Cleidocranial Dysplasia

RUNX2-associated cleidocranial dysplasia spectrum, CBFB-associated cleidocranial dysplasia and metaphyseal dysplasia with maxillary hypoplasia and brachydactyly (an allelic RUNX2 disorder)

Literature and regulatory information updated: 10 October 2026.

0. Initial Reference Table

Table 0.1 brings together the three entities within scope: cleidocranial dysplasia (CCD), caused by heterozygous loss-of-function variants in RUNX2, which GeneReviews treats as a continuous spectrum; CBFB-associated cleidocranial dysplasia, a phenocopy described in 2023 and catalogued in OMIM (Online Mendelian Inheritance in Man) as cleidocranial dysplasia 2; and metaphyseal dysplasia with maxillary hypoplasia and brachydactyly (MDMHB), an allelic RUNX2 disorder caused by an intragenic duplication. ORPHA denotes the Orphanet code, ICD-10 the 10th revision of the International Classification of Diseases, ERN the European Reference Networks, DXA dual-energy X-ray absorptiometry, CNV copy number variants and EUROCAT the European network of congenital anomaly registries.

Table 0.1 — Reference information for the entities within scope

1. Introduction

1.1. Historical Description and Nomenclature

Cleidocranial dysplasia is an autosomal dominant skeletal dysplasia whose classic triad combines delayed closure of cranial sutures and fontanelles, hypoplasia or aplasia of the clavicles and dental anomalies, with retained deciduous teeth, supernumerary teeth and failure of permanent tooth eruption [2]. Its history is long: Smith's Recognizable Patterns of Human Malformation refers to a possible example identified in the skull of a Neanderthal man, attributed to Greig (1933), whom the book cites indirectly, and records that well over 500 cases have been published [14]. Scheuthauer's description in 1871 and that of Pierre Marie and Paul Sainton in 1897–1898 are the foundational references cited by skeletal dysplasia atlases, and the eponym “Marie–Sainton disease” or “Scheuthauer–Marie–Sainton syndrome” derives from them [1, 14]. Marie and Sainton coined the term “cleidocranial dysostosis” to designate combined involvement of the clavicles and skull [14]. GeneReviews adds that the condition was also described as dento-osseous dysplasia in a large pedigree [2]. The original descriptions from 1871 and 1898 were not consulted directly; their characterisation is based on the secondary sources cited.

The designation “dysostosis” was abandoned in favour of “dysplasia” because the defect is not limited to individual bones: the transcription factor encoded by the responsible gene acts in both formation and maintenance of the entire skeleton, and the disease also affects bones formed by endochondral ossification, such as the pelvis, spine and hands, although bones formed by intramembranous ossification (calvaria and clavicles) are most affected [2, 14]. The list of synonyms recorded in the Bone Dysplasias atlas includes pelvic-cleidocranial dysostosis, cleido-cranio-digital dysostosis and osteodental dysplasia, reflecting the extent of involvement [1].

1.2. Molecular Discovery

CCD was mapped to the short arm of chromosome 6 (6p21), and a microdeletion in that region was described by Mundlos and colleagues in 1995, according to the citation in Smith [14]. In 1997, a group of five studies established the role of the transcription factor then designated CBFA1 (core-binding factor alpha-1) or OSF2, now RUNX2. Ducy and colleagues cloned Osf2/Cbfa1 as the factor binding the OSE2 element of the osteocalcin promoter, with expression beginning in skeletal mesenchymal condensations and subsequently restricted to the osteoblast lineage, and showed that its forced expression in non-osteoblastic cells induces osteoblast genes [15]. Komori and colleagues and Otto and colleagues showed that mice lacking Cbfa1 die shortly after birth without any intramembranous or endochondral ossification because osteoblast maturation is arrested, and Otto and colleagues observed that heterozygotes reproduce the skeletal anomalies of CCD, also present in the murine Ccd mutant, in which the gene is deleted [16, 17]. In the same year, Mundlos and colleagues described deletions removing one copy of CBFA1 and insertions, deletions and missense variants introducing stop codons in the DNA-binding domain or transactivation region in families with CCD, concluding that heterozygous loss of function is sufficient to cause the disease; in one family, an in-frame expansion of the polyalanine tract segregated with brachydactyly and minor signs of CCD [18]. Lee and colleagues described two de novo missense variants (published designations Met175Arg and Ser191Asn; NM_001024630.4:c.524T>G p.(Met175Arg) and NM_001024630.4:c.572G>A p.(Ser191Asn)) at conserved residues of the DNA-binding domain that abolish this binding, and characterised CCD as one of the most frequent skeletal dysplasias without disproportionate short stature [19].

In subsequent years, studies by Quack and colleagues (1999) and Zhou and colleagues (1999) showed that missense variants cluster in the runt domain, with a hotspot at arginine 225, and identified hypomorphic alleles responsible for mild forms and an exclusively dental phenotype, underpinning the concept of a spectrum [20, 21]. Ott and colleagues (2010) demonstrated that RUNX2 deletions are present in approximately 10% of patients [22]. In 2023, Beyltjens and colleagues described heterozygous variants in CBFB, the heterodimerisation partner of RUNX2, in eight individuals from five families with a CCD-like phenotype and no RUNX2 alteration, and the entity was incorporated into OMIM as cleidocranial dysplasia 2 (OMIM 620099) [3, 11]. In 2026, the first consensus recommendations for CCD diagnosis and management, obtained through a modified Delphi panel, were published [10].

1.3. Definition of the Entities Within Scope

RUNX2-associated CCD is defined by GeneReviews as a continuous spectrum ranging from classic CCD, with the complete triad, to mild CCD and isolated dental anomalies without other skeletal findings; diagnosis is established by characteristic clinical and radiographic findings or identification of a heterozygous pathogenic or likely pathogenic RUNX2 variant [2]. Under entry 119600, OMIM records the subentries “forme fruste, dental anomalies only” and “forme fruste, with brachydactyly” [6]. CBFB-associated CCD (OMIM 620099) is an example of locus heterogeneity, clinically indistinguishable from the classic form in the skeleton and dentition, but usually with normal stature and, in some patients, developmental delay [3, 4]. MDMHB (OMIM 156510) is an allelic RUNX2 disorder caused by a partial intragenic duplication, with a gain-of-function mechanism according to Orphanet and a distinct phenotype: short stature, long bone and spinal abnormalities, dystrophic teeth and widening of the medial half of the clavicles [2, 5]. Pycnodysostosis, CDAGS syndrome, Yunis–Varon syndrome, mandibuloacral dysplasia and the other entities with clavicular hypoplasia or large fontanelles remain in the differential diagnosis (section 6); pycnodysostosis belongs to the sclerosing bone dysplasia group.

2. Molecular Genetics and Pathophysiology

2.1. The RUNX2 Gene and RUNX2 Protein

The RUNX2 gene is located at 6p21.1 and encodes a transcription factor of the RUNX (runt-related) family, whose current HGNC (HUGO Gene Nomenclature Committee) approved name is RUNX family transcription factor 2; historical aliases include CBFA1, AML3, PEBP2A1 and CCD1 [6]. The protein has three main functional regions: an N-terminal Q/A domain, with polyglutamine and polyalanine repeats; the 128-amino-acid runt domain, named for homology with the Drosophila runt gene, which binds DNA and independently mediates heterodimerisation; and a C-terminal activation domain rich in proline, serine and threonine (PST) [2]. The systematic review by Thaweesapphithak and colleagues also describes a nuclear localization signal (NLS), a nuclear matrix targeting signal (NMTS) and the C-terminal VWRPY motif [23].

The gene generates isoforms through two alternative promoters and alternative splicing [2]. The MANE Select (Matched Annotation from NCBI and EMBL-EBI) transcript is NM_001024630.4, the longest, with 9 exons, encoding the 521-amino-acid isoform NP_001019801.3 initiated at the distal P1 promoter (initial sequence MASNSLFSTV); transcript NM_004348.3, used in some of the literature, encodes the 507-amino-acid isoform NP_004339.3 (initial sequence MRIPVDPSTS), shifting protein numbering by 14 residues [2, 7]. The historical designations Met175, Ser191, Thr200, Arg225 and Arg391, used in the foundational studies, coincide with residues in the MANE protein; by contrast, the designation p.R211W in a 2026 report refers to the short isoform and corresponds to p.(Arg225Trp) in MANE numbering [7, 24]. GeneGuidance gives NM_004348 as the reference RefSeq, making it necessary always to confirm the transcript used in a report [6].

2.2. Physiological Function and Disease Mechanism

RUNX2 is the master regulator of osteoblast differentiation and skeletal morphogenesis: it is essential for intramembranous ossification and participates in chondrocyte maturation during endochondral ossification [2, 25]. It binds DNA as a heterodimer with CBFβ and activates promoters of bone matrix genes, such as those encoding osteocalcin, osteopontin, bone sialoprotein and type I collagen; it also participates in adult bone maintenance [6, 15]. Mice lacking Runx2 form neither bone nor osteoblasts and die at birth from respiratory failure, whereas heterozygotes reproduce the clavicular hypoplasia and delayed cranial ossification of CCD [16, 17]. This gradient explains the selectivity of the human phenotype: bones formed by intramembranous ossification (calvaria and clavicles), which depend more directly on osteoblast differentiation, are most affected, whereas involvement of endochondral bones (pubis, femoral neck, phalanges, metacarpals) is less pronounced [2]. RUNX2 also acts in odontoblast differentiation and tooth formation, underpinning eruption anomalies and supernumerary teeth [26].

The mechanism of CCD is loss of function through haploinsufficiency. Most variants in classic CCD affect the runt domain and abolish DNA binding and transactivation, or introduce a premature stop codon upstream of or within this domain [2, 18, 19]. Quack and colleagues found no phenotypic difference between deletions or frameshift variants and other intragenic variants, reinforcing the haploinsufficiency model [20]. Missense variants cluster at arginine 225, a critical residue of the runt domain, and interfere with nuclear protein accumulation in vitro [2, 20]. In the mouse model with the variant equivalent to p.(Arg225Gln) (p.Arg232Gln in the mouse), heterozygotes have an open anterior fontanelle and clavicular hypoplasia, and the mutant protein almost completely loses its ability to transactivate the osteocalcin promoter and has partially impaired nuclear localisation; the same study described small root protrusions on the pulp chamber floor of the maxillary first molars, a preclinical finding with no established human correlate [27].

2.3. Variant Spectrum

Table 2.1 summarises the proportion of probands with a detectable variant by method, according to GeneReviews. These values are diagnostic yields in series, not the clinical sensitivity of the tests.

Table 2.1 — Molecular Tests in RUNX2-Associated CCD (GeneReviews, 2023)

Other series give lower yields: the Delphi panel reports RUNX2 variants in 60%–70% of individuals with CCD, Farrow and colleagues in 65% of clinically diagnosed patients, and Beyltjens and colleagues report 10%–30% of patients without an identified genetic cause [3, 10, 30]. The difference reflects methods from different periods and variable clinical inclusion criteria. Ott and colleagues showed that RUNX2 deletions are present in approximately 10% of patients, and Smith records the same proportion for intragenic deletions or duplications [14, 22]. In the Turkish series by Berkay and colleagues, with 51 patients from 30 families, 90% of patients with a variant had small sequence changes and 10% large deletions; 23 variants were identified, 11 novel, including a deletion of exons 6 to 9 and the variant c.29_30insT, which affects only the P1-dependent long isoform and was found in two patients with discordant phenotypes [31].

The systematic review by Thaweesapphithak and colleagues collected 569 variants and 453 patients from 103 articles: 48.68% of variants were in-frame and 51.32% null; 55.54% were located in the runt domain, 16.34% in PST, 6.33% in NMTS, 4.75% in Q/A, 1.41% in NLS, 1.23% in VWRPY and 10.19% in non-coding regions; in-frame variants were concentrated in the runt domain (90.97%) [23]. A ClinVar search on 10 October 2026 returned 158 records of RUNX2 variants classified as pathogenic, a figure counting records rather than patients [32].

Studies from 2025–2026 extended the mutational spectrum to regions not interrogated by exome sequencing. Mijovic and colleagues described a de novo 4.2-kb deletion confined to the RUNX2 3' untranslated region (3'UTR) (chr6:45518143-45522390, GRCh37) in a Serbian girl with classic CCD and negative microarray and exome results, detected by genome sequencing and confirmed by trio qPCR, initially classified as a variant of uncertain significance and reclassified as likely pathogenic after demonstration of its de novo origin [29]. Stojanovic and colleagues studied an Austrian family with CCD and negative exome results for RUNX2 and CBFB: complementary DNA from fibroblasts showed inclusion of a 78-nucleotide pseudoexon from intron 6 (published designation NM_001024630.4:c.859_860ins78, in HGVS r.859_860insN[78]; p.(Asp287Glyfs*3)), subject to nonsense-mediated decay (NMD), and long-read sequencing identified the deep intronic variant NC_000006.12:g.45509900G>T, corresponding to NM_001024630.4:c.860-2346G>T, absent from gnomAD v4.1.0 and cosegregating with the disease [7, 28].

2.4. Notable Variants and Nomenclature Normalisation

Table 2.2 presents variants of clinical or historical relevance, in HGVS (Human Genome Variation Society) nomenclature using MANE transcript NM_001024630.4. Conversions from other transcripts were performed in VariantValidator; the ClinVar classification is given when consulted.

Table 2.2 — Notable RUNX2 Variants (NM_001024630.4 / NP_001019801.3)

The missense variants p.(Met175Arg) and p.(Ser191Asn), described as de novo by Lee and colleagues in 1997, retain the same residue numbering in the MANE protein [7, 19]. Variants in GeneReviews Table 6 were provided by the chapter authors, without classification review by GeneReviews; for c.90dupC, c.598A>G and c.674G>T, the sources consulted do not include a ClinVar record [2].

2.5. Genotype–Phenotype Correlations, Penetrance and Expressivity

Penetrance is high and there are no reports of incomplete penetrance, but expressivity varies widely, even among relatives with the same variant, from dental and clavicular changes alone to severe skeletal defects [2, 14]. Established correlations chiefly concern dental manifestations; no clear correlation between genotype and degree of clavicular involvement has been demonstrated [2]. Yoshida and colleagues observed that short stature and dental anomalies are milder in patients with classic CCD whose runt domain is intact and who retain greater residual activity than in those with variants in this domain [2, 35]. Hypomorphic alleles c.90dupC, c.598A>G and c.1171C>T, which retain some function, produce the spectrum of classic CCD, mild CCD and isolated dental anomalies, with marked intrafamilial variability [2, 21]. Quack and colleagues described a family with a missense variant in which affected individuals had only supernumerary teeth and a radiographically normal skeleton [20].

In the systematic review of 453 patients, missense variants, particularly in the runt domain, were significantly associated, compared with nonsense variants, with supernumerary teeth, macrocephaly, a metopic groove, short ribs and hypoplastic iliac wings; compared with in-frame variants in other regions, they were associated with delayed fontanelle closure, metopic synostosis, hypertelorism, limited shoulder abduction, pubic symphysis abnormalities and hypoplastic iliac wings; null variants in the NLS region had weaker associations, and not all alter nuclear localisation [23]. The Delphi panel summarises these data by stating that loss-of-function variants are generally associated with a milder phenotype than missense variants in certain regions, and that the same variant can produce different phenotypes within and between families [10]. In a Thai cross-sectional study of 9 genetically confirmed individuals assessed by cephalometry and cone-beam computed tomography (CBCT), truncating or structural variants were associated with a shorter cranial base and smaller nasolabial angle, without an association between variant type and dental anomaly burden; the sample is small [36].

RUNX2 deletions extending into adjacent genes may combine the CCD phenotype with developmental delay and other findings, distinguishing them from intragenic variants [2, 6]. The variant c.1205dupC was associated with osteoporosis, recurrent fractures and scoliosis in an adult, the first evidence of RUNX2's role in adult bone maintenance [2, 20]. Parental somatic and germline mosaicism is documented (section 8), and Muurinen and colleagues described mosaic deletions of known genes as a cause of skeletal dysplasias with high and low bone mass [37, 38].

2.6. Allelic RUNX2 Disorders

Partial intragenic duplication of RUNX2 causes MDMHB (OMIM 156510), with short stature, long bone and spinal abnormalities, dystrophic teeth and widening of the medial half of the clavicles, a phenotype contrasting with the clavicular hypoplasia of CCD; Orphanet classifies the mechanism as gain of function [2, 5]. Complete RUNX2 duplications have been described in patients with craniosynostosis and oligodontia, suggesting a dosage effect opposite to haploinsufficiency [2]. The in-frame expansion of the polyalanine tract described by Mundlos and colleagues in a family with brachydactyly and minor CCD signs corresponds to the OMIM subentry “forme fruste with brachydactyly” [6, 18].

2.7. CBFB and Cleidocranial Dysplasia 2

CBFB, at 16q22.1, encodes the β subunit of core-binding factor (CBFβ), which does not bind DNA but heterodimerises with RUNX1, RUNX2 and RUNX3, stabilises them and allosterically increases their affinity for the TGTGGT sequence [4]. Mouse models showed that Cbfβ interacts with Runx2 and is necessary for bone development, and conditional knockouts in chondrocytes, osteoblasts or mesenchymal cells impair intramembranous and endochondral ossification [3, 39]. Goto and colleagues had already noted in 2004 that large fontanelles are a feature shared by RUNX2 and CBFB haploinsufficiency [40].

In 2023, Beyltjens and colleagues described eight individuals from five families, partly assembled through GeneMatcher, all without a RUNX2 alteration and with heterozygous CBFB (NM_022845.3) variants affecting the RUNX-binding domain: c.78+1G>T (r.74_78del, p.(Cys25Tyrfs*2), de novo); the 9003-nucleotide intragenic deletion c.283-1039_400-7568del, which removes exon 4 without altering the reading frame (p.(Val95_Gln133del), de novo); c.295_296dup p.(Pro100Leufs*3), in a mother and daughter; c.247C>T p.(Arg83*), in a mother and daughter; and c.283-2A>G, in a mother and daughter [3]. The authors suggest that the truncated protein may fail to interact with RUNX or have a neomorphic effect [3]. In the UK 100,000 Genomes Project, Pagnamenta and colleagues identified a mother and daughter with RUNX2-negative CCD and a cryptic CBFB rearrangement (deletion–inversion–deletion removing exons 1 and 2; NC_000016.10:g.67026846_67030652delins67028156_67028717inv), mediated by a palindromic AluSx repeat less than 1 kb from the transcription start; reduced transcript levels on qRT-PCR supported a loss-of-function rather than neomorphic mechanism [11]. The main phenotypic distinction from the RUNX2 form is normal stature (10th to 90th centiles in the initial series; adults 165 to 172 cm) and a higher frequency of developmental delay (3 of 8 individuals, mild in 2 and moderate in 1), which is not universal: there was no delay in the family reported by Pagnamenta and colleagues [3, 11].

CBFB haploinsufficiency also occurs in 16q22.1 microdeletion syndrome (OMIM 614541), a multiple congenital anomaly syndrome with failure to thrive, psychomotor delay and craniofacial dysmorphism. In the comparison by Beyltjens and colleagues, the 8 published cases with a deletion including CBFB had growth below the 3rd centile in 6 of 8, developmental delay in 8 of 8 and no reported dental or clavicular anomaly, whereas the 8 patients with an intragenic variant had abnormal clavicles in 6 of 7 and dental anomalies in 4 of 7 [3]. Khan and colleagues described an infant with an interstitial 16q21q22 deletion, delayed cranial ossification, cleft palate, congenital heart disease and feeding difficulties, with a favourable course [41].

Germline loss-of-function CBFB variants should not be confused with somatic alterations of the same gene. The inversion inv(16)(p13.1q22) or translocation t(16;16)(p13.1;q22), which generate the CBFB::MYH11 fusion, are acquired and define favourable-risk acute myeloid leukaemia in the 2022 ELN (European LeukemiaNet) classification [4, 42]. The only association recorded in GenCC (Gene Curation Coalition) between CBFB and acute myeloid leukaemia is classified as “limited”, and the germline variant associated with CCD does not indicate hereditary leukaemia predisposition or a need for haematological surveillance [4].

3. Clinical Spectrum and System-Specific Manifestations

3.1. Sources of Natural History Data

The natural history description is based on a few cohorts, all with limitations. Cooper and colleagues recorded the natural history of 90 probands and 56 first- and second-degree relatives in the USA in 2001 and are the source of most frequencies cited by GeneReviews [2, 43]. Roberts and colleagues reviewed experience with more than 100 affected individuals in South Africa [2, 44]. Berkay and colleagues described 51 patients from 30 families in Turkey [31]. Kashiki and colleagues compiled 605 published individuals, but frequencies in that compilation depend on what each report described and probably underestimate findings that were not sought, so they are not systematically ascertained prevalences [34]. None of these series allows rigorous separation of the spontaneous course from the course under dental and orthopaedic treatment, because most patients were treated; the skeletal pattern described is nevertheless congenital and largely unmodifiable. The sources consulted include no series from Portugal or Brazil; the Brazilian book Síndromes Genéticas: Reconhecer para Tratar cites only Brazilian case reports [26].

Table 3.1 — Frequency of Manifestations in CCD Patient Series

Table sources: Cooper 2001 via GeneReviews [2, 43]; Berkay 2021 [31]; Kashiki 2026 [34]. The difference between 39% conductive hearing loss in GeneReviews and 1.8% in the published-case compilation illustrates underreporting in the latter, in which hearing was rarely assessed systematically [2, 34].

3.2. Skull and Face

The skull is brachycephalic, with frontal, parietal and occipital bossing and a broad, flat forehead; fontanelles are large and open at birth and may remain open throughout life, and persistence of the metopic suture produces a midline metopic groove [2, 26]. In the Bone Dysplasias atlas, a three-generation family illustrates this persistence: the 34-year-old father still had an open anterior fontanelle and metopic suture, with basilar impression, and the “inverted pear” cranial profile is visible at several ages [1]. The face is small relative to the skull, with midface hypoplasia, a low nasal root, hypertelorism, exophthalmos in some, a small maxilla, a narrow high palate and sometimes cleft palate and micrognathia [14, 26]. Pneumatisation of the paranasal sinuses, frontal sinuses and mastoid air cells is deficient or absent [2]. A Japanese series of 11 patients studied by three-dimensional computed tomography found wormian bones and hypoplastic mastoid processes in all and described new findings, always bilateral: aplastic or hypoplastic styloid processes, thin orbital walls, inferior displacement of the temporal process of the maxilla and an atypical coronoid process [45]. Some families have severely delayed cranial ossification with cranium bifidum [48]. Absent parietal ossification, atlantoaxial subluxation and syringomyelia occur occasionally [14]. The atlas records a single patient with myelopathy due to atlantoaxial dislocation [1].

3.3. Clavicles, Shoulder Girdle and Thorax

Clavicular hypoplasia is typically bilateral but not necessarily symmetrical, ranging from complete absence to hypoplasia or discontinuity, with more marked involvement of the lateral than medial portions; two unfused fragments may be present [1, 2]. Clinically, the spectrum ranges from a skin dimple to narrow, sloping shoulders with unusual mobility allowing them to be brought together at the midline; an impalpable clavicle is accompanied by associated muscular defects [1, 14]. The scapulae are small, sometimes winged or deformed, and the thorax is narrow and cone-shaped, with a reduced upper diameter; short oblique ribs, a cervical rib and pectus excavatum may be present [1, 2, 14]. The narrow chest may cause respiratory difficulty in early childhood [14]. Habitual dislocation of the shoulder and radial head and elbow dislocation are less frequent complications [1, 2].

3.4. Dentition

Dental anomalies have the greatest lifelong impact and are present in up to 94% of patients, in the form of delayed permanent tooth eruption and failure of deciduous tooth exfoliation [2, 46]. The most consistent findings are the presence of the permanent second molar alongside the primary dentition (80%), wide spacing in the lower incisor region, supernumerary tooth buds (70%) and parallel-sided mandibular rami [2]. Permanent teeth may show aplasia, malformed roots, retention cysts and enamel hypoplasia, with increased caries, and gingival cysts around supernumerary teeth and a Class III reverse bite are more frequent [2, 14]. Dental age is typically delayed by approximately 3 years, and supernumerary teeth and delayed eruption become apparent only from 6 years, contributing little to diagnosis in younger children [10]. In a study of 11 Chinese patients with 37 unerupted second premolars, impaction was vertical in 62.16%, palatal in 27.03% and buccal in 10.81% [49]. Palatal bone thickness measured by CBCT is greater in patients with CCD than in controls [50]. Dental problems become more evident in adolescence and adulthood and affect chewing, speech and quality of life [10].

3.5. Axial Skeleton, Pelvis and Limbs

The pelvis is narrow, with delayed pubic ossification (absent in early childhood), wide pubic symphysis and sacroiliac joints, hypoplastic iliac wings and an elongated femoral head with a short neck and “chef's hat” epiphysis; the femoral neck may be in valgus or, less frequently, varus [1, 2]. The spine may show scoliosis, kyphosis, spondylolysis and spondylolisthesis, and vertebral body ossification is delayed [1, 14]. In the hands, the 2nd metacarpal is long, there are numerous large pseudoepiphyses in the metacarpals and metatarsals, short tapering distal phalanges, short middle phalanges of the 2nd and 5th digits (or the 3rd to 5th) with cone-shaped epiphyses, short broad thumbs and sometimes hypoplastic nails [2, 14, 26]. Carpal and tarsal ossification is delayed [1]. The most frequent orthopaedic complications in Cooper's cohort were flat feet (57%), genu valgum (28%) and scoliosis (18%) [2, 43]; hip dislocation is less frequent [1]. Joint hypermobility and muscular hypotonia are occasional findings [1].

3.6. Growth

Growth deficiency is postnatal and mild to moderate. In Cooper's cohort, males were on average approximately 15 cm shorter than their unaffected brothers, with a mean height of 165 ± 8 cm, and females approximately 7.5 cm shorter than their sisters, at 156 ± 10 cm [1, 2, 43]. IGF-1 (insulin-like growth factor 1) may be low [2, 26]. Stature is normal in the CBFB-associated form [3].

3.7. Bone Health

Osteopenia and osteoporosis are part of the phenotype. In the series by Dinçsoy Bir and colleagues (15 patients from 11 families), DXA showed osteoporosis in 8 of 14 (57%) and osteopenia in 3 of 14 (21%); low vitamin D was frequent but without a consistent association with osteoporosis [2, 47]. Serum alkaline phosphatase is rarely low, and urinary phosphoethanolamine may be elevated, prompting consideration of hypophosphatasia [1, 2, 51]. The Delphi panel considers a history of unexpected or premature fractures and low bone density typical and emphasises that bone density does not necessarily improve over time and that risk increases after menopause [10]. The adult patient described by Quack and colleagues had osteoporosis with recurrent fractures [20].

3.8. Otorhinolaryngology and Airway

Recurrent sinusitis and other upper airway complications are significantly more frequent than in the general population, conductive hearing loss affects 39% of patients and recurrent otitis media occurs at any age [2, 52]. Upper airway obstruction may cause obstructive sleep apnoea [2]. Deafness is recorded as an occasional anomaly in Smith [14].

3.9. Development, Cognition and Psychosocial Aspects

Intelligence is normal. Before 5 years, there may be mild motor delay, predominantly gross motor, possibly related to flat feet and genu valgum, without significant differences at school age [2, 26]. The Delphi panel recommends that delayed walking, language or puberty should not be attributed to CCD without assessment, because RUNX2 may be included in a contiguous gene deletion associated with neurocognitive impairment [10]. Patients report anxiety, depression, isolation and bullying, including cyberbullying, and stress related to their smile and the number of dental procedures [10]. In an oral quality-of-life study (OHIP-14, Oral Health Impact Profile) of 61 patients aged 15 years or older, posterior crossbite was associated with worse physical pain, anterior open bite with worse disability and anterior crossbite with worse functional limitation [53].

3.10. Pregnancy

The narrow pelvis, with a small pelvic inlet, predisposes to cephalopelvic disproportion. In Cooper's cohort, the primary caesarean section rate in women with CCD was 69%, higher than in controls [2, 43]. The atlases and Smith record that caesarean section may be necessary because of pelvic narrowing [1, 14].

3.11. Spectrum Forms and Associated Entities

In classic CCD, the triad is complete. In mild CCD, clavicular and cranial changes are subtle, and in the form with isolated dental anomalies, there is only delayed permanent tooth eruption and supernumerary teeth, without other skeletal findings; these forms are chiefly associated with hypomorphic alleles and coexist within the same family [2, 21]. The severe form may mimic hypophosphatasia in early childhood, with a markedly undermineralised skull and low alkaline phosphatase [51]. RUNX2 deletions involving adjacent genes add developmental delay [6].

In CBFB-associated CCD, clavicular findings range from a unilateral bipartite clavicle to bilateral aplasia; supernumerary teeth (4 and 13 in two individuals) and failure of eruption, metacarpal pseudoepiphyses, short distal phalanges and delayed carpal ossification occur; stature is normal, some patients have developmental delay and hearing loss occurred in 2 of 8 [3]. In the British family reported by Pagnamenta and colleagues, the proband had a right clavicular fracture or pseudarthrosis at 8 weeks, a large anterior fontanelle, mild motor delay, hypermobility and supernumerary teeth with abscesses, requiring extraction of 8 deciduous teeth, gingival surgery and successful orthodontic treatment; final height was 163.4 cm (50th centile), and DXA at 14 and 15 years gave lumbar spine Z-scores of −1.4 and −0.9; her mother had a large fontanelle, supernumerary teeth, unerupted central incisors, recurrent otitis, mild scoliosis and clinically normal clavicles, with a height of 165 cm [11].

In MDMHB, the medial half of the clavicles is widened rather than hypoplastic, and short stature, metaphyseal dysplasia and dystrophic teeth predominate [2].

4. Clinical Recognition and Warning Signs

4.1. When to Suspect the Diagnosis

The Delphi panel recommends considering and assessing a clinical diagnosis of CCD in patients with short or absent clavicles, delayed fontanelle closure and/or dental complications (supernumerary teeth and/or non-spontaneous or delayed eruption of permanent teeth), with 100% consensus (13/13) [10]. The panel did not reach consensus on criteria distinguishing CCD from other conditions [10]. Primary care professionals, dentists and orthodontists are most likely to see the initial presentation, and the specialists most involved in diagnosis are those in primary care, genetics, dentistry and orthodontics [10].

Table 4.1 — Contexts Raising Suspicion by Age

The anterior fontanelle normally closes at 12 ± 4 months and the posterior fontanelle around birth, with variation of ± 2 months [48]. For a large fontanelle, the Oxford Desk Reference proposes a three-generation pedigree (consanguinity suggests pycnodysostosis or autosomal recessive Robinow syndrome; other affected relatives suggest CCD or parietal foramina), examination of the clavicles, a skull radiograph and possibly a complete skeletal survey, bone age, a bone profile (calcium, phosphate and parathyroid hormone) and exclusion of hypothyroidism, rickets, congenital infections and intracranial hypertension [48].

4.2. Warning Signs and Emergencies

CCD has no characteristic acute decompensations, but some situations require prompt action. Respiratory difficulty in a newborn or infant due to a narrow chest, signs of obstructive sleep apnoea, recurrent otitis or sinusitis with hearing loss, head trauma with significant calvarial defects, neurological symptoms suggesting myelopathy (basilar impression or atlantoaxial instability) and low-impact fractures are warning signs [1, 2, 10, 14]. A predictably difficult airway due to dental and craniofacial anomalies should be flagged before any anaesthesia [2]. Cephalopelvic disproportion should be anticipated in pregnancy [2]. Global developmental delay or additional malformations should prompt investigation for a contiguous gene deletion or 16q22 microdeletion [6, 10].

4.3. Newborn Screening

There is no newborn screening for CCD. Prenatal detection by ultrasonography is possible in fetuses of an affected parent (section 5.6) [2].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Diagnostic Criteria

There are no formal clinical diagnostic criteria. Diagnosis is established in a proband with characteristic clinical and radiographic findings or by identification of a heterozygous pathogenic or likely pathogenic RUNX2 variant; a variant of uncertain significance (VUS) neither establishes nor excludes the diagnosis [2]. The Delphi panel considers genetic testing sufficient but not necessary: if unavailable, CCD can be diagnosed clinically [10].

Table 5.1 — Suggestive Findings (GeneReviews)

Source: [2]. The Delphi panel adds that dental screening by panoramic radiography or CBCT may be necessary to identify changes not visible on oral examination (91%, 10/11) and that radiographs of the skull, chest and/or clavicles should be considered according to symptoms and physical examination (100%, 13/13) [10].

5.2. Imaging

Radiology is the cornerstone of clinical diagnosis [6]. Skull radiographs show delayed ossification, partial lack of calvarial and skull base ossification, multiple wormian bones, platybasia, a foramen magnum enlarged or deformed by a midline occipital cleft and a small sphenoid [1, 14]. Chest radiographs show partial, less often complete, absence of the clavicles, more pronounced in the lateral and middle thirds [1]. Pelvic radiographs show absent pubic ossification in early childhood and later underdevelopment of the pubis and ischium, with wide symphysis, sacroiliac joints and triradiate cartilage, and lateral notching of the proximal femoral ossification centres [1, 14]. In the hands, metacarpal pseudoepiphyses fuse during childhood [14]. DXA documents bone density and should be interpreted with adjustment for stature (section 11) [10, 56]. Cranial computed tomography and magnetic resonance imaging show delayed ossification and open sutures, with normal parenchyma [24].

5.3. Biomarkers

There is no specific biomarker. Serum alkaline phosphatase and IGF-1 may be low [2]. Kashiki and colleagues described a Japanese infant with a 6-cm anterior fontanelle and serum alkaline phosphatase of 126 U/l (IFCC method), in whom tissue-nonspecific alkaline phosphatase (TNSALP) activity in a dried blood spot was normal (590.7 pmol/h/disc), excluding hypophosphatasia before exome sequencing confirmed CCD [34]. The calcium–phosphate profile, parathyroid hormone and thyroid function are used to exclude rickets and hypothyroidism [48].

5.4. Molecular Tests and Algorithm

The GeneReviews strategy starts with RUNX2 sequencing and, if negative, targeted deletion and duplication analysis (qPCR, long-range PCR, MLPA or targeted microarray); alternatively, a multigene panel including RUNX2 and differential diagnosis genes may be used. Karyotyping is indicated when RUNX2 testing is non-diagnostic in a patient with multiple congenital anomalies and/or developmental delay, to detect rearrangements or translocations at 6p21.1 without a copy number change [2]. When CCD is suspected, the Delphi panel recommends testing for pathogenic RUNX2 variants including gene deletions (100%, 13/13); testing CBFB, particularly when RUNX2 is negative; and microarray for 16q21q22 microdeletion when the triad coexists with congenital heart disease, developmental delay, cognitive impairment, microcephaly or other dysmorphic features [10]. Inconclusive or non-diagnostic results require follow-up with a geneticist or genetic counsellor (92%, 12/13) [10]. The Brazilian book proposes a panel or exome sequencing when the clinical presentation is atypical [26].

Table 5.2 — Diagnostic Algorithm

5.5. Test Limitations

A negative result does not exclude CCD. Exome sequencing does not interrogate deep intronic or untranslated regions: the variant c.860-2346G>T was detected only by fibroblast RNA studies and long-read genome sequencing, despite the exome covering 100% of target bases with more than 20 reads [28]. Small CNVs outside coding regions, such as the 4.2-kb 3'UTR deletion, escaped microarray and exome analysis and were detected only by genome sequencing [29]. Complex structural CBFB rearrangements may escape array comparative genomic hybridisation (array-CGH) and read-depth-based CNV calling algorithms: the rearrangement reported by Pagnamenta and colleagues was detected by neither array-CGH nor the Canvas algorithm, and only 3.5% of gnomAD genomes had coverage of 30× or more at the breakpoint [4, 11]. Targeted CNV methods may not detect large deletion breakpoints or deletions of adjacent genes, and balanced rearrangements require karyotyping [2]. Testing leucocyte DNA may miss low-level somatic mosaicism and does not detect exclusively germline mosaicism [2]. The proportions in Table 2.1 are fractions of probands with a detectable variant and do not equate to the clinical sensitivity of the test.

The report should specify the gene, transcript with version (for RUNX2, NM_001024630.4; for CBFB, NM_022845.3), HGVS variant, zygosity, ACMG/AMP (American College of Medical Genetics and Genomics/Association for Molecular Pathology) classification and criteria used. Comparison with the literature requires conversion of designations published using other transcripts, as illustrated by p.R211W (Table 2.2) [7, 24]. A VUS may be reclassified by demonstrating de novo origin with confirmed paternity, as occurred with the 3'UTR deletion, and by segregation in families with several affected individuals [28, 29]. The Delphi panel recommends genetic testing of relatives with compatible features (85%, 11/13) [10].

5.6. Prenatal and Preimplantation Diagnosis

Ultrasonographic diagnosis of classic CCD is possible from 14 weeks of gestation (gestational age in menstrual weeks) in fetuses of an affected parent; the most consistent findings are short clavicles, below the 5th centile for gestational age according to GeneReviews, or partially or completely absent clavicles, and less specific findings are an undermineralised brachycephalic skull, frontal bossing and generalised immature ossification [2, 54, 55]. The Brazilian book gives the 50th centile as the threshold, differing from its own source; the threshold adopted is that of GeneReviews [2, 26]. Molecular prenatal testing and preimplantation genetic testing are possible when the familial variant is known [2, 26]. When CCD is suspected prenatally, the Delphi panel recommends close surveillance and referral to perinatology and neonatology [10].

6. Differential Diagnosis

6.1. Approach

Clavicular hypoplasia, delayed fontanelle closure and dental anomalies are each individually non-specific. Table 2 of the Delphi panel groups differential diagnoses by the shared key feature: short or absent clavicles (CDAGS syndrome, Crane–Heise syndrome, mandibuloacral dysplasia, Yunis–Varon syndrome); delayed fontanelle closure (CDAGS syndrome, Cornelia de Lange syndrome, hypothyroidism, mandibuloacral dysplasia, pycnodysostosis, other skeletal dysplasias and vitamin D deficiency); and dental complications (familial supernumerary teeth and Gardner syndrome) [10]. The panel presented this list as exploratory, without formal consensus [10]. The combination of the three domains, with the radiological pattern of the pelvis and hands, points towards CCD.

Table 6.1 — Differential Diagnosis of Cleidocranial Dysplasia

AD: autosomal dominant; AR: autosomal recessive.

6.2. Pycnodysostosis

Pycnodysostosis is the differential diagnosis with the greatest craniofacial overlap. It is an autosomal recessive disorder caused by biallelic CTSK variants, encoding cathepsin K, a lysosomal cysteine protease of osteoclasts [58]. It shares with CCD an open anterior fontanelle even in adults, delayed suture fusion, wormian bones, non-pneumatised mastoids, maxillary hypoplasia and retained deciduous teeth [58, 59]. It is distinguished by generalised osteosclerosis due to increased trabecular rather than cortical bone, bone fragility with fractures (on average 0.2 per year, with the first at approximately 10 years), progressive acro-osteolysis of distal phalanges in more than 90% and of the acromial end of the clavicle in 25%, an obtuse mandibular angle, blue sclerae, a “parrot-beak” nose and short-limbed short stature, with adult heights of 134 to 152 cm [58, 59]. Clavicular hypoplasia is usually more pronounced in CCD, and the abnormal shape of the pelvis and pubis and thoracic deformities are characteristic of CCD [58, 59]. Approximately half of patients with pycnodysostosis have growth hormone deficiency and almost all have low IGF-1; growth hormone almost normalised adult height in a published series, contrasting with the absence of a growth hormone recommendation in CCD [59]. Mandibular osteomyelitis following tooth extraction and obstructive sleep apnoea are frequent complications [60]. GeneReviews estimates a frequency of 1 to 1.7 per million and approximately 200 reported affected individuals [59]. The incidence of 1:2000 live births given in a Brazilian book is incompatible with these data and was not adopted [59, 71].

6.3. Hypophosphatasia

Severe CCD may mimic hypophosphatasia in early childhood because of the markedly undermineralised calvaria, widened sutures and low alkaline phosphatase; Unger and colleagues described a child with severe CCD initially diagnosed with hypophosphatasia [2, 51]. In hypophosphatasia, the mineralisation defect is generalised, with rachitic changes, premature loss (rather than retention) of deciduous teeth, nephrocalcinosis and hypercalcaemia, and less affected clavicles [2]. In the case reported by Kashiki and colleagues, normal TNSALP activity in a dried blood spot, despite low serum alkaline phosphatase, allowed hypophosphatasia to be excluded [34].

6.4. Differential Diagnoses with Clavicular Hypoplasia and Extraskeletal Malformations

Yunis–Varon syndrome combines clavicular hypoplasia with absent thumbs and great toes, central nervous system malformations (arrhinencephaly, agenesis of the corpus callosum, vermian dysplasia), sparse hair, sclerocornea, cataracts and heart disease, with high neonatal mortality; absent thumbs are not obligatory, and Charcot–Marie–Tooth-like peripheral neuropathy may occur [48, 62]. CDAGS syndrome (craniosynostosis and clavicular and cranial ossification defects, anterior or imperforate anus, genitourinary anomalies and porokeratosis) paradoxically combines accelerated suture closure with delayed ossification; the 2018 Bone Dysplasias atlas listed an unknown gene at 22q12–q13, but Xing and colleagues identified biallelic variants in RNU12, encoding a small nuclear RNA, in 2021 [2, 63, 64]. Crane–Heise syndrome (OMIM 218090, number confirmed by GeneReviews and Orphanet; the Bone Dysplasias atlas uses divergent numbers in different chapters) is lethal and combines poorly mineralised calvaria, cleft lip and palate and absent cervical vertebrae [2, 64]. In mandibuloacral dysplasia, clavicular osteolysis produces narrow shoulders that can be brought together as in CCD, but distal phalanges are small because of osteolysis rather than hypoplasia, and lipodystrophy and skin changes are present [1, 65]. The atlas also records an isolated patient with hypoplastic clavicles, severe myopia, a single central incisor and peripheral neuropathy [1].

6.5. Molecular Phenocopies

When RUNX2 and CBFB are negative, microduplications upstream of MSX2 should be considered: Ott and colleagues associated them with a CCD phenocopy, sometimes with synpolydactyly, and in 2026 Yamada and colleagues described two unrelated families with persistent anterior fontanelles and non-coding duplications delineating a minimal regulatory region (chr5:173848716-173888281) within a single topologically associating domain; the causal relationship is based on four families and in silico chromatin analyses [57, 66]. Intragenic MSX2 variants cause parietal foramina with clavicular hypoplasia, without dental anomalies [2].

7. Treatments and Risk Management Strategies

7.1. General Principles and Sources of Recommendations

There is no disease-modifying therapy or authorised drug for CCD; management is supportive, multidisciplinary and symptom-driven [2, 10]. The 2023 GeneReviews stated that no clinical practice guidelines had been published and based its recommendations on the authors' opinions [2, 6]. In 2026, Hoover-Fong and colleagues published the first consensus recommendations, obtained through a three-round modified Delphi panel of 13 experts (10 from the USA, 2 from Israel and 1 from the United Kingdom; 4 orthodontists, 2 geneticists, 2 otorhinolaryngologists, 1 dentist, 1 rare bone disease endocrinologist, 1 family nurse practitioner, 1 orthopaedic surgeon and 1 paediatrician; two experts have CCD), with a consensus threshold of 70% [10]. Consensus was reached on 26 of 33 statements in the first round, 21 of 23 in the second and 12 of 12 in the third, totalling 37 recommendations (9 diagnostic, 7 dental and orthodontic, 17 non-dental medical and 4 concerning providers) [10]. The scope covers adults and children aged 2 years or older; infants younger than 2 years were excluded [10]. This is expert opinion without evidence grading; the work was funded by Costello Medical through a pro bono grant to the CCD Smiles Foundation, a patient association whose founder and a board member participated in the study, and the authors declared pharmaceutical company consultancies [10]. The panel is small and predominantly US-based, and the steering committee also answered the surveys [10].

Table 7.1 — Sources of Recommendations for CCD Management

No CCD-specific guideline issued by an ERN, ACMG or the European Society of Human Genetics was identified; the institutional pages of ERN BOND and ERN CRANIO do not name CCD [12, 13].

7.2. Assessment Following Diagnosis

Following diagnosis, GeneReviews recommends a clinical skeletal assessment with a complete skeletal survey including hands and feet, DXA in early adolescence, assessment by a dentist familiar with CCD, audiological assessment and genetic counselling [2]. The Delphi panel adds hearing testing at diagnosis (100%, 13/13), orthopaedic assessment of function and mobility at diagnosis (100%, 12/12), DXA as soon as the child can cooperate, typically at approximately 7 years (100%, 10/10), and orthodontic screening by an orthodontist experienced in CCD by 6 years (100%, 13/13) [10]. Patients should receive the same fundamental healthcare as the general population (paediatrics and primary care), in addition to specialist care (100%, 13/13) [10]. Coordination is multidisciplinary and symptom-driven, involving primary care or paediatrics, dentistry and orthodontics, oral and maxillofacial surgery, physiotherapy and occupational therapy, orthopaedics, genetics, otorhinolaryngology and endocrinology; in newborns and infants (0 to less than 2 years), coordination lies with primary care and/or genetics (100%, 13/13) [10].

7.3. Dental and Orthodontic Treatment

The Delphi panel defines dental management as medical rather than cosmetic because of its impact on nutrition, speech and overall health [10]. Oral hygiene should start with eruption of the first deciduous tooth: brushing twice daily with age-appropriate fluoride toothpaste, flossing, regular consultations and cleaning, because caries and gingivitis may delay or prevent surgery and orthodontic treatment [10]. Watchful waiting for spontaneous eruption is ineffective and contraindicated: extraction of deciduous teeth alone does not accelerate permanent tooth eruption, and permanent teeth with malformed roots are difficult to extract [2, 10, 14]. GeneReviews advocates early referral to a dental clinic familiar with CCD and an active approach, coordinating several oral operations to remove the primary dentition and expose the permanent dentition, aiming for functional appearance and chewing [2].

Table 7.2 — Dental and Orthodontic Recommendations of the Delphi Panel (2026)

Source: [10].

The Jerusalem approach, described by Becker and colleagues in 1997, involves two operations: the first at 10–12 years, extracting anterior deciduous and supernumerary teeth, exposing permanent incisors and applying traction; the second after 13 years, extracting the remaining deciduous teeth and exposing and applying traction to canines and premolars [10, 73, 74]. The Belfast–Hamburg, Bronx and Toronto–Melbourne paradigms expose all teeth in a single operation and are more intensive [10]. Because dental age is delayed by approximately 3 years, treatment should follow dental age; when teeth lie very deeply, extracting deciduous and supernumerary teeth may suffice to allow spontaneous eruption, and prior maxillary expansion and protraction may be required [10]. Treatment may continue for more than 10 years, and the panel acknowledges that “standard orthodontic treatment” in CCD is not well established [10].

Based on 8 French cases, Farrow and colleagues proposed always attempting orthodontically assisted eruption, adding orthognathic surgery to stabilise occlusion and using prostheses when orthodontics is only partially effective; if it fails, they recommend preserving native teeth and combining pre-prosthetic orthognathic surgery with implant-supported prostheses [30]. The systematic review by Zhu and colleagues (8 articles and 9 completed cases, ages 9 to 28 years) concluded that surgical exposure with orthodontic traction is effective, that early clearance of the eruption pathway facilitates spontaneous eruption, that adults can achieve an almost complete dentition at the cost of prolonged treatment, and that orthognathic surgery reduces difficulty and duration [75]. In a Chinese study of 11 patients with 37 unerupted second premolars, assisted closed eruption lasted 29.00 ± 6.84 months, with a success rate of 89.29%; 9 of the 37 teeth failed to erupt, 6 in a low palatal position with complete apices, and failure was associated with coronal and axial position and Nolla stage 9–10 [49]. Palatal bone thickness, greater in patients with CCD than controls in all measured regions, permits placement of palatal temporary anchorage mini-implants following individual assessment [50]. Sobti and colleagues described extraction of deciduous and supernumerary teeth and exposure, bonding and forced eruption of multiple unerupted teeth using a closed technique in a 21-year-old woman [76]. In Zinn and colleagues' oral quality-of-life study (61 patients aged 15 years or older), prosthetic treatment was associated with poorer overall quality of life than orthodontic treatment, in an observational design in which treatment selection may reflect severity [53]. Speech therapy is indicated during dental treatment and when hearing loss affects speech [2].

7.4. Skull, Face and Clavicles

When the calvarial defect is significant, the head should be protected from blunt trauma, with a helmet during high-risk activities, and these patients should be assessed by a craniofacial surgeon [2]. The Delphi panel considers that a helmet may benefit children with delayed fontanelle closure (100%, 13/13) but asks that psychosocial implications be discussed through shared decision-making [10]. Fontanelles eventually close in most patients, and cranial remodelling is generally unnecessary; the Oxford Desk Reference considers surgery generally contraindicated [2, 48]. Cosmetic correction of the metopic groove (with a bone substitute) and clavicular lengthening by distraction osteogenesis have been performed successfully in a very small number of patients [2, 77, 78]. In bilateral clavicular pseudarthrosis with preserved function, function-guided conservative management is acceptable, reserving surgery for pain, instability, progressive deformity, neurovascular or respiratory compromise and functional limitation [24].

7.5. Otorhinolaryngology, Hearing and Sleep

Sinusitis and otitis should be treated promptly; tympanostomy tubes should be considered for recurrent otitis, and routine vaccination should include influenza vaccination [2, 52]. A sleep study is indicated when symptoms of sleep-disordered breathing are present (100%, 13/13), and airway surgery may be necessary [2, 10].

7.6. Orthopaedics and Rehabilitation

The Delphi panel recommends proactive, frequent assessment of mobility and activities of daily living, particularly during growth, by a physiotherapist, occupational therapist or rehabilitation physician (100%, 13/13), and physiotherapy when function or mobility problems occur (100%, 13/13) [10]. GeneReviews recommends orthopaedic assessment of flat feet, genu valgum, scoliosis and shoulder and elbow dislocations at every visit throughout childhood [2]. Counselling, mental health support and pain management should be available (100%, 13/13) [10].

7.7. Bone Health and Pharmacological Therapy

GeneReviews recommends calcium and vitamin D when DXA shows below-normal bone density, without specifying a dose [2]. The Brazilian book, derived from earlier GeneReviews versions, phrases the recommendation more broadly, suggesting consideration of preventive calcium and vitamin D supplementation [26]. Before any drug, the Delphi panel recommends a diet with adequate calcium and vitamin D and daily physical exercise according to general age-specific recommendations, and additional bone health treatment when low bone density or multiple fractures are present (100%, 12/12), but acknowledges that no preclinical or clinical study has tested bone drugs in CCD [10]. The sources consulted do not give CCD-specific doses of calcium, vitamin D or any bone drug; applicable doses are those in general age-specific recommendations.

Bisphosphonates have not been evaluated in CCD. Because the defect concerns bone formation rather than resorption, the panel suggests that they may offer little benefit and could even impair skeletal development, and notes that they hinder orthodontic movement, relevant in a disorder requiring years of dental traction [10]. If anabolic therapy is preferred during traction of unerupted teeth, the panel suggests that parathyroid hormone analogues and anti-sclerostin antibodies could be considered, but emphasises that anabolic pathways (intermittent parathyroid hormone, analogues and anti-sclerostin) converge on RUNX2, so efficacy in CCD may be limited [10]. Any of these drugs would be off-label use, without efficacy or safety data in CCD.

Table 7.3 — Drugs Discussed in CCD

Growth hormone is the main point of divergence. GeneReviews does not recommend any treatment for short stature because growth hormone efficacy has not been proven and adverse effects on the growth plate are theoretically possible, given RUNX2's role in chondrocyte differentiation and growth plate maintenance [2, 25]. The Delphi panel issued no recommendation on this subject [10]. Data are limited to case reports. Çamtosun and colleagues described a 5.5-year-old girl with severe short stature and the variant c.463A>C p.(Thr155Pro), who started growth hormone at 7 years and 2 months and had increased growth velocity without adverse effects; the dose is not in the abstract consulted [33]. In the case reported by Xu and colleagues, a boy with treated hypothyroidism started somatropin at 0.05 mg/kg/day at approximately 8 years, and height standard deviation score rose from −1.56 at 7 years and 1 month to +0.09 at 12 years and 1 month, without documented adverse effects; the authors acknowledge that the effect cannot be separated from treated hypothyroidism, spontaneous growth and puberty (section 10) [24]. These reports do not allow estimation of efficacy or safety, and growth hormone should be treated as off-label use, with an individual, documented decision.

7.8. Regulatory Status

Table 7.4 — Regulatory Status on 10 October 2026

Sources: EMA and FDA searches recorded in the source collection record; [10, 79]. A negative search does not demonstrate the absence of designations or non-public programmes.

7.9. Clinical Trials

No therapeutic trials in CCD were identified. A ClinicalTrials.gov search for “cleidocranial” returned a single observational study [79].

Table 7.5 — Registered Studies

Source: [79]. In the absence of trials, all therapeutic evidence comes from case series, small observational studies and expert opinion. The cited dental studies (89.29% success in closed eruption of premolars; quality-of-life differences between prostheses and orthodontics) are observational, without a randomised control group, and do not allow surgical–orthodontic protocols to be compared with one another [49, 53]. No negative, suspended or withdrawn trials were identified.

7.10. Developments in the Last 12 Months

Table 7.6 — Publications from October 2025 to October 2026

7.11. Anaesthesia and Perioperative Care

Dental and craniofacial anomalies make the airway predictably difficult, and anaesthetic planning should include otorhinolaryngology consultation to secure the airway; alternatives include neuraxial blockade, taking possible spinal anomalies into account [2, 80]. Ioscovich and colleagues described anaesthetic management of a patient with CCD undergoing several obstetric procedures [80].

7.12. Pregnancy, Delivery and Lactation

The Delphi panel recommends that pregnant women with CCD have regular visits to assess pregnancy safety and plan delivery (100%, 13/13), and that a pregnant woman with CCD or suspected fetal CCD receive close surveillance with referral to perinatology and neonatology (100%, 13/13) [10]. Surveillance for cephalopelvic disproportion is necessary because the narrow pelvis may require caesarean section; the primary caesarean section rate was 69% in Cooper's cohort [2, 43]. The Delphi panel aligns with the guideline by Savarirayan and colleagues for skeletal dysplasias, advocating genetic testing and counselling, maternal–fetal medicine follow-up, early delivery planning and possible delivery at a tertiary centre [10, 72]. The sources consulted describe no breastfeeding restrictions or specific drugs to avoid during pregnancy in women with CCD.

7.13. Agents and Activities to Avoid

GeneReviews recommends helmets and protective equipment during high-risk sports and activities to prevent head injury [2]. Bisphosphonates should be considered cautiously during orthodontic treatment because they interfere with tooth movement [10]. The sources identify no drugs contraindicated by the disorder itself.

7.14. Transition to Adulthood

Orthodontic treatment usually spans adolescence and may extend beyond 18 years, and bone density screening should continue into adulthood, with additional risk after menopause [10]. Transition should ensure continuity of the dental and maxillofacial team, audiological follow-up agreed with the otorhinolaryngologist, preconception genetic counselling and obstetric referral for women [10].

7.15. Divergences Between Sources

There are four principal divergences. In orthodontics, the Delphi panel recommends the staged Jerusalem approach as the least intensive, whereas Farrow and colleagues favour orthodontic eruption combined with orthognathic surgery and prostheses, and other paradigms expose all teeth in a single operation [10, 30]. For bone density, GeneReviews proposes DXA from early adolescence and then every 5–10 years, whereas the Delphi panel brings it forward to approximately 7 years and maintains periodic assessment throughout life without defining an interval [2, 10]. For supplementation, GeneReviews restricts calcium and vitamin D to patients with below-normal DXA, whereas the Brazilian book suggests considering them preventively [2, 26]. For growth hormone, GeneReviews advises against treatment because of lack of proven efficacy and theoretical risk, whereas the Delphi panel makes no statement [2, 10].

8. Reproductive Options and Family Counselling

8.1. Mode of Transmission and Risk to Offspring

RUNX2-associated and CBFB-associated CCD are inherited in an autosomal dominant manner; each child of an affected person has a 50% probability of inheriting the variant [2, 3, 10]. Because expressivity is variable and incomplete penetrance has not been described, a child who inherits the variant will have the disorder, but severity may differ greatly from that in the parent [2]. The Delphi panel recommends offering genetic counselling to all patients, including the probability of transmission to biological children (100%, 13/13), and preconception counselling to the patient and partner about testing, inheritance and assisted reproduction (100%, 13/13) [10].

8.2. Parental Status and De Novo Cases

Some probands have an affected parent, and GeneReviews describes the proportion of de novo cases as “high”, without a numerical value [2]. The Brazilian book gives a de novo proportion of approximately 20%–40%, a figure absent from the 2023 GeneReviews version and for which the primary source was not identified [26]. Recent reports confirm de novo origin through parental testing in several cases [24, 29, 34], and maternal transmission was documented in 3 of 5 CBFB families and in the British family with the cryptic rearrangement [3, 11].

In an apparently isolated case, GeneReviews recommends detailed clinical examination of the parents and consideration of craniofacial and skeletal radiographs if dental or bone signs are present, because the phenotype can differ between parent and child with the same variant, and molecular testing of the parents when the proband's variant is known [2]. An apparently negative family history is confirmed only by clinical examination and skeletal radiographs of the parents and/or molecular testing; in the family reported by Pagnamenta and colleagues, the carrier mother had clinically normal clavicles [2, 11]. The Delphi panel recommends genetic testing of relatives with compatible features (85%, 11/13) [10].

8.3. Mosaicism and Risk to Siblings

If the variant is not found in parental leucocytes, with maternity and paternity confirmed, it is either de novo in the proband or inherited from a parent with germline or somatic and germline mosaicism [2, 37, 38]. Testing leucocyte DNA may not detect all somatic mosaicism and does not detect exclusively germline mosaicism; a parent with somatic and germline mosaicism may be mildly or minimally affected [2]. Pal and colleagues demonstrated germline mosaicism in a family with three affected siblings and an apparently unaffected mother [37].

Table 8.1 — Recurrence Risk by Scenario

8.4. Cascade Testing, Minors and DNA Banking

Cascade testing applies to at-risk relatives, particularly those with compatible signs, and also helps determine dental and hearing surveillance for children [10]. Because CCD has childhood manifestations amenable to intervention (dental, hearing, orthopaedic), diagnosing at-risk children has direct clinical utility. The best time to assess risk and discuss prenatal and preimplantation testing is before pregnancy, and counselling should be offered to affected young adults [2]. For probands without a confirmed molecular diagnosis, GeneReviews recommends considering DNA banking because testing methods and knowledge of mechanisms will evolve [2].

8.5. Prenatal and Preimplantation Diagnosis

When the familial variant is known, prenatal testing and preimplantation genetic testing are possible [2, 26]. Ultrasonography can diagnose classic CCD in fetuses of an affected parent from 14 weeks, through short clavicles (below the 5th centile for gestational age) or absent clavicles and, less specifically, an undermineralised skull and immature ossification [2, 54, 55]. Hermann and colleagues described the use of three-dimensional ultrasonography in prenatal diagnosis [55]. The same approach applies in CBFB-associated CCD if the familial variant is known; the germline CBFB variant does not confer hereditary leukaemia risk, and this distinction should be explained to families [4].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence and Incidence

CCD frequency estimates vary by more than an order of magnitude because they measure different things. The only population-based ascertainment cited is that of Stevenson and colleagues, who found a frequency of 0.12 per 10,000 individuals (1.2 per 100,000, or approximately 1 in 83,000) in the Utah population, USA; this is a Utah-specific skeletal dysplasia study based on phenotype rather than molecular confirmation, suggesting a higher frequency than previously recognised [2, 6, 9]. Smith cites the same value, and Beyltjens and colleagues refer to it as an incidence of up to 1 in 80,000 births in a Utah population [3, 14]. According to GeneGuidance, Orphanet records a birth prevalence of 1.2 per 100,000 in the USA (Stevenson 2012), worldwide point prevalence of 0.1 per 100,000 (expert estimate) and European birth prevalence of 0.4 per 1,000,000, based on EUROCAT registries from 2005 to 2012 [4, 8]. The Delphi panel and Brazilian book give approximately 1 per 1,000,000 worldwide, without a population study of their own [10, 26]. The difference between Utah ascertainment and European congenital anomaly registries may reflect underdetection at birth, because mild and dental forms become apparent only later and escape neonatal registries. The disorder occurs in all ethnic groups, without sex predominance [26].

Table 9.1 — CCD Frequency Estimates

9.2. Populations with Published Series

South Africa has the largest reported experience, with more than 100 affected individuals reviewed by Roberts and colleagues, who discussed dental, historical and practical implications [2, 44]. The Turkish series of 51 patients from 30 families, the Thai series, the Japanese computed tomography series and Chinese cases show that the phenotype is the same across populations [31, 36, 45, 49]. In Europe, series and reports include Farrow and colleagues' French cases, the Austrian family with the deep intronic variant, the Serbian 3'UTR deletion case, the British family from the 100,000 Genomes Project and the families with CBFB variants described by Beyltjens and colleagues [3, 11, 28–30]. The sources consulted describe no RUNX2 founder variants.

The sources consulted include no epidemiological data from Portugal or Brazil. The Brazilian book cites four Brazilian case reports published between 2007 and 2013, which were not read [26]. The Delphi panel cites a systematic review of South American cases (Cano-Pérez, 2022), which was also not consulted [10]. The absence of data in the sources consulted does not mean there are no publications.

9.3. Special Populations

In pregnancy, the narrow pelvis influences the mode of delivery (section 7.12) [2, 10]. For infants younger than 2 years, for whom the Delphi panel made no recommendations, coordination lies with primary care and genetics [10]. In adults, particularly postmenopausal women, the risk of low bone density increases [10]. Barriers to genetic testing identified by the panel include lack of insurance, lack of knowledge about the disorder and testing process, and patients' and families' fears (100%, 13/13) [10].

9.4. European Networks and Patient Associations

ERN BOND, the European Reference Network for rare bone diseases, coordinated by the Istituto Ortopedico Rizzoli in Bologna, is the relevant network for the skeletal component [12]. ERN CRANIO, covering craniosynostosis, craniofacial anomalies, clefts, orodental anomalies and otorhinolaryngology, is relevant to the craniofacial, dental and hearing components [13]. Neither names CCD on the institutional pages consulted [12, 13]. The CCD Smiles Foundation, a non-profit organisation founded in the USA on 25 January 2017, connects patients, maintains a state-by-state professional directory and promotes research and advocacy [81]. GeneReviews also lists the Children's Craniofacial Association, Human Growth Foundation, MAGIC Foundation, the About Kids Health page (Canada) and UCLA's International Skeletal Dysplasia Registry (ISDR), and the Delphi panel mentions myFace, a US organisation for craniofacial differences [2, 10].

10. Illustrative Clinical Case

10.1. CCD Due to a De Novo p.(Arg225Trp) Variant with Preserved Function Under Conservative Management

In 2026, Xu and colleagues published the case of a Chinese boy followed at West China Second University Hospital, born at term by uncomplicated vaginal delivery and without a family history of CCD [24]. He was initially followed for hypothyroidism diagnosed at approximately 7 months and treated with levothyroxine, with controlled thyroid function throughout follow-up. Between 7 months and 4 years, paediatric and endocrine consultations documented delayed anterior fontanelle closure. At approximately 4 years, persistent fontanelle opening, short stature, delayed bone age and clavicular abnormalities raised suspicion of CCD. Examination showed a persistent fontanelle, mild frontal bossing and mild pectus excavatum; on palpation, both clavicles had firm, cartilage-like midshaft prominences without abnormal mobility, with shortened proximal and distal segments, compatible with bilateral clavicular pseudarthrosis, and shoulder range of motion was preserved without pain or weakness. Neurological examination was normal [24].

Cranial computed tomography and magnetic resonance imaging showed delayed cranial ossification and incompletely closed sutures, with normal brain parenchyma and pituitary gland. Chest computed tomography showed bilateral clavicular hypoplasia with segmental discontinuity and mild sternal depression without cardiopulmonary compression, and lumbosacral radiography was normal. Panoramic radiography showed delayed deciduous tooth exfoliation, permanent tooth eruption and dental development, with uneven development and abnormal positioning of several teeth, without obvious supernumerary teeth. The article's figures (three-dimensional clavicular reconstruction, axial chest computed tomography, panoramic radiograph, Sanger electropherograms of the proband and parents, and chest wall photograph) were consulted only through the captions in the full text [24].

Exome sequencing at approximately 4 years identified the heterozygous variant published as RUNX2 c.631C>T (p.R211W), in the runt domain, absent in both parents on Sanger sequencing and classified as pathogenic under ACMG/AMP criteria [24]. The authors' designation corresponds to transcript NM_004348.3; in the MANE transcript, the variant is NM_001024630.4:c.673C>T p.(Arg225Trp), the arginine 225 hotspot, recorded in ClinVar as pathogenic by multiple submitters without conflicts [7, 32].

Table 10.1 — Growth Before and During Somatropin Treatment (Xu 2026)

SD: standard deviation. Source: [24].

Because short stature and delayed bone age persisted despite adequate thyroid replacement, somatropin was started at 0.05 mg/kg/day at approximately 8 years, following counselling about risks and benefits, without documented adverse effects; at the same age, a left indirect inguinal hernia was repaired laparoscopically [24]. The anterior fontanelle was clinically closed at approximately 10 years. At the last follow-up, at 12 years, shoulder and upper limb function remained preserved, he played basketball and swam, had no fractures, pain, functional limitation or neurological symptoms, and cognitive and school development was normal; dental surveillance continued without any need by that date for orthodontic or oral surgical intervention, and neither clavicular reconstruction nor pectus excavatum correction had been performed [24].

The case illustrates delayed diagnosis when the triad is incomplete (without obvious supernumerary teeth and with the fontanelle initially attributed to other causes), conversion of nomenclature between transcripts and function-guided conservative management of clavicular pseudarthrosis. Somatropin treatment is recommended by neither GeneReviews nor the Delphi panel, and the report itself cannot distinguish its effect from treated hypothyroidism, spontaneous growth and puberty; it should not be interpreted as routine management [2, 10, 24]. The absence of dental intervention at 12 years also differs from the usual course, in which staged tooth exposure begins at 10–12 years [10].

10.2. Comparison Case: Diagnosis in Infancy

Kashiki and colleagues described a Japanese boy born at 39 weeks and 4 days, weighing 3348 g, in whom a 6-cm anterior fontanelle, a 2.5-cm-wide sagittal suture and an enlarged posterior fontanelle were noted at 2 weeks, with initially normal chest and abdominal radiographs [34]. Serum alkaline phosphatase was low (126 U/l), but TNSALP activity in a dried blood spot was normal, excluding hypophosphatasia. At 2 years and 4 months, he had a hypoplastic right clavicle, a narrow chest, cone-shaped epiphyses and short distal phalanges, with normal development (developmental quotient 102). Exome sequencing identified the likely pathogenic de novo variant NM_001024630.4:c.685+2T>G. At 3 years and 9 months, he had neither intellectual impairment nor hearing loss and had 20 erupted deciduous teeth with mild misalignment [34]. This case shows that before the relevant dental age, diagnosis depends on the skull and clavicles, and that clavicular hypoplasia may be unilateral and subtle.

Other reports from 2026 show the reverse pathway, in which the diagnosis is suspected at an adolescent dental consultation: a 16-year-old girl with previously investigated short stature and a normal growth hormone stimulation test, treated with an interim removable partial denture [82]; a 14-year-old Moroccan girl diagnosed in paediatric dentistry [83]; and an Indian family with three affected individuals across two generations, in which pericoronitis of an unerupted third molar was described [70].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

Physical capacity and life expectancy are normal, and cognitive function is usually normal [1, 14]. Prognosis is generally good, with dental, hearing and skeletal dysfunction mostly amenable to correction [26]. Morbidity centres on four domains. Dental disease requires prolonged orthodontic and surgical treatment, often exceeding 10 years, with an impact on oral quality of life [1, 10, 53]. Growth results in a mean adult height of 165 cm in males and 156 cm in females [1, 2, 43]. Bone health includes osteoporosis in 8 of 14 patients in the series by Dinçsoy Bir and colleagues and unexpected fractures described as typical by the Delphi panel [10, 47]. Otorhinolaryngological morbidity includes conductive hearing loss in 39% and recurrent sinusitis and otitis [2]. Orthopaedic complications (scoliosis, genu valgum, shoulder, radial head and hip dislocations) are frequent but rarely disabling [1]. Fontanelles close over time in most patients but may remain open in adulthood [1, 2]. The data do not allow estimation of survival or fracture incidence by age, and the series are small.

In CBFB-associated CCD, final stature is normal and developmental prognosis depends on the presence of delay, which was mild or moderate in affected patients [3, 11].

11.2. Surveillance

Table 11.1 — Recommended Surveillance

DXA interpretation is difficult in short stature and after fractures and should use the height adjustment described by Zemel and colleagues [10, 56]. DXA frequency, growth modulation and orthopaedic assessment remained undefined in the Delphi panel [10]. Surveillance in CBFB-associated CCD is the same, with additional developmental assessment; haematological surveillance is not indicated [4, 10].

12. Summary and Final Considerations

Cleidocranial dysplasia is an autosomal dominant skeletal dysplasia caused in most patients by haploinsufficiency of the transcription factor RUNX2, the master regulator of osteoblast differentiation and tooth formation. The triad of persistently open fontanelles and sutures, hypoplastic or absent clavicles and dental anomalies (retained deciduous teeth, supernumerary teeth and failure of permanent tooth eruption) defines the classic form, but the disorder is a spectrum including mild and exclusively dental forms, with highly variable expressivity within the same family. The frequency measured in Utah was 1.2 per 100,000, and European birth registries give lower values, probably because of underdetection of mild forms.

Diagnosis is clinical and radiological, supported by RUNX2 sequencing with deletion and duplication analysis, which identifies the cause in approximately 85%–95% of patients in GeneReviews series and 60%–70% in other series. A negative result does not exclude the disorder: deep intronic variants, small non-coding deletions and structural CBFB rearrangements escape exome sequencing and array-CGH and may require genome sequencing and RNA studies. CBFB-associated cleidocranial dysplasia reproduces the skeletal and dental phenotype, with normal stature and sometimes developmental delay, and its germline variant should not be confused with the somatic CBFB::MYH11 fusion in acute myeloid leukaemia. Conversion of published designations to MANE transcript NM_001024630.4 avoids interpretation errors, as illustrated by the equivalence between p.R211W and p.(Arg225Trp).

Differential diagnosis is based on the combination of the three domains and the radiological pattern of the pelvis and hands. Pycnodysostosis is distinguished by osteosclerosis, acro-osteolysis and recessive inheritance; hypophosphatasia by premature loss of deciduous teeth and very low alkaline phosphatase with reduced TNSALP activity; and Yunis–Varon, CDAGS and Crane–Heise syndromes and mandibuloacral dysplasia by extraskeletal malformations.

There is no disease-modifying therapy or authorised drug. Management, systematised in 2026 by the first Delphi consensus, which constitutes expert opinion, is multidisciplinary: orthodontic screening by 6 years, staged surgical–orthodontic treatment guided by dental age (Jerusalem approach), audiometry at diagnosis and annually until normal hearing is documented, height-adjusted DXA from the age at which the child can cooperate, a sleep study when symptoms are present, a helmet when warranted by cranial defects, anaesthetic planning for a difficult airway and obstetric surveillance for cephalopelvic disproportion. Bone drugs and growth hormone lack evidence in CCD and can only be used off-label, with the additional caution that bisphosphonates interfere with orthodontic movement. Life expectancy is normal, and morbidity centres on dentition, hearing, stature and bone health. Genetic counselling should address the 50% risk to offspring, clinical and radiological examination of apparently unaffected parents, the possibility of germline mosaicism and prenatal and preimplantation diagnostic options when the familial variant is known.

13. References

1. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. Cleidocranial dysplasia (MIM 119600), p. 767–772.
2. Machol K, Mendoza-Londono R, Lee B. Cleidocranial Dysplasia Spectrum Disorder. 2006 Jan 3 [updated 2023 Apr 13]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1513/. PMID: 20301686.
3. Beyltjens T, Boudin E, Revencu N, Boeckx N, Bertrand M, Schütz L, et al. Heterozygous pathogenic variants involving CBFB cause a new skeletal disorder resembling cleidocranial dysplasia. J Med Genet. 2023;60(5):498-504. doi:10.1136/jmg-2022-108739. PMID: 36241386. PMC10176335.
4. GeneGuidance. CBFB — gene and conditions (version dated 2026-09-26).
5. Orphanet. Metaphyseal dysplasia-maxillary hypoplasia-brachydactyly syndrome (ORPHA:2504). https://www.orpha.net/en/disease/detail/2504.
6. GeneGuidance. RUNX2 — gene and conditions (version dated 2026-09-26).
7. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-8. doi:10.1002/humu.23348. Tool: https://variantvalidator.org (accessed 2026-10-10). PMID: 28967166. PMC5765404.
8. Orphanet. Cleidocranial dysplasia (ORPHA:1452). https://www.orpha.net/en/disease/detail/1452.
9. Stevenson DA, Carey JC, Byrne JL, Srisukhumbowornchai S, Feldkamp ML. Analysis of skeletal dysplasias in the Utah population. Am J Med Genet A. 2012;158A(5):1046-54. doi:10.1002/ajmg.a.35327. PMID: 22461456.
10. Hoover-Fong J, Anderson-Cermin C, Artinian F, Chaushu S, Harrison M, Jan de Beur SM, et al. Diagnosis, treatment, and management recommendations for cleidocranial dysplasia: a modified Delphi panel. J Bone Miner Res. 2026;41(9):972-87. doi:10.1093/jbmr/zjag004. PMID: 41504382. PMC13525201.
11. Pagnamenta AT, Hashim M, Kennedy J, Lawton B, Offiah AC, Taylor JC, et al. A cryptic CBFB deletion-inversion expands the mutational spectrum of variants associated with cleidocranial dysplasia. Clin Genet. 2025;108(1):86-91. doi:10.1111/cge.14709. PMID: 39894570. PMC12136864.
12. European Reference Network on Rare Bone Diseases (ERN BOND). About ERN BOND. https://ernbond.eu/about-ern-bond [accessed 2026-10-10].
13. European Reference Network for rare and/or complex craniofacial anomalies and ear, nose and throat disorders (ERN CRANIO). https://www.ern-cranio.eu/ [accessed 2026-10-10].
14. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Cleidocranial dysplasia, p. 566–569.
15. Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G. Osf2/Cbfa1: a transcriptional activator of osteoblast differentiation. Cell. 1997;89(5):747-54. doi:10.1016/s0092-8674(00)80257-3. PMID: 9182762.
16. Komori T, Yagi H, Nomura S, Yamaguchi A, Sasaki K, Deguchi K, et al. Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell. 1997;89(5):755-64. doi:10.1016/s0092-8674(00)80258-5. PMID: 9182763.
17. Otto F, Thornell AP, Crompton T, Denzel A, Gilmour KC, Rosewell IR, et al. Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell. 1997;89(5):765-71. doi:10.1016/s0092-8674(00)80259-7. PMID: 9182764.
18. Mundlos S, Otto F, Mundlos C, Mulliken JB, Aylsworth AS, Albright S, et al. Mutations involving the transcription factor CBFA1 cause cleidocranial dysplasia. Cell. 1997;89(5):773-9. doi:10.1016/s0092-8674(00)80260-3. PMID: 9182765.
19. Lee B, Thirunavukkarasu K, Zhou L, Pastore L, Baldini A, Hecht J, et al. Missense mutations abolishing DNA binding of the osteoblast-specific transcription factor OSF2/CBFA1 in cleidocranial dysplasia. Nat Genet. 1997;16(3):307-10. doi:10.1038/ng0797-307. PMID: 9207800.
20. Quack I, Vonderstrass B, Stock M, Aylsworth AS, Becker A, Brueton L, et al. Mutation analysis of core binding factor A1 in patients with cleidocranial dysplasia. Am J Hum Genet. 1999;65(5):1268-78. doi:10.1086/302622. PMID: 10521292. PMC1288279.
21. Zhou G, Chen Y, Zhou L, Thirunavukkarasu K, Hecht J, Chitayat D, et al. CBFA1 mutation analysis and functional correlation with phenotypic variability in cleidocranial dysplasia. Hum Mol Genet. 1999;8(12):2311-6. doi:10.1093/hmg/8.12.2311. PMID: 10545612.
22. Ott CE, Leschik G, Trotier F, Brueton L, Brunner HG, Brussel W, et al. Deletions of the RUNX2 gene are present in about 10% of individuals with cleidocranial dysplasia. Hum Mutat. 2010;31(8):E1587-93. doi:10.1002/humu.21298. PMID: 20648631.
23. Thaweesapphithak S, Termteerapornpimol K, Wongsirisuwan S, Chantarangsu S, Porntaveetus T. The impact of RUNX2 gene variants on cleidocranial dysplasia phenotype: a systematic review. J Transl Med. 2024;22(1):1099. doi:10.1186/s12967-024-05904-2. PMID: 39627759. PMC11613773.
24. Xu H, Ma J, Huang X, Pu S, Hou F, Liu W, et al. Cleidocranial dysplasia with preserved function under conservative management: a case report. Front Pediatr. 2026;14:1844281. doi:10.3389/fped.2026.1844281. PMID: 42643437. PMC13503619.
25. Zheng Q, Sebald E, Zhou G, Chen Y, Wilcox W, Lee B, et al. Dysregulation of chondrogenesis in human cleidocranial dysplasia. Am J Hum Genet. 2005;77(2):305-12. doi:10.1086/432261. PMID: 15952089.
26. Timbó SS, Oliveira CM, Ribeiro EM, Carvalho EDF. Displasia cleidocraniana. In: Ribeiro EM, editor. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 261–264.
27. Ogawa S, Higuchi S, Yoshimoto Y, Hoshino M, Miura S, Hamada A, et al. Functional impact of pathogenic mutations in the runt homology domain of mouse Runx2 on skeletal and dental phenotypes in cleidocranial dysplasia. J Bone Miner Res. 2026;41(8):874-89. doi:10.1093/jbmr/zjaf201. PMID: 41454777.
28. Stojanovic D, Garczarczyk-Asim D, Vodopiutz J, Janecke AR. Deep-intronic variant in RUNX2 causing pseudo-exon inclusion in a family with cleidocranial dysplasia. Clin Genet. 2026;110(2):268-9. doi:10.1111/cge.70158. PMID: 41797213. PMC13327211.
29. Mijovic M, Cuturilo G, Ruml Stojanovic J, Brankovic M, Miletic A, Bosankic B, et al. A novel 4.2 kb deletion of the 3'UTR of RUNX2 gene causes cleidocranial dysplasia: further delineation of the role of the 3'UTR. Balkan J Med Genet. 2025;28(2):107-12. doi:10.2478/bjmg-2025-00026. PMID: 42148000. PMC13175645.
30. Farrow E, Nicot R, Wiss A, Laborde A, Ferri J. Cleidocranial dysplasia: a review of clinical, radiological, genetic implications and a guidelines proposal. J Craniofac Surg. 2018;29(2):382-9. doi:10.1097/SCS.0000000000004200. PMID: 29189406.
31. Berkay EG, Elkanova L, Kalaycı T, Uludağ Alkaya D, Altunoğlu U, Cefle K, et al. Skeletal and molecular findings in 51 cleidocranial dysplasia patients from Turkey. Am J Med Genet A. 2021;185(8):2488-95. doi:10.1002/ajmg.a.62261. PMID: 33987976.
32. National Center for Biotechnology Information. ClinVar: VCV000009303 (NM_001024630.4(RUNX2):c.673C>T p.Arg225Trp) and VCV000009302 (c.674G>A p.Arg225Gln). https://www.ncbi.nlm.nih.gov/clinvar/variation/9303/ ; https://www.ncbi.nlm.nih.gov/clinvar/variation/9302/ [accessed 2026-10-10].
33. Çamtosun E, Akıncı A, Demiral E, Tekedereli İ, Sığırcı A. A case of cleidocranial dysplasia with a novel mutation and growth velocity gain with growth hormone treatment. J Clin Res Pediatr Endocrinol. 2019;11(3):301-5. doi:10.4274/jcrpe.galenos.2018.2018.0211. PMID: 30468148. PMC6745455.
34. Kashiki H, Kido J, Misumi Y, Sugawara K, Tsuchida N, Matsumoto N, et al. Cleidocranial dysplasia caused by a novel de novo RUNX2 splice-site variant. Hum Genome Var. 2026;13(1):19. doi:10.1038/s41439-026-00353-3. PMID: 42243079. PMC13478401.
35. Yoshida T, Kanegane H, Osato M, Yanagida M, Miyawaki T, Ito Y, et al. Functional analysis of RUNX2 mutations in Japanese patients with cleidocranial dysplasia demonstrates novel genotype-phenotype correlations. Am J Hum Genet. 2002;71(4):724-38. doi:10.1086/342717. PMID: 12196916.
36. Chantarawaratit PO, Thaweesapphithak S, Pittayapat P, Porntaveetus T. Correlation of RUNX2 variants with craniofacial-dental phenotypes in cleidocranial dysplasia. Clin Exp Dent Res. 2026;12(2):e70351. doi:10.1002/cre2.70351. PMID: 41995300. PMC13088869.
37. Pal T, Napierala D, Becker TA, Loscalzo M, Baldridge D, Lee B, et al. The presence of germ line mosaicism in cleidocranial dysplasia. Clin Genet. 2007;71(6):589-91. doi:10.1111/j.1399-0004.2007.00812.x. PMID: 17539909.
38. Muurinen M, Taylan F, Tournis S, Eisfeldt J, Balanika A, Vastardis H, et al. Mosaic Deletions of Known Genes Explain Skeletal Dysplasias With High and Low Bone Mass. JBMR Plus. 2022;6(8):e10660. doi:10.1002/jbm4.10660. PMID: 35991531.
39. Kundu M, Javed A, Jeon JP, Horner A, Shum L, Eckhaus M, et al. Cbfbeta interacts with Runx2 and has a critical role in bone development. Nat Genet. 2002;32(4):639-44. doi:10.1038/ng1050. PMID: 12434156.
40. Goto T, Aramaki M, Yoshihashi H, Nishimura G, Hasegawa Y, Takahashi T, et al. Large fontanelles are a shared feature of haploinsufficiency of RUNX2 and its co-activator CBFB. Congenit Anom (Kyoto). 2004;44(4):225-9. doi:10.1111/j.1741-4520.2004.00043.x. PMID: 15566413.
41. Khan A, Hyde RK, Dutra A, Mohide P, Liu P. Core binding factor beta (CBFB) haploinsufficiency due to an interstitial deletion at 16q21q22 resulting in delayed cranial ossification, cleft palate, congenital heart anomalies, and feeding difficulties but favorable outcome. Am J Med Genet A. 2006;140(21):2349-54. doi:10.1002/ajmg.a.31479. PMID: 17022082.
42. Döhner H, Wei AH, Appelbaum FR, Craddock C, DiNardo CD, Dombret H, et al. Diagnosis and management of AML in adults: 2022 recommendations from an international expert panel on behalf of the ELN. Blood. 2022;140(12):1345-1377. doi:10.1182/blood.2022016867. PMID: 35797463.
43. Cooper SC, Flaitz CM, Johnston DA, Lee B, Hecht JT. A natural history of cleidocranial dysplasia. Am J Med Genet. 2001;104(1):1-6. doi:10.1002/ajmg.10024. PMID: 11746020.
44. Roberts T, Stephen L, Beighton P. Cleidocranial dysplasia: a review of the dental, historical, and practical implications with an overview of the South African experience. Oral Surg Oral Med Oral Pathol Oral Radiol. 2013;115(1):46-55. doi:10.1016/j.oooo.2012.07.435. PMID: 23102800.
45. Hung CY, Matsumoto H, Tsuji M, Moriyama K. Computed tomography analysis of craniofacial features in Japanese patients with cleidocranial dysplasia. Congenit Anom (Kyoto). 2026;66(1):e70044. doi:10.1002/cga.70044. PMID: 41645457. PMC13422770.
46. Golan I, Baumert U, Hrala BP, Müssig D. Dentomaxillofacial variability of cleidocranial dysplasia: clinicoradiological presentation and systematic review. Dentomaxillofac Radiol. 2003;32(6):347-54. doi:10.1259/dmfr/63490079. PMID: 15070835.
47. Dinçsoy Bir F, Dinçkan N, Güven Y, Baş F, Altunoğlu U, Kuvvetli SS, et al. Cleidocranial dysplasia: Clinical, endocrinologic and molecular findings in 15 patients from 11 families. Eur J Med Genet. 2017;60(3):163-168. doi:10.1016/j.ejmg.2016.12.007. PMID: 28027977.
48. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Large fontanelle, p. 204–205.
49. Chen JL, Yu LN, Ran QC, Wu QY, Cai YY, Li C, et al. Impacted second premolars in cleidocranial dysplasia: three-dimensional position and morphology characteristics and factors affecting the success rate of closed eruption. Am J Orthod Dentofacial Orthop. 2026;169(2):177-88. doi:10.1016/j.ajodo.2025.09.016. PMID: 41148103.
50. Bous R, Ahmed M, Lyamichev A, Elnaghy R, Valiathan M. Anatomic assessment of palatal temporary skeletal anchorage devices insertion sites among patients with cleidocranial dysplasia vs controls: a retrospective cone-beam computed tomography analysis. Am J Orthod Dentofacial Orthop. 2026;169(1):67-74. doi:10.1016/j.ajodo.2025.09.007. PMID: 41201411.
51. Unger S, Mornet E, Mundlos S, Blaser S, Cole DE. Severe cleidocranial dysplasia can mimic hypophosphatasia. Eur J Pediatr. 2002;161(11):623-6. doi:10.1007/s00431-002-0978-9. PMID: 12424591.
52. Visosky AM, Johnson J, Bingea B, Gurney T, Lalwani AK. Otolaryngological manifestations of cleidocranial dysplasia, concentrating on audiological findings. Laryngoscope. 2003;113(9):1508-14. doi:10.1097/00005537-200309000-00017. PMID: 12972925.
53. Zinn S, Schneider M, Mitchell S, Hoover-Fong J, Doughan M, Jolly LR, et al. Oral health-related quality of life in patients with cleidocranial dysplasia: impact of malocclusion traits and treatment modality. Am J Orthod Dentofacial Orthop. 2026;169(3):379-87. doi:10.1016/j.ajodo.2025.10.013. PMID: 41410626.
54. Stewart PA, Wallerstein R, Moran E, Lee MJ. Early prenatal ultrasound diagnosis of cleidocranial dysplasia. Ultrasound Obstet Gynecol. 2000;15(2):154-6. doi:10.1046/j.1469-0705.2000.00041.x. PMID: 10776001.
55. Hermann NV, Hove HD, Jørgensen C, Larsen P, Darvann TA, Kreiborg S, et al. Prenatal 3D ultrasound diagnostics in cleidocranial dysplasia. Fetal Diagn Ther. 2009;25(1):36-9. doi:10.1159/000195634. PMID: 19169035.
56. Zemel BS, Leonard MB, Kelly A, Lappe JM, Gilsanz V, Oberfield S, et al. Height adjustment in assessing dual energy x-ray absorptiometry measurements of bone mass and density in children. J Clin Endocrinol Metab. 2010;95(3):1265-73. doi:10.1210/jc.2009-2057. PMID: 20103654. PMC2841534.
57. Yamada M, Cleghorn M, Krishnamurthy P, French T, Francis D, Kosaki K, et al. Two unrelated families with noncoding duplications upstream of MSX2 refine the critical regulatory region likely involved in cranial bone development and a cleidocranial dysplasia-like phenotype. Hum Mutat. 2026;2026:1258291. doi:10.1155/humu/1258291. PMID: 42609732. PMC13478909.
58. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. Pyknodysostosis (MIM 265800), p. 601–603.
59. LeBlanc S, Savarirayan R. Pycnodysostosis. 2020 Nov 5 [revised 2023 Apr 6]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–. Available from: https://www.ncbi.nlm.nih.gov/books/NBK563694/.
60. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Pyknodysostosis, p. 564–565.
61. Orphanet. Pycnodysostosis (ORPHA:763). https://www.orpha.net/en/disease/detail/763.
62. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. Yunis-Varon syndrome (MIM 216340), p. 773–775.
63. Xing C, Kanchwala M, Rios JJ, Hyatt T, Wang RC, Tran A, et al. Biallelic variants in RNU12 cause CDAGS syndrome. Hum Mutat. 2021;42(8):1042-1052. doi:10.1002/humu.24239. PMID: 34085356.
64. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. CDAGS (MIM 603116), p. 776–778.
65. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. Mandibuloacral dysplasia (MIM 248370, 608612), p. 695–698.
66. Ott CE, Hein H, Lohan S, Hoogeboom J, Foulds N, Grünhagen J, et al. Microduplications upstream of MSX2 are associated with a phenocopy of cleidocranial dysplasia. J Med Genet. 2012;49(7):437-41. doi:10.1136/jmedgenet-2012-100825. PMID: 22717651.
67. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. Spondylo-megaepiphyseal-metaphyseal dysplasia (MIM 613330), p. 765–766.
68. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. SAMS syndrome (MIM 602471), p. 793–794.
69. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. Ischio-pubic-patellar dysplasia (MIM 147891), p. 782–784.
70. Wadewale MB, Sachdev SS, Ranganath S. Two-generation cleidocranial dysplasia involving three individuals of a family: a rare case report. Contemp Clin Dent. 2026;17(2):142-6. doi:10.4103/ccd.ccd_4_25. PMID: 42428297. PMC13349056.
71. Bastos Filho FAG, Ribeiro EM, Andrade MDFC. Picnodisostose. In: Ribeiro EM, editor. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 249–250.
72. Savarirayan R, Rossiter JP, Hoover-Fong JE, Irving M, Bompadre V, Goldberg MJ, et al. Best practice guidelines regarding prenatal evaluation and delivery of patients with skeletal dysplasia. Am J Obstet Gynecol. 2018;219(6):545-62. doi:10.1016/j.ajog.2018.07.017. PMID: 30048634.
73. Becker A, Lustmann J, Shteyer A. Cleidocranial dysplasia: Part 1—General principles of the orthodontic and surgical treatment modality. Am J Orthod Dentofacial Orthop. 1997;111(1):28-33. doi:10.1016/s0889-5406(97)70298-1. PMID: 9009920.
74. Becker A, Shteyer A, Bimstein E, Lustmann J. Cleidocranial dysplasia: Part 2—Treatment protocol for the orthodontic and surgical modality. Am J Orthod Dentofacial Orthop. 1997;111(2):173-83. doi:10.1016/s0889-5406(97)70213-0. PMID: 9057617.
75. Zhu Y, Zou Y, Yu Q, Sun H, Mou S, Xu S, et al. Combined surgical-orthodontic treatment of patients with cleidocranial dysplasia: case report and review of the literature. Orphanet J Rare Dis. 2018;13(1):217. doi:10.1186/s13023-018-0959-3. PMID: 30514338. PMC6280340.
76. Sobti R, Upadhyay S, Paige A, Panchura J, Lin Y, Kapila S. Optimized approach for retrieval of multiple impacted teeth in cleidocranial dysplasia. Cleft Palate Craniofac J. 2026. doi:10.1177/10556656261447538. PMID: 42138615.
77. Kang N, Kim SZ, Jung SN. Correction of depressed forehead with BoneSource in cleidocranial dysplasia. J Craniofac Surg. 2009;20(2):564-6. doi:10.1097/SCS.0b013e31819ba361. PMID: 19305258.
78. Sewell MD, Higgs DS, Lambert SM. Clavicle lengthening by distraction osteogenesis for congenital clavicular hypoplasia: case series and description of technique. J Pediatr Orthop. 2013;33(3):314-20. doi:10.1097/BPO.0b013e3182784e79. PMID: 23482270.
79. Johns Hopkins University. Cleidocranial Dysplasia (CCD): From Genotype to Phenotype and Considerations for Care. ClinicalTrials.gov ID NCT05368064. https://clinicaltrials.gov/study/NCT05368064 [accessed 2026-10-10].
80. Ioscovich A, Barth D, Samueloff A, Grisaru-Granovsky S, Halpern S. Anesthetic management of a patient with cleidocranial dysplasia undergoing various obstetric procedures. Int J Obstet Anesth. 2010;19(1):106-8. doi:10.1016/j.ijoa.2009.07.002. PMID: 19945847.
81. CCD Smiles Foundation. CCD Smiles. https://ccdsmiles.org/ [accessed 2026-10-10].
82. Çitaku E, Çitaku E, Reshitaj A. Cleidocranial dysplasia with complex oral manifestations: a case report. BMC Oral Health. 2026;26(1):1871. doi:10.1186/s12903-026-08697-6. PMID: 42410568. PMC13625300.
83. Mansy H, Oumensour K, Ikiss JA, Elarabi S. Cleidocranial dysplasia in a Moroccan patient: a case report. Pan Afr Med J. 2026;53:51. doi:10.11604/pamj.2026.53.51.48261. PMID: 42180632. PMC13197682.