Craniosynostoses: General Approach, Crouzon Syndrome and Apert Syndrome

Literature and regulatory information updated: 6 October 2026.

Isolated and syndromic craniosynostosis (craniostenosis); fibroblast growth factor receptor (FGFR)-associated craniosynostosis syndromes; Crouzon syndrome (craniofacial dysostosis) and Crouzon syndrome with acanthosis nigricans; Apert syndrome (acrocephalosyndactyly type I); with Pfeiffer, Muenke and Saethre–Chotzen syndromes as neighbouring entities

0. Initial Reference Table

1. Introduction

Craniosynostosis, also termed craniostenosis, is premature fusion of one or more cranial vault sutures. Sutures are the main skull growth zones, and bone growth occurs perpendicular to their line; when a suture closes early, growth perpendicular to it is prevented and the skull expands compensatorily in other directions, producing predictable skull shapes [1, 27]. Premature fusion also alters facial growth, may restrict brain growth and exposes the child to intracranial hypertension [1]. The Oxford manual estimates prevalence at 1 in 2,500 children: approximately 15% of cases correspond to a recognisable syndrome and another 10% have a genetic cause identifiable through targeted testing, whereas in most isolated craniosynostoses the aetiology remains unknown and environmental factors probably play a role [1].

This article covers craniosynostosis as a general finding (FGFR1, FGFR2, FGFR3, TWIST1, EFNB1, TCF12, ERF), Crouzon syndrome (FGFR2, with the Crouzon with acanthosis nigricans variant caused by FGFR3) and Apert syndrome (FGFR2 p.Ser252Trp and p.Pro253Arg). As these entities share the same clinical reasoning, Pfeiffer, Muenke and Saethre–Chotzen syndromes are included to the extent covered by the consulted sources, being the main diagnostic alternatives in a child with coronal synostosis. Muenke syndrome and Crouzon syndrome with acanthosis nigricans also belong to the FGFR3 allelic spectrum, which shares the common genetics of the receptor with achondroplasia, hypochondroplasia and thanatophoric dysplasia.

Craniosynostosis with radiohumeral fusions due to CYP26B1 deficiency was described in 2011 by Laue and colleagues, who identified biallelic variants in the gene encoding the retinoic acid-degrading enzyme in three siblings with a lethal phenotype and in a Turkish infant previously diagnosed with Antley–Bixler-like syndrome [20]. It was the first human disease attributed to a genetic defect of retinoic acid catabolism and demonstrated a physiological role of this morphogen in maintaining cranial suture patency [20]. The entity is included here because it is a differential diagnosis of Pfeiffer and Antley–Bixler syndromes, the latter caused by POR or FGFR2 [22, 28].

Historically, Apert syndrome was described by Wheaton in 1894; in 1906 Eugène Apert brought together nine cases under the designation acrocephalosyndactyly and, in 1920, Park and Powers published a classical account of the entity [10]. In 1912, Octave Crouzon described, in a mother and daughter, the “hereditary craniofacial dysostosis” that now bears his name [3]. Saethre (1931) and Chotzen (1932) independently described the syndrome combining brachycephaly, ptosis, characteristic external ears and cutaneous syndactyly [29], and Pfeiffer reported dominant acrocephalosyndactyly with broad thumbs and great toes in 1964 [30]. The molecular era began in 1994, when Reardon and colleagues demonstrated that FGFR2 variants cause Crouzon syndrome and Muenke and colleagues identified a recurrent FGFR1 variant in families with Pfeiffer syndrome [31, 32]. In 1995, Wilkie and colleagues found adjacent amino acid substitutions (Ser252Trp and Pro253Arg) in the linker between the second and third immunoglobulin domains of FGFR2 in all 40 unrelated Apert syndrome cases studied, demonstrating that Apert and Crouzon are allelic [33]. In the same year, Meyers and colleagues described the transmembrane FGFR3 variant p.Ala391Glu in three families with Crouzon syndrome and acanthosis nigricans [34], and in 1997 Muenke syndrome was defined on the basis of a single variant, FGFR3 p.Pro250Arg, in 61 individuals from 20 families [35].

Modern management is multidisciplinary and centralised. The Dutch guideline on craniosynostosis treatment and management, published in 2015 and updated in 2021 with support from 21 scientific societies, was translated and published with funding from the European Reference Network for rare craniofacial anomalies and ear, nose and throat disorders (ERN CRANIO), which endorses it as a European guideline; a patient and family version has existed since 2023 [16, 36, 37]. In the United States, the American Academy of Pediatrics (AAP) published a clinical report in 2020 on recognising the abnormally shaped head, focused on distinguishing synostosis from positional deformation [14].

2. Molecular Genetics and Pathophysiology

2.1 Suture biology and signalling pathways

The vault bones form by intramembranous ossification: differentiating osteoblasts accumulate at the osteogenic fronts and undifferentiated cells between these fronts constitute the suture, which remains patent while the brain grows [14]. Growth is controlled by a signalling system including ephrins, which mark sutural boundaries, fibroblast growth factor receptors (FGFR) and the transcription factor TWIST1 [1]. In FGFR- and MSX2-associated syndromes, the mechanism is gain of function, whereas in TWIST1 it is haploinsufficiency [1]. The coronal suture has a distinct embryonic origin at the boundary between neural crest and cephalic mesoderm, which has been invoked to explain its distinctive molecular epidemiology [38].

FGFRs are tyrosine kinase receptors with three extracellular immunoglobulin-like domains (IgI, IgII, IgIII), a transmembrane segment and an intracellular kinase domain. FGF binding induces receptor dimerisation, PLCG1 and FRS2 phosphorylation, and RAS-MAPK and AKT signalling activation [6, 11]. FGFR2 (10q26.13) has 18 exons, and the reference transcript encodes an 821-amino-acid protein with a 21-residue signal peptide; residues Ser252 and Pro253 lie in the linker between IgII and IgIII [4]. The IgIII domain exists as two isoforms through alternative splicing, IIIb (exon 8) and IIIc (exon 9), and some literature confuses exon numbering with domain numbering, calling these alterations “exon IIIb” and “exon IIIc” [4]. Linker variants increase affinity and alter FGF-binding specificity, reducing ligand dissociation; the FGFR3 variant p.Pro250Arg, situated in the same linker, increases affinity for FGF9 and accelerates bone differentiation [4, 6]. GeneGuidance summarises the mechanism as constitutive or dysregulated FGFR2 signalling, a gain in pathway activity rather than a deficiency, without a biochemical marker [11].

2.2 Genes and entities

The relative importance of these genes was measured in a cohort of 326 children operated on for craniosynostosis in Oxford (born between 1993 and 2002): 21% of children born between 1998 and 2002 had a confirmed genetic cause and, among the 84 genetic cases from all years, contributions were 32% for FGFR2, 25% for FGFR3, 19% for TWIST1, 7% for EFNB1 and 15% for chromosomal anomalies; the FGFR3 variant p.Pro250Arg was the most frequent single alteration (24% of genetic cases) [38]. This cohort predates the description of TCF12 and ERF as craniosynostosis genes (2013), limiting comparison with later panel-based series. In a 20-gene panel applied to 309 patients from Australia and New Zealand without variants in FGFR1-3 hotspots or TWIST1, yield was 14% (retrospective) and 15% (prospective), with TCF12 (22 patients) and EFNB1 (8) the most frequent genes [47].

2.3 Genotype–phenotype correlations and FGFR2 pleiotropy

In Apert syndrome, the two recurrent variants explain the vast majority of cases; sequencing detects approximately 99% and deletion/duplication analysis has rare yield (a 1.9 kb deletion including exon 8, a 1.3 kb deletion of exon 9 with a chimeric IIIb/IIIc domain, and three de novo Alu element insertions near exon 9, all interpreted as gain of function) [4]. Relative frequencies vary between sources: GeneReviews gives 62–71% for p.Ser252Trp and 26–33% for p.Pro253Arg, the bone dysplasia atlas reports approximately two-thirds and one-third, and Smith states that the two variants explain 98% of cases [4, 10, 12]. Variant p.Pro253Arg has been associated with more severe syndactyly and p.Ser252Trp with more frequent cleft palate; no other feature varies consistently with genotype and some studies find no correlation [4, 10].

In Crouzon syndrome, all variants found are in FGFR2, mainly in the IgIII domain; residue Cys342 is the most frequent hotspot, with substitutions by arginine, tyrosine, serine, phenylalanine and tryptophan [3, 9]. The same variant may produce Crouzon or Pfeiffer phenotypes, and families have members classified as Pfeiffer, Jackson–Weiss or Crouzon, although the phenotypic category tends to remain consistent within each family [3, 5]. The bone dysplasia atlas even considers Crouzon syndrome a mild expression of FGFR2-related Pfeiffer syndrome [8]. Some variants are associated with particular phenotypes: p.Trp290Cys with radiohumeral synostosis and central apnoea in Pfeiffer syndrome, and p.Ser267Pro and p.Val274_Glu275delinsLeu with cutis gyrata-like linear skin creases in individuals with Crouzon syndrome [5].

Crouzon syndrome with acanthosis nigricans is not a variant of classical Crouzon syndrome: it is caused by a specific variant of the FGFR3 transmembrane domain, p.Ala391Glu, and patients may show subtle radiological features of achondroplasia (narrow interpedicular distances and sacrosciatic notches, short metacarpals and phalanges) [3, 34]. GeneGuidance notes that attributing this phenotype to the wrong gene is a known diagnostic pitfall and counselling should not use data from FGFR2-related Crouzon syndrome [11].

2.4 TWIST1 haploinsufficiency and partner genes

TWIST1 has two exons and encodes a 202-amino-acid basic helix–loop–helix transcription factor; all pathogenic variants cause functional haploinsufficiency, whether truncating variants, missense variants in functional domains or large deletions, and TWIST1 loss disinhibits RUNX2 and accelerates osteogenesis [40]. GeneReviews states that no splicing, intronic or second-exon variants have been described, but Smith mentions variants outside the coding region, including 5' untranslated region variants disrupting translation [29, 40]. Variants in the TWIST box (C-terminal, binding to and inhibiting RUNX2) are associated with isolated sagittal or unicoronal synostosis rather than the complete syndrome [40]. TCF12 is a class I E protein that heterodimerises with TWIST1: the two act synergistically and mice double heterozygous for Tcf12 and Twist1 loss of function have severe coronal synostosis, making TCF12-TWIST1 heterodimer dosage critical to the coronal suture [41]. ERF is an ETS repressor directly linked to ERK1/2; reduced dosage causes postnatal multisuture craniosynostosis in humans and mice, linking RAS-ERK signalling to osteogenesis regulation [43].

2.5 Paternal origin and the paternal age effect

The association between advanced paternal age and de novo dominant mutations, as in achondroplasia and Apert syndrome, has been known since 1957 (Penrose); on average, the increase in paternal age in these disorders is only 3 to 5 years, so many cases are born to young fathers [48]. In Apert syndrome, the new mutation occurs in foci of increased frequency in most testes and confers a gain of function promoting clonal expansion of spermatogonia; this mechanism, rather than a high mutation rate at the nucleotide, explains increased risk [48]. Goriely and colleagues quantified substitutions at FGFR2 nucleotide 755 in sperm and showed that, although levels increase with age, even high levels result from rare mutational events, proposing a selective advantage in spermatogonia [49]. The concept was generalised to “paternal age effect” disorders (Apert, achondroplasia, thanatophoric dysplasia, Costello syndrome), in which mutations are positively selected and expand clonally in the testes of all men, in a process resembling oncogenesis through the growth factor receptor-RAS pathway [50].

Advanced paternal age has been clinically associated with de novo variants in Crouzon, Apert, Pfeiffer, Beare–Stevenson and Muenke syndromes, and the effect was demonstrated molecularly in Apert syndrome [5]. In Muenke syndrome, all ten informative de novo mutations studied by Rannan-Eliya were paternal in origin (95% confidence interval: 74–100%), with mean paternal age 34.7 years; the transversion c.749C>G occurs at a rate of 7.6 to 8 × 10⁻⁶ per haploid genome, one of the highest known for a human transversion [6, 51]. In Apert syndrome, sperm variant frequency is higher in men over age 60 [4]. Parental somatic and germline mosaicism has been described in Crouzon syndrome, which the GeneReviews FGFR overview presents as the only FGFR syndrome with documented parental mosaicism; however, the GeneReviews Apert syndrome chapter cites parental germline mosaicism in a family with two affected sisters born to unaffected parents [4, 5].

2.6 Pathophysiology of complications

Skull shape results from pressure of the growing brain against a container with fused sutures; expansion occurs perpendicular to the fused suture and, when several sutures are closed, towards regions of least resistance [5]. Prenatal pansynostosis produces a cloverleaf skull (Kleeblattschädel), whereas postnatal pansynostosis may manifest only as arrest of head circumference growth [5]. In syndromic forms, intracranial hypertension is multifactorial: insufficient intracranial volume, moderate-to-severe obstructive sleep apnoea, hydrocephalus and intracranial venous hypertension, alone or in combination [15]. Different development of the foramen magnum and early fusion of intraoccipital synchondroses may contribute to hydrocephalus and abnormal skull shape [5]. Maxillary retrusion and hypoplasia reduce orbital depth (exorbitism) and upper airway calibre, and tracheal anomalies, such as tracheal cartilaginous sleeves, add a risk of sudden obstruction by secretions [5].

2.7 Somatic FGFR2 and FGFR3 alterations

The same pathways are somatically altered in tumours. Isolated tumours, such as gastric carcinoma, frequently have somatic FGFR2 variants without hereditary significance, and tumour-acquired FGFR2 fusions are a therapeutic target in intrahepatic cholangiocarcinoma [4, 11]. GeneGuidance emphasises that none of these indications applies to germline variant carriers: a germline FGFR2 variant causes a craniofacial or skeletal phenotype, does not confer predisposition to cholangiocarcinoma or gastric cancer, and does not justify cancer surveillance; conversely, an FGFR2 fusion in a tumour panel should not be communicated to the family as hereditary [11].

2.8 CYP26B1 deficiency: retinoic acid catabolism and suture patency

Retinoic acid (RA), derived from vitamin A, is a morphogen whose tissue concentration results from the balance between synthesising enzymes (RALDH1-3) and degrading enzymes (CYP26A1, CYP26B1 and CYP26C1), in a source-and-sink mechanism that creates local gradients during development [20]. The CYP26 enzymes are cytochrome P450 monooxygenases anchored in the endoplasmic reticulum membrane that oxidise all-trans RA into inactive metabolites [19, 20]. CYP26B1 maps to 2p13.2 (MANE Select transcript NM_019885.4) and is expressed in the developing hindbrain and in the skeleton, where it regulates limb patterning, ossification and the formation of some synovial joints [19]. The enzyme receives electrons from P450 oxidoreductase (POR), its only established functional interaction, which explains the overlap with Antley–Bixler syndrome due to POR deficiency: the authors of the original description proposed that the joint fusions and craniosynostosis of POR deficiency result from reduced CYP26B1 activity and RA accumulation, and not only from impaired cholesterol synthesis [19, 20, 52].

Exposure to excess exogenous RA is teratogenic to the limbs and craniofacial skeleton, and nutritional vitamin A excess causes premature epiphyseal closure and craniosynostosis [20, 53]. The Cyp26b1−/− mouse has shortening of the proximal limb segments, radiohumeral synostosis, oligodactyly, craniofacial defects and reduced calvarial ossification, and the null and hypomorphic zebrafish mutants have hypermineralisation of the facial and axial bones [20]. Laue and colleagues proposed two mechanisms. In bones formed by endochondral ossification, the lack of local RA degradation allows unrestricted chondrogenesis at sites of joint formation, producing cartilaginous fusions that later ossify. In the calvaria, formed by membranous ossification, RA accelerates the transition of osteoblasts to pre-osteocytes and osteocytes, with reduced osteoid production and premature mineralisation [20]. The model explains the coexistence of calvarial hypoplasia and craniosynostosis: with severe enzyme deficiency, the frontal and parietal plates do not grow enough to form sutures and the ossification defect predominates (midline defect, occipital encephalocele), whereas with residual activity the calvaria grows but the coronal sutures, where Cyp26b1 expression is highest, fuse prematurely [20]. The authors noted that FGF2 induces an osteocyte-like phenotype in vitro and suggested that constitutive FGFR signalling and RA might converge on the osteoblast-osteocyte transition in specific sutures, a hypothesis not demonstrated in patients [20].

Variants with a nucleotide equivalent were confirmed in VariantValidator on the MANE Select transcript; those from the series of Grand and colleagues are given at protein level, as in the source. Silveira and colleagues compiled the published biallelic genotypes and proposed a gradient: higher residual activity is associated with craniosynostosis, ear malformations with hearing loss and intellectual disability, and lower activity with neural tube defects (encephalocele, spina bifida), deficient mineralisation (calvarial defects, absent clavicles), femoral hypoplasia and oligodactyly [19]. Structural modelling by Grand and colleagues placed the variants of the non-lethal forms further from the catalytic centre [25]. The c.1083C>A splicing variant impaired the enzyme more than p.Arg363Leu (p = 0.0063), which led the authors to question whether the latter is a complete null allele; as no patients with two truncating variants have been described, they suggested that total loss of function may be incompatible with human embryonic development, a hypothesis not demonstrated [19].

Dominant inheritance remains to be demonstrated. Sarma and colleagues described the truncating variant c.86C>A, p.(Ser29Ter), in the heterozygous state in an affected father and daughter of Indian origin and in a carrier grandfather without phenotype, and proposed haploinsufficiency [21]. Silveira and colleagues noted that the proband's parents were first cousins, that a second variant was not excluded, and that a monoallelic CYP26B1 deletion is recorded in a healthy individual in the Database of Genomic Variants [19]. Recurrent 2p13.2 microdeletions including CYP26B1 and EXOC6B have been described in two individuals with intellectual disability, language delay, hyperactivity, facial asymmetry, ear malformations and vertebral or craniofacial anomalies, with RA-induced CYP26B1 expression reduced to 8% of control levels in lymphoblasts [55]; the relative contribution of the two genes has not been established [19, 55].

3. Clinical Spectrum and Manifestations by System

3.1 Single-suture synostoses and skull shapes

The suture involved determines head shape, and descriptive terminology is a frequent source of confusion. Dolichocephaly or scaphocephaly (increased length relative to width) may be a postural deformation associated with prematurity and breech presentation, but is also the phenotype of sagittal synostosis; brachycephaly (occipital and frontal flattening with increased width) is a common postural deformation but also occurs in bicoronal synostosis; plagiocephaly (asymmetry) requires distinction between deformation and unicoronal or lambdoid synostosis; trigonocephaly (triangular forehead) is characteristic of metopic synostosis and should be distinguished from a simple metopic ridge; turricephaly, oxycephaly and acrocephaly describe the tall, narrow “tower” skull of severe coronal or multisuture synostoses; and cloverleaf skull (Kleeblattschädel), due to fusion of all sutures, is a severe situation requiring urgent surgery [27].

Sources: [1, 14].

Single-suture synostoses have their own functional consequences. Intracranial hypertension before surgery probably occurs in 2.5–14% of sagittal synostoses, increasing with age, in 2–8% of metopic and approximately 16% of isolated coronal synostoses [15]. Chiari malformation, frequently asymptomatic, has been described in 3–8% of sagittal, 0% of metopic, 6–18% of coronal and 25–60% of lambdoid synostoses [15]. Visual and ocular motility abnormalities are frequent in metopic and unicoronal synostoses [15]. At 6–18 months, children with single-suture synostosis have a slightly increased risk of speech and language problems, persisting mildly to moderately at 36 months; at age 7, metopic, coronal or lambdoid synostoses are associated with mild-to-moderate risk, whereas sagittal synostosis does not appear associated with increased risk [15]. At school age, mean intelligence quotient is comparable to or slightly below that of the population, but values below 80–85 are more common in metopic (21–31% versus 9–16% in the norm), coronal (27% versus 9%) and lambdoid synostoses [15].

3.2 Syndromic craniosynostoses: shared features

FGFR-associated craniosynostosis syndromes form a spectrum from severe prenatal multisuture synostosis, with feeding and airway problems, to isolated unicoronal synostosis; more than 500 individuals had been described at the date of GeneReviews [5]. Most have congenital craniosynostosis, but fusion may arise later in childhood and other sutures may fuse progressively; Muenke syndrome tends to involve only the coronal sutures, whereas Apert, Crouzon and Pfeiffer syndromes are associated with progressive postnatal fusion [5]. The shared pattern includes bicoronal synostosis, midface hypoplasia and shallow orbits with exorbitism and strabismus [14].

Sources: [4–6, 40, 56, 57].

3.3 Apert syndrome

Coronal synostosis is almost universal (nearly 100%), and many patients have multisuture synostosis or pansynostosis with cloverleaf skull; the anterior fontanelle is usually large and displaced anteriorly towards the forehead [4]. Children with the remaining phenotype without craniosynostosis at diagnosis have been described, but, because bone fusions are progressive, whether they develop it later is unknown [4]. The face is flat, with a high, full forehead, horizontal supraorbital groove, shallow orbits, interrupted eyebrows, hypertelorism, downslanting palpebral fissures, a small nose, maxillary hypoplasia, a narrow palate with a midline groove, with or without cleft palate or bifid uvula, and a trapezoidal mouth at rest [10]. Unlike Crouzon syndrome, in which the midface is normally formed but retruded, in Apert syndrome it is underdeveloped and retruded, with greater vertical impaction and a shorter maxilla, resembling Pfeiffer syndrome; this results in shallow orbits, malocclusion and relative mandibular prognathism [4]. Cleft palate is frequent in Apert syndrome and rare in Crouzon syndrome [4].

The hand always includes fusion of the three central digits, with a single nail for digits 2–4 (synonychia, “mitten hand”); the thumb and fifth digit may be included; feet have syndactyly of the three lateral toes, toes 2–5 or all toes, and upper limbs are usually more affected than lower limbs, without foot synonychia [4]. Radiologically, hand synostoses are progressive (distal phalanges of digits 2–4, bases of metacarpals 4 and 5, sometimes capitate and hamate), with brachymesophalangy, proximal then distal symphalangism, and a thumb with a delta proximal phalanx and radial deviation; in the feet syndactyly is cutaneous, but progressive tarsal and metatarsal fusion occurs (particularly 1st and 2nd), with hallux varus [12]. Radiohumeral synostosis, pre- or postaxial polydactyly and a broad distal thumb or great-toe phalanx may occur [4]. Glenohumeral dysplasia causes progressive restriction of forward flexion and abduction, limiting overhead tasks, and the feet develop progressive deformities with first-metatarsal shortening, load transfer to the second, painful callosities and difficulty finding footwear [4].

In the central nervous system, approximately 60% have non-progressive ventriculomegaly and 6–13% true hydrocephalus; jugular foramen stenosis was found in 93% [4]. Structural malformations include corpus callosum anomalies (23%), absence of the septum pellucidum (17%), Chiari I malformation or low-lying cerebellar tonsils (17%), posterior fossa arachnoid cyst (7%) and limbic malformations; only 2% have chronic tonsillar herniation, compared with 73% in Crouzon syndrome [4]. Smith also describes megalencephaly, gyral and hippocampal anomalies, abnormal semicircular canals, cochlear hypoplasia and malformed or fused ossicles [10].

Ocularly, strabismus occurs in 60%, refractive error in 34% and anisometropia in 19%; exposure keratopathy with corneal scarring (8%) and optic atrophy (8%) are secondary complications potentially preventable through surveillance and aggressive treatment of incomplete eyelid closure and ICH [4]. Hearing loss is common (80%), usually conductive, due to middle ear disease, ossicular anomalies and external auditory canal stenosis or atresia; semicircular canal anomalies occur in 70% [4]. Airway obstruction may be multilevel (nasal or choanal narrowing, tongue-base obstruction, tracheal cartilaginous sleeves and fused rings), ranging from mild symptoms to obstruction requiring tracheostomy, and cleft palate repair may unmask pharyngeal obstruction and worsen obstructive apnoea [4].

Cervical fusions occur in 68%, particularly C5–C6 (half with a single fusion, half multiple), compared with 25% in Crouzon syndrome (particularly C2–C3); atlantoaxial subluxation (7%) and occult C1 spina bifida (7%) have been described [4]. Dental agenesis (particularly upper canines) and enamel opacities occur in more than 40%, with ectopic eruption of the upper first molars and lateral palatal swellings [4]; a 2025 meta-analysis estimated permanent tooth agenesis (excluding third molars) at 37% (95% CI: 28–48%) in Apert syndrome and 31% (17–46%) in Crouzon syndrome, with predominantly bilateral absence of the lower second premolars [58].

Approximately 10% have structural heart disease (ventricular septal defect, overriding aorta; complex heart defects with greater early mortality) and 9.6% genitourinary anomalies (hydronephrosis, cryptorchidism) [4]. Intestinal malrotation is the most frequent digestive anomaly (one in 15 children in one series, probably underestimated because of lack of systematic radiological investigation), with distal oesophageal stenosis, pyloric stenosis, oesophageal atresia and ectopic anus described, and congenital diaphragmatic hernia in five infants [4, 5]. Hyperhidrosis is constant; adolescence brings oily skin and extensive acne on the face, chest, back and arms, sometimes excessive forehead wrinkles, and nail dystrophy is common [4]. Growth at birth is above percentile 50, slowing in childhood to percentiles 5–50 and further after adolescence [10].

The risk of intellectual developmental disorder is significant. Older series recorded mean IQs of 74 (52–89) and 61 (44–90), and another found IQ below 70 in 52%; more recent studies described means of 76.7 (59–94) in 6 individuals and 76.9 (47–108) in 8 [10]. GeneReviews indicates that most have normal intelligence or mild impairment, with moderate-to-severe impairment in some, and identifies first craniectomy after the first year of life and structural brain malformations as risk factors (septum pellucidum anomalies are associated with lower IQ; corpus callosum data are contradictory) [4]. Children raised in families have better cognitive outcomes than institutionalised children, and current outcomes may be better than those in historical series [4]. Early surgery does not necessarily prevent intellectual disability, although outcomes are better with early decompression [10].

3.4 Crouzon syndrome

Crouzon syndrome usually presents with isolated craniofacial anomalies: proptosis due to shallow orbits, with or without divergent strabismus, hypertelorism, frontal bossing, maxillary hypoplasia with a curved “parrot-beak” nose, inverted V-shaped palate, conductive hearing loss and craniosynostosis, particularly coronal, lambdoid and sagittal, with a palpable ridge [3]. Craniosynostosis is a variable feature; infants without synostosis may have a normal face at birth and develop significant proptosis, external strabismus, midface retrusion, convex nasal bridge and relative prognathism in the first one to two years, with considerable variability among affected relatives [5]. A high palate is common and cleft palate less frequent; hearing loss occurs in 74%, mostly conductive [5]. Hands are normal, although radiologically shorter phalanges than in unaffected relatives have been described [5].

Ocular manifestations include strabismus, exposure conjunctivitis or keratitis, reduced visual acuity, optic atrophy, nystagmus and optic nerve hypoplasia; keratoconus, glaucoma, iris coloboma and globe subluxation occur occasionally [3]. Other occasional findings are intellectual disability, hydrocephalus, seizures, agenesis of the corpus callosum, Chiari I, syringomyelia, jugular foramen stenosis, auditory canal atresia, cleft lip and/or palate, dental agenesis, pulmonary valve stenosis, tracheobronchomalacia, tracheal cartilaginous sleeve, radial head subluxation and sacrococcygeal appendage [3]. Upper airway obstruction frequently causes obligatory mouth breathing and rarely acute respiratory difficulty, but obstructive sleep apnoea is common and more than one-third of patients will need midface advancement to treat it [3]. Progressive hydrocephalus occurs in fewer than one-third according to Smith and approximately 30% according to GeneReviews, frequently with tonsillar herniation [3, 5]. ICH is very frequent and may recur after initial surgery [3]. In a London series of 49 patients treated only after ICH diagnosis, 30 (61.2%) had at least one episode, at a mean age of 1.42 years, attributed to craniocerebral disproportion or venous hypertension (19), hydrocephalus (8) or airway obstruction (3); 14 had a second episode and 3 a third, frequently with a different cause, and diagnosis of the first episode before age one increased recurrence risk [59].

Intelligence is usually normal, although there is a risk of delay, particularly with hydrocephalus and ICH [3, 5]. A study of 31 Swedish adults recorded lower educational attainment, a lower probability of having a partner and fewer children, without differences in housing type and with self-rated health similar to controls, except greater antiepileptic use; depressed mood was more frequent, but a positive attitude to life was similar [5].

3.5 Crouzon syndrome with acanthosis nigricans

The syndrome combines craniosynostosis with variable skull shape, significant proptosis, external strabismus, prognathism, midface hypoplasia, choanal atresia and hydrocephalus, considered frequent by the Oxford manual [1, 3, 5]. Acanthosis nigricans (hyperpigmentation of skin folds) may be present neonatally or develop later, usually in childhood, and predominantly involves the eyelids, perioral and perialar regions and neck [3, 5]. Hearing loss was recorded in 14%, intelligence is usually normal and odontogenic tumours have been described [5]. GeneGuidance adds that, in this context, acanthosis nigricans forms part of the syndrome and does not indicate insulin resistance, and that hydrocephalus and cervical fusions are reported more frequently than in FGFR2-related Crouzon syndrome [11].

3.6 Pfeiffer syndrome

Pfeiffer syndrome combines brachycephaly due to coronal synostosis, with or without sagittal involvement, a high forehead, hypertelorism, shallow orbits, proptosis, a small nose, choanal stenosis, a narrow maxilla and high palate, with broad, medially deviated distal phalanges of the thumbs and great toes, a frequently delta-shaped proximal phalanx, and partial syndactyly of digits 2–3 and toes 2–4 [30]. Cohen delineated three prognostically relevant types: type 1 is the classical form, with normal or almost normal intelligence and favourable prognosis; type 2 combines cloverleaf skull, severe proptosis, severe central nervous system involvement, elbow ankylosis or synostosis and visceral anomalies; type 3 resembles type 2 without cloverleaf skull [8, 30]. Types 2 and 3 carry a high risk of exposure keratopathy, obstructive apnoea and persistent ICH after suture release; most had a poor course, but survival is possible with aggressive medical and surgical treatment, particularly of the airway [30]. Hearing loss occurs in 92%, approximately 28% of children require surgery for hydrocephalus, and approximately 50% with cloverleaf skull have Chiari I [5]. In children with the FGFR2 variant p.Trp290Cys, all reported survivors required mechanical ventilation during sleep at some point because of central apnoea [5]. A 2019 study reported that all previously described individuals with p.Trp290Cys had severe intellectual disability or early death, but three children treated with aggressive surgical techniques addressing cephalocranial disproportion had approximately normal neurocognitive development [5]. Hand and foot changes tend to be less severe with FGFR1 than FGFR2 [30].

3.7 Muenke syndrome

Muenke syndrome has marked phenotypic variability, even within families, and incomplete penetrance [6, 56]. In the largest international study (106 patients from 71 families), 85% had craniosynostosis (bilateral in 47.5%, unilateral in 28.2%), without a sex difference in bicoronal synostosis; hearing loss was recorded in 70.8%, developmental delay in 66.3%, intellectual developmental disorder in 35.6%, ADHD in 23.7% and seizures in 20.2%, and 75% of patients with a complete skeletal survey had at least one additional radiological abnormality [57]. Hearing loss is typically bilateral sensorineural, mild to moderate, affecting low-to-middle frequencies, sometimes progressive, may develop after normal neonatal screening and occur in carriers without craniosynostosis [6, 56]. Other findings include macrocephaly, temporal bossing, hypertelorism, mild ptosis or proptosis, strabismus (39–66%), brachydactyly, broad thumbs and great toes, short, broad middle phalanges (“thimble-like”), conical epiphyses and carpal (capitate–hamate, trapezium–trapezoid) and tarsal (calcaneus–cuboid) fusions [6, 56]. Obstructive apnoea is less prevalent than in other syndromes [6], and preoperative ICH is rare (0–4%) [15]. A tendency towards greater severity in females and mesial temporal lobe dysplasia with temporal seizures, resembling hypochondroplasia, have been described [60]. The AAP report gives hearing loss in approximately one-third of patients, below the international multicentre study's figure [14, 57].

3.8 Saethre–Chotzen syndrome

The classical form combines uni- or bilateral coronal synostosis, facial asymmetry (particularly in unicoronal cases), strabismus, ptosis and a small external ear with a prominent helical root crossing the concha; partial cutaneous syndactyly of digits 2–3 is common and may be subtle [29, 40]. Other findings are a low frontal hairline, late-closing fontanelles, hypertelorism, lacrimal duct stenosis, maxillary hypoplasia with a narrow palate, mild brachydactyly, fifth-finger clinodactyly, broad great toes in valgus, parietal foramina, cervical vertebral anomalies, radioulnar synostosis and a duplicated or triangular-epiphysis distal great-toe phalanx [29, 40]. Craniosynostosis is not obligatory; intelligence is usually normal, but mild-to-moderate intellectual disability is associated with large TWIST1 deletions [29, 40]. Mild obstructive apnoea has been described in 5% and ICH in 21% (papilloedema persisting more than one year after surgery) [40], and Smith recommends annual surveillance for papilloedema and language regression [29]. A severe extreme is indistinguishable from Baller–Gerold syndrome (severe craniosynostosis, radial hypoplasia or agenesis and vertebral segmentation defects), and Robinow–Sorauf syndrome is now considered the mild extreme of the spectrum [29, 40].

3.9 Other forms to recognise

ERF-associated craniosynostosis has Crouzon-like facies but typically presents later and may involve the sagittal and lambdoid sutures in addition to coronals, with ICH and Chiari malformation risk [1]. In a series of 16 probands and 20 relatives, the sagittal–lambdoid (“Mercedes-Benz”) pattern predominated; synostosis was frequently postnatal, insidious and progressive, with median presentation age 42 months and, in some cases, permanent visual loss from unsuspected ICH; occlusion was class I, unlike Crouzon syndrome, and speech delay, poor motor coordination, hyperactivity and attention deficit were common, with variable expressivity and non-penetrance in relatives [44]. Craniofrontonasal syndrome (EFNB1) is X-linked dominant with paradoxically mild manifestations in males: the classical presentation is a female with marked hypertelorism, a midline nasal groove or bifid nasal tip, coronal synostosis, sloping shoulders, longitudinally grooved nails and, occasionally, syndactyly or polydactyly [1, 45]. TCF12-related coronal synostosis has a phenotype overlapping Saethre–Chotzen and non-syndromic coronal synostoses [1]. Beare–Stevenson syndrome (FGFR2) is characterised by cutis gyrata, acanthosis nigricans, cloverleaf skull and high neonatal mortality, with intellectual disability in all survivors [1, 5]. Jackson–Weiss syndrome (FGFR2) combines multisuture synostosis with broad, medially deviated great toes, foot syndactyly 2–3, tarsal fusions and normal hands, and is probably limited to members of the original pedigree [5, 6].

3.10 CYP26B1 deficiency

The clinical spectrum of biallelic CYP26B1 variants ranges from fetal death to survival into adulthood [19, 25]. In the review by Leduc and colleagues, which compiled 12 patients published up to 2025, craniosynostosis was present in 6 of 8 assessable patients (coronal, lambdoid, sagittal or multiple), delayed or deficient calvarial ossification in 8 of 10, humeroradial synostosis in 5 of 10, and absence or shortening of the first ray of the hands or feet in 9 of 11 [23]. Death occurred in utero in three pregnancies (one spontaneous fetal death and two terminations for medical reasons) or shortly after birth in 2 of 9, without a determined cause, and all surviving patients assessed (7 of 7) had developmental delay or intellectual disability, frequently with conductive hearing loss [23]. The denominators vary because many reports do not describe all findings; the series is small and of heterogeneous ascertainment, and the proportions do not equate to frequencies in the patient population.

Lethal end of the spectrum. The proband of the Laue family, the child of first cousins, died in utero at 35 weeks with a large occipital encephalocele resulting from hypoplasia of the occipital and parietal bones, arachnodactyly, aplasia of the terminal phalanges of the thumbs, severely bowed femora, elongated toes with absent halluces, elbow ankylosis due to complex radio-ulno-humeral fusion, narrow thorax, hypoplastic scapulae and pelvis, and advanced bone age; the next two siblings were recognised on ultrasound and the pregnancies terminated at 15 and 12 weeks, both with occipital encephalocele, short limbs with femoral bowing and absent halluces [20]. The Brazilian fetus described by Silveira and colleagues added generalised spina bifida occulta, absent clavicles and severe femoral hypoplasia to the spectrum (section 10.3) [19].

Intermediate and mild forms. The Turkish infant of the original description, born to consanguineous parents and initially diagnosed with Antley–Bixler-like syndrome without variants in FGFR2, POR or FGF9, had exorbitism, maxillary retrusion, brachycephaly, coronal and lambdoid synostosis with a widened sagittal suture, limited elbow extension and arachnodactyly, and died at 5 months of unknown cause [17, 20]. Morton and colleagues described the first patient to survive beyond infancy, a 22-year-old woman with multisuture synostosis, radiohumeral synostosis, normal bone mineral density and intellectual disability, with similarities to Antley–Bixler and Pfeiffer syndromes [28]. Grand and colleagues described four living patients from two families with multisuture synostosis, conductive hearing loss, joint contractures, long slender fingers, camptodactyly, broad fingertips and developmental delay or intellectual disability, and imaging showed gracile long bones and ribs, radioulnar synostosis and carpal or tarsal fusions [25]. In the two Canadian siblings described by Silveira and colleagues, aged 11 and 15 years at publication, the abnormal head shape was detected on prenatal ultrasound; both had turribrachycephaly with a high forehead, triangular face, shallow orbits, prominent nose, high palate, malar hypoplasia, low-set posteriorly rotated ears with narrow ear canals, recurrent otitis and conductive hearing loss requiring hearing aids, marfanoid habitus with arachnodactyly, limited pronation-supination, episodic headaches and learning difficulties; CT excluded craniosynostosis and showed partial absence of the greater wings of the sphenoid and maxillary hypoplasia, and the sister underwent craniofacial reconstruction at 10 years [19]. Caliskan and colleagues described a 3-year-old Turkish child with craniosynostosis and dysmorphism without limb synostoses, showing that joint fusions are not universal [24].

Skeletal maturation varies: the fetus with p.Arg363Leu had advanced bone age, and a review of disorders of RA metabolism associates biallelic CYP26B1 loss of function with advanced bone age and premature epiphyseal closure, whereas in the Silveira siblings bone age was appropriate [19, 20, 53]. None of the described patients had genital anomalies or abnormalities of steroid metabolism [22].

4. Clinical Recognition and Warning Signs

The aim in a child with abnormal skull shape is to distinguish a postural deformation, which generally regresses, from syndromic or non-syndromic craniosynostosis [27]. In postural deformation, the entire face and skull appear to rotate around a central axis: forehead, eyebrow, cheek and ear move backwards on the affected side, with reciprocal contralateral occipital flattening, resulting in a parallelogram-shaped head with the ipsilateral ear displaced anteriorly; in unicoronal synostosis the head is trapezoidal and the ear on the affected side lies more anteriorly [27]. The AAP report adds that in occipital positional deformation the ipsilateral ear is displaced anteriorly, there is no harlequin orbit or retroauricular prominence, deformation is absent at birth in 80% of cases and 80% are right-sided; in true lambdoid synostosis the ear moves backwards and downwards and there is mastoid prominence [14]. The European guideline recommends the Bredero-Boelhouwer flowchart for primary and secondary care triage, with three questions: was the abnormal shape present immediately after birth? is there a preferred position? has the shape improved? [15]. In a Dutch study, 13% of 107 referred children had been misclassified in primary and secondary care, and the flowchart eliminated false negatives [15].

Relevant history includes a three-generation pedigree (considering variable penetrance of dominant disorders) and consanguinity; pregnancy (breech presentation, multiple pregnancy, oligohydramnios, persistent discomfort and a feeling of the “head being stuck”, uterine constraint); maternal hyperthyroidism; exposure to drugs such as valproate (metopic synostosis) or fluconazole (Antley–Bixler phenocopy); paternal age; age at presentation and progression; and neurodevelopmental abnormalities, including seizures [1, 27]. Obstructed labour suggests pre-existing synostosis [27].

On examination, head circumference is measured and skull shape described from above, the side and the front, palpating sutural ridges, assessing fontanelle position, size and tension, and seeking compensatory bossing; intercanthal distance is measured and exorbitism, proptosis and capacity for eyelid closure assessed; the combination of midface hypoplasia, beaked nose and exorbitism constitutes “crouzonoid” facies; the palate, ears (small with a prominent helical root in Saethre–Chotzen), hands and feet, and skin (acne, acanthosis nigricans, cutis gyrata) are examined, as are both parents, reviewing childhood photographs [1, 27]. Bony fusion of the digits is diagnostic of Apert syndrome, and broad thumbs or great toes, with medial deviation in severe cases, of Pfeiffer syndrome; mild cutaneous syndactyly occurs in fewer than half of patients with Saethre–Chotzen and Pfeiffer, polydactyly in Carpenter syndrome and elbow restriction in FGFR2 variants [1]. In Apert and Pfeiffer syndromes, the metopic and sagittal sutures remain open to allow brain growth, which may produce a large fontanelle [61].

5. Diagnosis: Criteria, Tests and Algorithms

5.1 Diagnostic criteria

There are no formal consensus clinical criteria for FGFR craniosynostosis syndromes, so the specific diagnosis should be confirmed by genetic testing [5]. For Apert syndrome, GeneReviews accepts diagnosis with classical features (multisuture craniosynostosis, midface retrusion and syndactyly) or suggestive features with a heterozygous pathogenic or likely pathogenic FGFR2 variant; a variant of uncertain significance neither establishes nor excludes diagnosis [4]. Muenke syndrome is defined by the FGFR3 variant c.749C>G, p.(Pro250Arg) and cannot be diagnosed without molecular testing, given overlap with other coronal synostoses [1, 6]. Saethre–Chotzen syndrome is established with typical findings and a heterozygous pathogenic TWIST1 variant; syndactyly, even subtle, is virtually diagnostic when associated with craniosynostosis, a low frontal hairline and small ears [40].

5.2 Imaging

Craniosynostosis may be suspected clinically and confirmed by skull radiography, showing loss of the suture line; computed tomography (CT) with a bone window and three-dimensional reconstruction provides greater precision in doubtful cases [1]. Radiographic signs are partial or complete suture absence, indistinct areas along the suture and perisutural sclerosis; in coronal synostosis, the harlequin sign (elevation of the lesser sphenoid wing) is sought and, if the clinical impression is synostosis, specialist opinion should be requested even with a radiograph reported as normal [27]. Digital impressions (“copper beating”) indicate longstanding elevated intracranial pressure, although normal markings on the inner vault surface are maximal at ages 3–4 [1]. In Apert syndrome, radiography shows increased vertical and reduced anteroposterior diameter, a short anterior fossa and deep middle and posterior fossae, elevation of the orbital roof and sphenoid wing (harlequin appearance), and maxillary hypoplasia [12].

The European guideline recommends referring a child with suspected craniosynostosis to the reference centre as early as possible, without additional investigations, because prior radiological examinations were the main factor delaying referral [15]. At the centre, skull ultrasound or radiography is performed in children at moderate risk and low-dose 3D CT, without prior examination, in those with high clinical suspicion [15]. Suture ultrasound had 100% sensitivity (95% confidence interval [CI]: 56.1–100%) and 100% specificity (90.2–100%) in two studies, and 100% sensitivity with 86% specificity in another, being usable until age 8–9 months and dependent on radiologist experience; 3D CT with low-dose techniques reduces effective dose to 0.02–0.05 mSv, comparable to radiography (0.01–0.04 mSv) [15]. 3D CT is strongly recommended for surgical planning because a diagnostic error would only be recognised after opening the scalp, and magnetic resonance imaging (MRI) adds value in syndromic forms for detecting brain anomalies, ICH signs and aberrant venous drainage; black bone MRI is a promising CT alternative but generally requires anaesthesia [15]. The AAP report discourages routine imaging during initial skull shape assessment because both single-suture synostosis and positional deformation are generally diagnosed clinically [14].

5.3 Genetic testing: indications and strategy

In the Oxford study, no non-syndromic sagittal (n = 27), metopic (n = 28) or lambdoid (n = 7) synostosis case tested positive, unlike 3 in 8 (37.5%) bicoronal and 7 in 40 (17.5%) unicoronal cases, mainly with FGFR3 p.Pro250Arg [38]. Nine of the 77 children with a syndrome had p.Pro250Arg and five had been misclassified as Crouzon or Saethre–Chotzen; no clinically significant chromosomal anomaly was found in 92 non-syndromic children without a mutation [38]. In a series of 204 individuals with apparently non-syndromic and non-familial single-suture synostosis cited by GeneReviews, unicoronal synostosis was syndromic in 17% (Muenke in 10%) and bicoronal synostosis had Muenke syndrome in 38%, without variants in patients with lambdoid, sagittal or metopic synostosis; a trigonocephaly study found no variants in FGFR1, CER1 or CDON [5]. The Oxford manual states that 30% of children with non-syndromic coronal synostosis have FGFR3 p.Pro250Arg and another 10–30% a TCF12 variant (10% unicoronal and 30% bicoronal), and, in another section, approximately 40% for the two genes combined [1, 27]. In the original TCF12 exome series, 38 heterozygous variants were found in 347 probands [41].

5.4 Test limitations and report content

Sequence analysis detects single-nucleotide variants and small insertions/deletions, but generally not exon or whole-gene deletions; yield varies by gene: approximately 99% of Apert syndrome by sequencing, 100% of Muenke syndrome (defined by the variant), 72% of Saethre–Chotzen by sequencing, with another 23% by targeted deletion/duplication or array and 5% by karyotype [4, 6, 40]. TWIST1 deletion breakpoints may not be characterised by targeted methods, and array resolution depends on probe density at 7p21 [40]. These percentages describe the proportion of variants detectable by each method among patients with an established diagnosis, not the clinical sensitivity of the test in an unselected child. In mosaicism, Sanger sequencing may fail: in craniofrontonasal syndrome, mosaic EFNB1 variants (15–69% mutant cells) in six severely affected males were detected only with more sensitive methods (denaturing high-performance liquid chromatography, massively parallel sequencing and multiplex ligation-dependent probe amplification, MLPA) [46]. In Muenke syndrome, parental leucocyte testing may not detect somatic mosaicism and does not detect exclusively germline mosaicism [6]. A negative panel result does not exclude diagnosis: TCF12 and ERF were only described in 2013 and panel genes vary between laboratories [5, 41, 43]. DNA banking should be considered when molecular diagnosis is unconfirmed [5].

The report must specify gene, transcript with version, Human Genome Variation Society (HGVS) nomenclature and zygosity. For Apert syndrome, GeneReviews uses NM_000141.4/NP_000132.3; VariantValidator confirms that c.755C>G, p.(Ser252Trp), c.758C>G, p.(Pro253Arg) and c.755_756delinsTT, p.(Ser252Phe) have the same numbering in MANE Select transcript NM_000141.5 (GRCh38 coordinates NC_000010.11:g.121520163G>C and g.121520160G>C for the first two) [4, 7]. In historical literature, these variants were published as C934G and C937G [4]. For FGFR3, the MANE Select transcript is NM_000142.5: c.749C>G, p.(Pro250Arg) (NC_000004.12:g.1801844C>G) and c.1172C>A, p.(Ala391Glu) (g.1804426C>A) [6, 7]. The Oxford manual writes the Crouzon with acanthosis nigricans variant as “pAla391Gln”, which does not correspond to the substitution described (alanine to glutamic acid, p.Ala391Glu) in the other sources and VariantValidator [1, 3, 7, 34]. For TWIST1, the MANE Select transcript is NM_000474.4 and for FGFR1 it is NM_023110.3, in which the Pfeiffer variant p.(Pro252Arg) corresponds to c.755C>G [7, 8]. According to American College of Medical Genetics and Genomics and Association for Molecular Pathology (ACMG/AMP) standards, likely pathogenic variants are equivalent to pathogenic variants for clinical decision-making, whereas a variant of uncertain significance neither establishes nor excludes diagnosis [4].

5.5 Assessment after diagnosis

In Muenke syndrome, initial assessment includes skull radiographs or preferably 3D CT, CT or MRI for hydrocephalus, audiology, developmental assessment (motor, adaptive, cognitive, language), ADHD screening after 12 months, electroencephalography if seizures are suspected, and ophthalmology with fundoscopy [6]. In Saethre–Chotzen, additional measures include height and growth velocity measurement (with endocrinology referral if short stature is present), investigation for lacrimal duct stenosis and papilloedema, cardiac assessment, and flexion–extension cervical radiography from age 2 before high-risk activities [40]. The European guideline recommends MRI in all patients with Crouzon syndrome and multisuture synostosis at referral, with repeat MRI if ventriculomegaly is present, and MRI for Chiari at the first visit in patients with Crouzon/Pfeiffer, multisuture synostosis with lambdoid involvement and non-syndromic unilateral lambdoid synostosis [15].

5.6 Prenatal diagnosis

Craniosynostosis is rarely recognised prenatally; cranial sutures do not develop before 16 weeks and single-suture synostosis is difficult to detect on ultrasound [1, 15]. Apert syndrome is increasingly diagnosed prenatally, with abnormal skull shape (including cloverleaf), hypertelorism, proptosis, agenesis of the corpus callosum and hand and foot syndactyly, and sometimes echogenic lungs and other visceral malformations [10]. In Pfeiffer syndrome types 2 and 3, the fetus presents with polyhydramnios, cloverleaf skull, ventriculomegaly, proptosis, hypertelorism, frontal bossing, nasal malformation and limb anomalies [30]. Three-dimensional ultrasound and CT and MRI help characterise suspicious findings; in a pregnancy without prior risk, testing for FGFR1-3 variants has low yield and does not clarify prognosis, which is determined by clinical findings [5]. In Western Australia (1980–2010), prenatal identification of rare craniofacial anomalies was rare (0.6%) [63].

5.7 Summary algorithm

1. Abnormal skull shape: apply the triage flowchart (present at birth? preferred position? improvement?) and clinically distinguish postural deformation (parallelogram, anteriorly displaced ear) from synostosis (trapezoid, ridges, harlequin orbit) [15, 27].
2. Suspected synostosis: refer to the reference centre without prior investigations; at the centre, ultrasound/radiography for moderate risk or low-dose 3D CT for high risk [15].
3. Confirmed synostosis: targeted dysmorphological examination (hands, feet, ears, skin, palate), examination of parents and a three-generation pedigree [1].
4. Evident phenotype: targeted test. Isolated coronal synostosis: FGFR3 p.Pro250Arg and TCF12. Without an evident phenotype: NGS panel with deletion/duplication analysis. Malformations or delay: array and panel or trio [1, 15].
5. Established diagnosis: systemic assessment (airway, eyes, hearing, cervical spine, brain), surgical plan and syndrome-specific surveillance protocol; genetic counselling with parental testing [5, 15].

5.8 Diagnosis of CYP26B1 deficiency

There are no formal criteria and no clinical biomarker: enzyme activity has only been measured in research assays (luciferase reporter and minigenes), and confirmation is molecular [19]. Suspicion arises from the association of craniosynostosis or deficient calvarial ossification with limb synostoses, especially of the elbow, hypoplasia or aplasia of the first ray and arachnodactyly, with normal genitalia and steroid profile [22, 23]. In the presence of humeroradial synostosis, the review by Leduc and colleagues proposes a full skeletal survey, platelet count, electrolytes and creatinine, cardiac and renal ultrasound when indicated, and CT when craniosynostosis is suspected; in the group with multiple synostoses and craniosynostosis, an undermineralised calvaria or first-ray aplasia points to CYP26B1, genital anomalies to POR, and craniosynostosis without these features to FGFR2 or FGF9 [23]. The steroid profile is useful in a negative sense, because it is abnormal in POR deficiency and no steroid abnormalities have been described in CYP26B1 deficiency [22].

Molecular testing requires the gene to be included in the panel or the use of exome sequencing. In the Canadian family of Silveira and colleagues, genomic array and targeted testing of FGFR1 p.Pro252Arg, the IIIa and IIIc domains of FGFR2, FGFR3 and TWIST1 were negative, and only quartet exome sequencing (parents and two siblings) identified the two variants in trans [19]. The missense variants were initially classified as of uncertain significance (criteria PP1, PM2 and PP3), and reclassification depended on research functional studies [19]. The synonymous variant c.1083C>A was recognised only through prediction of a new splice donor site and confirmed in a minigene assay, being reclassified as pathogenic (PVS1_strong, PS3, PM2 and PP4); annotation restricted to the coding consequence would have filtered it out [19]. A report should state the transcript NM_019885.4, the HGVS nomenclature, zygosity, the phase of the variants (confirmed by parental testing) and the ACMG/AMP class; a 2p13.2 deletion including the gene requires copy-number analysis [55].

Prenatal diagnosis of the severe forms is possible by ultrasound: in the siblings of the Laue family, the occipital encephalocele, short limbs with bowed femora and absent halluces were visible at 12 and 15 weeks [20]. In the Brazilian fetus, ultrasound at 31 weeks and 5 days showed polyhydramnios, marked hydrocephalus, short long bones, severe kyphosis, narrow thorax and reduced mineralisation, and the initial hypotheses were hypophosphatasia and osteogenesis imperfecta [19]. Once the familial variants have been identified, molecular prenatal diagnosis and preimplantation genetic testing are applicable.

6. Differential Diagnosis

The first distinction is between primary craniosynostosis, in which abnormal suture biology causes premature closure (isolated or syndromic), and secondary craniosynostosis, in which the suture is normal but external forces or deficient brain growth cause closure: in children with microcephaly due to deficient brain growth, all sutures fuse and the head is symmetrical and microcephalic [5]. Craniosynostosis occurs in more than 150 genetic diseases [5]. Postural deformation (positional plagiocephaly or brachycephaly) is the most common skull shape anomaly, estimated in 20–50% of 6-month-old children, more frequent in males and associated with torticollis in 15–20%; it is purely cosmetic, requires no surgery and improves with acquisition of motor milestones and alternating positions, while retaining supine sleeping [14]. According to the Dutch guideline, approximately 20% of infants have a preferred position in the first months [15]. Bathrocephaly (a prominent occiput due to a persistent mendosal suture) and dolichocephaly in hospitalised premature infants (“NICUcephaly”) require no treatment [14]. Early anterior fontanelle closure does not imply suture closure: the fontanelle normally closes between 4 and 26 months, and normal cranial growth with a normal shape argues against craniosynostosis; hyperthyroidism, hyperparathyroidism, hypophosphatasia and rickets are associated with premature closure [14].

AD, autosomal dominant; AR, autosomal recessive; ICH, intracranial hypertension.

Terminal 10q26 deletion may also be associated with craniosynostosis, in the context of microcephaly and intellectual disability.

With a “mitten” hand, Fraser syndrome (cryptophthalmos) and type IV syndactyly (Haas polysyndactyly) have similar hands but without craniosynostosis; Cenani–Lenz syndrome (LRP4, autosomal recessive) has hands and feet resembling Apert syndrome, with radial and ulnar hypoplasia and metacarpal fusion [12, 70]. In syndactyly classification, complete syndactyly of digits 2–5 of the hands and feet, sometimes with the thumb, is seen mainly in Apert syndrome [70]. Mild Apert and Saethre–Chotzen forms, or Pfeiffer with mild great-toe broadening, may require molecular differentiation; in Carpenter syndrome without postaxial polydactyly, distinction from Pfeiffer relies on more severe brachymesophalangy, a prominent metopic ridge with a shallow supraorbital rim, and molecular testing [62, 65]. In FGFR2-related Antley–Bixler syndrome, endocrine assessment and molecular testing are mandatory to exclude potentially fatal POR-associated adrenal insufficiency [52]. Muenke syndrome differs from Saethre–Chotzen in the absence of a low frontal hairline, marked ptosis, short ear, cutaneous syndactyly and broad great toes, and greater prevalence of developmental delay (35% versus 5%) and sensorineural hearing loss (34%, rare in Saethre–Chotzen), according to the GeneReviews differential table [40, 60]. The Apert syndrome GeneReviews classifies Carpenter syndrome as autosomal dominant, contradicting Smith, the bone dysplasia atlas, Oxford and the FGFR overview itself, which describe it as autosomal recessive [1, 4, 5, 64, 65].

CYP26B1 deficiency is distinguished from POR deficiency by the absence of genital anomalies, maternal virilisation and steroid profile abnormalities, and from the FGFR2-related Antley–Bixler form by recessive inheritance and the undermineralised calvaria [22, 23]. It is separated from Pfeiffer syndrome by hypoplasia or aplasia of the thumb and hallux, instead of broad and deviated first rays [23, 28]. Shprintzen–Goldberg syndrome was the initial hypothesis in the siblings of Silveira and colleagues, because of the marfanoid habitus and arachnodactyly [19]. The multiple synostoses syndromes (NOG, GDF5, FGF9, GDF6; autosomal dominant) cause progressive synostoses of the interphalangeal joints, elbow and spine with conductive hearing loss, but without an undermineralised calvaria, and only FGF9 is associated with craniosynostosis [23]. Other causes of humeroradial synostosis in the Leduc classification include Cousin syndrome (TBX15), the WNT7A-, DONSON-, ESCO2- and RBM8A-related forms and entities without a known molecular basis [23]. In the fetus with deficient mineralisation, hypophosphatasia and osteogenesis imperfecta enter the ultrasound differential [19]. Fluconazole embryopathy, caused by an inhibitor of P450 enzymes, is a phenocopy of Antley–Bixler syndrome, and exposure to excess RA reproduces the limb malformations and craniosynostosis in animal models [20, 52].

7. Treatments and Risk Management Strategies

7.1 Organisation of care

Treatment of craniosynostosis syndromes is surgical, supportive and multidisciplinary; there is no receptor-targeted therapy, and FGFR inhibitors are oncology drugs without an indication in germline disorders [5, 11]. The European guideline recommends treating children only in accredited reference centres, with a minimum team comprising a paediatrician, clinical geneticist, paediatric anaesthetist, paediatric intensivist, neurosurgeon, ophthalmologist, paediatric radiologist, plastic surgeon, maxillofacial surgeon, psychologist, social worker, speech and language therapist, educator and care coordinator, with an additional paediatric neurologist, orthodontist, otorhinolaryngologist, MRI and polysomnography for multisuture or syndromic forms; it recommends at least 20 intracranial operations per surgeon per year, annually reviewed protocols, annual internal audit and an annual outcomes report (complications, reoperations, ICH, apnoea, neurocognition and quality of life) [15]. In the Netherlands, syndromic surgery is centralised in a single centre [15]. The GeneReviews-recommended Apert syndrome team also includes a dermatologist, dietitian, hand and foot orthopaedic specialist, respiratory or sleep physician and spinal surgeon [4].

7.2 Cranial vault surgery

Source for recommendations in the middle column: [15].

Evidence is predominantly observational. In Apert and Crouzon, occipital expansion with distraction was associated, after more than 5 years, with greater head circumference increase (+1.09 versus +0.32 standard deviations) and lower prevalence of tonsillar herniation (3/11 versus 7/8 patients with Crouzon) and papilloedema (4/19 versus 11/18) than fronto-orbital advancement [15]. There appears to be no clinically relevant difference in ICH prevalence at age 5 between routine expansion in the first year and treatment only after ICH diagnosis: with the latter strategy, 83% of children with Apert and at least 61% with Crouzon eventually require surgery [15]. In Paris studies, 17% of children with Apert and 81% with Crouzon operated on before age 1 had normal IQ, versus 0% and 56% when operated on later, and in bicoronal synostosis mean IQ was 99 versus 89, with the bias that late referral increases surgical age [15]. Endoscopic craniectomy with helmet therapy in syndromic bicoronal synostosis had greater first-year reoperation risk (fronto-orbital advancement in 5 of 9 syndromic versus 1 of 9 non-syndromic children), although with less blood loss [15]. For Muenke syndrome, GeneReviews describes greater need for a second operation (7 of 29, 24.1%, versus 2 of 47, 4.3%, without the variant; p = 0.048), particularly for postoperative ICH or temporal bulging, although another series found fewer revisions and related revision need to younger age at initial surgery [6]. The European guideline, citing greater need for additional correction with early surgery and low ICH prevalence (4%), recommends postponing surgery in Muenke syndrome until 9–12 months, differing from the 3–6 months given by GeneReviews [6, 15]. GeneReviews notes that infants with Apert syndrome may have minimal physiological reserve and that later operations provide more stable bony correction [5].

7.3 Perioperative care

The European guideline recommends tranexamic acid during surgical correction to reduce blood loss, considering intraoperative blood salvage (cell saver) and administering fresh frozen plasma and/or fibrinogen as soon as coagulopathy signs appear [15]. The meta-analysis of three randomised trials cited by the guideline showed reduced transfused volume with tranexamic acid (58 versus 133 mL; 185 versus 258 mL; 376 versus 655 mL), without consistent information on unit numbers; the guideline specifies no dose, so this article does not provide one [15]. Controlled hypotension (mean arterial pressure between 55 and 65 mmHg) did not reduce transfusion requirements, and the value of vitamin K1 is uncertain [15]. Anaesthetic risks are greater in Apert syndrome, particularly upper airway obstruction (6.1%), and anaesthetic assessment is recommended before any intervention, with cervical precautions (spinal surgery consultation for positioning) and, in cervical fusions, possible fibreoptic intubation [4, 5]. Intracardiac shunts increase embolism risk during cranial surgery, so prior closure may be considered [5].

7.4 Midface, airway and sleep

Obstructive sleep apnoea occurs in approximately 70% of children with multisuture or syndromic synostosis, with greater prevalence and severity in Apert, Crouzon and Pfeiffer syndromes; central apnoeas (approximately 4%) decrease with age [15]. Only moderate-to-severe apnoea was associated with ICH and altered sleep architecture, so mild apnoea is treated only if symptomatic, favouring less invasive interventions such as adenotonsillectomy, which is rarely definitive treatment [15]. In a series of 29 children (18 Apert, 11 Crouzon), the apnoea–hypopnoea index did not change significantly after adenotonsillectomy and apnoea persisted in 11 of 13 reassessed children [15]. The guideline recommends airway endoscopy for moderate or severe apnoea to identify obstruction levels (nose, nasopharynx, oropharynx and hypopharynx, larynx, trachea), and considering Le Fort III or monobloc advancement, if necessary combined with mandibular advancement [15].

GeneReviews recommends avoiding continuous positive airway pressure (CPAP) or bilevel pressure (BiPAP) as long-term treatment because pressure on the maxilla worsens midface retrusion; the European guideline permits CPAP or BiPAP as an interim solution, recognising difficulty fitting the mask, the risk of injury to prominent eyes from airflow, and interference of occipital straps with occipital expansion [5, 15]. Temporary measures include nasopharyngeal tubes or nasal stents, intubation and tracheostomy, sometimes with ventilation; the guideline discourages nasopharyngeal tubes as a prolonged bridge because of frequent changes and septal necrosis risk, and in children below age 2 with tongue-base obstruction a tracheal cannula may be considered, postponing monobloc with mandibular distraction until ages 3–4 [5, 15]. Placement and care of tracheostomies require extreme caution with tracheal cartilaginous sleeves because of abnormal healing and granulation tissue formation [5]. Nasal cannula oxygen is sometimes beneficial [5].

7.5 Hydrocephalus, Chiari and ICH

The European guideline recommends treating ICH according to the predominant causal factor (insufficient intracranial volume, moderate-to-severe apnoea, hydrocephalus or venous hypertension); cranial expansion corrects insufficient volume and appears to improve venous hypertension [15]. Hydrocephalus may be treated by cranial expansion with or without foramen magnum decompression, ventriculoperitoneal shunting or endoscopic third ventriculostomy, choosing case by case according to MRI; shunting opposes the aim of cranial expansion and frequently makes the patient lifelong shunt-dependent, and ventricular enlargement after each expansion must be distinguished from true hydrocephalus [15]. In London, 9 of 30 children with Crouzon treated for hydrocephalus received a shunt and 6 still required cranial expansion; in Rotterdam, of 19 children with Crouzon treated first by expansion, only one required a shunt [15]. Chiari malformation should be operated on only if symptomatic, with annual neurological surveillance by a neurosurgeon or paediatric neurologist and spinal MRI for syringomyelia if herniation increases or becomes symptomatic; 17–50% of patients with Crouzon/Pfeiffer or multisuture synostosis with Chiari become symptomatic [15]. Stable ventriculomegaly, present in more than half of children with Apert, rarely requires surgery [5].

7.6 Eyes, hearing, speech, teeth and limbs

Ocular treatment aims to detect and treat amblyopia early, perform decompression before optic atrophy and protect the cornea: early topical lubrication if eyelid closure is inadequate, tarsorrhaphy for globe subluxation or prevention of exposure keratopathy, and strabismus treatment by an ophthalmologist experienced in craniosynostosis [4, 5]. In Saethre–Chotzen, ptosis and strabismus should be corrected early (occlusion or surgery), and papilloedema detection leads to consideration of cranioplasty [40]. In Muenke syndrome, strabismus correction is frequently delayed because cranial surgery takes priority, but early correction should be considered to obtain binocularity [6]. Hearing loss is treated with tympanostomy tubes for chronic effusions, hearing aids, bone-conduction processors, tympanoplasty and atresia correction; in Muenke syndrome, with predominantly sensorineural loss, cochlear implants may be needed [5, 6]. Cleft palate is repaired before pressure consonants develop and, in Saethre–Chotzen, usually after cranioplasty; craniofacial-experienced speech and language therapy assesses articulation, resonance and velopharyngeal insufficiency, which may also change with orthognathic surgery [4, 5, 40]. An external dentist or orthodontist should never treat a patient with syndromic craniosynostosis without consulting the craniofacial team [15].

In Apert syndrome, syndactyly surgery type and timing depend on thumb syndactyly and soft tissue deficiency; the aim is to improve function with the fewest operations, in early (separations) and late (functional osteotomies) stages, and most require multiple procedures and revisions throughout childhood, with physiotherapy and occupational therapy before and afterwards [4, 10]. Progressive foot deformity requires orthopaedics, and scoliosis is treated according to usual standards [4].

7.7 Pharmacological therapy, regulatory status and trials

For refractory acne in Apert syndrome, GeneReviews indicates that oral isotretinoin may be more effective than standard therapies and cites biological models supporting a role in androgen regulation and FGFR2 signalling; this is limited evidence rather than a controlled trial [4]. Parents should avoid factors increasing hearing loss, including ototoxic drugs and loud noise [4].

There are no registered therapeutic trials for these syndromes; GeneReviews directs readers to ClinicalTrials.gov and the European registry, cautioning that trials may not exist [4, 5]. The ClinicalTrials.gov search in October 2026 identified no therapeutic trials in these syndromes: returned records were oncology trials with FGFR inhibitors (matching terms), a completed surgical trial in isolated sagittal synostosis (NCT04721769) and observational studies [72]. NAUTILUS (NCT07535372; Fondazione Policlinico Universitario Agostino Gemelli, Rome, funded by AFM-Téléthon), recruiting since 20 April 2026, aims to design personalised antisense oligonucleotides to silence gain-of-function alleles or restore expression in loss-of-function alleles, testing them in suture cells obtained as surgical waste from 12 children aged 0 to 5 with a molecular diagnosis (FGFR1-3, TWIST1, TCF12, EFNB1, ERF, MSX2, ALX4), with nanoparticle delivery in hydrogel and mouse validation; this is preclinical research in human material, without administration to patients [72]. In the Muenke mouse model, FGF10 reduction rescued cochlear function, remaining animal research [6]. Recent clinical studies are surgical and observational: a quantitative Apert syndrome severity scale (airway, intracranial pressure, eyes, midface retrusion and turribrachycephaly domains, score 5–15) applied retrospectively to 37 patients followed to a mean age of 17 correlated with operation number (r = 0.851), and a score of 9 or more predicted six or more procedures with 100% sensitivity and 86.7% specificity [73]. These results come from a single centre without external validation and should not guide individual decisions in isolation.

7.8 Pregnancy in affected women and situations to avoid

There are no published studies of pregnancy in women with Apert syndrome; monitoring for worsening obstructive apnoea and anaesthetic assessment before labour are recommended to identify airway and vertebral anomalies affecting anaesthetic techniques, with possible need for fibreoptic intubation and planning of delivery location and timing [4]. Children with cervical anomalies may be at risk of spinal cord injury during hyperextension and should avoid contact sports and activities involving cervical hyperflexion or extension until spinal assessment; those with exophthalmos may need eye protection during ball sports [4, 5].

7.9 CYP26B1 deficiency: management

There is no targeted therapy, and a ClinicalTrials.gov search for “CYP26B1” and “radiohumeral synostosis” (November 2026) returned no studies [74]; management follows the principles for syndromic craniosynostoses in the European guideline (reference centre, correction of the synostosis when indicated, surveillance for ICH, airway, hearing and development), adapted to the extent of the anomalies [15, 23]. In newborns with a narrow thorax, respiratory failure is a described cause of death and justifies neonatal respiratory assessment [23]. Conductive hearing loss, associated with narrow ear canals and recurrent otitis, required hearing aids in the two siblings of Silveira and colleagues, and reconstructive craniofacial surgery was performed at 10 years in one of them [19]. Humeroradial synostosis limits elbow function; the consulted sources do not describe outcomes of orthopaedic surgery in these patients [23].

Because the disease mechanism is local RA accumulation, and vitamin A excess and retinoids cause craniosynostosis and premature epiphyseal closure, additional exposure to systemic retinoids (isotretinoin, acitretin) and to vitamin A supplements above nutritional requirements lacks justification in these patients [20, 53]. This is an inference from the mechanism, not evaluated in clinical studies; none of the consulted sources makes a dietary recommendation, and there is no drug dose to state because no pharmacological therapy exists for the disease.

8. Reproductive Options and Family Counselling

8.1 Inheritance and risk to relatives

FGFR craniosynostosis syndromes are autosomal dominant. In milder forms (Muenke, Crouzon, Pfeiffer, Jackson–Weiss), transmission from an affected parent is common, whereas in more severe forms (bent bone dysplasia) de novo variants predominate [5]. In Crouzon syndrome, approximately one-quarter of reported cases result from a new mutation [3]; in Apert syndrome, the vast majority are sporadic, although the exact proportion is unknown [4, 10]; in Muenke syndrome, 33 of 51 informative probands (64.7%) inherited the variant, and de novo variants appear exclusively paternal and associated with advanced paternal age [6, 57]; in Saethre–Chotzen, more than half of cases are familial [29].

After a de novo FGFR variant, recurrence risk is substantially below 1% because most correspond to sporadic events associated with advanced paternal age; after a de novo TWIST1 variant, risk is unknown and germline mosaicism should be considered [1]. Family history may appear negative because of lack of recognition or reduced penetrance, so an apparently negative family history is confirmed only after molecular and clinical–radiological assessment of parents [5]. In Muenke syndrome, Saethre–Chotzen syndromes and ERF- and TCF12-associated forms, marked variability requires careful parental assessment supported by molecular testing [1]. In the Oxford series, 84 genetic cases led to identification of 64 affected relatives in 36 families, demonstrating the need for extended counselling [38]. Lajeunie and colleagues, in a French series of 260 probands with non-syndromic coronal synostosis, found familial aggregation in 14.4% and a dominant model with penetrance of 0.60, with significantly increased mean paternal age (32.7 years) [75].

The phenotypic category is usually constant within each family (for example, siblings of a proband with Pfeiffer will have a Pfeiffer phenotype), but severity may vary greatly, including sutures fused at birth and respiratory support required; rare families have members with Pfeiffer, Jackson–Weiss or Crouzon phenotypes [5]. In Muenke syndrome, uni- or bilateral synostosis and absence of synostosis may coexist within a family, and a child may be more or less affected than the parent [6]. GeneGuidance recommends formal audiological assessment of all p.Pro250Arg carriers, even with normal skull shape, and parental testing before assuming a de novo variant [11].

8.2 Paternal age in counselling

The paternal age effect is relevant to explaining variant origin but, without a family history, does not justify additional paternal-age-specific testing; men with a child affected by a genetic disease should receive counselling on risk in a subsequent pregnancy, which for age-related mutations is substantially below 1% [48]. The mean paternal age increase in these diseases is only 3 to 5 years, and many affected children are born to young fathers [48]. This counselling should distinguish slightly increased population risk with paternal age from recurrence risk after an affected child, which depends on germline mosaicism.

8.3 Prenatal and preimplantation diagnosis

Once the familial variant is known, prenatal diagnosis (by chorionic villus sampling) and preimplantation genetic testing are possible; severity is difficult to predict and detection of an FGFR variant does not predict occurrence or severity of manifestations in most phenotypes [1, 5]. Prognosis in historically severe syndromes is difficult to predict because surgical treatment has changed expected outcomes, as shown by aggressively treated children with p.Trp290Cys [5]. Without a definitive prenatal diagnosis excluding an affected fetus, Oxford recommends ultrasound at 36 weeks to exclude cephalopelvic disproportion before delivery [1]. The European guideline recommends referring a pregnant woman with ultrasound-suspected craniosynostosis to a tertiary centre for prenatal diagnosis and, once multisuture or syndromic synostosis is confirmed, transferring obstetric care to the reference centre because of high risk of non-spontaneous delivery and neonatal respiratory problems, with counselling by a geneticist, surgeon (plastic, neurosurgeon or maxillofacial) and obstetrician [15]. GeneReviews recommends delivery in a hospital with immediate access to paediatric otorhinolaryngology, plastic surgery, neurosurgery and respiratory medicine, and a professional experienced in intubation and resuscitation in the delivery room, given the frequency of choanal atresia [5]. The ideal time to discuss genetic risk and reproductive options is before pregnancy, and affected or at-risk young adults should receive counselling [5]. Professionals and families may differ in perspective regarding use of prenatal diagnosis for pregnancy termination, which is a personal decision [5].

8.4 Testing in at-risk relatives

Clarifying the genetic status of apparently asymptomatic older or younger relatives is justified to identify those benefiting from early surveillance and intervention, through familial variant testing or clinical and radiological assessment, recognising that manifestations may not be evident [5]. In Muenke syndrome, before testing siblings or other relatives, risks, benefits and limitations should be discussed with the genetic counsellor; in school-age individuals without delay, hearing loss, craniosynostosis or other signs, the probability of Muenke syndrome is low, although apparently unaffected carriers exist [6].

8.5 CYP26B1 deficiency

In autosomal recessive inheritance, the parents of a proband are obligate heterozygotes and, in each pregnancy, the risk is 25% for an affected child, 50% for a carrier and 25% for a non-carrier [19, 20]. Consanguinity is frequent in the described families: first cousins in the Laue family and in the Indian family of Sarma, consanguineous parents in the Turkish family and a coefficient of inbreeding of 1/32 in the Brazilian family [19–21]. Severity was similar between siblings (three lethal fetuses in the Laue family, two siblings with comparable presentations in the Canadian family), which supports the use of the proband's phenotype in counselling, with the caveat of a small series [19, 20]. Carrier testing can be offered to the parents' siblings and, in consanguineous unions, to the partner; with the known variants, molecular prenatal diagnosis, targeted ultrasound (which detected the lethal forms at 12–15 weeks) and preimplantation genetic testing are applicable [20]. The Brazilian couple had a healthy child four years later [19]. If the haploinsufficiency hypothesis from the Sarma family is confirmed, carriers of truncating variants would have a risk of their own and would transmit the phenotype with a probability of 50%; until confirmation, information given to carriers should make this uncertainty explicit [19, 21].

9. Special Considerations: Population Diversity and Epidemiology

9.1 Craniosynostosis in general

Frequency estimates depend on case definition and ascertainment. Oxford reports 1 in 2,500 live births [1, 27]; Lajeunie's French series estimated 1 in 2,100 children [75]; and ERN CRANIO gives 1 in 2,100 to 1 in 2,500 births [16]. In the Atlanta surveillance programme (1989–2003), 281 children had craniosynostosis, a birth prevalence of 4.3 per 10,000 births, of which 84% were isolated, 7% had multiple defects and 9% were syndromic; Apert syndrome accounted for 40% of syndromic cases and sagittal synostosis for 39% of non-syndromic cases, and maternal age 35 or older, multiple pregnancy, male sex and birth weight above 4,000 g were risk factors [13]. Orphanet records, for craniosynostosis, 24.3 per 100,000 in European EUROCAT surveillance (2012, not validated), 47.6 in France, 49.4 in Australia and 43.0 in the United States (validated) [2]. In Western Australia (1980–2010), metopic synostosis prevalence increased over time (annual percentage change 5.59), and rare craniofacial anomalies were more likely in premature infants, twins, children with growth restriction, offspring of older parents, pregnancies following fertility treatment and mothers with epilepsy, diabetes or hypothyroidism [63]. The AAP report also notes a possible recent increase in metopic synostosis without increases in the others [14].

The syndromic proportion depends on the series: 15% with a recognisable syndrome and another 10% with a genetic cause (Oxford), 21% with a confirmed genetic cause in the Oxford surgical cohort (1998–2002) and 9% syndromic in Atlanta [1, 13, 38]. These proportions are not directly comparable, because one is a surgical cohort with molecular testing and the other a population-based registry with clinical classification.

9.2 Specific syndromes

Discrepancies between countries, for example in Crouzon syndrome, exceed an order of magnitude and reflect differences in ascertainment and measure type (birth or point prevalence) [2, 11]. The Muenke syndrome estimate of 1 in 30,000 derived from the Oxford cohort, assuming craniosynostosis prevalence of 1 in 2,100 and 80% penetrance [38]. Muenke syndrome frequency refers to p.Pro250Arg and should not be transferred to all coronal synostoses, just as overall craniosynostosis frequency should not be attributed to a single gene.

9.3 Population data and special populations

The consulted sources identify no founder effect for these syndromes, whose mutational recurrence is explained by spermatogonial selection rather than common ancestry [50]. In the Apert case reported in Nanjing, the authors found only p.Ser252Trp and p.Pro253Arg in published Chinese series, and in the familial Crouzon series from Hebei almost all FGFR2 variants described in Chinese patients had already been identified in European and North American cohorts [9, 77]. In Atlanta, maternal race was analysed, but significant risk factors were maternal age, plurality, sex and birth weight [13]; in a US study cited by the European guideline, belonging to a racial minority was associated with late referral [15]. Regarding Portugal, GeneReviews cites a Muenke syndrome case with osteochondroma described by the Centro de Genética Médica Doutor Jacinto Magalhães (Porto), adding osteochondroma to the syndrome spectrum [5, 78]. Regarding Brazil, GeneReviews cites a large trigonocephaly cohort investigated with participation from the Universidade de São Paulo, without pathogenic variants in FGFR1, CER1 or CDON, and the European guideline cites a review published in Revista Paulista de Pediatria on diagnosis of synostotic and non-synostotic cranial deformities [5, 15, 79]. No national prevalence estimates for Portugal or Brazil were found in the consulted sources.

Muenke syndrome has a greater proportion of unicoronal synostosis in males in one series and a tendency towards greater severity in females in another, but the multicentre study of 106 patients found no sex difference in penetrance [5, 6, 60]. In craniofrontonasal syndrome, heterozygous females are more severely affected than hemizygous males, an exception explained by cellular interference resulting from random X inactivation; severely affected males are mosaic [45, 46]. Transition to adult care is a critical point: the Dutch GROW UR GENES study (NCT04463316) includes Saethre–Chotzen syndrome among syndromes whose adult medical needs are being characterised [72].

For CYP26B1 deficiency there are no population data: the published patients come from isolated families reported by groups from New Zealand and Europe (the Laue family, of unstated origin, and the Turkish infant), the United Kingdom, the United States, Canada (a non-consanguineous family of European ancestry), Brazil (a consanguineous couple from a town in the South settled by Polish immigrants), India, China and Turkey, with no recurrent variant or founder effect identified [19–21, 24, 25, 28, 54]. The Brazilian variant was absent from the gnomAD and ABraOM (Brazilian exome database) databases [19].

9.4 Reference networks and patient associations

The relevant European network is ERN CRANIO, launched in March 2017, with 36 member hospitals in 15 European Union/European Economic Area countries, endorsing the European craniosynostosis guideline and its patient version and providing explanatory videos of surgical techniques [16]. Patient organisations cited in the sources include Headlines Craniofacial Support (United Kingdom) [1], LAPOSA (the Dutch organisation for patients and parents with craniofacial anomalies) [15], and, in the United States and internationally, Born a Hero, Children's Craniofacial Association, American Cleft Palate-Craniofacial Association, FACES: The National Craniofacial Association and Face Equality International [4, 5, 40].

For the forms with skeletal dysplasia, such as CYP26B1 deficiency, ERN BOND, the European Reference Network for rare bone diseases, is also relevant [26].

10. Illustrative Clinical Case

10.1 Apert syndrome due to a de novo p.Pro253Arg variant (Li et al., 2018)

A 4-day-old male newborn was referred for genetic consultation at a prenatal diagnosis centre in Nanjing because of craniosynostosis and hand and foot syndactyly [77]. He was the first child of healthy, non-consanguineous Chinese parents aged 25 (mother) and 30 (father), born at 38 weeks and 2 days by caesarean section after an uneventful pregnancy, with weight 3,500 g, length 52 cm and Apgar scores 9 and 10 [77]. Examination showed syndactyly with hand contractures and severe toe syndactyly (Figure 1 of the article), craniosynostosis, midface retrusion, a high, broad forehead, increased head circumference, marked nasal root depression, a short, broad nose and proptosis (Figure 2), with slightly low-set ears and no cleft lip or palate; the remaining examination was normal [77]. Amplification and Sanger sequencing of FGFR2 exons 7 and 9 identified c.758C>G, p.(Pro253Arg), absent in both parents (Figure 3) [77].

The case illustrates several points of this article. The hand and foot phenotype enabled neonatal clinical diagnosis, and severe syndactyly without cleft palate is consistent with the correlation described for p.Pro253Arg, although this correlation is not consistent across all studies [4, 77]. The variant, absent in parental leucocytes, was interpreted as de novo; the authors do not mention confirmation of parentage, which GeneReviews lists as an alternative explanation to consider [4, 77]. The father was 30, highlighting that the paternal age effect is a population phenomenon and many cases are born to young fathers [48]. Parental recurrence risk is slightly above population risk because of possible germline mosaicism, and the patient's own transmission risk will be 50% [4]. The report does not describe progression or treatment; subsequent management should follow the multidisciplinary protocol in section 7, including assessment of the airway, cervical spine, ventricles, eyes and hearing.

10.2 Familial Crouzon syndrome with optic atrophy (Shi et al., 2021)

A 28-year-old man, assessed in 2017 at a Hebei ophthalmological hospital for progressive visual deterioration over the preceding six months and poor right-eye acuity since childhood, had craniofacial dysostosis, mandibular prognathism, hypertelorism, proptosis (exophthalmometry 25 mm on the right and 20 mm on the left), a short upper lip, scoliosis and thoracic deformity, with normal hands and feet (Figure 1) [9]. Best-corrected acuity was 0.05 on the right and 0.4 on the left; the fundus showed pale discs, and fluorescein angiography, optical coherence tomography and visual evoked potentials confirmed bilateral optic atrophy with myelinated retinal nerve fibres; skull radiography showed deepened, widened digital impressions and orbital CT showed tortuous optic nerves (Figure 2) [9]. His mother had a similar phenotype, with strabismus and ptosis. Sequencing of FGFR2 exons 8 and 10 identified the heterozygous variant c.1026C>G, p.(Cys342Trp) in both, absent in seven unaffected relatives (Figure 3) [9]. VariantValidator confirms this nomenclature in MANE Select transcript NM_000141.5 [7]. The authors describe the alteration as a transition, whereas substitution of cytosine by guanine is a transversion.

This case demonstrates vertical transmission with variable expressivity and the outcome surveillance seeks to prevent: visual loss from optic atrophy in an adult with radiological signs consistent with chronic intracranial hypertension. No intervention was described. Attribution of optic atrophy to p.Cys342Trp is the authors' hypothesis, which they acknowledge is premature as a correlation; the mechanism most consistent with this article's sources is prolonged ICH, which may cause papilloedema and optic atrophy when not treated promptly and aggressively [5, 9].

10.3 Lethal CYP26B1 deficiency due to a synonymous splicing variant (Silveira et al., 2023)

The proband of family 2 of Silveira and colleagues resulted from the first pregnancy of a healthy, consanguineous Brazilian couple (coefficient of inbreeding 1/32), of European ancestry and from a town in southern Brazil settled by Polish immigrants, with no family history of malformations [19]. Ultrasound at 31 weeks and 5 days showed polyhydramnios, marked hydrocephalus (head circumference 35.1 cm, z = +4.92), an elliptical skull, agenesis of the corpus callosum, absent posterior fossa structures, reduced abdominal circumference (z = −3.47), long bones below the 2.5th centile, severe kyphosis, narrow thorax with pulmonary hypoplasia, reduced mineralisation, bowing of the humeri, radii and ulnae, apparent absence of the femora, short tibiae, and hands and feet with missing digits; the sonographer considered hypophosphatasia and osteogenesis imperfecta [19]. The female fetus was stillborn by caesarean section at 33 weeks, with an occipitofrontal circumference of 41.0 cm (z = +7.57), dolichocephaly, a very small low-set right ear, mild ocular protrusion, tongue protrusion, short neck, thoracic kyphosis, short limbs, oligodactyly with four digits including the right thumb, and feet with a single median toe (Figure 2 of the article) [19].

The radiographic study showed generalised osteopenia, an unossified vault, absent clavicles, hypoplastic scapulae, generalised spina bifida occulta, severe thoracic kyphosis with tall vertebrae, short thin ribs, short humeri with a right diaphyseal fracture, elbow synostosis, four metacarpals, a thumb with a single phalanx, severely hypoplastic femora, bowed tibiae and fibulae, and three metatarsals [19]. Autopsy confirmed hydrocephalus with cerebral atrophy, skull base deformity, spina bifida occulta from C1 to S1 with an apparently normal spinal cord, and hypoplasia of the lower pulmonary lobes; the femoral growth plate showed hypercellularity of the resting zone and short proliferative and hypertrophic zones [19]. Sanger sequencing of the CYP26B1 exons in the parents showed both to be heterozygous for c.1083C>A in exon 5, and sequencing of exon 5 in formalin-fixed paraffin-embedded fetal tissue confirmed homozygosity; the variant, synonymous at protein level, was absent from population and clinical variant databases, created a new splice donor site and, in the minigene assay, removed 66 nucleotides of exon 5, with an in-frame deletion of 22 amino acids that removes the ExxR motif, p.(Val361_Asp382del); the figure legend of the article refers to 66 amino acids, in disagreement with the text and with the size of the deletion [19]. The variant was reclassified as pathogenic, and four years later the couple had a healthy child [19].

The case illustrates three points. A synonymous variant can cause a severe recessive disease and escapes filters restricted to the coding consequence. The lethal CYP26B1 phenotype overlaps on ultrasound with dysplasias with deficient mineralisation, and the elbow synostosis and first-ray oligodactyly point to the diagnosis. The strategy of sequencing the parents first, given limited fetal material, enabled the retrospective diagnosis and counselling for the next pregnancy.

11. Prognosis and Long-Term Follow-up

11.1 Prognosis

Long-term complications include restenosis of an operated suture, ICH, deficient midface growth, sleep apnoea, exposure keratitis, hearing loss, malocclusion and, rarely, syringomyelia; patients should remain under craniofacial team care until facial and cranial growth ends [1]. In the Oxford surgical cohort (mean follow-up 8 to 9.5 years depending on group), children with a confirmed genetic diagnosis had more major procedures (mean 1.74 versus 1.13 in non-syndromic cases), more than one major procedure in 52% (versus 12%), and more ventricular shunts, secondary ICH (26% versus 9%), tracheostomies, ptosis corrections, hearing aids, tympanostomy tubes and upper limb operations; midface surgery, shunting and tracheostomy requirements were associated mainly with FGFR2, and ptosis, hearing aid and upper limb surgery requirements with Saethre–Chotzen, Muenke and Apert syndromes, respectively [38]. Repeat surgery was required in 58% of children with a monogenic mutation and 17% with a chromosomal anomaly; a developmental quotient below 70 occurred in 26% of genetic cases, particularly with chromosomal anomalies (8 of 11) and Apert syndrome (5 of 9); the six cohort deaths (1.8%) occurred in syndromic children, two with FGFR2-related Pfeiffer syndrome, and four were directly attributed to multisuture craniosynostosis complications [38].

In Apert syndrome, childhood mortality is increased, particularly from upper and sometimes lower airway compromise (tracheal sleeves); many patients do well long term despite residual functional limitations, and multiple operations are usually necessary [10, 12]. In a follow-up study of 24 British patients described by Blank in 1960, traced through the National Health Service central registry until 2013 (53 years), 12 (50%) were alive without recorded cancer, two were untraced and ten had died, four from neoplasms; the authors conclude that Apert syndrome is not associated, in many individuals, with substantial lifespan shortening or high risk of specific cancers, although the same variant is oncogenic when somatic (enriched in endometrial carcinoma) [80]. This is a small historical series, consulted through the abstract, that cannot estimate absolute risks. Adults with Apert syndrome have greater social and relationship development difficulties than controls and adults with Crouzon syndrome [4]. Most adults with Crouzon syndrome report satisfactory quality of life and many are fully independent [3, 5]. In Pfeiffer type 1, long-term prognosis is excellent, whereas types 2 and 3 may cause early death [8, 30]. In Muenke syndrome, the course is generally benign, although feeding and swallowing difficulties may occur in early childhood, and many adults are cognitively normal [56, 60]. In Saethre–Chotzen, facial appearance tends to improve during childhood [29]. Parent-reported health-related quality of life is below the norm in syndromic or multisuture forms, particularly in vision, hearing and speech, and in Apert syndrome also in physical functioning, emotional impact, family activities and cognition; parents have reduced psychosocial quality of life and approximately 10% of children admitted to intensive care and their parents develop post-traumatic stress symptoms [15].

11.2 Intracranial hypertension surveillance

ICH may cause irreversible visual loss and there is no consensus on normal intracranial pressure in children [15]. Preoperative prevalence is probably 9–83% in Apert syndrome, 53–64% in Crouzon, 19–43% in Saethre–Chotzen and 0–4% in Muenke, and postoperatively 35–45%, 20–47%, 17–42% and 0–5%, respectively, with 58–67% in multisuture synostoses and approximately 31% in bicoronal synostosis [15]. Fundoscopy is specific but has only 22% sensitivity below age 8, so absence of papilloedema does not exclude ICH; optical coherence tomography (OCT) had 60% sensitivity and 90% specificity for predicting papilloedema, and 89% and 62% against invasive measurement, but requires cooperation (usually from age 3–4); diffuse radiographic impressions have 0% sensitivity below 18 months (98% specificity), and optic nerve ultrasound has 11% sensitivity [15]. Slowing of the head circumference curve is a useful predictor in Apert and Crouzon syndromes, but ICH from apnoea or other causes requires fundoscopy or OCT [15].

ICH is more severe and frequent in Crouzon syndrome, followed by Apert and Saethre–Chotzen, justifying syndrome-specific protocols in the first years and prolonged surveillance in syndromic forms, in which other ICH causes may arise later [15]. In insidious-onset ERF-associated craniosynostosis, surveillance should continue well beyond childhood [11, 44]. Parents should urgently contact the team for unexplained persistent vomiting, headache or head circumference change; papilloedema may occur before and after decompression [4].

11.3 Multisystem surveillance plan

A recent US consensus from the SynRG research group proposed a neurodevelopmental screening schedule from infancy to adolescence, with instruments such as the Ages and Stages Questionnaire, NIH Toolbox and PROMIS measures, aligned with developmental milestones and school transitions; the authors acknowledge barriers such as neuropsychologist availability and funding [81].

11.4 Transition to adulthood

Comprehensive care at a specialist craniofacial centre with psychosocial support and holistic transition planning may improve quality of life in adults with Apert syndrome [4]. Midface surgery may be performed from age 17 if no early indication existed, and in adulthood nasal septal surgery may improve nasal patency and apnoea-related complaints [15]. Reproductive counselling of young adults should include the 50% risk to offspring and prenatal and preimplantation diagnostic options [5].

11.5 CYP26B1 deficiency: prognosis and surveillance

Prognosis depends on residual enzyme activity: the most severe genotypes cause fetal or neonatal death, whereas patients with non-lethal forms reach adolescence and adulthood (a 22-year-old patient; siblings aged 11 and 15 years), with intellectual disability or learning difficulties and conductive hearing loss in most [19, 23, 25, 28]. The Turkish infant died at 5 months of unknown cause [20]. There are no specific guidelines; the following table applies the protocol for syndromic craniosynostoses (sections 11.2 and 11.3) and adds the domains specific to the entity.

12. Summary and Final Considerations

Craniosynostosis affects approximately 1 in 2,100 to 1 in 2,500 children, and approximately one-fifth to one-quarter of cases have an identifiable genetic cause; the involved suture guides reasoning because coronal and multisuture synostoses concentrate molecular testing yield, whereas isolated sagittal and metopic synostoses rarely have a monogenic cause [1, 38]. Clinical distinction between synostosis and postural deformation relies on head shape viewed from above (trapezoid versus parallelogram), ear position and progression, and the European guideline recommends early reference-centre referral without prior investigations [15, 27].

Apert and Crouzon syndromes are allelic FGFR2 disorders. Apert syndrome, almost always de novo and caused by p.Ser252Trp or p.Pro253Arg, is distinguished by bony hand syndactyly and carries a greater burden of brain malformations, cervical fusions, multilevel airway obstruction and intellectual disability; Crouzon syndrome, frequently familial, has normal limbs and greater propensity for hydrocephalus, chronic tonsillar herniation and ICH [4, 5]. Crouzon syndrome with acanthosis nigricans is a distinct FGFR3 entity (p.Ala391Glu), and Muenke syndrome (FGFR3 p.Pro250Arg) and Saethre–Chotzen syndrome (TWIST1) should be excluded in any coronal synostosis, together with TCF12 [1, 6]. De novo FGFR variants are paternal in origin and increase with paternal age through spermatogonial selection, explaining low recurrence risk after an affected child born to non-carrier parents [1, 50].

There is no targeted pharmacological treatment; management is surgical and multidisciplinary, with cranial expansion in the first year (6–9 months in most syndromic forms and 9–12 months in Muenke syndrome, according to the European guideline), midface advancement with distraction between ages 8 and 12 or after 17, proactive airway management and ICH surveillance at syndrome-specific intervals [15]. The main disagreements between sources concern surgical age in Muenke syndrome and the role of CPAP, and supporting evidence is predominantly observational and from specialist centres [5, 6, 15]. Research into molecular therapies, such as personalised antisense oligonucleotides, remains preclinical [72].

CYP26B1 deficiency is an autosomal recessive syndromic craniosynostosis caused by the lack of local retinoic acid degradation; it combines craniosynostosis or an undermineralised calvaria with elbow synostoses, first-ray hypoplasia and arachnodactyly, with severity ranging from fetal death to adult survival with hearing loss and intellectual disability. It is distinguished from the POR-related Antley–Bixler forms by normal genitalia and steroids, and molecular diagnosis requires attention to splicing variants, including synonymous ones.

13. References

1. Firth HV, Hurst JA. Craniosynostosis. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 378–81.
2. Orphanet. Epidemiology records, definitions and ICD-10 mapping for ORPHA:1531 (Craniosynostosis), ORPHA:207 (Crouzon syndrome), ORPHA:93262 (Crouzon syndrome-acanthosis nigricans syndrome), ORPHA:87 (Apert syndrome), ORPHA:710 (Pfeiffer syndrome), ORPHA:794 (Saethre-Chotzen syndrome), ORPHA:53271 (Muenke syndrome) and ORPHA:1540 (Jackson-Weiss syndrome) [Internet]. Orphadata API and ORPHAcodes API; accessed 2026 Oct 6. Available at: https://api.orphadata.com/rd-epidemiology/orphacodes/
3. Jones KL, Jones MC, del Campo M. Crouzon syndrome (craniofacial dysostosis). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 582–3.
4. Wenger TL, Hing AV, Evans KN. Apert syndrome. 2019 May 30. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 31145570. Available at: https://www.ncbi.nlm.nih.gov/books/NBK541728/
5. Wenger T, Miller D, Evans K. FGFR craniosynostosis syndromes overview. 1998 Oct 20 [updated 2020 Apr 30]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301628. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1455/
6. Kruszka P, Rolle M, Kahle KT, Muenke M. Muenke syndrome. 2006 May 10 [updated 2023 Mar 30]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301588. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1415/
7. VariantValidator [Internet]. Validation on GRCh38 using MANE Select transcripts of NM_000141.5:c.755C>G, c.758C>G, c.755_756delinsTT and c.1026C>G; NM_000142.5:c.749C>G and c.1172C>A; NM_023110.3:c.755C>G; NM_000474.4. Accessed 2026 Oct 6. Available at: https://rest.variantvalidator.org/
8. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Pfeiffer syndrome (MIM 101600, 136350). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 834–6.
9. Shi H, Yang J, Guo Q, Zhang M. Clinical assessment and FGFR2 mutation analysis in a Chinese family with Crouzon syndrome: a case report. Medicine (Baltimore). 2021;100(10):e24991. doi:10.1097/MD.0000000000024991. PMID: 33725872. PMC7969214.
10. Jones KL, Jones MC, del Campo M. Apert syndrome (acrocephalosyndactyly type 1). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 578–81.
11. GeneGuidance. Clinical curation of genes FGFR1, FGFR2, FGFR3, TWIST1, EFNB1, TCF12 and ERF and interpretation rules for FGFR3 variants p.Pro250Arg and p.Ala391Glu [database]. Export of 2026 Sep 26; accessed 2026 Oct 6.
12. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Apert syndrome (MIM 101200). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 829–33.
13. Boulet SL, Rasmussen SA, Honein MA. A population-based study of craniosynostosis in metropolitan Atlanta, 1989-2003. Am J Med Genet A. 2008;146A(8):984-91. doi:10.1002/ajmg.a.32208. PMID: 18344207.
14. Dias MS, Samson T, Rizk EB, Governale LS, Richtsmeier JT; Section on Neurologic Surgery, Section on Plastic and Reconstructive Surgery. Identifying the misshapen head: craniosynostosis and related disorders. Pediatrics. 2020;146(3):e2020015511. doi:10.1542/peds.2020-015511. PMID: 32868470.
15. Mathijssen IMJ; Working Group Guideline Craniosynostosis. Updated guideline on treatment and management of craniosynostosis. J Craniofac Surg. 2021;32(1):371-450. doi:10.1097/SCS.0000000000007035. PMID: 33156164. PMC7769187.
16. European Reference Network on rare and/or complex craniofacial anomalies and ear, nose and throat disorders (ERN CRANIO). Craniosynostosis; Standards of care; About ERN CRANIO [Internet]. Accessed 2026 Oct 6. Available at: https://www.ern-cranio.eu/
17. Online Mendelian Inheritance in Man. #614416 Radiohumeral fusions with other skeletal and craniofacial anomalies; RHFCA [Internet]. Baltimore (MD): Johns Hopkins University; last edited 2024 Mar 26; accessed 2026 Nov 1. Available at: https://www.omim.org/entry/614416
18. Orphanet. Lethal occipital encephalocele-skeletal dysplasia syndrome (ORPHA:293925): classification, inheritance, prevalence and ICD-10 mapping [Internet]. Paris: INSERM; accessed 2026 Nov 1. Available at: https://www.orpha.net/en/disease/detail/293925
19. Silveira KC, Fonseca IC, Oborn C, Wengryn P, Ghafoor S, Beke A, et al. CYP26B1-related disorder: expanding the ends of the spectrum through clinical and molecular evidence. Hum Genet. 2023;142(11):1571-86. doi:10.1007/s00439-023-02598-2. PMID: 37755482. PMC10602971.
20. Laue K, Pogoda HM, Daniel PB, van Haeringen A, Alanay Y, von Ameln S, et al. Craniosynostosis and multiple skeletal anomalies in humans and zebrafish result from a defect in the localized degradation of retinoic acid. Am J Hum Genet. 2011;89(5):595-606. doi:10.1016/j.ajhg.2011.09.015. PMID: 22019272. PMC3213388.
21. Sarma AS, Peter Mathew R, Dalal A, Bhat V, Patil SJ. Familial monoallelic CYP26B1 truncating variant causes a syndromic craniosynostosis due to haploinsufficiency? Eur J Med Genet. 2023;66(7):104772. doi:10.1016/j.ejmg.2023.104772. PMID: 37100236.
22. Idkowiak J, Cragun D, Hopkin RJ, Arlt W. Cytochrome P450 oxidoreductase deficiency. 2005 Sep 8 [updated 2017 Aug 3]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1419/
23. Leduc F, Vanlerberghe C, Escande F, Brunelle P, Petit F, Dieux A. Humeroradial synostosis: an updated classification and differential diagnosis based on genetic aetiology. Clin Genet. 2025;108(4):379-92. doi:10.1111/cge.70023. PMID: 40673520. PMC12405067.
24. Caliskan BO, Demir M, Oktem S, Ozturk S, Canpolat M, Dundar M. A rare craniosynostosis phenotype associated with a homozygous CYP26B1 pathogenic variant in the absence of extremity synostosis. Am J Med Genet A. 2026;200(2):490-5. doi:10.1002/ajmg.a.64257. PMID: 40999913.
25. Grand K, Skraban CM, Cohen JL, Dowsett L, Mazzola S, Tarpinian J, et al. Nonlethal presentations of CYP26B1-related skeletal anomalies and multiple synostoses syndrome. Am J Med Genet A. 2021;185(9):2766-75. doi:10.1002/ajmg.a.62387. PMID: 34160123.
26. European Reference Network on Rare Bone Diseases (ERN BOND) [Internet]. Accessed 2026 Nov 1. Available at: https://ernbond.eu/
27. Firth HV, Hurst JA. Plagiocephaly and abnormalities of skull shape. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 280–3.
28. Morton JE, Frentz S, Morgan T, Sutherland-Smith AJ, Robertson SP. Biallelic mutations in CYP26B1: a differential diagnosis for Pfeiffer and Antley-Bixler syndromes. Am J Med Genet A. 2016;170(10):2706-10. doi:10.1002/ajmg.a.37804. PMID: 27410456.
29. Jones KL, Jones MC, del Campo M. Saethre-Chotzen syndrome (acrocephalosyndactyly type III). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 570–3.
30. Jones KL, Jones MC, del Campo M. Pfeiffer syndrome (Pfeiffer-type acrocephalosyndactyly). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 574–7.
31. Reardon W, Winter RM, Rutland P, Pulleyn LJ, Jones BM, Malcolm S. Mutations in the fibroblast growth factor receptor 2 gene cause Crouzon syndrome. Nat Genet. 1994;8(1):98-103. doi:10.1038/ng0994-98. PMID: 7987400.
32. Muenke M, Schell U, Hehr A, Robin NH, Losken HW, Schinzel A, et al. A common mutation in the fibroblast growth factor receptor 1 gene in Pfeiffer syndrome. Nat Genet. 1994;8(3):269-74. doi:10.1038/ng1194-269. PMID: 7874169.
33. Wilkie AO, Slaney SF, Oldridge M, Poole MD, Ashworth GJ, Hockley AD, et al. Apert syndrome results from localized mutations of FGFR2 and is allelic with Crouzon syndrome. Nat Genet. 1995;9(2):165-72. doi:10.1038/ng0295-165. PMID: 7719344.
34. Meyers GA, Orlow SJ, Munro IR, Przylepa KA, Jabs EW. Fibroblast growth factor receptor 3 (FGFR3) transmembrane mutation in Crouzon syndrome with acanthosis nigricans. Nat Genet. 1995;11(4):462-4. doi:10.1038/ng1295-462. PMID: 7493034.
35. Muenke M, Gripp KW, McDonald-McGinn DM, Gaudenz K, Whitaker LA, Bartlett SP, et al. A unique point mutation in the fibroblast growth factor receptor 3 gene (FGFR3) defines a new craniosynostosis syndrome. Am J Hum Genet. 1997;60(3):555-64. PMID: 9042914. PMC1712518.
36. Mathijssen IMJ. Introduction to updated guideline on treatment and management of craniosynostosis. J Craniofac Surg. 2021;32(1):370. doi:10.1097/SCS.0000000000007301. PMID: 33306646. PMC7769178.
37. Faasse M, Mathijssen IMJ; ERN CRANIO Working Group on Craniosynostosis. Guideline on treatment and management of craniosynostosis: patient and family version. J Craniofac Surg. 2023;34(1):418-33. doi:10.1097/SCS.0000000000009143. PMID: 36472893. PMC9794150.
38. Wilkie AO, Byren JC, Hurst JA, Jayamohan J, Johnson D, Knight SJ, et al. Prevalence and complications of single-gene and chromosomal disorders in craniosynostosis. Pediatrics. 2010;126(2):e391-400. doi:10.1542/peds.2009-3491. PMID: 20643727. PMC3535761.
39. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bent bone dysplasia-FGFR2 type (MIM 614592). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 853–6.
40. Gallagher ER, Ratisoontorn C, Cunningham ML. Saethre-Chotzen syndrome. 2003 May 16 [updated 2019 Jan 24]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301368. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1189/
41. Sharma VP, Fenwick AL, Brockop MS, McGowan SJ, Goos JA, Hoogeboom AJ, et al. Mutations in TCF12, encoding a basic helix-loop-helix partner of TWIST1, are a frequent cause of coronal craniosynostosis. Nat Genet. 2013;45(3):304-7. doi:10.1038/ng.2531. PMID: 23354436. PMC3647333.
42. Foss-Skiftesvik J, Larsen CC, Stoltze UK, Kofod T, Hove H, Bøgeskov L, et al. The role of pathogenic TCF12 variants in children with coronal craniosynostosis—a systematic review with addition of two novel cases. Childs Nerv Syst. 2024;40(11):3655-71. doi:10.1007/s00381-024-06544-z. PMID: 39060747.
43. Twigg SR, Vorgia E, McGowan SJ, Peraki I, Fenwick AL, Sharma VP, et al. Reduced dosage of ERF causes complex craniosynostosis in humans and mice and links ERK1/2 signaling to regulation of osteogenesis. Nat Genet. 2013;45(3):308-13. doi:10.1038/ng.2539. PMID: 23354439. PMC3683605.
44. Glass GE, O'Hara J, Canham N, Cilliers D, Dunaway D, Fenwick AL, et al. ERF-related craniosynostosis: the phenotypic and developmental profile of a new craniosynostosis syndrome. Am J Med Genet A. 2019;179(4):615-27. doi:10.1002/ajmg.a.61073. PMID: 30758909. PMC6491982.
45. Jones KL, Jones MC, del Campo M. Craniofrontonasal dysplasia. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 588–91.
46. Twigg SR, Babbs C, van den Elzen ME, Goriely A, Taylor S, McGowan SJ, et al. Cellular interference in craniofrontonasal syndrome: males mosaic for mutations in the X-linked EFNB1 gene are more severely affected than true hemizygotes. Hum Mol Genet. 2013;22(8):1654-62. doi:10.1093/hmg/ddt015. PMID: 23335590. PMC3605834.
47. Lee E, Le T, Zhu Y, Elakis G, Turner A, Lo W, et al. A craniosynostosis massively parallel sequencing panel study in 309 Australian and New Zealand patients: findings and recommendations. Genet Med. 2018;20(9):1061-8. doi:10.1038/gim.2017.214. PMID: 29215649.
48. Firth HV, Hurst JA. Paternal age. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 772–3.
49. Goriely A, McVean GA, Röjmyr M, Ingemarsson B, Wilkie AO. Evidence for selective advantage of pathogenic FGFR2 mutations in the male germ line. Science. 2003;301(5633):643-6. doi:10.1126/science.1085710. PMID: 12893942.
50. Goriely A, Wilkie AO. Paternal age effect mutations and selfish spermatogonial selection: causes and consequences for human disease. Am J Hum Genet. 2012;90(2):175-200. doi:10.1016/j.ajhg.2011.12.017. PMID: 22325359. PMC3276674.
51. Rannan-Eliya SV, Taylor IB, De Heer IM, Van Den Ouweland AM, Wall SA, Wilkie AO. Paternal origin of FGFR3 mutations in Muenke-type craniosynostosis. Hum Genet. 2004;115(3):200-7. doi:10.1007/s00439-004-1151-5. PMID: 15241680.
52. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Antley-Bixler syndrome (MIM 201750). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 837–9.
53. Nilsson O, Isoherranen N, Guo MH, Lui JC, Jee YH, Guttmann-Bauman I, et al. Accelerated skeletal maturation in disorders of retinoic acid metabolism: a case report and focused review of the literature. Horm Metab Res. 2016;48(11):737-44. doi:10.1055/s-0042-114038. PMID: 27589347. PMC5534175.
54. Li J, Xie H, Jiang Y. Mucopolysaccharidosis IIIB and mild skeletal anomalies: coexistence of NAGLU and CYP26B1 missense variations in the same patient in a Chinese family. BMC Med Genet. 2018;19(1):51. doi:10.1186/s12881-018-0562-4. PMID: 29606097. PMC5880076.
55. Wen J, Lopes F, Soares G, Farrell SA, Nelson C, Qiao Y, et al. Phenotypic and functional consequences of haploinsufficiency of genes from exocyst and retinoic acid pathway due to a recurrent microdeletion of 2p13.2. Orphanet J Rare Dis. 2013;8:100. doi:10.1186/1750-1172-8-100. PMID: 23837398. PMC3710273.
56. Jones KL, Jones MC, del Campo M. FGFR3-associated coronal synostosis syndrome (Muenke syndrome). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 584–5.
57. Kruszka P, Addissie YA, Yarnell CM, Hadley DW, Guillen Sacoto MJ, Platte P, et al. Muenke syndrome: an international multicenter natural history study. Am J Med Genet A. 2016;170A(4):918-29. doi:10.1002/ajmg.a.37528. PMID: 26740388.
58. Becerril Santos MC, Ongkosuwito EM, Papadopoulou AK, Antonarakis GS. Prevalence and patterns of permanent tooth agenesis in patients with Crouzon or Apert syndrome: a systematic review and meta-analysis. Orthod Craniofac Res. 2026;29(1):1-11. doi:10.1111/ocr.70046. PMID: 41085212. PMC12779208.
59. Abu-Sittah GS, Jeelani O, Dunaway D, Hayward R. Raised intracranial pressure in Crouzon syndrome: incidence, causes, and management. J Neurosurg Pediatr. 2016;17(4):469-75. doi:10.3171/2015.6.PEDS15177. PMID: 26613275.
60. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Muenke syndrome (MIM 602849). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 851–2.
61. Firth HV, Hurst JA. Large fontanelle. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 204–5.
62. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Saethre-Chotzen syndrome (MIM 101400). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 840–2.
63. Junaid M, Slack-Smith L, Wong K, Bourke J, Baynam G, Calache H, et al. Epidemiology of rare craniofacial anomalies: retrospective Western Australian population data linkage study. J Pediatr. 2022;241:162-72.e9. doi:10.1016/j.jpeds.2021.09.060. PMID: 34626670.
64. Jones KL, Jones MC, del Campo M. Carpenter syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 592–3.
65. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Carpenter syndrome (MIM 201000, 614970). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 848–50.
66. Jones KL, Jones MC, del Campo M. Antley-Bixler syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 596–7.
67. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Baller-Gerold syndrome (MIM 218600, 266280). In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 843–7.
68. Jones KL, Jones MC, del Campo M. Greig cephalopolysyndactyly syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 594–5.
69. Jones KL, Jones MC, del Campo M. Curry-Jones syndrome. In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 586–7.
70. Firth HV, Hurst JA. Syndactyly (other than 2/3 toe syndactyly). In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 328–9.
71. U.S. Food and Drug Administration, openFDA drug label API (search for “craniosynostosis”, “Apert” and “Crouzon” in indications) and European Medicines Agency, authorised human medicines file (search for the same terms) [Internet]. Accessed 2026 Oct 6. Available at: https://api.fda.gov/drug/label.json and https://www.ema.europa.eu/en/medicines/download-medicine-data
72. ClinicalTrials.gov [Internet]. NCT07535372: Nano-ink based antisense oligonucleotide delivery for ultra-personalized treatment of syndromic craniosynostoses (NAUTILUS); NCT04463316: GROWing up with rare GENEtic syndromes (GROW UR GENES); searches for “Apert syndrome”, “Crouzon”, “Pfeiffer syndrome”, “Muenke”, “Saethre-Chotzen” and “syndromic craniosynostosis”. Bethesda (MD): National Library of Medicine; accessed 2026 Oct 6. Available at: https://clinicaltrials.gov/study/NCT07535372
73. Hu AC, Tolley PD, Han NA, Ryan IA, Liao EC, Swanson JW, et al. The APERT severity scale: a quantitative tool for risk stratification in Apert syndrome. Plast Reconstr Surg. 2026;158(1):120e-131e. doi:10.1097/PRS.0000000000012723. PMID: 41400369.
74. ClinicalTrials.gov [Internet]. Searches for “CYP26B1” and “radiohumeral synostosis”. Bethesda (MD): National Library of Medicine; accessed 2026 Nov 1. Available at: https://clinicaltrials.gov/
75. Lajeunie E, Le Merrer M, Bonaïti-Pellie C, Marchac D, Renier D. Genetic study of nonsyndromic coronal craniosynostosis. Am J Med Genet. 1995;55(4):500-4. doi:10.1002/ajmg.1320550422. PMID: 7762595.
76. Vogels A, Fryns JP. Pfeiffer syndrome. Orphanet J Rare Dis. 2006;1:19. doi:10.1186/1750-1172-1-19. PMID: 16740155. PMC1482682.
77. Li Y, Ma D, Sun Y, Meng L, Wang Y, Jiang T. Apert syndrome with FGFR2 758 C > G mutation: a Chinese case report. Front Genet. 2018;9:181. doi:10.3389/fgene.2018.00181. PMID: 29868125. PMC5966571.
78. Barbosa M, Almeida MR, Reis-Lima M, Pinto-Basto J, dos Santos HG. Muenke syndrome with osteochondroma. Am J Med Genet A. 2009;149A(2):260-1. PMID: 19097163.
79. Jehee FS, Alonso LG, Cavalcanti DP, Kim C, Wall SA, Mulliken JB, et al. Mutational screening of FGFR1, CER1, and CDON in a large cohort of trigonocephalic patients. Cleft Palate Craniofac J. 2006;43(2):148-51. PMID: 16526918.
80. Cairns BJ, Davidson DM, Smithson SF, Wilkie AOM. Long-term survival without high cancer risk in a cohort of 24 patients with Apert syndrome. Eur J Hum Genet. 2026;34(6):783-7. doi:10.1038/s41431-026-02122-w. PMID: 42082645. PMC13247221.
81. Baraya N, Thibodaux L, Martin A, Stone C, Strahle J, Patel K, et al. Expert recommendations for neurodevelopmental screening in children with craniosynostosis: a consensus report from the Craniosynostosis Research Neurodevelopmental Working Group (SynRG). Cleft Palate Craniofac J. 2026;63(6):1538-47. doi:10.1177/10556656251339879. PMID: 40368325.