Frontonasal Dysplasias (ALX1, ALX3, ALX4) and the Frontonasal Dysplasia Spectrum

Syndromic class of frontonasal dysplasia (median cleft face sequence): recessive forms related to the ALX genes — frontorhiny (FND1, ALX3), frontonasal dysplasia with alopecia and genital anomalies (FND2, ALX4) and frontonasal dysplasia with extreme microphthalmia and facial clefting (FND3, ALX1) —, acromelic frontonasal dysostosis (ZSWIM6), a dominant form due to SIX2 deletion, syndromic forms without a known gene and allelic disorders of ALX4 (enlarged parietal foramina, Potocki-Shaffer syndrome); craniofrontonasal syndrome (EFNB1) is part of the differential diagnosis.

Literature and regulatory information updated: 29 October 2026.

0. Initial Reference Table

Frontonasal dysplasia (FND) is not a single disease but a malformation pattern of the upper and middle thirds of the face whose forms are distinguished by gene, inheritance and extrafacial findings. The first table identifies each entity of the class and the second brings together the cross-cutting data [1–6].

1. Introduction

1.1. From “median cleft face” to frontonasal dysplasia

The modern description of the malformation begins in 1967, when DeMyer brought together 33 patients with successive degrees of the same defect of midfacial development — a midline frontal bone defect (cranium bifidum occultum), hypertelorism and a median cleft of the nose, lip and palate — and called it median cleft face syndrome (median cleft face syndrome) [26]. In 1970, Sedano and colleagues proposed the designation frontonasal dysplasia, which prevailed, and in 1988 Sedano and Gorlin interpreted it as a developmental field defect, isolated or part of syndromic associations, and established the criteria still in use: true ocular hypertelorism, broadening of the nasal root, median facial cleft of the nose and/or upper lip and palate, unilateral or bilateral clefting of the nasal alae, lack of formation of the nasal tip, anterior cranium bifidum occultum and a V-shaped frontal hairline or widow's peak; the presence of two or more of these signs defines FND [2, 3, 26]. Smith's Recognizable Patterns of Human Malformation retains the entity as a sequence — a primary midfacial defect of unknown cause that prevents the anterior apposition of the eyes and triggers the remaining anomalies — and lists nine multiple malformation syndromes in which FND is one of the components [26].

1.2. Milestones of molecular discovery

For four decades, almost all cases were sporadic, and the only consistent genetic cause was the EFNB1 gene in X-linked craniofrontonasal syndrome [2, 27]. In 2000 and 2001, Andrew Wilkie's group in Oxford showed that haploinsufficiency of two homeobox genes, MSX2 and ALX4, caused enlarged parietal foramina, symmetrical defects of parietal ossification; ALX4 was identified as the gene responsible for the foramina in the proximal 11p deletion syndrome (Potocki-Shaffer), in which loss of EXT2 explains the osteochondromas [28, 29]. In 2009, two independent studies linked FND to the “aristaless-like” genes: Twigg and colleagues described 11 patients from 7 families with a recessive FND with a distinctive facies, which they called frontorhiny, caused by homozygous variants in ALX3 — the first isolated FND with a defined genetic aetiology [27]; and Kayserili and colleagues identified, in two consanguineous Turkish families, a recessive frontofacial dysostosis with total alopecia, a large skull defect, coronal craniosynostosis and hypogonadism, caused by the homozygous variant ALX4 c.793C>T p.(Arg265Ter) [30]. In 2010, Uz and colleagues completed the family with ALX1, whose biallelic loss causes extreme microphthalmia and severe facial clefts [31]. The OMIM nomenclature was consolidated as FND1 (ALX3), FND2 (ALX4) and FND3 (ALX1) [3–5].

Subsequent milestones broadened the class. In 2014, exome sequencing revealed that acromelic frontonasal dysostosis (AFND), the form with brain and limb malformations, results from a single recurrent de novo variant in ZSWIM6, c.3487C>T p.(Arg1163Trp) [32]; in 2016, Twigg and colleagues demonstrated low-level parental mosaicism and cases without limb anomalies [21]. In 2016, Hufnagel and colleagues described a mild dominant FND due to a heterozygous deletion of SIX2 [33], and Farlie and colleagues proposed, in a review, the provisional designation “FND4” for this form [2]. In 2017, Sweeney-Cox syndrome was attributed to de novo substitutions at residue Glu117 of TWIST1 [34]. From 2013 onwards, several families showed that heterozygous ALX4 variants can cause, in addition to parietal foramina, mild FND with vertical transmission, and in 2023 and 2026 dominant forms with hypotrichosis and ectodermal defects attributed to gain of function were described [35–39]. In acrofrontofacionasal dysostosis, described by Richieri-Costa and colleagues in 1985 in two Brazilian siblings, exome sequencing identified in 2018 a biallelic NBAS variant in a family of Indian origin and in 2021 a homozygous intronic PIGB variant in the original Brazilian families, which separated types 1 and 3 genetically and placed type 1 among the defects of glycosylphosphatidylinositol (GPI) anchor biosynthesis [14, 19, 40].

1.3. Definition of the entities in this article

The FND sequence designates the morphological phenotype, whatever the cause; most isolated cases remain without a molecular diagnosis, and the association with discordant monozygotic twinning and with insults to fetal development suggests non-genetic causes in many of them [2, 27, 41]. FND1 (frontorhiny, OMIM 136760, ORPHA:391474) is the recessive form due to ALX3, with a narrow spectrum and normal cognition [27]. FND2 (OMIM 613451, ORPHA:228390) is the recessive form due to ALX4, with extensive skull defects, alopecia and genital anomalies; in the heterozygous state the same gene causes enlarged parietal foramina type 2 (OMIM 609597) and mild dominant forms of FND [4, 9]. FND3 (OMIM 613456, ORPHA:306542) is the recessive form due to ALX1, the most severe of the ALX group [5, 31]. AFND (OMIM 603671, ORPHA:1827) is the dominant form due to ZSWIM6 with brain and limb malformations [32]. Craniofrontonasal syndrome (EFNB1, OMIM 304110) belongs to the FND spectrum but appears here only in the differential diagnosis. Orphanet classifies FND as a group (ORPHA:250) that includes frontorhiny, FND2, FND3 and AFND [1, 42–45]. Acrofrontofacionasal dysostosis (ORPHA:1784) is a recessive group in which FND is associated with limb anomalies and severe intellectual disability: type 1 results from PIGB variants (OMIM 201180), type 2 has no known gene (OMIM 239710, ORPHA:2211) and type 3 results from NBAS variants (OMIM 621629) [12, 13, 15, 16, 18]. Enlarged parietal foramina (ORPHA:60015) are included in this article as allelic disorders of ALX4 (type 2, OMIM 609597) and MSX2 (type 1, OMIM 168500), with a third locus at 4q21-q23 supported by limited evidence [8, 9].

2. Molecular Genetics and Pathophysiology

2.1. Embryological origin of the frontonasal process

The middle and upper thirds of the face are built from three blocks: the two maxillary processes, which form the upper jaw, the cheeks and the lateral nasal structures, and the frontonasal process, which gives rise to the frontal bones, the bridge and tip of the nose and joins the bilateral maxillary structures [2]. The mesenchyme of the frontonasal process derives from neural crest cells of the caudal forebrain and rostral midbrain, which migrate as a sheet over the future telencephalon and the eyes before neural tube closure; unlike the maxillary process, this mesenchyme is in direct contact with the forebrain neuroepithelium, which conditions its growth and patterning [2]. Between the 4th and 8th weeks of gestation, the frontonasal, medial and lateral nasal, maxillary and mandibular processes proliferate and fuse to form the orbital, nasal and oral structures; disruption of this sequence produces combinations of hypertelorism, altered nasal configuration and oral, palatal or facial clefts [27]. Coordinated expression of SHH (sonic hedgehog), BMP4 (bone morphogenetic protein 4) and FGF8 (fibroblast growth factor 8) in a frontonasal ectodermal zone regulates the growth of the primordia, and fusion of the processes precedes osteogenic differentiation, which allows overt clefts and hypoplasia of intact bony elements to result from temporally distinct events [2]. Excessive Hedgehog pathway activity, induced in mice by eliminating primary cilia in cranial neural crest cells, produces hypertelorism and FND, which supports the hypothesis that narrowing and excessive expansion of the facial midline are extremes of the same disturbance of this pathway [3].

2.2. The ALX family

The genes ALX1 (formerly CART1, 12q21.31), ALX3 (1p13.3) and ALX4 (11p11.2) encode transcription factors with a paired-type homeodomain and a conserved C-terminal domain (aristaless domain or “paired tail”), expressed in neural crest-derived mesenchyme and in lateral plate mesoderm, where they regulate cell determination and differentiation [26, 33, 46]. In mice, the three genes are expressed in the frontonasal mesenchyme in overlapping domains during neural crest migration [2]. Redundancy between paralogues explains the differences between species: the Alx3−/− mouse is normal and the Alx4−/− mouse has only delayed formation of the parietal bones, but the Alx3/Alx4 double mutant develops a complete median facial cleft owing to increased cell death in the frontonasal mesenchyme, whereas in humans biallelic loss of each individual gene is sufficient to cause FND [2, 27]. ALX1 encodes a protein of 326 amino acids [47]. The Cart1 (Alx1)-deficient mouse is born with acrania and meroanencephaly and dies shortly afterwards, with penetrance dependent on the genetic background (about 65% on a hybrid background and 100% on an isogenic 129/SvEv background) and suppression of the phenotype by prenatal folic acid, whereas humans with total loss of ALX1 have no neural tube defects [2, 31, 47]. In mice, Alx3 is strongly expressed in the frontonasal mesenchyme between embryonic days 9.5 and 11.5, and mutants for three of the four Alx3/Alx4 alleles have a facial cleft with increased apoptosis at day 10.0 [48]. A mouse model of Alx1 deletion generated by CRISPR/Cas9 reproduced the median orofacial cleft and the ocular and nasal alar anomalies, with apoptosis of the periocular mesenchyme and loss of frontonasal mesenchyme identity partly compensated by ALX4 [49]. In 2026, a study in zebrafish and lamprey proposed that the alx genes act as homeotic selector genes that confer segmental identity on frontonasal neural crest cells, whose mutations produce mirror-image transformations [50].

ALX4 has 4 coding exons and encodes a protein of 410 amino acids [51]. The protein binds DNA as a dimer at a palindromic TAAT-NNN-ATTA site; in 2025, the crystal structure of the dimer showed that binding is independent of the cofactor TWIST1 in cranial neural crest cells, that seven residues participate in dimerisation, that the two molecules use different residues in asymmetric interactions and that cooperativity is required for transcriptional activation; pathogenic homeodomain variants cause distinct molecular defects, including loss of cooperativity, reduced DNA contact, reduced nuclear localisation and reduced stability [9, 52]. The skull is particularly sensitive to ALX4 and MSX2 dosage; in the absence of ALX4, midline facial development, hair follicle growth and genital development are also affected [9].

2.3. ALX3 and frontorhiny (FND1)

Frontorhiny is caused by biallelic loss-of-function variants in ALX3. In the original series, the 7 homozygous variants identified in 7 families included missense substitutions at critical positions of the homeodomain — position 16 of helix I (p.Leu168Val), position 31 of helix II (p.Arg183Trp), positions 44 and 51 of helix III (p.Arg196Trp and p.Asn203Ser; asparagine residue 51 directly contacts an adenine in the DNA) — and nonsense, frameshift and splicing variants that eliminate the DNA-binding helix III [27]. None of the variants was present in at least 226 Northern European control chromosomes, none of the 18 genotyped heterozygotes had facial features of a manifesting carrier, and no genotype–phenotype correlation was observed; the phenotypic spectrum of biallelic ALX3 loss is narrow [27]. ALX3 sequencing was negative in 14 other frontonasal malformations (including one possible Pai syndrome, one OAFNS and three AFND) and in 93 patients with non-syndromic cleft lip and/or palate [27]. In 2018 the nonsense variant NM_006492.3:c.604C>T p.(Gln202Ter) was described in a consanguineous Pakistani family with 4 affected members [53, 54], and in 2011 the variant p.Tyr214Ter in two siblings assessed at the University of California, Davis, at a time when the genetic change had been found in 14 patients [55].

2.4. ALX4: one gene, several inheritance patterns

ALX4 illustrates a dosage effect with two inheritance patterns, both documented in primary sources and in OMIM [9, 51]. Heterozygous loss of function — whole-gene deletions, nonsense or frameshift variants at any position and, in part, homeodomain missense variants — causes enlarged parietal foramina type 2 (PFM2) through haploinsufficiency, with autosomal dominant inheritance [9, 29]. Biallelic loss causes FND2, with autosomal recessive inheritance, described mainly in consanguineous families from Turkey, Iran, Egypt, Pakistan and Bangladesh, and in a patient with meningoencephalocele followed at the Federal University of São Paulo [30, 46, 56–60]. The original founder variant, c.793C>T p.(Arg265Ter), truncates 146 amino acids (part of the homeodomain and the C-terminal domain) without triggering nonsense-mediated mRNA decay; skin biopsy showed a hypomorphic interfollicular epidermis with reduced suprabasal layers and follicular structures with altered differentiation [30]. The same variant was found again in the homozygous state in a child from Bangladesh, whose heterozygous father had enlarged parietal foramina [46]. Severity correlates with residual function: the homozygous missense variant p.(Gln225Glu) produced a mild phenotype (parietal foramina, notched nasal alae, normal cognition), and the heterozygous parents had small parietal foramina [4, 56].

Between the two extremes there is a zone of overlap. Heterozygotes from families with FND2 may have a mild form of FND, as in the Pakistani family with ALX4 c.652C>T p.(Arg218Trp), in which the three homozygotes had severe anomalies and the heterozygotes mild FND2 [59]. In addition, certain heterozygous variants cause FND with vertical transmission: a complex frameshift change distal to the homeodomain, in a family with mother-to-son transmission studied in São Paulo (Bertola and colleagues, 2013); the missense variant c.646C>G p.(Arg216Gly) in a three-generation family with 4 affected members, in which mild notching of the nasal alae accompanied the parietal foramina and was proposed as a predictive sign of heterozygosity [35, 36]. Peled and colleagues showed in 2023 that frameshift variants generating an elongated C-terminal tail increase ALX4 transcriptional activity in a reporter assay and alter the WNT/β-catenin pathway in keratinocytes, which explains a dominant FND2 with hypotrichosis and hypodontia through gain of function [37]. In 2026 two new reports of heterozygous variants with congenital diffuse hypotrichosis and frontonasal dysmorphism were published [38, 39]. GeneReviews summarises this heterogeneity: heterozygous loss of function causes parietal foramina; certain homeodomain missense variants and complex changes with dominant-negative or gain-of-function potential cause broader phenotypes in the heterozygous state; classic FND2 is biallelic [9]. Very rare ALX4 missense variants have also been proposed — for example NM_021926.4:c.19G>T p.(Val7Phe) and NM_021926.4:c.631A>G p.(Lys211Glu), found in 2 of 203 patients with isolated craniosynostosis, with gain of function in vitro and also present in an unaffected parent, which classifies them as low-penetrance variants [61] — as a susceptibility factor for non-syndromic craniosynostosis (CRS5, OMIM 615529), a plausible association but without genotype–phenotype correlation or clear functional separation from the variants that cause foramina [9, 51]. ALX4 is therefore among the coronal craniosynostosis genes to be distinguished from Saethre-Chotzen syndrome [62].

The contiguous 11p11.2 deletion that includes ALX4 and EXT2 gives rise to Potocki-Shaffer syndrome (enlarged parietal foramina and multiple osteochondromas); variable deletion of PHF21A largely explains the intellectual disability and facial dysmorphism [9, 29]. A female patient with severe FND, partial alopecia, hypogonadism and intellectual disability had a de novo 11p11.12p12 deletion including ALX4 and negative sequencing of the ALX genes and EFNB1, and the authors considered a second undetected variant on the remaining allele [63].

2.5. ALX1 and FND3

Total loss of ALX1 compromises the orchestrated fusion of the frontonasal, medial nasal, lateral nasal and maxillary processes [31]. In family 1 of Uz and colleagues, analysis of 250,000 polymorphisms revealed a Mendelian segregation error within a homozygous region of about 27 Mb, corresponding to a 3.7 Mb homozygous deletion at 12q21.31 containing seven genes (TMTC2, SLC6A15, LRRIQ1, ALX1, PAMCI, NTS and MGAT4C); the parents were heterozygous carriers with normal cranial tomography [31]. In family 2, the homozygous variant NM_006982.3:c.531+1G>A, at the invariant donor site of intron 2, absent from 171 Turkish controls, affects glutamine residue 177 of helix III [5, 31, 54]. An acceptor-site variant, c.661-1G>C, caused a mild form in a Pakistani family, with hypertelorism, ptosis, blepharophimosis, eyebrows present and normal cognition, attributed to residual activity [2, 5, 64]. In neural crest cells derived from induced pluripotent stem cells of patients homozygous for p.(Leu165Phe), Pini and colleagues observed increased sensitivity to apoptosis and impaired migration, with low BMP2 and raised BMP9, a defect reversed in vitro by soluble antagonists of both — a preclinical result [65]. OMIM keeps the ALX1–FND3 relationship as provisional [5].

2.6. ZSWIM6 and acromelic frontonasal dysostosis

ZSWIM6 encodes a protein of 1215 amino acids with a SWIM zinc finger domain and a C-terminal SIN3-like domain, expressed in the central nervous system and, in the mouse embryo, in the lens, cochlear hair cells, the outer root sheath of the hair follicle and the ameloblasts [66]. All genotyped patients with AFND have the same heterozygous substitution NM_020928.2:c.3487C>T p.(Arg1163Trp), in the SIN3-like domain of exon 14 [21, 32, 54]. Deletions encompassing the gene do not cause AFND, which points to a gain of function; analysis of osteoblast and fibroblast lines from two patients suggested Hedgehog pathway activation [2, 32]. Other variants in the same gene produce distinct phenotypes: the recurrent de novo nonsense variant c.2737C>T p.(Arg913Ter), in the penultimate exon, escapes mRNA decay and causes severe to profound intellectual disability without frontonasal or limb malformations (OMIM 617865), with a proposed dominant-negative effect [66, 67]; and the first homozygous variant, c.3119G>A p.(Arg1040His), described in 2026 in two siblings from a consanguineous Kurdish family, causes severe short stature with a growth hormone deficiency profile, microcephaly and delay, with frontal dysmorphism but without classic FND [68]. In luciferase assays, p.Arg1040His reduced activation of the ZIC2 promoter (loss of function), p.Arg1163Trp increased it (gain of function) and p.Arg913Ter increased activation of HECW2 (derepression) [68]. The AFND variant almost always arises de novo and may be mosaic, in the patient or in a mildly affected parent [21, 32].

2.7. Other genes of the spectrum: SIX2 and TWIST1

A heterozygous 108.3 kb deletion at 2p21, limited to SIX2 and adjacent non-coding DNA, segregated with a mild dominant FND in one family; the brachyrrhine mouse model, with dose-dependent reduction of Six2 expression owing to a probable regulatory change, reproduces FND and skull base anomalies [2, 33]. Sweeney-Cox syndrome results from the de novo substitutions NM_000474.4:c.350A>T p.(Glu117Val) and NM_000474.4:c.350A>G p.(Glu117Gly) in the basic domain of TWIST1, which never occur among the approximately 100 variants of Saethre-Chotzen syndrome (haploinsufficiency); the Caenorhabditis elegans model favours a dominant-negative mechanism [34]. The ClinGen panel classified this relationship as limited (3 to 4 points) [25].

2.8. Normalised variants

Normalisation in VariantValidator (GRCh38) corrected the description of the variant of Peled and colleagues: the insertion published as c.985_986insGTGC corresponds, under the HGVS 3' alignment rules, to the duplication c.982_985dup, and the predicted protein terminates 116 rather than 115 codons later [37, 54]. The variant c.793C>T p.(Arg265Ter), presented as novel in 2024, is the same as that described in 2009 [30, 46]. The 2009 ALX3 variants were described on transcript NM_006492.2 and are kept here with the authors' nomenclature.

2.9. MSX2: haploinsufficiency in parietal foramina and gain of function in craniosynostosis

MSX2 (5q35.2, OMIM 123101) encodes a homeodomain transcription factor of the Msx family, phylogenetically distant from ALX4 and with different biochemical properties; despite this, heterozygous variants in the two genes produce similar skull defects. Both are expressed in the ectomeninx, the mesenchymal condensation around the neural tube that gives rise to the bones of the vault and the dura mater, and in mice they have similar and additive roles in vault differentiation; the similarity of the bony and dural defects suggests a common disturbance of ectomeningeal patterning [10]. Whole-gene deletions, nonsense and frameshift variants at any position in the coding region and, in part, homeodomain missense variants cause loss of function and parietal foramina through haploinsufficiency; because the second exon of MSX2 is the terminal exon, truncating variants located there probably escape nonsense-mediated mRNA decay [9, 10].

In 2006, the Oxford group collated seven intragenic variants and one whole-gene deletion of MSX2 from the literature and added two frameshift variants, c.417_418del and c.456_465dup (authors' numbering, from the initiation codon) [10]. The variant NM_002449.5:c.515G>A p.(Arg172His), described in the first demonstration of haploinsufficiency of the gene, was found in two unrelated families with parietal foramina and headaches and is classified in ClinVar as likely pathogenic [10, 28, 70]. The variant NM_002449.5:c.443C>A p.(Pro148His), by contrast, increases the affinity of the homeodomain for DNA without altering sequence specificity and causes autosomal dominant Boston-type craniosynostosis (OMIM 604757), without parietal foramina; an additional copy of MSX2 in 5q duplications has also been associated with craniosynostosis, and non-coding microduplications upstream of the gene with a phenocopy of cleidocranial dysplasia [9, 10, 72]. The gene thus shows a dosage effect in both directions: reduced activity delays ossification of the vault and increased activity brings forward suture fusion.

There is no clear genotype–phenotype correlation between the different MSX2 loss-of-function variants; single variants in isolated families have been associated with congenital aplasia cutis of the scalp and with clavicular hypoplasia, which may indicate mild dominant-negative effects [9]. In the family with dental anomalies and an MSX2 variant studied in Oxford, the dental defects segregated independently of the variant, and MSX2 was not found to be altered in a study of Adams-Oliver syndrome, which combines scalp and skull defects [10]. In screening of ALX4 and MSX2 in 181 patients with craniosynostosis without point variants in FGFR1, FGFR2, FGFR3 and TWIST1, there were no rare MSX2 variants and only two ALX4 variants of uncertain significance, one of them present in the unaffected father; the authors concluded that study of these genes is not justified routinely in craniosynostosis [10]. Penetrance of the foramina is high but incomplete, because the defect decreases with age and may close completely; residual variability within families is compatible with genetic modifiers, and screening for coding variants in the reciprocal gene (ALX4 in carriers of MSX2 variants and vice versa) did not distinguish carriers with and without a defect [10].

2.10. PIGB and NBAS: acrofrontofacionasal dysostosis outside the ALX pathway

PIGB (15q21.3, OMIM 604122) encodes GPI mannosyltransferase III, a 554-amino-acid transmembrane protein of the endoplasmic reticulum that adds the third mannose to the core of the GPI anchor [73]. The anchor attaches a large number of surface proteins to the membrane; when the intermediate steps of synthesis fail, such as the one catalysed by PIGB, the substrate proteins are cleaved and released into the medium instead of being anchored [14, 73]. Biallelic PIGB variants mainly cause a developmental and epileptic encephalopathy (DEE80, OMIM 618580): in ten families, most patients had global developmental delay or intellectual disability, all had seizures, two had polymicrogyria and four axonal peripheral neuropathy, eight children died before 4 years of age, two had a clinical diagnosis of DOORS syndrome (deafness, onychodystrophy, osteodystrophy, intellectual disability and seizures), most of those tested had raised alkaline phosphatase, and lymphocytes and fibroblasts showed reduced GPI-anchored proteins on the cell surface [74]. Defects of GPI anchor biosynthesis, including DEE80, are part of the differential diagnosis.

In the two original Brazilian AFFND1 families, studied at the Hospital for Rehabilitation of Craniofacial Anomalies of the University of São Paulo (Bauru), without NBAS variants, exome sequencing found the intronic variant NM_004855.5:c.795-19T>G in the homozygous state, which segregated with the disease in both families, was absent from the Brazilian polymorphism databases BIPMed and ABraOM and was rare in gnomAD [14, 54, 73]. A minigene assay in HEK293T cells and patient RNA showed skipping of exon 7, with a frameshift and a premature termination codon in exon 8 (p.(Phe266GlyfsTer9) at transcript level); no normally spliced transcript was detected, indicating complete loss of function [14, 73]. The variant is classified in ClinVar as likely pathogenic [70]. The authors interpret AFFND1 as the null extreme of the spectrum of GPI biosynthesis defects, in keeping with the severity of the phenotype, and consider a founder effect in the Brazilian population, from which almost all described patients come [14].

NBAS (2p24.3, OMIM 608025) encodes a component of the syntaxin 18 tethering complex in the endoplasmic reticulum, which interacts with the SNARE proteins p31, BNIP1, STX18 and Sec22b and participates in retrograde transport from the Golgi apparatus to the reticulum; it also acts in nonsense-mediated mRNA decay, regulating genes of protein trafficking, of reticulum-coupled protein modifications and of bone mineralisation [19, 71]. In the two sisters of Indian origin described by Prontera and colleagues in 2011, daughters of consanguineous parents, exome sequencing identified in the homozygous state a splice-site variant in intron 47, described in OMIM as c.6237-3G-C; normalisation in VariantValidator shows that, on the coding strand of the MANE transcript, the change is NM_015909.4:c.6237-3C>G, the same as that recorded in ClinVar, with conflicting classifications [18, 54, 70]. The truncated protein, overexpressed in HEK293T cells, impaired nonsense-mediated decay and retrograde transport; in the mouse embryo, NBAS is expressed in chondrogenic and osteogenic structures of the skeleton and in the cortex, hippocampus and cerebellum, and depletion of the orthologue in zebrafish disturbed the morphogenesis of the cranial cartilaginous elements [19]. OMIM classifies the NBAS–AFFND3 relationship as provisional, because it rests on a single family [18].

Biallelic NBAS variants cause a spectrum much more frequent than AFFND3. SOPH syndrome (short stature, optic atrophy and Pelger-Huët anomaly, OMIM 614800) was described in 33 Yakuts, an isolated population of the far east of the Russian Federation, homozygous for NM_015909.4:c.5741G>A p.(Arg1914His); infantile liver failure syndrome type 2 (ILFS2, OMIM 616483) presents with episodes of acute liver failure triggered by fever, with a mean onset at 13 months [54, 71]. Patient fibroblasts have 18% to 36% of normal NBAS levels, which fall further when the temperature rises from 37 °C to 40 °C, supporting the hypothesis that fever triggers hepatocyte apoptosis through endoplasmic reticulum stress [71]. In an international cohort of 110 patients from 97 families, with 86 variants, the location of missense variants and in-frame deletions defined three subgroups with significant differences in the main manifestations: β-helix (combined phenotype), Sec39 domain (ILFS2) and C-terminal region (short stature, optic atrophy and Pelger-Huët anomaly, or SOPH) [75]. Recurrent acute liver failure due to NBAS is likewise part of the differential diagnosis.

3. Clinical Spectrum and Manifestations by System

3.1. The FND sequence: common manifestations

The phenotype varies along a continuum between mild forms, with hypertelorism and a bifid nasal tip, and severe forms with a median or oblique facial cleft and encephalocele [26, 76]. In severe cases, the craniofacial region may show a widow's peak, a midline frontal bone defect (cranium bifidum occultum) and a median facial cleft of the nose and/or upper lip and palate; the eyes, true hypertelorism (interpupillary distance above the 97th centile) with lateral displacement of the inner canthi; the nose, anything from a broad notched tip to completely separated nostrils, with hypoplasia or absence of the prolabium and premaxilla, notching of the nasal alae, a broad nasal root and lack of formation of the tip [26]. Occasional anomalies include accessory nasal tags, microphthalmia, coloboma, cataract, preauricular tags, low-set ears, conductive hearing loss, frontal cutaneous or corpus callosum lipoma, agenesis of the corpus callosum, basal encephalocele and tetralogy of Fallot [1, 26]. The association with tetralogy of Fallot was described in three unrelated children, all with hypertelorism, a median nasal groove and absence of the nasal tip and without cognitive impairment [3]. The widow's peak results from the wide separation of the eyes: the bilateral periocular zones of hair growth suppression overlap at a lower point than usual, leaving a point of hair implantation in the midline [77].

Most individuals have normal intelligence: in DeMyer's series, 8% of the 33 patients had severe intellectual disability and 12% impaired intelligence [26]. Intellectual disability is more likely when hypertelorism is severe or there is extracranial involvement [1]. A review of 104 cases grouped patients into seven phenotypic subtypes in addition to isolated FND [3]. These figures describe historical clinical series, almost all without a molecular diagnosis, and do not allow the cognitive prognosis of each genetic form to be estimated.

3.2. Frontorhiny (FND1, ALX3)

The major features of frontorhiny are hypertelorism, a broad nasal bridge, a short nasal dorsum, widely separated nasal bones with a bifid tip, a broad columella inserting into the face above the alae, widely separated slit-like nostrils, a long philtrum with prominent ridges that sometimes form bilateral swellings up to the nostrils, and midline notching of the upper lip and alveolus [27]. In the original series of 11 patients, occasional findings were upper eyelid ptosis (3 patients), inclusion dermoid cysts of the craniofacial structures, philtral pits or rugose folds behind the pinnae (2 each), iris coloboma, strabismus and corpus callosum lipoma (1 each); none of the 5 patients with cranial imaging had craniosynostosis [27]. Growth and development were normal in all, but most required several operations [27]. Two findings have clinical consequences: one patient had six episodes of pneumococcal meningitis with cerebrospinal fluid (CSF) rhinorrhoea owing to multiple defects of the cribriform plate of the ethmoid, and another had frontal subcutaneous swellings over paramedian foramina of the frontal bone, suggestive of dermoids with potential transcranial extension [27]. Orphanet adds a low hairline and swellings at the base of the columella [42]. In a Pakistani family, the 4 affected members had hypertelorism, a bifid nasal tip, a broad bridge and a widened philtrum, and all had normal cognition [3, 53]. In a Turkish surgical series, a widened philtrum with prominent philtral columns was the feature that best distinguished FND1 from the other ALX forms [22].

3.3. FND2 (biallelic ALX4)

Classic FND2 combines a large skull defect (cranium bifidum with extensive parietal foramina and aplasia of the ethmoid and nasal bones), coronal craniosynostosis, brachycephaly, severe hypertelorism, a very depressed nasal bridge and dorsum, a bifid tip, notched nasal alae, total alopecia or very sparse hair, sparse eyebrows and eyelashes, hypogonadism, cryptorchidism, hypospadias, agenesis or anomalies of the corpus callosum, a midline intracranial lipoma and variable intellectual disability [2, 26, 30, 43]. Notching of the nasal alae is the constant sign in all biallelic patients described up to 2014 [36]. The Iranian girl of Kariminejad and colleagues, at 2 years of age, had very sparse hair, eyebrows and eyelashes, absent body hair, hypertelorism, short upslanting palpebral fissures, telecanthus, a thick columella, hypoplastic alae and notched nostrils; imaging showed cranium bifidum with parietal foramina and small, abnormal occipital lobes; skin biopsy revealed mild hyperkeratosis and poorly developed hair follicles, with reduced lacrimation and sweating and normal cognition [4]. The two Egyptian boys homozygous for c.291del had a large skull defect, alopecia, hypertelorism, notched alae and impalpable gonads; in one of them, bilateral parietal meningocele-like cysts enlarged with age; neuroimaging showed small, dysplastic occipital lobes with dysgyria and a midline subarachnoid cyst [58]. In the series of Vargel and colleagues, FND2 was the ALX form with the most severe deformities: severe hypertelorism, brachycephaly, large parietal defects, a broad nasal dorsum and alopecia [22].

The spectrum includes mild forms. The boy homozygous for p.(Gln225Glu) had hypertelorism, a broad nasal bridge, dorsum and tip, a bifid tip, notched alae, a broad columella, a unilateral preauricular tag, a shallow labiogingival sulcus and large parietal foramina, with an angulated corpus callosum, a small vermis and normal psychomotor development [4, 56]. At the opposite extreme, a child homozygous for p.(Arg265Ter) had biparietal meningoencephaloceles, dural venous anomalies, preaxial polydactyly of the foot and a fatal outcome (section 10) [46]. The dominant forms due to heterozygous variants with a dominant-negative or gain-of-function effect present with a mild nasal cleft, a broad columella, a broad nasal bridge, sparse hair or hypotrichosis, hypodontia, broadening of the thumb or hallux, genital anomalies and, in some families, neurobehavioural changes, almost always with parietal foramina [9, 35–37].

3.4. FND3 (ALX1)

FND3 is the most severe ALX form, because of the ocular involvement associated with the facial clefts [22, 76]. The proband of Uz and colleagues was born with hypertelorism, extreme bilateral microphthalmia, upper eyelid colobomas, sparse eyelashes, absent eyebrows, a broad nasal base with hypoplastic alae, non-median bilateral cleft lip, a prominent glabella, complete cleft palate, low posteriorly rotated ears and a sacral caudal appendage without a vertebral defect; magnetic resonance imaging (MRI) showed brachycephaly and a large midline cranial bone defect [31]. At 11 years of age he had mild intellectual disability with special education, followed commands and had normal hearing, but the facial anatomy limited expressive speech; he walked at 12 months [31]. The two sibling fetuses, terminated at 25 and 23 weeks, had an identical craniofacial phenotype, agenesis of the corpus callosum, asymmetrical or hypoplastic optic nerves, mild cerebellar hypoplasia, neuronal heterotopia of the pons and cystic dilatations or delayed renal maturation [31]. The patient from the second family had extreme right-sided microphthalmia, cranium bifidum, a frontal soft-tissue mass (dermoid or fibroma), markedly delayed myelination and a thin corpus callosum; she underwent surgery and died in infancy from a pulmonary infection [31]. The authors placed the phenotype on the boundary between frontofacionasal dysplasia and Fryns microphthalmia syndrome [5]. The splicing variant c.661-1G>C was associated with a mild phenotype, with eyes and eyebrows present and normal development [5, 64]. In the Turkish series of 89 FND patients, 8 had FND3 [22].

3.5. Acromelic frontonasal dysostosis (ZSWIM6)

Verloes and colleagues delineated AFND in 1992 as a variant of FND with central nervous system and limb malformations; Prescott and colleagues had already grouped, in 1989, eight cases of median cleft face at the severe extreme, with agenesis of the nasal root, very widely separated slit-like nostrils, tibial aplasia, polydactyly of the hallux and intellectual disability, and Slaney and colleagues described five infants in 1999 [6]. AFND combines severe FND — severe hypertelorism, ptosis, downslanting palpebral fissures, median cleft face with nasal bifurcation and very widely separated alae, slit-like nostrils, a “carp-shaped” mouth with a midline lip notch, cleft palate — with cranium bifidum or parietal foramina, brain malformations (interhemispheric lipoma, agenesis or dysgenesis of the corpus callosum, hydrocephalus, periventricular nodular heterotopia, absent olfactory bulbs) and lower limb anomalies (symmetrical preaxial polydactyly of the feet, bifid hallux, tibial hypoplasia or aplasia, clubfoot); intellectual disability is severe, and hypopituitarism and cryptorchidism may coexist [2, 21, 26, 45, 78]. In the series of Twigg and colleagues, the boy with classic AFND, at 7 years of age, could neither walk nor talk; another patient with the same variant had no limb anomalies and was recognised only by genotyping; and a 20-week fetus had tibial hypoplasia and clubfoot without polydactyly [21]. Limb anomalies are therefore not mandatory for the diagnosis [21]. The mosaic mother of the first patient had only hypertelorism, a short broad nose with a bifid tip, normal intelligence and no extracranial features, with a phenotype similar to frontorhiny [21].

3.6. Enlarged parietal foramina and Potocki-Shaffer syndrome

Enlarged parietal foramina are symmetrical, paired radiolucencies of the parietal bones, near the intersection of the sagittal and lambdoid sutures, due to deficient ossification around the parietal notch, which normally closes by the 5th fetal month [9, 28]. They are usually asymptomatic and appear as a prenatal ultrasound finding, a large posterior fontanelle in an infant or an incidental radiographic finding; in infants, a single central defect (cranium bifidum) tends to divide into separate foramina during the first years through ossification of a midline bony bridge, and the foramina decrease with age but may persist throughout life [9]. Meningeal, cortical and vascular malformations of the posterior fossa occasionally accompany the defects and may predispose to epilepsy; a minority have headaches, vomiting or severe local pain, especially with pressure on the unprotected cortex; scalp defects and skull fracture from direct trauma have been described [9]. With severe defects, the picture may mimic an occipital encephalocele in the prenatal period [9, 79]. The skull defects do not differ between the MSX2 and ALX4 forms [9]. The obsolete eponym “Catlin marks” comes from the family studied by Goldsmith in 1922, with 16 affected members over 5 generations, many of them with circumscribed aplasia of the scalp; familial radiographic series documented the transition from cranium bifidum in infancy to separate parietal foramina in adulthood, showing that the two presentations are age-dependent; associations with seizures and with cleft lip or palate have also been described, as well as a third locus (PFM3, 4q21-q23, OMIM 609566) with limited evidence [8, 9]. In a study of 250 adult skulls from anatomical collections published in 2026, 2 (0.8%) had enlarged bilateral foramina, with smooth corticated margins, abrupt cortical thinning and fibrovascular tissue without osteoclastic activity, confirming that this is a congenital ossification defect and not erosion [80]. Potocki-Shaffer syndrome adds multiple osteochondromas, intellectual disability and craniofacial dysmorphism (brachycephaly, large fontanelles, ptosis, epicanthus, a broad depressed nasal tip, a short philtrum, micrognathia) [9, 81].

In the Oxford series published in 2006, which collated radiographic measurements from 35 carriers of ALX4 variants (11 families) and 17 of MSX2 variants (7 families), the width of the defect as a proportion of maximum skull width decreased significantly with age for both genes (P < 0.01), with no difference between them after adjustment for age (P = 0.30) [10]. The exception was the ALX4 variant NM_021926.4:c.653G>A p.(Arg218Gln), present in 7 individuals from 2 families, with disproportionately wide defects (P < 0.001), a tendency to persistent cranium bifidum and posterior fossa malformations — polymicrogyria, separation of the occipital lobes, persistence of the embryonic median prosencephalic vein with hypoplasia of the straight sinus and high insertion of the tentorium cerebelli —, suggesting an additional dominant-negative effect; the analysis was post hoc and requires replication [10]. In a family with the variant NM_021926.4:c.620C>A p.(Ser207Ter), the five affected members studied had, in addition to the foramina, absent or malpositioned teeth, short broad thumbs and halluces and mild facial dysmorphism; in a family with an MSX2 variant, one of the children had a small muscular ventricular septal defect and patent ductus arteriosus, without an established causal relationship [10]. Normal small parietal foramina, which transmit the emissary veins, are present in 60% to 70% of skulls and are distinguished from enlarged foramina, for which a threshold of 5 mm in diameter has been proposed [69, 82]. In a cross-sectional study of 60 adult skulls from northern India, carried out in 2025, the foramen was absent in 20%, bilateral in 60% and unilateral in 16.6%, with a mean diameter of 0.82 mm (0.3–1.5 mm) and a mean distance from lambda of 31.9 ± 5.15 mm [83]. Outside the clinical setting, bilateral foramina of about 12 mm in an adult skull required, in a Brazilian forensic examination, distinction from old bullet holes by their blunt margins, bilateral distribution and absence of residues and radiating fractures [84].

3.7. Other syndromic forms of the spectrum

FND due to SIX2 deletion presented in a girl assessed at 22 months with a head circumference at the 95th centile, an anterior fontanelle of 6.4 × 4.0 cm, inner and outer intercanthal distances of 3.4 cm and 9.1 cm (both above the 95th centile), ptosis, epicanthus inversus, a flat nasal bridge and a broad tip; computed tomography (CT) showed a persistent metopic suture, bilateral parietal foramina, complete sagittal synostosis at 31 months, premature lateral spheno-occipital synchondrosis and a persistent craniopharyngeal canal; macrocephaly progressed to +2.8 standard deviations, development remained normal and renal function was normal; the carrier mother had a similar facies [33]. Oculoauriculofrontonasal syndrome combines FND with features of the oculo-auriculo-vertebral spectrum (preauricular tags, auricular dysplasia, mandibular asymmetry, epibulbar dermoids, eyelid coloboma, costovertebral anomalies); in a cohort of 130 patients with FND, 18 had OAFNS, 56% of whom had nasal polyps, and micro- or anophthalmia, cataract, thyroid agenesis, polymicrogyria, olfactory bulb hypoplasia and mandibular cleft were described for the first time [85]. Frontofacionasal dysplasia, described by Gollop in 1984, combines posterior encephalocele, blepharophimosis, lagophthalmos, telecanthus, S-shaped palpebral fissures, eyelid coloboma, widow's peak, cranium bifidum, frontal lipoma, nasal hypoplasia and paramedian facial clefts, with suspected recessive inheritance; Gollop's original patients were siblings born to related parents [26, 31, 86]. OAFNS was proposed in 1983 by Golabi and colleagues in two siblings, children of a diabetic mother, with FND, ocular dermoids and preauricular tags [87]. FND with alar clefts (OMIM 203000) corresponds to a single observation of two sisters with hypoplasia and coloboma of the alar cartilages and telecanthus [88]. FND with optic disc anomalies (6 sporadic patients) combines sphenoethmoidal encephalocele, agenesis of the corpus callosum, optic disc anomalies (peripapillary staphyloma, hypoplastic discs, absent chiasm, morning glory disc) and pituitary dysfunction (growth hormone deficiency, hypothyroidism, diabetes insipidus, adrenal insufficiency) [3, 26]. Acrofrontofacionasal dysostosis type 1 (AFFND1, OMIM 201180) is a recessive disorder with severe psychomotor delay, hypertelorism, a broad nose with a notched tip, cleft lip and palate, broad prominent upper central incisors, colobomas and other ocular anomalies, late-onset hearing loss, mesomelic limb shortening, distal digital hypoplasia, fibular hypoplasia and clubfoot; it was described in three Brazilian families (Richieri-Costa and colleagues, 1985 and 1992; Guion-Almeida and Richieri-Costa, 2003) and attributed in 2021 to biallelic variants in PIGB, a gene of glycosylphosphatidylinositol anchor biosynthesis; AFFND3 (OMIM 621629) results from NBAS variants, and AFFND2 (OMIM 239710), with genitourinary anomalies, remains without a gene [2, 12]. Guion-Almeida and Richieri-Costa described in 2009 10 Brazilian boys with FND, cleft lip and palate, severe neuropsychological delay without language acquisition and midline central nervous system anomalies (agenesis of the corpus callosum, large interhemispheric cysts, gyral anomalies), proposed as a distinct syndrome of unknown cause [3]; in another Brazilian series with frontonasal dysgenesis, callosal agenesis, basal encephalocele and ocular anomalies, sequencing of TGIF, ALX3 and ALX4 was negative [89].

3.8. Comparative summary of the genetic forms

3.9. Acrofrontofacionasal dysostosis (types 1, 2 and 3)

Acrofrontofacionasal dysostosis combines FND with acral anomalies and, in types 1 and 3, with severe neurodevelopmental impairment, which separates it from the ALX forms, in which cognition is usually normal. Orphanet defines it as a congenital malformation syndrome with facial and skeletal anomalies, severe intellectual disability and, occasionally, genitourinary anomalies, and counts 12 cases or families, with a prevalence below 1/1,000,000 [13]. All described families were consanguineous or had affected siblings, and there are no natural history studies; the following data come from case reports.

Type 1 (PIGB). Richieri-Costa and colleagues described in 1985 two siblings, children of Brazilian first cousins, with intellectual disability, short stature, hypertelorism, a broad notched nasal tip, cleft lip and palate, postaxial campto-brachy-polysyndactyly, fibular hypoplasia and marked anomalies of foot structure; they considered that, despite the facial similarity to FND, the condition was a multiple congenital anomaly syndrome with intellectual disability, probably of recessive inheritance [40]. In 1992, the same team described a 26-year-old Brazilian woman with severe intellectual disability, ocular anomalies, cleft lip, a high-arched palate, camptobrachysyndactyly of the hands and feet and foot anomalies, whose similarly affected brother died at 1 year of age [90]. The third Brazilian family, published in 2003, had a brother and sister, children of second cousins [12]. At the 2021 molecular reassessment, the 1992 patient, aged 55, had brachycephaly, a long face, a broad nasal base, hypertelorism, distal digital hypoplasia with hypoplastic nails, genu valgum, tibial hypoplasia and clubfoot, intellectual disability with language delay, bilateral blindness from scarred corneal ulcers, polar cataracts and colobomas of the iris, retina and optic nerve, and sensorineural hearing loss beginning in the fifth decade (complete deafness on the right and severe on the left at 52 years of age); she had breast carcinoma in the sixth decade [12, 14]. The boy from 2003, re-examined at 26 years of age, had macrocephaly, a coarse face, hypertelorism, a broad nasal base and tip, repaired cleft lip and palate, large ears, digital contractures, distal digital hypoplasia, short tibiae, flexed lower limbs and clubfoot; vision was reduced by colobomas of the iris, retina and optic nerve, and severe intellectual disability with aggressive behaviour prevented further ophthalmological examination; he had no sphincter control, did not feed himself or eat solids owing to chewing and swallowing difficulties, and did not attend school [12]. The OMIM synopsis describes severe psychomotor delay, hypertelorism, a broad nose with a notched tip, cleft lip and palate, broad prominent upper central incisors, colobomas and other ocular anomalies, late-onset hearing loss, mesomelic limb shortening, distal digital hypoplasia, fibular hypoplasia and clubfoot [12]. The descriptions consulted do not mention epilepsy, in contrast with DEE80 due to the same gene [12, 74].

Type 3 (NBAS). Prontera and colleagues described in 2011 two sisters of Indian origin, assessed in Italy, with a condition they classified as AFFND1 [18]. The elder had at birth bilateral cleft lip and palate, complete bilateral syndactyly of the 3rd and 4th fingers and clubfoot; in the first month, mixed hearing loss, optic nerve atrophy and an abnormal electroencephalogram were documented, and radiographs showed hypomineralisation of the calvaria, a large anterior fontanelle, hip dislocation, short long bones, brachymetacarpia, hypoplastic distal phalanges, polysyndactyly (a supernumerary digit between the 3rd and 4th) and hypoplastic fibulae [18]. At 11 years of age, height, weight and head circumference were below the 3rd centile; she had profound intellectual disability, could neither walk nor talk, and had brachycephaly, a broad forehead, hypertelorism, ptosis, a broad nasal bridge with a broad notched tip, oligodontia, hypoplastic gums, broad prominent upper central incisors, drooping ears with broad conchae, thin helices and absent lobules, diffuse muscle hypotrophy, scoliosis, mesomelia of all four limbs, broad proximally implanted thumbs, camptodactyly and clinodactyly of the fifth fingers and short metatarsals [18]. The younger sister had at birth microcephaly with suture diastasis, hypertelorism, epicanthus, cleft palate, short arms with skin creases, flexed and adducted thumbs and a sacral dimple with a hair tuft, without polysyndactyly; auditory evoked potentials showed a partial deficit and visual evoked potentials a severe deficit, with bilateral optic atrophy, pupils unreactive to light and exotropia; at 8 months there was severe growth retardation, extreme axial hypotonia with distal hypertonia and hyperreflexia, and magnetic resonance imaging showed only a slightly thinned corpus callosum [18, 19]. At 2 years and 7 months she had trunk control but showed pyramidal signs, stereotyped movements, poor social interaction, absent verbal language and an intelligence quotient of 20; at 19 and 18 years of age, the two sisters remained in a stable but very severe clinical state [18]. The sources consulted do not describe hepatic crises in this family, although fever is the usual trigger of liver failure in the other NBAS disorders [71].

Type 2 (no gene). Richieri-Costa and colleagues described in 1989 a 6-month-old boy, the son of first cousins, with microbrachycephaly, a broad forehead, marked hypertelorism, a broad nose with a median groove and a small blind pit on each side, hypospadias, syndactyly between the 3rd and 4th fingers and broad thumbs and halluces, with a normal karyotype; a sister with an identical phenotype died at 15 days of age from respiratory distress [15, 17]. Teebi proposed in 1992 that this was the same entity described by Naguib in 1988 in a Kuwaiti girl and her two brothers, children of first cousins, with hypertelorism, proptosis, ptosis, polysyndactyly, hypospadias and normal stature, and designated it Naguib-Richieri-Costa syndrome [15]. In 2008, a 22-day-old Tunisian boy, the son of first cousins, had a frontal encephalocele, a wide anterior fontanelle, marked hypertelorism, upper eyelid colobomas, proptosis, congenital glaucoma, a broad nose, syndactyly of the 3rd and 4th fingers, hypoplasia of the 3rd to 5th toes with nail dysplasia, hypospadias with a cleft glans and a bifid scrotum; magnetic resonance imaging showed a right frontal encephalocele and gyration anomalies, without a corpus callosum anomaly [91]. Orphanet records three families [16].

4. Clinical Recognition and Warning Signs

4.1. When to suspect

In the prenatal period, second-trimester ultrasound may show hypertelorism, median nasal bifidity, cleft lip, a midline or posterior skull defect and, in AFND, lower limb anomalies; only very marked hypertelorism is reliably predicted, although reference values for ocular distance by gestational age exist [76, 92]. Large parietal defects are detectable from mid-gestation and may mimic an occipital encephalocele [9, 79]. At birth, FND is recognised by the combination of true hypertelorism with changes of the nasal root and tip; the presence of two Sedano and Gorlin criteria is sufficient to define it [2]. In mild forms, the diagnosis may only be made in adolescence, as in a girl assessed at 14 years of age for cosmetic surgery [93]. Enlarged parietal foramina are suspected when there is a soft, flattened, symmetrical area palpable behind the vertex of the skull, or a persistently large posterior fontanelle [9].

4.2. Clues to the genetic form

Some signs point to the genetic form. A long, widened philtrum with prominent ridges and a broad columella inserted above the alae suggest ALX3 [22, 27, 76]. Alopecia, notching of the nasal alae and large parietal defects suggest ALX4; in a setting of consanguinity, the presence of parietal foramina in a parent strengthens the hypothesis [36, 46, 58]. Extreme microphthalmia with oblique facial clefts and eyelid colobomas suggests ALX1 [31]. Severe symmetrical FND with very widely separated nostrils, especially with preaxial polydactyly of the feet, tibial hypoplasia or an interhemispheric lipoma, suggests ZSWIM6, even when the limbs are normal [21]. Coronal craniosynostosis with facial asymmetry, nails with longitudinal grooves or splits and sloping shoulders suggests craniofrontonasal syndrome [76]. The configuration of the eyebrows and renal agenesis point to the FREM1 phenotypes [76]. FND with cleft lip and palate, mesomelic shortening, fibular hypoplasia, clubfoot, polysyndactyly and severe intellectual disability suggests acrofrontofacionasal dysostosis; Brazilian origin and colobomas point to PIGB, optic atrophy with early hearing loss to NBAS, and hypospadias with syndactyly of the 3rd and 4th fingers to type 2 [12, 15, 18]. A palpable bilateral parietal defect in an infant without other anomalies suggests parietal foramina, even when the parents have no clinical signs, because the defects of adult carriers may have closed [10, 69].

4.3. Warning signs

Warning signs include watery rhinorrhoea and recurrent meningitis, particularly pneumococcal, which require investigation for defects of the cribriform plate or anterior skull base (described in frontorhiny and severe FND) [3, 27]; midline frontal or nasal swellings, which may be dermoids with an intracranial tract [27]; progressive enlargement of parietal meningoceles and CSF leaks in FND2 [46, 58]; nasal breathing difficulty, obstructive apnoea or feeding difficulty in the newborn with facial clefts [23]; corneal exposure due to lagophthalmos or eyelid colobomas [26, 31]; and, in FND with optic disc anomalies, polyuria, failure to thrive or signs of hypothyroidism or adrenal insufficiency, which indicate pituitary dysfunction [3]. In parietal foramina, seizures, headaches with vomiting and local pain on pressure warrant neurological assessment and imaging [9]. There is no newborn screening for these disorders.

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Clinical criteria

There are no consensus criteria for enlarged parietal foramina; confusion with normal small parietal foramina is unlikely given the size, location, natural history and family history [9].

5.2. Clinical and imaging assessment

Ocular measurement requires the inner and outer intercanthal and interpupillary distances, related to head circumference; in clinic, the interpupillary distance is difficult to measure accurately and there are parallax errors, and radiological orbital measurements are an alternative [76]. It is essential to separate true hypertelorism (increased distance between the orbits) from telecanthus (lateral displacement of the inner canthi and lacrimal puncta, as in Waardenburg syndrome type 1) [76]. The history includes a three-generation family history, consanguinity, twinning, threatened miscarriage, maternal illness and teratogens; the examination includes stature, skull shape and symmetry, frontal encephalocele, the implantation of the hair and eyebrows, the nose viewed from above, clefts, the philtral pillars, the heart, the genitalia, the hands, the feet and the nails [76, 78]. Before counselling an apparently sporadic case, the parents should be examined and their ocular measurements recorded [76].

Skull radiography identifies craniosynostosis and frontal defects; three-dimensional CT with a bone window defines the cranial and facial defects and plans surgery; brain MRI assesses lipomas, agenesis of the corpus callosum, encephaloceles and cortical and vascular malformations [9, 27, 76]. In parietal foramina, the posteroanterior radiograph shows the “pair of spectacles” pattern (less evident on the lateral view), 3D CT with a bone window reveals the defect and MRI is superior for meningeal, vascular and cortical changes [9]. The assessment also includes ophthalmology (strabismus, microphthalmia, colobomas, optic disc), hearing, renal ultrasound when FREM1 or FND3 is considered and, in forms with midline brain anomalies, endocrine assessment [3, 23, 76].

5.3. Genetic tests

The parameters of the American Cleft Palate-Craniofacial Association (ACPA, 2024) recommend that the genetic evaluation be carried out by a clinical geneticist with a genetic counsellor, with a full history, a three-generation pedigree and a whole-body examination, and that the test be chosen according to the presentation — chromosomal microarray, targeted test, extended panels, exome or genome —, with pre- and post-test counselling [23]. The 2021 guideline of the American College of Medical Genetics and Genomics (ACMG) recommends exome or genome sequencing as a first- or second-tier test in children with congenital anomalies with onset before 1 year of age [20]. For parietal foramina, GeneReviews proposes a multigene craniofacial panel including ALX4 and MSX2, or the exome when the presentation is atypical [9].

5.4. Limitations of the tests

In parietal foramina, about 70% of cases are attributed to ALX4 and 30% to MSX2; in each gene, about 90% of pathogenic variants are detected by sequencing and about 10% require targeted deletion and duplication analysis (quantitative polymerase chain reaction, multiplex ligation-dependent probe amplification, or MLPA, and targeted array), because the exome and genome detect these variations with lower sensitivity [9]. These proportions describe the distribution of the variants found and not the clinical sensitivity of the test in a patient with FND. In FND3, the cause in one of the two founding families was a 3.7 Mb homozygous deletion recognised through a Mendelian segregation error and copy number analysis of polymorphism array data, confirmed by array comparative genomic hybridisation [31]. In AFND, Sanger sequencing only detects mosaicism above about 6%: in a mildly affected mother, the variant was negative in blood and only deep sequencing (mean depth >152,800 reads) found it, with an 11% variant allele fraction in buccal mucosa (about 22% of cells), 3% in saliva, about 2% in urine and blood and less than 1% in skin [21]. Most FND remain without a molecular diagnosis: the six genes associated with the spectrum in 2016 explained a small fraction of cases, and among the patients with OAFNS studied, sequencing of the ALX genes (4 patients), next-generation panels (4), the exome (4) and the genome (1) were negative [2, 85]. A negative result does not exclude a genetic cause, particularly mosaicism, a regulatory variant or a gene not yet described.

The report should state the gene, the MANE transcript with version, the HGVS nomenclature, the zygosity, the class according to the criteria of the ACMG and the Association for Molecular Pathology (ACMG/AMP) and, in the recessive forms, confirmation of phase by parental study; a variant of uncertain significance in ALX4 or MSX2 neither establishes nor excludes the diagnosis [9]. Family segregation and cranial imaging of the parents help to interpret ALX4 variants, whose heterozygotes may have parietal foramina with reduced, age-dependent penetrance [9, 58]. For MSX2, there is a processed pseudogene at 17q22 (MSX2P1), to be considered in assay design [9]. In the ClinGen panel, only the MSX2–parietal foramina relationship had been curated by 2026; the relationships of the ALX genes and ZSWIM6 are under curation, and according to the ACMG technical standards only genes with moderate or stronger evidence should be included in diagnostic panels [25].

5.5. Prenatal diagnosis

Detailed ultrasound between 18 and 20 weeks usually detects enlarged parietal foramina in a fetus with an a priori risk of 50%, and fetal MRI is an option; the information is used to plan delivery [9]. In FND3, two successive affected pregnancies were diagnosed by ultrasound and terminated at 23 and 25 weeks [31]. In AFND, a fetus was identified at 19+3 weeks by hypertelorism, a broad glabella, nasal hypoplasia and bilateral clubfoot [21]. In a case of severe FND with limb anomalies, hydrocephalus, a hypoplastic corpus callosum and a ventricular septal defect, three-dimensional ultrasound at 20 weeks defined the phenotype, but the karyotype, CNV analysis and trio exome and genome were negative; according to the authors, only about 10 non-syndromic FND had been diagnosed in the prenatal period in 30 years [92]. In 2026 the use of 3D-printed models from ultrasound volumes was proposed [95]. When the familial variant is known, molecular prenatal diagnosis and preimplantation genetic testing are possible [9].

5.6. Particular features of acrofrontofacionasal dysostosis and isolated parietal foramina

In acrofrontofacionasal dysostosis, trio exome or genome sequencing is the diagnostic route, as in the study of all families with a molecular diagnosis [14, 19]. The PIGB founder variant lies in intron 6, 19 nucleotides from exon 7: it is only captured when the analysis includes the adjacent intronic regions, and demonstration of its effect required a minigene assay and study of patient RNA [14, 73]. In DEE80 due to the same gene, flow cytometry of lymphocytes and fibroblasts showed variable reduction of surface GPI-anchored proteins (CD16 and CD59, among others), down to about 25% to 50% of controls for some variants, and most patients tested had raised alkaline phosphatase; these functional data may support the interpretation of a PIGB variant of uncertain significance, but the sources consulted do not describe them in patients with AFFND1 [73, 74]. In the NBAS disorders there is no specific laboratory marker; NBAS immunohistochemistry on liver biopsy or Western blot in fibroblasts help to validate variants of uncertain significance, and the peripheral blood smear may show Pelger-Huët anomaly [71]. AFFND2 has no targeted molecular test; the diagnosis is clinical, with a normal karyotype in the cases described, and an exome negative for the known genes does not exclude it [15, 91].

In parietal foramina, the Oxford series found ALX4 or MSX2 variants in 11 of 13 familial cases and in only 1 of 6 sporadic cases; the difference reflects the benign natural history of carriers, who reproduce, and the higher frequency of atypical or complex presentations in sporadic cases [10]. The authors recommended direct sequencing, because the numerous ALX4 polymorphisms hamper heteroduplex screening methods, and MLPA dosage analysis in all negative cases; in the seven negative cases of the series, MLPA found no deletions, and they considered additional heterogeneity [10]. A carrier parent without a radiographic defect does not invalidate the diagnosis, because penetrance is age-dependent: in the infant described by Kahl and colleagues, the father had the same MSX2 variant without clinical signs of a skull defect [69].

6. Differential Diagnosis

6.1. Differential diagnosis of hypertelorism with a bifid nasal tip

Isolated hypertelorism is rare and almost always forms part of a chromosomal, monogenic or environmental syndrome; it may also be secondary to major facial clefts [76]. The table brings together the main differential diagnoses of FND and the entities with which it overlaps.

Craniofrontonasal syndrome is the main genetic differential diagnosis and is distinguished by coronal craniosynostosis, facial asymmetry and the nails; viewed from above, the nasal tip is broad with a shallow groove [78]. In isolated FND, the Oxford Desk Reference: Clinical Genetics and Genomics (Oxford Desk Reference) recommends carefully distinguishing sporadic forms from all the listed syndromes before giving a low recurrence risk [78]. Within the spectrum, ALX3 sequencing was negative in cases of Pai syndrome, OAFNS and AFND, and the phenotype of frontorhiny is narrow enough to guide testing [27]. On the other hand, a parent mosaic for ZSWIM6 may have a facies similar to frontorhiny, which justifies testing for mosaicism when ALX3 is negative [21]. The ZSWIM6 variant p.(Arg913Ter) causes severe intellectual disability without FND and should not be confused with AFND [67].

6.2. Differential diagnosis of parietal defects and of cranial ossification

Isolated parietal defects also require exclusion of mainly non-genetic causes of deficient ossification: meningoencephalocele, ventricular or arachnoid cyst, ectopic glial tissue, tumours, scalp defects, craniolacunae, osteoporosis, local inflammation, trauma and infections [9]. The multiple craniolacunae of a dysplastic vault should not be confused with parietal foramina [94]. A parietal scalp defect may be associated with foramina, but MSX2 is not a frequent cause of scalp defects [104].

7. Treatments and Risk Management Strategies

7.1. Principles and organisation of care

There is no pharmacological treatment or gene-directed therapy for any form of FND; treatment is surgical, rehabilitative and supportive, staged throughout growth [9, 22]. No society has published specific guidelines for FND, and GeneReviews states that its recommendations for parietal foramina are based on the authors' experience and a review of the literature [9]. The most complete general framework is the ACPA parameters of care, updated in 2024, which organise recommendations by health need rather than by specialty: the patient with a craniofacial anomaly should be followed by an interdisciplinary team (craniofacial and plastic surgery, neurosurgery, ophthalmology, otorhinolaryngology, genetics, orthodontics, speech and language therapy, psychology and social work, among others), with periodic assessments whose frequency and content the team defines, and teams with sufficient procedure volume and available intensive care have better outcomes [23]. In Europe, the relevant reference network is ERN CRANIO [24]. The Oxford Desk Reference recommends that a child with significant hypertelorism have a multidisciplinary assessment in a specialist craniofacial team, with careful ophthalmological assessment of strabismus [76].

7.2. Neonatal period and early childhood

The ACPA parameters provide, immediately after referral, for contact with the family for feeding support, education on airway maintenance techniques, assessment of the safety of oral feeding and intervention in infants with upper airway obstruction, which may require endoscopy, polysomnography and procedures ranging from adenoidectomy to skeletal surgery [23]. Airway assessment should precede any surgery, with an experienced paediatric anaesthetist, because these children are at risk of a difficult airway and difficult extubation [23]. In forms with microphthalmia, eyelid colobomas or lagophthalmos (FND3, frontofacionasal dysplasia), protection of the ocular surface and correction of the colobomas form part of the initial plan [22, 26]. CSF rhinorrhoea, meningitis or midline frontonasal swellings require CT and MRI to identify defects of the anterior skull base and dermoids with intracranial extension before any nasal intervention [27].

7.3. Craniofacial surgery by subtype

Surgical treatment of hypertelorism and nasal deformities follows two routes. Intracranial approaches — box osteotomy, which mobilises and approximates the orbits, and facial bipartition, which divides the midface and rotates it towards the midline — correct true orbital hypertelorism, but carry high morbidity [22, 105]. Fronto-orbital advancement corrects craniosynostosis and frontal retrusion [22]. For selected patients, especially those with less marked hypertelorism, a team in Ankara described in 2026 an extracranial nasal reconstruction with an “umbrella” graft [105]. The timing of surgery follows the ACPA principles: vault surgery depends on current or anticipated functional compromise, ophthalmological concerns, intracranial hypertension and family preferences; midface and mandibular surgery is deferred until the end of growth, unless there is compromise of the airway, masticatory function, speech or psychosocial adjustment; definitive septorhinoplasty waits for skeletal maturity and orthognathic surgery, when the latter is indicated [23].

These data come from retrospective single-centre series, without a comparison group or long follow-up, and the inclusion criterion relating to interorbital distance is not legible in the abstract consulted; they do not allow the technique to be compared with intracranial approaches [22, 105]. The series of Vargel and colleagues, also retrospective and single-centre, proposes a genotype-guided surgical roadmap for the ALX forms, but with only 13 genotyped patients [22]. The GeneGuidance curation database classifies this evidence as level 4 and records the absence of a guideline and of a ClinGen actionability assertion [106].

7.4. Dental, orthodontic and orthognathic rehabilitation

Facial clefts and midface deficiency give rise to malocclusion, maxillary constriction and dental anomalies. In a patient with complex FND (Tessier 0–14 cleft, mild hypertelorism, absent medial nasal process and frontonasal encephalocele) treated at the Hospital for Rehabilitation of Craniofacial Anomalies of the University of São Paulo, in Bauru, after the cranial and plastic surgery, orthodontic treatment began at 12 years of age with rapid maxillary expansion and a fixed appliance, bimaxillary advancement orthognathic surgery was performed at 21 years and treatment was completed with veneers on the upper incisors [107]. The ACPA recommends joint planning between surgeon and orthodontist (photographs, orthopantomography or cone-beam CT, cephalometry and virtual planning), speech assessment before and after maxillary advancement in patients with a repaired cleft palate and polysomnography when there is a risk of obstructive sleep apnoea [23]. The hypodontia described in the dominant ALX4 forms and the diastema and malocclusion of FND2 justify early dental assessment [37, 46].

7.5. Skull defects and parietal foramina

In enlarged parietal foramina, management is conservative. After diagnosis, GeneReviews recommends skull radiographs and 3D CT, investigation for seizures, brain imaging (CT or MRI) for meningeal, cortical and vascular malformations of the posterior fossa, examination of the scalp and genetic counselling [9]. Surgical closure has been advocated and performed, but its role is controversial and it is not routinely indicated, given the benign natural history, the reduction of the defects with age and the uncertainty about improvement of headaches; persistent cranium bifidum may justify closure [9]. The child, parents and teachers should be informed in order to avoid risk-taking behaviour, because the risk of penetrating brain injury is small but generates anxiety; contact sports are avoided while a midline bone defect persists [9]. Seizures are treated with antiepileptic drugs chosen by an experienced neurologist, none with efficacy demonstrated specifically in this disorder; headaches are treated symptomatically; and scalp defects are closed surgically, with a technique depending on size [9]. In FND2, the defects are much larger and may be associated with meningoceles that enlarge with age; in the child described by Goswami and colleagues, repair of the meningoencephaloceles was complicated by CSF leak, a lumboperitoneal shunt with overdrainage, external ventricular drainage, meningitis and death, which shows the risk of these interventions when the dural and venous anatomy is abnormal [46].

In a series of four infants with large, persistent foramina, two treated with autologous calvarial bone graft and two with metal mesh and hydroxyapatite, there were no perioperative complications or infections over a mean follow-up of 36 months, but one graft failed late and required repeat cranioplasty with methyl methacrylate; the authors advocate cranioplasty in active young children at risk of trauma [82]. The Oxford group, whose two families with the ALX4 variant p.(Arg218Gln) and large persistent defects were managed conservatively without significant problems, advises weighing the surgical complications, especially graft failure and infection, against the individual benefit; in the same series, one child had the foramina covered with titanium mesh at 5 years of age and a father and son underwent surgery with bone grafting, the son before 1 year of age [10]. Surgery may be justified in symptomatic patients, those with very large and persistent defects, with encephalocele and intracranial hypertension, or at increased risk of injury owing to impulsive behaviour or epilepsy [69, 82]. In the infant described by Kahl and colleagues, clinical, ultrasound and CT follow-up over 20 months made any intervention unnecessary [69].

7.6. Pharmacological treatment, regulatory status and trials

No medicinal product has an approved indication for any form of FND or for parietal foramina. The drugs used are symptomatic and in common use (antiepileptic drugs, analgesics, antibiotics for meningitis), without a disease-specific dose, and follow the general standards of the specialty [9].

The absence of registered trials results from a search of public databases and does not prove the absence of studies in other registers.

7.7. Developments in the last 12 months and research

In vitro recovery of neural crest cell migration with BMP2 and BMP9 antagonists in cells from patients with an ALX1 variant is the only finding close to a targeted intervention, and it is exclusively preclinical [65].

7.8. Special situations

When prenatal ultrasound identifies a large skull defect, delivery should be planned, reviewing the indication for scalp electrodes, forceps and vacuum extraction; elective caesarean section may reduce the theoretical risk of traumatic injury [9]. In the perioperative period, airway assessment, paediatric anaesthetic experience and the availability of intensive care are prerequisites for craniotomies and procedures that compromise the airway [23]. The ACPA recommends psychosocial screening from early childhood to young adulthood, with specific questionnaires (CLEFT-Q, Psychosocial Assessment Tool — craniofacial version) and general ones (Ages and Stages Questionnaire, PedsQL, PROMIS, Strengths and Difficulties Questionnaire), and standardised neurodevelopmental assessment when screening is positive [23]. At transition to adult care, needs related to scars, nasal deformity, malocclusion, missing teeth, relapses, hearing and speech persist; prosthetic dental rehabilitation is carried out at the end of growth; adults may want genetic counselling in their own right; and teams should continue to coordinate care, ideally with a person responsible for transition [23].

7.9. Recommendations and divergences between sources

Sources: [9, 20, 23, 24, 76, 94, 105].

7.10. Acrofrontofacionasal dysostosis: supportive care

There is no targeted treatment and there are no registered trials for any of the three types; a ClinicalTrials.gov search for “acrofrontofacionasal” and for PIGB returned no studies, and the search for NBAS only returned records unrelated to the gene [109]. Treatment follows the general parameters for craniofacial anomalies, with an interdisciplinary team: repair of the cleft lip and palate and speech, feeding and dental rehabilitation; orthopaedic correction of clubfoot, digital contractures, hip dislocation and scoliosis; ophthalmological assessment of colobomas, cataract and corneal lesions; auditory rehabilitation; and neurodevelopmental and behavioural support [12, 18, 23]. The patient with AFFND1 followed up to 55 years of age shows that hearing loss may arise only in adulthood, which justifies lifelong audiometry; the 26-year-old patient, without effective chewing or swallowing, illustrates the need for nutritional and swallowing assessment [12].

In type 3, no hepatic crises were reported in the family described, but biallelic NBAS variants cause recurrent acute liver failure triggered by fever, which may be fatal [71]. In six patients with NBAS deficiency, early and consistent antipyretic treatment and induction of anabolism with intravenous glucose and parenteral lipids attenuated and shortened the crises, especially when started early, and effective antipyretic treatment may prevent them; an emergency card, letter or bracelet with instructions and contacts is recommended [71]. In hypogammaglobulinaemia, immunoglobulin replacement should be considered to reduce the infections that trigger the crises; one patient transplanted at 3 years of age had no further crises; and cervical instability due to C1 and C2 hypoplasia, described in two patients, may require surgical fixation [71]. These measures derive from the hepatic and multisystem NBAS forms and are applied to AFFND3 by extrapolation; the source describes the components of the emergency response (early antipyretic, intravenous glucose, parenteral lipids and, in hypogammaglobulinaemia, immunoglobulin) without specifying doses [71]. In the series of 14 patients of Staufner et al., from which the emergency protocol derives, high-dose intravenous glucose (10–12 g/kg/day in infants) ameliorated the crisis, especially when started early; parenteral lipids (1 g/kg/day, soybean-oil emulsion), tested in six patients, accelerated recovery; five patients received carnitine (100 mg/kg/day) during crises; and early, effective fever control often prevented the liver crisis, with no antipyretic dose specified in the text of the article [110]. In type 1, seizures are not part of the described phenotype; if they occur, the approach to the GPI deficiency encephalopathies is followed, including the experience with pyridoxine.

8. Reproductive Options and Family Counselling

8.1. Recurrence risks by scenario

The first step is to identify the entity, because the risk varies from almost nil to 50%. The Oxford Desk Reference recommends examining the parents and recording their ocular measurements before counselling an apparently sporadic case [76].

In the family of Uz and colleagues, after the first child with FND3, the next two pregnancies were affected [31]. In the family of Goswami and colleagues, a previous pregnancy had a fetus with a large occipital encephalocele and intrauterine death at about 5.5 months, presumed affected [46].

In acrofrontofacionasal dysostosis with identified variants, parental study confirms phase, carrier testing is extended to relatives of reproductive age and prenatal or preimplantation diagnosis becomes possible; in type 2, without a gene, counselling is based on the empirical recessive risk and on the anomaly scan, with no prenatal markers described [14, 15].

8.2. Particular features of ALX4

Heterozygotes for ALX4 variants are not silent carriers: they may have parietal foramina (with high but reduced and age-dependent penetrance, because the defects decrease with growth) and, in some families, mild signs of FND [9, 59]. The parents of a child with FND2 should therefore have skull radiography or CT in addition to molecular testing; in several families, one or both parents had parietal foramina [4, 46, 56, 58]. Conversely, when parietal foramina occur in a consanguineous or endogamous family, the risk of severe malformations due to homozygosity for ALX4 (and, theoretically, MSX2) should be considered in a couple in which both are heterozygous [9]. A child with parietal foramina and clinically normal parents of unknown genetic status remains at increased risk for siblings owing to reduced penetrance and gonadal mosaicism; confirmation of de novo status requires parental testing [9]. The intrafamilial variability of the dominant forms (from subtle nasal changes to mild FND2) complicates prediction of the phenotype in future pregnancies [36, 58].

8.3. Mosaicism in AFND

AFND is almost always de novo, but a mildly affected or even apparently normal parent may be mosaic [21, 32]. Because the mutant fraction varies between tissues and Sanger sequencing in blood may be negative, study of the parent should use deep sequencing of several tissues; in the mother studied by Twigg and colleagues, the fraction was highest in buccal mucosa [21]. One of the four patients in the original series had somatic mosaicism (60:40 ratio between normal and mutant allele) [32].

8.4. Prenatal and preimplantation diagnosis

With the familial variant identified, prenatal diagnosis by chorionic villus sampling or amniocentesis and preimplantation genetic testing are possible; in parietal foramina, requests are rare given the benign prognosis [9]. A detailed anomaly scan (18–20 weeks) and fetal MRI detect the larger skull defects and severe facial malformations, but mild hypertelorism and subtle facial forms escape ultrasound [9, 76]. Perspectives on the use of prenatal diagnosis vary between professionals and families, and the discussion should preferably take place before pregnancy [9]. The ACPA parameters include, within genetic services, information on prenatal testing and reproductive options for future pregnancies and discussion of the ethical and legal implications of the results [23].

8.5. Other aspects of counselling

Counselling should include factual information, relief of guilt, correction of misconceptions, support for decision-making and the process of adaptation, and identification of support groups [23]. Affected or at-risk young adults should receive genetic counselling in their own right on risks to offspring and reproductive options [9, 23]. When no molecular diagnosis is obtained, DNA storage for future reanalysis should be offered [9].

9. Special Considerations: Population Diversity and Epidemiology

9.1. What each figure measures

The EUROCAT birth prevalence refers to FND as a morphological group and not to each genetic form; GeneGuidance stresses that the frequency of frontorhiny is much lower than this figure, because the group includes the ALX1, ALX4 and SIX2 forms and the unexplained ones [106, 111]. Orphanet case counts are numbers of publications and not population rates. The proportions in the series of Vargel, Lehalle and Twigg derive from samples from referral centres, biased towards severe or atypical forms, and do not estimate the aetiological distribution in the population. The surveys underlying the prevalence of parietal foramina predate molecular diagnosis and CT [9, 106].

9.2. Founder populations and consanguinity

The recessive ALX forms have been described mainly in consanguineous families or families of restricted geographical origin: frontorhiny in families from Morocco (parents from adjacent valleys), Algeria, Ireland, the Netherlands (two families without known consanguinity, of Dutch ancestry), Turkey, India and Pakistan [27, 53]; FND2 in Turkey, Iran, Egypt, Pakistan and Bangladesh [30, 46, 57–59]; FND3 in Turkey (neighbouring villages in family 2) and Pakistan [31, 64]. The ALX4 variant c.793C>T p.(Arg265Ter) has been found in the homozygous state in Turkey and Bangladesh [30, 46]. In the United States, frontorhiny was described in two siblings with the ALX3 variant p.Tyr214Ter treated at the University of California, Davis, and the SIX2 deletion in a family assessed in Cincinnati [33, 55]. In Europe, the Oxford series on twins (5 discordant monozygotic pairs over two decades) and the multicentre cohort of French and Turkish authors on OAFNS are the largest case series [41, 85]. Homozygosity mapping was decisive in identifying rare recessive genes from few samples from consanguineous families [27].

Outside the ALX pathway, the intronic PIGB variant c.795-19T>G explains AFFND1 in two Brazilian families with no known relationship, and the authors consider a founder effect in this population [14]; the NBAS variant c.5741G>A p.(Arg1914His) is a founder variant in the Yakuts of eastern Siberia, where it accounts for 33 cases of SOPH syndrome without a hepatic phenotype [71]. AFFND2 has been described in consanguineous families from Brazil, Kuwait and Tunisia [15, 17, 91]. In parietal foramina, a partially mummified skeleton of a young woman of the Chinchorro culture, in northern Chile, about 4000 years old, shows that the anomaly existed in ancient Andean populations; the authors consider a contribution from endogamy [112].

9.3. Portugal and Brazil

In Brazil, the sources consulted include a São Paulo family with dominant FND due to heterozygous ALX4 (Instituto da Criança, Faculty of Medicine of the University of São Paulo) [35], a patient with FND2 and meningoencephalocele studied at the Federal University of São Paulo [60], the consanguineous Brazilian families with acrofrontofacionasal dysostosis type 1, later explained by PIGB variants [12], the series of 10 boys with FND, severe neuropsychological impairment and midline central nervous system anomalies [3], the exclusion of ALX3/ALX4 variants in the frontonasal dysgenesis syndrome with callosal agenesis and basal encephalocele studied at the Hospital for Rehabilitation of Craniofacial Anomalies (HRAC-USP, Bauru) [89] and the multidisciplinary rehabilitation of a complex FND at the same institution [107]. Outside the clinical setting, a forensic anthropological examination in Goiás described bilateral parietal foramina of about 12 mm in an adult skull [84]. In AFFND1, the PIGB variant of the Bauru families was not present in the Brazilian polymorphism databases BIPMed and ABraOM [14]. The sources consulted contain no Portuguese data on FND; this absence does not demonstrate that there are no published cases.

9.4. Reference network and patient associations

In the European Union, ERN CRANIO brings together the reference centres for rare craniofacial anomalies [24]. Patient associations include Headlines Craniofacial Support (United Kingdom, https://www.headlines.org.uk), the Children's Craniofacial Association (United States, https://ccakids.org) and FACES: The National Craniofacial Association (United States, https://www.faces-cranio.org), in addition to the general organisations Contact a Family (United Kingdom) and the National Organization for Rare Disorders (NORD, United States) [9, 24, 76].

10. Illustrative Clinical Case

10.1. Frontorhiny with recurrent meningitis (Twigg and colleagues, 2009)

Patient 1 of the series that defined frontorhiny is a boy from family 1, born to Moroccan parents without known consanguinity, originating from adjacent valleys, with a sister who was also affected [27]. The photograph at 2 days of age (figure 1A of the article) shows the facies of frontorhiny: hypertelorism, a broad nasal bridge, a short nasal dorsum, a bifid tip, a broad columella, widely separated slit-like nostrils and a long philtrum with prominent ridges [27]. He also had a cleft palate and a bone defect of the anterior cranial fossa; coronal CT at 6 years of age (figure 1B) showed a widened, dysmorphic ethmoid and apparent continuities between the nasal cavity and the brain [27]. He had six episodes of pneumococcal meningitis associated with CSF rhinorrhoea, explained by multiple defects of the cribriform plate of the ethmoid [27]. Homozygosity mapping with his sister and an Algerian patient identified a 653 kb interval at 1p13.3, with a maximum heterogeneity LOD score (HLOD) of 4.58, and sequencing revealed in both siblings the homozygous ALX3 variant c.595-2A>T (authors' nomenclature, NM_006492.2), which alters the splicing acceptor site of intron 2 and predicts loss of the DNA-binding helix III [27]. Treatment included palatoplasty, Le Fort I advancement, Le Fort III advancement with medialisation of the orbits, repair of the anterior skull base defect and reconstruction of the nasal airway; growth and development were normal [27]. The figures were consulted in the open-access article: https://pmc.ncbi.nlm.nih.gov/articles/PMC2681074/figure/fig1/ [27].

The case shows that an apparently “isolated” FND may conceal a skull base defect with a risk of meningitis, and that frontorhiny, being recessive, carries a 25% risk for siblings even without known consanguinity when the parents come from a restricted geographical area. The sequence of operations described reflects the decisions of one centre and not a standardised protocol.

10.2. Severe FND2 due to ALX4 (Goswami and colleagues, 2024)

A girl from Bangladesh, the daughter of consanguineous parents, was seen at 8 years of age for two soft parietal swellings with a bony gap, present since birth as small reddish areas of the scalp, which had enlarged and had been draining CSF intermittently for 8 months [46]. She had absent hair, absent eyebrows, sparse eyelashes, preaxial polydactyly of the right foot, hypertelorism, telecanthus, blepharophimosis, strabismus, nystagmus, a bifid depressed nasal tip, notched nasal alae, a short broad columella and a broad philtrum; 3D CT showed brachycephaly, cranium bifidum, craniosynostosis, a frontal bony gap, wormian bones and a trifoliate nasal bone; MRI revealed biparietal meningoencephalocele, dysgyria, dysplastic occipital lobes, an angulated corpus callosum, a parietal porencephalic cyst and cerebellar atrophy; venography showed absence of the posterior part of the superior sagittal sinus, the straight sinus and the right transverse sinus; the ovaries were hypoplastic and polycystic; developmental milestones were delayed, but intellect was appropriate for age [46]. Trio genome sequencing revealed the homozygous ALX4 variant NM_021926.4:c.793C>T p.(Arg265Ter), with heterozygous parents; the father's CT showed bilateral enlarged parietal foramina and the mother's normal foramina [46, 54]. The meningoencephaloceles were repaired in the same operation; a CSF leak led to a lumboperitoneal shunt, later revised for overdrainage with sunken flap syndrome and removed; recurrence of the leaks was treated with external ventricular drainage, complicated by meningitis, seizures and coma, and the child died while a ventriculoperitoneal shunt was being planned [46]. Figures 1 and 2 of the article document the facies, the polydactyly, the parents' CT and the MRI [46].

The case illustrates the severe extreme of the ALX4 spectrum, the usefulness of cranial imaging of the parents and the complexity of CSF management when the dural and venous anatomy is abnormal. The succession of shunts described should not be read as recommended practice, and the authors worked with resources limited by financial constraints and by the COVID-19 pandemic.

10.3. Parietal foramina due to MSX2 in an infant (Kahl and colleagues, 2024)

A girl born at 39 weeks, after a normal pregnancy and a vacuum-assisted delivery with an occipital cephalohaematoma, was seen in Germany for a skull malformation and facial asymmetry; no signs of intracerebral haemorrhage or cerebral dysplasia were recorded [69]. At 4 weeks, at a routine paediatric examination, bilateral symmetrical parietal openings near the lambda were documented, 18 mm on the right and 14 mm on the left, with palpable focal protrusion of parenchyma covered only by the scalp; the first hypothesis was bilateral diastasis of the lambdoid suture with encephalocele [69]. The parents and siblings had no clinical signs of a skull defect, and there were no neurological or bone disorders in the family [69]. At 17 weeks, ultrasound showed enlargement of the defects (19 × 18 mm on the left and 25 mm on the right), but the child remained well and the protrusions were diminishing; at 6 months, three-dimensional CT showed that the defects, measuring 32 and 28 mm, communicated with the sagittal suture, and the coronal, lambdoid and sagittal sutures were open [69]. Molecular testing revealed a heterozygous MSX2 variant published as c.515G>C p.R172H; the protein change corresponds to NM_002449.5:c.515G>A p.(Arg172His), the variant already described by Wilkie and colleagues and classified in ClinVar as likely pathogenic, whereas c.515G>C would predict p.(Arg172Pro) [28, 54, 69, 70]. The father, clinically without a defect, had the same variant [69]. Over the following 20 months, head circumference moved from the 3rd–10th to the 10th–25th centiles, ossification progressed to symmetrical defects of 18 mm at 2.5 years of age, development remained appropriate for age and no intervention was required; regular paediatric surveillance was continued [69].

The case shows the typical presentation of a large defect in an infant that decreases with growth, the risk of confusing the foramina with encephalocele, the value of molecular testing in avoiding unnecessary surgery and the incomplete, age-dependent penetrance in a carrier parent without clinical signs. The conservative approach coincides with the GeneReviews recommendation [9].

10.4. Acrofrontofacionasal dysostosis type 1 followed into adulthood (Palagano and colleagues, 2021)

The patient described at 26 years of age by Richieri-Costa and colleagues in 1992, from a Brazilian family whose affected brother died at 1 year of age, was reassessed at 55 years in the study that identified the homozygous PIGB variant NM_004855.5:c.795-19T>G [14, 90]. She had brachycephaly, a long face, a broad nasal base, hypertelorism, distal digital hypoplasia with hypoplastic nails, genu valgum, tibial hypoplasia and clubfoot, intellectual disability with language delay and learning difficulties, and bilateral blindness from scarred corneal ulcers, bilateral polar cataracts and colobomas of the iris, retina and optic nerve [12]. Hearing loss began in the fifth decade and, at 52 years of age, there was complete deafness on the right and severe sensorineural hearing loss on the left; in the sixth decade she had breast carcinoma, treated with mastectomy and chemotherapy, without an established relationship with the underlying disease [12]. The case documents survival into the sixth decade and progressive sensory loss in adulthood, data absent from the initial paediatric descriptions.

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

Prognosis depends on the form. In frontorhiny, growth and cognitive development were normal in all 11 patients of the original series and in the 4 Pakistani patients; morbidity results from the nasal deformity, the multiple operations, the dermoids and, in one patient, recurrent meningitis [3, 27]. In FND2, cognition ranges from normal to intellectual disability, and extensive skull defects may be complicated by progressive meningoceles, CSF leaks and infection [2, 46, 58]. In FND3, one patient had mild intellectual disability at 11 years of age and another died in infancy from a pulmonary infection; the form with an acceptor splicing variant had normal development [5, 31]. In AFND, intellectual disability is severe, with children who neither walk nor talk at 7–8 years of age [21]. In the FND sequence in general, most individuals have normal intelligence (DeMyer's series: 8% with severe impairment and 12% with impairment among 33 patients) [26]. Parietal foramina have a benign course, with reduction of the defects with age [9]. Prognostic data come from case reports and small series, without longitudinal cohorts, and describe surgically treated patients.

11.2. Follow-up

There are no published surveillance protocols specifically for FND1, FND2, FND3, AFND or acrofrontofacionasal dysostosis; the intervals derive from the general parameters of craniofacial care and from the GeneReviews entry on parietal foramina [9, 23, 106].

12. Summary and Final Considerations

Frontonasal dysplasia is a malformation pattern defined clinically by at least two of seven Sedano and Gorlin criteria — true hypertelorism, a broad nasal root, a median facial cleft, clefting of the nasal alae, lack of a nasal tip, anterior cranium bifidum occultum and a widow's peak —, with a birth prevalence in Europe of about 0.7 per 100,000. Most isolated cases are sporadic, frequently associated with discordant monozygotic twinning, and have no molecular diagnosis. A minority correspond to recognisable monogenic forms: frontorhiny (FND1) due to biallelic loss of ALX3, with a narrow spectrum and normal cognition, marked by a long philtrum with prominent ridges and a broad columella; FND2 due to biallelic loss of ALX4, with large skull defects, alopecia and genital anomalies; FND3 due to biallelic loss of ALX1, with extreme microphthalmia and facial clefts; and acromelic frontonasal dysostosis due to a single gain-of-function variant of ZSWIM6, with brain and limb malformations and severe intellectual disability. ALX4 has two inheritance patterns: haploinsufficiency causes dominant, benign enlarged parietal foramina, and certain heterozygous variants with a dominant-negative or gain-of-function effect cause mild FND with hypotrichosis; heterozygotes in families with FND2 should be studied by cranial imaging. Diagnosis rests on the dysmorphological examination with ocular measurements, on CT and MRI and on exome or genome sequencing with copy number analysis, bearing in mind that homozygous deletions and low-level mosaicism escape conventional methods. Treatment is surgical and staged within an interdisciplinary craniofacial team — correction of hypertelorism by box osteotomy or facial bipartition, nasal reconstruction, correction of clefts and colobomas, orthodontics and orthognathic surgery at the end of growth —, with attention to the airway, to skull base defects that predispose to meningitis and to psychosocial support up to the transition to adult care. There are no approved medicinal products or clinical trials; recent developments are structural, embryological and classificatory, and the inclusion of the ALX genes and ZSWIM6 in ClinGen curation should clarify the validity of each gene–disease relationship.

The class also extends to two families of entities with a different mechanism. Enlarged parietal foramina, due to haploinsufficiency of ALX4 or MSX2, are as a rule benign, decrease with age and are managed conservatively; MSX2 shows the inverse dosage effect in Boston-type craniosynostosis due to gain of function. Acrofrontofacionasal dysostosis adds limb anomalies and severe intellectual disability to FND; type 1 is a null defect of GPI anchor biosynthesis due to an intronic founder variant of PIGB in Brazil, type 3 a disorder of retrograde trafficking due to NBAS, with the need to prevent the febrile hepatic crises characteristic of this gene, and type 2 remains without a gene.

13. References

1. Orphanet. Frontonasal dysplasia (ORPHA:250). https://www.orpha.net/en/disease/detail/250
2. Farlie PG, Baker NL, Yap P, Tan TY. Frontonasal dysplasia: towards an understanding of molecular and developmental aetiology. Mol Syndromol. 2016;7(6):312-21. doi:10.1159/000450533
3. OMIM. #136760 Frontonasal dysplasia 1; FND1. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/136760 (accessed 2026-10-29).
4. OMIM. #613451 Frontonasal dysplasia 2; FND2. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/613451 (accessed 2026-10-29).
5. OMIM. #613456 Frontonasal dysplasia 3; FND3. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/613456 (accessed 2026-10-29).
6. OMIM. #603671 Acromelic frontonasal dysostosis; AFND. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/603671 (accessed 2026-10-29).
7. OMIM. #609597 Parietal foramina 2; PFM2. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/609597 (accessed 2026-10-29).
8. OMIM. #168500 Parietal foramina 1; PFM1. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/168500 (accessed 2026-10-29).
9. Mavrogiannis LA, Wilkie AOM. Enlarged Parietal Foramina. 2004 Mar 30 [Updated 2025 Jun 26]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1128/
10. Mavrogiannis LA, Taylor IB, Davies SJ, Ramos FJ, Olivares JL, Wilkie AO. Enlarged parietal foramina caused by mutations in the homeobox genes ALX4 and MSX2: from genotype to phenotype. Eur J Hum Genet. 2006;14(2):151-8. doi:10.1038/sj.ejhg.5201526
11. OMIM. #601224 Potocki-Shaffer syndrome. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/601224 (accessed 2026-10-29).
12. OMIM. #201180 Acrofrontofacionasal dysostosis 1; AFFND1. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/201180 (accessed 2026-10-29).
13. Orphanet. Acrofrontofacionasal dysostosis (ORPHA:1784). https://www.orpha.net/en/disease/detail/1784
14. Palagano E, Gordon CT, Uva P, Strina D, Dimartino C, Villa A, et al. A novel intronic variant in PIGB in Acrofrontofacionasal dysostosis type 1 patients expands the spectrum of phenotypes associated with GPI biosynthesis defects. Bone. 2021;153:116152. doi:10.1016/j.bone.2021.116152
15. OMIM. 239710 Acrofrontofacionasal dysostosis 2. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/239710 (accessed 2026-11-01).
16. Orphanet. Hypertelorism-hypospadias-polysyndactyly syndrome (ORPHA:2211). https://www.orpha.net/en/disease/detail/2211
17. Richieri-Costa A, Montagnoli L, Kamiya TY. Autosomal recessive acro-fronto-facio-nasal dysostosis associated with genitourinary anomalies. Am J Med Genet. 1989;33(1):121-4. doi:10.1002/ajmg.1320330118
18. OMIM. #621629 Acrofrontofacionasal dysostosis 3; AFFND3. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/621629 (accessed 2026-11-01).
19. Palagano E, Zuccarini G, Prontera P, Borgatti R, Stangoni G, Elisei S, et al. Mutations in the Neuroblastoma Amplified Sequence gene in a family affected by Acrofrontofacionasal Dysostosis type 1. Bone. 2018;114:125-36. doi:10.1016/j.bone.2018.06.013
20. Manickam K, McClain MR, Demmer LA, Biswas S, Kearney HM, Malinowski J, et al. Exome and genome sequencing for pediatric patients with congenital anomalies or intellectual disability: an evidence-based clinical guideline of the American College of Medical Genetics and Genomics (ACMG). Genet Med. 2021;23(11):2029-37. doi:10.1038/s41436-021-01242-6
21. Twigg SR, Ousager LB, Miller KA, Zhou Y, Elalaoui SC, Sefiani A, et al. Acromelic frontonasal dysostosis and ZSWIM6 mutation: phenotypic spectrum and mosaicism. Clin Genet. 2016;90(3):270-5. doi:10.1111/cge.12721
22. Vargel I, Canter HI, Kucukguven A, Aydin A, Ozgur F. ALX-related frontonasal dysplasias: clinical characteristics and surgical management. Cleft Palate Craniofac J. 2022;59(5):637-43. doi:10.1177/10556656211019621
23. American Cleft Palate-Craniofacial Association. Parameters of care for evaluation and treatment of individuals with cleft lip/palate and/or other craniofacial conditions. 2024 update. Chapel Hill (NC): ACPA; 2024. Available from: https://acpacares.org/wp-content/uploads/2025/02/2024-ACPA_ParametersOfCare_Final.pdf
24. European Reference Network for rare and/or complex craniofacial anomalies and ear, nose and throat disorders (ERN CRANIO). https://www.ern-cranio.eu (accessed 2026-10-29).
25. Edoh EYA, Mighton C, Broeren EC, Gitau VN, Ratliff J, DiStefano MT, et al. Evidence-based classification of genes implicated in craniosynostosis disorders using the ClinGen curation framework. Genet Med. 2026;28(9):102587. doi:10.1016/j.gim.2026.102587
26. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 342–345.
27. Twigg SR, Versnel SL, Nürnberg G, Lees MM, Bhat M, Hammond P, et al. Frontorhiny, a distinctive presentation of frontonasal dysplasia caused by recessive mutations in the ALX3 homeobox gene. Am J Hum Genet. 2009;84(5):698-705. doi:10.1016/j.ajhg.2009.04.009
28. Wilkie AO, Tang Z, Elanko N, Walsh S, Twigg SR, Hurst JA, et al. Functional haploinsufficiency of the human homeobox gene MSX2 causes defects in skull ossification. Nat Genet. 2000;24(4):387-90. doi:10.1038/74224
29. Mavrogiannis LA, Antonopoulou I, Baxová A, Kutílek S, Kim CA, Sugayama SM, et al. Haploinsufficiency of the human homeobox gene ALX4 causes skull ossification defects. Nat Genet. 2001;27(1):17-8. doi:10.1038/83703
30. Kayserili H, Uz E, Niessen C, Vargel I, Alanay Y, Tuncbilek G, et al. ALX4 dysfunction disrupts craniofacial and epidermal development. Hum Mol Genet. 2009;18(22):4357-66. doi:10.1093/hmg/ddp391
31. Uz E, Alanay Y, Aktas D, Vargel I, Gucer S, Tuncbilek G, et al. Disruption of ALX1 causes extreme microphthalmia and severe facial clefting: expanding the spectrum of autosomal-recessive ALX-related frontonasal dysplasia. Am J Hum Genet. 2010;86(5):789-96. doi:10.1016/j.ajhg.2010.04.002
32. Smith JD, Hing AV, Clarke CM, Johnson NM, Perez FA, Park SS, et al. Exome sequencing identifies a recurrent de novo ZSWIM6 mutation associated with acromelic frontonasal dysostosis. Am J Hum Genet. 2014;95(2):235-40. doi:10.1016/j.ajhg.2014.07.008
33. Hufnagel RB, Zimmerman SL, Krueger LA, Bender PL, Ahmed ZM, Saal HM. A new frontonasal dysplasia syndrome associated with deletion of the SIX2 gene. Am J Med Genet A. 2016;170A(2):487-91. doi:10.1002/ajmg.a.37441
34. Kim S, Twigg SRF, Scanlon VA, Chandra A, Hansen TJ, Alsubait A, et al. Localized TWIST1 and TWIST2 basic domain substitutions cause four distinct human diseases that can be modeled in Caenorhabditis elegans. Hum Mol Genet. 2017;26(11):2118-32. doi:10.1093/hmg/ddx107
35. Bertola DR, Rodrigues MG, Quaio CR, Kim CA, Passos-Bueno MR. Vertical transmission of a frontonasal phenotype caused by a novel ALX4 mutation. Am J Med Genet A. 2013;161A(3):600-4. doi:10.1002/ajmg.a.35762
36. Altunoglu U, Satkın B, Uyguner ZO, Kayserili H. Mild nasal clefting may be predictive for ALX4 heterozygotes. Am J Med Genet A. 2014;164A(8):2054-8. doi:10.1002/ajmg.a.36578
37. Peled A, Sarig O, Mohamad J, Eskin-Schwartz M, Vodo D, Bochner R, et al. Dominant frontonasal dysplasia with ectodermal defects results from increased activity of ALX4. Am J Med Genet A. 2023;191(12):2806-12. doi:10.1002/ajmg.a.63408
38. Cesarato N, Kumar S, Wehner M, Aldisi R, Thiele H, Hamm H, et al. From hypotrichosis to frontonasal dysplasia: expanding the phenotypic spectrum of ALX4 variants. Am J Med Genet A. 2026 Jul 24 [Epub ahead of print]. doi:10.1002/ajmg.a.70249
39. Zhang F, Li F, Song D, Wang S. A novel heterozygous ALX4 frameshift variant causing congenital diffuse hypotrichosis and frontonasal dysmorphism. J Dermatol. 2026 Sep 30 [Epub ahead of print]. doi:10.1111/1346-8138.70522
40. Richieri-Costa A, Colletto GM, Gollop TR, Masiero D. A previously undescribed autosomal recessive multiple congenital anomalies/mental retardation (MCA/MR) syndrome with fronto-nasal dysostosis, cleft lip/palate, limb hypoplasia, and postaxial poly-syndactyly: acro-fronto-facio-nasal dysostosis syndrome. Am J Med Genet. 1985;20(4):631-8. doi:10.1002/ajmg.1320200409
41. Mohammed SN, Swan MC, Wall SA, Wilkie AO. Monozygotic twins discordant for frontonasal malformation. Am J Med Genet A. 2004;130A(4):384-8. doi:10.1002/ajmg.a.30280
42. Orphanet. Frontorhiny (ORPHA:391474). https://www.orpha.net/en/disease/detail/391474
43. Orphanet. Frontonasal dysplasia-alopecia-genital anomalies syndrome (ORPHA:228390). https://www.orpha.net/en/disease/detail/228390
44. Orphanet. Frontonasal dysplasia-severe microphthalmia-severe facial clefting syndrome (ORPHA:306542). https://www.orpha.net/en/disease/detail/306542
45. Orphanet. Acromelic frontonasal dysplasia (ORPHA:1827). https://www.orpha.net/en/disease/detail/1827
46. Goswami B, Rahman A, Jahan I, Akter S, Banu TA, Osman E, et al. A rare homozygous ALX4 mutation in a Bangladeshi girl with frontonasal dysplasia type-2 (FND2). Heliyon. 2024;10(15):e34929. doi:10.1016/j.heliyon.2024.e34929
47. OMIM. *601527 Aristaless-like homeobox 1; ALX1. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/601527 (accessed 2026-10-29).
48. OMIM. *606014 Aristaless-like homeobox 3; ALX3. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/606014 (accessed 2026-10-29).
49. Iyyanar PPR, Wu Z, Lan Y, Hu YC, Jiang R. Alx1 deficient mice recapitulate craniofacial phenotype and reveal developmental basis of ALX1-related frontonasal dysplasia. Front Cell Dev Biol. 2022;10:777887. doi:10.3389/fcell.2022.777887
50. Mumme-Monheit A, Mitchell JM, Bailon-Zambrano R, Wright N, Neukirch LA, Keating MK, et al. The alx gene family confers segmental identity to frontonasal cranial neural crest cells. Nat Commun. 2026;17(1):7713. doi:10.1038/s41467-026-74434-w
51. OMIM. *605420 Aristaless homeobox 4; ALX4. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/605420 (accessed 2026-10-29).
52. Cain B, Yuan Z, Thoman E, Kovall RA, Gebelein B. The ALX4 dimer structure provides insight into how disease alleles impact function. Nat Commun. 2025;16(1):4800. doi:10.1038/s41467-025-59728-9
53. Ullah A, Umair M, E-Kalsoom U, Shahzad S, Basit S, Ahmad W. Exome sequencing revealed a novel nonsense variant in ALX3 gene underlying frontorhiny. J Hum Genet. 2018;63(1):97-100. doi:10.1038/s10038-017-0358-y
54. 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. Accessed (GRCh38) on 2026-10-29 and 2026-11-01.
55. Pham NS, Rafii A, Liu J, Boyadjiev SA, Tollefson TT. Clinical and genetic characterization of frontorhiny: report of 3 novel cases and discussion of the surgical management. Arch Facial Plast Surg. 2011;13(6):415-20. doi:10.1001/archfacial.2011.684
56. Kayserili H, Altunoglu U, Ozgur H, Basaran S, Uyguner ZO. Mild nasal malformations and parietal foramina caused by homozygous ALX4 mutations. Am J Med Genet A. 2012;158A(1):236-44. doi:10.1002/ajmg.a.34390
57. Kariminejad A, Bozorgmehr B, Alizadeh H, Ghaderi-Sohi S, Toksoy G, Uyguner ZO, et al. Skull defects, alopecia, hypertelorism, and notched alae nasi caused by homozygous ALX4 gene mutation. Am J Med Genet A. 2014;164A(5):1322-7. doi:10.1002/ajmg.a.36008
58. El-Ruby M, El-Din Fayez A, El-Dessouky SH, Aglan MS, Mazen I, Ismail N, et al. Identification of a novel homozygous ALX4 mutation in two unrelated patients with frontonasal dysplasia type-2. Am J Med Genet A. 2018;176(5):1190-4. doi:10.1002/ajmg.a.38655
59. Hussain S, Umm-E-Kalsoom, Ullah I, Liaqat K, Nawaz S, Ahmad W. A novel missense variant in the ALX4 gene underlies mild to severe frontonasal dysplasia in a consanguineous family. Genet Test Mol Biomarkers. 2020;24(4):217-23. doi:10.1089/gtmb.2019.0203
60. Meloni VA, Moysés-Oliveira M, Melo MC, Caneloi TP, Dantas AG, Soares MF, et al. Novel homozygous ALX4 mutation causing frontonasal dysplasia-2 in a patient with meningoencephalocele. Clin Genet. 2015;88(6):593-6. doi:10.1111/cge.12595
61. OMIM. #615529 Craniosynostosis 5, susceptibility to; CRS5. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/615529 (accessed 2026-10-29).
62. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 570.
63. Ferrarini A, Gaillard M, Guerry F, Ramelli G, Heidi F, Keddache CV, et al. Potocki-Shaffer deletion encompassing ALX4 in a patient with frontonasal dysplasia phenotype. Am J Med Genet A. 2014;164A(2):346-52. doi:10.1002/ajmg.a.36140
64. Ullah A, Kalsoom UE, Umair M, John P, Ansar M, Basit S, et al. Exome sequencing revealed a novel splice site variant in the ALX1 gene underlying frontonasal dysplasia. Clin Genet. 2017;91(3):494-8. doi:10.1111/cge.12822
65. Pini J, Kueper J, Hu YD, Kawasaki K, Yeung P, Tsimbal C, et al. ALX1-related frontonasal dysplasia results from defective neural crest cell development and migration. EMBO Mol Med. 2020;12(10):e12013. doi:10.15252/emmm.202012013
66. OMIM. *615951 Zinc finger SWIM domain-containing protein 6; ZSWIM6. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/615951 (accessed 2026-10-29).
67. Palmer EE, Kumar R, Gordon CT, Shaw M, Hubert L, Carroll R, et al. A recurrent de novo nonsense variant in ZSWIM6 results in severe intellectual disability without frontonasal or limb malformations. Am J Hum Genet. 2017;101(6):995-1005. doi:10.1016/j.ajhg.2017.10.009
68. Amaratunga SA, Bezdicka M, Tayeb TH, Soucek O, Lebl J. Homozygous variants in ZSWIM6 cause severe syndromic short stature and developmental delay. Horm Res Paediatr. 2026 [Epub ahead of print]. doi:10.1159/hrp/adaag023
69. Kahl N, Lüsebrink N, Schubert-Bast S, Freiman TM, Kieslich M. Bilateral foramina parietalia permagna – a calvarial defect caused by haploinsufficiency of the Msh homeobox 2 gene: a case report and current literature review. Neuropediatrics. 2024;55(3):205-8. doi:10.1055/s-0044-1781465
70. Landrum MJ, Lee JM, Benson M, Brown GR, Chao C, Chitipiralla S, et al. ClinVar: improving access to variant interpretations and supporting evidence. Nucleic Acids Res. 2018;46(D1):D1062-7. doi:10.1093/nar/gkx1153. Variation records 16963, 16961, 1800864 and 2005289 accessed on 2026-11-01.
71. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 802–806 (NBAS/RALF deficiency).
72. Yamada M, Cleghorn M, Krishnamurthy P, French T, Francis D, Kosaki K, et al. Two unrelated families with noncoding duplications upstream of MSX2 refine the critical regulatory region likely involved in cranial bone development and a cleidocranial dysplasia-like phenotype. Hum Mutat. 2026;2026:1258291. doi:10.1155/humu/1258291
73. OMIM. *604122 Phosphatidylinositol glycan anchor biosynthesis class B protein; PIGB. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/604122 (accessed 2026-11-01).
74. Murakami Y, Nguyen TTM, Baratang N, Raju PK, Knaus A, Ellard S, et al. Mutations in PIGB cause an inherited GPI biosynthesis defect with an axonal neuropathy and metabolic abnormality in severe cases. Am J Hum Genet. 2019;105(2):384-94. doi:10.1016/j.ajhg.2019.05.019
75. Staufner C, Peters B, Wagner M, Alameer S, Barić I, Broué P, et al. Defining clinical subgroups and genotype-phenotype correlations in NBAS-associated disease across 110 patients. Genet Med. 2020;22(3):610-21. doi:10.1038/s41436-019-0698-4
76. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 258–259.
77. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 970.
78. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 242–244.
79. Saraç Sivrikoz T, Altunoglu U, Kalelioglu İH, Yüksel A, Uyguner OZ, Has R, et al. ALX4 related parietal foramina mimicking encephalocele in prenatal period. Prenat Diagn. 2016;36(6):591-3. doi:10.1002/pd.4826
80. Seltzer LA, Hines BL, Weisberg ZS, Ezra D, Lockwood J, Kawai H, et al. New insights into enlarged parietal foramina: an anatomical, radiological, and histological study. Childs Nerv Syst. 2026;42(1):119. doi:10.1007/s00381-026-07197-w
81. Orphanet. Potocki-Shaffer syndrome (ORPHA:52022). https://www.orpha.net/en/disease/detail/52022
82. Kortesis B, Richards T, David L, Glazier S, Argenta L. Surgical management of foramina parietalia permagna. J Craniofac Surg. 2003;14(4):538-44. doi:10.1097/00001665-200307000-00028
83. Yadav S, Das RS, Mishra A, Singh D, Mishra PP. Morphological and topographical characteristics of parietal foramen in the adult human skull bones of North Indian population and its clinical correlations – a cross-sectional study. Asian J Neurosurg. 2026;21(4):725-31. doi:10.1055/s-0046-1819648
84. Silva Santos TK, De Oliveira RG, Souza Oliveira AB, Rodrigues LG, Franco A, Silva RF. 12mm wide parietal foramina: morphological aspects and forensic application – a case report. Morphologie. 2025;109(366):100973. doi:10.1016/j.morpho.2025.100973
85. Lehalle D, Altunoglu U, Bruel AL, Assoum M, Duffourd Y, Masurel A, et al. The oculoauriculofrontonasal syndrome: further clinical characterization and additional evidence suggesting a nontraditional mode of inheritance. Am J Med Genet A. 2018;176(12):2740-50. doi:10.1002/ajmg.a.40662
86. OMIM. %229400 Frontofacionasal dysplasia. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/229400 (accessed 2026-10-29).
87. OMIM. %601452 Oculoauriculofrontonasal syndrome; OAFNS. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/601452 (accessed 2026-10-29).
88. OMIM. 203000 Frontonasal dysplasia with alar clefts. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/203000 (accessed 2026-10-29).
89. Ribeiro-Bicudo LA, Quiezi RG, Guion-Almeida ML, Legnaro C, Richieri-Costa A. Exclusion of mutations in TGIF, ALX3, and ALX4 genes in patients with the syndrome of frontonasal dysgenesis, callosal agenesis, basal encephalocele, and eye anomalies. Am J Med Genet A. 2012;158A(5):1233-5. doi:10.1002/ajmg.a.35305
90. Richieri-Costa A, Guion-Almeida ML, Pagnan NA. Acro-fronto-facio-nasal dysostosis: report of a new Brazilian family. Am J Med Genet. 1992;44(6):800-2. doi:10.1002/ajmg.1320440616
91. Chaabouni M, Maazoul F, Ben Hamida A, Berhouma M, Marrakchi Z, Chaabouni H. Autosomal recessive acro-fronto-facio-nasal dysostosis associated with genitourinary anomalies: a third case report. Am J Med Genet A. 2008;146A(14):1825-7. doi:10.1002/ajmg.a.32349
92. Guo C, Zhang T, Ma Y, Yue S, Sun L. Prenatal diagnosis of a severe form of frontonasal dysplasia with severe limb anomalies, hydrocephaly, a hypoplastic corpus callosum, and a ventricular septal defect using 3D ultrasound: a case report and literature review. BMC Pregnancy Childbirth. 2024;24(1):420. doi:10.1186/s12884-024-06619-4
93. Ainuz BY, Wolfe EM, Wolfe SA. Surgical management of a mild case of frontonasal dysplasia: a case report and review of literature. Cureus. 2021;13(1):e12821. doi:10.7759/cureus.12821
94. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 204–205.
95. Pang H, Gao Q. A three-dimensional-printed model from ultrasound volumes: a new tool for the prenatal diagnosis of frontonasal dysplasia. Quant Imaging Med Surg. 2026;16(9):756. doi:10.21037/qims-2026-0871
96. OMIM. #304110 Craniofrontonasal syndrome; CFNS. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/304110 (accessed 2026-10-29).
97. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 588–591.
98. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 314–316.
99. Orphanet. SPECC1L-related hypertelorism syndrome (ORPHA:1519). https://www.orpha.net/en/disease/detail/1519
100. OMIM. #145420 Teebi hypertelorism syndrome 1; TBHS1. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/145420 (accessed 2026-10-29).
101. OMIM. #617746 Sweeney-Cox syndrome; SWCOS. Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/617746 (accessed 2026-10-29).
102. OMIM. %155145 Cleft, median, of upper lip with polyps of facial skin and nasal mucosa (Pai syndrome). Online Mendelian Inheritance in Man. Baltimore (MD): Johns Hopkins University. https://omim.org/entry/155145 (accessed 2026-10-29).
103. 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. 776–778.
104. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 306–307.
105. Yildizdal S, Boynuyogun E, Altunal SK, Kose B, Koçer U, Vargel I. A novel umbrella graft approach in nasal reconstruction of patients with frontonasal dysplasia. J Plast Reconstr Aesthet Surg. 2026;114:205-11. doi:10.1016/j.bjps.2026.01.019
106. GeneGuidance. Gene records for ALX1, ALX3, ALX4 and ZSWIM6 (version of 2026-09-26).
107. Silva RCE, Tonello C, Bastos JCDC, Azevedo RMG, Zechi-Ceide RM, Garib D, et al. Multidisciplinary oral rehabilitation of complex frontonasal dysplasia: a case report. Cleft Palate Craniofac J. 2026;63(4):860-7. doi:10.1177/10556656241307738
108. European Medicines Agency. Medicines output: orphan designations. Amsterdam: EMA; accessed 29 October 2026. https://www.ema.europa.eu/en/documents/report/medicines-output-orphan_designations-json-report_en.json
109. ClinicalTrials.gov. Search by condition “frontonasal dysplasia”, “parietal foramina” and “hypertelorism” (29 October 2026) and by “acrofrontofacionasal”, “PIGB” and “NBAS” (1 November 2026). Bethesda (MD): National Library of Medicine. https://clinicaltrials.gov
110. Staufner C, Haack TB, Köpke MG, Straub BK, Kölker S, Thiel C, et al. Recurrent acute liver failure due to NBAS deficiency: phenotypic spectrum, disease mechanisms, and therapeutic concepts. J Inherit Metab Dis. 2016;39(1):3-16. doi:10.1007/s10545-015-9896-7. PMID 26541327
111. Orphanet. Orphadata: epidemiological data (product 9, prevalence), version 2026-06-23. https://www.orphadata.com/epidemiology/
112. Standen VG, Monsalve S, Arriaza B, Verano J, Rivera M. A case of enlarged parietal foramina or foramina parietalia permagna in an individual from the Chinchorro Culture of northern Chile (4000 BP). Int J Paleopathol. 2025;48:34-42. doi:10.1016/j.ijpp.2024.11.004