Weaver Syndrome and PRC2-Related Overgrowth (EZH2, EED, SUZ12)

Literature and regulatory information updated: 29 October 2026.

An overgrowth–intellectual disability spectrum caused by germline variants in the genes encoding the three core subunits of Polycomb repressive complex 2 (PRC2): Weaver syndrome or EZH2-related overgrowth, Cohen-Gibson syndrome or EED-related overgrowth, and Imagawa-Matsumoto syndrome or SUZ12-related overgrowth. Sotos syndrome (NSD1) and the other overgrowth syndromes form part of the differential diagnosis.

0. Reference Summary Table

1. Introduction

1.1. Original description and clinical delineation

In 1974, David Weaver, with Graham, Thomas and David Smith, described two unrelated boys with overgrowth of prenatal onset, accelerated skeletal maturation, an unusual facies and camptodactyly, and proposed that this association constituted a new entity [1, 29]. Photographs of the two boys, at 18 and 11 months, and radiographs showing accelerated skeletal maturation and distal widening of the femora were reproduced in Smith's atlas [30]. Over the following decades, more than 50 patients with similar features were published under the eponym Weaver syndrome, also called Weaver-Smith syndrome [2, 23]. The subtlety of the phenotype and its overlap with Sotos syndrome made clinical diagnosis difficult, even for experienced dysmorphologists, and the robustness of many clinical associations published during this period was uncertain [23].

1.2. Molecular discovery

In 2003, the description of NSD1 variants in some patients with a clinical diagnosis of Weaver syndrome suggested that the two syndromes were allelic [31]. This association was later refuted, and NSD1 is now regarded as the gene for Sotos syndrome and not for Weaver syndrome [1, 30, 32]. In December 2011, two independent groups identified germline EZH2 variants as the cause of Weaver syndrome. Gibson and colleagues, in the FORGE Canada consortium, performed trio exome sequencing in two families, one of them one of the original 1974 families, and found two distinct de novo EZH2 variants; a third de novo variant was found in a third patient with the classic phenotype [33]. Tatton-Brown and colleagues, in the UK Childhood Overgrowth study, described 19 individuals with EZH2 variants and linked the gene not only to the syndrome but also to the variation in human height [23, 34]. In 2013, the same team extended the series to 48 individuals from 39 families and defined the mutational spectrum and the phenotype associated with the gene [23].

The discovery extended to the other subunits of the complex. In 2015, Cohen and colleagues described the first EED variant associated with overgrowth, in a 27-year-old Turkish man with the de novo variant NM_003797.5:c.906A>C p.(Arg302Ser) [9, 18, 35], and the entity received the eponym Cohen-Gibson syndrome [4]. In 2017, Imagawa and colleagues performed exome sequencing in eight Japanese patients with suspected Weaver syndrome and found a 25.4 kb deletion encompassing part of EZH2, an EED missense variant and the first germline SUZ12 variant described in humans, inherited from a father who carried the variant in mosaic form; functional assays showed reduced H3K27 trimethylation [14]. Two new SUZ12 variants, described in 2018 in a 19-year-old Brazilian man and a 9-year-old French girl by teams from Japan, Brazil (São Paulo and Curitiba) and France [20, 36], and a series of ten patients from nine families published in 2019 consolidated the third entity [13, 36], which Matsumoto's team reviewed in 2023 under the name Imagawa-Matsumoto syndrome [7]. In 2020, Choufani and colleagues described an EZH2-specific DNA methylation signature that also classifies EED and SUZ12 variants [25].

1.3. Nomenclature, classification and scope

GeneReviews uses the term "EZH2-related overgrowth" for the entire phenotypic spectrum of heterozygous variants in the gene, ranging from classic Weaver syndrome to tall stature without other features [1]. Orphanet groups the three entities within the "PRC2 complex-related overgrowth spectrum" (ORPHA:659387) and has merged the former entry "Weaver-like syndrome" into that group [8]. Weaver-Williams syndrome (ORPHA:3448), described in two siblings in 1977 with microcephaly, skeletal hypoplasia and delayed skeletal maturation, shares only the eponym and does not belong to this group [37]. In the Oxford Desk Reference, Weaver syndrome appears in a table of causes of overgrowth as an EZH2 and EED entity, alongside the other disorders of epigenetic regulators (NSD1, DNMT3A) and separate from the disorders of the PI3K/mTOR pathway and the imprinting defects [38].

The three PRC2 syndromes share the mechanism, much of the phenotype and the diagnostic methods; Weaver syndrome is the most frequent and the most studied. For the purposes of this text, "overgrowth" denotes regional or global growth more than 2 standard deviations (SD) above the mean for age, approximately the 98th centile [38], and "overgrowth–intellectual disability" (OGID) denotes the association of height and/or head circumference (HC) ≥ +2 SD with intellectual disability (ID) [24, 39].

Table sources: [1, 3–6, 13, 24, 26].

2. Molecular Genetics and Pathophysiology

2.1. Polycomb repressive complex 2

PRC2 is a chromatin-modifying complex that maintains the transcriptional silencing of developmental genes. Its core consists of EZH2 (or its paralogue EZH1), EED, SUZ12 and RBBP4 or RBBP7, and its catalytic activity consists of adding up to three methyl groups to lysine 27 of histone H3 (H3K27me1, H3K27me2 and H3K27me3) [4, 40, 41]. H3K27 trimethylation is associated with local transcriptional repression; its correct propagation to daughter cells after mitosis is required for repressed domains to persist in newly synthesised chromatin [4, 42]. Silenced target genes include HOXC8, HOXA9, MYT1 and CDKN2A, and the H3K27me3 mark is required for embryonic stem cell identity and orderly differentiation [10]. The same complex can recruit DNA methyltransferases, linking two repressive systems [10]. EZH2 is also inhibited by AKT, a component of the PI3K/mTOR pathway, through phosphorylation of serine 21, which prevents binding to histone H3; this is one of the molecular bridges between the two major families of overgrowth syndromes, those of epigenetic regulators and those of the PI3K/mTOR pathway [23, 38].

2.2. The three genes

EZH2 (enhancer of zeste 2 polycomb repressive complex 2 subunit), at 7q36.1, encodes the catalytic subunit. The protein has a SET domain (su(var)3-9, enhancer of zeste, trithorax), which catalyses methylation, a cysteine-rich pre-SET domain (CXC) and two SANT domains [1]. EZH2 also methylates non-histone substrates such as GATA4 and RORA [10]. The Oxford Desk Reference places the gene at 7q35, which differs from OMIM, GeneReviews and GeneGuidance (7q36.1); the French protocol writes "7p36.1", which is also a transcription error [1, 11, 38].

EED (embryonic ectoderm development), at 11q14.2, encodes a ubiquitous, conserved protein with seven WD40 repeats forming a β-propeller. The C-terminal domain binds specifically to histone tails carrying H3K27me3 and allosterically stimulates EZH2. This is the "reader" function of the complex, complementary to the "writer" function of EZH2, and allows the mark to be propagated to daughter cells [4, 10, 42].

SUZ12 (SUZ12 polycomb repressive complex 2 subunit), at 17q11.2, encodes a structural subunit required for the assembly and activity of the complex [10]. It includes regions that interact with JARID2/EPOP and AEBP2/PHF19, a C2H2 zinc finger and the VEFS domain, which interacts with EZH2 [21, 22]. The gene lies about 560 kb from NF1 and is included in the common 1.4 Mb type 1 NF1 microdeletion; type 2 deletions have a breakpoint within SUZ12 [7, 10].

2.3. Disease mechanism

In all three syndromes, PRC2 activity is reduced. The EZH2 variants associated with Weaver syndrome reduce the methyltransferase activity of PRC2 complexes reconstituted in vitro [43]. The variant NM_004456.5:c.1876G>A p.(Val626Met), expressed in growth plate chondrocytes, has lower enzymatic activity; in mice carrying the same variant, homozygotes die perinatally and heterozygotes are viable, fertile and show mild overgrowth, with reduced H3K27me in both genotypes [44]. The recurrent variant p.Arg684Cys, studied in mouse stem cells with a humanised region, is hypomorphic rather than null [45]. These data support partial loss of function as the mechanism [1, 41].

The mutational spectrum points in the same direction. More than 90% of variants are missense, and the few truncating variants lie in the last exon (exon 20), after the SET domain, so they probably escape nonsense-mediated mRNA decay (NMD) and produce a truncated protein that retains the SET domain [1, 23, 38]. The absence of early truncating variants suggests that null alleles do not cause the syndrome [44]. Few individuals with large deletions including EZH2 have been described: a 25.4 kb deletion involving part of EZH2 and CUL1 in a Japanese patient with a Weaver phenotype, and a 1.2 Mb deletion at 7q36.1 in a child with tall stature and ID [14, 46]. Conversely, five earlier cases with deletions including EZH2 had no reported tall stature [23]. GeneReviews therefore considers simple haploinsufficiency unlikely and regards it as uncertain whether deletions produce the same phenotype [1]. The French protocol nevertheless includes heterozygous deletion of the gene among the causes of the syndrome [11].

In 2025, Deevy and colleagues proposed a dominant-negative mechanism. Structural analysis of all known variants and isogenic modelling of ten representative variants in embryonic stem cells showed global reduction of H3K27me2 and H3K27me3, increased H3K27ac and chromatin decompaction. The reduction of H3K27me2/3 suggested dominant-negative interference with PRC2 activity, even when the variants were expressed at low levels. Weakly Polycomb-bound genes lost occupancy by canonical PRC1 and were derepressed, including growth-control genes [47]. The two models are not mutually exclusive: the variants reduce the activity of the complex, and some appear to do so by interfering with the normal allele. GeneReviews also acknowledges that other mechanisms are theoretically possible [1, 41].

Gain-of-function EZH2 variants produce the opposite effect: the methylation signature and chromatin studies associate them with growth restriction, with transcriptional changes reciprocal to those of Weaver variants [1, 25, 47]. The EED reader domain binds H3K27me3 through an aromatic "cage" formed by Phe97, Trp364 and Tyr365 [18]; the variant NM_003797.5:c.1091G>T p.(Trp364Leu), described in mosaic form in 2026, affects one of these residues [12]. For EED, a lymphoblastoid cell line from a patient with p.(Arg236Thr) showed reduced H3K27me3, consistent with partial loss of function [4, 14]. A Drosophila assay, with the substitutions introduced into the orthologue esc, distinguished probably benign variants, which behaved like the wild type, from pathogenic variants, which behaved as loss of function [48]. For SUZ12, most variants are truncating, and peripheral blood mononuclear cells from a child and mother carrying c.1244_1248del showed global reduction of H3K27me3 [13, 22]. GeneGuidance, however, reports a ClinGen haploinsufficiency score of 0 for SUZ12, that is, no evidence evaluated [10].

At tissue level, loss of repression of growth and skeletal maturation programmes explains overgrowth with advanced bone age [10]. The overlap between the 14 OGID genes and the 611 genes in height-associated regions from genome-wide association studies (p = 6.84 × 10⁻⁸), and between these genes and 260 cancer driver genes (p = 1.75 × 10⁻¹⁴), illustrates the link between growth control and tumorigenesis. The germline and somatic mutational spectra, however, differ [39].

2.4. Variant spectrum and nomenclature

GeneReviews uses the reference sequences NM_004456.4 (EZH2) and NM_003797.4 (EED). The current MANE Select transcripts are NM_004456.5 (NP_004447.2), NM_003797.5 (NP_003788.2) and, for SUZ12, NM_015355.4 (NP_056170.2). The variants in the following table were validated in VariantValidator against the MANE transcripts and retain the published c. numbering [1, 4, 9].

Source of the genomic conversions and nomenclature confirmation: [9].

The variant p.Arg684Cys is described in a 2025 preclinical study as "c.2035C>T". This notation does not match the NM_004456.5 reference sequence, in which the same protein change is c.2050C>T; the description should be read as belonging to another reference [9, 45]. The SUZ12 variant c.1783_1786del p.(Lys595Profs*18), located in the VEFS domain, was reported on NM_015355.4 by the authors [21].

The first three variants described by Gibson and colleagues were NM_004456.5:c.457_459del p.(Tyr153del), c.2080C>T p.(His694Tyr) and c.394C>T p.(Pro132Ser), all de novo; residue His694 and neighbouring residues are somatically mutated in myeloid neoplasms [16, 33]. In the 2013 UK series (48 variants), 44 were missense and 4 truncating; missense variants were distributed throughout the gene, with clustering in the SET domain (12/44), and NM_004456.5:c.2050C>T p.(Arg684Cys) and c.1955C>G p.(Ser652Cys) were recurrent [23]. The compilation of 56 patients published in 2024 included 49 missense, 1 frameshift, 1 nonsense, 1 in-frame duplication, 1 in-frame deletion, 2 splicing variants and 1 large deletion [51]. The French protocol counts fewer than 100 published pathogenic or likely pathogenic variants, with enrichment in the SET domain [11]. On the basis of few cases, residue conservation and domain function, missense variants in the SET domain are considered likely pathogenic [1]. Among the 19 SUZ12 variants reviewed in 2026, frameshift, nonsense and splicing variants predominate, together with some missense variants and 17q11.2 deletions including the gene [13, 21, 26]. In ClinVar, at the time of GeneGuidance, EZH2 had 77 pathogenic, 35 likely pathogenic and 267 uncertain significance submissions. These numbers reflect testing volume and not disease frequency [10].

2.5. Genotype–phenotype correlations, penetrance and expressivity

No genotype–phenotype correlation has been established for EZH2. The entire clinical spectrum has been observed with truncating or missense variants, inside or outside the SET domain [1, 11, 23]. The enzymatic activities of different variants do not correlate directly with clinical severity [43]. For EED (19 probands) and SUZ12 the data are insufficient [4, 7]. A 2025 case study linked the severity of a phenotype with brain malformations to a SANT-domain variant, but this interpretation rests on a single patient [50].

Penetrance is estimated differently depending on the source. GeneReviews considers the data insufficient and accepts reduced penetrance for some variants, given the subtlety of certain phenotypes; in the 2013 series, a variant NM_004456.5:c.1852C>T p.(His618Tyr) was inherited from a father without evident overgrowth [1, 23]. The French protocol considers penetrance close to 100%, with variable expressivity and tall stature as an almost constant sign [11]. For EED, expected penetrance is high, because most variants are de novo, but there are no estimates for inherited variants [4]. For SUZ12, there are carriers with a mild phenotype, minimally affected parents and one carrier without macrocephaly or overgrowth [7, 13]. Mosaicism is documented in the unaffected father of an 11-year-old Japanese girl with a SUZ12 variant (present at low level in the father) and in a girl with post-zygotic EED mosaicism, with a variant allele fraction of 31% in blood, 11–17% in buccal mucosa and 30% in saliva [12, 14].

ClinGen classifies the relationship between EZH2 and Weaver syndrome as definitive, with autosomal dominant inheritance (2022 evaluation) [54]. For SUZ12, GenCC records definitive evidence (Illumina, G2P), strong evidence (Labcorp/Invitae, PanelApp Australia) and moderate evidence (Ambry), without a ClinGen validity curation [10].

2.6. Allelic disorders and somatic alterations

No germline EZH2 phenotypes other than the overgrowth spectrum are known, except the gain-of-function variants associated with growth restriction [1, 25]. Somatic alterations are different, non-hereditary events. In EZH2, the recurrent activating Tyr646 variant occurs in diffuse large B-cell and follicular lymphomas, and inactivating variants, mono- or biallelic, occur in myelodysplastic syndromes and myeloproliferative neoplasms with a poor prognosis, including chronic myelomonocytic leukaemia [2, 23, 55]. The variants NP_004447.2:p.(Gly159Arg) (published protein designation), NM_004456.5:c.2050C>T p.(Arg684Cys) and c.2199C>G p.(Tyr733*) have been described as somatic in myeloid neoplasms and as germline in eight individuals in the UK series, none of whom had developed a neoplasm. Myeloid neoplasms occur mainly after the age of 60 years, and the series consisted mostly of children [23]. Loss-of-function variants and deletions of EZH2 and SUZ12 have been described in 25% of T-cell acute lymphoblastic leukaemias, and somatic 7q36.1 deletions including EZH2 occur in myelodysplastic syndromes [16, 20]. EED variants have also been identified in large autism studies, without information on growth [4]. Somatic loss of function of EED or SUZ12 occurs in 92% of sporadic malignant peripheral nerve sheath tumours (MPNST), 70% of those associated with neurofibromatosis type 1 (NF1) and 90% of radiation-induced ones, with complete loss of H3K27me3 [10, 56]. The JAZF1-SUZ12 fusion is recurrent in endometrial stromal sarcoma [10]. In patients with the NF1 microdeletion, co-deletion of SUZ12 modifies the phenotype and is thought to contribute to the higher lifetime risk of MPNST, estimated at 16–26%, versus 8–13% in carriers of intragenic NF1 variants [7, 10].

3. Clinical Spectrum and Manifestations by System

3.1. Data sources and limitations of the series

The clinical frequencies of Weaver syndrome come from series of patients identified through clinical suspicion and from published cases, which biases the spectrum towards typical forms and underestimates mild ones [1, 10]. The 2013 UK series (Childhood Overgrowth Genetics study) brought together 48 individuals from 39 families, recruited from genetics centres in several countries: 25 from the United Kingdom, 10 from the United States, 3 from France, 2 from the Netherlands, 2 from Portugal, 2 from Réunion and 1 from each of Austria, Belgium, Ireland and Japan [23]. The 2024–2025 GeneReviews compiles at least 70 individuals [1]. A 2024 Turkish compilation brought together 56 [51], and the 2025 French protocol reports 72 published patients and calculates tumour frequencies on the basis of 86 carriers of pathogenic or likely pathogenic variants [11]. The following table summarises the reference frequencies for EZH2.

Sources: [1, 11, 23]. The GeneReviews ID percentages refer to individuals with ID; in the text of the same chapter, among 61 individuals with data, 8 had normal intelligence and 52 had ID (30 mild, 13 moderate, 3 severe, 6 unclassified), a sum that does not add up to 61 [1]. The frequencies of umbilical hernia in the French protocol vary between sections (52%, 50% and "about one third") [11].

3.2. Growth

Overgrowth usually begins before birth. In 23 newborns, mean length was +2.2 SD (−0.5 to +4.9), and in 45 mean weight was +1.7 SD (−1.6 to +4.6) [1]. In the UK series, only 38% (15/39) of newborns had a weight above +2 SD, with a median of +1.3 SD, whereas 12 of 18 had a length above +2 SD; parents described the babies as "long and thin", and length, less often measured and less often remembered than weight, discriminates better [23]. About 53% have excessive length at birth across the literature [11]. Smith's atlas states that, in early childhood, weight increases more than height and that, over time, skeletal growth predominates; in a minority, overgrowth becomes apparent only a few months after birth [30]. The atlas of skeletal dysplasias reports birth lengths of about 56 cm in boys and 53 cm in girls [2].

Tall stature is the most constant sign: 59 of 65 individuals aged 1 to 70 years had height ≥ +2 SD, with a postnatal mean of +3.5 SD [1]. In the UK series, 14 children and 2 adults had height ≥ +4 SD. Of the four individuals below +2 SD, three were adults who had been tall as children: one developed severe scoliosis at 18 months, operated on in adolescence, and another was at +5.7 SD up to age 3 years and then decelerated for no known reason [23]. The final heights cited in the atlases are 194.2 cm in men and 176.3 cm in women (above +3.5 SD), with HC of 61 cm and 59.5 cm [2, 30]. Among the adults described in OMIM, a woman followed for more than 20 years measured 187 cm at 25 years, and an affected father measured 193 cm, with an HC of 62.8 cm (+5 SD). The first 1974 patient was reassessed at 30 years, with mild ID, poor coordination, macrocephaly and scoliosis of 10 to 20 degrees [15]. The French protocol describes harmonious growth, without body disproportion and without acceleration of growth velocity, unlike endocrine causes of tall stature such as acromegaly or hyperthyroidism. Height lies on average about +1.5 SD above the channel expected from parental height, endocrine investigations are normal and puberty appears to occur at a normal age despite the advanced bone age [11].

HC is large but proportionate to height [30]. Among 59 individuals, 32 had macrocephaly (> +2 SD), up to +5.5 SD, and 27 did not. HC was ≥ +2 SD at birth in 5 of 6 newborns with data, suggesting macrocephaly present from birth [1]. In the UK series, median HC was +1.8 SD. Comparison with Sotos syndrome illustrates the difference: in Weaver syndrome, about 50% have increased height only and 5% increased HC only, whereas in Sotos syndrome increased HC is more characteristic [23].

3.3. Facies

The facies is characteristic but subtle and changes with age. Before age 3 years, retrognathia, large fleshy ears and a "stuck-on chin" appearance predominate, with a horizontal skin crease and sometimes a central dimple. At all ages there is a broad forehead (increased bifrontal diameter), round face, widely spaced eyes, almond-shaped palpebral fissures and a long or prominent philtrum [1]. Smith's atlas adds a flat occiput, epicanthus, depressed nasal bridge, downslanting palpebral fissures, strabismus and a thin upper lip [30]. The retrognathia of young children resolves with age [1]. The Oxford Desk Reference places the age of easiest recognition between 1 and 3 years [38].

In the UK series, a panel of five dysmorphologists classified photographs of 43 individuals: 20 had classic Weaver syndrome, 11 possible Weaver syndrome and 12 a non-typical facies. The youngest were concentrated in the classic group and the oldest in the non-typical group (χ² = 19.6; p = 0.003). Fewer than half (47%) had the classic gestalt, and 5 of the 12 non-typical cases were mistaken for Sotos syndrome [23]. The French protocol estimates that the facies is recognisable by experienced clinicians in about half of patients and recommends retrieving photographs from early childhood [11]. The voice may be low-pitched and hoarse, with late and sometimes slurred speech [30].

3.4. Development, cognition and behaviour

ID is present in about 80–85% of patients and is mostly mild [1, 23]. GeneReviews describes the degrees in functional terms. In mild ID, the child attends mainstream school with additional support and independent living in adulthood is expected. In moderate ID, the child develops language but needs high levels of support or special education and, in adulthood, supported residential living. In severe ID, special education and considerable support in adulthood are needed [1]. The French protocol estimates preserved intelligence in 10%, mild ID in about 50% and a need for greater support in about 20%, with severe ID being rare [11]. Language delay is classic, and poor motor coordination (80% in the UK series) is partly attributable to ligamentous laxity [1, 11, 23]. Smith's atlas notes that, with age, few adults are described with clear ID [30].

Behavioural disorders are uncommon. Autistic traits, phobias and anxiety have been reported anecdotally [1], tantrums in 3 and attention deficit in 2 individuals in the UK series [23], and occasional attention deficit hyperactivity disorder [30]. The French protocol considers that autism spectrum disorder and anxiety are not classically described [11].

3.5. Nervous system and tone

Abnormal tone is frequent in infancy and generally resolves during childhood. Hypotonia, mainly central, was reported in 22 of 47 individuals, and hypertonia, mainly peripheral, with limb stiffness and brisk reflexes, in 14 of 51; three had peripheral hypertonia with central hypotonia [1]. Smith's atlas describes axial hypotonia with mild limb hypertonia and reports progressive spasticity [2, 30]. Neonatal hypotonia may significantly delay early motor milestones [11].

Brain magnetic resonance imaging (MRI) is not performed routinely and its findings are of uncertain significance [1]. Ventriculomegaly (7 individuals, with normal cerebrospinal fluid pressure and no need for shunting), neuronal migration defects with polymicrogyria, periventricular leukomalacia (2), cerebellar anomalies (2), cerebellar infarction and persistent cavum septum pellucidum have been described [1, 23]. ID was mild in 7 of the individuals with MRI abnormalities and moderate in 3. It was severe in 2 individuals with polymicrogyria and pachygyria, but one patient with polymicrogyria had normal developmental milestones and body asymmetry [1]. The French protocol estimates brain abnormalities in about a quarter of patients studied by imaging and does not ascribe prognostic value to them [11]. Recent cases have added callosal dysgenesis, frontal hypoplasia and arachnoid cyst [50]. The atlases also mention delayed myelination, fatty filum terminale, cerebral atrophy and hypervascularisation [30].

Epilepsy is rare and is not always associated with brain abnormalities. Four individuals had afebrile seizures, with tonic-clonic seizures beginning at 13 years or brief absences at 15, and two had seizures in a febrile context [1, 11].

3.6. Musculoskeletal system and connective tissue

Bone age is advanced in virtually all patients assessed: 33 of 33 in GeneReviews and 25 of 25 in the UK series [1, 23]. The carpus is more advanced than the metacarpals and phalanges, the reverse of what occurs in Sotos syndrome [2, 30]. The figures in the atlas of skeletal dysplasias show, at 15 months, a carpal bone age of 4 years 2 months and a metacarpal and phalangeal bone age of 3 years 6 months, and, at 5 years 3 months, 7 years 10 months and 6 years 10 months, respectively [2]. The discrepancy was recorded in 3 patients in the UK series, probably owing to under-reporting, because the questionnaire did not ask for it [23]. Other radiological findings include wide metaphyses, especially of the distal femur and the distal humerus or ulna, mild elongation of the short tubular bones, a narrow thorax and coxa valga [2, 11].

Camptodactyly affects the fingers, toes or both, is present mainly in the neonatal period and sometimes requires surgical correction in the feet [1, 38]. In adults a mild boutonnière-like deformity appears, with flexion of the proximal interphalangeal joint and hyperextension of the distal joint, without loss of function; it was observed in five adults in the UK series and is a diagnostic clue at older ages [1, 23]. Broad thumbs, thin deep-set nails, prominent fingertip pads, clinodactyly (9 in the UK series), overlapping toes and limited extension of the elbows and knees are also found [23, 30]. Scoliosis ranges from mild, requiring only monitoring, to severe early-onset scoliosis requiring surgery; it was reported in 13 individuals, and pectus excavatum or carinatum in 4 [1]. Talipes equinovarus occurred in 8, ranging from fixed and bilateral, requiring surgery, to mild and unilateral, resolved with physiotherapy [1].

The cervical spine deserves attention, although it is a less frequent problem in Weaver syndrome than in Cohen-Gibson syndrome. Published reports include anterior subluxation of C2 on C3 in flexion with a pyramidal syndrome, treated with posterior arthrodesis; a case in which the cervical collar was replaced by C1–C3 arthrodesis for impingement and stenosis; familial vertebral anomalies without deficit; and multilevel spondylolysis with late-diagnosed cervical myelopathy [11]. Kelly and colleagues described cervical spine anomalies and tumours in a family with Weaver syndrome in 2000 [30]. Bilateral congenital hip dislocation with talar hypoplasia and "neurological" hip dislocations in patients with hypertonia and contractures have also been described [11].

Ligamentous laxity with hypermobility and flat feet is common, but is usually reported only when there is pain. It contributes to poor coordination and delayed walking [1, 11]. Soft, "doughy" or redundant skin was reported in 23 of 43 children [1]. These signs should not suggest other connective tissue disorders, such as Ehlers-Danlos syndrome [11]. Umbilical hernia occurred in 28 of 57 children and was large enough to require surgery in 8 newborns; inguinal hernias are also described [1, 30].

3.7. Neonatal period

The neonatal period is notable for feeding difficulty (10 of 28 newborns; one needed a nasogastric tube for two weeks; hypotonia was reported in only 3 of these) and a hoarse, low-pitched cry (16 of 35) [1]. Neonatal hypoglycaemia and neonatal hypocalcaemia were each reported in 2 individuals [1]. Prenatally, fetal macrosomia, hydramnios and camptodactyly may occur, all non-specific [11]. A 2024 case had polyhydramnios from 20 weeks [51], and a Chinese newborn reported in 2025 weighed 5.04 kg at birth and was admitted for persistent neonatal jaundice [52].

3.8. Other organs

Congenital heart defects occur in about 10% and are not complex: ventricular septal defect, mitral regurgitation or prolapse and patent ductus arteriosus [1, 11]. Renal anomalies (nephromegaly, horseshoe kidney, hydronephrosis) were reported in 4 patients; hydrocele in 2–3, cryptorchidism in 1–2 and hypospadias in 1 [1, 11]. Cleft lip, cleft palate or velopalatal clefts occurred in 3, bifid uvula with agenesis of two teeth in 1 and conductive or sensorineural deafness in 3 [1, 11]. Ophthalmological findings occur in about 9% (strabismus, hypermetropia, myopia, sometimes high-grade) [11]. Other isolated findings, of uncertain relationship to the syndrome, were café-au-lait macules, haemangiomas, pigmented naevi, gastro-oesophageal reflux, hiatus hernia, hepatosplenomegaly and gingival hypertrophy [1, 11, 51]. Smith's atlas also mentions inverted nipples, fine hair, hypothyroidism and adult growth hormone deficiency, the frequency of which is not quantified [30].

3.9. Tumours

Neuroblastoma is the only tumour with a demonstrated increased incidence [1, 30]. In GeneReviews, 7 of 70 individuals had tumours (one of them two) [1]:

Sources: [1, 9, 49].

The French protocol counts tumours in 8 of 86 carriers (9%): neuroblastoma in 5 (6%), mainly abdominal (adrenal or paravertebral), and haematological neoplasms in 3 (3%) [11]. The AACR cites a study that brought together 56 individuals with Weaver syndrome, of whom 8 had cancer, 6 of them neuroblastoma, and assigns neuroblastoma in carriers of inactivating EZH2 variants a penetrance of 11%. It acknowledges, however, that the number of cases is small and that publication bias is likely [27]. Before the molecular era, the following were reported: a sacrococcygeal teratoma resected at birth in one of two half-siblings with cervical spine anomalies (Kelly and colleagues, 2000), an endodermal sinus tumour of the ovary at 17 years in a mother with a probable diagnosis of Weaver syndrome (Derry and colleagues, 1999) and acute lymphoblastic leukaemia in a 4.5-year-old girl (Basel-Vanagaite, 2010) [15, 30]. There are no data on cancer in adulthood, and the lifetime risk is unknown [11, 30]. None of the 48 individuals in the UK series had a chronic haematological neoplasm [23].

3.10. Cohen-Gibson syndrome (EED)

The 2025 GeneReviews brings together 19 individuals published between 2015 and 2024, including a second male patient, patients with a previous clinical diagnosis of Weaver syndrome, three patients with possible mutational hotspots and one patient with epilepsy [4, 35, 57–62]. Overgrowth, present in 15 of 16 with data, begins in the fetal period or early childhood. Gestational age at delivery ranged from 36 to 42 weeks. Birth weight ranged from 3550 g (boy, 38 weeks) to 5160 g (boy, 40 weeks), with z-scores of 0 to 4.2 and none below the mean; length ranged from 52 cm (z 0.5) to 57 cm (z 3.0) and HC from 35 cm (z 0) to 37.2 cm (z 2.0). In childhood and adolescence, z-scores ranged from 0.84 to 5.1 for height, 1.0 to 3.8 for weight and 0.6 to 4.0 for HC; body mass index (BMI) had a z-score of about 1.4 to 1.8. In adults, height had a z-score of 1.85 to 3.1 and HC of 1.46 to 3.9. Two adults in their mid-20s had a BMI of 32.6 and 38.2 kg/m² [4].

Psychomotor delay and ID are present in all, from mild to severe, and coordination and balance difficulties may persist into adulthood. Autism spectrum disorder and major behavioural problems have not been described and, if present, should prompt a search for another cause [4]. The most distinguishing features of the syndrome are musculoskeletal and their persistence into adulthood: large hands and feet, long fingers, camptodactyly and contractures, pes planovalgus or clubfoot, thoracic scoliosis or kyphoscoliosis (6/19), hypermobility of the small joints, recurrent patellar subluxation and skin fragility [4]. Cervical spinal stenosis was described in 3 patients, one of whom needed laminectomy and arthrodesis and another had myelopathy at C3; in a fourth, C1–C2 atlantoaxial instability required laminectomy and fusion. Lumbar stenosis or spondylolisthesis occurred in 2 [4]. In one patient, neck movement during gymnastics caused neurological compromise requiring surgery [4]. The first patients illustrate the evolution of growth. A Turkish man had a height of +4.6 SD at 8 years, which fell to +1.85 SD in adulthood; a Hispanic girl was born with a weight of +4.2 SD and had cleft palate, tracheomalacia, heart disease (ductus arteriosus occluded by coil, atrial septal defect, mitral regurgitation), cervical stenosis and moderate ID; a Japanese boy had, at 5.4 years, a height of +4.5 SD, weight of +3.8 SD and HC of +2.6 SD, with an intelligence quotient between 57 and 65 [17]. Comparison with 56 patients with Weaver syndrome showed more heart disease, more cervical spine anomalies, cryptorchidism in boys and a generally distinguishable facial gestalt [24]. Orphanet describes the ID as more severe than in the other entities of the group, with dysmorphism that remains recognisable in adulthood [5].

Source: [4]. ASD, atrial septal defect; VSD, ventricular septal defect; PDA, patent ductus arteriosus.

The facies includes a round face, prominent forehead, hypertelorism, low and broad nasal bridge, large ears (with or without posterior helical pits and lobe creases), long philtrum, horizontal chin crease and retrognathia, with an open mouth due to facial hypotonia. The palpebral fissures are narrow or short, with downward slanting of the lateral portion of the upper eyelid and eversion of the lateral lower eyelid [4, 30]. The voice may be hypernasal, low-pitched or hoarse. Bilateral cleft palate, bifid uvula, epilepsy (6 patients, one with hyperinsulinaemic hypoglycaemia), enlarged ventricles, white matter loss, arachnoid and septum pellucidum cysts, pituitary microadenoma and a thin corpus callosum have been reported [4]. Hyperinsulinaemic hypoglycaemia occurred in a newborn, who required a glucose infusion until resolution at 21 days, and in a child, treated with octreotide. Other findings described were adrenal calcification with hypocortisolism, tracheomalacia requiring surgery, intestinal malrotation with Hirschsprung disease, cataract at 30 years, chorioretinal degeneration and hearing loss [4]. No childhood tumours have been described in germline carriers. One adult had anal squamous cell carcinoma at 61 years [4].

3.11. Imagawa-Matsumoto syndrome (SUZ12)

The manifestations overlap with those of the other two syndromes, but generalised overgrowth, ID, scoliosis and lax skin are less frequent [6, 7]. Retrognathia, a hoarse, low-pitched cry and a low nasal bridge are usually absent, according to Orphanet [6], and the EZH2 GeneReviews identifies hypertrichosis and normal skin texture as distinguishing features [1]. The 2019 series described variable prenatal and postnatal growth, dysmorphism, musculoskeletal anomalies, developmental delay or ID, genitourinary anomalies and structural brain anomalies in some patients and a possible association with respiratory problems [13]. Prenatal growth appears less marked and postnatal growth more marked than in Cohen-Gibson syndrome [4].

A systematic review published in 2026 brought together 19 cases: 5 from the United States, 3 from Japan, 3 from Turkey and 2 from China; 7 (36.8%) were female, and the median age at diagnosis was 12 years (from the fetal period to 41 years) [21]. The table summarises the frequencies.

Source: [21].

In an analysis of 13 patients published in 2019, mean birth length was +1.99 SD, mean height in childhood +2.53 SD and mean final adult height +3.75 SD; mean HC at birth was +0.85 SD, and two patients had no overgrowth. Most had advanced bone age, with large hands and feet, and 6 of 8 boys had genitourinary anomalies, 4 with cryptorchidism. In the series of 10 patients, 8 had developmental delay, 6 ID, 2 autism spectrum disorder and 4 structural brain anomalies (thin, hypoplastic or absent corpus callosum, polymicrogyria). Missense variants appeared to be associated with more marked growth (mean maximum weight +4.6 SD, versus +2.15 SD for truncating variants), in a sample too small for conclusions [13, 19]. Expression is highly variable. A patient with the splicing variant c.1023+1G>C had neither macrocephaly nor overgrowth [7]. A 19-year-old woman with a de novo 1.5 Mb deletion at 17q11.2 including SUZ12 had a behavioural disorder, ID, macrocephaly, agenesis of the corpus callosum and polymicrogyria [26]. The 2026 cases added cervical stenosis with spinal cord compression in an 11-year-old girl [21], hypertrichosis and macrocephaly in a family with maternal transmission [22] and keloids [63]. No tumours have been described in children with germline SUZ12 variants, and cancer predisposition in isolated IMMAS has not been established [1, 10, 21].

3.12. Adulthood

In Weaver syndrome, the facies becomes less marked, the retrognathia disappears and the hypotonia resolves; tall stature, the mild boutonnière deformity and poor coordination persist [1, 3, 23]. There are few data on older adults. The UK series considered an increased risk of haematological neoplasms late in life to be possible, given that the same variants occur somatically in myeloid neoplasms of older people, but this risk has not been quantified [23]. In Cohen-Gibson syndrome, the facies remains recognisable in adulthood, and obesity, early cataract and persistent motor difficulties have been described [4, 5].

4. Clinical Recognition and Warning Signs

4.1. When to suspect, by age

Before birth, the syndrome is rarely suspected. Fetal macrosomia, hydramnios and camptodactyly are non-specific signs that carry weight only when a first-degree relative has the diagnosis [11]. In the newborn and infant, suspicion arises from increased length (more than weight), macrocephaly, camptodactyly, large umbilical hernia, soft lax skin, a hoarse, low-pitched cry, feeding difficulty and abnormal tone, with axial hypotonia and peripheral hypertonia [1, 23, 38]. As several of these signs improve or disappear, GeneReviews recommends taking a detailed neonatal history when the child is assessed later [1].

In childhood, the most frequent picture combines delayed walking and language, hypotonia and advanced stature (about 80%), and the facies is most recognisable between 1 and 3 years [11, 38]. In adolescence and adulthood, tall stature with advanced bone age, scoliosis, camptodactyly or the boutonnière deformity, and mild to moderate ID predominate [11, 23]. At this stage, photographs from early childhood help to recognise the facies [1].

Tall stature is, in most cases, a normal variant of polygenic origin. A Turkish case published in 2025 summarises the signs that should lead to consideration of a syndromic cause: height above the parental target, increased length at birth, normal growth velocity with a stable high centile, mild dysmorphism and mild ID [53]. In the general context of overgrowth, a syndrome should be suspected when global overgrowth (increased height and HC) is associated with ID or a congenital anomaly [38].

The GeneReviews suggestive criteria for EZH2 include height ≥ +2 SD, macrocephaly ≥ +2 SD, ID, characteristic facies, poor coordination, soft or redundant skin, camptodactyly, umbilical hernia, abnormal tone, a hoarse, low-pitched cry and advanced bone age [1]. For EED, the suggestive criteria are height, HC and hand and foot length with z ≥ 2, psychomotor delay and ID, bone age with z ≥ 2, radiological changes (widened or lucent metaphyses, flattened epiphyses and acetabula, coxa valga) and cervical spine anomalies. An adult of normal height who was relatively tall, or who had advanced bone age in childhood, may also meet the overgrowth criterion [4].

4.2. Warning and emergency signs

Detection of neuroblastoma symptoms may be difficult in children with a neurodevelopmental disorder, because changes in behaviour and sleep are confused with those of the syndrome itself. For this reason, the French protocol recommends systematic blood pressure measurement [11].

4.3. Newborn screening

There is no newborn screening for any of the PRC2 syndromes. Population screening for neuroblastoma in infants, carried out in Japan, North America and Europe by measurement of urinary catecholamines, detected many tumours that would have regressed spontaneously and did not reduce high-risk neuroblastoma at later ages; it should not be confused with surveillance targeted at risk groups [27].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Criteria

There are no validated clinical criteria for Weaver syndrome. Before the discovery of the gene, Khosravi and colleagues proposed, on the basis of 34 published patients, that the diagnosis should require at least 8 of 10 features: excessive postnatal growth, macrocephaly, increased bifrontal diameter, hypertelorism, prominent or long philtrum, micrognathia, large ears, hoarse or low-pitched cry, advanced general and carpal skeletal maturation and wide metaphyses [15]. The 2013 UK series showed that fewer than half of carriers of an EZH2 variant have the classic gestalt, and in 2018 two SUZ12 variants were described in patients who did not meet the criteria for Weaver syndrome; both findings show the limits of clinical criteria [23, 36]. The diagnosis of all three entities is established by identifying a heterozygous germline pathogenic or likely pathogenic variant in the respective gene in an individual with suggestive findings. Under the standards of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP), the two classes have the same value for clinical decision-making. A variant of uncertain significance (VUS) neither confirms nor excludes the diagnosis [1, 4, 11].

5.2. Non-molecular investigations

Auxological assessment includes height (or length), weight and HC with the respective SD, the growth chart and target height calculated from parental heights [11, 38]. Bone age is assessed on a radiograph of the left hand and wrist (Greulich and Pyle or BoneXpert methods). The dissociation between the carpus and the metacarpals should be recorded, because it points to Weaver syndrome and away from Sotos syndrome [2, 11]. As bone age does not reflect actual maturation, methods for predicting final height from bone age, such as Bayley-Pinneau, do not apply in this syndrome [11].

To exclude an endocrine cause, the French protocol proposes minimal laboratory testing with thyroid-stimulating hormone (TSH), free thyroxine (FT4), insulin-like growth factor 1 (IGF-1) and IGF-binding protein 3 (IGFBP3). FSH, LH, oestradiol and testosterone are added if there are pubertal signs, and Δ4-androstenedione, 17-hydroxyprogesterone, testosterone and dehydroepiandrosterone sulphate (DHEA-S) if there is pubic hair or signs of hyperandrogenism [11]. In Weaver syndrome, these investigations are normal [11, 53]. The karyotype and FBN1 analysis were normal in the Turkish case, before the exome clarified the diagnosis [53].

5.3. Molecular tests

The first-line test is a multigene overgrowth–ID panel including EZH2, EED, SUZ12, NSD1, NFIX, DNMT3A and the other genes in the differential, or exome or genome sequencing when the phenotype is atypical [1, 4]. The choice is justified by the clinical overlap: in the UK series, 5 of 48 carriers of an EZH2 variant were mistaken for Sotos syndrome, and the authors recommended studying NSD1 if EZH2 was negative and vice versa, preferably simultaneously [23]. The Oxford Desk Reference recommends studying EZH2 in cases of Sotos syndrome without an NSD1 variant [38]. Some overgrowth–ID panels do not include EED [4]. Detection of 17q11.2 deletions including SUZ12 requires microarray or copy number variation analysis [26].

In a UK series of 710 individuals with OGID, a causal variant in one of 14 genes was found in 50% (353/710); six epigenetic regulation genes (NSD1, EZH2, DNMT3A, CHD8, HIST1H1E and EED) accounted for 44% (311/710) [39]. In a Brazilian cohort of 37 patients with height above the 97.7th centile and syndromic features, the exome yielded a diagnosis in 29.7% (11/37), including one SUZ12 case [64].

Sources: [1, 4, 13, 26].

5.4. DNA methylation episignature

Using genome-wide methylation data from 187 individuals with OGID and 969 controls, Choufani and colleagues defined an EZH2 DNA methylation signature. The signature is sensitive and specific, distinguishes loss-of-function missense variants from gain-of-function variants, detects somatic mosaicism, classifies EED and SUZ12 variants and predicted pathogenic variants in undiagnosed individuals with OGID [25]. GeneReviews recommends considering it, on peripheral blood DNA, when there is clinical suspicion without an identified variant or when a VUS has been identified [1, 4]. In the case of EED mosaicism published in 2026, the high-confidence PRC2 signature supported the classification of a variant with an allele fraction of 31% in blood [12]. For the interpretation of EED VUS, a functional assay in Drosophila has been calibrated with known variants, but it remains a research tool. Up to 1% of the general population may carry a rare variant in one of these three genes, most classified as VUS [48].

5.5. Limitations of the tests

Sequencing does not detect large deletions or duplications, and the exome and genome detect them with lower sensitivity than targeted methods (MLPA, quantitative PCR, microarray) [1]. A negative result does not exclude the diagnosis: there may be non-coding variants or low-level mosaicism in blood [4, 12]. In the girl with EED mosaicism, the allele fraction was 31% in blood and 11–17% in buccal mucosa samples. Deep sequencing and the study of several tissues were needed to confirm the mosaicism [12]. The proportion of variants detected by a method (97% for EZH2 sequencing) is not equivalent to the clinical sensitivity of the test, which depends on the phenotype and on the proportion of cases due to other causes. In patients with clinical suspicion and a negative test, the French protocol recommends reassessment and continued aetiological investigation [11]. In the Turkish case, the diagnosis was made only at 14 years, after extensive endocrine investigation [53].

5.6. Report content and variants of uncertain significance

The report should state the gene, transcript and version (preferably MANE: NM_004456.5, NM_003797.5 or NM_015355.4), the HGVS c. and p. nomenclature, zygosity, the allele fraction when it suggests mosaicism, the ACMG/AMP class and the criteria used [1, 9, 21]. Variants receive different names depending on the transcript; the same p.Arg684Cys change appears with different c. numberings in the literature [9, 45]. Family segregation, de novo occurrence and the evolution of knowledge allow most VUS to be reclassified [11]. The episignature and functional assays are the reclassification tools available [25, 48]. The author of the EED GeneReviews offers to review VUS in EED, EZH2 and SETD1B [4].

5.7. Practical algorithm

5.8. Prenatal and preimplantation diagnosis

When the familial variant is known, prenatal diagnosis (chorionic villus sampling between 12 and 14 weeks of gestation, or amniocentesis from 15 weeks, with a risk of fetal loss estimated at about 0.5%) and preimplantation genetic diagnosis after in vitro fertilisation are possible [1, 11]. In France, non-invasive prenatal exclusion diagnosis is possible when the index case's variant arose de novo or was inherited from the father [11]. Ultrasound findings (macrosomia, hydramnios, camptodactyly) do not allow the diagnosis, and fetal neuroimaging in Imagawa-Matsumoto syndrome has been described in only one report [11, 65].

6. Differential Diagnosis

6.1. Framework

The differential diagnosis of Weaver syndrome includes the other syndromes of global overgrowth with ID, especially those of epigenetic regulators, syndromes with macrosomia of genetic cause and non-genetic causes of macrosomia and tall stature [1, 11, 38]. Non-genetic macrosomia should be borne in mind in the prenatal period, for example in gestational diabetes, maternal obesity or major weight gain during pregnancy [11]. Macrosomia is a minor and inconsistent criterion in the RASopathies (Noonan, cardiofaciocutaneous, Costello), in Pallister-Killian syndrome (supernumerary isochromosome 12p), in Phelan-McDermid syndrome (22q13 monosomy) and, rarely, in mosaic trisomy 8 [11]. Other chromosomal anomalies with overgrowth are duplication or triplication of 15q including IGF1R and 4p16.3 duplication including FGFR3 [38].

6.2. Comparative table

6.3. Weaver and Sotos

Sotos syndrome is the differential diagnosis with the greatest overlap. The two share prenatal and postnatal overgrowth, variable ID, advanced bone age, scoliosis and hypermobility [1]. The facial difference is clearest between 1 and 3 years: in Sotos syndrome, the face is long and narrow, with a prominent chin and malar flushing; in Weaver syndrome, the face is round, with retrognathia and a "stuck-on" chin [1, 23]. The hand radiograph helps: in Weaver syndrome, the carpus is more advanced than the metacarpals and phalanges, and in Sotos syndrome it is delayed relative to them [2]. The growth profile also differs. About 50% of patients with Weaver syndrome have increased height only and 5% increased HC only; in Sotos syndrome, increased HC is more characteristic (about 20% with increased HC only) [23]. ID is mild in about half of patients with Weaver syndrome and moderate in 45–50% of patients with Sotos syndrome [23]. Despite these differences, clinical distinction is not reliable, and 5 of 48 EZH2 carriers were mistaken for Sotos syndrome [23].

6.4. Differences within PRC2

Comparison between EED and EZH2 suggests more heart disease, more cervical anomalies, more frequent cryptorchidism and a more persistent facies in the former [5, 24]. Both show frontal bossing and a horizontal chin crease. Cohen-Gibson syndrome also shows telecanthus, narrow palpebral fissures with lateral ptosis and eversion of the lateral portion of the lower eyelid [30]. In Imagawa-Matsumoto syndrome, the common features are present less frequently [6, 7]. The number of patients with EED and SUZ12 variants is still too small to establish robust distinguishing features [1]. As the three entities share the same methylation signature, the distinction is made by sequencing [25].

7. Treatment and Risk Management Strategies

7.1. Principles and evidence base

There is no curative treatment or therapy directed at the epigenetic defect in any of the three syndromes. Management consists of treating the manifestations, monitoring for complications and supporting development, preferably with a multidisciplinary team coordinated by a clinical geneticist or a reference centre [1, 4, 11]. The evidence base is limited. The GeneReviews chapters on EZH2 and EED state that there are no published guidelines and that the recommendations are based on the authors' experience [1, 4]. The French National Diagnosis and Care Protocol (PNDS) for Weaver syndrome, published in December 2025 by the Caen reference centre for developmental anomalies (AnDDI-Rares network), is the first guideline dedicated to the syndrome. It is based on publications and expert opinion and has not been validated by the Haute Autorité de Santé, which did not take part in its preparation [11]. The only recommendation based on formal international consensus is the neuroblastoma surveillance of the American Association for Cancer Research (AACR) [27]. There are no randomised trials on any aspect of management [10].

7.2. Assessments after diagnosis

7.3. Treatment of manifestations

Development and behaviour. Regular developmental monitoring and early referral to intervention (physiotherapy, occupational therapy, speech and language therapy and feeding therapy) are recommended. Psychomotor therapy helps the hypotonic child to achieve stable sitting and walking earlier [1, 11]. Speech and language therapy treats delay in spoken and written language, drooling and oral difficulties. Occupational therapy supports independence in daily activities, and neuropsychology helps when categorical diagnosis is difficult [11]. When expressive language is limited, augmentative and alternative communication should be assessed; it does not hinder speech development [1]. Children with autistic traits may benefit from behavioural interventions, and medication for attention deficit hyperactivity disorder (ADHD) follows the usual indications [1].

Epilepsy. It is treated with antiseizure medications according to general principles, by an experienced neurologist. No drug has been shown to be more effective in these syndromes, and education of carers about the forms in which seizures present is recommended [1, 4, 11]. In Cohen-Gibson syndrome, hyperinsulinism should be excluded in the presence of seizures [4].

Feeding. Feeding difficulty is treated with feeding therapy. Swallowing assessment should be requested with a low threshold if there is choking, poor weight gain or recurrent respiratory infections. Foods may be thickened or chilled, and a nasogastric tube or gastrostomy is reserved for severe or persistent dysfunction [1].

Hands and feet. Camptodactyly is treated with physiotherapy and orthoses to limit worsening and stiffness. Surgery is reserved for significant functional impairment or difficulty with footwear, which is rare; it is occasionally needed for the toes [1, 11]. Clinodactyly is only monitored, unless it worsens or causes functional impairment [11]. Talipes equinovarus is treated as in the general population, and painful flat feet with insoles [1, 11]. In nail dysplasia, the nails should not be cut too short; ingrown toenail is treated with local care and, rarely, with surgical repair [11].

Spine and thorax. The spine should be examined at every visit. A radiograph of the whole spine, anteroposterior and lateral, preferably standing and with a low-dose system (EOS®), is requested if there is asymmetry of the waist creases or shoulders or a rib hump on forward bending [11]. Scoliosis is followed by a paediatric orthopaedic surgeon or a physical medicine and rehabilitation physician. A brace is used if the curve exceeds 15° and is progressing, and surgery for major deformities. Physiotherapy addresses tone, posture and hamstring extensibility, and respiratory physiotherapy is added if thoracic scoliosis is severe [1, 11]. Treatment of pectus is cosmetic; pectus carinatum may be corrected with a brace if there is associated scoliosis [11].

Cervical spine. The neurological examination should look for hypertonia, reflex changes and Babinski sign. When instability is suspected, lateral radiographs in flexion and extension and/or MRI of the craniocervical junction and cervical spine are requested; computed tomography (CT) helps surgical planning [11]. Treatment is decided by a team experienced with the craniocervical junction. A collar limiting flexion may suffice, and cervical fixation is indicated when there is stenosis, spinal cord compromise or impingement [11]. In Cohen-Gibson syndrome, surgery is indicated if instability is severe with present or probable neurological compromise, and specialist neurosurgical advice is recommended particularly for prophylactic intervention [4]. There are insufficient data to recommend routine cervical imaging, but the threshold for requesting it should be low in the presence of gait deterioration, especially of rapid onset. Elective imaging should be considered before activities involving abrupt head and neck movements [4]. In the girl with Imagawa-Matsumoto syndrome and cervical stenosis with spinal cord oedema, posterior decompression with canal expansion and fixation did not bring significant improvement in gait [21].

Tone, contractures and hip. For peripheral hypertonia and contractures, physiotherapy, orthoses and, at the discretion of the rehabilitation physician or orthopaedic surgeon, intramuscular botulinum toxin or surgery are used. For hypertonia or dystonia, specialists may consider baclofen, tizanidine or antiparkinsonian agents [1, 4, 11]. If hip mobility changes or there is pain, a pelvic radiograph is requested, because of the risk of neurological dislocation [11]. Painful laxity is treated with proprioceptive and postural physiotherapy, muscle strengthening and, sometimes, compression garments; if pain persists, referral to a chronic pain clinic should be made [11].

Hernias. In children, umbilical hernia may be monitored, because it often closes spontaneously before age 5 years. Surgery is indicated if it persists after that age or if there is a complication [11]. Some neonatal hernias are large and require early surgery [1]. Inguinal hernias and cryptorchidism are treated as in the general population [4].

Other organs. Heart defects, renal anomalies, cleft palate (surgery, and orthodontics for retrognathia), refractive errors, strabismus, cataract and deafness are treated according to the general principles of each specialty [1, 4, 11].

Hyperinsulinaemic hypoglycaemia (EED). It is treated with glucose and, if it persists, octreotide is considered, with endocrinological assessment if hypoglycaemia persists or insulin is raised [4]. Octreotide was used in a single patient with Cohen-Gibson syndrome, and the reference publication does not state the dose [4].

7.4. Tall stature

There is no consensus on the treatment of tall stature without an endocrine cause [11]. In 2002, the use of high-dose oestrogens in a patient with Weaver syndrome was published, but this treatment is now highly controversial, because of doubts about an increased tumour risk (melanoma, breast cancer) and impaired fertility in treated adolescent girls [11]. Bilateral percutaneous epiphysiodesis, used in constitutional tall stature, might reduce final height by 4 to 6 cm relative to predicted height, but it has never been described in Weaver syndrome. Prediction of final height from bone age, on which the decision depends, is not reliable in this syndrome [11]. Psychological support may be offered according to how well the stature is tolerated [11].

7.5. Tumour surveillance: recommendations and divergences

Neuroblastoma surveillance in Weaver syndrome is the point on which the sources diverge most. In 2017, the AACR did not recommend surveillance. It pointed to a possibly only slightly increased risk of neuroblastoma, with numbers too small to calculate the absolute risk, and advised clinical vigilance and investigation of symptoms [55]. The 2013 UK series had reached the same conclusion [23]. In the 2024 update, resulting from the 2023 workshop, the AACR moved to recommending routine surveillance for carriers of inactivating EZH2 variants. The inclusion criterion is a neuroblastoma prevalence ≥ 1% among carriers, or segregation with neuroblastoma in at least two individuals from at least two families, and the AACR estimates a penetrance of 11% for EZH2, acknowledging publication bias [27]. GeneReviews adopted this recommendation in the June 2025 revision, but notes that surveillance has been inconsistent between countries and that there are no data supporting the modality, interval or duration [1]. The German portal for cancer predisposition syndromes, at Hannover Medical School (Medizinische Hochschule Hannover), adopts the AACR protocol [67].

The French protocol of December 2025 considers the information insufficient for specific surveillance recommendations. It advises systematic enquiry for tumour symptoms and signs, including blood pressure, discussion of a full blood count with platelets and a low threshold for further investigations and referral to the paediatric haematology-oncology clinic [11]. The 2017 Oxford Desk Reference also did not recommend screening, but anticipated that the recommendation might change with more data [38].

The common AACR protocol for neuroblastoma predisposition is as follows [27]:

The AACR rationale is based on the epidemiology of neuroblastoma. The median age at clinical diagnosis is about 20 months, about 80% are diagnosed by age 6 years and 98% by age 10 years. About 80% of tumours arise in the adrenal gland or the paravertebral sympathetic chain below the diaphragm (hence the ultrasound) and 20% in the neck and thorax (hence the radiograph, followed by contrast-enhanced MRI if there is a mass). About 90% secrete catecholamines intermittently. MRI is indicated if ultrasound is technically inadequate or if another investigation raises suspicion. Surveillance should begin as soon as the variant is identified, because neuroblastoma may be present at birth, and families should be taught to recognise the signs of the disease. Compared with the 2017 version, the 2024 protocol reduced the frequency of chest radiographs [27]. The AACR itself acknowledges that surveillance will detect some tumours that would have regressed or matured spontaneously, especially in infants. Its mortality benefit has not been demonstrated, and previous population screening did not reduce high-risk neuroblastoma [27]. In practice, the decision between the US protocol and French clinical vigilance should be shared with the family, with information on the absolute risk (5–11% depending on the source), anxiety and false positives [1, 11, 27].

For EED and SUZ12, no guideline recommends tumour surveillance [4, 10]. As somatic EED and SUZ12 variants are frequent in MPNST, the EED GeneReviews recommends that lesions suspicious for MPNST be followed by clinical examination and/or imaging, as in the NF1 guidelines [4]. A Chinese team that published a SUZ12 case proposed tumour surveillance, but this is the authors' opinion, without risk data [21].

There are no data on cancer risk in adults with Weaver syndrome [11]. The UK series suggested clinical vigilance for haematological neoplasms in older adults, without a specific programme [23].

7.6. Pharmacotherapy and agents to avoid

No drug has an indication in these syndromes. The drugs used (antiseizure medications, baclofen, tizanidine, botulinum toxin, ADHD medication, octreotide for hyperinsulinism in EED syndrome) are the usual ones for each manifestation [1, 4, 11].

The specific pharmacological precaution concerns the EZH2 and PRC2 inhibitors used in oncology. GeneReviews and the French protocol warn that the efficacy and adverse effects of these inhibitors may be altered in people with germline EZH2 or EED variants. Adverse effects may be increased, and there are no clinical data on their use in these syndromes. If a cancer treatable with these drugs arises, such as a haematological malignancy or non-small cell lung cancer, treatment should be adapted to this risk [1, 4, 11]. Tazemetostat (TAZVERIK®) was approved by the Food and Drug Administration (FDA) on 23 January 2020, under accelerated approval. It is indicated for metastatic or locally advanced unresectable epithelioid sarcoma, in adults and adolescents aged 16 years and older, and for relapsed or refractory follicular lymphoma in adults. The label warns of second primary malignancies: among 758 adults treated with 800 mg twice daily as monotherapy, 1.7% developed myelodysplastic syndrome, acute myeloid leukaemia or B-cell acute lymphoblastic leukaemia, and one paediatric patient developed T-cell lymphoblastic lymphoma. The label also warns of embryo-fetal toxicity [68]. The French protocol also cites valemetostat [11].

In Cohen-Gibson syndrome, and by extension when cervical anomalies are documented, caution is advised with activities involving rapid neck movement or a risk of head and neck trauma, such as contact sports and extreme amusement rides [4].

7.7. Clinical trials and developments in the last 12 months

On 29 October 2026, ClinicalTrials.gov listed no interventional trials for Weaver, Cohen-Gibson or Imagawa-Matsumoto syndromes, or for EZH2-related overgrowth; the only related study is observational, on body perception and representation in overgrowth syndromes (NCT04993235, status unknown) [70]. The available results are all preclinical or observational.

In the last 12 months, the development with the greatest clinical impact was the publication of the French PNDS (December 2025), a national guideline based on expert consensus [11]. In July 2025, the AACR published a synthesis of the 2023–2024 updates of its paediatric surveillance recommendations [71]. Case reports from 2025–2026 expanded the phenotype of EZH2 (severe camptodactyly and callosal dysgenesis; macrosomia of 5.04 kg with persistent jaundice) and SUZ12 (cervical stenosis, hypertrichosis, keloids) [21, 22, 50, 52, 63]. There were no sponsor announcements, regulatory submissions, opinions or authorisations for these syndromes. A 2025 review of overgrowth–ID syndromes describes the mobilisation of patient organisations and researchers to develop treatments, but deals mainly with Sotos, Malan and Tatton-Brown-Rahman syndromes [72].

7.8. Pregnancy, perioperative care and transition

Pregnancies in which the mother and/or the fetus carry an EZH2 variant are generally uncomplicated. The obstetric team should know that the newborn may be large, in order to plan the delivery [1]. Hydramnios has been described in isolated cases [11, 51]. The sources consulted do not describe specific anaesthetic recommendations. In the presence of known cervical anomalies (especially EED and SUZ12), or of signs of myelopathy, the cervical spine should be assessed before procedures requiring neck manipulation [4, 11, 21]. In late adolescence, transition to adult services should be organised to avoid interruption of follow-up. In adults, the French protocol provides for reference centre visits every 2–3 years and follow-up by the general practitioner, neurology, orthopaedics and physical medicine and rehabilitation, as needed [11]. In France, the protocol also provides for legal protection measures for vulnerable adults, when necessary [11].

7.9. Summary of guidelines and divergences

Sources: [1, 4, 11, 27].

8. Reproductive Options and Family Counselling

8.1. Inheritance and proportion of de novo cases

All three syndromes have autosomal dominant inheritance [1, 4, 6]. For EZH2, among 70 individuals compiled by GeneReviews, 35 were isolated cases with a de novo variant, 18 had a familial variant and in 17 the origin could not be confirmed [1]. The French protocol estimates 70% de novo and 30% inherited variants among cases with information [11]. The 2024 compilation of 56 patients included 27 de novo, 13 familial and 16 without data [51]. Smith's atlas notes that, although most cases are sporadic, familial cases are not rare, and that part of the mild spectrum is identified by testing relatives of classic cases [30]. For EED, most variants are de novo; two families had an affected mother with more than one affected child, and post-zygotic mosaicism has been described [4, 12, 62]. For SUZ12, variants arise de novo or are inherited, sometimes from a minimally affected parent, and one was inherited from a mosaic father [13, 14, 22].

8.2. Recurrence risks

Parental testing is recommended when the proband appears to be the only affected individual, because an apparently negative family history may conceal a parent with a mild phenotype [1]. A parent with somatic and germline mosaicism may have only subtle signs, such as tall stature, above-average HC, mild dysmorphism or a history of mild delay, which are detected only if actively sought [4]. Testing of leucocytes may not detect somatic mosaicism and does not detect exclusively germline mosaicism [1, 4]. In the mosaic EED case, the allele fraction ranged from 11% to 31% depending on the tissue [12]. If the parent is a carrier, the parent's relatives may also be at risk [1].

8.3. Prenatal and preimplantation diagnosis and alternatives

The ideal time to discuss risk and options is before pregnancy, and affected or at-risk young adults should receive genetic counselling [1]. With the familial variant identified, prenatal diagnosis by chorionic villus sampling (12–14 weeks of gestation) or amniocentesis (from 15 weeks), with a risk of fetal loss of about 0.5%, and preimplantation genetic diagnosis are possible [1, 11]. The French protocol accepts non-invasive prenatal exclusion diagnosis when the index case's variant is de novo or paternal, and mentions gamete donation as an alternative for couples who do not wish to resort to termination of pregnancy or embryo selection. Requests are discussed in a multidisciplinary prenatal or preimplantation diagnosis centre [11]. For the parents of a child with a de novo variant, prenatal diagnosis may be offered in subsequent pregnancies because of the risk of germline mosaicism [11]. Perspectives on the use of these tests differ between professionals and between families, and the decision is a personal one [1].

8.4. Disclosure of the diagnosis, family information and minors

The French protocol recommends a dedicated disclosure consultation, held jointly by the physician who ordered the test and a geneticist and, where possible, with a psychologist. It also recommends a follow-up consultation a few weeks later, because the initial information is difficult to take in [11]. There is no consensus on the age at which to explain the diagnosis to the child; this should be decided case by case, according to the child's maturity and questions, and psychological support may be extended to siblings [11]. In France, the law places on the carrier the obligation to inform the family (Decree No. 2013-527 of 20 June 2013); if the carrier declines to do so, they may ask the physician to contact the relatives [11]. Testing of asymptomatic siblings is clinically useful in Weaver syndrome, because early diagnosis allows tumour surveillance to be planned during the neuroblastoma risk window, which the AACR extends up to age 10 years [27]. In adults with a mild phenotype, testing clarifies reproductive risk [1].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence and published cases

There is no population-based prevalence estimate for any of the syndromes. Orphanet places all three in the prevalence class of less than 1 in 1,000,000. This class is an upper bound and not a measured rate, and is based on 48 cases or families for Weaver syndrome (2013 UK series), 12 for Cohen-Gibson syndrome and 15 for Imagawa-Matsumoto syndrome [3, 5, 6, 10]. GeneReviews does not estimate the prevalence of Weaver syndrome because mild forms escape diagnosis [1]. The number of published individuals with an EZH2 variant is at least 70 in GeneReviews, 72 in the French protocol (with 86 carriers used in calculating tumour frequencies) and fewer than 70 according to a 2025 Chinese report [1, 11, 52]. There are 19 individuals with EED variants in the 2025 GeneReviews, to which a mosaic case was added in 2026 [4, 12]. For SUZ12, there are about 19 to 20 cases [21, 22]. Ascertainment depends on clinical suspicion of Weaver syndrome, which makes underestimation of mild forms likely, and the volume of ClinVar submissions measures testing frequency and not disease frequency [10].

9.2. Geographical distribution in the series

The 2013 UK series included patients from several European countries, among them two from Portugal, and from the United States and Japan, recruited by clinical genetics centres [23]. The Japanese series identified the first EED and SUZ12 variants [14]. There are published cases from Korea, Turkey, China and Taiwan [1, 50–53]. In Brazil, teams from São Paulo (University of São Paulo) and Curitiba (Federal University of Paraná) took part in the description of SUZ12 variants, one of them in a 19-year-old Brazilian man [7, 20, 36]. In a Brazilian cohort of 37 patients with syndromic tall stature, one of the 11 diagnostic variants was in SUZ12 [64]. In the 2026 review of the 19 SUZ12 cases, most came from the United States (5), Japan (3), Turkey (3) and China (2) [21]. No founder variants have been described, and the sources consulted provide no incidence data for Portugal or Brazil.

9.3. Special populations

In adults, information is scarce. Life expectancy does not appear to be reduced, except when there is onco-haematological disease [11]. Adults with Weaver syndrome may have a very mild phenotype and be identified only through the diagnosis of a child [1, 30]. In the UK series, the germline EZH2 variants also described in myeloid neoplasms led the authors to suggest looking for undiagnosed Weaver syndrome in patients with haematological neoplasms and an EZH2 variant in the tumour, by analysing DNA from another tissue, such as buccal mucosa [23]. Patients with an NF1 microdeletion including SUZ12 form a distinct group, with a higher risk of MPNST than carriers of intragenic NF1 variants [7, 10].

9.4. European network and patient organisations

The relevant European Reference Network is ERN ITHACA, for rare congenital malformations and rare intellectual disability [28]. In France, follow-up is coordinated by the centres of the AnDDI-Rares, DéfiScience and OSCAR networks [11]. Organisations and resources: Child Growth Foundation, in the United Kingdom [1, 74], MAGIC Foundation, in the United States [1, 75], Weaver Syndrome Network, in the United States (cited in the PNDS) [11], and the Overgrowth Syndromes Alliance, which brings together organisations for overgrowth–ID syndromes [72]. The French protocol notes that, at the time of its publication, there was no specific organisation in France [11]. The sources consulted do not identify specific organisations in Portugal or Brazil.

10. Illustrative Clinical Case

The case was published by Ünsal and colleagues in 2025 and describes 13 years of follow-up of a Turkish girl with Weaver syndrome [53].

The patient was the third child of non-consanguineous parents and was born at 37 weeks weighing 2600 g (−0.8 SD) and measuring 54 cm (+2.4 SD), that is, with increased length and normal weight. At 20 months she was referred to clinic for tall stature. She weighed 15.2 kg (+2.5 SD) and measured 95 cm (+3.1 SD), with normal body proportions. The target height calculated from the parents (mother 160 cm, father 165 cm) was 156 cm (−1.1 SD). On examination she had hypertelorism, frontal bossing and camptodactyly, without joint hyperextension, laxity, scoliosis or flat feet [53].

Endocrine investigations were normal: blood glucose 89 mg/dL, insulin 2 mIU/L, basal growth hormone 0.21 ng/mL, IGF-1 47.6 ng/mL (−0.6 SD), TSH 1.32 mIU/L, FT4 1.4 ng/dL, together with normal corticotrophin, cortisol, prolactin and pituitary MRI. She had hypermetropia and normal audiometry. The karyotype (46,XX) and FBN1 analysis were normal [53]. During follow-up she grew 11 cm in the first year and then 6.7 cm per year, parallel to the 97th centile. At 8 years, the echocardiogram showed mitral valve prolapse. Because of rapid growth of the hands and feet with height above +3 SD, she underwent an oral glucose tolerance test, with normal suppression of growth hormone (0.15 ng/mL at 120 minutes) [53].

At 14 years she was referred to genetics for tall stature, mild dysmorphism and mild ID. She weighed 62.9 kg (+1.9 SD), measured 177.2 cm (+3.1 SD), had an HC of 57 cm (90th–97th centiles) and hands (22 cm) and feet (30 cm) above the 97th centile. She had a broad forehead, mild hypertelorism, long philtrum, thin upper lip, prominent chin dimple and bilateral deviation of the fifth finger [53]. Exome sequencing with CNV analysis identified in exon 20 of EZH2 the heterozygous variant NM_004456.5:c.2213C>G p.(Ala738Gly), classified as likely pathogenic, never previously described in patients with Weaver syndrome and confirmed by Sanger sequencing as de novo [9, 53]. At 15 years she measured 179.6 cm (+3.1 SD) and weighed 64.9 kg, was at Tanner breast stage 3 and had not yet reached menarche. The authors planned clinical and radiological follow-up for neoplasms [53].

Figure 1 of the article shows the patient at 8, 15 and 30 months, and at 10 and 14.5 years. It also includes a hand radiograph at 13.5 years, with a bone age of 13 years, the clinodactyly of the fifth finger and a radiograph showing scoliosis (https://pmc.ncbi.nlm.nih.gov/articles/PMC11961133/figure/F1/). Figure 2 shows the growth chart up to 15 years (https://pmc.ncbi.nlm.nih.gov/articles/PMC11961133/figure/F2/) [53].

The case brings together several typical features of the syndrome: increased length at birth with normal weight, stable tall stature on a high centile without acceleration of growth velocity, height well above the parental target, subtle dysmorphism and mild ID [11, 23]. It also shows the limitations of a late diagnosis. Repeated endocrine investigation, including the glucose tolerance test, is not the recommended approach when the picture suggests a syndrome, and the French protocol proposes minimal laboratory testing followed by genetic investigation [11]. Some observations in the case diverge from the literature. At 13.5 years, bone age was not advanced, unlike what is described in most series [1, 23]. The text of the article describes camptodactyly of the fifth finger, but the figure legend calls it clinodactyly. The initial examination showed no scoliosis, but one of the figures shows it. The article also attributes the cause of the syndrome to somatic variants, whereas the variants are germline, and cites a lifetime cancer risk of 11% that does not correspond to the primary sources consulted, in which 11% is the penetrance estimated by the AACR for neuroblastoma [1, 27]. Tumour surveillance, begun at 15 years, fell outside the neuroblastoma risk window considered by the AACR [27].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

In Weaver syndrome, life expectancy does not appear to be reduced, except when there is an onco-haematological neoplasm [11]. Tall stature persists into adulthood in about 91% of patients, with cited final heights of 194.2 cm in men and 176.3 cm in women [11, 30]. ID is mild in about half of patients, and most of those with mild ID attend mainstream school with support; independent living in adulthood and, often, starting a family are expected [1, 11]. Abnormal tone, camptodactyly and the facies tend to improve with age [1]. The absolute risk of neuroblastoma is estimated at between 5% and 11%, depending on the source, and is concentrated in the first years of life (the tumours described arose between the prenatal period and 4 years). Haematological neoplasms are rare and it is not known whether their frequency is increased [1, 11, 27]. There are no data on cancer risk in adults [11]. In Cohen-Gibson syndrome, ID is present in all cases, motor difficulties may persist into adulthood and the cervical spine is the main cause of serious morbidity described [4]. In Imagawa-Matsumoto syndrome, the prognosis is highly variable, from mild forms without overgrowth to presentations with brain malformations or cervical stenosis [7, 21, 26].

11.2. Surveillance table

In older children and adolescents without complications, visits may be spaced further apart [1, 67]. In Weaver syndrome, the choice between the AACR neuroblastoma surveillance protocol and the clinical vigilance of the French protocol should be made explicit and recorded with the family (section 7.5) [11, 27].

12. Summary and Concluding Remarks

Weaver, Cohen-Gibson and Imagawa-Matsumoto syndromes form the PRC2-related overgrowth spectrum, PRC2 being the complex that keeps developmental genes silenced through H3K27 methylation. Germline EZH2, EED and SUZ12 variants reduce the activity of the complex, through partial loss of function and, for some EZH2 variants, through a dominant-negative effect; EZH2 null alleles do not appear to cause the syndrome [1, 44, 47]. Weaver syndrome is the best known, with at least 70 patients described. Tall stature, present in about 90%, is the most constant sign, ID is usually mild and bone age is advanced, with the carpus further advanced than the phalanges. The facies, with hypertelorism, retrognathia and a "stuck-on" chin, is best recognised between 1 and 3 years. Camptodactyly, soft skin, hoarse cry and umbilical hernia in the neonatal period are clues that tend to disappear [1, 23]. Cohen-Gibson syndrome has constant ID and more cervical and cardiac anomalies, and Imagawa-Matsumoto syndrome has more variable expression [4, 21, 24].

The diagnosis is molecular, by overgrowth–ID panel or exome. The PRC2 methylation episignature helps to classify variants of uncertain significance and to detect mosaicism, and the overlap with Sotos syndrome justifies studying the genes together [23, 25]. There is no targeted treatment and there are no clinical trials. Management is based on the initial assessment (including echocardiogram and renal ultrasound), developmental support, orthopaedic treatment and attention to the cervical spine [1, 4, 11]. Neuroblastoma is the only tumour with a demonstrated increased risk in Weaver syndrome. Since 2024 the AACR has recommended surveillance up to age 10 years, whereas the 2025 French protocol confines itself to clinical vigilance, and this divergence should be discussed with families [11, 27]. EZH2 and PRC2 inhibitors used in oncology require caution in these patients [1, 11]. Inheritance is autosomal dominant, with a 50% risk to offspring and about 1% to siblings when the parents are not carriers [1, 11]. Gaps remain regarding prevalence, penetrance, cancer risk in adults and the benefit of tumour surveillance.

13. References

1. Ocansey S, Tatton-Brown K. EZH2-Related Overgrowth. 2013 Jul 18 [updated 2024 Mar 21; revised 2025 Jun 26]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK148820/
2. 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: Oxford University Press; 2018. p. 824–825.
3. Orphanet. Weaver syndrome (ORPHA:3447) [Internet]. Paris: INSERM; [cited 2026 Oct 29]. Available from: https://www.orpha.net/en/disease/detail/3447
4. Cohen ASA, Gibson WT. EED-Related Overgrowth. 2019 Apr 11 [updated 2025 May 8]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK540017/
5. Orphanet. Cohen-Gibson syndrome (ORPHA:659396) [Internet]. Paris: INSERM; [cited 2026 Oct 29]. Available from: https://www.orpha.net/en/disease/detail/659396
6. Orphanet. Imagawa-Matsumoto syndrome (ORPHA:659463) [Internet]. Paris: INSERM; [cited 2026 Oct 29]. Available from: https://www.orpha.net/en/disease/detail/659463
7. Imagawa E, Seyama R, Aoi H, Uchiyama Y, Marcarini BG, Furquim I, et al. Imagawa-Matsumoto syndrome: SUZ12-related overgrowth disorder. Clin Genet. 2023;103(4):383-391. doi:10.1111/cge.14296. PMID: 36645289.
8. Orphanet. PRC-2 complex-related overgrowth spectrum (ORPHA:659387) [Internet]. Paris: INSERM; [cited 2026 Oct 29]. Available from: https://www.orpha.net/en/disease/detail/659387
9. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-68. doi:10.1002/humu.23348. PMID: 28967166.
10. GeneGuidance. Gene pages for EZH2, EED and SUZ12 and condition pages for Weaver syndrome and Cohen-Gibson syndrome [accessed 2026 Sep 26].
11. Centre de Référence Anomalies du Développement et Syndromes Malformatifs (CHU de Caen), Filière AnDDI-Rares; Vincent-Devulder A, Garin T, coordinators. Protocole National de Diagnostic et de Soins (PNDS): Syndrome de Weaver [Internet]. Saint-Denis: Haute Autorité de Santé; 2025 Dec [cited 2026 Oct 29]. Available from: https://www.has-sante.fr/upload/docs/application/pdf/2025-12/dir2/texte_du_pnds_syndrome_de_weaver.pdf
12. Song F, Michels M, Betz B, Vorhoff W, Neuhann T, Senninger L, et al. Mosaicism for Cohen-Gibson Syndrome in a Female Individual With Overgrowth and Neurodevelopmental Delay. Mol Genet Genomic Med. 2026;14(10):e70286. doi:10.1002/mgg3.70286. PMID: 42830684. PMCID: PMC13635803.
13. Cyrus SS, Cohen ASA, Agbahovbe R, Avela K, Yeung KS, Chung BHY, et al. Rare SUZ12 variants commonly cause an overgrowth phenotype. Am J Med Genet C Semin Med Genet. 2019;181(4):532-547. doi:10.1002/ajmg.c.31748. PMID: 31736240.
14. Imagawa E, Higashimoto K, Sakai Y, Numakura C, Okamoto N, Matsunaga S, et al. Mutations in genes encoding polycomb repressive complex 2 subunits cause Weaver syndrome. Hum Mutat. 2017;38(6):637-648. doi:10.1002/humu.23200. PMID: 28229514.
15. Online Mendelian Inheritance in Man, OMIM®. Johns Hopkins University, Baltimore, MD. MIM Number: 277590 — Weaver syndrome; WVS [Internet]. [cited 2026 Oct 29]. Available from: https://omim.org/entry/277590
16. Online Mendelian Inheritance in Man, OMIM®. Johns Hopkins University, Baltimore, MD. MIM Number: 601573 — Enhancer of zeste 2 polycomb repressive complex 2 subunit; EZH2 [Internet]. [cited 2026 Oct 29]. Available from: https://omim.org/entry/601573
17. Online Mendelian Inheritance in Man, OMIM®. Johns Hopkins University, Baltimore, MD. MIM Number: 617561 — Cohen-Gibson syndrome; COGIS [Internet]. [cited 2026 Oct 29]. Available from: https://omim.org/entry/617561
18. Online Mendelian Inheritance in Man, OMIM®. Johns Hopkins University, Baltimore, MD. MIM Number: 605984 — Embryonic ectoderm development; EED [Internet]. [cited 2026 Oct 29]. Available from: https://omim.org/entry/605984
19. Online Mendelian Inheritance in Man, OMIM®. Johns Hopkins University, Baltimore, MD. MIM Number: 618786 — Imagawa-Matsumoto syndrome; IMMAS [Internet]. [cited 2026 Oct 29]. Available from: https://omim.org/entry/618786
20. Online Mendelian Inheritance in Man, OMIM®. Johns Hopkins University, Baltimore, MD. MIM Number: 606245 — SUZ12 polycomb repressive complex 2 subunit; SUZ12 [Internet]. [cited 2026 Oct 29]. Available from: https://omim.org/entry/606245
21. Pang Z, Zhang T, Wang D, Xu P, Wu B, Shen Q, et al. Case Report: First report of spinal stenosis in Imagawa-Matsumoto syndrome: a novel SUZ12 variant in an 11-year-old Chinese child. Front Genet. 2026;17:1867197. doi:10.3389/fgene.2026.1867197. PMID: 42523552. PMCID: PMC13406684.
22. Xu X, Xu H, Huang S, He Y, Zhang L. Identification and functional characterization of a SUZ12 frameshift variant in a Chinese family with Imagawa-Matsumoto syndrome. BMC Med Genomics. 2026;19(1). doi:10.1186/s12920-026-02431-5. PMID: 42595994. PMCID: PMC13471259.
23. Tatton-Brown K, Murray A, Hanks S, Douglas J, Armstrong R, Banka S, et al. Weaver syndrome and EZH2 mutations: Clarifying the clinical phenotype. Am J Med Genet A. 2013;161A(12):2972-80. doi:10.1002/ajmg.a.36229. PMID: 24214728.
24. Griffiths S, Loveday C, Zachariou A, Behan LA, Chandler K, Cole T, et al. EED and EZH2 constitutive variants: A study to expand the Cohen-Gibson syndrome phenotype and contrast it with Weaver syndrome. Am J Med Genet A. 2019;179(4):588-594. doi:10.1002/ajmg.a.61066. PMID: 30793471.
25. Choufani S, Gibson WT, Turinsky AL, Chung BHY, Wang T, Garg K, et al. DNA Methylation Signature for EZH2 Functionally Classifies Sequence Variants in Three PRC2 Complex Genes. Am J Hum Genet. 2020;106(5):596-610. doi:10.1016/j.ajhg.2020.03.008. PMID: 32243864. PMCID: PMC7212265.
26. Yücel Z, Yüksel EB, Koç A. Imagawa-Matsumoto Syndrome: The First Case From Turkey. Noro Psikiyatr Ars. 2024;61(3):289-292. doi:10.29399/npa.28400. PMID: 39258127. PMCID: PMC11382558.
27. Kamihara J, Diller LR, Foulkes WD, Michaeli O, Nakano Y, Pajtler KW, et al. Neuroblastoma Predisposition and Surveillance-An Update from the 2023 AACR Childhood Cancer Predisposition Workshop. Clin Cancer Res. 2024;30(15):3137-3143. doi:10.1158/1078-0432.CCR-24-0237. PMID: 38860978.
28. European Reference Network on Rare Congenital Malformations and Rare Intellectual Disability (ERN ITHACA) [Internet]. [cited 2026 Oct 29]. Available from: https://ern-ithaca.eu/
29. Weaver DD, Graham CB, Thomas IT, Smith DW. A new overgrowth syndrome with accelerated skeletal maturation, unusual facies, and camptodactyly. J Pediatr. 1974;84(4):547-52. doi:10.1016/s0022-3476(74)80675-x. PMID: 4366187.
30. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 210–213.
31. Douglas J, Hanks S, Temple IK, Davies S, Murray A, Upadhyaya M, et al. NSD1 mutations are the major cause of Sotos syndrome and occur in some cases of Weaver syndrome but are rare in other overgrowth phenotypes. Am J Hum Genet. 2003;72(1):132-43. doi:10.1086/345647. PMID: 12464997. PMCID: PMC378618.
32. Tatton-Brown K, Douglas J, Coleman K, Baujat G, Cole TR, Das S, et al. Genotype-phenotype associations in Sotos syndrome: an analysis of 266 individuals with NSD1 aberrations. Am J Hum Genet. 2005;77(2):193-204. doi:10.1086/432082. PMID: 15942875. PMCID: PMC1224542.
33. Gibson WT, Hood RL, Zhan SH, Bulman DE, Fejes AP, Moore R, et al. Mutations in EZH2 cause Weaver syndrome. Am J Hum Genet. 2012;90(1):110-8. doi:10.1016/j.ajhg.2011.11.018. PMID: 22177091. PMCID: PMC3257956.
34. Tatton-Brown K, Hanks S, Ruark E, Zachariou A, Duarte Sdel V, Ramsay E, et al. Germline mutations in the oncogene EZH2 cause Weaver syndrome and increased human height. Oncotarget. 2011;2(12):1127-33. doi:10.18632/oncotarget.385. PMID: 22190405. PMCID: PMC3282071.
35. Cohen AS, Tuysuz B, Shen Y, Bhalla SK, Jones SJ, Gibson WT. A novel mutation in EED associated with overgrowth. J Hum Genet. 2015;60(6):339-42. doi:10.1038/jhg.2015.26. PMID: 25787343.
36. Imagawa E, Albuquerque EVA, Isidor B, Mitsuhashi S, Mizuguchi T, Miyatake S, et al. Novel SUZ12 mutations in Weaver-like syndrome. Clin Genet. 2018;94(5):461-466. doi:10.1111/cge.13415. PMID: 30019515.
37. Orphanet. Weaver-Williams syndrome (ORPHA:3448) [Internet]. Paris: INSERM; [cited 2026 Oct 29]. Available from: https://www.orpha.net/en/disease/detail/3448
38. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 272–275.
39. Tatton-Brown K, Loveday C, Yost S, Clarke M, Ramsay E, Zachariou A, et al. Mutations in Epigenetic Regulation Genes Are a Major Cause of Overgrowth with Intellectual Disability. Am J Hum Genet. 2017;100(5):725-736. doi:10.1016/j.ajhg.2017.03.010. PMID: 28475857. PMCID: PMC5420355.
40. Cao R, Wang L, Wang H, Xia L, Erdjument-Bromage H, Tempst P, et al. Role of histone H3 lysine 27 methylation in Polycomb-group silencing. Science. 2002;298(5595):1039-43. doi:10.1126/science.1076997. PMID: 12351676.
41. Cyrus S, Burkardt D, Weaver DD, Gibson WT. PRC2-complex related dysfunction in overgrowth syndromes: A review of EZH2, EED, and SUZ12 and their syndromic phenotypes. Am J Med Genet C Semin Med Genet. 2019;181(4):519-531. doi:10.1002/ajmg.c.31754. PMID: 31724824.
42. Margueron R, Justin N, Ohno K, Sharpe ML, Son J, Drury WJ 3rd, et al. Role of the polycomb protein EED in the propagation of repressive histone marks. Nature. 2009;461(7265):762-7. doi:10.1038/nature08398. PMID: 19767730. PMCID: PMC3772642.
43. Cohen AS, Yap DB, Lewis ME, Chijiwa C, Ramos-Arroyo MA, Tkachenko N, et al. Weaver Syndrome-Associated EZH2 Protein Variants Show Impaired Histone Methyltransferase Function In Vitro. Hum Mutat. 2016;37(3):301-7. doi:10.1002/humu.22946. PMID: 26694085. PMCID: PMC4832389.
44. Lui JC, Barnes KM, Dong L, Yue S, Graber E, Rapaport R, et al. Ezh2 Mutations Found in the Weaver Overgrowth Syndrome Cause a Partial Loss of H3K27 Histone Methyltransferase Activity. J Clin Endocrinol Metab. 2018;103(4):1470-1478. doi:10.1210/jc.2017-01948. PMID: 29244146. PMCID: PMC6276576.
45. Gibson WT, Lengyell TC, Korecki AJ, Janssen SM, Adair BA, Gamu D, et al. Minimally Humanized Ezh2 Exon-18 Mouse Cell Lines Validate Preclinical CRISPR/Cas9 Approach to Treat Weaver Syndrome. Hum Gene Ther. 2025;36(5-6):618-627. doi:10.1089/hum.2024.170. PMID: 39964768.
46. Suri T, Dixit A. The phenotype of EZH2 haploinsufficiency-1.2-Mb deletion at 7q36.1 in a child with tall stature and intellectual disability. Am J Med Genet A. 2017;173(10):2731-2735. doi:10.1002/ajmg.a.38356. PMID: 28696078.
47. Deevy O, Li J, Monger C, Matrà F, Tuck E, Davies M, et al. Dominant-negative effects of Weaver syndrome-associated EZH2 variants. Genes Dev. 2025;39(21-22):1355-1376. doi:10.1101/gad.351884.124. PMID: 40846643. PMCID: PMC12581829.
48. Cyrus SS, Medina-Giró S, Lian T, Allan DW, Gibson WT. Functional analysis of human EED variants using Drosophila. Genetics. 2025;231(1). doi:10.1093/genetics/iyaf120. PMID: 40539649. PMCID: PMC12405997.
49. Usemann J, Ernst T, Schäfer V, Lehmberg K, Seeger K. EZH2 mutation in an adolescent with Weaver syndrome developing acute myeloid leukemia and secondary hemophagocytic lymphohistiocytosis. Am J Med Genet A. 2016;170A(5):1274-7. doi:10.1002/ajmg.a.37562. PMID: 26762561.
50. Lee CL, Chuang CK, Chiu HC, Chang YH, Tu YR, Lo YT, et al. From Overgrowth to Complex Malformations: A Novel EZH2 Variant Reveals the Expanding Clinical Spectrum of Weaver Syndrome. Children (Basel). 2025;12(11). doi:10.3390/children12111487. PMID: 41300605. PMCID: PMC12650869.
51. Kendir-Demirkol Y, Yeter B, Jenny LA. Expanding the Phenotypic and Genotypic Spectrum of Weaver Syndrome: A Missense Variant of the EZH2 Gene. Mol Syndromol. 2024;15(2):161-166. doi:10.1159/000533733. PMID: 38585548. PMCID: PMC10996336.
52. Ren L, Jiang L, Jiang X, Wang H, Li Y, Ren X. Clinical and genetic characterization of Weaver syndrome: A case report of an EZH2 mutation and review of the literature. Medicine (Baltimore). 2025;104(36):e44080. doi:10.1097/MD.0000000000044080. PMID: 40922349. PMCID: PMC12419298.
53. Ünsal Y, Yıldırım N, Buluş AD, Kılıç E. A New Case with Weaver Syndrome: Delineating Natural Course and Growth Pattern, Further Clarifying Clinical Phenotype. Mol Syndromol. 2025;16(2):187-193. doi:10.1159/000541476. PMID: 40176837. PMCID: PMC11961133.
54. ClinGen. Gene-disease validity: EZH2 – Weaver syndrome (MONDO:0010193), autosomal dominant, Definitive [Internet]. 2022 Aug 3. Available from: https://search.clinicalgenome.org/kb/gene-validity/CGGV:assertion_5dcf5941-6076-4486-94be-7ba3944abaee-2022-08-03T160000.000Z
55. Villani A, Greer MLC, Kalish JM, Nakagawara A, Nathanson KL, Pajtler KW, et al. Recommendations for cancer surveillance in individuals with RASopathies and other rare genetic conditions with increased cancer risk. Clin Cancer Res. 2017;23(12):e83-e90. doi:10.1158/1078-0432.CCR-17-0631. PMID: 28620009.
56. Lee W, Teckie S, Wiesner T, Ran L, Prieto Granada CN, Lin M, et al. PRC2 is recurrently inactivated through EED or SUZ12 loss in malignant peripheral nerve sheath tumors. Nat Genet. 2014;46(11):1227-32. doi:10.1038/ng.3095. PMID: 25240281. PMCID: PMC4249650.
57. Cohen AS, Gibson WT. EED-associated overgrowth in a second male patient. J Hum Genet. 2016;61(9):831-4. doi:10.1038/jhg.2016.51. PMID: 27193220.
58. Cooney E, Bi W, Schlesinger AE, Vinson S, Potocki L. Novel EED mutation in patient with Weaver syndrome. Am J Med Genet A. 2017;173(2):541-545. doi:10.1002/ajmg.a.38055. PMID: 27868325.
59. Smigiel R, Biernacka A, Biela M, Murcia-Pienkowski V, Szmida E, Gasperowicz P, et al. Novel de novo mutation affecting two adjacent aminoacids in the EED gene in a patient with Weaver syndrome. J Hum Genet. 2018;63(4):517-520. doi:10.1038/s10038-017-0391-x. PMID: 29410511.
60. Spellicy CJ, Peng Y, Olewiler L, Cathey SS, Rogers RC, Bartholomew D, et al. Three additional patients with EED-associated overgrowth: potential mutation hotspots identified?. J Hum Genet. 2019;64(6):561-572. doi:10.1038/s10038-019-0585-5. PMID: 30858506.
61. Hetzelt KLML, Winterholler M, Kerling F, Rauch C, Ekici AB, Winterpacht A, et al. Manifestation of epilepsy in a patient with EED-related overgrowth (Cohen-Gibson syndrome). Am J Med Genet A. 2022;188(1):292-297. doi:10.1002/ajmg.a.62496. PMID: 34533271.
62. Goel H, O'Donnell S, Edwards M. EED related overgrowth: First report of multiple members in a single family. Am J Med Genet A. 2024;194(2):374-382. doi:10.1002/ajmg.a.63438. PMID: 37840385.
63. Guijarro-Llorca J, García-Payá E, Blanes-Martínez MM, Bañuls-Roca J. Keloids as a clinical manifestation of a pathogenic SUZ12 mutation and Polycomb Repressive Complex 2 dysfunction in Imagawa-Matsumoto syndrome. Actas Dermosifiliogr. 2026;:104728. doi:10.1016/j.ad.2026.104728. PMID: 42749126.
64. Kim GJ, Vasco de Albuquerque Albuquerque E, Rezende RC, De Polli Cellin L, Santillan Vasconez AM, Krepischi ACV, et al. Exome sequencing of patients with syndromic tall stature reveals four novel candidate genes. Endocr Connect. 2025;14(7). doi:10.1530/EC-25-0137. PMID: 40577202. PMCID: PMC12268985.
65. Simsek O, Vossough A. Fetal and postnatal neuroimaging of SUZ12-related overgrowth: Imagawa-matsumoto syndrome. J Neuroradiol. 2024;51(5):101210. doi:10.1016/j.neurad.2024.101210. PMID: 38850627.
66. 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: Oxford University Press; 2018. p. 815–816.
67. Medizinische Hochschule Hannover, CPS-Register und Ambulanz, Klinik für Pädiatrische Hämatologie und Onkologie. Weaver syndrome – compendium [Internet]. Hannover: MHH; [undated; cited 2026 Oct 29]. Available from: https://www.krebs-praedisposition.de/en/kompendium/weaver-syndrome/
68. U.S. Food and Drug Administration. TAZVERIK (tazemetostat) tablets: prescribing information [Internet]. Epizyme; NDA 211723; initial approval 2020 Jan 23; label revision 2026 Apr 29 [cited 2026 Oct 29]. Available from: https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=211723
69. European Medicines Agency. Medicines output — medicines report [Internet]. Amsterdam: EMA; [downloaded 2026 Oct 29]. Available from: https://www.ema.europa.eu/en/documents/report/medicines-output-medicines_json-report_en.json
70. ClinicalTrials.gov [Internet]. Bethesda (MD): National Library of Medicine; search "Weaver syndrome", "Cohen-Gibson", "Imagawa-Matsumoto", "EZH2 overgrowth", "Overgrowth Syndrome" [cited 2026 Oct 29]. Available from: https://clinicaltrials.gov/
71. Brodeur GM, Diller LR, Nichols KE, Plon SE, Porter CC, Malkin D. Pediatric Cancer Predisposition and Surveillance Update: Summary Perspective and Future Directions. Clin Cancer Res. 2025;31(13):2581-2588. doi:10.1158/1078-0432.CCR-25-0423. PMID: 40327330. PMCID: PMC12213195.
72. Atterton C, Trew I, Cale JM, Aung-Htut MT, Grens K, Kiernan J, et al. Overgrowth-intellectual disability disorders: progress in biology, patient advocacy and innovative therapies. Dis Model Mech. 2025;18(5). doi:10.1242/dmm.052300. PMID: 40353642. PMCID: PMC12091874.
73. Rahbari R, Wuster A, Lindsay SJ, Hardwick RJ, Alexandrov LB, Turki SA, et al. Timing, rates and spectra of human germline mutation. Nat Genet. 2016;48(2):126-133. doi:10.1038/ng.3469. PMID: 26656846. PMCID: PMC4731925.
74. Child Growth Foundation [Internet]. United Kingdom; [cited 2026 Oct 29]. Available from: https://childgrowthfoundation.org/
75. MAGIC Foundation [Internet]. USA; [cited 2026 Oct 29]. Available from: https://www.magicfoundation.org/