Nicolaides-Baraitser Syndrome (SMARCA2)

SMARCA2-related Nicolaides-Baraitser syndrome and the allelic blepharophimosis-intellectual disability syndrome

Literature and regulatory information updated: 27 October 2026; integration of the French national protocol (PNDS): 6 November 2026.

0. Initial Reference Table

OMIM, Online Mendelian Inheritance in Man; ORPHA, Orphanet code; ICD-10, International Classification of Diseases, 10th revision; MONDO, Mondo Disease Ontology; GARD, Genetic and Rare Diseases Information Center; HGNC, HUGO Gene Nomenclature Committee; MANE, Matched Annotation from NCBI and EMBL-EBI; GRCh38, human genome reference assembly, version 38.

1. Introduction

Definition and framework

Nicolaides-Baraitser syndrome (NCBRS) is a developmental disorder with multiple congenital anomalies, of autosomal dominant transmission and almost always caused by a de novo variant of the SMARCA2 gene, which encodes one of the two alternative catalytic ATPases (enzymes that hydrolyse adenosine triphosphate, ATP) of the BAF (BRG1-associated factors) chromatin remodelling complex, the human equivalent of the yeast SWI/SNF (switch/sucrose non-fermentable) complex [1, 18]. The picture combines intellectual disability (ID), usually severe and with marked impairment of language, sparse hair, prominent interphalangeal joints, brachydactyly, a face that becomes coarse with age, acquired microcephaly, short stature and epilepsy [1, 19]. Together with Coffin-Siris syndrome, it is the most recognisable form of the BAFopathy group, and the chromatin remodelling genes that cause them also produce less specific pictures of non-syndromic intellectual disability [19]. Coffin-Siris syndrome and the other BAFopathies enter the differential. The 2021 French national protocol treats the two syndromes together, under the designation BAFopathies, because they share clinical features and molecular basis; it describes NCBRS as much rarer than Coffin-Siris syndrome and caused only by missense variants of the helicase domain of SMARCA2, and attributes a Coffin-Siris phenotype to duplications of the gene [15]. The statement that only missense variants cause NCBRS diverges from GeneReviews, which attributes about 3% of pathogenic variants to in-frame intragenic deletions (section 5) [1, 15].

Historical overview

In 1993, Nicolaides and Baraitser, of the Hospital for Sick Children in London, described a 16-year-old girl with an undiagnosed syndrome of mental retardation, sparse hair, a prominent lower lip and brachydactyly with cone-shaped epiphyses [20]. In the original description, the eyebrows and eyelashes were normal and the interphalangeal joints appeared swollen [3]. Re-evaluated by Sousa and colleagues at 32 years, the same patient had had recurrent refractory seizures since the age of 3, had lost all language in early adulthood, walked only short distances, with a wide base and flexed hips and knees, used a wheelchair most of the time and had short stature, obesity, complete alopecia, wrinkled skin on the face, neck and extremities, eczema and an aged appearance, while remaining sociable [3]. In 1996, Krajewska-Walasek and colleagues described a 19-year-old boy with a similar picture and cryptorchidism; in both patients, the microscopic structure of the hair was normal, and the authors proposed that it was a distinct syndrome [3]. The original patient died at 33 years, in status epilepticus, after a deterioration in motor function and language [11, 21]. In 2003, Morin and colleagues reported two unrelated patients with short stature, hypotrichosis, brachydactyly with cone-shaped epiphyses, epilepsy and severe mental retardation, confirming its status as a distinct entity, and proposed the eponym after the authors of the initial report [1, 22]. In 2009, Sousa and colleagues delineated the phenotype in 23 patients (18 new and the follow-up of 5 already described), all sporadic, with no sex difference, and identified Coffin-Siris syndrome as the main differential diagnosis and the progressive nature of some features [23]. Between the original description and 2008 only five more cases had been published (Krajewska-Walasek 1996, Morin 2003, Witters and Fryns 2003, Castori 2008), and the patients and families of the 2009 delineation founded an international parent support group that has met annually since 2010 [18].

Milestones of molecular discovery

In 2012, Van Houdt and colleagues sequenced the exome of 10 individuals with NCBRS and found heterozygous SMARCA2 variants in 8; extended screening identified non-synonymous variants in 36 of 44 patients, de novo whenever parental samples were available, clustered in ultraconserved motifs of the catalytic ATPase region [24]. The 44 patients (22 of the 26 then published and 22 new) had been divided into 37 with a high-certainty diagnosis and 7 with low certainty, and variants were found in 34 of 37 and in 2 of 7, respectively; in 2 of the 3 patients of the first group without a variant (that of Krajewska-Walasek and patient 18 of the 2009 series) a pathogenic ARID1B variant was later identified, and both were reclassified as Coffin-Siris syndrome [18]. Simultaneously, Wolff and colleagues described a 32-kb in-frame deletion of the C-terminal helicase domain, detected by single nucleotide polymorphism array, and two de novo missense variants in the same domain [25]. In 2014, the international consortium (Sousa and Hennekam) analysed 61 molecularly confirmed cases (the 47 then published and 14 unpublished, through a questionnaire to the attending physicians and review of photographs) and concluded that NCBRS is clinically and genetically homogeneous, with non-truncating variants in the ATPase region and a probable dominant-negative effect [18]. In 2015, Mari and colleagues screened the six BAF complex genes then known in 11 patients with a phenotype compatible with NCBRS or Coffin-Siris syndrome and found heterozygous SMARCA2 missense variants in 8 unrelated patients with NCBRS [3, 26]. There followed the DNA methylation signatures (episignatures) shared by the BAFopathies (2018) and specific to NCBRS (2019) [27, 28], the demonstration in human embryonic stem cells that NCBRS variants reprogramme neural enhancers (2019) [29] and, in 2020, the delineation of a second syndrome caused by SMARCA2 variants located outside the helicase domains [10]. This syndrome was delineated by reverse phenotyping: the 20 individuals studied had SMARCA2 variants found by panel, exome or genome sequencing without prior clinical suspicion of NCBRS, were brought together through Matchmaker Exchange and ERN ITHACA, and 14 shared a recognisable face [10].

Nomenclature and reclassifications

GeneReviews uses the designation SMARCA2-related Nicolaides-Baraitser syndrome (SMARCA2-NCBRS), and Orphanet records the synonym Intellectual disability-sparse hair-brachydactyly syndrome [1, 2]. Several individuals with SMARCA2 variants were initially classified as Coffin-Siris syndrome, and re-evaluation showed a phenotype more consistent with NCBRS; GeneReviews on Coffin-Siris syndrome therefore includes SMARCA2 among the 14 genes of the syndrome, with about 2% of cases [19, 30]. Conversely, patients with a provisional diagnosis of NCBRS were reclassified as Coffin-Siris syndrome after molecular study [21, 31]. The GeneGuidance database associates NCBRS with ORPHA:637013 and code Q87.0 of the International Classification of Diseases, 10th revision (ICD-10); Orphanet data (Orphadata) show that this code corresponds to the blepharophimosis syndrome, with exact mapping to OMIM 619293, and that NCBRS is ORPHA:3051, with exact mapping to OMIM 601358 and ICD-10 Q87.1 [2, 4, 16]. OMIM records as alternative titles of NCBRS Sparse hair-impaired intellectual development syndrome and the abbreviation NBS, and notes that NBS also designates Nijmegen breakage syndrome (OMIM 251260), a distinct entity [3]. For the allelic entity, OMIM adopted the title Blepharophimosis-impaired intellectual development syndrome (BIS), instead of the name blepharophimosis intellectual disability syndrome proposed by Cappuccio and colleagues [6].

Entities within scope

NCBRS (OMIM 601358, ORPHA:3051) is defined by the association of suggestive findings with a heterozygous pathogenic SMARCA2 variant, typically non-truncating and located in the ATPase/helicase domain [1]. SMARCA2-related blepharophimosis-intellectual disability syndrome (BIS; OMIM 619293, ORPHA:637013) is an allelic entity defined by de novo variants clustered outside the helicase region, with global developmental delay, hypotonia, absent or severely impaired speech and a face with blepharophimosis, epicanthic folds and a drooping nasal tip [4, 10]. Also within scope, as ill-defined SMARCA2 phenotypes, are the neurological picture without the typical face or limbs associated with a QLQ domain variant (Pan 2021), the non-specific phenotypes of 6 of the 20 individuals in the Cappuccio series and the 9p24.3 duplication including SMARCA2 with a Coffin-Siris-like phenotype [10, 32, 33]. Somatic SMARCA2 variants in sporadic tumours are not hereditary and do not belong to this picture [1].

2. Molecular Genetics and Pathophysiology

SMARCA2 gene and protein

The SMARCA2 gene is located at 9p24.3 (GRCh38 chr9:2015347-2193624), has 34 exons and encodes the protein probable global transcription activator SNF2L2, also called BRM, hBRM, BAF190 or SNF2L2, of 1590 amino acids [1, 8, 11, 16]. OMIM also records the synonyms SNF2-like 2 and Drosophila brahma homologue, a genomic extent of 178.2 kb (Koga 2009) and the GRCh38 coordinates 9:2015347-2193620 obtained from NCBI, which differ by 4 bp at the 3′ end from those given above; the localisation evolved from 9p24-p23 (isotopic in situ hybridisation, 1994) to 9p24.1 (fluorescence in situ hybridisation, FISH, 1998) and to the current 9p24.3 [9]. Yang and colleagues (2025) give 38 exons, a number that diverges from the 34 of GeneReviews and OMIM and that the article does not substantiate [1, 34]. The human protein is 56% identical and 72% similar to Drosophila brahma, and the mouse orthologue (Snf2l2) is located on chromosome 19 [9]. The MANE Select (Matched Annotation from NCBI and EMBL-EBI) reference transcript is NM_003070.5, with the protein NP_003061.3 (CCDS34977.1); the coding sequence runs from nucleotide 95 to 4867 of a 5748-nucleotide transcript, and the LRG_882 record remains pending [7]. GeneReviews mentions a transcript of 5879 base pairs, isoforms from alternative splicing and a length polymorphism of a CAG trinucleotide repeat [1]. The SNF2 family ATPase domain occupies exons 15 to 25 (amino acids 508 to 1305) and contains two functional regions, the ATP-binding region (N-terminal helicase) and the C-terminal helicase, with the seven canonical helicase-related motifs characteristic of the SNF2 group and 14 additional conserved blocks; outside this domain are the QLQ domain, where residue 185 is located, and the bromodomain encoded by exon 30 [11, 32, 35].

Function in the BAF complex

The SWI/SNF family of ATP-dependent chromatin remodellers is essential for the regulation of gene expression, differentiation and development; SMARCA2 belongs to the family of helicase-related proteins whose ATPase domain couples ATP hydrolysis to DNA binding, displacing or repositioning nucleosomes [1]. SMARCA2 (BRM) is one of the two mutually exclusive ATPases of the BAF complexes; the other is SMARCA4 (BRG1), and UniProt places SMARCA2 in the npBAF (neural progenitor) and nBAF (neuron-specific) complexes, whose subunit switching accompanies cell cycle exit during brain development [16, 18]. The two ATPases are not redundant: BRG1 binds zinc finger proteins through an N-terminal domain absent from BRM, whereas BRM interacts with ankyrin repeat proteins of the Notch signalling pathway, which directs them to different promoters; BRM also forms, with MECP2, a corepressor complex recruited to methylated promoters, including that of FMR1, and is phosphorylated and excluded from condensed chromosomes during mitosis [9].

Molecular mechanism: dominant-negative effect or gain of function

Several arguments exclude haploinsufficiency as the mechanism of NCBRS. ClinGen dosage sensitivity curation assigns the gene a score of 0 for both haploinsufficiency and triplosensitivity; patients with 9p deletions including SMARCA2 do not have NCBRS, and mice without functional Smarca2 (Brm) are viable and fertile, although the knockout model of Koga and colleagues (2009) showed deficits in social interaction and prepulse inhibition [1, 8, 9, 11]. NCBRS variants are predominantly non-truncating and concentrated in the SNF2 ATPase domain; they probably do not prevent assembly of the complex, but disrupt ATPase activity, which suggests a dominant-negative or gain-of-function effect: the mutant protein is incorporated into the complex that binds downstream targets and nucleosome remodelling does not occur normally [1, 18, 24]. In favour of this model, variants of the conserved motifs of the yeast SNF2 ATPase had dominant-negative activity in a functional assay, two of them identical to NCBRS variants and at least four others in the same motifs; in mammalian cells, a variant of the lysine of the ATP-binding site (motif I, the residue affected by p.(Lys755Arg)) retained nuclear localisation and binding to the other BAF subunits, but impaired complex function [18]. The authors propose that complexes with inactive but structurally intact SMARCA2 occupy genomic sites and competitively prevent access of normal complexes, and that they may also sequester subunits shared with SMARCA4, with which SMARCA2 has 74% sequence identity [18]. The pLI (probability of loss-of-function intolerance) of 1 presented by ClinGen from DECIPHER does not contradict this model, because NCBRS does not result from loss of function [8].

Tissue mechanism: reprogramming of neural enhancers

In human embryonic stem cells (hESC) carrying the heterozygous NCBRS variants p.(Lys755Arg) or p.(Arg1159Gln), Gao and colleagues observed phenotypically normal cells, but with markedly impaired differentiation into neural progenitors; chromatin analysis showed loss of SOX3-dependent neural enhancers, emergence of astrocyte-specific de novo enhancers, increased binding of mutant SMARCA2 and global retargeting of SMARCA4, with overexpression of FRA2 (AP-1 family) acting as a pioneer factor [29]. The genes of the NCBRS methylation signature are enriched in calcium signalling, synaptic function and autism-associated genes from the SFARI database (22 of 225; p<0.001), which is consistent with the neurological phenotype [28].

Inheritance pattern

NCBRS is autosomal dominant (AD); almost all cases result from de novo variants, and the rare inherited cases come from unaffected mosaic parents, including low-level paternal germline mosaicism and maternal mosaicism [1, 32, 36, 37]. There is no record of people with NCBRS having had offspring [1].

Variant spectrum

In the 2014 consortium series, of the 61 variants compiled, 59 were missense (58 in the ATPase domain) and 2 were multiexon deletions [18, 21]. The full text of the series places all 61 changes in the ATPase region: the 59 missense variants corresponded to 48 different changes in 34 ultraconserved amino acids, all predicted as deleterious in silico and distributed across exons 15 (5 patients), 18 (16), 19 (4), 23 (2), 24 (11) and 25 (21), and the two multiexon deletions were in-frame (32 kb, exons 20 to 26, Wolff 2012; 55 kb, exons 20 to 27, Tsurusaki 2012); the patient with p.(Gly1420Arg) in the bromodomain, identified by exome without suspicion of NCBRS, was left out of the analysis [18]. The ATPase region is 100% conserved in the chimpanzee and mouse and 94.7% in the zebrafish; the variants affect helicase motifs I, Ia and II to VI and conserved blocks B, D and H of the SNF2 family, and eight residues affected in 13 patients lie outside the known conserved zones, mainly downstream of motif VI and blocks M/N (residues 1201 to 1205, 7 patients), suggesting functional zones not yet characterised [18]. By 2016, 62 missense variants and 3 in-frame deletions had been described, clustered in the ATPase domains of exons 15 to 25, with exceptions: p.(Ala1219Pro), next to but outside the domain, in a patient with NCBRS without seizures; p.(Gln1241Glu), in exon 26; p.(Gly1420Arg), in the bromodomain, in a patient without the typical face, with ID, absent speech and seizures; the 32-kb in-frame deletion of exons 20 to 26; and a 55-kb deletion of exons 20 to 27 in a patient with a Coffin-Siris phenotype [25, 35]. This 55-kb deletion, described by Tsurusaki and colleagues in 2012, was de novo and caused skipping of exons 20 to 27; the patient did not have sparse hair, was hirsute, with thick eyebrows, long eyelashes and coarse facies, and had microcephaly, hypotonia, seizures, visual and hearing deficits, vertebral and intestinal anomalies, but no hypoplasia of the fifth fingers or nails; he was clinically classified as Coffin-Siris syndrome, and OMIM includes him among the allelic variants of NCBRS (600014.0013) [9]. The 2014 series, which included this patient (redescribed by Kosho and colleagues in 2013 as SMARCA2-1), considers that he and Wolff's patient with the 32-kb deletion had the typical physical features of NCBRS, with severe ID, absent speech and seizures, the Japanese patient also having hypospadias and an episode of macroscopic haematuria due to IgA nephropathy; the clinical classification of this patient therefore diverges between sources [9, 18]. The compilation by Zhang and colleagues (2000–2021) gathered 88 variants; after excluding 29 repeated ones and one associated with Coffin-Siris syndrome, 58 de novo positions remained: 52 missense, 2 in-frame deletions, 2 frameshift, 1 intronic splicing and 1 nonsense, with exon 25 (27 entries) and exon 18 (23) the most affected and more than half of the variants in the C-terminal helicase region [11]. The truncating variants in that compilation come from a single study (Wieczorek 2013, described on NM_003070.3), which contrasts with the GeneReviews statement that NCBRS variants are exclusively non-truncating [1, 11]. GeneReviews itself attributes about 3% of variants to gene deletion/duplication analysis (Tsurusaki 2012 and Wolff 2012), and an intragenic SMARCA2 deletion (hg19 2094861_2141830) including the ATP-binding region was described in a child with NCBRS and hypertrophic obstructive cardiomyopathy [1, 38]. The coherent reading of the sources is that the known causal variants are mainly missense variants and in-frame deletions of the ATPase/helicase domain, that intragenic deletions may preserve the reading frame, and that whole-gene deletions do not cause NCBRS [1, 8, 11].

On 27 October 2026, ClinVar contained 309 pathogenic or likely pathogenic SMARCA2 records (217 pathogenic, 78 likely pathogenic and 14 pathogenic/likely pathogenic), including copy number variants (CNV) [5]. Table 2.1 presents recurrent variants or variants cited in the article, normalised in VariantValidator against the MANE transcript (all validated without error; the coordinates returned for GRCh37 coincide numerically with those of GRCh38 for these positions) [5, 7]. The table includes the allelic variants selected by OMIM (600014.0001 to 600014.0019): the nucleotide numbering of Van Houdt (2012) and Wolff (2012) and that of the NM_001289396 transcript used by OMIM for the BIS variants coincide, at these positions, with that of NM_003070.5 (verified in VariantValidator, with NM_001289396.2 for c.1514G>A and c.2786A>T), and OMIM uses the historical one-letter protein notation (for example, R1213W for p.(Arg1213Trp)) [7, 9].

Table 2.1. SMARCA2 variants normalised according to Human Genome Variation Society (HGVS) nomenclature (NM_003070.5; NP_003061.3; NC_000009.12)

VCV, ClinVar variation identifier; P, pathogenic; LP, likely pathogenic.

Other recurrent positions classified as pathogenic in ClinVar are Glu852 (c.2554G>A p.(Glu852Lys); c.2554G>C p.(Glu852Gln)), Gly853 (c.2558G>A p.(Gly853Asp)), His854, Arg855, Gly881, Pro883 (Ser, Ala, Gln), Arg1105 (Pro, Gly, His, Leu) and Ala1201 (c.3602C>A p.(Ala1201Glu)) [5]. The historical nomenclature of some articles uses earlier versions of the transcript (NM_003070.3 in Wieczorek 2013; NM_003070.4 in Zhang 2022) or only the one-letter protein notation (p.P883L, p.G853R, p.V1198M) [11, 34, 39].

In the Chinese multicentre cohort of Yang and colleagues (2025), sequencing of 4006 patients with epilepsy (1184 with a 518-gene epilepsy panel and 2822 by exome, between March 2015 and March 2024) identified 24 patients with a heterozygous SMARCA2 variant, carrying 22 unique variants: 20 missense, 1 intronic splicing variant (c.1522-4A>G, with no predictable protein consequence, p.?) and 1 exon deletion described as «deletion of exons 9 and 26», without coordinates or indication of whether it includes the intervening exons [34]. The variants were classified according to the American College of Medical Genetics and Genomics (ACMG) criteria, confirmed by Sanger, and none was present in the Genome Aggregation Database (gnomAD); p.(Pro883Leu) and p.(Gly853Arg) occurred in 2 patients each, and one patient (patient 16) had two de novo variants, p.(Pro1150Arg) and p.(Val1177Leu) [34]. The authors describe 17 variants as novel and use NM_003070.5 numbering with GRCh37 coordinates; normalisation in VariantValidator confirmed the correspondence between the c. and p. notations of 20 of the 21 single-nucleotide variants, with two discrepancies: c.3289G>A predicts p.(Asp1097Asn) and not the p.D1097H of the article, which would correspond to c.3289G>C (neither has a ClinVar record), and the variant written p.G1164P in the list of novel variants is c.3490G>A p.(Gly1164Arg), as in the article's own table [7, 34]. In ClinVar (search of 28 October 2026), p.(Pro883Leu) (VCV 30016, P), p.(Ala1201Val) (VCV 30017, P/LP), p.(Pro883Ser) (VCV 4282221, P), p.(Gly1164Arg) (VCV 1299136, LP) and, as variants of uncertain significance, p.(Arg812His) (VCV 3898793), p.(Pro1150Arg) (VCV 4902382) and p.(Lys1197Asn) (VCV 4854326, condition NCBRS) are recorded; the others have no record [5]. Table 3.3 presents the variants by patient.

For the domain analysis, the authors combined the 24 patients of the cohort with those of a literature review (147 patients from 82 studies published between 2012 and March 2024) and separated patients with epilepsy (70, of whom 66 with a missense variant) from patients without described epilepsy (101, of whom 85 with a missense variant) [34]. Among patients with epilepsy, 77.3% of missense variants (51 of 66) were located in the SNF2-ATPase domain, 15.2% (10 of 66) between the C-terminal helicase domain and the bromodomain, 3% (2 of 66) in the QLQ domain, 3% (2 of 66) between the HSA (helicase/SANT-associated) domain and the ATP-binding helicase domain and 1.5% (1 of 66) in the HSA domain; among patients without epilepsy, 54.1% (46 of 85) were in the SNF2-ATPase domain, 25.9% (22 of 85) in the C-terminal helicase domain, 5.9% (5 of 85) between the HSA and the ATP-binding domain, 3.5% (3 of 85) in the HSA, 2.3% (2 of 85) between the QLQ and the HSA and 1.2% (1 of 85) in the QLQ, and the authors consider that the variants of patients with epilepsy are more concentrated in the SNF2-ATPase domain (p<0.005) [34]. This association is exploratory: the domain boundaries used (QLQ, residues 172 to 208; HSA, 436 to 508; ATP-binding helicase, 720 to 912; C-terminal helicase, 1080 to 1164; bromodomain, 1380 to 1508) do not coincide with the definition of the ATPase domain as residues 508 to 1305 used by Zhang and colleagues, the location categories differ between the two groups, the percentages of the group without epilepsy add up to 79 of the 85 patients, the patient with p.(Gln185Glu) may have been counted twice (section 3) and the article itself considers the separation between patients with and without epilepsy provisional, because patients still without seizures may come to have them [11, 34].

Founder variants

The sources consulted do not describe founder variants; the known recurrences (Arg1159, Pro883, Arg1105, Gly853) correspond to recurrent de novo mutational positions in patients of various origins [5, 21, 34].

Genotype-phenotype correlations

GeneReviews does not recognise genotype-phenotype correlations of clinical utility, but notes that all individuals with a variant in the C-terminal helicase region of the ATPase domain have severe ID and epilepsy, more frequently than those with variants in other parts of the gene, and that more than half of the cases described have a variant in this region [1, 11]. The 2014 series suggested that domain VI variants are associated with severe ID and seizures and that those of residues Pro883, Leu946 and Ala1201 are associated with mild phenotypes, with numbers still too small for definitive conclusions, and a possible correlation between the severity of ID and the presence of seizures, absent speech, short stature and microcephaly [18]. In the full text, 36 of 61 patients had a variant in the C-terminal helicase region, with a tendency to severe ID and epilepsy; motif VI (exon 25) concentrated 11 patients, 6 with a variant of Arg1159, a residue that contacts the γ-phosphate of ATP and is required for its hydrolysis: all 6 had severe ID and epilepsy with first seizure between 1 and 2.5 years, 3 never spoke and 2 lost speech with the onset of seizures, 3 had short stature and 4 microcephaly [18]. Four patients with an Arg1105 variant (block D) had a moderate to severe phenotype, and in the 20 with a variant in the N-terminal ATP-binding helicase region (9 in motif II) there was no consistent correlation; p.(Ala1201Val) (4 patients; OMIM gives 3), p.(Pro883Leu) (3) and p.(Leu946Ser) or p.(Leu946Phe) (2) were associated with mild to moderate ID, less speech impairment and a lower probability of epilepsy, with the caveat that adjacent residues occur in severe patients [9, 18]. In the phenotype, the severity of ID tracked postnatal short stature (17 of 25 with severe ID, 12 of 22 with moderate ID and 1 of 9 with mild ID), microcephaly (20 of 23, 11 of 20 and 3 of 9), absent speech (15 of 27, 4 of 22 and 0 of 11) and seizures (23 of 28, 13 of 22 and 3 of 11); of the 39 patients with seizures, 23 had severe ID, 13 moderate and 3 mild, microcephaly was more frequent with seizures (25 of 35) than without (10 of 35), and facial features did not correlate with developmental level [18]. These correlations have exceptions: the patient with p.(Gln1165Lys) in domain VI had moderate ID, seizures that ceased at 4 years and almost normal language at 20 years, and a patient with p.(Gly1132Asp) in the C-terminal helicase domain had a borderline full-scale intelligence quotient [21, 25]. The methylation signature shows a domain-dependent genotype-epigenotype-phenotype correlation: variants in the ATPase/helicase domain produce the full signature and the classic picture, variants proximal to the domain produce neither the signature nor clinical NCBRS, and the variant p.(Trp1283Cys), 77 amino acids distal to the domain, produced an intermediate profile in a 16-year-old girl with mild NCBRS [28]. In the epilepsy cohort of Yang and colleagues, the authors associate the intronic variant c.1522-4A>G with a more severe phenotype, with drug-resistant seizures, more marked developmental delay and behavioural comorbidity, on the basis of a single patient (patient 4: epileptic spasms from the 3rd day of life, microcephaly with a head circumference of 41 cm, below −3 standard deviations, and widened subarachnoid spaces on magnetic resonance imaging, MRI) [34].

Penetrance and expressivity

The sources do not describe unaffected heterozygous carriers of germline variants; expressivity is variable, with ID from severe to mild and sparse hair that may be present only at certain ages [1, 18].

Allelic disorders

Blepharophimosis-intellectual disability syndrome (BIS)

Cappuccio and colleagues studied 20 individuals from 18 families with a syndromic neurodevelopmental disorder not compatible with NCBRS and candidate SMARCA2 variants, with functional assays in yeast, transcriptomics and methylation in leucocytes [10]. Fourteen had a recognisable phenotype with epicanthic folds, blepharophimosis, a drooping nasal tip and ID of variable degree, and all their variants were located outside the helicase domains; the yeast assays distinguished the NCBRS variants from the others; transcriptomics grouped NCBRS with BIS and the non-specific phenotype with the controls, and DNA methylation separated BIS from NCBRS and the controls; in the other 6 individuals, with non-specific dysmorphisms, it was not possible to reclassify the variants [10].

All had been referred for ID or behavioural problems with dysmorphism and congenital anomalies, after a negative work-up that included array comparative genomic hybridisation (aCGH); they were recruited through Matchmaker Exchange and ERN ITHACA, with ethical approval in Naples (Federico II University Hospital) and in Dijon, and through the Deciphering Developmental Disorders (DDD; United Kingdom and Republic of Ireland), 100,000 Genomes Project (England) and BUILD (London) projects [10]. Sequencing was targeted in 3 individuals and, in the rest, performed by TruSight One panel, exome or trio genome; all variants were confirmed by Sanger and, according to the article, 19 were de novo (ACMG criterion PS2) [10]. The variants were absent from population databases, including gnomAD (PM2), affected conserved amino acids and were predicted to be deleterious by several programs (PP3), and no other candidate variant was found in any individual; p.(Ile932Thr) and p.(Leu529Val), with variants of the same codon recorded in ClinVar as VUS without clinical description, were classified as likely pathogenic by PS2, PM2 and PP3 [6, 10].

The BIS variants clustered in exons 8 and 9 (p.(Arg505Gln), p.(Gly513Val), p.(Arg525His), p.(Arg525Cys), p.(Leu529Val) and p.(Asp534Asn)), in the region between the HSA (helicase/SANT-associated) domain and the ATP-binding helicase domain, or in exon 19 (p.(Glu929Val), p.(Arg937His), p.(Arg937Leu) and p.(Arg937Cys)), in the linker between the ATP-binding helicase domain (DExx) and the C-terminal helicase domain; residue Arg525 was affected in 4 patients and Arg937 in 5 [9, 10]. In NCBRS, by contrast, most variants are located in exons 15 to 25, which encode the ATPase domain, with exon 25 as a hot spot [10, 18]. In the 6 individuals with a non-specific phenotype, the variants were scattered across the gene: exon 8 (p.(His484Asn) and p.(Asn486Lys)), exon 14 (p.(Leu719Pro)), exon 19 (p.(Ile932Thr)) and exon 21 (p.(Lys1014Glu)) [10]. The text of the article gives only the protein notation; Table 2.2 presents the nucleotide correspondence of the ClinVar records with the same protein change, validated in VariantValidator against the MANE transcript (the other BIS variants are in Table 2.1) [5, 7, 10].

Table 2.2. SMARCA2 variants from the Cappuccio series (2020) absent from Table 2.1 (NM_003070.5; NP_003061.3; NC_000009.12)

VCV, ClinVar variation identifier; P, pathogenic; LP, likely pathogenic; VUS, variant of uncertain significance; ARP, actin-related proteins.

Of the 57 positions with pathogenic SMARCA2 variants aligned with the Saccharomyces cerevisiae Snf2 protein, 50 (87.7%) are identical or very similar; in the structure of Snf2 bound to a nucleosome (PDB 5z3l), the NCBRS residues lie in the core of the helicase domain, and those of BIS in an alpha helix probably at the interface with other subunits of the SWI/SNF complex [6, 10]. Among the residues of the non-specific group, Ile932 (Leu977 of Snf2) lies at the interface defined by the BIS residues, Lys1014 (Lys1057) probably disrupts DNA binding, Leu719 (Ile762) is peripheral, and His484 and Asn486 (His637 and His639) are at the interface between the HSA domain and the actin-related protein module [10]. In yeast strains edited by CRISPR/Cas9, complete deletion of SNF2 and the Snf2-Pro926Leu mutant, corresponding to the NCBRS variant p.(Pro883Leu), had very severe and severe growth defects, respectively, on fermentable (galactose, raffinose) and non-fermentable (glycerol, acetate) carbon sources, in medium with hydroxyurea and at 37 °C, whereas the mutants corresponding to the BIS variants p.(Glu929Val) and p.(Arg937Leu) and to p.(Asn486Lys) of the non-specific group grew like the wild-type strain, on solid and in liquid medium [9, 10]. The assay separated the typical NCBRS variants from the others, but did not distinguish BIS from the non-specific phenotype or from the controls, nor did it clarify the molecular mechanism that differentiates the two syndromes [10].

Total blood RNA sequencing (2 patients with BIS, 3 individuals from the non-specific group, 2 patients with NCBRS as positive controls and 3 healthy controls) separated, on the first principal component (60% of the variation), NCBRS and BIS on one side from the non-specific group and the controls on the other; the second component (13% of the variation) barely distinguished NCBRS from BIS, and hierarchical clustering of the 5000 most differentially expressed genes divided the NCBRS/BIS group into two subgroups [6, 10]. SMARCA2 expression was reduced in NCBRS and BIS relative to controls (base-2 logarithm of the expression ratio, log2FC, of −0.74; adjusted p of 0.0025), a result confirmed by quantitative polymerase chain reaction (qPCR) in 20 samples, which included 7 additional controls, 2 additional patients with BIS and 1 additional individual from the non-specific group; expression was not reduced in the individuals with p.(Ile932Thr) and p.(Lys1014Glu) [10]. With thresholds of |log2FC| ≥0.58 and adjusted p ≤0.01, there were 3051 differentially expressed genes in NCBRS and 1832 in BIS relative to controls, of which 1375 were shared, 1676 specific to NCBRS and 457 specific to BIS, with enrichment in terms of transcriptional regulation and protein translation in both groups [10]. DNA methylation (Infinium EPIC array) in 4 patients with BIS separated NCBRS and BIS from the controls, but BIS formed its own subgroup; comparison with 40 age- and sex-matched healthy controls identified 163 differentially methylated CpG sites (methylation difference >10%, p corrected for multiple comparisons <0.01), which separated BIS from the controls and from NCBRS [10]. The number of RNA samples was small, and the authors present the molecular results as support for the clinical separation of the two entities, possibly with distinct mechanisms [10].

Before this series, a patient with p.(Trp1283Cys), between the ATPase domain and the bromodomain, had been described as mild NCBRS without convincing dysmorphism and with a partial methylation signature; a patient with p.(Gly1420Arg), in the bromodomain, had facial features partially overlapping with those of BIS; and an individual without NCBRS features with p.(Asp534Tyr), in exon 9, had a picture described as suggestive of Cornelia de Lange or Hallermann-Streiff syndrome [10, 28]. SMARCA2 binds directly to ADNP, and 4 individuals with variants in the bipartite nuclear localisation signal of ADNP had blepharophimosis and epicanthic folds instead of the typical phenotype of Helsmoortel-Van der Aa syndrome; the authors proposed that the blepharophimosis-ID phenotype reflects the molecular interaction of the two proteins and that SMARCA2 variants may give rise to a spectrum of phenotypes similar to that of ARID1B, from non-syndromic ID to Coffin-Siris syndrome [10]. The pathogenicity of the variants of the non-specific group remained to be established [10]. In 15 individuals with BIS and 12 with class II Helsmoortel-Van der Aa syndrome (truncating ADNP variants) a shared methylation signature was identified, the first description of a phenotype-specific epigenetic biomarker common to distinct genetic conditions, each with its own signature [13]. Helsmoortel-Van der Aa syndrome (ADNP) enters the differential. In ClinVar, the variants classified for BIS include p.(Glu464Lys), p.(Asp510Gly), p.(Gly513Asp), p.(Arg525Cys), p.(Arg525His), p.(Asp534Gly), p.(Pro625Leu), p.(Arg937Cys), p.(Arg937His), p.(Arg937Leu), p.(Gln957Arg) and p.(Tyr1489Cys), as well as a single-submitter deletion NM_003070.5:c.(-37+1_-36-1)_(3981+1_3982-1)del; some variants, such as p.(Ala1219Pro), p.(Val1141Met) and p.(Arg1105His), have been attributed by different submitters to NCBRS and to BIS, which requires clinical correlation [5].

Other SMARCA2 phenotypes

A missense variant c.553C>G p.(Gln185Glu) of the QLQ domain, inherited through paternal germline mosaicism, caused ID, early-onset epilepsy and autistic features, without the face or limb anomalies of NCBRS; GeneReviews considers it uncertain whether it represents a poorly known allelic disorder or an expansion of the phenotype [1, 32]. A 15.8-Mb duplication (9p24.3-p22.2, 46 coding genes, including SMARCA2) was found in 2 of 37 patients with a Coffin-Siris phenotype without a variant on exome; the phenotype, closer to Coffin-Siris syndrome than to NCBRS, was attributed to SMARCA2 overdosage, although ClinGen does not recognise evidence of triplosensitivity [8, 33].

Outside the Mendelian disorders, OMIM records an association between schizophrenia and three SMARCA2 single nucleotide polymorphisms in the Japanese population (Koga 2009), including the missense polymorphism rs2296212 (D1546E in the study's notation; = NM_003070.5:c.4638C>G p.(Asp1546Glu)), whose risk allele reduced the nuclear localisation of BRM, and intronic polymorphisms associated with lower expression in the post mortem prefrontal cortex; SMARCA2 is part of a genomically supported schizophrenia gene network regulated by REST/NRSF [9]. This is a susceptibility association with common variants, unrelated to the risk of NCBRS or BIS.

Somatic variants and tumour risk

Somatic and germline variants of SWI/SNF complex genes are associated with tumour suppression, and isolated sporadic tumours often contain somatic SMARCA2 variants absent from the germline, without hereditary predisposition [1]. No cancers have been reported in people with NCBRS; most patients described are children or adolescents without follow-up into adulthood, so the oncological risk has not been established [1, 11]. Unlike SMARCA4 and SMARCB1, associated with rhabdoid tumour predisposition syndromes, SMARCA2 is not among the genes of these syndromes [30]. The selective SMARCA2 degraders in phase 1 and 2 trials are intended for SMARCA4-deficient cancers and have no relation to the treatment of NCBRS [14, 40]. The 2021 French PNDS concludes that the available data do not allow an increased tumour risk to be asserted in people with constitutional BAFopathies: the tumours described occurred in nine patients with Coffin-Siris syndrome due to minor genes, and the risk cannot be excluded only for missense variants of SMARCA4 and SMARCB1; the protocol does not include oncological surveillance in follow-up and calls for long-term cohort studies [15].

Modifiers and dual diagnoses

The sources do not identify modifier genes. Dual diagnoses that alter the phenotype or the response to treatment have been described: NCBRS with a POLG1 variant and hepatotoxicity to valproic acid; NCBRS with nocturnal frontal lobe epilepsy due to CHRNB2; and siblings with NCBRS in whom the proband also had a de novo TCF12 variant responsible for craniosynostosis [37, 41, 42].

3. Clinical Spectrum and Manifestations by System

Reference cohorts

The clinical description rests on small series and literature compilations: 23 patients in the 2009 delineation, 61 molecularly confirmed cases in the 2014 consortium series, at least 80 individuals with a pathogenic variant in the compilation by Zhang and colleagues (up to 2021), 25 patients in an international study focused on epilepsy and 24 patients in a Chinese epilepsy cohort [11, 18, 23, 34, 43]. In the molecular series of Van Houdt and colleagues (2012), of the 36 individuals assessed, 34 with a clinical diagnosis considered certain, all had ID (mild in 3, moderate in 9 and severe in 24), 35 had sparse hair, 22 of 35 had seizures, 19 of 35 microcephaly, 28 of 35 prominent interphalangeal joints, 21 of 35 prominent distal phalanges and 16 of 32 short metacarpals or metatarsals; about half had excessive skin wrinkling, one third prenatal growth restriction and more than half postnatal growth restriction; the most constant facial features were a wide mouth (34 of 36), a thick lower vermilion (32 of 36), a broad (31 of 36) and long (29 of 36) philtrum, a thin upper vermilion (27 of 36) and thick, anteverted alae nasi, present in the vast majority [3]. In the Italian series of Mari and colleagues (8 patients with a SMARCA2 variant), all had developmental delay with very poor speech, coarse facies, a low anterior hairline, sparse hair, a wide mouth with a thick lower vermilion and prominent interphalangeal joints, and thick eyebrows, synophrys, long eyelashes, a broad nose with anteverted nares, hypotonia, seizures, feeding difficulties and eczema were variable [3, 26]. The cardinal features (ID, short stature, microcephaly, typical face, sparse hair, brachydactyly, prominent interphalangeal joints, behavioural problems and seizures) are present in almost all, with no sex difference [18, 23]. Table 3.1 summarises the GeneReviews frequencies. In the 2014 series, the 61 patients comprised 35 males and 26 females with a median age of 10 years at last assessment (2 to 33 years), and Table 3.2 shows the frequencies by age group, which make it possible to see the evolution of each sign with age [18].

Table 3.1. Frequency of selected findings in NCBRS (basis: at least 80 individuals with a pathogenic SMARCA2 variant)

Source: [1, 11].

Table 3.2. Frequencies by age group in the series of 61 patients with NCBRS and a SMARCA2 variant (Sousa and Hennekam, 2014)

Source: Sousa and Hennekam (2014), Table I; birth data refer, for the means, only to newborns of 38 to 42 weeks; ID was classified by estimate of the intelligence quotient (IQ), without formal testing in most (mild, IQ 50 to 69; moderate, 35 to 49; severe, <35) [18]. The «printed» values are those of the article when they differ from the division of the numbers presented; for gingival hypertrophy, the value 9/10 in the ≥17 years group is incompatible with the total of 2/48. SD, standard deviation; —, no data by age group.

An evolving syndrome

NCBRS evolves in phases: the face is not very characteristic at young ages and becomes evident with age, with rough facial skin, wrinkles and laxity due to scarce subcutaneous tissue, especially when smiling; the prominent interphalangeal joints and the typical face become evident mainly after 6 years, and intellectual disability and behavioural changes dominate the later phases [11, 21, 23]. This evolution explains why the overlap with Coffin-Siris syndrome in infancy leads to underdiagnosis or misclassification [21].

Neurodevelopment

Developmental delay and ID are severe in almost half of patients, moderate in one third and mild in the rest [1, 11, 20, 44]. In the Van Houdt series, ID was severe in 24 of 36 (67%), moderate in 9 and mild in 3, a proportion of severe forms higher than that of the GeneReviews summary, a difference the sources do not explain [1, 3]. About 80% have severe speech delay and almost one third never develop spoken language; about half of 60 patients described have absent speech or speech regression [1, 21]. The mean age at sitting is about 9 months and at independent walking 21 months (range 10 months to 5 years) [1]. Psychomotor regression is not typical, but progressively worsening seizures are associated with loss of speech, and regression or arrest of progress coincides with seizure onset, in a pattern of epileptic encephalopathy [1, 35]. About one third have hypotonia, and at least half of infants have feeding difficulties, usually without the need for tube feeding [1]; the patient of Ejaz and colleagues required a gastrojejunostomy from 4 months to 12 years [21]. In the 2014 series, ID was severe in 28 of 61 (45.9%), moderate in 22 (36.1%) and mild in 11 (18.0%); speech was absent in 19 of 60, first words appeared at a mean of 29.5 months (median 24; 10 to 96 months) and 9 of 42 later lost or reduced the words acquired, in some cases at the same time as the first seizure [18]. Hypotonia affected 19 of 51, but sitting (mean 9 months; median 8; 6 to 20) and walking unsupported (mean 21 months; median 18; 10 to 60) were not much delayed; brain imaging, performed in at least 42 patients, was usually normal, with enlarged ventricles in 3 and a small corpus callosum in 2 [18]. Choking was very frequent, feeding problems (23 of 49) usually mild and without the need for a nasogastric tube, and many patients prefer completely pureed food into adulthood [18]. The French PNDS (French National Diagnosis and Care Protocol), which summarises the same series, concludes that the neurodevelopmental phenotype is more marked in NCBRS than in Coffin-Siris syndrome, especially that associated with ARID1B; in the data the protocol gathers for the latter, ID is present in more than 95% and is moderate to severe, 20 to 30% do not acquire speech, children sit at a mean of 12 months, walk at 30 and say their first words at 24, and epilepsy affects about 25% of patients with an ARID1B variant, with few seizures and good response to treatment [15]. In the 2019 ARID1B cohort, from which these figures come, seizures affected 27.5% of 142 patients, with a lifetime risk estimated at about 40% by Kaplan-Meier, and 4 of 17 patients with recorded frequency had monthly or more frequent seizures; the 18 treated patients all responded to medication [45]. Feeding difficulties also differ: in NCBRS they are frequent (46% in the 2014 series) but rarely require nutritional support, whereas in Coffin-Siris syndrome they affect more than 70% and lead to temporary nasogastric tube feeding in 17% [15, 18].

Epilepsy

Epilepsy affects two thirds of patients in the GeneReviews summary (65%), 22 of 35 (63%) in the molecular series of Van Houdt, about half in the Zhang compilation and 23 of 25 patients in the targeted study by Hofmeister [1, 3, 11, 43]. Onset ranges from birth to 14 years, with a mean of 20.9 months and a median of 16.5 months; in the 2014 series, the median age at first seizure was 18 months [1, 35]. In that series, seizures affected 39 of 61 (63.9%), from 8 of 14 below 7 years to 10 of 14 between 12 and 16 years, and the first seizure occurred at a mean of 23.9 months (0 to 168 months); the seizure type varied, even in the same patient, the frequency increased in some despite several antiepileptic drugs, the EEG showed no specific changes and sodium valproate was the drug of first choice of many clinicians [18]. In the monozygotic twins of the series, cognitive decline coincided with seizure onset in one of them and occurred in the same way in the other, whose seizures only began two years later; the authors therefore attribute the decline and the seizures to the same process, rather than the decline to the seizures [18]. Seizure types include absences, tonic-clonic, atonic seizures, epileptic spasms and focal status epilepticus; generalised seizures occur in about half according to GeneReviews and in 85% (17 of 20) in the Hofmeister series, where the electroencephalogram (EEG) showed generalised epileptogenic abnormalities in 53% (9 of 17) [1, 11, 43]. Brain magnetic resonance imaging (MRI) is usually normal, and the authors consider the epilepsy more genetic than structural [11, 43]. Myoclonic-atonic epilepsy (Doose syndrome) has been described, with axial negative myoclonus and generalised polyspike-wave bursts on the EEG at 17 months [35]. Orphanet lists status epilepticus, absences and epileptic spasms as frequent [2]. Initial response to antiepileptic drugs is frequent, especially to levetiracetam and valproic acid, but complete remission is very rare (section 7) [1, 43].

In the Chinese cohort of Yang and colleagues (24 patients, 16 male and 8 female, with a median age of 4 years, from 1 year and 5 months to 11 years, followed in clinic or by telephone every 3 months), the first seizure occurred between the 1st day of life and 75 months (mean 18.67 ± 15.73 months; median 14.5 months in the results text and 15 months in the abstract): before 1 year in 9 of 24 (37.5%), between 1 and 2 years in 10 of 24 (41.7%) and after 2 years in 5 of 24 (20.8%) [34]. Focal seizures predominated (15 of 24, 62.5%; the abstract gives 60.9%), followed by generalised tonic-clonic seizures (10 of 24, 41.7%), myoclonic seizures with spike-wave (3 of 24, 12.5%), epileptic spasms (2 of 24, 8.3%) and atonic and tonic seizures (1 of 24 each) [34]. Two patients had infantile epileptic spasms syndrome (IESS), 1 Lennox-Gastaut syndrome, 1 developmental and epileptic encephalopathy with spike-wave activation in sleep (DEE-SWAS), 5 non-specific developmental and epileptic encephalopathy (DEE) and 13 epilepsy with developmental delay, with seizure onset between 10 and 75 months (median 15 months), mainly focal and generalised tonic-clonic seizures and delay that, according to the authors, became apparent after the onset of epilepsy [34]. The patient with DEE-SWAS, the first description of electrical status epilepticus in slow sleep (ESES) with a SMARCA2 variant, had seizures in sleep from 3 years and 6 months, slow spike-wave at 1.5 to 2 Hz in non-REM sleep (without rapid eye movements) and regression with reduced verbal fluency, poorer fine motor skills, drooling without dysphagia and falls without loss of consciousness, while retaining verbal communication and higher cognitive functions; Fig. 3 of the article shows her EEG at 5 years and 3 months [34]. Video-EEG (4 to 24 hours) showed diffuse slowing of background activity in 8 patients, interictal discharges in 16 (focal slow spike-waves in 8 and multifocal in 3), hypsarrhythmia in 2 and electroclinical seizures recorded in 4 (16.7%), and was normal in 7 at last follow-up [34]. Brain MRI was abnormal in 3 of 24 (12.5%): widening of the frontal and temporal subarachnoid spaces with deepened sulci and enlarged lateral ventricles (patient 4, at 1 year and 2 months, Fig. 4 of the article), rounded posterior horns of the lateral ventricles (patient 2) and an arachnoid cyst of the cisterna magna (patient 18) [34]. In neurodevelopment, 23 of 24 (95.8%; 95.3% in the discussion) had delay, 13 of 24 (54.2%) poor language development, 4 microcephaly, 2 hypotonia and 1 hypertonia, 2 attended special education and 2 had regression or stagnation after 2 years (at 3 years and 3 months and at 2 years and 4 months); one patient had normal psychomotor development at 4 years [34]. The cohort was selected for epilepsy and the article does not indicate which patients met the clinical criteria for NCBRS: the extraneurological findings recorded are scarce (hypotrichosis, brachydactyly and hypertrichosis in 1 patient each, short stature in 2), which brings part of the cohort closer to the SMARCA2 phenotype without NCBRS [34].

Table 3.3. Chinese epilepsy cohort with a SMARCA2 variant (Yang 2025): genotype, epilepsy and antiepileptic drugs by patient (NM_003070.5; NP_003061.3)

Variants confirmed in VariantValidator against NM_003070.5 (GRCh38 and GRCh37 with the same genomic numbering at these positions) [7]; clinical data from Table 1 of Yang 2025, in which underlining indicates clinical response (reduction in seizure frequency or severity) and italics worsening, which is not indicated in any patient [34]. y, years; m, months; M, male; F, female; GTC, generalised tonic-clonic seizures; ED, epilepsy with developmental delay; DEE, developmental and epileptic encephalopathy; DEE-SWAS, DEE with spike-wave activation in sleep; IESS, infantile epileptic spasms syndrome; VPA, valproic acid; LEV, levetiracetam; OXC, oxcarbazepine; TPM, topiramate; CLB, clobazam; CZP, clonazepam; NZP, nitrazepam; LCM, lacosamide; LTG, lamotrigine; PER, perampanel; ZNS, zonisamide; PB, phenobarbital; VGB, vigabatrin; ACTH, adrenocorticotropic hormone.

In the literature review of the same article (46 patients with epilepsy and a SMARCA2 variant, from 11 studies according to the results and 23 according to the methods), the first seizure occurred between 8 months and 18 years, between 1 and 2 years in 27 of 38 (71%; 68% in the discussion); generalised tonic-clonic seizures were the most frequent (6 of 11, 54.5%, in the results; 11 of 16, 69%, with focal seizures in 4 of 16, 25%, in the discussion), all had developmental delay (46 of 46), 4 had ASD, 6 hypotonia and 3 suspected attention deficit hyperactivity disorder (ADHD) [34]. The most frequent variants in this review were those at positions Glu852 (Lys, Asp or Gln; 6 patients), Arg855 (Gln, Gly or Leu; 4), Arg1159 (Gln, Gly or Leu; 4) and p.(Gly752Ala) (3) [34].

Behaviour

Behavioural problems have been described in at least 19 individuals, with autistic features (perseveration, hyperacusis), attention deficit and oral sensitivities; at least 3 had a clinical diagnosis of autism spectrum disorder (ASD) [1]. There are reports of formally diagnosed ASD at 5 and 8 years, and a report of two cases with ASD [35, 46, 47]. Orphanet lists echolalia, mutism and aphasia as very frequent [2]. In a patient with moderate ID, self-injury increased after puberty and behaviour became the main source of distress for the family [21]. In the 2014 series, patients were typically cheerful and sociable, with tantrums and periods of aggressiveness, autism spectrum traits without a formal diagnosis in any, short attention span, a high pain threshold, sensitivity to noise that decreased with age and oral sensitivity, with a liking for salty and spicy food; in the older ones, parents noticed slowing of movements [18].

Growth

About half have low birth weight and the same proportion has short stature; only one known individual was above the 50th centile for height [1]. Microcephaly tends to be acquired: present in almost one third at birth and in two thirds at follow-up [1]; in the Van Houdt series, 19 of 35 (54%) had microcephaly [3]. There are exceptions, such as a patient with a head circumference between the 98th and 99.6th centiles at 17 months [35]. In the 2014 series, one third (19 of 57) were born with a weight below −2 standard deviations (SD), a proportion lower than the approximately half reported by GeneReviews; in term newborns, the means were 48.7 cm length, 2732 g weight and 33.1 cm head circumference, and birth length was below −2 SD in 8 of 38, but above +2 SD in 5 of 38 [1, 18]. Postnatal short stature affected 30 of 56, almost all below the 50th centile, with normal body proportions; adult height ranged between 1.44 and 1.80 m in men (mean 1.58 m; n=5) and between 1.30 and 1.69 m in women (mean 1.57 m; n=6) [18]. Microcephaly went from 7 of 30 at birth to 34 of 52 at follow-up, with a mean adult head circumference of 52.8 cm in men (50 to 56 cm; n=4) and 53.2 cm in women (50.5 to 54 cm; n=6); weight was below the 50th centile in 36 of 46 and below −2 SD in 24 of 46, and no patient was overweight [18].

Hair, skin and subcutaneous tissue

Scalp hair is usually sparse at birth and becomes sparser with age, especially in the second decade, although in some patients the sparseness improves [1]. A low hairline, thick eyebrows and hirsutism of the neck and back have been described in some patients [11]. Skin pigmentation appears reduced, without true albinism, and the poor distribution of subcutaneous adipose tissue makes the veins visible; translucent skin with visible vessels and a markedly everted lower lip are more characteristic of NCBRS than of Coffin-Siris syndrome [1, 19]. Orphanet lists eczematoid dermatitis and excessively wrinkled skin as frequent [2]; eczema of the elbows and knees, frontal baldness and, in a 4-year-old girl, persistent severe eczematous dermatitis with diffuse non-scarring alopecia have been described [35, 48]. At 20 years, one patient had cystic acne of the back [21]. In the 2014 series, sparse hair (59 of 61) could appear in the first months of life or only in the second half of the first decade or in the second decade, with normal growth and texture and no abnormalities under the microscope; at birth there may be facial hypertrichosis and a low anterior hairline, the eyebrows, normal or dense initially, tend to thin like the hair, the eyelashes remain prominent, pubic hair is normal and facial hair is sparse in adult men [18]. The skin was wrinkled in 33 of 61, mainly at the distal extremities, with very visible subcutaneous veins; 44% had hypertrichosis, mainly on the neck and back, and one third had eczema, mainly on the distal extremities and the face, sometimes severe, which almost always decreased in the second half of the first decade; the skin was often sensitive and pale, 6 patients had generally small nails and none had small nails only on the fifth digits; hypohidrosis, hypopigmented patches, a café-au-lait spot, patchy hyperpigmentation and a supernumerary nipple occurred in isolated patients [18].

Face

The face includes prominent eyelashes, thick eyebrows, downslanting palpebral fissures, ptosis, anteverted nares, a long philtrum, a wide mouth, a thin upper and thick lower vermilion [1]. Orphanet adds a triangular face, rounded premaxilla, broad nasal base, thick alae nasi and an upturned nasal tip [2]. The features are subtle in the neonatal period and early childhood and become coarse and wrinkled over time; after puberty deep-set eyes, an overhanging columella and prognathism appear [1, 21]. Micrognathia with limited mouth opening due to trismus and, in a 22-month-old child, a very small mouth, midface hypoplasia and a V-shaped cleft palate have been described [49, 50]. In the 2014 series, the face was triangular, with dense, prominent eyelashes, a broad nasal base, thick nares, an upturned nasal tip, rounded premaxilla, broad philtrum, thin upper vermilion, thick and everted lower vermilion and a wide mouth; in the young child, the broad philtrum is often accompanied by protrusion of the central portion of the upper lip, in a Cupid's bow [18]. The face was coarse in 43 of 56 and became progressively coarser in 18 of 31; with the loss of subcutaneous adipose tissue, the infraorbital skin becomes lax and wrinkled, especially on the cheeks when smiling, although some keep full cheeks, and in some adults the lower third of the face widens, especially at the angle of the mandible; several patients were initially suspected of having Williams syndrome, and there was one case of congenital choanal atresia and another of cleft palate [18].

Teeth

Delayed tooth eruption is common and often requires surgical extraction of the primary dentition to allow eruption of the permanent teeth; in one series, 58.2% had widely spaced teeth and 18.6% hypodontia [1]. That series is the 2014 one, with denominators of 32 of 53 for widely spaced teeth and 8 of 43 for hypodontia or oligodontia; enamel was abnormal in 3 of 42 and there was gingival hypertrophy in 2 of 48, and delayed eruption, present in one third, affected the permanent dentition more often than the deciduous [18]. In the Chinese epilepsy cohort of Yang and colleagues, the oral cavity was assessed in 11 of the 24 patients: 8 of 11 (73%) had widely spaced teeth, 3 of 11 (27%) fused teeth, all in the deciduous dentition of the lower arch, 1 of 11 hypodontia (absence of one deciduous incisor, in a patient with fused teeth) and 1 of 11 a high-arched palate; fused teeth are described for the first time in SMARCA2 disease, and Fig. 2 of the article shows the oral cavity of 8 patients [34]. The proportion of 52.8% (19 of 36) with dental disease, which the abstract attributes to the cohort, corresponds in the text to the patients of the literature review with oral data (15 with widely spaced teeth, 3 with delayed eruption and 1 with abnormalities of the upper teeth) [34]. The authors cite, in the literature, widely spaced teeth in 45.8% to 58.2% and attribute the higher proportion in the cohort to selective recording of oral findings; the discussion states that delayed eruption, hypodontia and enamel changes did not occur in their patients, mostly young children with deciduous dentition, which contradicts the hypodontia recorded in one patient, and recommends referral for dental assessment and treatment [34].

Hands, feet and skeleton

The distal phalanges widen with age, becoming oval and broad, the interphalangeal joints appear swollen because of scarce subcutaneous tissue, and the space between the first and second toes may increase (sandal gap); the nails are not hypoplastic, although small nails, usually of the fifth digit, have been described [1, 11, 19]. Hand radiographs show cone-shaped epiphyses, metaphyseal widening and shortening of the phalanges, metacarpals and metatarsals (especially of the 4th and 5th rays), with bone age more often delayed but sometimes advanced; other findings are platyspondyly, flat intervertebral discs, a small pelvis, hypoplasia of the pubis, small femoral heads and a short femoral neck [1, 51]. Scoliosis occurred in 9 of 22 patients in one series, and Orphanet lists a short palm, joint dislocation and metacarpal changes as very frequent [2, 21]. In the 2014 series, the hands and feet were usually normal at birth; over time, the distal phalanges widened in all directions in three quarters of patients, almost half had fetal fingertip pads and prominent interphalangeal joints (50 of 59) were the most characteristic sign; finger mobility, initially normal or increased, tends to decrease, some older patients do not tolerate passive mobilisation of the fingers and none had true arthritis [18]. Sandal gap was present in about half (26 of 51), with distal thickening of the toes less marked than in the hands; cone-shaped epiphyses, initially considered a major sign, occurred in 9 patients (of 34 with radiographs), some had ivory epiphyses, and shortening of the metacarpals and phalanges could affect all digits, especially the 4th and 5th rays; few had small fifth fingers and none had small or absent distal phalanges or absent fifth-digit nails [18]. Bone age was normal in 53.9%, delayed in 41% and advanced in 5.1%; osteoporosis was not rare and several patients had fractures at or after puberty, and congenital hip dislocation (4 of 45), patellar dislocation (1 patient), generalised joint laxity and scoliosis (17 of 60) complete the skeletal picture [18].

Eyes and hearing

Myopia affects 15% and astigmatism 6%; glaucoma has been described in 2 individuals in GeneReviews [1]. A systematic review of molecularly confirmed cases highlighted long eyelashes, thick eyebrows, lax periorbital skin, downslanting fissures and ptosis, which may require surgical correction [52]. Bilateral glaucoma with cataracts and degenerative vitreoretinopathy in a 12-year-old girl and bilateral congenital glaucoma in an 11-year-old boy, presented as the third case of this association, have been described [53, 54]. Hearing loss affects 5 to 10%, predominantly conductive, with one sensorineural case [1]. In the 2014 series, hearing loss affected 4 of 59 (3 conductive and 1 congenital sensorineural), and myopia was diagnosed in 10 patients and astigmatism in 4 [18].

Heart, kidney, genital tract and other organs

Cryptorchidism affects most boys (60%), and hypospadias and umbilical hernia have been described [1, 21]. Six individuals had congenital heart disease (atrial septal defect, pulmonary artery stenosis, coarctation, patent ductus arteriosus, double aortic arch), and a child with an intragenic SMARCA2 deletion had hypertrophic obstructive cardiomyopathy [1, 38]. A 22-month-old child had bilateral dysplastic kidneys with an estimated glomerular filtration rate of about 40 ml/min/1.73 m², nocturnal oxygen and combined apnoea, but the report also mentions signs of partial 9p monosomy [50]. Hashitoxicosis in early adolescence has been described, with a suggested link between SMARCA2 and the thyroid [1, 55]. There are no published frequencies of endocrine disease in NCBRS; the French PNDS includes hypothyroidism and diabetes mellitus among the complications that the attending physician should monitor in both syndromes, based on frequencies measured in patients with ARID1B variants (hypothyroidism in 19%, growth hormone deficiency in 13% and diabetes in 7%), the application of which to NCBRS is an extrapolation [15]. In the 2014 series, cryptorchidism affected 20 of 34 boys, there was hypospadias in 1 of 36, inguinal and/or umbilical hernias in 26 of 57, one omphalocele and 3 cases of pectus excavatum; the 6 heart defects were a double aortic arch, a small ostium secundum atrial septal defect with mild stenosis of the left pulmonary artery that resolved spontaneously, mild pulmonary stenosis with mild left ventricular hypertrophy that resolved spontaneously, an atrial septal defect with patent ductus arteriosus, mild aortic coarctation with patent ductus arteriosus and a brachiocephalic trunk crossing the midline with tracheal compression [18]. Menarche occurred at a median age of 14.8 years (11 to 19), with normal menstruation except in one patient with oligomenorrhoea, and at least half of adult women had reduced breast development; recurrent infections, mainly urinary, were infrequent, and vesicoureteric reflux, IgA nephropathy, marked constipation, villous atrophy, mild dyslipidaemia, an abnormal carnitine profile and spastic paraparesis occurred in isolated patients [18].

Prenatal period

A fetus with a de novo pathogenic SMARCA2 variant had caudal regression (sacral agenesis) and congenital vertical talus, the first findings of this type in NCBRS; it is not certain whether they are a rare part of the syndrome or a co-occurrence [1, 56].

Blepharophimosis-intellectual disability syndrome (BIS)

BIS manifests with global developmental delay, often with severe hypotonia and limited mobility, mild to severe ID with absent or severely impaired speech and behavioural problems [1, 4]. The face includes blepharophimosis, epicanthic folds, sparse eyebrows and eyelashes, a broad nasal bridge, a short nose with a drooping tip, a wide diastema, a tented upper lip, an open mouth with a thin upper vermilion and abnormal ears [1, 4]. The features characteristic of NCBRS are rare or absent in BIS (sparse hair in 4 of 14, prominent interphalangeal joints in 0 of 12, seizures in 3 of 14; Table 3.4), and the phenotype and methylation signature bring it closer to Helsmoortel-Van der Aa syndrome [10, 13]. The OMIM synopsis of the 14 patients of the original series, aged 3 to 19 years, adds data by system: walking was delayed in most and several never walked independently; many had hypotonia, sometimes progressing to peripheral spasticity; 3 had seizures and 6 microcephaly (down to −4.4 standard deviations); some had poor weight gain and feeding difficulties with gastro-oesophageal reflux; visual problems were frequent (astigmatism, hypermetropia, esotropia, myopia and cortical visual impairment); many had widely spaced teeth or enamel defects; 3 boys had cryptorchidism and one girl hypoplastic labia; almost all had variable skeletal anomalies (contractures, flat or deformed feet, long tapering fingers, clinodactyly); about half had recurrent respiratory infections; and neuroimaging, performed in a few, showed enlarged ventricles, pontine hypoplasia or non-specific white matter signs [6]. Other facial features described were frontal bossing, hypertelorism, a long or short philtrum, a short nose with small nares and hypoplastic alae nasi, a mental crease and a broad, flat face [6]. The OMIM synopsis and the original article agree on the counts of seizures (3 of 14) and microcephaly (6 of 14, postnatal) [6, 10].

In the series of Cappuccio and colleagues, the 14 patients with BIS shared blepharophimosis (14 of 14), epicanthic folds (14 of 14), sparse eyebrows and eyelashes (11 of 14), highly arched eyebrows (10 of 14), a broad nasal bridge (10 of 14) and a drooping nasal tip (8 of 14); also frequent were frontal bossing, hypertelorism, a narrow nose with hypoplastic alae nasi, a tented upper lip with an accentuated Cupid's bow and an open mouth with a U-shaped upper vermilion [10]. The nose distinguishes the two syndromes, with a drooping tip in BIS and an upturned tip with a broad base in NCBRS, and BIS has no coarse facies, thick lower vermilion or prominent interphalangeal joints; computerised facial analysis (Face2Gene) confirmed the difference between BIS, the non-specific phenotype and NCBRS [10]. Dental anomalies (widely spaced teeth, enamel hypoplasia, early tooth loss) occurred in 8 of 12, and joint contractures and thin or tapering fingers were frequent [10]. ID was moderate to severe in all 14, speech was severely delayed (7 with few or no words), walking was acquired between 22 months and 8 years and 6 patients never walked independently, using a wheelchair or support devices; 2 had limb spasticity, and refractive errors, recurrent respiratory infections, gastro-oesophageal reflux and feeding difficulties were frequent [10]. Orphanet describes the ID of BIS as mild to severe, whereas the original series included only moderate to severe forms; the sources consulted do not indicate whether the difference comes from later cases [4, 10]. Table 3.4 compares the frequencies in BIS with those in NCBRS and the non-specific phenotype.

Among the 20 individuals with SMARCA2 variants without NCBRS, hypotonia (13 of 20), postnatal microcephaly (8 of 20) and behavioural problems (13 of 20) were frequent, and the features of NCBRS rare: sparse hair in 6, coarse facies in 2, thick lower vermilion in 6 and prominent interphalangeal joints in 2; none had enough findings for a diagnosis of NCBRS, not even of a mild form [10]. In the 6 individuals with a non-specific phenotype, the facial, dental and limb changes of BIS were rare, and the common features partially overlapped with those of BIS or NCBRS without suggesting either [10].

Table 3.4. Comparative frequencies in NCBRS, BIS and the non-specific SMARCA2 phenotype

Source: Cappuccio and colleagues (2020), Table 1; the NCBRS column reproduces the percentages of Sousa and colleagues (2014) cited in the article, without denominators [10, 18]. Comparison with the original table of the 2014 series (Table 3.2) shows transcription differences: ptosis is 20.7% (12 of 58), and 21.8% is the value for synophrys; prominent interphalangeal joints are 84.7% (50 of 59); «small distal phalanges» correspond to short phalanges (20.9%), because the series recorded no patient with small distal phalanges; «respiratory infections» correspond to frequent infections, mainly urinary (27.1%); «genital anomalies» correspond to cryptorchidism (20 of 34 boys); and short metacarpals and metatarsals are 39.0% and 9.8% [10, 18]. The denominators vary because not all findings were assessed in all individuals. In the alae nasi row, the denominator printed in the BIS column («0/149») is incompatible with the series of 14 patients. Narrow palpebral fissures are given as 10 of 14 in the table, whereas the text of the article describes blepharophimosis in 14 of 14.

SMARCA2 phenotype without NCBRS

The patient with the p.(Gln185Glu) variant of the QLQ domain had ID, early-onset epilepsy and autistic features, without a characteristic face or limb anomalies [32]. In the cohort of Yang and colleagues, patient 1 has the same variant, the same paternal germline mosaicism of 2.8% in sperm and the same seizure onset at 2 years, and probably corresponds to the follow-up of Pan's patient: at 11 years he had Lennox-Gastaut syndrome with generalised tonic-clonic, focal, myoclonic and tonic seizures, severe delay, attendance at special education and persistent seizures after five antiepileptic drugs; the article does not identify the two cases as the same patient and also includes Pan's patient in the literature review [32, 34]. Patients with variants proximal to the ATPase domain classified as benign by the methylation signature did not have clinical NCBRS [28].

Natural history and treated disease

There is no disease-modifying treatment, so the natural history described is that of patients on symptomatic treatment. Two deaths have been published: the original British patient at 33 years, from status epilepticus and respiratory complications, and a Polish patient at 25 years, from ruptured oesophageal varices [11]. At 32 years, one year before her death, the original patient had lost all language and depended on a wheelchair for most of her mobility, which documents the course of a severe form under symptomatic treatment [3]. In contrast, the patient of Ejaz and colleagues, with moderate ID, was seizure-free from 4 years and, at 20, was independent in daily activities, with almost normal language [21].

4. Clinical Recognition and Red Flags

When to suspect

There are no consensus clinical diagnostic criteria [1, 56]. NCBRS should be considered in the presence of developmental delay or ID (usually severe, with disproportionate impairment of language), sparse hair, prominence of the interphalangeal joints and distal phalanges due to poor distribution of subcutaneous fat, a characteristic face that becomes coarse and wrinkled over time, microcephaly, seizures, the hand radiographic findings described in section 3 and a usually negative family history [1]. The microscopic structure of the hair was normal in the first two patients studied, so the sparse hair of NCBRS does not correspond to a shaft abnormality identifiable under the microscope [3]. In differential diagnosis lists by sign, NCBRS appears among the causes of coarse facies with severe developmental delay without regression (seizures, sparse hair, brachydactyly and swelling of the interphalangeal joints, with clinical and radiological changes of the phalanges) and among the causes of changes in hair, teeth, nails and skin [51, 57]. The practical distinction from Coffin-Siris syndrome rests on sparse hair instead of hypertrichosis, brachydactyly with widened interphalangeal joints and distal phalanges, epilepsy, severe language impairment and the absence of nail and fifth-digit hypoplasia [1, 19]. Epilepsy panels and exome also identify patients with a SMARCA2 variant without the full syndromic picture, with epilepsy onset in the first or second year of life, predominantly focal seizures, developmental delay and, frequently, widely spaced teeth [34].

By age

In the prenatal period there is no characteristic ultrasound sign; the only prenatal report describes caudal regression and vertical talus [56]. In the newborn and infant, the facial features are subtle and may be absent: a patient without identified dysmorphism at 17 months was only diagnosed after an exome requested for myoclonic-atonic epilepsy and re-examination at 5 years and 5 months [1, 35]. The earliest clues are feeding difficulties, poor weight gain, sparse hair, cryptorchidism and motor delay [1, 21]. At preschool and school age, the onset of seizures with regression or arrest of language and the appearance of prominent interphalangeal joints, especially after 6 years, increase the suspicion [1, 21]. In adolescents and adults, the coarse and wrinkled face, progressively sparse hair and sandal gap make the phenotype more recognisable; adults are probably underdiagnosed because they have not had genomic testing [1]. The French PNDS (French National Diagnosis and Care Protocol) adds that NCBRS is rarely suspected in the prenatal period, because agenesis of the corpus callosum, which points to Coffin-Siris syndrome in a fetus, is not part of the picture, and that the diagnosis arises mainly in the presence of developmental delay or learning difficulties with morphological particularities, sometimes with epilepsy; at any age, including adulthood, the typical morphological features suffice to raise suspicion, and the hypotonic infant with laryngomalacia and feeding difficulties or the child with delayed milestones justify assessment by a paediatric neurologist or a geneticist [15].

Red flags and emergency signs

Status epilepticus is a real risk, listed as frequent by Orphanet and the cause of death of the original patient at 33 years [2, 11]. Loss of language or motor skills coinciding with worsening seizures justifies neurological and treatment reassessment [1]. Coughing or choking during feeding, poor weight gain, frequent respiratory infections and unexplained food refusal are signs of dysphagia with a risk of aspiration [1]. Any new ocular complaint or sign justifies ophthalmological assessment, because glaucoma, including bilateral congenital glaucoma, and cataracts have been described [52–54]. Before general anaesthesia, micrognathia, limited mouth opening and cleft palate indicate a difficult airway [49, 50]. On valproic acid, signs of liver dysfunction require immediate assessment, because the label associates the drug with fatal hepatotoxicity, especially in the first 6 months and below 2 years, and contraindicates it in POLG-related mitochondrial disease; surveillance is all the more necessary if POLG-related disease is suspected [41, 58].

Newborn screening

There is no newborn screening for NCBRS or BIS; the sources consulted describe no programme.

5. Diagnosis: Criteria, Tests and Algorithms

Diagnostic criteria

The diagnosis is established in a proband with suggestive findings and a heterozygous pathogenic (or likely pathogenic) SMARCA2 variant, according to the criteria of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP); in a clinical context, the two classes have the same value [1]. A variant of uncertain significance (VUS) in SMARCA2 neither establishes nor excludes the diagnosis [1]. There are no consensus clinical criteria for NCBRS or BIS [1, 56]. Sousa and Hennekam advocate reserving the diagnosis of NCBRS for patients with sufficient clinical resemblance to the typical phenotype, which is almost always due to a de novo variant in the ATPase domain of SMARCA2; in their experience, all patients with a convincing clinical diagnosis assessed in the previous two years had a variant, and a patient of Wieczorek 2013 classified as NCBRS without a variant in BAF complex genes had an unconvincing picture [18]. The French PNDS (French National Diagnosis and Care Protocol) requires that a clinical diagnosis of NCBRS or Coffin-Siris syndrome be followed by a search for the genetic cause, but accepts maintaining the clinical diagnosis when the picture is convincing and no pathogenic variant is found, because introns, 5′ and 3′ untranslated regions and GC-rich regions of the genes involved are difficult to study completely, a limitation that genome sequencing may overcome [15]. The same protocol leaves the choice of test (chromosomal microarray, ID gene panel, exome or genome) to the experience of the prescriber and asks that the pathogenicity of the variant be discussed between the molecular biologist and the clinician before the results consultation [15]. The acceptance of a clinical diagnosis without a variant diverges from GeneReviews, for which the diagnosis is established by identifying a pathogenic or likely pathogenic variant (Table 7.6) [1, 15].

Algorithm

When the phenotype suggests NCBRS, GeneReviews proposes single-gene testing (SMARCA2 sequencing for missense, nonsense, splicing variants and small insertions or deletions and, if negative, gene deletion/duplication analysis) or a multigene panel including SMARCA2 and the genes of the differential diagnosis, in particular those of Coffin-Siris syndrome [1]. When the phenotype is indistinguishable from other disorders with dysmorphism and ID, the indicated test is the exome or genome, in agreement with the ACMG recommendation to use exome or genome as a first- or second-tier test in children with ID or multiple congenital anomalies; most variants described are coding and detectable by exome [1]. Trio exome identified the cause in 9 of 10 patients with a provisional diagnosis of Coffin-Siris or NCBRS and revealed unexpected diagnoses (Wiedemann-Steiner, Kabuki, Adams-Oliver, de novo GRIN2A and SHANK3 variants), as well as SMARCA2 and SMARCA4 deletions not detected on a previous panel [59]. In an Italian cohort of 1161 patients with ID, strong clinical suspicion of NCBRS or Coffin-Siris in 11 patients was confirmed in all (8 de novo SMARCA2 missense variants, 2 ARID1B frameshift and 1 SMARCB1 missense) [26]. Figure 5.1 summarises the pathway.

Before the widespread use of exome, Sousa and Hennekam proposed a stepwise approach: clinical examination of the limbs with hand radiographs, because small distal phalanges or small fifth-digit nails have never been described in NCBRS and point to Coffin-Siris syndrome; Sanger sequencing of exons 15 to 25 of SMARCA2 (first exons 18, 24 and 25, then 15, 19 and 23 and finally 16, 17, 20, 21 and 22); if negative, a search for in-frame deletions of this region by multiplex ligation-dependent probe amplification (MLPA) or detailed array comparative genomic hybridisation (array CGH); then, ARID1B sequencing and a search for its deletions; clinical reassessment and, if the phenotype remained compatible, trio exome in a research context; where available, diagnostic exome from the outset, targeted to the BAF complex genes or the ID genes [18]. This sequence now has mainly historical value and value for settings without access to exome; the authors stress that, for SMARCA2, the distinction between pathogenic and non-pathogenic variants is difficult without parental DNA [18].

Figure 5.1. Diagnostic algorithm (text)

Yield of molecular tests

Table 5.1. Molecular tests in NCBRS

These values are proportions of variants detected by each method and not clinical sensitivity [1]. In patients with a clinical diagnosis of NCBRS, SMARCA2 variants are found in more than 80% [19]; in the discovery series, in 36 of 44 [24]. In a study of 46 individuals with Coffin-Siris or NCBRS (exome, 23-gene SWI/SNF panel and molecular karyotype), 60% (28 of 46) had a variant in ARID1A, ARID1B, SMARCB1, SMARCE1, SMARCA2 or PHF6, and ARID1B and SMARCB1 variants were found in patients with an initial diagnosis of NCBRS [31]. In a population selected for epilepsy, a 518-gene panel or exome identified a heterozygous SMARCA2 variant in 24 of 4006 Chinese patients (about 0.6%); one of the 22 variants was an exon deletion, whose detection method the article does not describe [34].

DNA methylation signature (episignature)

The BAFopathy episignatures in peripheral blood overlap between Coffin-Siris subtypes (ARID1B, SMARCB1, SMARCA4) and NCBRS, sometimes more than within Coffin-Siris syndrome itself, and a machine learning model resolved ambiguous cases and reclassified VUS [27]. The specific NCBRS-SMARCA2 episignature has 429 differentially methylated CpG sites (8 patients versus 23 neurotypical controls, MethylationEPIC array); on validation, it correctly classified 8 of 8 patients with ATPase/helicase domain variants and 94 of 94 independent controls (sensitivity and specificity of 100%; the article abstract gives 96 controls), reclassified as pathogenic 3 ATPase domain VUS in patients with clinical NCBRS (p.(Thr880Ile), p.(Arg855Pro), p.(Gly853Val)) and as benign 5 variants proximal to the domain in patients without NCBRS [28]. The methylation signature is indicated when there are suggestive findings without a pathogenic SMARCA2 variant or with a VUS in the gene; it is an adjunct and does not replace molecular testing, a normal profile does not exclude the diagnosis (sensitivity depends on the domain affected) and the signature is not exclusive, because the BIS episignature is shared with Helsmoortel-Van der Aa syndrome [1, 13, 16, 28]. BIS has its own episignature of 163 CpG sites (4 patients versus 40 controls), distinct from that of NCBRS, which the authors propose as an additional test to classify SMARCA2 VUS; blood transcriptomics and qPCR of SMARCA2 expression were used only in research [10]. The French PNDS accepts both clinical uses of the methylation signature: deciding the pathogenicity of a VUS and identifying the BAFopathy episignature in patients without a pathogenic variant after sequencing of the known genes [15].

In clinical practice, the EpiSign implementation in a European and North American network validated 211 samples and tested 207 subjects, with 27.6% positive results: 35.3% (48 of 136) in the cohort targeted at ambiguous findings or VUS and 11.3% (8 of 71) in the screening cohort, and 1.9% inconclusive [61]. At Erasmus MC (298 patients, 2019–2023), targeted analysis was positive in 18% of VUS (10 of 55), 91% of likely pathogenic variants (10 of 11) and 89% of pathogenic variants (8 of 9); full analysis found a signature in 9.0% (20 of 223), with full or partial phenotypic match in 55% and a confirmed diagnostic yield of 4.0% [62]. In 400 patients with a neurodevelopmental disorder studied by exome and EpiSign, 17% had variants in chromatinopathy genes, and 43% of the 60 individuals analysed by methylation had a concordant specific episignature [63].

Other investigations

Echocardiography is indicated if heart disease is clinically suspected and EEG if seizures are suspected [1]. Hand radiographs show the changes described in section 3, but may be normal in mild forms, as in Ejaz's patient at 6.5 years [1, 21]. Karyotype and microarray are usually normal (normal in 14 patients of the 2009 series), and brain MRI is usually normal [23, 43].

Limitations of the tests

A negative result does not exclude the diagnosis. Sequencing does not detect the in-frame intragenic deletions of tens of kb that cause NCBRS, which require CNV analysis; exome detected SMARCA2 deletions not identified on a capture panel [1, 25, 59]. CNV analysis is also needed for 9p24.3 duplications with a Coffin-Siris-like phenotype [33]. Most variants described are coding; the sources consulted do not describe the frequency of deep intronic variants or complex structural rearrangements in NCBRS [1]. Low-level parental mosaicism (2.8% of alleles in sperm) was not detected in the parents' blood or saliva and was only demonstrated by deep amplicon sequencing of sperm [32]. The methylation signature depends on the position of the variant and may give intermediate profiles outside the ATPase/helicase domain [28]. A SMARCA2 splicing variant classified as likely pathogenic in a child with macrocephaly and megalencephalic leukoencephalopathy with subcortical cysts (MLC) showed that molecular findings should complement, not replace, clinical and radiological evidence [60].

Content of the molecular report

The report should state the gene, the reference transcript with version (NM_003070.5, MANE Select), the variant in HGVS nomenclature (c. and p.), the genomic coordinate with the genome version, the zygosity (heterozygous), the ACMG/AMP class and its justification, the origin (de novo or inherited) when the parents have been studied and the protein domain affected, which has implications for the distinction between NCBRS and BIS [1, 7, 10]. ClinVar classifications for BIS and NCBRS depend on the submitter and some variants are assigned to both entities [5].

VUS and segregation

In the presence of a VUS, parental testing to confirm de novo origin (with confirmed paternity and maternity), reassessment of the phenotype over time and the methylation signature are the tools available [1, 28]. De novo origin is not sufficient when the phenotype is discordant [60]. In the BIS series, the combination of de novo origin (PS2), absence from population databases (PM2) and computational predictions (PP3) supported the classification as likely pathogenic of variants in codons with VUS in ClinVar, but the scattered variants of the 6 individuals with a non-specific phenotype remained unreclassified, because neither the clinical picture nor transcriptomics distinguished them from the controls [10].

Prenatal and preimplantation diagnosis

Once the variant has been identified in an affected relative, prenatal diagnosis and preimplantation genetic testing (PGT) are technically possible [1]. Without a known variant, prenatal diagnosis depends on non-specific ultrasound findings; in the only fetus described, amniocentesis with karyotype and microarray was normal, rapid exome failed due to maternal cell contamination and the diagnosis was only made after termination of pregnancy [56]. The French PNDS provides that the detection of a variant in a BAFopathy gene in a fetus, increasingly frequent with prenatal genomic tests requested for malformations, should be communicated in a genetics clinic dedicated to the possible consequences of the variant, which may lead to a decision on termination of pregnancy for medical reasons [15].

6. Differential Diagnosis

Table 6.1. Differential diagnosis of NCBRS

AD, autosomal dominant; AR, autosomal recessive; GPI, glycosylphosphatidylinositol; CNV, copy number variation.

Coffin-Siris syndrome and the other BAFopathies

The overlap between NCBRS and Coffin-Siris syndrome probably results from both involving genes of the SWI/SNF complex, and the episignatures show a functional continuum between the two [1, 27]. The face may be similar, and the digital findings are the most helpful: NCBRS has prominent interphalangeal joints and no hypoplasia of the fifth digits, and the hair is sparse rather than showing the hypertrichosis of Coffin-Siris syndrome [1, 30]. There are people with molecularly confirmed Coffin-Siris syndrome and digital findings more compatible with NCBRS, such as monozygotic twins with an intragenic ARID1B deletion and an overlapping gestalt confirmed by computerised facial analysis, and the authors therefore propose studying ARID1B in patients with an NCBRS phenotype without a SMARCA2 variant [1, 64]. It has been suggested that other genes of the SWI/SNF complex explain cases suggestive of NCBRS without a SMARCA2 variant [1, 57]. Coffin-Siris syndrome has its own genetic classification, by gene. In the 2014 series, the phenotype of NCBRS proved narrower than that of Coffin-Siris syndrome; small distal phalanges or small fifth-digit nails were never observed in NCBRS and point to Coffin-Siris syndrome, where SMARCA4 variants are also missense or in-frame deletions, mostly in the ATPase domain, and in the young child Williams syndrome was the initial suspicion in several patients [18].

Coffin-Lowry syndrome

It shares the coarse facies that becomes more marked over time and the severe delay, but X-linked inheritance, the large, soft hands with tapering distal phalanges and kyphoscoliosis point to the diagnosis, which radiographs of the hands, chest and spine may support [51].

Other syndromes with ID and dysmorphism

Williams-Beuren syndrome is distinguished by the cognitive profile with relative strength in language and the characteristic heart disease, Cornelia de Lange syndrome by the face and limb changes and Börjeson-Forssman-Lehmann syndrome by the PHF6 gene [1, 30]. Exome in patients with a provisional diagnosis of Coffin-Siris or NCBRS revealed Wiedemann-Steiner, Kabuki and Adams-Oliver syndromes, and GRIN2A and SHANK3 variants [59]. A patient with micrognathia and trismus was described with a face similar to that of Treacher Collins syndrome [49]. The French PNDS organises the differential of the BAFopathies by mode of presentation: in the infant with hypotonia, feeding difficulties and particular morphological features, the RASopathies, especially Noonan syndrome; in the presence of nail hypoplasia with a neurodevelopmental disorder, DOORS (TBC1D24), Mabry and other glycosylphosphatidylinositol anchor deficiencies and Temple-Baraitser (KCNH1) syndromes, in which the affected nail is classically that of the thumb; in the presence of hypertrichosis with a neurodevelopmental disorder, Cornelia de Lange (NIPBL), Wiedemann-Steiner (KMT2A) and FHEIG (KCNK4) syndromes [15]. In NCBRS, hypertrichosis outside the scalp is present in 44% (22 of 50) and small nails affect the thumbs and halluces (2 of 33) or all nails (6 of 34), and never only the fifth digits (0 of 34), which keeps these diagnoses in the differential [18]. In the presence of fetal agenesis of the corpus callosum, the most frequent cause among syndromic forms is Coffin-Siris syndrome and not NCBRS, and the PNDS calls for a genetics clinic for the aetiological work-up [15].

Biotinidase deficiency

Seizures, hypotonia, developmental delay, skin changes and alopecia bring biotinidase deficiency close to NCBRS; the face, ataxia (possible in biotinidase deficiency and absent in NCBRS) and measurement of biotinidase activity distinguish them [1].

Trichorhinophalangeal syndrome

Fine, sparse hair with brachyphalangy and cone-shaped epiphyses is common to both entities; the bulbous pear-shaped nose and dystrophic nails point to trichorhinophalangeal syndrome, which may be part of the Langer-Giedion contiguous gene syndrome [57].

2q37 deletion syndrome

The 2q37 deletion syndrome (OMIM 600430), also called brachydactyly-mental retardation syndrome (BDMR) or Albright hereditary osteodystrophy-like syndrome, is a contiguous gene syndrome with variable expression according to the size and content of the deletion: type E brachydactyly of the metacarpals and metatarsals in about half of patients, short stature, mild to moderate ID, behavioural changes and facial dysmorphism, with many carriers without cognitive impairment [65]. Wilson and colleagues (1995) described 5 patients with an Albright hereditary osteodystrophy-like appearance (short stature, stocky build, ID, brachymetaphalangism and eczema), without soft-tissue ossification and with normal parathyroid hormone, calcium metabolism and Gsα protein, and found 2q37 deletions in all 5, cytogenetically visible and de novo in 4 [65]. HDAC4 haploinsufficiency explains the type E brachydactyly, but is not sufficient to cause ID: families have been described with inherited 2q37.3 deletions including HDAC4 transmitted with brachydactyly and short stature and normal cognition [65]. Some patients were referred with suspected Smith-Magenis syndrome [65]. Short metacarpals and eczema occur in both entities; possible familial inheritance, sometimes normal cognition and detection of the deletion by microarray point to the 2q37 deletion, whereas NCBRS requires SMARCA2 sequencing [1, 3, 65].

Differential of BIS

BIS is distinguished from NCBRS by blepharophimosis, the absence of the digital findings and the position of the variant, and shares phenotype and episignature with Helsmoortel-Van der Aa syndrome (ADNP) [10, 13]. In the comparison with NCBRS (Table 3.4), the features with most discriminating value are narrow palpebral fissures (71% versus 15.8%), a broad nasal bridge (71% versus 32.7%) and the absence of coarse facies (0 of 14 versus 76.6%), prominent interphalangeal joints (0 of 12 versus 84.5%) and a thick lower vermilion (0 of 14 versus 83.3%), together with less frequent sparse hair (29% versus 96.7%) [10]. Within the ADNP differential, the picture with blepharophimosis and epicanthic folds has been described in 4 individuals with variants in the bipartite nuclear localisation signal of the gene [10]. Rare ptosis (2 of 14) and the absence of epicanthus inversus and limb anomalies distinguish BIS from other recognisable conditions with blepharophimosis, such as blepharophimosis, ptosis and epicanthus inversus syndrome (BPES) [10]. An individual with p.(Asp534Tyr), in the BIS region, was described with a picture suggestive of Cornelia de Lange or Hallermann-Streiff syndrome, which broadens the clinical differential of variants in this region [10].

Dual diagnoses and discordant results

The debate around a patient with SMARCA2 (c.2556A>C) and POLG1 (c.3708G>T) variants showed that a second diagnosis may explain part of the phenotype and modify the therapeutic choice, because valproic acid is contraindicated in POLG-related disease [41, 58, 66, 67]. POLG-related disorders enter the differential. The dual diagnoses with nocturnal frontal lobe epilepsy due to CHRNB2 and with craniosynostosis due to TCF12 reinforce the value of trio exome and of reassessment of findings that NCBRS does not explain [37, 42]. The 55-kb intragenic deletion described by Tsurusaki and colleagues shows the boundary between the two syndromes: the patient was classified as Coffin-Siris syndrome, without hypoplasia of the fifth digits or sparse hair, and OMIM lists the variant as allelic to NCBRS [9].

7. Treatment and Risk Management Strategies

General principles

There is no cure or treatment targeting the mechanism of NCBRS or BIS; care is supportive and multidisciplinary, coordinated by a reference physician, with early developmental intervention, treatment of epilepsy, feeding support and surveillance of the organs affected [1]. The GeneReviews recommendations are based on the authors' personal experience, because there are no published clinical practice guidelines [1]. The French PNDS (French National Diagnosis and Care Protocol) confirms that there is no specific treatment, recommends no particular diet and states that vaccinations are not contraindicated; care aims at the best possible intellectual development, the prevention of complications and the detection of comorbidities, should be early, sustained and continuous, and the frequency of consultations is adjusted to each patient, with an annual consultation for stable patients [15]. Follow-up is shared between the general practitioner or paediatrician, who coordinates local care and watches for the onset of epilepsy, scoliosis, behavioural changes, hypothyroidism and diabetes, and a hospital developmental specialist, who sees the patient once a year [15]. In addition to the specialties in Table 7.1, the protocol assigns roles to neonatology and paediatric intensive care (neonatal period, undernutrition, cardiac complications), paediatric visceral surgery (malformations), dermatology (changes of the skin appendages, hyperkeratosis and hyperhidrosis), nephrology and urology (renal and urinary malformations and vesicoureteric reflux), orthodontics, orthoptics and hearing aids, social services (liaison between the family and the teams and guidance on social support) and to the medico-social structures for early intervention, special education and adult accommodation [15].

Guidelines and level of evidence

The only clinical guidance covering NCBRS is the French PNDS for Coffin-Siris and Nicolaides-Baraitser syndromes, drafted by the reference centre Déficiences intellectuelles de causes rares (Pitié-Salpêtrière Hospital and Trousseau Hospital, Paris) and the DéfiScience network according to the 2012 HAS method, with a multidisciplinary working group and a reading group that included the French patient association, with text dated September 2021 and publication by the HAS on 4 November 2021 [15]. It applies to fetuses, infants, children, adolescents and adults and covers the signs of suspicion, genetic diagnosis, comorbidities, follow-up, transition to adulthood and genetic counselling; it is an expert consensus without grading of evidence, most of the data on which it rests come from ARID1B-associated Coffin-Siris syndrome, and it contains no drug doses or recommendations on antiepileptic drugs [15]. No international guideline or BIS-specific consensus was found, nor a ClinGen actionability assessment; ERN ITHACA, the relevant European reference network, has published its own guidelines for Beckwith-Wiedemann syndrome (2018), valproate embryopathy (2018), Cornelia de Lange syndrome (2019), Phelan-McDermid syndrome (2023), Rubinstein-Taybi syndrome (2024), spinal dysraphisms (2025), Kleefstra syndrome (2026) and transition to adult care (2026), and none of these nor the guidelines it endorses or the patient journeys cover NCBRS, BIS or the BAFopathies [1, 16, 17, 68]. For young people with NCBRS moving to adult care, the ERN ITHACA transition guideline of March 2026 is the European guidance applicable to all rare neurodevelopmental disorders [17]. The references for management are GeneReviews and the French PNDS, which agree on the symptomatic and multidisciplinary nature of care and diverge on the points gathered in Table 7.6; the therapeutic evidence available is observational (retrospective series and case reports) [1, 15, 43]. As the PNDS has no doses, the reference doses remain those of the official label (Table 7.3) [15].

Initial assessment after diagnosis

Table 7.1. Recommended assessments after the diagnosis of NCBRS

Treatment of manifestations

Table 7.2. Treatment of the manifestations of NCBRS

Development, motor skills and communication

From 0 to 3 years, GeneReviews recommends early intervention (OT, PT, speech and feeding therapy, infant mental health, special educators); from 3 to 5 years, preschool education with an individual education plan; at all ages, follow-up by a developmental paediatrician, with annual review of the education plan, the least restrictive environment possible and a transition plan in adolescence; the services cited are those of the United States and vary between countries [1]. PT seeks to maximise mobility and reduce late orthopaedic complications, with wheelchairs, walkers, orthoses and adapted bath chairs, and OT works on fine motor skills for feeding, hygiene, dressing and writing [1]. Augmentative and alternative communication (AAC) should be assessed, from picture exchange to speech-generating devices; AAC does not hinder but rather supports speech development [1]. Occupational therapy and speech therapy started before 1 year accompanied the favourable course of the patient of Ejaz and colleagues [21].

Feeding

Oromotor dysfunction should be assessed at every visit; choking, poor weight gain, frequent respiratory infections or unexplained food refusal justify clinical feeding assessment or a radiological swallowing study [1]. Feeding therapy (OT or speech therapist) applies when the oral route is safe; thickening or cooling food may help, and a nasogastric tube or gastrostomy is reserved for severe dysfunction [1]. Feeding difficulties may be prolonged: one patient had a gastrojejunostomy from 4 months to 12 years and kept a preference for purées and an increased gag reflex [21]. According to the French PNDS, growth delay and feeding difficulties justify gastroenterological assessment, which may include swallowing studies, upper gastrointestinal imaging or endoscopy, and an early dental assessment helps to identify crowding and problems linked to delayed dentition [15].

Behaviour

ASD interventions are used, including applied behaviour analysis (ABA) with a certified analyst, and the developmental paediatrician may propose behavioural strategies or medication, for example for attention deficit hyperactivity disorder (ADHD) [1]. The sources do not describe psychotropic drugs with specific efficacy in NCBRS. In mild forms with self-injury after puberty, the focus should be on behaviour modification [21].

Communication of the diagnosis and therapeutic education

The French PNDS recommends that the diagnosis be communicated in a private setting, with sufficient time and preferably to both parents, by the physician who requested the molecular test, possibly accompanied by another team member (psychologist, genetic counsellor or nurse) [15]. The consultation explains the diagnosis and the results, the disease and its mode of transmission to siblings, the need for regular multidisciplinary follow-up and the care plan; the presence of a psychologist is considered essential during the consultation, immediately after it or at follow-up, and the family is given the contact details of patient associations, the choice of contacting them being left to the family [15]. Therapeutic education, which seeks to give the patient and carers the skills to manage life with a chronic disease, has no particularities in the BAFopathies compared with other developmental disorders with ID, orthopaedic risk and feeding difficulties; the protocol refers to the 2016 collective expert report of the Institut national de la santé et de la recherche médicale (Inserm) on intellectual disabilities [15].

Epilepsy: data on response to antiepileptic drugs

The international retrospective study of the NETRE network (Network Therapy of Rare Epilepsies) and a parent group gathered 25 patients with NCBRS, 23 with epilepsy, by electronic questionnaire [43]. The most used antiepileptic drugs were valproic acid and levetiracetam (12 of 20 each), phenobarbital (8 of 20), topiramate (5 of 20) and carbamazepine (5 of 20) [43]. A reduction of more than 50% in seizure frequency occurred with levetiracetam in 9 of 12, phenobarbital in 6 of 8, topiramate in 4 of 5 and valproic acid in 9 of 12; temporary seizure freedom (more than 6 months) occurred with levetiracetam in 4 of 12, phenobarbital in 3 of 8, topiramate in 1 of 5 (only in combination with phenobarbital and nitrazepam) and valproic acid in 4 of 12 [43]. There was worsening with lamotrigine in 2 of 4 patients and, in isolated cases, with levetiracetam, phenobarbital and valproic acid; the ketogenic diet and vagus nerve stimulation reduced seizures in 1 of 2 patients each [43]. The authors' conclusion is that the response is mainly initial, in particular to levetiracetam and valproic acid, and that complete seizure freedom is very rare [1, 43]. In the Chinese cohort of Yang and colleagues, all 24 patients were treated with antiepileptic drugs, and age at last follow-up ranged from 1 year and 5 months to 11 years (the treatment section of the article gives 5 months to 12 years) [34]. At last follow-up, 10 of 24 (41.7%) had been seizure-free for 8 months to 5 years; the abstract and discussion report 11 of 24 (45.8%) seizure-free on valproate and/or levetiracetam, but the article's patient table shows 10 seizure-free patients, one of them on adrenocorticotropic hormone (ACTH) and not on those drugs (Table 3.3) [34]. The authors considered effective the drug with which the patient became seizure-free for about 6 months: levetiracetam monotherapy in 3 patients, valproate monotherapy in 2 and the combination of valproate with levetiracetam in the others; in one patient treated with valproate, levetiracetam and clobazam, there was no recurrence after clobazam was reduced [34]. In another 6 patients, seizure frequency decreased by more than 50%, most often with levetiracetam (6), valproate (5) and oxcarbazepine (2) and, in isolated cases, with clonazepam, topiramate and nitrazepam; 2 patients treated with topiramate had no significant benefit, and no drug is marked as worsening seizures [34]. Infantile epileptic spasms syndrome was treated with ACTH, attributed in the text to patient 23 and in the table to patient 24, who became seizure-free at 1 year [34]. The article does not give doses, serum levels or treatment durations [34]. The proportion of seizure-free patients contrasts with the rarity of complete remission in the Hofmeister study; differences in selection (cohort identified through epilepsy panels, with a poorly described extraneurological phenotype), age and duration of follow-up preclude comparison of the two studies [34, 43]. Yang and colleagues' summary of the Hofmeister study also includes responses to less used drugs (acetazolamide in 1 of 1, clobazam in 2 of 3, felbamate in 1 of 1, lacosamide in 2 of 2, oxcarbazepine in 3 of 4, phenytoin in 2 of 3 and zonisamide in 1 of 1), in addition to the responses to levetiracetam, phenobarbital, topiramate and valproate already cited, and the first seizure occurred at a mean of 20.9 months (7 to 108 months) [34, 43]. In a cohort of 50 patients with variants in genes of chromatin remodelling complexes (including SMARCA2), 56.3% of those with epilepsy needed 3 or more antiepileptic drugs and 69% had been seizure-free for more than 1 year at last follow-up [69]. In Ejaz's patient, carbamazepine started at 21 months was followed by cessation of seizures at 4 years [21].

Lamotrigine worsened seizures in half of the patients who received it in the Hofmeister series, a number too small to contraindicate it, but which justifies surveillance if it is used [43].

Dosing of the antiepileptic drugs with most data in NCBRS

No source publishes doses specific to NCBRS, and there is no guideline with doses for this condition; the reference doses are therefore those of the official label for the seizure type, which constitutes an extrapolation: NCBRS is not an authorised indication, but use is within the label when the seizure type and age coincide with the approved indication [43, 58, 70, 71]. The only report with doses describes, in a boy with seizures from 1 year and 3 months, levetiracetam 187.5 mg twice daily, then 187.5 mg and 250 mg, without control, followed by the addition of valproic acid oral solution (2 ml twice daily, increased with weight to 3.5 ml twice daily; concentration and dose in mg/kg not given), with seizure reduction to 1 to 2 per month; this regimen is an individual report and not a recommendation [11]. The Portuguese Summary of Product Characteristics (SmPC) of valproate (Depakine®) and the Brazilian package inserts of Depakote® and Depakene® serve as comparison, with the differences described after Table 7.3; for levetiracetam, the SmPC applicable in Portugal is that of the European public assessment report (EPAR) for Keppra®, as it is a centralised authorisation [71–74].

Table 7.3. Levetiracetam and valproic acid: labelled doses applicable to the seizure types of NCBRS

The main differences between labels concern levetiracetam from 6 months to 4 years, for which the target of the US Food and Drug Administration (FDA) is 25 mg/kg twice daily (recommended daily dose of 50 mg/kg, which may be reduced if not tolerated; no paediatric maximum dose defined in the source beyond the recommended dose) and that of the European Medicines Agency (EMA) 30 mg/kg twice daily, and primary generalised tonic-clonic seizures, approved as adjunctive therapy from 6 years by the FDA and from 12 years in the EU [70, 71]. The FDA label gives, for children aged 1 to <6 months, 7 mg/kg twice daily up to 21 mg/kg, for 4 to <16 years the target of 30 mg/kg twice daily and, for tablets in children of 20 to 40 kg, a start of 250 mg twice daily up to a maximum of 750 mg twice daily [70]. In paediatric renal impairment (<50 kg), the EMA reduces the dose to 5 to 15 mg/kg twice daily with creatinine clearance (CrCl) of 30 to 49 ml/min/1.73 m² and to 5 to 10 mg/kg twice daily below 30, with specific doses on dialysis [71]. With valproate, children aged 3 months to 10 years have clearance 50% higher than adults, and the label asks for extremely cautious use and monotherapy below 2 years [58]. The behavioural warning for levetiracetam is relevant in a syndrome with frequent behavioural problems [1, 70]. In Brazil, the active registrations with the Brazilian Health Regulatory Agency (ANVISA) are Keppra® (UCB Biopharma Ltda., registration 1.2361.0083) and Keppra® XR (registration 1.2361.0093), Depakote® (divalproex sodium, Abbott Laboratórios do Brasil Ltda., registration 1.0553.0203) and Depakene® (sodium valproate/valproic acid, Abbott, registration 1.0553.0315) [75]. The healthcare professional package insert for Depakote®, approved by ANVISA on 9 April 2021, and that for Depakene® 50 mg/ml syrup, approved on 15 March 2019, differ from the FDA and EU labels on two points with consequences in NCBRS: both contraindicate use below 10 years of age, which excludes most children with NCBRS, whose epilepsy begins on average at around 21 months, and both set the maximum recommended dose at 60 mg/kg/day, with a start of 10 to 15 mg/kg/day (15 mg/kg/day in absences), increases of 5 to 10 mg/kg weekly, division above 250 mg/day and a therapeutic range of 50 to 100 µg/ml [1, 73, 74]. Use of valproate below 10 years in Brazil is therefore outside the conditions of the package insert. The 2021 Depakote® package insert predates the 2024 European measures on paternal exposure and states only that divalproex may affect male fertility, reversibly in most of the cases described [73, 76]. In Portugal, the SmPC of Depakine® 200 mg/ml oral solution (sodium valproate; Sanofi – Produtos Farmacêuticos, Lda.; first authorisation on 27 November 1989, renewed on 20 December 2011) differs from the FDA and Brazilian labels on points with consequences in NCBRS: it indicates valproate in generalised and partial seizures and in West and Lennox-Gastaut syndromes without a lower age limit, proposes faster titration (increases every 2 to 3 days), a usual paediatric dose of about 30 mg/kg/day with no fixed maximum, with close monitoring above 50 mg/kg/day, and a therapeutic range of 40 to 100 mg/l [58, 72, 73]. The SmPC identifies as the group at highest risk of severe liver injury children under 3 years with severe epilepsy, brain lesions, mental retardation or metabolic disease, especially on polytherapy, a profile that matches that of many children with NCBRS, and contraindicates valproate, in addition to POLG disease, in acute or chronic hepatitis, a personal or family history of severe hepatitis, hepatic porphyria, urea cycle disorders and uncorrected primary systemic carnitine deficiency [72]. The same SmPC incorporates the 2024 measures in men (information on the potential risk, effective contraception, including for the partner, during treatment and for at least 3 months afterwards, abstention from sperm donation, regular review and patient guide), describing the evidence as inconsistent and the causal role as uncertain; in girls, it asks parents to contact the specialist at menarche, that the need for valproate be reassessed annually from then on and that a switch to an alternative be sought before adulthood [72, 76]. For levetiracetam, the SmPC applicable in Portugal is that of the Keppra® EPAR [71]; in Brazil, the healthcare professional package insert for Keppra® was not found on the marketing authorisation holder's website.

Valproic acid, POLG and restrictions in women

A patient with NCBRS and an additional POLG1 variant had hepatotoxicity on valproic acid, which led the authors to discuss a POLG-related effect [41]. The FDA label contraindicates valproate in patients with mitochondrial disease due to POLG variants and in children under 2 years with suspected POLG-related disease, and recommends POLG screening according to clinical practice [58]. In the EU, since the review by the Pharmacovigilance Risk Assessment Committee (PRAC) endorsed by the Coordination Group for Mutual Recognition and Decentralised Procedures (CMDh) on 23 March 2018, valproate is banned in pregnancy for epilepsy unless there is no effective alternative, and contraindicated in girls and women of childbearing potential outside a pregnancy prevention programme (assessment of childbearing potential, pregnancy tests, counselling, effective contraception, specialist review at least annually and an annual risk acknowledgement form); treatment should not be started unless alternatives are not suitable, including in girls before puberty [77]. On 12 January 2024, the PRAC also recommended precautionary measures in men, following a retrospective observational post-authorisation study (EUPAS34201) with registries from Denmark, Norway and Sweden: in the children of fathers treated with valproate in the 3 months before conception, the pooled adjusted hazard ratio for neurodevelopmental disorders was 1.50 (95% confidence interval 1.09 to 2.07) compared with fathers treated with lamotrigine or levetiracetam, with an adjusted cumulative risk of about 5% versus about 3% and no difference in congenital malformations [76]. The PRAC could not establish whether the association was causal, because of differences in indication and follow-up between the groups, and the study did not assess discontinuation more than 3 months before conception [76]. The recommended measures are initiation and supervision of treatment by a specialist, information to the patient, discussion of effective contraception for the patient and partner during treatment and for at least 3 months after discontinuation, regular review of the appropriateness of valproate, especially when conception is planned, and abstention from sperm donation during the same period; the recommendations went forward for a CMDh position and for updating of the product information [76]. As in NCBRS epilepsy begins on average at around 21 months and patients are not known to reproduce, the practical issue is mainly that of girls who reach adolescence on valproate, for whom the annual review applies [1, 77].

Regulatory table

Table 7.4. Regulatory status of the drugs cited

Clinical trials and treatment studies

There is no therapeutic trial in NCBRS or BIS: the ClinicalTrials.gov search in October 2026 found for «Nicolaides-Baraitser» only the observational rare disease registry CoRDS (Coordination of Rare Diseases at Sanford), and the studies with «SMARCA2» are all in oncology (SMARCA2 degraders in SMARCA4-deficient cancers); the search «BAFopathy» returned no studies [14, 40]. Table 7.5 brings together the studies that inform treatment.

Table 7.5. Studies with treatment data in NCBRS

LEV, levetiracetam; VPA, valproic acid; PB, phenobarbital; LTG, lamotrigine; CI, confidence interval.

These studies are observational and small, without a control group, with non-standardised doses and combinations and with data collected by questionnaire or record review, so the response percentages cannot be compared between drugs or between studies; in Yang 2025, some counts differ between the abstract, the text and the patient table [34, 43]. The response rates are absolute (patients with seizure reduction among those treated with each drug) and not relative reductions versus a comparator. The Chen cohort mixes several genes, and the proportion of patients with SMARCA2 is not given in the abstract [69]. The negative results (worsening with lamotrigine and isolated cases with other drugs) are included [43].

Developments in the last 6 to 12 months

Between October 2025 and October 2026 there were no therapeutic developments, authorisations or regulatory opinions for NCBRS or BIS [14]. The publications of the period are case reports and retrospective cohorts (low level of evidence): bilateral congenital glaucoma in a child, siblings with NCBRS due to probable paternal germline mosaicism and a dual diagnosis of craniosynostosis, a dual diagnosis with epilepsy due to CHRNB2, a discordant SMARCA2 variant in a patient with MLC, the cohort of 50 patients with variants in chromatin remodelling genes and the utility of methylation signatures in chromatinopathies [37, 42, 54, 60, 63, 69]. In oncology, the first selective SMARCA2 degrader in humans (PRT3789) was described in cancers with SMARCA4 variants; this is preclinical and phase 1 research with no application to NCBRS [14, 40]. In basic research, the team of the neurogenetics laboratory of the UCL Queen Square Institute of Neurology announced in March 2026 that it is seeking to identify the most frequent SMARCA2 variants in order to introduce them into mouse models under development, that it is studying patient-derived induced pluripotent stem cells (iPSC) and that it is preparing a patient registry; this is an institutional announcement about preclinical research, with no published results [78].

Skin and hair

In a single case, severe eczematous dermatitis was treated with mometasone cream twice daily during flares, 0.03% tacrolimus ointment for maintenance, wet-wrap dressings, oral antihistamines and barrier-repair moisturisers, and the alopecia with topical mometasone solution at night, ketoconazole shampoo and a multivitamin; the abstract does not describe efficacy, and these measures are not routine [48].

Anaesthesia and perioperative care

Micrognathia, limited mouth opening and cleft palate make the airway potentially difficult. In a 22-month-old child undergoing palatoplasty, the authors recommended a trial laryngoscopy with a videolaryngoscope under spontaneous ventilation and deep inhalational anaesthesia, before muscle relaxant, to detect a difficult airway while preserving protective reflexes [50]. In a 9-year-old child with trismus, oral intubation with a videolaryngoscope failed and nasal intubation with a flexible fibreoptic scope succeeded [49]. Dental extractions, frequent in NCBRS, are common occasions for general anaesthesia [1, 49]. The sources consulted have no specific recommendations on the perioperative management of antiepileptic drugs in NCBRS.

Growth

The sources consulted do not describe the use of growth hormone in NCBRS; short stature is monitored through growth parameters and nutritional status [1].

Agents to avoid

Valproic acid should be avoided in patients with suspected or confirmed POLG-related disease, and in girls should only be started if alternatives are not suitable [41, 58, 77]. The Portuguese SmPC recommends avoiding salicylates in combination with valproate below 3 years and does not recommend the combination with carbapenems, which reduces valproic acid concentration [72]. Lamotrigine worsened seizures in 2 of 4 patients in one series [43]. Because of the difficult airway, the authors of an anaesthetic report advise against muscle relaxant before trial laryngoscopy confirms that intubation is possible [50].

Pregnancy and lactation

People with NCBRS are not known to reproduce, and the sources have no data on pregnancy or lactation in affected women [1]. The restrictions on valproate in women of childbearing potential apply to treated adolescents [77]. The Portuguese SmPC quantifies the risk that justifies them: major congenital malformations in about 11% of children of women with epilepsy exposed to valproate monotherapy, versus 2 to 3% in the general population, with dose-dependent risk and no safe threshold, and early developmental delays in up to 30 to 40% of preschool children exposed in utero to monotherapy [72]. In boys who reach adolescence on valproate, the 2024 European precautionary measures for men (regular review, information on the potential risk to offspring, contraception and no sperm donation) apply equally, although reproduction has not been described in NCBRS [1, 76].

Transition to adult care

Transition should begin at around 10 years, with realistic plans for adult life, using as a model the American Epilepsy Society document on the transition from paediatric epilepsy to adult care [1]. Adults are probably underdiagnosed, and the natural history in adulthood is known only from isolated cases [1, 21].

Blepharophimosis-intellectual disability syndrome

The sources describe no specific treatment or BIS-specific management recommendations; the general principles of supportive care for ID, hypotonia and behavioural changes apply [1, 4]. The original series makes no management recommendations; the problems it described most frequently (refractive errors, recurrent respiratory infections, gastro-oesophageal reflux and feeding difficulties, dental anomalies in 8 of 12, joint contractures and lack of independent walking in 6 of 14) correspond to areas covered by the general surveillance of Table 11.1, applied here by extrapolation [10].

Divergences between the French PNDS and GeneReviews

Table 7.6. Divergences between the French PNDS (2021) and GeneReviews (2025)

P/LP, pathogenic or likely pathogenic; VUS, variant of uncertain significance; CNV, copy number variation.

8. Reproductive Options and Family Counselling

Mode of transmission

NCBRS is autosomal dominant and, to date, all diagnosed patients have a de novo variant or, rarely, one inherited from an unaffected mosaic parent [1]. In the Van Houdt and Wolff series (2012), the variants were de novo in all cases with parental samples available [3]. In the 2014 series, no parent studied had the child's variant, there was no familial recurrence apart from a pair of monozygotic twins and no consanguinity, and mean ages at birth were 32.9 years for the father and 30.1 years for the mother, the significance of which the authors could not assess because of the mixed origin of the families [18]. BIS also results from de novo variants, demonstrated in the 14 patients of the original series [6, 10]. Across the 20 individuals of the series, the article counts 19 de novo variants; the non-specific phenotype group included a mother and her son (individuals 18 and 17), both with a SMARCA2 variant, and the mother was the only individual without parents available for study; the text does not detail by individual the count of 19 de novo variants, which is difficult to reconcile with maternal transmission in this family [10].

Parents

Molecular testing of both parents for the variant identified in the proband is recommended; it is very unlikely to find the variant in constitutional heterozygosity in an apparently asymptomatic parent [1]. If the variant is not found in either parent, with confirmed paternity and maternity, the explanation is a de novo variant or parental gonadal (or somatic and gonadal) mosaicism; leucocyte DNA may not detect all somatic mosaicism and does not detect mosaicism confined to the gonads [1]. Maternal mosaicism in a Chinese case and low-level paternal germline mosaicism have been documented: in the family described by Pan and colleagues, the variant c.553C>G p.(Gln185Glu) was present in 2.8% of alleles in paternal sperm, absent from the blood and saliva of both parents, and was transmitted to the proband and to a subsequent pregnancy [32, 36]. In the cohort of Yang and colleagues, none of the 24 patients had a family history, 23 had de novo variants (the abstract gives 21 de novo variants) and the remaining one had the same variant c.553C>G p.(Gln185Glu), with 2.8% paternal germline mosaicism quantified by deep amplicon sequencing of sperm DNA [34]. In 2026, two siblings with NCBRS and the same novel variant p.(Ile833Phe) led to suspicion of paternal germline mosaicism, which may explain higher-than-expected recurrences [37].

Recurrence risk

Table 8.1. Recurrence risk by scenario

Cascade testing and carriers

As no asymptomatic carriers have been described and there is no recessive inheritance, cascade testing is limited to the parents, to clarify the origin of the variant, and the concept of carrier does not apply [1]. Deep sequencing of sperm can document paternal germline mosaicism when blood is negative [32].

Prenatal diagnosis and preimplantation genetic testing

The ideal time to estimate the risk and discuss prenatal diagnosis or PGT is before pregnancy, and counselling should be offered to the parents of affected children [1]. Once the variant has been identified, prenatal diagnosis and PGT are possible; perspectives on their use differ between professionals and families, and the decision is personal [1]. In the family with paternal germline mosaicism, preimplantation genetic testing for monogenic disorders (PGT-M) with in vitro fertilisation and intracytoplasmic sperm injection (IVF/ICSI) biopsied 8 blastocysts, analysed by karyomapping and Sanger sequencing; an unaffected embryo was transferred, amniocentesis at 16 weeks (CNV sequencing, karyotype and Sanger) confirmed the absence of the variant and a healthy girl was born, the first live birth described after PGT for low-level paternal germline mosaicism [79]. Without a known familial variant, ultrasound findings are not specific, and in the only fetus described the prenatal exome failed due to maternal contamination [56]. The French PNDS considers prenatal diagnosis targeted at the proband's variant an option to be discussed with the parents, because of the risk of germline mosaicism and the severity of the disability, and recommends parental testing to detect somatic mosaicism, noting that a negative result does not exclude germline mosaicism, which puts the risk at about 1% [15]. When the variant is inherited from a parent with an attenuated form, the 50% risk should be communicated and prenatal diagnosis discussed in the same way [15]. In the French framework, preimplantation diagnosis is reserved for cases in which one of the parents has proven germline mosaicism, a condition more restrictive than that of GeneReviews, which accepts it once the familial variant is known; the family of Pan and colleagues, with 2.8% of mutated alleles in sperm, met this criterion [1, 15, 32].

Counselling of minors and the family

As affected people usually have severe ID and are not known to have offspring, counselling is directed mainly at the parents and adult siblings; the healthy siblings of a proband with a de novo variant have a presumably low risk [1]. For minor siblings, the sources consulted describe no indication for testing. Renewed genetic counselling, for example for family planning, is part of follow-up [1].

9. Special Considerations: Population Diversity and Epidemiology

Prevalence and what the figures measure

The prevalence of NCBRS is unknown and estimated as extremely low; the figures available are counts of published cases and not population studies [1]. GeneReviews mentions fewer than 100 people described and at least 80 with an identified pathogenic variant (up to 2021), and Orphanet records a worldwide point prevalence class of less than 1/1,000,000, which is a ceiling and not a measurement, with a count of 61 cases or families [1, 2, 11, 16]. Orphanet attributes this count to the review of the BAFopathies by Kosho and colleagues (PMID 25169878), and not to the series of 61 cases of Sousa and Hennekam (PMID 25169058) cited by GeneReviews [12, 18]. Zhang and colleagues report a prevalence of 1/1,000,000 cited from an earlier source, without their own measurement [11]. A University College London (UCL) announcement of March 2026 mentions more than 300 confirmed cases worldwide; this is an unpublished count, without method or denominator, which indicates a number of diagnoses well above that in the literature but does not allow prevalence to be estimated [78]. For BIS, the only count is that of 14 cases or families in the original series, with an Orphanet class of less than 1/1,000,000 and no prevalence estimate [4, 16]. The series was gathered in European centres, with recruitment or funding in Italy, France, the United Kingdom, the Republic of Ireland and the Czech Republic, through Matchmaker Exchange, ERN ITHACA and the DDD, 100,000 Genomes Project and BUILD projects; the article does not give the distribution of patients by country nor does it include patients from Portugal or Brazil identified as such [10].

Frequency in ID and epilepsy cohorts

In a cohort of 1161 patients with ID from three Italian centres, 8 cases of NCBRS and 3 of Coffin-Siris syndrome were confirmed, and the authors estimated that these two conditions may account for up to about 1% of ID; this is the frequency in a population referred for ID and not the population prevalence [26]. By comparison, Coffin-Siris syndrome has more than 530 molecularly confirmed individuals in its registry and a rough estimate of 1/100,000 births [30]. The French PNDS starts from the same order of magnitude as the Italian series (up to 1% of unexplained ID) and from an ID prevalence of 2% to estimate Coffin-Siris syndrome at about 2/10,000, which would correspond to more than 100 births per year in France and more than 800 in Europe; this is an estimate by calculation, without measurement, 20 times higher than that of GeneReviews, and the protocol gives no figure for NCBRS, which it describes as much rarer [15, 30]. Of 4006 Chinese patients with epilepsy tested by panel or exome, 24 (about 0.6%) had a SMARCA2 variant; this is a frequency in a population selected for epilepsy and referred to specialised centres, not a prevalence [34].

Geographical distribution

Published cases come from Europe, North America, Asia, the Middle East and Latin America; no founder variants or populations with increased frequency have been described [11, 18, 21, 34, 37, 54, 80]. In China, the case of Zhang and colleagues was the fourth described in 2022, and in 2025 a cohort of 24 patients (16 male and 8 female) with epilepsy and SMARCA2 variants was published, gathered between March 2015 and March 2024 in epilepsy centres in Beijing, Wuhan, Zhengzhou, Urumqi and Quanzhou [11, 34]. In Latin America, the first molecularly confirmed case was a young Colombian adult with a mild to moderate phenotype [80]. In the 2014 series, the 61 patients came from five continents and various ethnic groups, two thirds from European countries, with no difference in severity or frequency between sexes; the consortium brought together centres in Europe (including Spain), North America, Japan, Turkey and Australia, and the article does not single out patients from Portugal or Brazil [18].

Portugal and Brazil

The sources consulted contain no Portuguese epidemiological data. From a bibliographical point of view, Sousa SB, of the Medical Genetics Unit of the Coimbra Paediatric Hospital, is first author of the 2009 phenotype delineation and the 2014 consortium series, and co-author of the description of BIS and of the episignature shared with ADNP [10, 13, 18, 23]. In Brazil, an 8-year-old boy with NCBRS and ASD was described in 2023 as the first genetically confirmed Brazilian case, after an earlier case without confirmation [46].

Special populations

Adults are probably underdiagnosed and under-represented, because many have not had genomic testing [1]. Young children are also underdiagnosed, because of the uncharacteristic face and the overlap with Coffin-Siris syndrome [21]. There is no difference in frequency between sexes [23].

European networks and patient associations

The relevant ERN is ERN ITHACA, dedicated to rare congenital anomalies and intellectual disability [17]. ERN ITHACA has not published a guideline or patient journey specific to NCBRS [17, 68]. In France, NCBRS is followed at the reference centre Déficiences intellectuelles de causes rares, coordinated at the Pitié-Salpêtrière Hospital (Paris), with 10 constituent centres (Lyon, Marseille, Trousseau Hospital, two at Necker Hospital, Dijon, Robert-Debré Hospital, Brest, Rennes and Strasbourg) and 13 competence centres (Tours, Kremlin-Bicêtre, Lille, Amiens, Besançon, Nancy, Nantes, Nice, Pointe-à-Pitre, Reims, Toulouse, Bordeaux and Montpellier), integrated into the national network (filière) DéfiScience for rare diseases of brain development and ID and in liaison with the AnDDI-Rares network for developmental anomalies [15]. The PNDS and the accompanying list of procedures and services serve as a reference for the attending physician for the care protocol underpinning the application for exemption from patient co-payment as a long-term condition not on the list (affection hors liste), and the family is informed about MDPH benefits and the medico-social structures (early medico-social action centres, medico-educational institutes, special education and home care services and specialised residential homes for adults) [15]. The patient resources cited by the protocol are the Association Coffin-Siris France, the Coffin-Siris Syndrome Foundation, a Dutch website dedicated to the ARID1B gene, the Alliance Maladies Rares and the UNAPEI federation of associations of people with intellectual disability; the protocol does not cite an NCBRS-specific association [15]. The NCBRS Worldwide Foundation, a charity registered in England and Wales under no. 1190194 and based in London (contactus@ncbrs.com), has the mission of supporting and informing families, carers and professionals, promoting awareness of the syndrome and advocating for research; it maintains, in partnership with the CoRDS programme of Sanford Research, a registry of people with NCBRS, and its declared activities are support, education, practical advice, annual conferences, welcome packs and small grants, with income of 20,002 pounds in the financial year ending 31 March 2026 [81, 82]. Since 2024, a UCL researcher has sat on its scientific board [78]. The Nicolaides-Baraitser Syndrome International Consortium brings together researchers; the NETRE network recruited patients for the epilepsy study, and the Sanford CoRDS registry includes NCBRS [1, 14, 43, 81]. Other resources cited by GeneReviews are GARD, the Leiden Open Variation Database (LOVD) and ClinVar for SMARCA2, and materials from the Epilepsy Foundation and the American Epilepsy Society [1, 5].

10. Illustrative Clinical Case

20-year follow-up of a mild form

Ejaz, Babul-Hirji and Chitayat described a boy born to non-consanguineous parents of English descent, followed for 20 years [21]. The pregnancy was complicated by a urinary infection in the first trimester, placental abruption at 30 weeks and pre-eclampsia at 39; delivery was at term, with a weight of 3.4 kg (50th centile), length of 48 cm (10th centile), head circumference of 34 cm (50th centile) and Apgar score of 9/9 [21].

At 6 weeks he was assessed for poor weight gain (3.57 kg, below the 3rd centile; head circumference 34.5 cm, −3 standard deviations) and had left torticollis, cryptorchidism, hypospadias, umbilical hernia, sparse hair, drooping eyelids, curly eyelashes, a prominent nasal root, a bulbous nose, malar hypoplasia and a thin upper vermilion; karyotype (46,XY), basic metabolic work-up, brain MRI, abdominal ultrasound and EEG were normal [21]. At 4 months a gastrojejunostomy was placed for feeding difficulties, and occupational therapy and speech therapy began before 1 year [21].

At 19 months he weighed 8.12 kg and measured 76 cm (both below the 3rd centile), with a head circumference of 45 cm (−3 standard deviations), a more prominent lower lip and mild scoliosis; he said no words, but comprehension was normal, and he began walking at 18 months [21]. At 21 months he had his first generalised tonic-clonic seizure, with a normal EEG, and started carbamazepine; seizures ceased at 4 years [21]. At 3 years he said 5 to 10 words and had attention deficit and tactile sensitivity [21].

At 6.5 years he weighed 18.1 kg (3rd to 10th centile), measured 106 cm (below the 3rd centile) and had a head circumference of 49.5 cm (−2 standard deviations), with widening of the distal phalanges of the hands and feet and prognathism, but radiographs of the hands, feet and spine showed no changes; he spoke in full sentences, rode a bicycle and attended the 1st year of mainstream school with one-to-one support, functioning at 3 years and 6 months at the age of 6 years and 1 month, and the Wechsler Intelligence Scale for Children (WISC-IV) showed moderate ID at 11 years [21]. The tube was removed at 12 years, with a continued preference for purées and an increased gag reflex; puberty, at 15 to 16 years, was accompanied by increased self-injury [21].

At 20 years he weighed 59.9 kg (10th to 25th centile), measured 163.4 cm (5th centile) and had a head circumference of 52.5 cm (−2 standard deviations); the face had become coarser after puberty, with sparse hair, deep-set eyes, an overhanging columella, full lips, prognathism and cystic acne of the back [21]. He was seizure-free, independent, rode a skateboard, spoke with normal language and mild articulation difficulty, prepared meals and read and wrote basic English, and was about to attend special education at a local college, without any regression [21].

The clinical diagnosis was made at 20 years and confirmed by a SMARCA2 panel (Illumina and Sanger sequencing), which identified the heterozygous de novo variant c.3493C>A p.(Gln1165Lys) in exon 25, helicase domain VI [21]. Normalised, the variant is NM_003070.5:c.3493C>A, NP_003061.3:p.(Gln1165Lys), NC_000009.12:g.2115858C>A, classified as likely pathogenic in ClinVar (VCV 212227) [5, 7]. The same patient was included in the methylation signature study, with a positive NCBRS-SMARCA2 score of 0.37 and pathogenic classification [28].

The figures of the article consulted (Figure 1, hands and feet with prominent interphalangeal joints and widened distal phalanges: https://pmc.ncbi.nlm.nih.gov/articles/PMC4831382/figure/ccr3425-fig-0001/; Figure 3, facial evolution from 3 months to 20 years, with progressive coarsening, sparse hair, eversion of the lower lip and prognathism: https://pmc.ncbi.nlm.nih.gov/articles/PMC4831382/figure/ccr3425-fig-0003/; Figure 4, diagram of the SMARCA2 gene with the location of the 59 known germline missense variants, adapted from Sousa and Hennekam, and the arrow at the patient's variant: https://pmc.ncbi.nlm.nih.gov/articles/PMC4831382/figure/ccr3425-fig-0004/) show the evolving nature of the phenotype [21].

Lessons from the case

The case runs counter to the correlation between domain VI variants and severe ID with epilepsy, described in six patients with Arg1159 variants and in one patient with p.(Ala1156Pro) and absent speech, and shows that early diagnosis is difficult because the features evolve with age [18, 21]. The measures in the report (carbamazepine, prolonged gastrojejunostomy) were individual decisions: carbamazepine is not one of the drugs with most data in NCBRS, and GeneReviews does not recommend a specific antiepileptic drug [1, 21, 43]. A contrasting case, of a girl of Ghanaian descent with myoclonic-atonic epilepsy from 14 months, without recognised dysmorphism at 17 months and with relative macrocephaly, was only diagnosed after trio exome at 5 years (c.3721C>G p.(Gln1241Glu), exon 26), which illustrates late recognition when the face is not yet characteristic [35]. In this patient, review at 5 years and 5 months, after the exome, showed wrinkled skin, frontal baldness, a broad and long philtrum, a broad nasal base with an upturned tip and thick alae nasi, anterior projection of the upper lip, a thin upper and thick lower vermilion, slightly thickened distal phalanges and eczema of the elbows and knees [35]. The figures of the article document the de novo variant (Figure 1, IGV alignment of the patient and parents and Sanger chromatogram: https://pmc.ncbi.nlm.nih.gov/articles/PMC5516167/figure/ajmga37935-fig-0001/), the phenotype at 5 years (Figure 2, face, hands and feet: https://pmc.ncbi.nlm.nih.gov/articles/PMC5516167/figure/ajmga37935-fig-0002/) and the position of the variant, the first described in exon 26, outside but close to the ATPase domain, in a diagram of the protein with the published variants (Figure 3: https://pmc.ncbi.nlm.nih.gov/articles/PMC5516167/figure/ajmga37935-fig-0003/); the authors proposed including exon 26 in targeted laboratory testing [35].

11. Prognosis and Long-Term Follow-Up

Prognosis

It is unknown whether life expectancy is reduced; the original patient lived to 33 years, and survival into adulthood is possible [1]. The two published deaths occurred at 33 years (status epilepticus and respiratory complications) and at 25 years (ruptured oesophageal varices) [11]. The syndrome is described as phased and progressive: the hair becomes sparser (especially in the second decade, although sometimes it improves), the face coarser and more wrinkled, the distal phalanges broader, the sandal gap more marked, and microcephaly goes from about one third at birth to two thirds at follow-up [1, 11]. Worsening seizures are associated with loss of speech, and complete remission of epilepsy is very rare [1, 43]. In the Chinese epilepsy cohort, with ages at last follow-up of up to 11 years, 10 of 24 patients were seizure-free, and a patient with DEE and congenital heart disease died at 1 year and 5 months, from influenza; the difference from the Hofmeister study reflects different populations and follow-up (section 7) [34]. There is, however, a favourable course in mild forms, with medical and developmental stability after 6 years and progressive gains, and a child with p.(Val1198Met) had global improvement over 4 years of follow-up [21, 39]. Analysis of public methylation data from 12 patients and 27 controls showed accelerated epigenetic age (PhenoAge and GrimAge) and shortened methylation-estimated telomere length, suggesting accelerated biological ageing, with clinical implications that remain hypothetical [1, 83]. The oncological risk has not been established and GeneReviews does not recommend tumour screening [1, 11]. In the 2014 series (median age 10 years, maximum 33), none of the 61 patients had developed neoplasia, but the authors note that the cohort was young and that tumours at older ages cannot be excluded; osteoporosis with fractures at or after puberty and slowing of movement in older patients are part of the course described, whereas eczema tends to improve in the second half of the first decade [18]. The French PNDS adds no prognostic data specific to NCBRS; for Coffin-Siris syndrome, it records the absence of long-term studies on life expectancy and the existence of ongoing prospective studies [15]. For BIS, the data are limited to the original series, aged 3 to 19 years: moderate to severe ID in all, walking acquired between 22 months and 8 years and no independent walking in 6 of 14; the sources consulted have no long-term follow-up or mortality data [6, 10].

Surveillance

Table 11.1. Recommended surveillance in NCBRS

12. Summary and Final Considerations

Nicolaides-Baraitser syndrome is an autosomal dominant BAFopathy, almost always de novo, caused by predominantly missense variants or in-frame deletions of the ATPase/helicase domain of SMARCA2, which act through a dominant-negative or gain-of-function effect and not through haploinsufficiency; whole-gene deletions do not cause the disease [1, 8, 29]. The picture combines usually severe ID with disproportionate impairment of language, sparse hair, prominent interphalangeal joints, brachydactyly, a face that becomes coarse with age, acquired microcephaly, short stature and epilepsy in about two thirds of patients, often difficult to control [1, 18]. Because the features evolve with age and overlap with those of Coffin-Siris syndrome, diagnosis rests on molecular study (sequencing with CNV analysis, or exome or genome), complemented by the methylation signature in the presence of VUS or negative tests [1, 28]. The allelic blepharophimosis-intellectual disability syndrome results from variants outside the helicase domains, has its own face and episignature and shares the signature with Helsmoortel-Van der Aa syndrome [10, 13]. Treatment is supportive: in epilepsy, levetiracetam and valproic acid give the most frequent initial responses, with labelled doses, attention to the behavioural warning for levetiracetam and to the restrictions on valproate (POLG-related disease, girls and women of childbearing potential); the airway may be difficult, and ophthalmological, hearing and dental surveillance should be regular [1, 43, 50, 58, 70, 77]. The recurrence risk for siblings is low but not zero, because of documented parental germline mosaicism, which allowed the first birth after PGT-M [1, 32, 79]. The 2021 French PNDS, the only clinical guidance covering NCBRS, adds to GeneReviews systematic echocardiography and abdominal ultrasound, early refraction under cycloplegia, targeted endocrine and immunological surveillance and the organisation of care in reference centres, but rests mainly on data from Coffin-Siris syndrome and has no doses [15]. International guidelines, therapeutic trials and data on adulthood, life expectancy and oncological risk are lacking [1, 14].

13. References

1. Abdul-Rahman O, May F. SMARCA2-Related Nicolaides-Baraitser Syndrome. 2015 Oct 15 [updated 2025 Mar 6]. In: Adam MP, Bick S, Mirzaa GM, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK321516/ PMID: 26468571.
2. Orphanet. Nicolaides-Baraitser syndrome (ORPHA:3051) [Internet]. Paris: INSERM; [cited 2026 Oct 27]. Available from: https://www.orpha.net/en/disease/detail/3051
3. OMIM® Online Mendelian Inheritance in Man. #601358 NICOLAIDES-BARAITSER SYNDROME; NCBRS. Johns Hopkins University; last edited 12 June 2024. https://omim.org/entry/601358. Accessed 27 October 2026.
4. Orphanet. SMARCA2-related blepharophimosis-intellectual disability syndrome (ORPHA:637013) [Internet]. Paris: INSERM; [cited 2026 Oct 27]. Available from: https://www.orpha.net/en/disease/detail/637013
5. National Center for Biotechnology Information. ClinVar: SMARCA2[gene] pathogenic/likely pathogenic [Internet]. Bethesda (MD): NCBI; [cited 2026 Oct 27]. Available from: https://www.ncbi.nlm.nih.gov/clinvar/?term=SMARCA2%5Bgene%5D
6. OMIM® Online Mendelian Inheritance in Man. #619293 BLEPHAROPHIMOSIS-IMPAIRED INTELLECTUAL DEVELOPMENT SYNDROME; BIS. Johns Hopkins University; last edited 30 April 2021. https://omim.org/entry/619293. Accessed 27 October 2026.
7. VariantValidator [Internet]. Manchester: University of Manchester; [cited 2026 Oct 27]. Available from: https://variantvalidator.org (validation of NM_003070.5, GRCh38, mane_select; gene2transcripts_v2 SMARCA2).
8. ClinGen. Dosage Sensitivity Curation: SMARCA2 (HGNC:11098) [Internet]. [cited 2026 Oct 27]. Available from: https://search.clinicalgenome.org/kb/gene-dosage/HGNC:11098
9. OMIM® Online Mendelian Inheritance in Man. *600014 SWI/SNF-RELATED, MATRIX-ASSOCIATED, ACTIN-DEPENDENT REGULATOR OF CHROMATIN, SUBFAMILY A, MEMBER 2; SMARCA2. Johns Hopkins University; last edited 4 May 2021. https://omim.org/entry/600014. Accessed 27 October 2026.
10. Cappuccio G, Sayou C, Tanno PL, Tisserant E, Bruel AL, Kennani SE, et al. De novo SMARCA2 variants clustered outside the helicase domain cause a new recognizable syndrome with intellectual disability and blepharophimosis distinct from Nicolaides-Baraitser syndrome. Genet Med. 2020;22(11):1838-1850. doi:10.1038/s41436-020-0898-y. PMID: 32694869.
11. Zhang X, Chen H, Song Y, Chen Z, Liu X, Rong P, et al. Ten-year follow-up of Nicolaides-Baraitser syndrome with a de novo mutation and analysis of 58 gene loci of SMARCA2-associated NCBRS. Mol Genet Genomic Med. 2022;10(9):e2009. doi:10.1002/mgg3.2009. PMID: 35811451. PMC9482403.
12. Kosho T, Miyake N, Carey JC. Coffin-Siris syndrome and related disorders involving components of the BAF (mSWI/SNF) complex: historical review and recent advances using next generation sequencing. Am J Med Genet C Semin Med Genet. 2014;166C(3):241-51. doi:10.1002/ajmg.c.31415. PMID: 25169878.
13. Sarli C, van der Laan L, Reilly J, Trajkova S, Carli D, Brusco A, et al. Blepharophimosis with intellectual disability and Helsmoortel-Van Der Aa Syndrome share episignature and phenotype. Am J Med Genet C Semin Med Genet. 2024;196(4):e32089. doi:10.1002/ajmg.c.32089. PMID: 38884529.
14. ClinicalTrials.gov [Internet]. Bethesda (MD): National Library of Medicine; [cited 2026 Oct 27]. Pesquisas «Nicolaides-Baraitser», «SMARCA2», «BAFopathy», «Coffin-Siris»; NCT01793168 (Rare Disease Patient Registry & Natural History Study – CoRDS). Available from: https://clinicaltrials.gov/study/NCT01793168
15. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Syndromes de Coffin-Siris et de Nicolaides-Baraitser (BAFopathies). Saint-Denis La Plaine: HAS; 2021 Nov 4 (text of September 2021; Centre de Référence Déficiences intellectuelles de causes rares, filière DéfiScience). https://www.has-sante.fr/jcms/p_3295021. Accessed 6 November 2026.
16. GeneGuidance. SMARCA2 — gene and condition record [database; version of 2026-09-26].
17. ERN ITHACA. ERN ITHACA guidelines [Internet]. [cited 2026 Oct 27]. Available from: https://ern-ithaca.eu/guidelines/ern-ithaca-guidelines/
18. Sousa SB, Hennekam RC; Nicolaides-Baraitser Syndrome International Consortium. Phenotype and genotype in Nicolaides-Baraitser syndrome. Am J Med Genet C Semin Med Genet. 2014;166C(3):302-14. doi:10.1002/ajmg.c.31409. PMID: 25169058.
19. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 816–819 (Coffin-Siris Syndrome).
20. Nicolaides P, Baraitser M. An unusual syndrome with mental retardation and sparse hair. Clin Dysmorphol. 1993;2(3):232-6. doi:10.1097/00019605-199307000-00007. PMID: 8287185.
21. Ejaz R, Babul-Hirji R, Chitayat D. The evolving features of Nicolaides-Baraitser syndrome - a clinical report of a 20-year follow-up. Clin Case Rep. 2016;4(4):351-5. doi:10.1002/ccr3.425. PMID: 27099726. PMC4831382.
22. Morin G, Villemain L, Baumann C, Mathieu M, Blanc N, Verloes A. Nicolaides-Baraitser syndrome: confirmatory report of a syndrome with sparse hair, mental retardation, and short stature and metacarpals. Clin Dysmorphol. 2003;12(4):237-40. doi:10.1097/00019605-200310000-00005. PMID: 14564210.
23. Sousa SB, Abdul-Rahman OA, Bottani A, Cormier-Daire V, Fryer A, Gillessen-Kaesbach G, et al. Nicolaides-Baraitser syndrome: Delineation of the phenotype. Am J Med Genet A. 2009;149A(8):1628-40. doi:10.1002/ajmg.a.32956. PMID: 19606471.
24. Van Houdt JK, Nowakowska BA, Sousa SB, van Schaik BD, Seuntjens E, Avonce N, et al. Heterozygous missense mutations in SMARCA2 cause Nicolaides-Baraitser syndrome. Nat Genet. 2012;44(4):445-9. doi:10.1038/ng.1105. PMID: 22366787.
25. Wolff D, Endele S, Azzarello-Burri S, Hoyer J, Zweier M, Schanze I, et al. In-Frame Deletion and Missense Mutations of the C-Terminal Helicase Domain of SMARCA2 in Three Patients with Nicolaides-Baraitser Syndrome. Mol Syndromol. 2012;2(6):237-244. doi:10.1159/000337323. PMID: 22822383. PMC3362220.
26. Mari F, Marozza A, Mencarelli MA, Lo Rizzo C, Fallerini C, Dosa L, et al. Coffin-Siris and Nicolaides-Baraitser syndromes are a common well recognizable cause of intellectual disability. Brain Dev. 2015;37(5):527-36. doi:10.1016/j.braindev.2014.08.009. PMID: 25249037.
27. Aref-Eshghi E, Bend EG, Hood RL, Schenkel LC, Carere DA, Chakrabarti R, et al. BAFopathies' DNA methylation epi-signatures demonstrate diagnostic utility and functional continuum of Coffin-Siris and Nicolaides-Baraitser syndromes. Nat Commun. 2018;9(1):4885. doi:10.1038/s41467-018-07193-y. PMID: 30459321. PMC6244416.
28. Chater-Diehl E, Ejaz R, Cytrynbaum C, Siu MT, Turinsky A, Choufani S, et al. New insights into DNA methylation signatures: SMARCA2 variants in Nicolaides-Baraitser syndrome. BMC Med Genomics. 2019;12(1):105. doi:10.1186/s12920-019-0555-y. PMID: 31288860. PMC6617651.
29. Gao F, Elliott NJ, Ho J, Sharp A, Shokhirev MN, Hargreaves DC. Heterozygous Mutations in SMARCA2 Reprogram the Enhancer Landscape by Global Retargeting of SMARCA4. Mol Cell. 2019;75(5):891-904.e7. doi:10.1016/j.molcel.2019.06.024. PMID: 31375262. PMC7291823.
30. Schrier Vergano S, Santen G, Wieczorek D, Matsumoto N. Coffin-Siris Syndrome. 2013 Apr 4 [updated 2025 May 15]. In: Adam MP, Bick S, Mirzaa GM, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK131811/ PMID: 23556151.
31. Wieczorek D, Bögershausen N, Beleggia F, Steiner-Haldenstätt S, Pohl E, Li Y, et al. A comprehensive molecular study on Coffin-Siris and Nicolaides-Baraitser syndromes identifies a broad molecular and clinical spectrum converging on altered chromatin remodeling. Hum Mol Genet. 2013;22(25):5121-35. doi:10.1093/hmg/ddt366. PMID: 23906836.
32. Pan N, Chen S, Cai X, Li J, Yu T, Huang HF, et al. Low-level germline mosaicism of a novel SMARCA2 missense variant: Expanding the phenotypic spectrum and mode of genetic transmission. Mol Genet Genomic Med. 2021;9(9):e1763. doi:10.1002/mgg3.1763. PMID: 34296532. PMC8457699.
33. Miyake N, Abdel-Salam G, Yamagata T, Eid MM, Osaka H, Okamoto N, et al. Clinical features of SMARCA2 duplication overlap with Coffin-Siris syndrome. Am J Med Genet A. 2016;170(10):2662-70. doi:10.1002/ajmg.a.37778. PMID: 27264538.
34. Yang Y, Xin T, Sun D, Zhang X, Sun Y, Zhuang J, et al. Genotype and phenotype correlation in epilepsy patients with SMARCA2 variants. Seizure. 2025;131:73-83. doi:10.1016/j.seizure.2025.05.019. PMID: 40513420.
35. Tang S, Hughes E, Lascelles K, Simpson MA, Pal DK. New SMARCA2 mutation in a patient with Nicolaides-Baraitser syndrome and myoclonic astatic epilepsy. Am J Med Genet A. 2017;173(1):195-199. doi:10.1002/ajmg.a.37935. PMID: 27665729. PMC5516167.
36. Liu X, Yang Q, Shi C, Hao H, Xiao X, Li S. [Analysis of clinical characteristics and genetic variant in a child with Nicolaides-Baraitser syndrome due to maternal mosaicism]. Zhonghua Yi Xue Yi Chuan Xue Za Zhi. 2022;39(12):1366-1369. doi:10.3760/cma.j.cn511374-20211220-01006. PMID: 36453960.
37. Bizzari S, Mehawej C, Chouery E, Nair P, Corbani S, Lefranc G, et al. Blended Phenotypes in Siblings: Dual Diagnoses of Nicolaides-Baraitser and Craniosynostosis Syndromes. Mol Syndromol. 2026;17(4):400-408. doi:10.1159/000549014. PMID: 41409309. PMC12707963.
38. Foley R, Duignan S, McArdle L, Betts DR, Green A, McMahon CJ. Nicolaides-Baraitser syndrome in a patient with hypertrophic cardiomyopathy and SMARCA2 gene deletion. Cardiol Young. 2022;32(5):821-823. doi:10.1017/S1047951121003826. PMID: 34521483.
39. Akella R. SWI/SNF-Related SMARCA2 Gene Mutation Associated with Nicolaides-Baraitser's Syndrome: Follow-up Study. J Pediatr Genet. 2024;13(2):154-157. doi:10.1055/s-0041-1740458. PMID: 38721580. PMC11076067.
40. Hulse M, Wang M, Xu C, Carter J, Agarwal A, Lu L, et al. PRT3789 Is a First-in-Human SMARCA2-Selective Degrader That Induces Synthetic Lethality in SMARCA4-Mutated Cancers. Cancer Res. 2026;86(1):213-235. doi:10.1158/0008-5472.CAN-25-1141. PMID: 40991405.
41. Hofmeister B, von Stülpnagel C, Berweck S, Abicht A, Kluger G, Weber P. Cooccurrence of Two Different Genetic Diseases: A Case of Valproic Acid Hepatotoxicity in Nicolaides-Baraitser Syndrome (SMARCA2 Mutation)-Due to a POLG1-Related Effect?. Neuropediatrics. 2020;51(1):49-52. doi:10.1055/s-0039-1694976. PMID: 31541998.
42. Cardwell EB, Mathews JM, Hildebrandt CC. Pediatric epilepsy: a dual diagnosis of SCMARCA2 -Nicolaides-Baraitser syndrome and CHRNB2 -nocturnal-frontal-lobe epilepsy. Clin Dysmorphol. 2026;35(1):32-34. doi:10.1097/MCD.0000000000000544. PMID: 41198078.
43. Hofmeister B, von Stülpnagel C, Betzler C, Mari F, Renieri A, Baldassarri M, et al. Epilepsy in Nicolaides-Baraitser Syndrome: Review of Literature and Report of 25 Patients Focusing on Treatment Aspects. Neuropediatrics. 2021;52(2):109-122. doi:10.1055/s-0041-1722878. PMID: 33578439.
44. Pretegiani E, Mari F, Renieri A, Penco S, Dotti MT. Nicolaides-Baraitser syndrome: defining a phenotype. J Neurol. 2016;263(8):1659-60. doi:10.1007/s00415-016-8194-0. PMID: 27286846.
45. van der Sluijs PJ, Jansen S, Vergano SA, Adachi-Fukuda M, Alanay Y, AlKindy A, et al. The ARID1B spectrum in 143 patients: from nonsyndromic intellectual disability to Coffin-Siris syndrome. Genet Med. 2019;21(6):1295-1307. doi:10.1038/s41436-018-0330-z. PMID: 30349098. PMC6752273.
46. Lopes DC, Mendes LST, Ferreira ICH. Autism spectrum disorder in a patient with Nicolaides-Baraitser Syndrome: case report. Einstein (Sao Paulo). 2023;21:eRC0480. doi:10.31744/einstein_journal/2023RC0480. PMID: 37970954. PMC10631755.
47. Gana S, Panizzon M, Fongaro D, Selicorni A, Memo L, Scandurra V, et al. Nicolaides-Baraitser syndrome: two new cases with autism spectrum disorder. Clin Dysmorphol. 2011;20(1):38-41. doi:10.1097/MCD.0b013e32833edaa9. PMID: 20802310.
48. Moreno BA, Torres K, Duarte A. Expanding the Cutaneous Spectrum of Nicolaides‑Baraitser Syndrome: Eczema and Generalized Hair Loss. Cureus. 2025;17(7):e88854. doi:10.7759/cureus.88854. PMID: 40881542. PMC12381894.
49. Taharabaru S, Sato T, Nishiwaki K. Difficult Airway Management in a Patient With Nicolaides-Baraitser Syndrome Who Had a Small Jaw and Limited Mouth Opening. Anesth Prog. 2021;68(1):47-49. doi:10.2344/anpr-67-04-05. PMID: 33827121. PMC8033586.
50. Goehring M, Choorapoikayil S, Zacharowski K, Messroghli L. Anaesthesia and orphan disease: management of a case of Nicolaides-Baraitser syndrome undergoing cleft palate surgery. BMC Anesthesiol. 2021;21(1):162. doi:10.1186/s12871-021-01380-z. PMID: 34039274. PMC8152345.
51. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 107–108 (Coarse facial features).
52. Simmers R, Goodwin A, Al Saif H, Couser N. Ophthalmologic and facial abnormalities of Nicolaides-Baraitser syndrome. Ophthalmic Genet. 2022;43(5):699-702. doi:10.1080/13816810.2022.2089358. PMID: 35762114.
53. Sethi S, Thau A, Kaplan P, Asher SB, Levin AV. Glaucoma and degenerative vitreoretinopathy in a girl with Nicolaides-Baraitser syndrome. J AAPOS. 2019;23(3):169-171. doi:10.1016/j.jaapos.2018.12.002. PMID: 30707941.
54. Alyahya YA, Alsirhy EY, Alwadani SF, Al-Anazi TA, Kondkar AA, Azad TA. Bilateral congenital glaucoma in a child with Nicolaides-Baraitser syndrome: a case report. Ann Med Surg (Lond). 2026;88(1):902-904. doi:10.1097/MS9.0000000000004499. PMID: 41497020. PMC12768013.
55. Henriquez-Lopez CJ, McLean S. Hashitoxicosis in a patient with Nicolaides-Baraitser Syndrome: a case report. Clin Dysmorphol. 2023;32(2):77-79. doi:10.1097/MCD.0000000000000446. PMID: 36779796.
56. Wang MJ, Febres-Cordero DA, Poorvu T, Delerme P, Hecht J, Oyelese Y, et al. Caudal regression in fetus with de novo SMARCA2 pathogenic variant. Prenat Diagn. 2024;44(9):1111-1114. doi:10.1002/pd.6627. PMID: 38877377.
57. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 332 (Unusual hair, teeth, nails, and skin).
58. AbbVie Inc. DEPAKOTE (divalproex sodium) delayed-release tablets. Prescribing information [Internet]. Silver Spring (MD): U.S. Food and Drug Administration; revised 2026 Mar [cited 2026 Oct 27]. NDA018723. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=08a65cf4-7749-4ceb-6895-8f4805e2b01f
59. Bramswig NC, Lüdecke HJ, Alanay Y, Albrecht B, Barthelmie A, Boduroglu K, et al. Exome sequencing unravels unexpected differential diagnoses in individuals with the tentative diagnosis of Coffin-Siris and Nicolaides-Baraitser syndromes. Hum Genet. 2015;134(6):553-68. doi:10.1007/s00439-015-1535-8. PMID: 25724810.
60. Wang CH, Lin SP, Huang JK, Chiu NC. Diagnostic Pitfall in Genomic Era: Discordant SMARCA2 Finding in Patient with Megalencephalic Leukoencephalopathy with Subcortical Cysts. Diagnostics (Basel). 2026;16(15):2314. doi:10.3390/diagnostics16152314. PMID: 42587552. PMC13463989.
61. Sadikovic B, Levy MA, Kerkhof J, Aref-Eshghi E, Schenkel L, Stuart A, et al. Clinical epigenomics: genome-wide DNA methylation analysis for the diagnosis of Mendelian disorders. Genet Med. 2021;23(6):1065-1074. doi:10.1038/s41436-020-01096-4. PMID: 33547396. PMC8187150.
62. Smits DJ, Debuy C, Brooks AS, Schot R, Ferraro F, Rots D, et al. Clinical utility of DNA-methylation signatures in routine diagnostics for neurodevelopmental disorders. Eur J Hum Genet. 2025;33(10):1281-1289. doi:10.1038/s41431-025-01919-5. PMID: 40730689. PMC12480637.
63. Koparir A, Kerkhof J, Rzasa J, Metzger E, Bahena Carbajal P, Kolokotronis K, et al. Chromatinopathies: clinically overlapping disorders, revealing novel variants and their DNA methylation signatures. Clin Epigenetics. 2026;18(1):69. doi:10.1186/s13148-026-02120-1. PMID: 41957673. PMC13088768.
64. Pascolini G, Valiante M, Bottillo I, Laino L, Fleischer N, Ferraris A, et al. Striking phenotypic overlap between Nicolaides-Baraitser and Coffin-Siris syndromes in monozygotic twins with ARID1B intragenic deletion. Eur J Med Genet. 2020;63(3):103739. doi:10.1016/j.ejmg.2019.103739. PMID: 31421289.
65. OMIM® Online Mendelian Inheritance in Man. #600430 CHROMOSOME 2q37 DELETION SYNDROME. Johns Hopkins University; last edited 25 September 2015. https://omim.org/entry/600430. Accessed 27 October 2026.
66. Finsterer J. Overlapping Phenotype from Double Trouble SMARCA2 and POLG1 Variants c.2556A > C and c.3708G > T, Respectively. Neuropediatrics. 2020;51(3):235-236. doi:10.1055/s-0039-3402008. PMID: 31887771.
67. von Stülpnagel C, Hofmeister B, Berweck S, Kluger G, Weber P. Response To: Overlapping Phenotype from Double Trouble SMARCA2 and POLG1 Variants c.2556A > C and c.3708G > T, Respectively. Neuropediatrics. 2020;51(5):375-376. doi:10.1055/s-0040-1712486. PMID: 32464671.
68. ERN ITHACA. Endorsed guidelines [Internet]. [cited 2026 Oct 27]. Available from: https://ern-ithaca.eu/guidelines/endorsed-guidelines/
69. Chen S, Yin F, He F, Zhang C, Deng X, Yang L, et al. Genotypic and Phenotypic Profile of 50 Cases With Chromatin Remodeling Complexes-Related Neurological Disorders. CNS Neurosci Ther. 2026;32(8):e71061. doi:10.1002/cns.71061. PMID: 42528014. PMC13420320.
70. UCB, Inc. KEPPRA (levetiracetam) tablets and oral solution. Prescribing information [Internet]. Silver Spring (MD): U.S. Food and Drug Administration; revised 2026 Jan [cited 2026 Oct 27]. NDA021035/NDA021505. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=3ca9df05-a506-4ec8-a4fe-320f1219ab21
71. European Medicines Agency. Keppra (levetiracetam): EPAR — product information (Keppra 100 mg/ml oral solution, SmPC) [Internet]. Amsterdam: EMA; first authorised 2000 Sep 29 [cited 2026 Oct 27]. Available from: https://www.ema.europa.eu/en/documents/product-information/keppra-epar-product-information_en.pdf
72. INFARMED. Depakine 200 mg/ml solução oral (valproato de sódio): resumo das características do medicamento [Internet]. Lisboa: INFARMED, Infomed; marketing authorisation holder Sanofi – Produtos Farmacêuticos, Lda.; first authorisation 1989 Nov 27, renewal 2011 Dec 20 [cited 2026 Oct 27]. Registration no. 9729400. Available from: https://extranet.infarmed.pt/INFOMED-fo/download-ficheiro.xhtml?med_id=2413&tipo_doc=rcm
73. Abbott Laboratórios do Brasil Ltda. Depakote® (divalproato de sódio) comprimidos revestidos de liberação entérica. Bula para o profissional da saúde, approved by Anvisa on 09/04/2021 [Internet]. São Paulo: Abbott; 2021 [cited 2026 Oct 27]. Reg. MS 1.0553.0203. Available from: https://dam.abbott.com/pt-br/documents/pdfs/nossas-bulas/D/BU-21-Depakote-Bula-Profissional-Final.pdf
74. Abbott Laboratórios do Brasil Ltda. Depakene® (valproato de sódio) xarope 50 mg/mL. Bula para o profissional da saúde, approved by Anvisa on 15/03/2019 [Internet]. São Paulo: Abbott; 2019 [cited 2026 Oct 27]. Reg. MS 1.0553.0315. Available from: https://dam.abbott.com/pt-br/documents/pdfs/nossas-bulas/D/BU14_Depakene_Bula_Profissional_Xpe_Final.pdf
75. Agência Nacional de Vigilância Sanitária (ANVISA). Dados abertos: medicamentos registrados no Brasil (DADOS_ABERTOS_MEDICAMENTOS.csv) [Internet]. Brasília: ANVISA; [cited 2026 Oct 27]. Available from: https://dados.anvisa.gov.br/dados/DADOS_ABERTOS_MEDICAMENTOS.csv
76. European Medicines Agency. Potential risk of neurodevelopmental disorders in children born to men treated with valproate medicines: PRAC recommends precautionary measures [Internet]. Amsterdam: EMA; 2024 Jan 12 [cited 2026 Oct 27]. Available from: https://www.ema.europa.eu/en/news/potential-risk-neurodevelopmental-disorders-children-born-men-treated-valproate-medicines-prac-recommends-precautionary-measures
77. European Medicines Agency. New measures to avoid valproate exposure in pregnancy endorsed (EMA/145600/2018) [Internet]. London: EMA; 2018 Mar 23 [cited 2026 Oct 27]. Available from: https://www.ema.europa.eu/en/documents/referral/valproate-article-31-referral-new-measures-avoid-valproate-exposure-pregnancy-endorsed_en.pdf
78. Xu M; UCL Faculty of Brain Sciences. Family visit highlights global collaboration in Nicolaides-Baraitser syndrome research [Internet]. London: University College London; 2026 Mar 26 [cited 2026 Oct 27]. Available from: https://www.ucl.ac.uk/brain-sciences/news/2026/mar/family-visit-highlights-global-collaboration-nicolaides-baraitser-syndrome-research
79. Pan J, Li J, Chen S, Xu C, Huang H, Jin L. Living birth following preimplantation genetic testing for monogenic disorders to prevent low-level germline mosaicism related Nicolaides-Baraitser syndrome. Front Genet. 2022;13:989041. doi:10.3389/fgene.2022.989041. PMID: 36160002. PMC9500527.
80. Sánchez AI, Rojas JA. A SMARCA2 Mutation in the First Case Report of Nicolaides-Baraitser Syndrome in Latin America: Genotype-Phenotype Correlation. Case Rep Genet. 2017;2017:8639617. doi:10.1155/2017/8639617. PMID: 28948053. PMC5602489.
81. NCBRS Worldwide Foundation [Internet]. London: NCBRS Worldwide Foundation; 2020-2026 [cited 2026 Oct 27]. Available from: https://ncbrs.com
82. Charity Commission for England and Wales. Nicolaides Baraitser Syndrome (NCBRS) Worldwide Foundation, charity number 1190194 [Internet]. [cited 2026 Oct 27]. Available from: https://register-of-charities.charitycommission.gov.uk/charity-search/-/charity-details/5162706
83. Shinko Y, Okazaki S, Otsuka I, Horai T, Kim S, Tanifuji T, et al. Accelerated epigenetic age and shortened telomere length based on DNA methylation in Nicolaides-Baraitser syndrome. Mol Genet Genomic Med. 2022;10(3):e1876. doi:10.1002/mgg3.1876. PMID: 35092358. PMC8922957.