinal pigment epithelium (RPE) and choriocapillaris, with coarse hyperpigmentation and preservation of the large choroidal vessels; digital absences, polydactyly, syndactyly and camptodactyly accompany it variably, with the hands more affected than the feet [4, 5]. According to OMIM, the absence of digital anomalies and normal eyebrows and eyelashes distinguish HJMD from EEM [2].

The separation is, in practice, a convention. The same variant, c.830del p.(Gly277AlafsTer20), has been found in the homozygous state in patients with HJMD without limb anomalies and in a family with EEM, and mild limb anomalies (cutaneous syndactyly, a hypoplastic nail on one digit) may go unnoticed and lead to a diagnosis of HJMD [11, 12, 29]. Hull and colleagues therefore proposed that all patients be seen as part of a spectrum with constant hypotrichosis and macular dystrophy and variable additional limb and ectodermal anomalies [11]. The expression «CDH3-related disorder» designates the whole, and the names HJMD and EEM indicate the clinical poles.

1.3. Classification and nosological framework

CDH3 disorders are classified differently depending on the perspective. In the molecular classification of ectodermal dysplasias by signalling pathways, which updates the 2019 classification [30] and whose first five-yearly update was approved by an international panel in Paris in June 2022, EEM syndrome appears in the p63 pathway group, alongside the syndromes caused by TP63 and KDF1; HJMD does not appear in that consensus's table of core ectodermal dysplasias, which requires changes in at least two structures of ectodermal origin [16]. In dermatological approaches to paediatric hypotrichosis, HJMD is an autosomal recessive syndromic hypotrichosis [31] and, in a review of ectodermal dysplasias, a disease of a membrane structural protein [32]. In ophthalmology, CDH3 belongs to the group of retinal cadherinopathies, alongside CDHR1, CDH23 and PCDH15, with the particularity of being the only one of these cadherins expressed in the RPE and not in the photoreceptors [33]. The American Academy of Ophthalmology (AAO) guidelines for the assessment of patients with IRD place the disease among the dystrophies involving the macula and the chorioretinal degenerations [13].

Table 1.1 summarises the forms described in the CDH3 allelic spectrum, including the two borderline phenotypes whose causal relationship with the gene rests on few families.

2. Molecular Genetics and Pathophysiology

2.1. The CDH3 gene and P-cadherin

The CDH3 gene is located at 16q22.1, in a cluster of cadherin genes in the 16q21–q22.1 region that includes CDH1, CDH5, CDH8 and CDH11 [6]. It spans about 54.8 kb of genomic DNA and has 16 exons, all coding; the reference mRNA is 4276 bp long and the coding sequence about 2.5 kb [12, 38]. The MANE Select transcript is NM_001793.6 (ENST00000264012.9), which encodes the 829-amino-acid protein NP_001784.2; UniProt (P22223) records two isoforms that differ at residues 761–829 [7, 12]. CDH3 and CDH1 are adjacent, about 38 kb apart, which has consequences for the interpretation of large deletions (section 2.7) [35, 39].

P-cadherin (placental cadherin, so called because it was identified in the mouse placenta) is a classic calcium-dependent cell–cell adhesion glycoprotein. The human protein has a signal peptide, five extracellular cadherin domains (EC1 to EC5) joined by calcium-binding linkers, a transmembrane segment and a conserved cytoplasmic tail that binds β-catenin and, through it, the actin cytoskeleton [6, 12, 40]. Binding is homophilic: the EC1 domain mediates trans dimerisation between neighbouring cells and the EC4 domain cis dimerisation in the same membrane [32]. β-catenin also activates the CDH3 promoter in a cell type-dependent manner [12].

2.2. Tissue expression and function

Three tissues explain the phenotype. In the RPE, absolute cDNA quantification of the three classic cadherins showed that P-cadherin accounts for 92–93% of the total in the human RPE and 82–85% in the mouse RPE in vivo, located at the cell borders; cultured RPE expresses mainly N-cadherin and therefore does not reproduce the mature RPE [41]. During RPE differentiation, N-cadherin expression is switched off and P-cadherin expression switched on, and P-cadherin is not expressed in the neural retina [11]. In the hair follicle, P-cadherin is the only cadherin expressed in the inner hair matrix during morphogenesis, whereas the epidermis and the other appendages express mainly E-cadherin, which explains why the hair is affected without relevant skin disease [35]. In the developing limb, mouse Cdh3 is expressed in the apical ectodermal ridge between embryonic days 10.5 and 12.5 and then in the interdigital mesenchyme, a pattern compatible with the digital anomalies of EEM syndrome [27].

Shimomura and colleagues showed, by promoter assays and chromatin immunoprecipitation, that the transcription factor p63 binds directly to two regions of the CDH3 promoter; P-cadherin is thus an effector of the p63 pathway, which explains why CDH3 and TP63 disorders share hypotrichosis and split hand/foot malformation. The Cdh3-null mouse does not reproduce the human phenotype, and P-cadherin is required for follicular and apical ectodermal ridge morphogenesis in humans, but not in mice [28]. In human hair follicles in organ culture, CDH3 silencing by RNA interference inhibited shaft growth, induced premature catagen and reduced matrix keratinocyte proliferation, with less membrane β-catenin and less axin 2, more active GSK3β and more TGF-β2; GSK3β inhibition and TGF-β neutralisation partially antagonised these effects, which indicates that P-cadherin maintains the anagen phase through the canonical Wnt pathway and suppression of TGF-β2 [42]. Scalp histology from six patients homozygous for p.(Arg503His), aged 9 to 21 years, showed more follicles in catagen and telogen, an increased ratio of vellus to terminal follicles in two cases and no inflammation or scarring, a pattern similar to that of chronic telogen effluvium [43].

2.3. Mechanism of the retinal disease

The dominant hypothesis is that photoreceptor loss is secondary to RPE dysfunction. The arguments are the absence of P-cadherin in the neural retina, the predominance of this cadherin in the RPE and the observation, on optical coherence tomography (OCT), that the outer retina does not atrophy before RPE loss, contrary to what would be expected in a primary photoreceptor disease [11, 41]. Outer retinal tubulations, frequent in CDH3 disease, also occur in other diseases with RPE dysfunction, such as Stargardt disease and age-related macular degeneration [11]. The normal electro-oculogram (EOG) in most patients studied and the pattern of simultaneous loss of RPE and photoreceptors, similar to choroideremia, leave open the possibility that the disease is not an isolated primary RPE dysfunction [3, 11]. The greater susceptibility of the posterior pole has been attributed to the density of cones and the metabolic demand of the macula on the RPE, which phagocytoses the shed outer segments, and to the thinner Bruch membrane in that region, without experimental demonstration [11]. Marziali and colleagues interpret the disease as an RPE-driven entity, due to dysfunction or malformation of this epithelium, rather than as a diffuse retinal degeneration [40]. Oxidative stress dissociates P-cadherin and β-catenin from the mouse RPE membrane, with nuclear translocation of β-catenin and activation of the canonical Wnt pathway [41].

2.4. Inheritance pattern

All sources describe autosomal recessive inheritance due to biallelic loss-of-function variants, without divergence between OMIM, Orphanet, the GeneReviews on Druze founder variants and the clinical series [1, 2, 5, 8]. Heterozygous parents are healthy, as confirmed by segregation studies in several families [12, 29, 44–46]. Consanguinity is frequent (eight of the ten families in Hull's series), but not obligatory: there is compound heterozygosity in non-consanguineous families from England, Spain, Italy, Japan and China [9, 11, 44–47]. In the Brazilian family with EEM, a brother heterozygous for the familial deletion had mild bilateral syndactyly between digits 1 to 4 and small, widely spaced teeth, with normal eye and hair examination; the sources consulted do not clarify whether this is expression in the heterozygous state or coincidence [5, 6]. In the Lebanese family with isolated hypotrichosis, a homozygous mother and daughter gave a pattern of pseudodominance [35].

2.5. Variant spectrum

The Human Gene Mutation Database (HGMD Professional) recorded 35 pathogenic variants in January 2020 and 95 causal or probably causal variants in May 2026, 75 of them associated with retinal disease [40, 44]. In October 2026, ClinVar contained 955 records with CDH3 in the gene field (including large CNVs encompassing the gene), 93 classified as pathogenic or likely pathogenic and 423 as variants of uncertain significance (VUS), raw counts without curation [48]. Most variants are truncating (nonsense, frameshift, splice site), distributed throughout the protein; 14 of the 18 variants known in 2010 lie between exons 5 and 14 [11, 12]. All missense variants published by 2016 were located in the extracellular domains, and almost all in the calcium-binding interdomain linkers: p.(Asn322Ile) and p.(Arg503His) close to calcium ions, p.(Glu504Lys) in EC4 and p.(His575Arg) in EC5 close to residue Asp542 of EC4; the exception is p.(Gly277Val), buried in the core of the EC2 domain, which destabilises the structure through the hydrogen bond network [11, 12]. Hull and colleagues noted that four of the seven missense variants known in 2016 were associated with EEM, and hypothesised that an abnormal protein being produced is more harmful to limb development than the absence of protein [11].

Table 2.1 presents the variants described in the sources consulted, normalised in VariantValidator to the MANE transcript. Several older publications used historical numbering or inconsistent notations; the conversions given result from consulting VariantValidator and not from manual calculation [7].

Table 2.1 has three nomenclature problems worth noting. The variant of patient 9 in Hull's series is published as «c.2157C>T (p.Arg720*)»; in the MANE transcript, c.2157C>T is synonymous, p.(His719=), and the substitution that generates p.(Arg720Ter) is c.2158C>T, so the original notation is internally inconsistent and cannot be corrected without the raw data [7, 11]. The Pakistani variant cited as «c.1308_1310delACT p.255delTyr» normalises to p.(Ala437del), incompatible with the protein stated [7, 63]. In the Lebanese family, the article uses two notations for the same allele (p.Ser233 and p.Ser223), and the normalised form is c.665_666dup p.(Ser223GlyfsTer4) [7, 35]. As for the EEM families of Kjaer 2005, OMIM (with segregation data for both) and Karti and colleagues assign the c.830del deletion to the Brazilian family and the missense p.(Asn322Ile) to the Danish family; table 1 of Basel-Vanagaite 2010 assigns them the other way round, and Kjaer's abstract numbers only families 1 and 2 [6, 12, 27, 29]. The assignment adopted follows OMIM.

2.6. Splicing variants and functional validation

Canonical splicing variants (c.160+1G>A, c.1425-1G>T/A, c.1796-2A>G, c.1795+1G>C, c.2280+1G>T) are classified as loss of function. Singh and colleagues showed that RNA from peripheral blood lymphocytes allows the effect of these variants to be studied: in c.1796-2A>G, reverse transcription followed by amplification (RT-PCR) produced a 404 bp fragment instead of the 611 bp of the control, and sequencing confirmed exon 13 skipping, with loss of the extracellular calcium-binding domain and part of the transmembrane segment [38]. Non-canonical intronic variants require validation: in c.391-3C>G, a minigene assay in COS7 cells showed partial retention of intron 4 (with frameshift and predicted nonsense-mediated decay) and exon 5 skipping (in-frame deletion of 52 amino acids in EC1) [46]. In the same patient, structural modelling predicted that p.(Asn542Ser) lengthens the hydrogen bond between Asn542 and His575 from about 3.5 to 3.8 Å, with instability of the EC4–EC5 interface [46].

2.7. Structural variants and the CDH1–CDH3 locus

Whole-exon deletions in the homozygous state have been described: a deletion of about 1.8 kb mediated by recombination between Alu elements in introns 2 and 3, which removes exon 3 (c.161-811_246+1044del, authors' notation) and joins exon 2 to exon 4 out of frame, in an Iranian Azeri family with 12 affected individuals [14]; an 8815 bp deletion encompassing exons 12 and 13 [32, 66]; and a deletion of exons 1 and 2 in an Iranian patient with rod-cone dystrophy, detected only by exome CNV analysis [37]. In compound heterozygosity, an exon 2 deletion was identified in a North American patient of African ancestry, after a first non-diagnostic genetic test [15]. In the Iranian family, the exome did not detect the deletion and pointed to a homozygous missense variant in FHOD1, located 1.4 Mb away within the same region of homozygosity and in linkage disequilibrium; a knock-in mouse for this variant had no phenotype, and only the whole genome identified the cause [14].

The proximity of CDH1 has two consequences. In families with hereditary diffuse gastric cancer, an Alu-mediated germline deletion of 193,593 bp has been described that removes all of CDH3 and exons 1 and 2 of CDH1; carriers are heterozygous and no CDH3 phenotype is described in them, as expected in a recessive disease [39]. Conversely, Kadhi and colleagues hypothesised that E-cadherin, co-expressed with P-cadherin in the mouse apical ectodermal ridge, partially compensates for the loss of P-cadherin and explains normal limbs in patients with truncating variants, without having found a CDH1 variant in the family studied [35]. Somatic overexpression of P-cadherin is associated with high-grade breast carcinoma and melanoma progression, and P-cadherin is the target of antibodies and radiopharmaceuticals in phase 1 and 2 oncology trials; these data concern tumours and bear no relation to the germline disease [6, 17].

2.8. Genotype–phenotype correlations and modifiers

No genotype–phenotype correlation has been established. The severity of limb anomalies varies between and within families: c.830del caused EEM in the Brazilian family and HJMD without limb anomalies in Turkish, Portuguese and Spanish patients, and, in one of Kjaer's families, one brother had severe ectrodactyly and another only syndactyly [11, 12]. Syndactyly is the most constant limb finding among patients with limb anomalies, regardless of variant type [12]. Early truncating variants predictably lead to mRNA decay and absence of protein; the only exception in Hull's series, p.(Gly786AlafsTer7), located beyond the predicted boundary for that decay, was associated with the latest visual onset in the series (17 years), with a retinal phenotype similar to the others [11]. The p.(Arg503His) variant has been associated by some authors with a milder phenotype, without a comparative study to demonstrate it [32]. Intrafamilial variability in hair morphology, skin findings and age at onset of visual loss has been documented since 2003 in 22 patients [49].

2.9. Borderline allelic phenotypes

Basit and colleagues described two consanguineous Pakistani families with autosomal recessive hypotrichosis, sparse eyebrows and eyelashes, normal fundoscopy and electrophysiology and homozygous truncating CDH3 variants; linkage analysis was compatible with digenic inheritance, with a second locus at 12q21.2–q22 without an identified gene [12, 34]. Kadhi and colleagues reported a non-consanguineous Lebanese family with generalised simple hypotrichosis, without macular degeneration and with a normal full-field electroretinogram (ERG) at several time points, in a 19-year-old female proband and her mother, both homozygous for c.665_666dup; the authors recommended repeat eye examination every two years [35]. Because the full-field ERG can be normal in patients with maculopathy confirmed by multifocal ERG (mfERG), the absence of retinopathy in these families should be read as limited by the method and the follow-up [35, 51].

At the opposite pole, three Druze siblings aged 30–34 years with severe autosomal recessive pigmentary retinopathy, residual ERG and normal dermatological examination were homozygous for p.(Val734Met), in the intracellular domain; the variant was found in the heterozygous state in 4 of 94 controls from the same village, and the evidence rests on a single nuclear family [36]. In a consanguineous Iranian cohort of 192 patients with IRD, one patient with early rod-cone dystrophy and macular dystrophy had a homozygous deletion of exons 1 and 2, but it was not possible to re-examine the hair [37]. Khan and Bolz described a man without frank hypotrichosis, only slow hair growth, with typical retinopathy [60]. These cases justify including CDH3 in IRD panels and, when a CDH3 variant is found in a patient with apparently isolated retinopathy, actively looking for the ectodermal signs.

3. Clinical Spectrum and Manifestations by System

3.1. Temporal sequence

The disease has a constant order of appearance. Hair is the first sign: in Sprecher's Druze families, patients were born with apparently normal hair, lost their scalp hair at about 3 months and had partial regrowth of short, sparse hair at puberty [2]. In other patients, hair is sparse from birth or never needs cutting [52, 63]. Central visual disturbance appears in childhood or adolescence: between 3 and 21 years in the Druze families, between 3 and 17 years in Hull's British series, with most patients presenting in the first decade [2, 11]. In EEM syndrome, the limb anomalies are present at birth [4]. Reviews place legal blindness between the second and fourth decades, with great variability between and within families (section 11) [29, 40, 49].

All available data describe the natural history, because there is no disease-modifying treatment. The cohorts are small and almost all retrospective: 16 Israeli patients followed for up to 14 years (Leibu, 2006), 22 patients compared by Indelman in 2003, 10 British patients assessed between 2013 and 2016 (Hull) and 4 Saudi patients seen between 2010 and 2014 (Khan and Bolz); the remaining knowledge comes from reports of one to three patients [3, 11, 49, 60].

3.2. Hair, skin and appendages

Scalp hair is short, fine, sparse and fragile, breaks easily and can be pulled out painlessly with gentle traction [9, 52]. In Hull's series, all patients had fine scalp hair at least since early childhood, with normal hair in other regions, including the eyebrows [11]. Microscopy of the shaft shows pili torti (periodic twists of the shaft about its axis), considered one of the most frequent changes and a useful marker, although absent in some patients; pseudomonilethrix, fusiform flattening of the shaft, longitudinal grooves, scaling and absence of cuticle have also been described [2, 12, 14, 26]. Trichoscopy in recent patients has shown variation in shaft calibre, coiled and circle hairs and a predominance of follicular units with a single hair [40, 46]. Microscopy may be normal, as in a 4-year-old Portuguese child in whom light and electron microscopy of the shafts showed no changes [52]. Scalp biopsy is not very specific: reduced ratio of terminal to vellus follicles and an excess of follicles in catagen and telogen, without inflammation [2, 43].

Severity varies greatly within the same family. In the Iranian family with 12 affected individuals, aged 1 to 76 years, some had almost normal growth at 17 years and others almost total loss at 13 [14]. Eyebrows, eyelashes and body hair are generally normal in HJMD, but there are exceptions: a Spanish boy had hypotrichosis of the face (no beard or moustache at 16 years) and arms, and reduced eyebrows and eyelashes in childhood that later normalised [9]. In EEM syndrome, hypotrichosis typically affects the scalp, eyebrows and eyelashes [4, 5].

The skin has no change specific to the disease. Atopic dermatitis, keratosis pilaris, centrofacial lentiginosis, eczematous dermatitis, dry skin and follicular hyperkeratosis have been described, in unknown proportions and without a demonstrated causal relationship [9, 14, 32, 49]. Sweating is normal, which distinguishes these diseases from hypohidrotic ectodermal dysplasia [12]; the hypohidrosis reported in a Colombian boy is an isolated exception [32]. Transient nail dystrophy, slow nail growth and hypoplastic nails have been observed in patients with HJMD [12, 14, 50, 60].

3.3. Retina and visual function

The fundus finding is a bilateral and symmetrical chorioretinal atrophy of the posterior pole. In Hull's series, atrophy of the retina, RPE and choroid extended nasally to the optic disc in all patients except the youngest, went beyond the vascular arcades in seven and exposed the sclera in the two oldest patients; all had hyperpigmented spots or clumps in the macula, which increased over time in three of five patients with serial imaging [11]. Khan and Bolz described a circumscribed maculopathy whose borders do not respect the arcades and which contains polygonal pigment clumps, a pattern that allows the disease to be suspected even without evident hypotrichosis [60]. The periphery is sharply spared in most patients, although progression may reach the mid-periphery [10, 40, 50].

On fundus autofluorescence (FAF), the atrophic area is confluently hypoautofluorescent and surrounded by a ring of relatively increased autofluorescence (six of seven of Hull's patients), sometimes with hypoautofluorescent dots within; there are no hyperautofluorescent lesions or lipofuscin accumulation as in Stargardt disease [11, 40]. On OCT, there is variable atrophy of the outer retina, RPE and choroid, outer retinal tubulations (seven of nine of Hull's patients) and, sometimes, cysts of the outer nuclear layer; the ellipsoid zone may be preserved at the fovea for years [11, 44]. Lin and colleagues observed, in a Chinese boy followed for two years, replacement of the tubulations by diffuse outer retinal atrophy, and suggest that tubulations are a transient pre-atrophic stage [46]. OCT angiography shows marked attenuation of choriocapillaris flow beneath the lesion, with retinal plexuses preserved outside the atrophy [45, 46]. Adaptive optics, in two German siblings, showed loss of central photoreceptors with a preserved peripheral cone mosaic, correlated with FAF [51].

Electrophysiology shows mainly macular dysfunction. In Hull's series, the pattern ERG (PERG) was undetectable in three of five patients and subnormal in a fourth, and the full-field ERG showed mild generalised cone and rod dysfunction in two; in one patient, the PERG deteriorated over nine years while the full-field ERG remained stable, and the EOG was normal in the two patients tested [11]. In Japan, the full-field ERG of a 22-year-old woman was normal for rods and at the lower limit for cones, whereas the mfERG was greatly reduced centrally [44]. Leibu and colleagues found, in 16 patients, a subnormal ERG with a similar deficit in light and dark adaptation, declining gradually over up to 14 years, a normal EOG and markedly subnormal pattern visual evoked potentials (VEP) even when visual acuity was little affected; it was this diffuse dysfunction that led them to propose the designation cone-rod dystrophy [3]. A Syrian girl examined at 9 and 14 years progressed from moderate to severe reduction of the scotopic and photopic ERG, with an undetectable PERG at 9 years and protan-deutan defects, a cone-rod dystrophy phenotype [50]. In a 6-year-old Saudi girl, the full-field ERG showed marked cone dysfunction and minimal rod reduction [63].

Visual acuity (VA) was reduced in all of Hull's patients at presentation; at the last visit (mean age 27.2 years; range 5 to 57), it ranged from 0.18 logMAR (Snellen 20/30) to hand movements, and the oldest patients had 1.0 logMAR (20/200) or worse in the better eye [11]. There are exceptions with preserved VA: two Arab siblings homozygous for p.(Tyr615Ter) had no visual complaints despite marked macular pigmentary changes, and the eldest son of one family had normal VA at 14 years [11, 55]. Complaints include loss of central vision, central scotoma with a preserved peripheral field, photophobia, altered colour vision, reduced contrast sensitivity and glare [10, 45, 52]. Nyctalopia is rare: one patient in Hull's series, some patients in the Iranian family and the Saudi girl [11, 14, 63]. No nystagmus was reported in Hull's patients [11]. High hypermetropia (up to +9.00 dioptres), strabismus and amblyopia are modes of presentation in the young child [11, 32, 40, 63].

The complications described are macular neovascularisation (MNV), identified by OCT angiography in an 11-year-old Italian boy with a fibrovascular membrane in the left eye [45], serous detachment of the residual retina over the atrophic area in a 57-year-old patient [11] and macular hole in a patient progressing to severe central atrophy [60]. Intraretinal cysts were observed in the Portuguese child from Braga and in a British patient [11, 52]. In cases of EEM described in Turkey, cataracts were reported, together with alopecia, absent eyebrows and oligodontia [67].

3.4. Limbs

Limb anomalies define EEM syndrome, but they also occur, minimally, in patients classified as HJMD. Table 3.1 brings together the patients with limb anomalies and CDH3 variants reviewed by Basel-Vanagaite and colleagues, with the assignment of the variants to Kjaer's families corrected according to OMIM (section 2.5).

In Hull's series, one patient had mild webbing between the second and third fingers of the left hand and camptodactyly of the second toes, and the other nine had normal limbs [11]. Other minimal findings were absence of the left index fingernail in an Iranian boy, bilateral clinodactyly of the fifth finger in an Albanian child, a hypoplastic nail on the second finger of the left hand and onychodystrophy of the toes in a Mexican girl, and hypoplasia of the distal phalanx with a hypoplastic nail of the second finger in a Yemenite patient [12, 32, 40, 62]. The hands are more affected than the feet, and ectrodactyly of the feet has been described in only one patient in Ohdo's series [12]. Syndactyly is the most constant finding and may be the only expression in the limbs [12].

3.5. Teeth

Dental changes are characteristic of EEM syndrome: selective tooth agenesis or partial anodontia, microdontia, widely spaced teeth and caries, and enamel hypoplasia [4, 5, 9]. In patients with HJMD, discoloured deciduous teeth, agenesis of an upper deciduous molar with the successor's germ present in a Saudi girl and enamel anomalies with serrated incisors in a Colombian boy have been described [12, 32, 63]. Most patients with HJMD have normal dentition [11, 29, 52].

3.6. Growth and findings of uncertain association

Orphanet annotates «short stature» as very frequent in HJMD. None of the series read documents it: Hull's series does not mention it, and the only detailed anthropometric data are those of a Yemenite patient with EEM who, at 20 years, weighed 46.5 kg (−2 standard deviations), was 158 cm tall (−1 standard deviation) and had a head circumference of 51.5 cm (−2 standard deviations) [1, 11, 12]. The annotation may reflect the case of Kuster and colleagues with alopecia, macular degeneration and growth retardation, which the authors considered a distinct entity [5]. Cognitive development is normal in the patients assessed [12, 32].

Several findings have been described in isolated patients without a demonstrated relationship with CDH3: gastrointestinal nodular lymphoid hyperplasia and a colorectal polyp at 2 years [9]; severe neonatal micrognathia in a boy with gestational exposure to isotretinoin in the first month [32]; infertility in a Saudi man [60]; and recurrent sinopulmonary infections with hearing loss in a girl who also had a homozygous DNAH5 variant, explained by concomitant primary ciliary dyskinesia [64]. The Orphanet frequencies for «immunological hypersensitivity», «melanocytic naevus» and «skeletal dysplasia» in HJMD have, in the sources consulted, no series that quantifies them [1].

3.7. EEM syndrome

Reports of EEM are few and, for the most part, predate molecular diagnosis. In the Brazilian family of Balarin Silva, two siblings had the complete form (hypotrichosis, dental anomalies, syndactyly and retinal changes with prominent pigmentation at the posterior pole), another sibling only syndactyly and a first cousin ectrodactyly of the right hand, without dental, skin or ocular changes; over ten years of follow-up, the VA of the affected siblings declined gradually and the degenerative changes progressed [5, 23]. In a 41-year-old Japanese woman with EEM, the retinal degeneration progressed minimally over 11 years [22]. The Yemenite Jewish siblings had a similar but milder picture [24]. The Yemenite patient described by Basel-Vanagaite was diagnosed with macular degeneration at 9 years and, at 18, had symmetrical macular atrophy and scarring with perimacular haemorrhages; her brother had visual disturbance at 7 years and VEP and ERG with asymmetric dysfunction at 10 years [12]. In Brazil, two sisters of consanguineous parents had bilateral retinal dystrophy, alopecia and bone changes or syndactyly [68]. In Turkey, three siblings of consanguineous parents had extensive retinochoroidal atrophy with diffuse pigmentation at the posterior pole and mild arteriolar attenuation, which the authors considered the feature distinguishing EEM from other syndromes with ectrodactyly [67].

3.8. Frequency of signs according to Orphanet

Tables 3.2 and 3.3 reproduce the frequencies annotated by Orphanet in Human Phenotype Ontology (HPO) terms. They are estimates by class, not measurements in cohorts, and some do not coincide with the published series.

Source of the frequencies: [1].

Source: [4].

3.9. Patients published up to 2010

Table 3.4 summarises the patients with CDH3 variants reviewed by Basel-Vanagaite and colleagues in 2010, with the variants in current nomenclature.

Source: table 2 of [12], with VariantValidator nomenclature [7]. The row for the Brazilian family (c.830del, EEM) does not appear in this table in the original publication.

4. Clinical Recognition and Warning Signs

4.1. When to suspect

The sign that allows the diagnosis is the association of sparse scalp hair from birth or the first months of life with loss of central vision in childhood. Hull and colleagues concluded that a child with macular dystrophy can be diagnosed from examination of the scalp, and hypotrichosis distinguishes this disease from other juvenile macular dystrophies [11]. In practice, suspicion arises in three contexts.

In the infant or young child seen for hair that «does not grow», has never needed cutting or is easily pulled out, congenital hypotrichosis should prompt careful fundoscopy, because macular pigmentary changes point to CDH3; this recommendation, formulated by Indelman and colleagues and taken up by other authors, applies even when the child has no visual complaints [52, 55, 69]. The maculopathy may not yet be visible in the young child, as in a 6-year-old Colombian boy without pigmentary changes, which requires follow-up [32].

In the child with low central vision, photophobia, high hypermetropia, strabismus or amblyopia without an evident cause, and with atrophic maculopathy, the scalp hair should be examined. Hypermetropia, strabismus and amblyopia were the mode of presentation in children aged 8 months to 6 years [10, 32, 40, 63]. The fundus pattern (posterior pole atrophy contiguous with the disc, extending to or beyond the arcades, with polygonal pigment clumps and a spared periphery) and FAF without hyperautofluorescent lesions point to CDH3 rather than ABCA4 [11, 40, 60].

In the adult with pigmentary retinopathy or cone-rod dystrophy and a CDH3 variant on the panel, slow hair growth or hair fragility, hypoplastic nails, dental changes and subtle hand anomalies should be sought, because there are patients without frank hypotrichosis [36, 37, 60].

4.2. Signs suggesting EEM syndrome

In a child with HJMD, mild syndactyly of the hands or feet, camptodactyly, clinodactyly, hypoplastic nails, hypoplasia of the distal phalanges, scars from syndactyly surgery and tooth agenesis should be actively sought, because EEM syndrome may pass for HJMD when the limb anomalies are subtle [11, 12, 40]. A child with ectrodactyly and ectodermal dysplasia, without cleft lip or palate, should undergo fundoscopy, because the maculopathy is what separates EEM from the TP63 syndromes [5, 67].

4.3. Diagnostic pitfalls

Published reports show long diagnostic delays and repeated causes. Dermatology reassures the family that the hair will grow; ophthalmology does not link the hair to the maculopathy; and a hairpiece or wig conceals the key sign. In a family of Russian ancestry, the diagnosis was only made about two decades after the first sign, when the patients themselves found the association on a patient website [10]. The misdiagnoses described include Stargardt disease, North Carolina macular dystrophy, «inverse retinitis pigmentosa» and keratosis follicularis spinulosa decalvans (Siemens syndrome) [9, 10, 45]. A single ABCA4 allele found on targeted testing does not explain a maculopathy and may delay diagnosis [9].

4.4. Urgent signs

The disease progresses slowly and has no systemic emergencies. Sudden loss of central vision, metamorphopsia or macular haemorrhage in a known patient justify prompt ophthalmological assessment with OCT, OCT angiography or fluorescein angiography, to exclude macular neovascularisation, which has been described in this disease [38, 45]. Signs of depression or isolation in the face of visual loss should be sought at each visit [13].

4.5. Newborn screening

There is no newborn screening for CDH3 disorders, and none of the sources consulted proposes it. There is no biochemical marker, and the diagnosis depends on clinical recognition and genetic testing.

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Criteria

There are no formal published diagnostic criteria for HJMD or EEM syndrome, nor any specific guideline. The clinical diagnosis rests on the association of congenital- or neonatal-onset scalp hypotrichosis with childhood-onset macular dystrophy or posterior pole chorioretinal atrophy; the presence of ectrodactyly, syndactyly or other limb anomalies, with or without dental changes, classifies the picture as EEM syndrome [2, 5, 11]. Confirmation requires biallelic pathogenic or likely pathogenic CDH3 variants, because there is no biochemical marker or routine functional test [11, 13].

5.2. Ophthalmological assessment

The 2022 AAO guidelines on the assessment of patients with IRD are the available reference; they are not specific to CDH3, and their application to these diseases is an extrapolation from the classes of dystrophies involving the macula and chorioretinal degenerations [13]. At the first visit, they recommend ocular, medical and family history with a pedigree, attention to mood and signs of depression, VA with refraction and standardised charts (ETDRS, Early Treatment Diabetic Retinopathy Study), slit-lamp examination with intraocular pressure, ophthalmoscopy under mydriasis, fundus photography, FAF, OCT with high-density cubes and recording of central thickness as a baseline, visual fields and electrophysiology according to the standards of the International Society for Clinical Electrophysiology of Vision (ISCEV) [13]. In maculopathies, fundus photographs should be requested sparingly, because of the risk of light toxicity, only when useful for documenting progression or eligibility for trials; FAF should use reduced illumination or longer wavelengths, and near-infrared autofluorescence or reflectance are alternatives [13].

The choice of electrophysiology has consequences in CDH3 disease. The full-field ERG may be normal or almost normal, as in two German siblings and a Japanese patient, whereas the mfERG or PERG show marked central dysfunction; the mfERG or PERG are therefore the sensitive tests for documenting the maculopathy [11, 44, 51]. The AAO recommends the full-field ERG in young patients with disease apparently limited to the macula to exclude generalised dysfunction, and the mfERG or PERG to detect and monitor macular disease, with limited value when central loss is marked or fixation unstable [13]. Microperimetry is useful with eccentric fixation and detected, in an Albanian child, functional progression not visible in VA or anatomy over four years [13, 40]. The EOG is normal in most patients [3, 11].

In the young child, functional tests are unreliable. The AAO indicates, below 2 years of age, ERG with skin electrodes or handheld OCT with the infant swaddled and drowsy in the dark, and, between 2 and 6 years, examination under sedation when the quality of the examination justifies it, weighing the risks and knowing that sedation alters the ERG; visual fields before 7 years are difficult and improve with repetition [13]. In the face of acute visual loss, fluorescein angiography or OCT angiography allows neovascularisation to be excluded [38, 45].

5.3. Dermatological, limb and dental assessment

Assessment of the hair includes the history of neonatal hair, haircuts, fragility and colour, the pull test, trichoscopy and light microscopy of the shafts, which may show pili torti or pseudomonilethrix but may also be normal [31, 40, 52]. Scalp biopsy is not very specific and, in the cases described, did not advance the diagnosis, and even suggested Siemens syndrome [9, 43]. In the investigation of congenital hypotrichosis, shaft microscopy excludes loose anagen syndrome (ruffled cuticles, hockey-stick bulbs), and vitamin, amino acid and biotinidase assays exclude metabolic causes [52]. Assessment of a possible ectodermal dysplasia also includes the nails, sweating, heat intolerance, tooth eruption, the number and shape of the teeth (20 deciduous and 32 permanent) and examination of the hands and feet, with radiographs when there are anomalies [70].

5.4. Genetic testing

The AAO recommends genetic testing for most patients with IRD, to confirm the diagnosis and optimise management, with a next-generation sequencing panel for retinal dystrophies as the first step, and indicates that the panel should include the genes for syndromic forms even without syndromic signs, because these may appear later or go unnoticed [13]. The genetic cause is identified in 56–76% of patients with IRD in general [13]. In the Portuguese national IRD registry (IRD-PT), with 1369 patients from 1125 families enrolled up to March 2024, the overall yield was 72.3%, and the four cases of HJMD, from four families, all had a genetic diagnosis [71]. A macular dystrophy panel used in the United Kingdom, with 17 genes, already included CDH3 in 2016 [11]. The exome is the option when the panel is inconclusive, and the genome is used mainly in a research context [13].

When EEM syndrome is suspected, CDH3 sequencing should be performed together with that of TP63, which accounts for most syndromes with ectrodactyly and ectodermal dysplasia; both genes belong to the p63 pathway group of the 2022 classification [16].

5.5. Limitations of testing

A negative result does not exclude the disease. Whole-exon deletions of CDH3 (exon 2, exon 3, exons 1–2, exons 12–13) are not detected by point variant analysis of a panel or exome: in the Iranian family, only homozygosity mapping and the whole genome identified the exon 3 deletion, and in the North American patient of African ancestry the exon 2 deletion was only found on reanalysis with comprehensive genomic testing and CNV analysis, after a first result suggesting another diagnosis [14, 15, 37, 66]. The request should therefore include CNV analysis, and a result with a single variant in a patient with a typical phenotype should lead to a search for a deletion on the other allele.

Apparent homozygosity without study of both parents does not distinguish true homozygosity from hemizygosity due to deletion of the other allele; several recent reports have homozygosity without complete segregation (father unavailable, adopted daughter, exome without Sanger confirmation or parental study) [32, 40]. Intronic variants outside the canonical splicing positions require RNA study, which can be performed on peripheral blood lymphocytes, or a minigene assay [38, 46]. Underrepresentation of populations of African ancestry in allele frequency databases hampers interpretation of variants in these patients [15]. A test targeted at the Druze founder variant does not exclude the disease in that population, in which at least one other variant has been described [8]. No mosaicism was described in the sources consulted.

5.6. Interpretation and report content

The report should state the gene, the versioned transcript (NM_001793.6), the HGVS nomenclature of each variant, the zygosity, the phase (cis or trans) when studied and the classification according to the criteria of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) [7, 40]. Published classifications mainly use the criterion of loss of function in a gene whose mechanism is loss of function (PVS1), rarity in the population (PM2), phenotypic concordance (PP4) and previous classifications in databases (PP5) [35, 40, 44]. The c.1837dup variant was published as a VUS and reclassified as pathogenic with new cases, an example of how classification changes as data accumulate [40]. A rare missense variant in trans with a pathogenic variant, in a patient with a typical phenotype, remains a VUS without functional or segregation data, as with p.(Ile553Asn) [15]. Segregation in the parents confirms the phase, and analysis of affected siblings strengthens the evidence [9, 46].

Interpretation and communication of the result should be carried out by an experienced physician or genetic counsellor, with counselling after the result and repeated later in life in early-onset diseases; a negative or inconclusive result may be reviewed with new genes, new technologies or new interpretation, and a definitive result does not need to be repeated [13].

5.7. Algorithm

Table 5.1 summarises a diagnostic pathway built from the sources cited; it is not a validated algorithm.

5.8. Prenatal and preimplantation diagnosis

In retinal dystrophies with molecular confirmation, prenatal diagnosis is available and carrier detection is possible when the causal variant has been identified [72]; by the same principle, preimplantation genetic testing is technically applicable when both familial variants are known. Basel-Vanagaite and colleagues recommend fetal ultrasound of the limbs in pregnancies of parents of patients with HJMD or EEM, because of the possibility of limb anomalies even in families with HJMD [12]. The sources consulted do not describe cases of prenatal or preimplantation diagnosis in these diseases.

6. Differential Diagnosis

6.1. Comparative table

The differential diagnosis depends on the mode of presentation: congenital hypotrichosis before visual loss, juvenile macular dystrophy without attention to the hair, or ectrodactyly with ectodermal dysplasia. Table 6.1 brings together the entities described in the sources consulted (AD, autosomal dominant; AR, autosomal recessive; XL, X-linked).

6.2. Syndromes with ectrodactyly and ectodermal dysplasia

EEC syndrome and the other TP63 syndromes share ectodermal dysplasia and split hand/foot malformation with EEM, which is explained by the same pathway: p63 activates CDH3 transcription [16, 28]. Three criteria distinguish them: the autosomal dominant inheritance of the TP63 syndromes, against the recessive inheritance of EEM; cleft lip or palate, present in about two thirds of patients with EEC and absent in EEM; and macular dystrophy, which exists only in EEM [67, 73, 75]. EEC syndrome and the classification of ectodermal dysplasias are part of the differential. The Oxford Desk Reference: Clinical Genetics and Genomics (Oxford Desk Reference) places TP63 at 7q21–22, an error that probably confuses the gene with the DLX5 split hand/foot malformation locus; the gene is at 3q28 [70, 73, 75].

6.3. Congenital hypotrichoses before visual loss

In the child with hypotrichosis and still normal vision, the differential diagnosis is that of the paediatric hypotrichoses. The algorithmic approach of So and colleagues begins by separating isolated from syndromic forms, through observation of the hair, teeth, nails and sweating, and uses the history of neonatal hair and the inheritance pattern to guide the genetic panel [31]. Hypohidrotic ectodermal dysplasia is distinguished by hypohidrosis and severe hypodontia with conical teeth, which are not part of CDH3 disorders [12, 31]. Hereditary hypotrichosis simplex and the hypotrichoses due to LIPH, LPAR6 or DSG4 can only be reliably distinguished by genetics and by the absence of maculopathy during follow-up, because CDH3 maculopathy may only appear after the first years [10, 32, 35]. Fundoscopy is therefore mandatory in the assessment of any congenital hypotrichosis [55, 69].

6.4. Juvenile macular dystrophies

ABCA4-related Stargardt disease is the most frequent cause of juvenile-onset macular dystrophy and the most common misdiagnosis. Like CDH3 disease, it may present with loss of central vision, macular atrophy and electrophysiological dysfunction limited to the macula (about a third of children) or generalised (the remaining two thirds), and not all children have flecks at presentation [11]. Examination of the hair is the simplest way to separate them; in addition there is the classic (but not universal) peripapillary sparing in ABCA4 disease, against atrophy contiguous with the disc in CDH3, the smaller extent of macular atrophy in the child with Stargardt and the absence of hyperautofluorescent lipofuscin lesions in CDH3 [11, 40]. A single ABCA4 allele is frequent in the population and does not confirm Stargardt disease [9]. Best disease, North Carolina macular dystrophy and dominant drusen are autosomal dominant and have no ectodermal changes [10, 72].

When CDH3 disease presents as cone-rod dystrophy, the differential diagnosis broadens to the cone and cone-rod dystrophies and the other retinal cadherinopathies; no other known macular dystrophy is associated with hair loss [10, 33, 50]. Sjögren-Larsson syndrome combines juvenile macular dystrophy with ichthyosis and neurological signs, which separates it easily [80, 81]. In the Druze population, the ABCA4 founder variant and the CFAP418 variant cause retinal dystrophies in the same community in which CDH3 has a founder effect, which makes examination of the hair and the full panel necessary even when there is a familial diagnosis [8].

7. Treatments and Risk Management Strategies

7.1. Principles

There is no treatment that modifies the course of CDH3 disorders. No medicine is authorised for HJMD or EEM syndrome in any jurisdiction, there is no approved gene therapy for CDH3 and no clinical trial is registered on ClinicalTrials.gov for these diseases [13, 17]. No published series describes an intervention that has altered the retinal, hair or limb phenotype [11, 40]. Management is supportive and multidisciplinary: ophthalmology with experience in IRD, medical genetics, dermatology, hand surgery and orthopaedics in EEM syndrome, dentistry, visual rehabilitation, psychological support and liaison with the school [10, 13, 40]. As there is no drug directed at the disease, management rests on non-pharmacological interventions.

7.2. Ophthalmological management and visual rehabilitation

Optical correction of refractive errors is part of management, and significant hypermetropia or astigmatism is frequent in the child [11, 40, 63]. The AAO guidelines ask that the patient and family be informed about the hereditary and probably progressive nature of the disease and that the ophthalmologist encourage low vision rehabilitation, work with the school, promote orientation and mobility training and assess emotional status; they recommend low vision assessment at baseline and every 1–2 years, information about patient registries and clinical trials, vocational rehabilitation services, assessment for assistive technologies and, when indicated, a guide dog [13]. Patients with central loss and preserved periphery are candidates for optical and electronic low vision aids and digital assistive technologies at school, at work and at home [10, 11].

Narayan and Moosajee also advised a balanced diet rich in fruit and green leafy vegetables, sunglasses with ultraviolet protection in bright light and blue light filters on screens; these are measures based on the authors' opinion, without evidence of efficacy in this disease [10]. The AAO recommends that exposure to retinotoxic medicines be recorded in the clinical history and that, in maculopathies, fundus photographs be used sparingly because of the risk of light toxicity [13].

Macular neovascularisation is a described but rare complication: the Italian boy in whom it was identified by OCT angiography at 11 years maintained the same VA at 12 years, and the authors do not describe treatment [45]. The sources on CDH3 do not describe the treatment of this complication; acute visual loss should be investigated with angiography to exclude neovascularisation, and treatment, when indicated, follows medical retina practice for macular neovascularisation, without data specific to this disease [38, 45].

7.3. Hair, skin and teeth

No effective treatment for the hypotrichosis is described in the sources consulted. A hairpiece or wig are cosmetic options used by adult patients, with the caveat that they conceal the sign that allows the diagnosis in siblings and in the eye clinic [10]. Atopic dermatitis, keratosis pilaris and dry skin, when present, are treated as in other patients [9, 32]. Sweating is normal, so the hyperthermia prevention measures typical of hypohidrotic ectodermal dysplasia are not necessary [12].

In EEM syndrome, tooth agenesis, microdontia, widely spaced teeth and enamel hypoplasia require dental follow-up and oral rehabilitation [4, 9]. The oral health care parameters for patients with ectodermal dysplasias of the National Foundation for Ectodermal Dysplasias (NFED) were not obtained, and the sources consulted do not define intervals or specific techniques for EEM. In ectodermal dysplasias in general, the Oxford Desk Reference recommends referral to dermatology and dentistry and artificial tears if there is xerophthalmia [70].

7.4. Limbs

Cutaneous syndactyly can be corrected surgically, as in the Yemenite brother with syndactyly between the second and third fingers operated on in childhood [12]. In the syndromes with split hand/foot malformation of the EEC group, surgical reconstruction and rehabilitation focused on limb structure allow good functional adaptation; extrapolation to EEM is based on the similarity of the malformations and not on series of patients with EEM [75]. The sources do not describe the ideal age for surgery or functional outcomes in EEM.

7.5. Psychological, educational and social support

Progressive visual loss in childhood and hypotrichosis affect self-esteem and quality of life; reports describe emotional burden and low self-esteem, and the AAO recommends mental health support from diagnosis and throughout progression, with support groups and individual counselling [10, 13]. Children with retinal dystrophies may need psychological and educational support with a multidisciplinary team [40, 72].

7.6. Regulatory status

Table 7.1 summarises the regulatory status at the date of this update.

7.7. Clinical trials

A search of ClinicalTrials.gov in October 2026, with the terms «CDH3», «hypotrichosis juvenile macular», «P-cadherin» and «EEM syndrome», found no trials in HJMD, EEM syndrome or CDH3 retinal dystrophy [17]. The records with CDH3 are oncology trials in which P-cadherin is a tumour target, with no therapeutic relationship to the germline disease (Table 7.2).

Source: [17].

The absence of registrations does not demonstrate the absence of preclinical research; no published experimental gene therapy study for CDH3 was found in the sources consulted either [17].

7.8. Approaches under investigation

Singh and colleagues proposed the disease as a candidate for gene augmentation therapy: the CDH3 coding sequence, of about 2.5 kb, fits within the limit of about 4.7 kb of a recombinant adeno-associated virus (rAAV) with promoter and inverted terminal repeats; rAAV2 has a favourable safety profile in retinal trials and has already been used with an RPE-specific promoter in a trial for MERTK pigmentary retinopathy; and the relative preservation of central foveal thickness and the slow progression suggest a therapeutic window, dependent on the extent of foveal sparing. This is a conceptual proposal, without experimental vector data [38]. The review by Yusuf and colleagues considers the retinal cadherinopathies a potentially treatable group and summarises the likely molecular approaches for each, with no approved therapy [33]. Marziali and colleagues point to subretinal gene therapy as a possible strategy, given the pattern localised to the posterior pole, the limited progression and the preserved periphery [40]; Lin and colleagues suggest that outer retinal tubulations may mark a window for intervention before atrophy [46].

Narayan and Moosajee mentioned, speculatively, premature termination codon read-through drugs, such as ataluren, for patients with nonsense variants, without data in this disease [10]. A patient family website in Germany publishes information on these approaches and, in November 2025, argued that only variants escaping mRNA decay would be candidates; this is an opinion not peer-reviewed, without support in the scientific sources consulted [82].

7.9. Developments in the last twelve months

Between October 2025 and October 2026, the publications on CDH3 were case reports or small case series, without a trial, sponsor announcement, opinion or regulatory application. Marziali and colleagues published in July 2026 the 4- and 15-year follow-up of two Italian children, with stable VA and full-field ERG and progression detected only by microperimetry [40]. Lin and colleagues validated a non-canonical intronic variant by minigene and documented the transition from tubulations to atrophy over two years [46]. Danish and colleagues described a novel homozygous missense variant in a Saudi girl with cone-rod dysfunction [63]. Vaidyanthan and colleagues published the first case in a person of African ancestry, diagnosed after a diagnostic odyssey through an exon 2 deletion detected on CNV analysis [15]. The level of evidence of all these publications is that of a case report.

7.10. Pregnancy, perioperative period and transition

The sources consulted do not describe pregnancies in women with HJMD or EEM, nor anaesthetic or perioperative risks specific to these diseases. Transition from paediatric to adult care has no specific recommendations; the AAO recommends repeating genetic counselling later in life in early-onset diseases, which coincides with the age of transition and with reproductive decisions [13]. Cataract surgery, when necessary, has no particularities described in this disease.

7.11. Guideline recommendations and disagreements

The only formal applicable recommendations come from generic documents: the 2022 AAO guidelines on the assessment of IRD and the 2012 AAO recommendations on genetic testing in inherited eye diseases [13, 83]. The 2022 AAO recommends follow-up every 1–2 years and low vision assessment at the same frequency [13]; recent paediatric series carry out annual follow-up [10, 40], and Kadhi and colleagues, for patients with a CDH3 variant and isolated hypotrichosis, propose eye examination every 2 years [35]. The 2012 recommendations advise against testing asymptomatic minors for untreatable diseases, except in extraordinary circumstances, and the 2022 AAO accepts testing of asymptomatic at-risk relatives only with genetic counselling [13, 83]. There is no disagreement between societies on treatment, because none issues therapeutic recommendations for this disease.

8. Reproductive Options and Family Counselling

8.1. Recurrence risks

Inheritance is autosomal recessive. When both parents are heterozygous, each pregnancy has a 25% probability of an affected child, 50% of a healthy carrier child and 25% of a child without the variant; a healthy sibling of a patient has a two-thirds probability of being a carrier. The heterozygous parents described are healthy, with the exception, of uncertain interpretation, of the heterozygous brother with mild syndactyly in the Brazilian family with EEM [5, 12]. The child of a patient is, as a rule, an obligate carrier; the risk of being affected depends on the probability of the other parent being a carrier, which is low in the general population and higher in consanguineous marriages or within founder populations (Druze, Israeli Arabs) [8, 49]. The Lebanese family with a homozygous mother and daughter illustrates pseudodominance when the unaffected parent is a carrier [35].

Basel-Vanagaite and colleagues stress that, because of the overlap between HJMD and EEM, families with a child with HJMD should be informed that a subsequent pregnancy may have limb anomalies, because the same variant and the same family can give both phenotypes [12]. In the Brazilian family, two siblings had the complete form, a third only syndactyly and a cousin only ectrodactyly [5, 23].

8.2. Cascade testing and carriers

When the familial variants have been identified, adult siblings and the reproductive partners of carriers can undergo carrier testing [13, 72]. In closed populations, alleles may have a high frequency: in the Druze village studied by Beryozkin and colleagues, with about 4300 inhabitants, the p.(Val734Met) variant was present in the heterozygous state in 4 of 94 controls [36]. For the Druze founder variant c.981del, the carrier frequency is unknown, and targeted screening does not exclude other variants in the same population [8]. The deletion that removes CDH3 and part of CDH1 in families with hereditary diffuse gastric cancer makes its carriers heterozygous for CDH3 as well, which is relevant if the partner carries a CDH3 variant [39].

8.3. Counselling of minors

Asymptomatic minor siblings of a patient raise two different questions. Carrier testing should not be performed in children, according to the international guidelines followed by Blanco-Kelly and colleagues, who proposed testing the healthy 9-year-old sibling in adulthood [9]. Diagnosis of a possibly affected sibling is different, because hypotrichosis is usually visible from the first months and clinical and ophthalmological examination of the sibling guides the need for testing; the 2012 AAO recommendations advise against genetic testing of asymptomatic minors for untreatable diseases, except in extraordinary circumstances, and the 2022 AAO accepts testing of at-risk relatives only with genetic counselling [13, 83]. Early diagnosis allows visual and educational support and possible inclusion in future studies.

8.4. Germline mosaicism

The sources consulted do not describe germline mosaicism or de novo variants in CDH3; the GeneReviews on Druze founder variants notes that the percentages of variants by population do not take into account the possibility of rare de novo variants [8]. In a recessive disease with confirmed heterozygous parents, mosaicism does not alter the 25% recurrence risk.

8.5. Prenatal and preimplantation diagnosis

With both variants known, prenatal diagnosis is available [72], and preimplantation genetic testing is technically applicable by the same principle; the sources consulted do not describe its use in this disease. Basel-Vanagaite and colleagues recommend fetal ultrasound targeted at the limbs in all pregnancies of parents of patients with HJMD or EEM, even when the affected child has no limb anomalies [12]. In retinal dystrophies without a molecular diagnosis, Oxford Desk Reference recommends storing DNA or a fibroblast line for future diagnostic use [72].

8.6. Content of counselling

Counselling should cover recessive inheritance and recurrence risks, variability between HJMD and EEM within the same family, the visual prognosis and its variability, the absence of targeted treatment and of ongoing trials, the limitations of a negative test (deletions, intronic variants), carrier testing in the family and in partners, reproductive options and patient resources; it should be repeated later in life, when the patient reaches adulthood [9, 12, 13].

9. Special Considerations: Population Diversity and Epidemiology

9.1. What the numbers measure

There is no population-based prevalence study for HJMD or EEM syndrome. Orphanet assigns both to the worldwide prevalence class below 1 per 1,000,000, an estimate by class and not a measurement [1, 4]. The published counts are counts of cases in the literature, not frequencies: about 50 cases of HJMD according to an Orphanet expert and according to reviews from 2017 and 2019; 21 families with molecular confirmation of HJMD and 6 of EEM by 2016; about 30 patients published since 1935 according to a 2017 review; about 15 cases of EEM documented by 2019; and a 2026 review that tabulates 45 patients or families with HJMD [1, 9–11, 29, 40]. Orphanet records «7 families» with EEM based on Kjaer 2005, whereas the abstract of that article describes two families with molecular study; the difference probably corresponds to the families reviewed in the full text, which could not be read [4, 27]. The differences between these counts reflect different inclusion criteria and dates, and underdiagnosis, suggested by the diagnostic delays described, prevents them from being read as a measure of the real frequency [10, 15].

9.2. Founder populations and geographical distribution

Most cases have been described in communities with high consanguinity in the Middle East and South Asia: Druze and Arabs in Israel, Turks, Pakistanis, Saudis, Jordanians, Syrians and Iranians [40, 44, 63]. There are two documented founder effects. In the Druze population of northern Israel, the variant c.981del p.(Met327IlefsTer23) was found in the homozygous state in the four families with 11 affected individuals that allowed the gene to be identified; GeneReviews includes it in the table of Druze founder variants, with an unknown carrier frequency and with at least one other variant described in this population [6, 8, 25]. In Israeli Arabs, p.(Arg503His) was found in three apparently unrelated families with a common haplotype [49]. The same variant later appeared in Syria, Jordan, Germany, a family of Russian ancestry, a Mexican Mennonite girl and a Colombian boy from northern Antioquia, a region of high consanguinity in which the authors allow for a founder effect [10, 11, 32, 50, 51, 84]. In the Druze population, p.(Val734Met), associated with pigmentary retinopathy without hypotrichosis, was found in the heterozygous state in 4 of 94 controls from one village [36]. The variant c.1425-1G>T has been described in several Pakistani families [11, 56, 58].

In the United States, published cases include a North American family with compound heterozygosity, a girl of Indian ancestry in Miami and the first patient of African ancestry, diagnosed in 2026; the authors of this last case stress the underrepresentation of this ancestry in genomic databases, which hampers interpretation of variants [15, 47, 64]. In Europe, in addition to the British and German patients, the first cases in Spain (2017) and Italy (2021) have been published, as well as cases of Albanian and French origin [9, 40, 45, 49]. In East Asia, the first cases were two Japanese siblings in 2021, followed by Chinese cases [44, 46, 85]. EEM has been described in Japan, Denmark, Brazil, Yemenite Jews and Turkey [5, 21, 22, 24, 67].

9.3. Portugal

There are several descriptions of Portuguese patients or patients of Portuguese origin. Souied and colleagues described in 1995 a brother and sister whose parents had been born in the same Portuguese village, with macular dystrophy evident only after 18 years of age [2]. Elfatoiki and colleagues reported in 2016 an 11-year-old girl of related Portuguese parents, with hypotrichosis from birth and macular dystrophy diagnosed at 5 years, without molecular study [69]. Hull's British series includes, according to the review table of Marziali and colleagues, two patients of Portuguese origin, homozygous for c.613G>A p.(Val205Met) and for c.830del p.(Gly277AlafsTer20), the latter with VA of 0.18 logMAR at 31 years after 20 years of follow-up [11, 40]. At Hospital de Braga, a 4-year-old girl homozygous for c.830del was described in 2018 [52], and Oliveira-Ferreira and colleagues published in 2019 another patient homozygous for the same variant, whose article was consulted only through the 2026 review table [40, 86]. The c.830del variant recurs in Portuguese, Spanish and Turkish patients and in a 23-year-old Jordanian woman, and in the Brazilian family with EEM; there is no haplotype study that would allow a founder effect to be asserted [6, 9, 11, 52, 53].

The Portuguese national registry IRD-PT, a cross-sectional multicentre study in six Portuguese centres with consecutive patients with a clinical diagnosis and a genetic result enrolled up to March 2024, included 1369 patients from 1125 families, with consanguinity in 20.2%. HJMD accounted for 4 patients from 4 families, all with a genetic diagnosis, which corresponds to 0.4% of solved cases and 0.5% of solved families. It is a hospital registry, and not a prevalence estimate; the authors estimate that the cohort represents about 39.7% of the 3448 patients with IRD expected in Portugal, assuming an IRD prevalence of 1 in 3000 [71].

9.4. Brazil

The first molecularly characterised family with EEM was Brazilian: two siblings with the complete form, followed for ten years, a sibling with isolated syndactyly and a cousin with ectrodactyly, described by Balarin Silva and colleagues in 1999, and later found to be homozygous for c.830del in Kjaer's study [6, 23, 27]. Matos and colleagues described two sisters of consanguineous parents with EEM [68]. In São Paulo, an 11-year-old Iranian boy with HJMD, absence of the left index fingernail and mild eczema, homozygous for c.640A>T p.(Lys214Ter), was described [62]. In Curitiba, exome sequencing identified three genotypes in Brazilian families with HJMD: c.160+1G>A in the homozygous state, c.160+1G>A with c.1063G>T p.(Asp355Tyr), and c.1795+1G>C in the homozygous state [40, 54]. The sources consulted have no frequency data for Brazil.

9.5. Special populations

Young children do not perform reliable functional tests, and assessment may require sedation or adapted examinations (section 5.2) [13]. In consanguineous families, coexistence of another recessive disease is possible, such as primary ciliary dyskinesia in a girl with HJMD, and atypical findings should prompt a search for a second diagnosis [64]. There are no data on pregnancy, older people or other special populations.

9.6. Reference networks and patient organisations

In the European Union, patients can be referred to ERN-EYE for the retinal disease, to ERN-Skin for the hypotrichosis and ectodermal dysplasias and to ERN ITHACA for the congenital malformations of EEM syndrome [18–20]. There is an information website created by a German family (hjmd.de), which played a role in the diagnosis of at least one family, but whose opinions are not peer-reviewed [10, 82]. In the United Kingdom, patient organisations for retinal dystrophies include the Royal National Institute of Blind People (RNIB), the Macular Society and Retinitis Pigmentosa Fighting Blindness [72]. The sources consulted did not confirm the addresses of Portuguese or Brazilian patient organisations for retinal dystrophies or ectodermal dysplasias.

10. Illustrative Clinical Case

Blanco-Kelly and colleagues published in 2017 the first Spanish case of HJMD, which shows how the separation between dermatology and ophthalmology delayed the diagnosis for more than a decade [9].

The patient, a boy born in 1998 in north-eastern Spain, the son of healthy non-consanguineous parents and with a healthy 9-year-old brother, was first seen in dermatology in 2000 for lack of hair growth since the first year of life. The dermatology report described pale skin, sparse scalp hair, punctate follicular keratosis on the face, forehead, eyebrows, arms and legs, erythema with follicular hyperkeratosis and reduced eyebrow density. A skin biopsy showed a reduced number of follicles, absence of sebaceous glands and an infundibular cyst, and led to the suspicion of keratosis follicularis spinulosa decalvans (Siemens syndrome). At 2 years, a pedunculated colorectal polyp with gastrointestinal nodular lymphoid hyperplasia was excised, without an established relationship to the disease [9].

The retinal change was noticed at 2 years. At 4–5 years the first complaints appeared, with reduced VA and central visual field (VA of 0.08 in the right eye and 0.1 in the left at 5 years, bilateral central scotomas) and macular spicules, and the diagnosis oscillated between maculopathy and «inverse retinitis pigmentosa». At 10 years, the VEP were abnormal and the ERG was asymmetric: normal under scotopic conditions and with a cone response at the lower limit in the right eye, and with a moderate decrease in all responses in the left. At 12 years, VA was 0.15 in both eyes, with loss of the central 20° in the right eye and of the central 10° in the left, a preserved peripheral field and macular atrophy with pigment accumulation [9].

In 2011, at 13 years, he was referred for genetic testing with the suspicion of two independent diseases, ABCA4 maculopathy and Siemens syndrome. The ABCA4 study (genotyping by arrayed primer extension, Sanger sequencing of all exons and multiplex ligation-dependent probe amplification, MLPA) found only one variant in the heterozygous state, c.6148G>C p.(Val2050Leu) (NM_000350), without a second allele. In 2012, at 14 years, the clinical genetics clinic reassessed the patient and unified the two presentations into a single diagnosis, HJMD. Examination showed pale and dry skin, hypotrichosis of the scalp, face (no beard or moustache at 16 years) and arms, blond and fragile hair with a pili torti appearance, normal eyebrows and eyelashes (reduced at a younger age), facial pigmented lesions suggestive of keratoses and centrofacial lentiginosis, probably sun-related, and no dental, hand or foot changes [9]. Figure 1 of the article shows these findings and the pedigree (https://pmc.ncbi.nlm.nih.gov/articles/PMC5219735/figure/Fig1/).

In 2013, at 15 years, Sanger sequencing of CDH3 identified two variants in compound heterozygosity, confirmed by segregation: c.830del p.(Gly277AlafsTer20), inherited from the father and already described, and c.613G>A p.(Val205Met), inherited from the mother, novel, absent from the frequency databases then available, classified as deleterious by most prediction programs and located at a calcium-binding site in the structural model of the protein [9]. The 9-year-old brother was not tested, because most international guidelines do not recommend carrier testing in children, and the authors proposed carrier testing for ABCA4 and CDH3 when he reached adulthood [9].

The case illustrates the practical points of this article. The congenital hypotrichosis was attributed to a different genodermatosis and the maculopathy to ABCA4, because the two signs were assessed separately; a single ABCA4 allele, frequent in the population, did not explain the disease; and the diagnosis resulted from joint assessment of the signs by a geneticist [9, 11]. The management described (carrier testing deferred to adulthood, counselling of the family) is in line with the AAO recommendations on testing of minors [13, 83]. An IRD panel that includes CDH3 and the genes for syndromic forms, as recommended by the 2022 AAO guidelines, allows the diagnosis to be reached even when the hypotrichosis is not given due weight [13].

11. Prognosis and Long-Term Follow-Up

11.1. Visual prognosis

The retinal disease is progressive but centrally limited. In Hull's series, eight patients had more than one assessment, with review intervals of 1 to 8 years: five had deterioration of VA, one maintained stable VA and two already had severe loss at presentation [11]. The area of atrophy did not increase in the two patients in whom FAF allowed it to be measured, but retinal thickness in the atrophic area decreased in all five patients measured over 2 to 10 years; in one patient, the PERG deteriorated over nine years and the full-field ERG remained stable [11]. The authors conclude that the disease is centrally progressive, with preserved peripheral function, and that the probable prognosis is preservation of peripheral vision [11]. Leibu and colleagues recorded a slow decline in VA and a gradual and consistent decrease in the ERG over follow-up of up to 14 years [3].

Variability is great. A British patient homozygous for c.830del maintained VA of 0.18 logMAR (20/30) at 31 years, after 20 years of follow-up and with preserved foveal photoreceptors on OCT, whereas the oldest patients in the same series had 1.0 logMAR or worse in the better eye [11]. Two Italian children maintained stable VA for 4 and 15 years, with progression detected only by microperimetry in one of them [40]. A Japanese patient went from decimal VA of 0.8 in both eyes at 22 years to 0.2 and 0.7 at 26, with worsening scotomas and foveal thinning, and her brother had a fall in right eye VA between 30 and 34 years without apparent progression of the atrophy [44]. In EEM syndrome, VA declined gradually over ten years in the Brazilian family and the retinal degeneration progressed minimally over 11 years in a Japanese patient [22, 23]. Reviews place legal blindness between the first and third decades, between the second and third, in the first 30 years of life or between the second and fourth decades, depending on the author [26, 29, 40, 49]. These estimates are reviews of selected cases and not survival curves, and the series with longitudinal data shows patients with useful vision in the fourth decade [11].

As regards the periphery, most patients retain the peripheral field, but there is progression to the periphery in some, with peripheral RPE atrophy and pigmentary changes, and cone-rod phenotypes with a progressively reduced ERG [3, 10, 50]. Macular neovascularisation is a possible complication that accelerates visual loss [45].

11.2. General prognosis

Life expectancy is not described as reduced, and cognitive development is normal in the patients assessed [12, 32]. The hypotrichosis is permanent, with partial regrowth at puberty in some patients [2, 14]. The functional prognosis of the hands in EEM syndrome depends on the severity of the malformation and has not been described in series [12].

11.3. Surveillance

There is no surveillance protocol specific to CDH3 disorders. Table 11.1 combines the generic AAO guidelines for IRD with the protocols described in the paediatric series and gives the source of each interval.

The AAO frequency of 1–2 years applies to the classes of chorioretinal degenerations and macular dystrophies; in the class of macular dystrophies, the AAO provides for the full-field ERG at the initial visit, but not at follow-up, and the mfERG at both [13]. The Italian and British paediatric series use annual follow-up [10, 40].

12. Summary and Final Considerations

Biallelic loss-of-function variants of CDH3, which encodes P-cadherin, cause an autosomal recessive spectrum whose poles have their own names: congenital hypotrichosis with juvenile macular dystrophy (HJMD) and EEM syndrome, which adds ectrodactyly, syndactyly and dental changes. The same variant can produce both pictures, and the limb anomalies may be minimal, so the distinction is clinical and not genetic [11, 12].

P-cadherin is the dominant cadherin of the RPE, the only cadherin of the inner matrix of the hair follicle and a transcriptional target of p63 in the apical ectodermal ridge of the limb, which explains the triad of hair, retina and limbs and the overlap with the TP63 syndromes [28, 35, 41]. The retinal disease is a chorioretinal atrophy of the posterior pole, contiguous with the disc, beginning in childhood, with progressive and variable loss of central VA and usually preserved peripheral function [11].

The diagnosis depends on bringing together two signs that are seen by different specialties. Every child with congenital hypotrichosis should undergo fundoscopy, and every child with macular dystrophy should have the scalp examined [11, 55]. Confirmation is by an IRD panel that includes CDH3 and CNV analysis, because there are exon deletions that conventional exome sequencing does not detect, and homozygosity without study of the parents may conceal a deletion [13–15].

There is no targeted treatment, authorised medicine or registered clinical trial; subretinal gene therapy is a plausible hypothesis given the size of the gene and the location of the disease, without experimental data [17, 38, 40]. Management rests on annual ophthalmological follow-up in children and every 1–2 years in adults, low vision rehabilitation, psychological and school support, hand surgery and oral rehabilitation in EEM syndrome, and genetic counselling, which should include the risk of limb anomalies in future pregnancies of families with HJMD [12, 13, 40]. The Portuguese cases described and the IRD-PT registry show that the disease exists in Portugal, where the c.830del variant recurs, and one of the two families in which the genetic cause of EEM syndrome was identified was Brazilian [23, 52, 71].

13. References

1. Orphanet. Hypotrichosis with juvenile macular degeneration (ORPHA:1573) [Internet]. Paris: INSERM; [accessed 2026-10-29]. Available from: https://www.orpha.net/en/disease/detail/1573
2. Online Mendelian Inheritance in Man, OMIM®. #601553 Hypotrichosis, congenital, with juvenile macular dystrophy; HJMD [Internet]. Baltimore (MD): Johns Hopkins University; [last edited 2024-10-01; accessed 2026-10-29]. Available from: https://omim.org/entry/601553
3. Leibu R, Jermans A, Hatim G, Miller B, Sprecher E, Perlman I. Hypotrichosis with juvenile macular dystrophy: clinical and electrophysiological assessment of visual function. Ophthalmology. 2006;113(5):841-7.e3. doi:10.1016/j.ophtha.2005.10.065. PMID: 16650681.
4. Orphanet. EEM syndrome (ORPHA:1897) [Internet]. Paris: INSERM; [accessed 2026-10-29]. Available from: https://www.orpha.net/en/disease/detail/1897
5. Online Mendelian Inheritance in Man, OMIM®. #225280 Ectodermal dysplasia, ectrodactyly, and macular dystrophy syndrome; EEMS [Internet]. Baltimore (MD): Johns Hopkins University; [last edited 2024-10-04; accessed 2026-10-29]. Available from: https://omim.org/entry/225280
6. Online Mendelian Inheritance in Man, OMIM®. *114021 Cadherin 3; CDH3 [Internet]. Baltimore (MD): Johns Hopkins University; [last edited 2024-10-01; accessed 2026-10-29]. Available from: https://omim.org/entry/114021
7. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-8. doi:10.1002/humu.23348. PMID: 28967166. Online tool: https://variantvalidator.org (accessed 2026-10-29).
8. Wallace SE, Bean LJH. Resources for Genetics Professionals — Genetic Disorders Associated with Founder Variants Common in the Druze Population. 2019 Nov 7 [updated 2023 Sep 28]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK549466/
9. Blanco-Kelly F, Rodrigues-Jacy da Silva L, Sanchez-Navarro I, Riveiro-Alvarez R, Lopez-Martinez MA, Corton M, et al. New CDH3 mutation in the first Spanish case of hypotrichosis with juvenile macular dystrophy, a case report. BMC Med Genet. 2017;18(1):1. doi:10.1186/s12881-016-0364-5. PMID: 28061825. PMC5219735.
10. Narayan A, Moosajee M. Sparse scalp hair and vision loss: think hypotrichosis with juvenile macular dystrophy. BMJ Case Rep. 2019;12(10):e232342. doi:10.1136/bcr-2019-232342. PMID: 31645385. PMC6827737.
11. Hull S, Arno G, Robson AG, Broadgate S, Plagnol V, McKibbin M, et al. Characterization of CDH3-Related Congenital Hypotrichosis With Juvenile Macular Dystrophy. JAMA Ophthalmol. 2016;134(9):992-1000. doi:10.1001/jamaophthalmol.2016.2089. PMID: 27386845.
12. Basel-Vanagaite L, Pasmanik-Chor M, Lurie R, Yeheskel A, Kjaer KW. CDH3-Related Syndromes: Report on a New Mutation and Overview of the Genotype-Phenotype Correlations. Mol Syndromol. 2010;1(5):223-230. doi:10.1159/000327156. PMID: 22140374. PMC3214945.
13. Duncan JL, Branham K, Birch DG, Daiger SP, Ferris FL 3rd, Fishman GA, et al. Guidelines on Clinical Assessment of Patients with Inherited Retinal Degenerations – 2022. Clinical Statement. San Francisco (CA): American Academy of Ophthalmology; 2022 Oct [accessed 2026-10-29]. Available from: https://www.aao.org/education/clinical-statement/guidelines-on-clinical-assessment-of-patients-with
14. Saeidian AH, Vahidnezhad H, Youssefian L, Sotudeh S, Sargazi M, Zeinali S, et al. Hypotrichosis with juvenile macular dystrophy: Combination of whole-genome sequencing and genome-wide homozygosity mapping identifies a large deletion in CDH3 initially undetected by whole-exome sequencing-A lesson from next-generation sequencing. Mol Genet Genomic Med. 2019;7(11):e975. doi:10.1002/mgg3.975. PMID: 31560841. PMC6825862.
15. Vaidyanthan A, Morales PC, Estrada-Puente C, Maldonado RS. First CDH3-related HJMD case in a black patient: novel variants and diagnostic odyssey. Ophthalmic Genet. 2026:1-6. doi:10.1080/13816810.2026.2716313. PMID: 42675022.
16. Peschel N, Wright JT, Koster MI, Clarke AJ, Tadini G, Fete M, et al. Molecular Pathway-Based Classification of Ectodermal Dysplasias: First Five-Yearly Update. Genes (Basel). 2022;13(12):2327. doi:10.3390/genes13122327. PMID: 36553593. PMC9778228.
17. ClinicalTrials.gov [Internet]. Bethesda (MD): National Library of Medicine. Searches «CDH3», «hypotrichosis juvenile macular», «P-cadherin», «EEM syndrome» [accessed 2026-10-29]. Available from: https://clinicaltrials.gov
18. European Reference Network for Rare Eye Diseases (ERN-EYE) [Internet]. [accessed 2026-10-29]. Available from: https://www.ern-eye.eu
19. European Reference Network for Rare and Complex Diseases of the Skin (ERN-Skin) [Internet]. [accessed 2026-10-29]. Available from: https://ern-skin.eu
20. European Reference Network on Rare Congenital Malformations and Rare Intellectual Disability (ERN ITHACA) [Internet]. [accessed 2026-10-29]. Available from: https://ern-ithaca.eu
21. Ohdo S, Hirayama K, Terawaki T. Association of ectodermal dysplasia, ectrodactyly, and macular dystrophy: the EEM syndrome. J Med Genet. 1983;20(1):52-57. doi:10.1136/jmg.20.1.52. PMID: 6302256. PMC1048986.
22. Hayakawa M, Yanashima K, Kato K, Nakajima A, Yamauchi H. Association of ectodermal dysplasia, ectrodactyly and macular dystrophy: EEM syndrome (case report). Ophthalmic Paediatr Genet. 1989;10(4):287-292. doi:10.3109/13816818909009884. PMID: 2628819.
23. Balarin Silva V, Simões AM, Marques-de-Faria AP. EEM syndrome: report of a family and results of a ten-year follow-up. Ophthalmic Genet. 1999;20(2):95-99. doi:10.1076/opge.20.2.95.2290. PMID: 10420194.
24. Senecky Y, Halpern GJ, Inbar D, Attias J, Shohat M. Ectodermal dysplasia, ectrodactyly and macular dystrophy (EEM syndrome) in siblings. Am J Med Genet. 2001;101(3):195-197. doi:10.1002/ajmg.1361. PMID: 11424132.
25. Sprecher E, Bergman R, Richard G, Lurie R, Shalev S, Petronius D, et al. Hypotrichosis with juvenile macular dystrophy is caused by a mutation in CDH3, encoding P-cadherin. Nat Genet. 2001;29(2):134-136. doi:10.1038/ng716. PMID: 11544476.
26. Indelman M, Bergman R, Lurie R, Richard G, Miller B, Petronius D, et al. A missense mutation in CDH3, encoding P-cadherin, causes hypotrichosis with juvenile macular dystrophy. J Invest Dermatol. 2002;119(5):1210-1213. doi:10.1046/j.1523-1747.2002.19528.x. PMID: 12445216.
27. Kjaer KW, Hansen L, Schwabe GC, Marques-de-Faria AP, Eiberg H, Mundlos S, et al. Distinct CDH3 mutations cause ectodermal dysplasia, ectrodactyly, macular dystrophy (EEM syndrome). J Med Genet. 2005;42(4):292-298. doi:10.1136/jmg.2004.027821. PMID: 15805154. PMC1736041.
28. Shimomura Y, Wajid M, Shapiro L, Christiano AM. P-cadherin is a p63 target gene with a crucial role in the developing human limb bud and hair follicle. Development. 2008;135(4):743-753. doi:10.1242/dev.006718. PMID: 18199584.
29. Karti O, Abali S, Ayhan Z, Gokmeydan E, Nalcaci S, Yaman A, et al. CDH3 gene related hypotrichosis and juvenile macular dystrophy - A case with a novel mutation. Am J Ophthalmol Case Rep. 2017;7:129-133. doi:10.1016/j.ajoc.2017.06.007. PMID: 29260097. PMC5722150.
30. Wright JT, Fete M, Schneider H, Zinser M, Koster MI, Clarke AJ, et al. Ectodermal dysplasias: Classification and organization by phenotype, genotype and molecular pathway. Am J Med Genet A. 2019;179(3):442-447. doi:10.1002/ajmg.a.61045. PMID: 30703280. PMC6421567.
31. So N, Yip L, Orchard D. Paediatric Hypotrichosis: A Clinical and Algorithmic Approach to Diagnosis. Australas J Dermatol. 2025;66(3):e109-e119. doi:10.1111/ajd.14429. PMID: 39992008. PMC12062735.
32. Prieto LE, Vasquez-Forero DM, Pachajoa H. The First Colombian Patient with CDH3-Related Hypotrichosis with Juvenile Macular Dystrophy. Skin Appendage Disord. 2025:1-7. doi:10.1159/000545293. PMID: 40330852. PMC12052343.
33. Yusuf IH, Garrett AM, MacLaren RE, Charbel Issa P. Retinal cadherins and the retinal cadherinopathies: Current concepts and future directions. Prog Retin Eye Res. 2022;90:101038. doi:10.1016/j.preteyeres.2021.101038. PMID: 35066146.
34. Basit S, Wali A, Aziz A, Muhammad N, Jelani M, Ahmad W. Digenic inheritance of an autosomal recessive hypotrichosis in two consanguineous pedigrees. Clin Genet. 2011;79(3):273-281. doi:10.1111/j.1399-0004.2010.01455.x. PMID: 20528890.
35. Kadhi A, Hamie L, Tamer C, Nemer G, Kurban M. A Novel Pathogenic CDH3 Variant underlying Heredity Hypotrichosis Simplex detected by Whole-Exome Sequencing (WES)-A Case Report. Cold Spring Harb Mol Case Stud. 2022:mcs.a006225. doi:10.1101/mcs.a006225. PMID: 35962736. PMC9528967.
36. Beryozkin A, Shevah E, Kimchi A, Mizrahi-Meissonnier L, Khateb S, Ratnapriya R, et al. Whole Exome Sequencing Reveals Mutations in Known Retinal Disease Genes in 33 out of 68 Israeli Families with Inherited Retinopathies. Sci Rep. 2015;5:13187. doi:10.1038/srep13187. PMID: 26306921. PMC4549705.
37. Vincke L, Van Schil K, Ahmadieh H, Moghaddasi A, Sabbaghi H, Daftarian N, et al. Uncovering the genetic architecture of inherited retinal disease in a consanguineous Iranian cohort. NPJ Genom Med. 2025;10(1):19. doi:10.1038/s41525-025-00473-9. PMID: 40055385. PMC11889135.
38. Singh MS, Broadgate S, Mathur R, Holt R, Halford S, MacLaren RE. Hypotrichosis and juvenile macular dystrophy caused by CDH3 mutation: A candidate disease for retinal gene therapy. Sci Rep. 2016;6:23674. doi:10.1038/srep23674. PMID: 27157923. PMC4860587.
39. Oliveira C, Senz J, Kaurah P, Pinheiro H, Sanges R, Haegert A, et al. Germline CDH1 deletions in hereditary diffuse gastric cancer families. Hum Mol Genet. 2009;18(9):1545-1555. doi:10.1093/hmg/ddp046. PMID: 19168852. PMC2667284.
40. Marziali E, Elia C, Bargiacchi S, Mancano G, Artuso R, Dirupo E, et al. CDH3 Retinopathy: Long-Term Multimodal Follow-Up with Pediatric Multidisciplinary Insights. J Clin Med. 2026;15(14):5393. doi:10.3390/jcm15145393. PMID: 42513307. PMC13412677.
41. Yang X, Chung JY, Rai U, Esumi N. Cadherins in the retinal pigment epithelium (RPE) revisited: P-cadherin is the highly dominant cadherin expressed in human and mouse RPE in vivo. PLoS One. 2018;13(1):e0191279. doi:10.1371/journal.pone.0191279. PMID: 29338041. PMC5770047.
42. Samuelov L, Sprecher E, Tsuruta D, Bíró T, Kloepper JE, Paus R. P-cadherin regulates human hair growth and cycling via canonical Wnt signaling and transforming growth factor-β2. J Invest Dermatol. 2012;132(10):2332-2341. doi:10.1038/jid.2012.171. PMID: 22696062.
43. Bergman R, Sapir M, Sprecher E. Histopathology of hypotrichosis with juvenile macular dystrophy. Am J Dermatopathol. 2004;26(3):205-209. doi:10.1097/00000372-200406000-00006. PMID: 15166507.
44. Hayashi T, Katagiri S, Kubota D, Mizobuchi K, Ishiuji Y, Asahina A, et al. The first Japanese family of CDH3-related hypotrichosis with juvenile macular dystrophy. Mol Genet Genomic Med. 2021;9(6):e1688. doi:10.1002/mgg3.1688. PMID: 33837674. PMC8222849.
45. Carnovale-Scalzo G, Carnevali A, Piccoli G, Ceravolo D, Bruzzichessi D, Iuliano R, et al. Multimodal imaging of Hypotrichosis with juvenile macular dystrophy: a case report. BMC Ophthalmol. 2021;21(1):284. doi:10.1186/s12886-021-02037-8. PMID: 34301208. PMC8299653.
46. Lin Y, Yu P, Cui S, Xiong S, Zhou Y, Chen Y, et al. Novel Compound Heterozygous Variants in CDH3 Cause Congenital Hypotrichosis with Juvenile Macular Dystrophy: A Case Report with Longitudinal Imaging and Functional Validation. Case Rep Ophthalmol. 2026;17(1):433-440. doi:10.1159/000551989. PMID: 42199197. PMC13200835.
47. Indelman M, Eason J, Hummel M, Loza O, Suri M, Leys MJ, et al. Novel CDH3 mutations in hypotrichosis with juvenile macular dystrophy. Clin Exp Dermatol. 2007;32(2):191-196. doi:10.1111/j.1365-2230.2006.02335.x. PMID: 17342797.
48. National Center for Biotechnology Information. ClinVar — CDH3[gene] [Internet]. Bethesda (MD): NCBI; [accessed 2026-10-29]. Available from: https://www.ncbi.nlm.nih.gov/clinvar/?term=CDH3%5Bgene%5D
49. Indelman M, Hamel CP, Bergman R, Nischal KK, Thompson D, Surget MO, et al. Phenotypic diversity and mutation spectrum in hypotrichosis with juvenile macular dystrophy. J Invest Dermatol. 2003;121(5):1217-1220. doi:10.1046/j.1523-1747.2003.12550_1.x. PMID: 14708629.
50. Nasser F, Mulahasanovic L, Alkhateeb M, Biskup S, Stingl K, Zrenner E. Hypotrichosis with cone-rod dystrophy in a patient with cadherin 3 (CDH3) mutation. Doc Ophthalmol. 2019;138(2):153-160. doi:10.1007/s10633-019-09675-w. PMID: 30710256.
51. Nasser F, Kempf M, Kurtenbach A, Stöhr H, Weber BHF, Neuhaus C, et al. Correlating Adaptive Optics Images to Clinical Findings in Juvenile Macular Dystrophy with Hypotrichosis in Siblings with Homozygous CDH3 Pathogenic Variation. Ophthalmic Res. 2020;63(2):141-151. doi:10.1159/000504757. PMID: 31927556.
52. Almeida FT, Carneiro-Freitas R, Caldas R, Vieira AP. Hypotrichosis with Juvenile Macular Dystrophy. Int J Trichology. 2018;10(5):234-236. doi:10.4103/ijt.ijt_60_18. PMID: 30607044. PMC6290288.
53. Al Zubi K, Mwafi N, Alrawashdeh HM, Al Sarireh F, Somkuwar A, Abdulmannan DM. The first reported case of CDH3-related hypotrichosis with juvenile macular dystrophy from Jordan: a case report. Ophthalmic Genet. 2022;43(3):420-424. doi:10.1080/13816810.2022.2028298. PMID: 35038959.
54. Schauren JS, Torres ACMBG, de Almeida RC, Santos PSC, Mulinari-Brenner F, Lima LH, et al. Novel CDH3 variants in Brazilian families with hypotrichosis and juvenile macular dystrophy revealed by exome sequencing. Clin Genet. 2020;97(3):529-531. doi:10.1111/cge.13659. PMID: 31696509.
55. Indelman M, Leibu R, Jammal A, Bergman R, Sprecher E. Molecular basis of hypotrichosis with juvenile macular dystrophy in two siblings. Br J Dermatol. 2005;153(3):635-638. doi:10.1111/j.1365-2133.2005.06734.x. PMID: 16120155.
56. Jelani M, Salman Chishti M, Ahmad W. A novel splice-site mutation in the CDH3 gene in hypotrichosis with juvenile macular dystrophy. Clin Exp Dermatol. 2009;34(1):68-73. doi:10.1111/j.1365-2230.2008.02933.x. PMID: 19076794.
57. Kamran-ul-Hassan Naqvi S, Azeem Z, Ali G, Ahmad W. A novel splice-acceptor site mutation in CDH3 gene in a consanguineous family exhibiting hypotrichosis with juvenile macular dystrophy. Arch Dermatol Res. 2010;302(9):701-703. doi:10.1007/s00403-010-1035-6. PMID: 20140736.
58. Shimomura Y, Wajid M, Kurban M, Christiano AM. Splice site mutations in the P-cadherin gene underlie hypotrichosis with juvenile macular dystrophy. Dermatology. 2010;220(3):208-212. doi:10.1159/000275673. PMID: 20203473. PMC2865484.
59. Avitan-Hersh E, Indelman M, Khamaysi Z, Leibu R, Bergman R. A novel nonsense CDH3 mutation in hypotrichosis with juvenile macular dystrophy. Int J Dermatol. 2012;51(3):325-327. doi:10.1111/j.1365-4632.2011.04973.x. PMID: 22348569.
60. Khan AO, Bolz HJ. Phenotypic observations in "hypotrichosis with juvenile macular dystrophy" (recessive CDH3 mutations). Ophthalmic Genet. 2016;37(3):301-306. doi:10.3109/13816810.2015.1071411. PMID: 26885695.
61. Ahmed A, Alali A, Alsharif O, Kaki A. Hypotrichosis with Juvenile Macular Dystrophy in Saudi Arabia: A Case Report. Skin Appendage Disord. 2021;7(1):75-79. doi:10.1159/000511741. PMID: 33614726. PMC7879292.
62. Vicente LP, Finzi S, Susanna R, Young TL. Hypotrichosis with juvenile macular dystrophy: a case report with molecular study. Arq Bras Oftalmol. 2017;80(1):49-51. doi:10.5935/0004-2749.20170013. PMID: 28380103.
63. Danish E, Alsulami R, Baeshen H. CDH3 Mutation in Saudi Arabia: A Case of Hypotrichosis With Juvenile Macular Dystrophy. Cureus. 2026;18(4):e106554. doi:10.7759/cureus.106554. PMID: 42109964. PMC13152264.
64. Fan KC, Patel NA, Yannuzzi NA, Prakhunhungsit S, Negron CI, Basora E, et al. A unique case of vision loss in a patient with hypotrichosis and juvenile macular dystrophy and primary ciliary dyskinesia. Am J Ophthalmol Case Rep. 2019;15:100486. doi:10.1016/j.ajoc.2019.100486. PMID: 31431935. PMC6579934.
65. Ahmad F, Ali RH, Muhammad D, Nasir A, Umair M, Wakil SM, et al. Novel homozygous sequence variants in the CDH3 gene encoding P-cadherin underlying hypotrichosis with juvenile macular dystrophy in consanguineous families. Eur J Dermatol. 2016;26(6):610-612. doi:10.1684/ejd.2016.2848. PMID: 27456782.
66. Halford S, Holt R, Németh AH, Downes SM. Homozygous deletion in CDH3 and hypotrichosis with juvenile macular dystrophy. Arch Ophthalmol. 2012;130(11):1490-1492. doi:10.1001/archophthalmol.2012.708. PMID: 23143461.
67. Yildirim MS, Ogun TC, Kamiş U. Ectrodactyly, ectodermal dysplasia, macular degeneration syndrome: a further contribution. Genet Couns. 2006;17(2):149-153. PMID: 16970031.
68. Matos AG, Gurgel VP, Yugar PJ, Yugar AS. Ophthalmological findings in ectodermal dysplasia, ectrodactyly, and macular dystrophy syndrome: a case report. Arq Bras Oftalmol. 2018;81(5):440-442. doi:10.5935/0004-2749.20180085. PMID: 30208149.
69. Elfatoiki FZ, Cordoliani F, Pascal Regane P, Afforitit-Demoge A. Hypotrichosis with juvenile macular dystrophy: Portuguese case. Dermatol Online J. 2016;22(5):13030/qt0w28m8h4. doi:10.5070/d3225030955. PMID: 27617529.
70. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 330–2 (Unusual hair, teeth, nails, and skin).
71. Marta A, Marques-Couto P, Vaz-Pereira S, Costa J, Cabral D, Estrela-Silva S, et al. Clinical and genetic landscape of IRD in Portugal: pooled data from the nationwide IRD-PT registry. NPJ Genom Med. 2025;10(1):11. doi:10.1038/s41525-025-00475-7. PMID: 39939324. PMC11821859.
72. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 302–5 (Retinal receptor dystrophies).
73. Online Mendelian Inheritance in Man, OMIM®. #604292 Ectrodactyly, ectodermal dysplasia, and cleft lip/palate syndrome 3; EEC3 [Internet]. Baltimore (MD): Johns Hopkins University; [last edited 2026-08-03; accessed 2026-10-29]. Available from: https://omim.org/entry/604292
74. Online Mendelian Inheritance in Man, OMIM®. #103285 ADULT syndrome [Internet]. Baltimore (MD): Johns Hopkins University; [last edited 2020-01-22; accessed 2026-10-29]. Available from: https://omim.org/entry/103285
75. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 422–3 (Ectrodactyly–ectodermal dysplasia–clefting syndrome).
76. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 869–71 (25.3 Ectrodactyly, ectodermal dysplasia, and cleft lip/palate syndrome).
77. Online Mendelian Inheritance in Man, OMIM®. %129900 Ectrodactyly, ectodermal dysplasia, and cleft lip/palate syndrome 1; EEC1 [Internet]. Baltimore (MD): Johns Hopkins University; [last edited 2026-08-10; accessed 2026-10-29]. Available from: https://omim.org/entry/129900
78. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 212–5 (Limb reduction defects).
79. Online Mendelian Inheritance in Man, OMIM®. #248200 Stargardt disease 1; STGD1 [Internet]. Baltimore (MD): Johns Hopkins University; [last edited 2025-07-02; accessed 2026-10-29]. Available from: https://omim.org/entry/248200
80. Online Mendelian Inheritance in Man, OMIM®. #270200 Sjogren-Larsson syndrome; SLS [Internet]. Baltimore (MD): Johns Hopkins University; [last edited 2023-12-04; accessed 2026-10-29]. Available from: https://omim.org/entry/270200
81. Richard G. Autosomal Recessive Congenital Ichthyosis. 2001 Jan 10 [updated 2023 Apr 20]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1420/
82. Hypotrichose mit juveniler Makuladystrophie – HJMD / CDH3 [Internet]. Patient family website (Germany); [accessed 2026-10-29]. Available from: https://hjmd.de/
83. Stone EM, Aldave AJ, Drack AV, Maccumber MW, Sheffield VC, Traboulsi E, et al. Recommendations for genetic testing of inherited eye diseases: report of the American Academy of Ophthalmology task force on genetic testing. Ophthalmology. 2012;119(11):2408-2410. doi:10.1016/j.ophtha.2012.05.047. PMID: 22944025.
84. Leal-Rodríguez R, Barragán-Arévalo T, Pérez-Torres A, Giraldo-Gómez DM, Zenteno JC. Clinical, genetic, and electron microscopy of hair findings in a patient with CDH3 -related hypotrichosis with juvenile macular dystrophy. Clin Dysmorphol. 2023;32(2):62-64. doi:10.1097/mcd.0000000000000449. PMID: 36779776.
85. Liu Y, Xu Z, Sun Y, Zhang B, Wang X, Ma L. Hypotrichosis with juvenile macular dystrophy with novel mutations in CDH3 gene. Australas J Dermatol. 2024;65(1):91-94. doi:10.1111/ajd.14170. PMID: 37850495.
86. Oliveira-Ferreira C, Leuzinger-Dias M, Tavares-Ferreira J, Silva SE, Brandão E, Falcão-Reis F, et al. Hypotrichosis with juvenile macular dystrophy. Ophthalmic Genet. 2019;40(6):574-577. doi:10.1080/13816810.2019.1688840. PMID: 31718366.