ZMYND11-Related Disorder and 10p15.3 Microdeletion Syndrome

Autosomal dominant neurodevelopmental disorder due to loss of function of a reader of the chromatin mark H3.3K36me3, caused by intragenic ZMYND11 variants or by 10p15.3 deletions that include the gene, with speech delay, behavioural abnormalities, epilepsy and subtle facial dysmorphism.

Literature and regulatory information updated: 28 October 2026.

0. Initial Reference Table

Table 0 brings the two aetiological forms together under the same entity, as Orphanet does, because the 10p15.3 deletion and the intragenic variant compromise the same ZMYND11 dosage and produce overlapping phenotypes [12, 15]. The code ORPHA 178469, cited in a 2024 report for this disease, corresponds in the official Orphanet nomenclature to the group entity “autosomal dominant non-syndromic intellectual disability”, and not to this disorder; the specific codes are 694304, 694308 and 687424 [12, 20]. The OMIM entries for the phenotype (616083) and for the gene (608668), both last edited in September 2022, confirm the numbers, the 10p15.3 cytoband, the GRCh38 coordinates and autosomal dominant inheritance with incomplete penetrance and variable expressivity; they are based on the series of Coe (2014), Yates (2020) and Oates (2021) and on the case of Cobben (2014), and do not include the literature after 2022 [1, 11].

1. Introduction

1.1. From 10p deletions to identification of the critical gene

The history of this disorder begins with the cytogenetics of the short arm of chromosome 10. Before the microarray era, two classical 10p deletion syndromes had been described, both associated with larger and more centromeric deletions: hypoparathyroidism, sensorineural deafness and renal dysplasia syndrome (HDR or Barakat syndrome, OMIM 146255), caused by haploinsufficiency of GATA3 at 10p14, and DiGeorge syndrome type 2 (DGS2, OMIM 601362), associated with 10p14-p13 [6]. The two entities correspond to contiguous segments of the same chromosome arm: comparison of terminal 10p deletions showed, in 2000, that loss of a more proximal region is associated with heart disease and T-lymphocyte deficiency and loss of a more distal region with the triad of hypoparathyroidism, deafness and renal dysplasia, and in the same year deletion mapping in patients with HDR narrowed the critical region to about 200 kb containing GATA3, whose loss of function then came to explain the syndrome also in patients with intragenic variants [21, 22]. For DGS2, by contrast, OMIM still records only a locus defined by overlapping deletions, with a minimal common region estimated at about 2 Mb at the 10p13/10p14 boundary and no established causal gene [22]. Studies of larger terminal deletions had proposed two critical regions, one of 1.6 Mb at 10p15.1 linked to ID and language abnormalities and another of 4.3 Mb at 10p14 linked to autism and dysmorphism, but until 2012 only two patients with an isolated submicroscopic 10p15.3 deletion had been published [6].

In 2012, DeScipio and colleagues described 19 unrelated individuals, from the United States and Europe, with 10p15.3 deletions identified by microarray, ranging in size from 154,558 bp to about 4 Mb [6]. The common phenotype included cognitive, behavioural and developmental abnormalities, speech and motor delay, craniofacial dysmorphism and hypotonia, and all deletions but one included ZMYND11, whereas all but another included DIP2C; the size of the deletion did not predict severity [6]. In the same year, a study of de novo variants in children on the autism spectrum identified a ZMYND11 splice variant in a child with autism spectrum disorder (ASD), without evident ID or dysmorphism [4, 23].

The distinction between the two genes of the minimal region was made in 2014. Coe and colleagues built a CNV morbidity map from 29,085 children with developmental delay and 19,584 controls and resequenced 26 candidate genes in 4716 additional cases and 2193 controls [24]. They found five truncating ZMYND11 variants in cases and none in DIP2C, whereas in controls only DIP2C truncations occurred; the integrated analysis of deletions and loss-of-function variants gave p = 2.81×10⁻⁵ for ZMYND11 and p = 0.48 for DIP2C, and ZMYND11 came to be considered the critical gene of the 10p15.3 microdeletion [24]. In that year, Cobben and colleagues published a boy with severe syndromic ID and a de novo missense variant, later recognised as c.1798C>T p.(Arg600Trp), arguing for pathogenicity on the basis of the deletions that included the gene, but acknowledging that more cases would be needed [25, 26].

1.2. Delineation of the phenotype and expansion of the spectrum

Isolated reports followed: a North American patient with the de novo missense variant p.(Ser421Asn) and a severe phenotype resembling Angelman syndrome (Moskowitz and colleagues, 2016), patients identified in sequencing studies of ID cohorts (Popp and colleagues, 2017) and a frameshift variant in one of 57 families studied for suspected Cornelia de Lange syndrome (Aoi and colleagues, 2019) [20, 26–28]. In 2017, Tumiene and colleagues compared 10p15.3 deletions with ZMYND11 loss-of-function variants in a Lithuanian family and in the literature and confirmed the phenotypic overlap [15].

In 2020, when only 11 patients had been published, Yates and colleagues described 16 new individuals from the United Kingdom, the Netherlands, France, Belgium and the United States, including a mother with three affected children, under the designation “ZMYND11-related syndromic intellectual disability”, later adopted by the patient organisation Unique [4, 17]. In 2021, Oates and colleagues gathered 47 individuals and described 20 with epilepsy in detail, establishing the gene as a cause of focal epilepsies with centrotemporal spikes and of generalised epilepsies associated with a neurodevelopmental disorder [9]. In the same year the smallest known deletion was published, an intragenic deletion that removes the 5′ untranslated region and the first two exons [7].

From 2022 the spectrum broadened to adulthood, with movement disorders (dystonia, ataxia, tremor) that appear or worsen after childhood [29, 30], to progressive myoclonic ataxia [31], to neuropsychiatric regressions triggered by infections [32] and to an association with schizophrenia risk in exome studies [33]. In 2024, ClinGen classified the gene–disease relationship as definitive [3], and in 2026 Yoshimatsu and colleagues reviewed 50 published cases and described the phenotype specific to carriers of p.(Arg600Trp) [10].

1.3. Nomenclature and definition of the entities

OMIM designates the phenotype intellectual developmental disorder, autosomal dominant 30, with speech delay and behavioural abnormalities (MRD30, 616083), a name that replaced the former “mental retardation, autosomal dominant 30”, and defines it as developmental delay evident from early childhood, with cognitive impairment of variable degree (many patients attend special schools), frequent behavioural abnormalities (attention deficit hyperactivity disorder, ADHD, autistic traits and aggressiveness) and, inconsistently, seizures of various types, hypotonia, mild skeletal anomalies, feeding difficulties and dysmorphism [1]. The gene has the number 608668, in an entry that also includes the BRAM1 isoform [2, 11]. ClinGen uses the broader label “ZMYND11 syndromic complex neurodevelopmental disorder” [2, 3]. Orphanet created a parent entity (ORPHA:694304) with two aetiological subtypes, one due to 10p15.3 microdeletion (ORPHA:687424) and the other due to a point variant (ORPHA:694308) [5, 12, 13].

Within the scope of this article, ZMYND11-related disorder is defined by the presence of a heterozygous intragenic pathogenic or likely pathogenic ZMYND11 variant, or of a deletion that removes all or part of the gene, in an individual with a neurodevelopmental disorder. The 10p15.3 microdeletion syndrome is defined by a submicroscopic deletion of the 10p15.3 region, terminal or interstitial, that includes ZMYND11 and, in most cases, the neighbouring genes [6, 15]. Larger deletions that extend into 10p14 or that coexist with other chromosomal imbalances have a phenotype augmented by the remaining genes and are treated as differential diagnosis and context [6, 14].

2. Molecular Genetics and Pathophysiology

2.1. The gene and the protein

ZMYND11 is located at 10p15.3, near the telomere of the short arm of chromosome 10 (GRCh38 chr10:130 088–254 637), and encodes a nuclear protein of about 74 kDa also known as BS69 [3, 34]. The MANE Select (Matched Annotation from NCBI and EMBL-EBI) reference transcript is NM_001370100.5, corresponding to ENST00000381604.9 and to the protein NP_001357029.1 [3]. The protein has six domains: a winged helix (WH), a PHD finger (plant homeodomain), a bromodomain, a PWWP domain, a coiled-coil region and, at the C-terminal end, the MYND-type zinc finger that gives the gene its name [10, 34].

The name BS69 comes from the discovery of the protein, in 1995, as a partner of the adenovirus E1A protein in a yeast two-hybrid screen: the sequence then deduced had 562 amino acids and migrated with an apparent mass of about 69 kDa; the protein bound the 289R isoform of E1A, but not the 243R isoform, and inhibited transactivation dependent on the former, and the transcript was detected in all mouse tissues studied, with the highest level in the kidney [11]. The existence of isoforms of different length explains why the older literature uses amino acid numberings that differ from those of the MANE transcript (section 2.5). In 1998 an alternatively spliced isoform, BRAM1, was described, formed by the 186 C-terminal amino acids of BS69 preceded by 12 amino acids of its own; unlike BS69, which is nuclear, BRAM1 is located in the cytoplasm and binds the kinase domain of bone morphogenetic protein (BMP) receptor type 1A, BMPR1A, and the protein TAB1, which led to the proposal that it participates in BMP–TAB1–TAK1 signalling [11]. In 2000 it was shown that complete repression of transcription by BS69 depends on the MYND domain, which recruits the corepressor NCoR; BRAM1 also binds NCoR but does not repress, so this interaction is necessary but not sufficient, and E1A abolishes the repression [11]. In 2002 the PxLxP motif, recognised by MYND, was identified in E1A, in the Epstein-Barr virus protein EBNA2 and in the cellular protein MGA, which is related to MYC, with the viral proteins competing for binding [11]. The first chromosomal localisation, by fluorescence in situ hybridisation (FISH) in 2000, placed the gene at 10p14; current genomic annotation places it at 10p15.3 [11].

The main function is chromatin reading. The bromodomain–PWWP tandem specifically recognises trimethylation of lysine 36 of histone H3 when it is incorporated in the H3.3 variant (H3.3K36me3): the trimethylated group is accommodated in an aromatic cage of the PWWP and the serine 31 residue, unique to H3.3, fits into a pocket formed by the two domains, which explains why the protein does not recognise the canonical histones H3.1 and H3.2 [2, 35]. These conclusions derive from the crystal structures of the bromodomain–PWWP tandem free and in complex with the H3.3K36me3 peptide (resolution 1.95 and 2.0 Å), compared with that of the complex with H3.1K36me3 (2.3 Å) [11, 35]. The H3K36me3 mark is deposited in gene bodies during transcriptional elongation, and ZMYND11 colocalises genome-wide with H3K36me3 and H3.3 in highly expressed genes, where it acts as an unconventional corepressor: it modulates RNA polymerase II at the elongation stage and contributes to repressing cryptic transcription from gene bodies [15, 35]. Binding to H3.3K36me3 chromatin also participates in the regulation of pre-mRNA processing [20, 36].

The MYND domain binds PxLxP motifs of partner proteins and recruits corepressors such as NCoR and transcriptional regulators such as EZH2, HDAC1, BRG1 and E2F6 [10, 34]. Structural studies in 2026 described intermolecular zinc fingers in the PHD and intermolecular disulfide bridges between the winged helix and the coiled-coil–MYND region, whose redox state alters binding to targets; the protein forms oligomers through the C-terminus, and the PHD and MYND domains are required for nuclear localisation and SUMOylation [9, 34]. Outside the nervous system, ZMYND11 behaves as a tumour suppressor, repressing a transcriptional programme required for tumour cell growth; in patients with breast cancer, low expression was associated with a worse prognosis, and overexpression reduced cell growth in vitro and tumour formation in mice [2, 11, 35]. A somatic ZMYND11::MBTD1 fusion has been described in acute myeloid leukaemia, but it belongs to somatic oncology and not to the tumour risk of germline carriers, which the sources consulted do not describe [37].

2.2. Expression and role in cortical development

Expression is ubiquitous, without regional specificity in the mammalian brain, and is enriched in the developing human somatosensory cortex, above all in corticothalamic projection neurons [9]. An old Northern blot study, cited in 2016, found the highest expression in the kidney and lower expression in the lung, brain, spleen, thymus and testis [26]. In mouse models, loss of Zmynd11 disrupts neurogenesis and neuronal maturation, with involvement of the PI3K pathway and of H3K36me3 methylation of the Epha2 gene [10, 38].

In human cortical organoids and isogenic embryonic stem cell lines with loss of function and with heterozygous variants similar to those of patients, lack of ZMYND11 reduces the expression of FOXG1 and TBR2 and the production of intermediate progenitors from radial glial cells, activates latent developmental programmes that should be silenced and alters an RBFOX2-dependent brain-specific RNA isoform switch [39]. The authors observed similar defects in other chromatin-related ASD risk genes, partially corrected by increasing ZMYND11 function; these data are preclinical in vitro [39]. Preprints from 2025 and 2026, not yet peer reviewed, propose that ZMYND11 restrains the methyltransferase KMT2A during the neuronal developmental programme and that excitatory cortical neurons are selectively vulnerable to its loss [40, 41].

For epilepsy, Oates and colleagues proposed, without experimental demonstration, that reduced ZMYND11 increases transcriptional elongation of H3.3-dependent synaptic genes and thereby excitatory drive; the hypothesis links the epileptic phenotype to the disorders of the H3.3 histones (H3-3A and H3-3B, formerly H3F3A and H3F3B), of CHD2, of SMARCA2 and of the paralogue ZMYND8, which reads other marks (H3.1K36me2 and H4K16ac) and is also associated with neurodevelopmental disorders with seizures [9].

2.3. The histone H3 lysine 36 pathway

ZMYND11 belongs to a group of genes that write or read H3K36 methylation and whose disorders overlap clinically. NSD1 and SETD2 encode enzymes that deposit H3K36 methylation (“writers”) and cause Sotos syndrome and the SETD2 neurodevelopmental disorders (including Luscan-Lumish syndrome); DNMT3A encodes a DNA methyltransferase whose PWWP domain reads H3K36me2 and causes Tatton-Brown-Rahman syndrome [42–45]. Unlike these, which are mainly overgrowth syndromes, the ZMYND11 disorder does not have macrosomia as a cardinal feature, and head circumference (HC) is usually normal [4, 10]. The functional relationship justifies the inclusion of these genes in the differential diagnosis and in panels, developed in section 6.

2.4. Disease mechanism and genetic constraint

The main mechanism is haploinsufficiency. ClinGen assigned a haploinsufficiency score of 3 (sufficient evidence for pathogenicity due to loss of dosage) and a triplosensitivity score of 0 [3]. The gene is highly intolerant to loss of function in the general population: in the ExAC database, 0 loss-of-function variants were observed against 28.6 expected (probability of loss-of-function intolerance, pLI = 1.0) and 90 missense variants against 217.2 expected (z = 4.22), and in gnomAD v4.0 the pLI is 1 and the LOEUF (loss-of-function observed/expected upper bound fraction) is 0.17; the DECIPHER haploinsufficiency percentage is 16.99 [3, 26]. Complete deletion of the gene, the intragenic deletion that removes the first and second exons and truncating variants give overlapping phenotypes, which reinforces loss of function as the common mechanism [4, 7, 15].

Two recurrent missense variants raise the possibility of an additional mechanism. p.(Arg600Trp) alters a key binding residue of the MYND domain and is associated with a more severe and distinct phenotype (microcephaly, growth restriction, cryptorchidism, nipple abnormalities); p.(Ser421Asn), located between the PWWP and MYND domains at a residue phosphorylated in mitosis, gave in two patients a severe picture with drug-resistant epilepsy and delayed myelination [10, 20, 26]. The authors allow for an effect different from simple haploinsufficiency, possibly gain of function, but no functional study has demonstrated it [4, 10]. By contrast, the in-frame deletion p.(Gln587del) and the missense variant p.(Cys574Arg), both in the MYND domain, give phenotypes overlapping those of truncating variants, which is compatible with loss of function [4, 24].

2.5. Variant spectrum and nomenclature

In the systematic review of 50 cases published in 2026, the variants were distributed as frameshift (17; 34%), nonsense (15; 30%), missense (13; 26%), splice (3; 6%) and in-frame (2; 4%) [10]. Oates and colleagues gathered 43 sequence variants and observed pathogenic or likely pathogenic variants in the four main domains: only frameshift in the PHD, frameshift and truncations in the bromodomain, missense and truncations in the PWWP, and missense, deletions and truncations in the MYND; they found no relationship between the type or location of the variant and the presence of epilepsy [9]. Truncating variants located in the last exon, such as p.(Gln586Ter), are expected to escape nonsense-mediated mRNA decay (NMD) and to disrupt the MYND domain [4]. No founder variants have been described.

The literature up to 2020 used the transcripts NM_006624 (Coe 2014, Cobben 2014) and NM_006624.5 (Yates 2020), and the recent literature uses the MANE NM_001370100.5 [4, 20, 24]. Checking in VariantValidator showed that NM_006624.4, .5, .6 and .7 have the same c. and p. numbering as the MANE for the variants tested, so the older designations are directly comparable; the exceptions are NM_006624.3 and NM_212479.1, whose numbering differs by −120 nucleotides (−40 amino acids), and NM_212479.2–.4, which differs by −3 nucleotides (−1 amino acid), for example p.(Val378Gly) or p.(Val417Gly) for the variant that is p.(Val418Gly) in the MANE [46]. Table 2.1 presents the variants of this article normalised according to Human Genome Variation Society (HGVS) nomenclature.

Table 2.1. ZMYND11 variants cited in the article, normalised on the MANE transcript NM_001370100.5 (VariantValidator, GRCh38 and GRCh37; ClinVar accessed on 28 October 2026; in the rows for the Coe and Yates series, the ClinVar and dbSNP records are those indicated by OMIM) [11, 46, 47]

The truncating variants of the Coe series, on NM_006624.5 and therefore with the same numbering as the MANE, were p.(Thr70AsnfsTer12) de novo, p.(Asn152ThrfsTer26), p.(Met187IlefsTer19) inherited from a father with delay, p.(Gln326Ter), p.(Glu416SerfsTer5) and the in-frame deletion p.(Gln587del) de novo [24]. In the Yates series, 10 variants were truncating, two missense — c.1720T>C p.(Cys574Arg) in the MYND domain and c.1246G>A p.(Glu416Lys) outside any known domain — and one splice variant, none present in gnomAD v2.1.1 [4]. Submission to VariantValidator of the GRCh37 genomic coordinates that OMIM gives for the Coe variants yielded, on the MANE, c.206dup, c.561del, c.976C>T, c.1246_1247del and c.1759_1761del, with the same protein designations as the article; the four Yates variants selected by OMIM were likewise confirmed, with one correction of form: the duplication published as c.1572dup is written c.1573dup according to the HGVS rule of shifting to the most 3′ position, without change in the consequence p.(Asp525GlyfsTer5) (Table 2.1) [4, 11, 46]. The Coe variant p.(Asn152ThrfsTer26), the missense variants p.(Cys574Arg) and p.(Glu416Lys) and the Yates splice variant were not submitted individually; numbering equivalence follows from the confirmed transcript identity [46].

2.6. 10p15.3 deletions

The deletions in the DeScipio series ranged from 154,558 bp to about 4 Mb and included ZMYND11 in all probands but one and DIP2C in all but another; the proband with a deletion of ZMYND11 only, confirmed by multiplex ligation-dependent probe amplification (MLPA) and fluorescence in situ hybridisation (FISH), had intrauterine growth restriction, HC −2.5 standard deviations (SD), cleft uvula and a wide mouth, and could not be distinguished clinically from the proband without ZMYND11 in the deletion [6]. In the Coe series, ZMYND11 deletions were found in 8 cases and in no controls [24]. In five prenatal cases from 2026, with deletions of 0.14 to 3.17 Mb (one intragenic and three including the whole gene), ZMYND11 was the only haploinsufficient gene common to all five [8]. The reviews indicate minimal clinical differences between large and small deletions that include ZMYND11, but deletions extending into 10p14 may add the GATA3 (HDR) phenotype or that of the DGS2 region [6, 14].

2.7. Penetrance, expressivity and modifiers

Most variants arise de novo, but vertical transmission is documented: an affected mother with three affected children, a carrier father with mild learning difficulties, a father with delay (walking at 3–4 years), learning difficulties, aggressiveness in childhood and mood swings, who learned to read, write and calculate and obtained a driving licence [11], and a mother with a 10p15.3 deletion and learning difficulties [4, 6, 9, 24]. Among five maternal transmissions gathered by Oates and colleagues, two mothers were mosaic and three were affected [9]. The authors concluded that there is reduced penetrance and variable expressivity, without a reliable genotype-phenotype correlation outside the two recurrent missense variants [9]. The sources consulted do not identify genetic modifiers.

2.8. Allelic phenotypes and associations outside the classical picture

The same gene appears in clinical contexts that broaden the spectrum. In an exome study of 1924 patients with dystonia, mostly isolated and of late onset, two ZMYND11 variants — p.(Arg495Ter), in a patient with childhood-onset generalised dystonia, and p.(His315dup), in a patient with cervical dystonia, tremor and polyneuropathy — were considered presumably pathogenic, and the authors allow for a role of the gene in dystonia even without neurodevelopmental features [30]. A consecutive Dutch cohort of 34 patients with progressive myoclonic ataxia includes ZMYND11 among the molecular causes [31]. In an exome meta-analysis of 28,898 schizophrenia cases and 103,041 controls, ZMYND11 was associated with schizophrenia at a false discovery rate of 5%, without reaching exome-wide significance [33]. These associations do not establish allelic entities of their own with a nosological code; they broaden the clinical spectrum associated with the gene and rest on small numbers of carriers.

3. Clinical Spectrum and Manifestations by System

3.1. The cohorts and what they measure

Clinical knowledge rests on small series, recruited through genotype-sharing networks (GeneMatcher, the DDD study — Deciphering Developmental Disorders, ClinVar, epilepsy networks) and on literature reviews that gather the same patients several times. The denominators change from feature to feature because not all reports describe all systems, and series recruited in epilepsy centres overestimate seizures. All the data describe natural history under usual care; there is no disease-modifying treatment that would allow treated and untreated disease to be compared.

Table 3.1. Main cohorts and series

Table 3.2. Frequency of features in the systematic review of 50 cases, by variant class (denominator: individuals with information) [10]

The text of the same article gives some denominators that do not match the table (for example, ID in 34/35 and feeding difficulties in 15/22, and the feeding problems row of the table, 39/48, matches the count of craniofacial dysmorphism given in the text), so the percentages in Table 3.2 describe orders of magnitude and not precise estimates [10].

3.2. Development and cognition

Developmental delay is universal in the published series and affects language above all. In the Yates series, all 14 patients with information had developmental and speech delay; independent walking occurred at a median of 24 months (17 months to 4 years) and first words at a median of 2.5 years (2 to 4 years), and two patients were not walking or speaking at 2.5 and 4 years [4]. ID was mild to moderate in all 13 assessed in that series, and four previously published patients had severe ID [4]. In the review of 50 cases, mild ID predominated with loss-of-function variants (13/17; 77%), whereas severe ID was more frequent with missense variants [10].

The spectrum includes people without ID. In the Coe series, a 17-year-old girl with the de novo frameshift variant p.(Thr70AsnfsTer12) had cognition classified as normal in the article's table, with speech delay, social difficulties and behavioural abnormalities [24]; the clinical description summarised in OMIM indicates, at 8 years, a full-scale intelligence quotient (IQ) of 73 (verbal 80, non-verbal 68 and perceptual organisation 62), that is, in the borderline range, and social and emotional developmental difficulties that suggested ASD without meeting the diagnostic criteria [11]. In the proband of the Coe family with p.(Met187IlefsTer19), IQ was 63 at 9 years, 66 at 18 and 55 at 25, the only longitudinal cognitive trajectory described in the sources consulted [11]. Oates and colleagues described a girl with a full-scale IQ of 87 and very low adaptive functioning (0.2nd percentile on the ABAS-III), and concluded that delay and ID are not always present and that the neurological examination may be normal [9]. Adult carriers with a mild phenotype, sometimes identified only after a child's diagnosis, show the same variability [4, 29].

3.3. Behaviour and mental health

Behavioural abnormalities are the feature that weighs most on family life. In the Yates series they occurred in 14/16 (88%): attention deficit, hyperactivity and impulsivity in 8/16 (50%), aggressiveness in 8/16 (50%) and ASD or autistic traits in 3/15 (20%) [4]. In the Coe series, the psychiatric phenotype was present in 3/5 patients with information, the three men had aggressiveness, tantrums and outbursts of anger, and the two women had autistic tendencies [24]. Among the 20 patients with epilepsy in the Oates series, 9 had ASD, sometimes without severe ID, and 13 had attention deficit hyperactivity disorder (ADHD) or challenging and aggressive behaviour [9]. The comparison by Minale and colleagues suggests autistic behaviour is more frequent with sequence variants (16/26; 61%) than with deletions (3/29; 10%), but the counts depend on what each report described [14].

In adulthood, the oldest man in the Coe series, at 45 years, had diagnoses of borderline personality disorder, bipolar disorder, psychosis, depression, low frustration tolerance with aggressiveness and ADHD; a de novo frameshift variant was described in a patient with schizophrenia [24]. In the table of the same article and in the OMIM entry, the carrier of p.(Gln326Ter) is 41 years old and, according to OMIM, has rapid-cycling bipolar disorder, borderline personality disorder, pervasive developmental disorder, psychosis and substance abuse, with a normal physical examination apart from hypertelorism and cubitus valgus; the sources do not clarify whether the age of 45 years in the article's text corresponds to another observation of the same patient [11, 24]. The proband with p.(Met187IlefsTer19), at 25 years, had low frustration tolerance and aggressive and provocative behaviour, treated with risperidone [11]. The authors of the 2020 series stress that these abnormalities may constitute the main psychosocial burden, particularly when ID is mild and social expectations are higher [4].

3.4. Epilepsy

Epilepsy was observed in 5/16 (31%) of the Yates series, in 39% of the 27 patients then known and in 22/47 (47%) in the review of 50 cases [4, 10]. In the Oates sample of 47 individuals, 20 (43%) had epilepsy, a figure influenced by recruitment through epilepsy networks [9]. In ASD in general, the risk of epilepsy increases in adolescence, particularly when there is significant ID, which justifies continued attention to seizures beyond childhood [48]. In the DeScipio deletion series, 3/7 had seizures; the Minale review counted seizures in 19/29 deletions and 18/26 variants, counts that depend on what the reports recorded [6, 14].

Oates and colleagues grouped the 20 patients into three sets [9]. Eight had definite or probable idiopathic focal epilepsy, four of them with atypical benign partial epilepsy (ABPE) and one with rolandic epilepsy; others had focal impaired-awareness seizures of parietal origin, well controlled, or focal seizures preceded by multiple febrile convulsions. Four had generalised or infantile-onset epilepsies: a probable juvenile absence epilepsy with 3–3.5 Hz generalised spike-wave, a possible childhood absence epilepsy, and two infantile-onset pictures with multiple drug-resistant generalised seizure types, generalised spike-wave and posterior disorganisation of the background activity, compatible with infantile epileptic encephalopathy. Eight had insufficient description for classification [9]. In the Yates family with p.(Gln586Ter), one of the three siblings had seizures and the mother had had seizures in childhood [4, 11].

ABPE initially resembles epilepsy with centrotemporal spikes, but it begins earlier (4–7 years), includes focal motor, myoclonic, myoatonic seizures or atypical absences and may evolve with continuous spike-wave activation in sleep, language and social communication deficits and transient drug resistance [9]. In the 2022 International League Against Epilepsy (ILAE) classification, rolandic epilepsy is called self-limited epilepsy with centrotemporal spikes (SeLECTS), and electrical status epilepticus in sleep (ESES) or continuous spike-wave in sleep (CSWS) is now designated (developmental and) epileptic encephalopathy with spike-wave activation in sleep (DEE-SWAS or EE-SWAS) [49]. The ABPE of the 2021 series corresponds, in this nomenclature, to the atypical SeLECTS spectrum with evolution to EE-SWAS in some patients; the correspondence is interpretative, because the 2022 classification was consulted only in abstract [9, 49]. In 2026, monozygotic twins with ZMYND11-related DEE-SWAS were published, known only from the title of the letter [50].

Response to treatment varies between spontaneous remission, control with monotherapy and drug resistance: in 4 of the 20 patients the data suggest drug resistance, in 3 there was spontaneous remission, control with monotherapy or seizure reduction, and in 13 the response was not known [9]. The cases with recurrent missense variants illustrate the severe extreme: the 2016 patient with p.(Ser421Asn) had seizures from 9 months of age, intractable to ten antiseizure medications and the ketogenic diet, and the adult patient with p.(Arg600Trp) had recurrence of seizures from the age of 20, with drug-resistant myoclonic, tonic-clonic and atonic seizures [26, 29].

3.5. Tone, movement and gait

Hypotonia was recorded in 5/16 (31%) of the Yates series, in 48% of the 27 patients known in 2020 and in 19/31 (61%) of the 2026 review, more frequently with missense variants (9/10) than with loss-of-function variants (10/21) [4, 10]. Broad-based ataxic gait has been described in patients with recurrent missense variants, and athetoid stereotyped movements appeared in a 24-year-old patient [20, 26, 29].

Movement disorders extend into adult life. In an Innsbruck series of four adults with rare neurodevelopmental disorders diagnosed late, all had early delay and a movement disorder (dystonia, ataxia or tremor) that appeared or worsened in adulthood, and the authors recommended neurological surveillance beyond childhood [29]. Up to 2022, ataxia had been described in one patient and dystonia in two others [29]. The deletion series includes tremor and myoclonus in one patient [6], and the gene is among the causes of progressive myoclonic ataxia, in which ataxia begins in the first two years and myoclonus in the first five [31].

3.6. Neuroimaging

Magnetic resonance imaging (MRI) is normal or shows non-specific abnormalities. In the Oates series, prenatal infarction, ventricular dilatation or atrophy, delayed myelination, Chiari type I malformation in two patients, glial scarring and, in one patient, posterior polymicrogyria were described [9]. Carriers of recurrent missense variants had cortico-subcortical atrophy and delayed myelination [20, 26], and the adult with p.(Arg600Trp) had global atrophy and turricephaly at 26 years [29]. In the deletion series, 4/6 patients with imaging had abnormalities, including arachnoid cyst and pachygyria [6]. The 2026 carrier of p.(Arg600Trp) had a normal MRI at 6 months [10].

3.7. Growth, feeding and digestive tract

Growth and HC are generally normal with loss-of-function variants; in the Yates series, birth weight was at or above the 98th percentile in 3/14 (21%) [4]. Short stature after 10 years of age was identified as a feature of deletions and of loss-of-function variants in the comparison by Tumiene and colleagues [15]. Microcephaly, birth small for gestational age and short stature are characteristic of p.(Arg600Trp) [10]. The two Polish patients with exon 13 missense variants were large at birth (weight +2.39 and +1.94 SD; length +4.03 and +4.77 SD), which shows that the growth pattern is not uniform [20].

Feeding difficulties appear in the first months. In the Yates series they affected 6/13 (46%), mainly excessive vomiting after meals and bottle feeds lasting more than an hour; in the set of 27 patients known in 2020 they reached 59%, usually mild, and only three required supplementary nasogastric tube feeding [4]. In the 2026 review, feeding problems were recorded in 39/48 (81%) [10]. The digestive spectrum includes gastro-oesophageal reflux disease (GORD), food aversion, constipation, hypertrophic pyloric stenosis, severe eosinophilic oesophagitis and, in a 32-year-old woman with a 10p15.3 deletion, refractory gastroparesis confirmed by gastric emptying scintigraphy [6, 26, 51]. A patient with p.(Val418Gly) required Nissen fundoplication and gastrostomy for GORD with aspiration pneumonias at 12 months [20].

3.8. Craniofacial dysmorphism and limbs

The dysmorphism is subtle and not constant: it was observed in 11/16 (69%) of the Yates series and in 12/20 of the Oates patients with epilepsy [4, 9]. The shared features are thick eyebrows, prominent eyelashes, a depressed nasal root with a bulbous nasal tip, anteverted nares, a thin upper lip vermilion and a wide mouth [4]. Coe and colleagues recorded hypertelorism (6 patients), ptosis (3) and a wide mouth (4) [24]. The individual descriptions summarised in OMIM add brachycephaly, upslanting palpebral fissures, an arched upper lip and diastema in the girl with p.(Gln587del); small ears with overfolded upper helix and prominent antihelix, mild ptosis, joint laxity, clinodactyly, hallux valgus and long second and fourth toes in the girl with p.(Thr70AsnfsTer12), with a normal brain computed tomography (CT) scan at 21 months; synophrys, ptosis and hypertelorism in the proband with p.(Met187IlefsTer19); and a long philtrum, asymmetric skull, deep-set eyes, hypertelorism and overfolded upper helix, with CT showing mild atrophy, in the man with p.(Glu416SerfsTer5) [11]. In 10p15.3 deletions, the Minale review counted highly arched eyebrows in 21/29 and low-set ears in 24/29, features that are rare with sequence variants (1/26 and 2/26) [14]. Brachydactyly has been described in two patients and dental enamel hypoplasia in two, in a context in which formal dental assessment was rarely documented [4]. Some patients with missense variants or with larger deletions have features reminiscent of Cornelia de Lange syndrome (synophrys, hypertrichosis, open mouth, low hairline) [14, 20].

3.9. Eyes, hearing and other systems

Strabismus was recorded in 8/23 (35%) of the 2026 review, mainly with missense variants (6/8 versus 2/15 with loss-of-function variants), and nystagmus, myopia and hypermetropic astigmatism appear in individual reports [10, 20, 26]. Mild to moderate hearing loss has been described in patients with missense variants, in one of whom a GJB2 variant and a family history of deafness coexisted, which precludes attributing the hearing loss to the gene [20, 26].

Genitourinary anomalies include cryptorchidism (mainly in carriers of p.(Arg600Trp)), hypospadias, transient bilateral hydronephrosis and inguinal hernia; in the deletion series, bicuspid aortic valve with mitral regurgitation, congenital diaphragmatic hernia, craniovertebral junction anomaly, kyphosis, hypoplastic or inverted nipples and alopecia totalis were also recorded, each in a single patient [6, 10, 15]. A tetralogy of Fallot was counted in one deletion [14].

Recurrent infections (otitis, sinusitis, lower respiratory tract infections) have been described in deletions and in a patient with p.(Ser421Asn), who also had eosinophilic oesophagitis, asthma, atopic dermatitis and multiple allergies with anaphylaxis; the hypothesis of a link to regulation of the NF-κB pathway through TRAF3 is only speculative [6, 15, 26]. A patient with p.(Val418Gly) had hyperinsulinaemic hypoglycaemia treated with diazoxide, which the authors themselves acknowledge may have been coincidental or secondary to the fundoplication [20]. A 48-year-old man with a 10p15.3 deletion had a stroke of unknown cause at 38 years [6]; this is an isolated finding.

3.10. Regression and decompensation triggered by infection

The disease is not degenerative: the published sources do not describe a progressive course [2]. There are, however, transient losses of skills in two contexts. Spike-wave activation in sleep and seizure clusters are accompanied by cognitive and language regression, sometimes over several days [9]. In an Australian series of eight children with de novo variants in chromatin genes, including one with ZMYND11 p.(Cys249SerfsTer2), all had neurological regression or abrupt-onset neuropsychiatric syndromes triggered by infection or vaccination, at a mean age of 4 years and 2 months [32]. This is a descriptive series without controls.

3.11. 10p15.3 microdeletion form

In the DeScipio series, with a full history in 12 individuals, cognitive, behavioural and developmental differences were present in 11/11, speech delay in 10/10, motor delay in 10/10, dysmorphism in 9/12, hypotonia in 7/11, brain anomalies in 4/6 and seizures in 3/7 [6]. Tumiene and colleagues added low birth weight, short stature after 10 years of age, genitourinary malformations and recurrent infections [15]. The complex partial seizures of two patients were controlled with low-dose carbamazepine, and one adult had tonic epilepsy from 14 months, then tonic-clonic seizures and absences on valproate, carbamazepine and primidone [6]. Deletions extending into 10p14 add features from other genes; an Italian woman with a 6 Mb deletion (10p15.3p14) and a 7p22 duplication had cleft palate, vesicoureteral reflux, patent ductus arteriosus, severe short stature (−5.1 SD) and a Cornelia de Lange spectrum phenotype [14].

Table 3.3. Comparison of 10p15.3 deletions and sequence variants in the 2026 review (counts over the total number of patients in each group, not over patients with information) [14]

3.12. Subtypes due to recurrent missense variants

The three published carriers of p.(Arg600Trp) — Cobben's boy, an Austrian woman seen at 34 years and a Japanese boy — share fetal growth restriction, microcephaly, short stature, moderate to severe ID and, in the boys, cryptorchidism; broad nasal alae and inverted, widely spaced nipples appeared in two [10, 25, 29]. The two carriers of p.(Ser421Asn) — a 24-year-old North American woman and a 5.5-year-old Polish boy — had hypotonia, epilepsy or paroxysmal episodes, severely delayed walking, absent expressive language, strabismus or nystagmus and delayed myelination; the 2016 patient had a sociable behaviour with an almost constant smile that prompted exclusion of Angelman syndrome [20, 26]. The number of carriers is too small to define formal clinical subtypes.

The first carrier of p.(Arg600Trp), described by Cobben and colleagues, was a 7-year-old boy, the son of unrelated Russian parents [1, 25]. The pregnancy was complicated by intrauterine growth restriction, increased nuchal translucency and oligohydramnios; he required tube feeding and had cryptorchidism, urethral stenosis and pyloric stenosis, with height at −4.2 SD and HC at −2.9 SD [1]. The dysmorphism included a mild metopic ridge, low-set ears, telecanthus, epicanthic folds, narrow upslanting palpebral fissures, a broad nasal root and broad alae, a thin upper lip, a smooth philtrum, retrognathia, widely spaced inverted nipples, short tapering fingers, deep palmar creases and broad feet, together with vertebral anomalies [1]. MRI at 13 months showed delayed myelination; at 6 years he walked with a walking aid, said two words and had pyramidal signs [1]. No functional studies were performed, and Yates and colleagues considered that the different phenotype suggested an effect distinct from haploinsufficiency, possibly gain of function [1, 4].

3.13. Prenatal period

Prenatal signs are non-specific. Of five fetuses with 10p15.3 microdeletion, two were identified in the first trimester because of increased nuchal translucency and three had normal ultrasound scans in the first and second trimesters and were diagnosed in the third because of fetal growth restriction [8]. Fetal growth restriction and placental dysfunction have been described in carriers of p.(Arg600Trp) [10, 29].

4. Clinical Recognition and Warning Signs

4.1. When to suspect the disorder

There is no pathognomonic sign, and dysmorphism is absent in about one third of patients, so most diagnoses result from broad-spectrum sequencing requested for developmental delay, and not from targeted clinical suspicion [4, 9]. The combination that should lead to the inclusion of ZMYND11 in panels and in exome reanalysis is the following: speech delay disproportionate to motor delay, mild to moderate ID, behavioural abnormalities with aggressiveness, impulsivity or autistic traits, hypotonia and early feeding difficulties with vomiting, and a face with thick eyebrows, prominent eyelashes, a bulbous nasal tip and a wide mouth [4, 24]. The dysmorphism is useful for reverse phenotyping, that is, for confirming the plausibility of a variant found by exome [4].

By age, the picture presents differently. In the infant, hypotonia, feeding difficulties with vomiting and delayed milestones predominate; in carriers of p.(Arg600Trp), fetal growth restriction, microcephaly and cryptorchidism [4, 10]. At preschool and school age, speech delay, hyperactivity, aggressiveness and focal or generalised seizures stand out [4, 9]. In the adult, the diagnosis may be made late, in the context of a movement disorder (dystonia, ataxia, tremor), worsening seizures, multiple psychiatric diagnoses or the assessment of an affected child [4, 24, 29].

In an epilepsy context, the gene should be considered in a child with SeLECTS-like focal epilepsy of early onset, with myoclonic or atonic seizures or atypical absences, spike-wave activation in sleep and a language disorder or autism spectrum disorder, particularly if ADHD or ID coexists [9, 49]. A 10p15.3 deletion found by microarray should be interpreted as a diagnosis of this disorder when it includes ZMYND11 [3, 6].

4.2. Warning and emergency signs

The signs that call for urgent or short-term assessment are: loss of language or cognitive skills, worsening behaviour or increased seizures, which require exclusion of spike-wave activation in sleep with a sleep EEG [9]; abrupt-onset regression after infection or vaccination, described in chromatin genes including ZMYND11 [32]; clusters of myoclonic or atonic seizures with falls [9]; persistent vomiting, food refusal, weight loss or aspiration pneumonias, which point to severe GORD, eosinophilic oesophagitis or gastroparesis [20, 26, 51]; hypoglycaemia in an infant or child fed by gastrostomy or after fundoplication [20]; and, in the adult, the onset of dystonia, ataxia, tremor or recurrence of seizures after a controlled period [29].

4.3. Newborn screening

There is no newborn screening, because there is no biochemical biomarker and no treatment whose presymptomatic initiation would alter the course [2].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Criteria

There are no formal clinical criteria or scoring system. The diagnosis is molecular: identification, in an individual with a neurodevelopmental disorder, of a heterozygous pathogenic or likely pathogenic ZMYND11 variant, according to the criteria of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP), or of a deletion that removes all or part of the gene [3, 4, 20]. ClinGen classified the gene–disease relationship as definitive and dosage sensitivity as sufficient haploinsufficiency, which allows deletions of the gene and null variants to be classified on the basis of the established mechanism [3]. There is no ClinGen actionability assertion and the gene is not on the ACMG secondary findings list v3.2 [2, 3].

5.2. Tests in order

The 2021 ACMG guideline strongly recommends considering exome or genome sequencing as a first- or second-tier test in patients with congenital anomalies with onset before 1 year of age or with developmental delay or ID with onset before 18 years, because of a higher yield than conventional tests [16]. In non-specific ID, a recognisable clinical diagnosis is made in about 25%, the conventional karyotype adds about 4% and high-resolution microarray a further 14%, with a mean yield of 18% as a first-line test; the exome identifies the causal monogenic variant in up to 50% of sporadic severe ID [52]. In a child with seizures and delay, MRI, microarray and a panel, exome or genome, and an EEG for specific patterns are recommended [53]. In ASD, microarray is indicated when there is ID or dysmorphism, and the exome or genome has a low yield in non-syndromic ASD and a higher yield when there are additional signs of monogenic disease, such as the ID, epilepsy or dysmorphism that characterise this disorder [48].

In practice, there are two routes to this disorder. If the first test is microarray, a 10p15.3 deletion identifies the microdeletion form, and the report should list ZMYND11 and the other genes included [6, 8]. If the first test is trio exome or genome sequencing, it identifies sequence variants; the exome also allows CNV calling, as in the Italian case in which the 6 Mb deletion was detected in the exome data and confirmed by microarray [14]. Several patients had a normal microarray before the exome revealed the variant, and a normal microarray does not exclude the disorder [10, 20, 26]. Investigation of ID should include testing for the FMR1 expansion, which the exome does not detect [52].

5.3. Test limitations

Several limitations must be taken into account. Small intragenic deletions, such as that of Huynh and colleagues, which covers the 5′ untranslated region and the first two exons, may escape low-resolution microarrays and exome pipelines without CNV calling [7]. In a carrier of a cryptic balanced translocation, the microarray is normal; in apparently isolated terminal deletions, study of the parents by karyotype, FISH or genome is required to exclude a parental balanced rearrangement, as the 46,XX,der(10) karyotype of the Italian case suggests [14, 54]. About 2% of apparently de novo interstitial CNVs result from balanced insertions in a parent, which MLPA, quantitative polymerase chain reaction (PCR) and microarray do not detect in the carrier [52]. Low-level mosaicism in a parent may not be seen by Sanger sequencing of blood; two mosaic mothers were identified in the 2021 series [9]. Microarray with single-nucleotide polymorphism (SNP) probes is more sensitive than comparative genomic hybridisation for mosaicism [54].

The clinical sensitivity of the tests for this disorder has not been measured: there are no studies testing a clinically defined cohort with several methods, and the proportion of variants detectable by each technique is not equivalent to sensitivity. A negative exome does not exclude variants in poorly covered regions, deep intronic variants or structural rearrangements; in the 2026 Japanese case, mean exome coverage was 60.88×, with 97% of coding regions covered by at least 10 reads and the whole of ZMYND11 by at least 20 [10].

5.4. Variant interpretation

The report should state the gene, the transcript with its version (NM_001370100.5), the HGVS c. and p. nomenclature, the zygosity, the origin (de novo, inherited, unknown) and the ACMG/AMP class; for CNVs, the genomic coordinates with the genome version, the size and the genes included [3, 20]. Null variants (nonsense, frameshift, canonical splice) in a gene with established haploinsufficiency can be classified as pathogenic, except those in the last exon, which escape NMD and require additional arguments [3, 4].

Missense variants are the main interpretive problem. In the two Polish cases of 2024, the variants p.(Ser421Asn) and p.(Val418Gly), absent from gnomAD, in a highly constrained coding region and with CADD (Combined Annotation Dependent Depletion) scores of 24.1 and 28.5, remained variants of uncertain significance (VUS) with 3 points without the PS2 criterion and became likely pathogenic with 7 points once they were confirmed to be de novo [20]. p.(Arg600Trp) is pathogenic by the criteria PS1, PM2, PS2 and PP3 [10]. The missense constraint of the gene (z = 4.22) supports the pathogenicity of variants in functional domains, but does not obviate segregation [26]. Because there are carriers with a mild phenotype and familial transmission, inheritance from an apparently unaffected parent does not exclude pathogenicity, and clinical assessment of the parents is part of the interpretation [4, 9].

There is no biochemical biomarker, enzyme activity or validated DNA methylation signature for ZMYND11 in the sources consulted, unlike other chromatin disorders [2].

5.5. Complementary assessment after diagnosis

After confirmation, the baseline assessment includes EEG (preferably with sleep) if there are seizures, paroxysmal episodes or regression; brain MRI if there is micro- or macrocephaly, focal neurological signs or epilepsy; formal assessment of development, language and behaviour with validated instruments (for example, ADOS-2, ADI-R, Conners-III, WISC-V and ABAS-III, used in the 2021 series); nutritional and gastroenterological assessment; ophthalmological examination; hearing assessment; dental examination; and genitourinary assessment in boys [4, 9, 10, 52]. The justification for each examination is the frequency of the findings described in section 3.

5.6. Practical algorithm

Table 5.1. Diagnostic algorithm

5.7. Prenatal and preimplantation diagnosis

Specific prenatal testing is only possible when the proband's variant or deletion has been identified; in that case, it is performed by targeted testing of a fetal sample, and without a laboratory diagnosis in the proband there is no specific prenatal test [52]. Preimplantation genetic testing depends on the same condition, and the sources consulted do not describe its application to this disorder. The 10p15.3 microdeletion may be identified by microarray in pregnancies with increased nuchal translucency or fetal growth restriction, but the ultrasound signs are non-specific and most fetuses will have no ultrasound findings [8].

6. Differential Diagnosis

6.1. Principles

The differential is broad because the core phenotype (speech delay, mild to moderate ID, behavioural abnormalities, hypotonia, epilepsy and subtle dysmorphism) is common to hundreds of monogenic and chromosomal causes of neurodevelopmental disorder; more than 950 ID genes had been identified by 2017 [52]. In practice, the differential is resolved by broad-spectrum genomic testing, and the usefulness of clinical comparison lies in three situations: deciding when a targeted panel is not enough, interpreting a variant of uncertain significance in the light of the phenotype, and recognising syndromes with different treatment, cancer surveillance or recurrence risk [16, 53]. The most useful differential groups are the other disorders of the H3K36 pathway and of chromatin, the syndromes with a similar epileptic phenotype and the larger 10p deletions.

Table 6.1. Differential diagnosis

6.2. The H3K36 pathway and overgrowth syndromes

The comparison with Sotos syndrome, SETD2 disorders and Tatton-Brown-Rahman syndrome is mainly conceptual: all three affect the deposition or reading of H3K36 methylation and share ID, ASD and behavioural abnormalities, but they are overgrowth syndromes, with macrocephaly and, in the last two, excess weight, and with DNA methylation signatures that allow VUS to be classified [43–45]. In the ZMYND11 disorder, growth and HC are generally normal, and microcephaly, when present, points to p.(Arg600Trp) [4, 10]. A patient with overgrowth, macrocephaly and ID points towards NSD1, SETD2, DNMT3A and the other overgrowth syndromes; a patient with normal growth, speech delay, aggressiveness and centrotemporal epilepsy points towards ZMYND11 and the focal epilepsy genes.

6.3. Other chromatin disorders with dysmorphism

Coffin-Siris, Rubinstein-Taybi, Sotos and Cornelia de Lange syndromes have been identified as overlapping phenotypes in patients with ZMYND11 missense variants [20]. The dysmorphism of some carriers — thick eyebrows, low hairline, open mouth, hypertrichosis — can resemble Cornelia de Lange syndrome, and a ZMYND11 variant was found in a family studied for that suspicion [20, 28]. The Italian authors of 2026 propose including ZMYND11 in the assessment of chromatinopathies with Cornelia de Lange spectrum features [14]. The distinction is made by molecular testing, and panels restricted to the classical Cornelia de Lange syndrome genes do not necessarily include ZMYND11.

6.4. Differential diagnosis of the epileptic phenotype

Focal epilepsy with centrotemporal spikes and encephalopathy with spike-wave activation in sleep have several genetic causes. GRIN2A is the most frequent, in up to 20% of patients with ABPE, Landau-Kleffner syndrome or EE-SWAS, and is associated with speech disorder and verbal dyspraxia; ZMYND11 was the first chromatin remodelling gene associated with idiopathic focal epilepsies [9, 60]. 16p11.2, 16p13.11, 1p36 and Xp22.31 rearrangements, SLC6A1 variants and regulatory variants of PAX6 and miR-328 have also been linked to this spectrum [9]. ABPE may be confused with Lennox-Gastaut syndrome or with epilepsy with myoclonic-atonic seizures, which has consequences for the choice of treatment [9, 49]. In typical SeLECTS, prior development is normal and the course is self-limited, whereas the presence of ADHD, ASD, ID or dysmorphism justifies a genomic test [9, 61].

6.5. 10p deletions and cytogenetic differential

Small 10p15.3 deletions are distinguished from larger deletions by the genes included. A deletion extending into 10p14 may include GATA3 and cause hypoparathyroidism, deafness and renal dysplasia, and one reaching 10p14-p13 may give the DiGeorge type 2 picture; in these cases surveillance is dictated by the additional genes [6]. Orphanet also classifies, as neighbouring entities, distal 10p deletion (ORPHA:1580) and 10p11.21p12.31 deletion (ORPHA:284169) [12]. In OMIM, the 10p12-p11 deletion has no entry of its own: it appears as an entity included in DeSanto-Shinawi syndrome (OMIM 616708), caused by heterozygous loss of function of WAC, a gene located in the minimal common region of about 360 kb of the published interstitial deletions, which ranged from 1 to 10 Mb [62]. This entity shares with the ZMYND11 disorder global delay, hypotonia, behavioural disorders and constipation, and is distinguished by the combination of a broad forehead, bulbous nasal tip, synophrys and refractive errors, and, in larger deletions, by a higher frequency of heart disease [62]. In HDR syndrome, the 2018 review of 180 published patients (123 with an identified GATA3 variant) found sensorineural deafness in 96.7%, hypoparathyroidism in 93.3% and renal disease in 72.2%, with the complete triad in 77 of the 123 patients with a variant, which shows that absence of one of the components does not rule out the diagnosis [21]. Additional components described include hypogonadotropic hypogonadism, uterine and vaginal malformations (among them the case of a 29-year-old Portuguese woman with uterovaginal agenesis, severe hypoparathyroidism and mild deafness, without renal disease), heart disease, pigmentary retinopathy, basal ganglia calcifications and, in some patients, cognitive impairment or autism [21]. One case is directly relevant to the 10p15.3 region: a child with the HDR triad, psychomotor delay, bilateral cleft lip and palate and tetralogy of Fallot had a de novo 6.5 Mb deletion at 10p15.3-p15.1 associated with a 1.9 Mb duplication that included GATA3, without a variant in the gene sequence, which led the authors to propose that both loss and gain of GATA3 dosage can produce the phenotype; this is an isolated observation [21]. For DGS2, a series of 22 patients diagnosed with DiGeorge syndrome without a 22q11.2 deletion revealed no deletion of the 10p critical region by FISH, which indicates that the 10p deletion is a rare cause of the phenotype and is nowadays identified by microarray [22]. A terminal deletion may result from an unbalanced translocation inherited from a parent carrying a balanced rearrangement, as suggested by the Italian case with der(10) and 7p22 duplication, which changes the recurrence risk [14, 54].

7. Treatments and Risk Management Strategies

7.1. Principles and organisation of care

There is no specific or disease-modifying treatment, and no scientific society has published a treatment guideline for the ZMYND11 disorder or for the 10p15.3 microdeletion [2, 17]. Management is symptomatic and directed at each manifestation, with the same drugs and interventions used for epilepsy, ADHD, irritability associated with autism and feeding disorders of other causes. The patient organisation Unique, with clinical review by the authors of the 2020 series, recommends a multidisciplinary team with a geneticist, paediatrician, paediatric neurologist or neurologist and epilepsy specialist, and therapies tailored to each child's needs, such as speech and language therapy [17]. The series describe speech and language therapy, occupational therapy and physiotherapy, and special educational needs in all members of one affected family, two of whom went on to be employed [4, 6]. The 2014 authors suggested that early diagnosis may influence prognosis, which is an opinion without intervention data [24].

The doses given in the following subsections come from the official United States Food and Drug Administration (FDA) labels for the approved indications, because no guideline or trial defines doses for this disorder; their application to patients with ZMYND11 variants is an extrapolation from the approved indication (seizure type, ADHD or irritability associated with autism) and should be read as such [63–67]. The summaries of product characteristics (SmPC) of the European Medicines Agency (EMA) and INFARMED and the ANVISA package inserts were not consulted for this edition, with the exception of the harmonised European SmPC for risperidone, so the US–EU comparison of doses and age indications is, for the other drugs, limited to the North American side [68].

7.2. Epilepsy

Treatment of seizures follows the classification of the seizure type and epilepsy syndrome, and not the gene, because there are no studies comparing drugs in this disorder [9]. In the published cases, levetiracetam, valproate, ethosuximide, carbamazepine, phenobarbital, lamotrigine and, in drug-resistant cases, the ketogenic diet and vagus nerve stimulation were used, with variable responses [6, 9, 20, 26, 29]. In the 2012 series, the complex partial seizures of two patients with deletions were controlled with low-dose carbamazepine [6]. In Oates patient #7, with ABPE and spike-wave activation in sleep, sodium valproate was added to levetiracetam and then to ethosuximide; in patient #10, with EE-SWAS, valproic acid, levetiracetam and the classic 4:1 ketogenic diet were unsuccessful [9]. A boy with p.(Ser421Asn) became seizure-free on levetiracetam, later discontinued at 4 years [20].

In drug-resistant epilepsy, vagus nerve stimulation, implanted at 15 years, reduced the atonic and tonic-clonic seizures of the 2016 patient, who had not responded to ten antiseizure medications or to the ketogenic diet; the adult patient with p.(Arg600Trp) was on five antiseizure medications and vagus nerve stimulation at 34 years [26, 29]. These are observations from isolated cases, without comparison.

Suspected EE-SWAS, suggested by regression of language, behaviour or cognition, requires a sleep EEG, because it changes the therapeutic strategy [9, 49]. The use of ethosuximide in spike-wave activation in sleep, as in patient #7, is outside the labelled indication, which is absence epilepsy [9, 65]. The sources consulted do not include the therapeutic recommendations of epilepsy guidelines for EE-SWAS; the choice rests with the epilepsy specialist.

7.3. Behaviour, attention and sleep

Behavioural abnormalities are frequent and are treated with a behavioural and educational approach and, when necessary, pharmacologically, without disorder-specific data [4, 17]. Aggressiveness, tantrums and irritability correspond to the indication of risperidone for irritability associated with autistic disorder, and ADHD, described in about half of the patients of the 2020 series, to the indication of methylphenidate [4, 66, 67]. In the European Union, risperidone has no indication in ASD: it is authorised from 5 years of age only for the short-term (up to 6 weeks) symptomatic treatment of persistent aggression in conduct disorder, in children and adolescents with below-average intellectual functioning or intellectual disability. The starting dose is 0.25 mg/day below 50 kg (optimum 0.5 mg/day; maximum 0.75 mg/day) and 0.5 mg/day from 50 kg (optimum 1 mg/day; maximum 1.5 mg/day) [68]. The only use of psychotropic drugs described in the sources for this disorder is that of risperidone in a young man from the Coe series with low frustration tolerance and aggressive and provocative behaviour, with no dose or response reported [11]. In the 2026 Japanese case, self-injury was attributed to the absence of expressive language, a reminder that augmentative communication and speech and language therapy are part of the treatment of behaviour [10].

Levetiracetam can worsen behaviour: the label describes behavioural abnormalities and psychotic symptoms, and in children aggressiveness and irritability are among the most frequent adverse reactions [63]. In a disorder in which aggressiveness is cardinal, this effect should be actively sought after initiation or dose increase. Severe sleep disturbance was described in one patient up to 5 years of age [20]; the sources consulted do not describe its treatment.

7.4. Feeding and digestive tract

Feeding difficulties in infancy are generally mild and resolve with usual measures, but a minority require supplementary nasogastric tube feeding, and severe GORD with aspiration pneumonias led to Nissen fundoplication and percutaneous endoscopic gastrostomy (PEG) at 12 months in one patient [4, 20]. A patient with suspected cow's milk protein allergy received an amino acid formula, macrogol for constipation and a low-FODMAP diet (fermentable oligo-, di-, mono-saccharides and polyols) from 2 years of age; the hypoglycaemia with transient hyperinsulinaemia that occurred at 3 years was attributed by the authors to the restrictive diet [20]. For severe eosinophilic oesophagitis, the sources do not describe the treatment used [26].

In the 32-year-old woman with a 10p15.3 microdeletion and daily vomiting for 1.5 years, omeprazole 20 mg daily titrated up to 40 mg twice daily, with famotidine, had no effect; repeated upper gastrointestinal endoscopy and 96-hour pH monitoring excluded pathological GORD and eosinophilic oesophagitis; scintigraphy showed abnormal gastric emptying; metoclopramide 10 mg three times daily had no effect, and prucalopride 1 mg daily produced significant improvement without adverse effects [51]. The use of prucalopride in gastroparesis is off-label and rests on a single case [51].

7.5. Hyperinsulinaemic hypoglycaemia

In a patient with p.(Val418Gly), gastrostomy and fundoplication, hyperinsulinaemic hypoglycaemia confirmed at 15 months (glucose 21 mg/dL with insulin 21.9 µIU/mL and C-peptide 2.2 ng/mL, without ketonuria) was treated with diazoxide started at 5 mg/kg/day and titrated up to 10.8 mg/kg/day in three doses, with disappearance of hypoglycaemia on continuous glucose monitoring [20]. Diazoxide is the first-line treatment in congenital hyperinsulinism and frequently causes hypertrichosis, fluid retention, pulmonary hypertension, neutropenia and thrombocytopenia [20]. The authors acknowledge that the hypoglycaemia may have been coincidental or secondary to the fundoplication, they found no variants in the hyperinsulinism genes, and the dose described is that of the case, not a recommendation for the disorder [20].

7.6. Regression triggered by infection

In the Australian series of eight children with variants in chromatin genes and neurological regression or acute neuropsychiatric syndrome after infection or vaccination, six improved with intravenous immunoglobulin (IVIG; three patients) or antibiotics (four patients) [32]. The series has no controls, and the hypothesis of immune dysregulation remains to be demonstrated; IVIG in this situation is off-label and does not constitute a recommendation [32].

7.7. Movement disorders, vision, hearing and teeth

The sources recommend neurological surveillance beyond childhood, because dystonia, ataxia and tremor may appear or worsen in adulthood, but they do not describe therapeutic responses in this disorder [29]. In the Japanese case, exotropia and hypermetropic astigmatism led to the prescription of glasses [10]. Dental enamel hypoplasia justifies dental review, which was poorly documented in the series [4].

7.8. Drugs used and reference dosing

Table 7.1. Symptomatic drugs used in the published cases or applicable by extrapolation from the approved indication (FDA label doses; no dose is specific to the ZMYND11 disorder)

The extrapolation has limits that the table does not conceal. The trials supporting these indications did not include patients with ZMYND11 variants, and ID, epilepsy and age alter tolerance to several of these drugs; the dose to be used is that of the label for the corresponding indication and of the specialist, and the individual response in the published cases does not allow drugs to be compared. The doses are those of the label of the formulation consulted; other formulations, especially of methylphenidate, have different doses.

7.9. Regulatory status

No medicinal product has an approved indication for the ZMYND11 disorder or for the 10p15.3 microdeletion, and the search of trial registries and the literature identified no drugs in clinical development and no ongoing regulatory procedures [69]. Table 7.2 summarises the status of the symptomatic drugs.

Table 7.2. Regulatory status of the symptomatic drugs

7.10. Clinical trials and recent developments

Table 7.3. Registered clinical trials (ClinicalTrials.gov, search of 28 October 2026) [69]

There are no registered interventional trials and no negative, suspended or withdrawn trials were identified; a negative search of a registry does not demonstrate that no studies are in preparation [69]. Without trials, there are no response rates, confidence intervals or comparisons to present; all the therapeutic evidence comes from isolated cases and descriptive series.

The publications of the last twelve months are observational or preclinical. On the clinical side, the systematic review of 50 cases and the p.(Arg600Trp) phenotype (February 2026), the comparison of 29 deletions with 26 sequence variants (April 2026), the prenatal series of five microdeletions (July 2026) and the letter on monozygotic twins with DEE-SWAS (2026, known only by its title) [8, 10, 14, 50]. On the preclinical side, the structure of the intermolecular zinc fingers and redox regulation (February 2026), the cortical organoids with partial rescue of defects of other ASD genes by increasing ZMYND11 function (October 2025) and two preprints not peer reviewed [34, 39–41]. A June 2026 review on “therapeutic horizons” is under embargo in PubMed Central until August 2027 and was not read [70]. None of these publications describes a human interventional study, a regulatory opinion or an authorisation.

7.11. Drugs to avoid, pregnancy and lactation

Valproate should not be used in women who are pregnant or planning pregnancy, unless alternatives have failed, because of the risk of congenital malformations, especially neural tube defects, and of reduced IQ and neurodevelopmental disorders after in utero exposure [64]. Before 2 years of age, with polytherapy or with severe epilepsy and ID, the risk of fatal hepatotoxicity is much higher, and the label calls for use as a single drug and with extreme caution [64]. In a disorder in which aggressiveness and irritability are cardinal, the behavioural effects of levetiracetam weigh in the choice [63]. No disorder-specific interactions or contraindications have been described.

In the 2022 review, most published adults had a milder phenotype, without impairment of reproductive fitness, and there are affected mothers with carrier children [4, 29]. The sources consulted do not describe pregnancies of affected women in obstetric detail; management follows that of epilepsy in pregnancy and the counselling in section 8. The fetal growth restriction and placental dysfunction described in fetuses with p.(Arg600Trp) justify fetal growth surveillance when the fetus is a known carrier [10, 29].

7.12. Perioperative context and transition

The sources do not describe anaesthetic complications specific to the disorder. The known risks are aspiration in patients with severe GORD and dysphagia, hypoglycaemia in gastrostomy-fed patients after fundoplication and the haematological effects of valproate [20, 64]. In the transition to adulthood, the sources recommend continued neurological surveillance, because seizures may reappear and movement disorders may appear or worsen after childhood [29]. In severe ID, life expectancy is reduced by the accumulation of health problems and by the delay in treating acute illness caused by communication difficulties, which justifies an attending physician who knows the patient [52].

7.13. Available recommendations and divergences

Table 7.4. Register of guidelines and consensus documents

There are no divergences between societies to record, because no society has published specific recommendations; the only management proposals come from the authors of the series (neurological surveillance in adults, dental assessment, attention to behavioural abnormalities) and from the patient organisation [4, 17, 29].

8. Reproductive Options and Family Counselling

8.1. Inheritance and recurrence risks

The disorder is autosomal dominant. In most patients the variant or deletion arises de novo, but there is transmission from affected parents, sometimes with a mild phenotype, and from mosaic mothers [4, 9, 17]. The recurrence risk depends on the scenario, as summarised in Table 8.1.

Table 8.1. Recurrence risks by scenario

8.2. Parental testing and cascade testing

Testing both parents is indispensable after a diagnosis, for three reasons: it distinguishes a de novo variant from an inherited one, which changes the recurrence risk; it supports the classification of missense variants by the PS2 criterion; and it identifies parents with a mild phenotype who may benefit from their own assessment [4, 20]. Interpretation of a carrier parent should include clinical assessment, because several carrier parents had undiagnosed learning difficulties, delay or behavioural abnormalities [6, 9, 24]. In terminal deletions, parental assessment includes karyotype or FISH for balanced rearrangements [54]. When a parent is a carrier, testing extends to symptomatic siblings and to at-risk relatives who wish to have reproductive information [4].

Testing asymptomatic minor siblings without therapeutic implications raises the issue of predictive testing in minors. Because there is no presymptomatic treatment, testing a sibling without signs has mainly future reproductive value, and the decision may be deferred until the person can participate; a sibling with delay, learning difficulties or behavioural problems should be assessed and tested on clinical grounds. In the 2026 Japanese case, the proband's sibling had poor feeding and articulation difficulty, but the parents were not carriers [10].

8.3. Prenatal and preimplantation diagnosis

With the familial variant or deletion known, targeted testing of a fetal sample is possible; in recurrences due to parental mosaicism the risk is low, but testing may be offered for reassurance [9, 52]. Ultrasound findings are non-specific (increased nuchal translucency, fetal growth restriction), and a normal ultrasound does not exclude the disorder [8]. A positive prenatal result does not allow severity to be predicted, given the variable expressivity among relatives with the same variant [4, 9]. The sources consulted do not describe the application of preimplantation genetic testing to this disorder.

8.4. Content of counselling

Counselling should explain that the disorder results from loss of function of a chromatin regulatory gene; that the spectrum ranges from ASD or ADHD with normal IQ to severe ID with drug-resistant epilepsy; that most cases are de novo but that parents with a mild phenotype may be carriers; that the course is not degenerative, although seizures and movement disorders may appear or worsen in adulthood; and that there is no specific treatment [4, 9, 17, 29]. Families should be informed of the existence of support groups and of the possibility of registration in databases such as DECIPHER, ClinVar and LOVD (Leiden Open Variation Database), which contributed to the delineation of the phenotype [4].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence and what each figure measures

Population prevalence is unknown, and no study has measured it. The available figures are counts of published cases and should not be read as prevalence:

Table 9.1. Available frequency data

The real frequency must be higher than the number of published cases, because patients with a mild phenotype, ASD without ID or isolated epilepsy may not be tested or published, and because many diagnoses made in clinical practice never reach the literature [4, 9]. In 2014, ZMYND11 was one of 10 genes with enrichment of loss-of-function variants in a developmental delay and autism cohort, which places it among the monogenic causes of neurodevelopmental disorder identifiable by exome, without quantifying its fraction [24]. In ID in general, the frequency is 1–2.5% for IQ <70 and 0.3–0.5% for IQ <50 [52].

9.2. Geography and ancestry

The published patients come from the United States, several European countries (United Kingdom, Netherlands, France, Belgium, Germany, Denmark, Switzerland, Lithuania, Poland, Italy, Austria), Japan, China, Argentina and Australia [4, 8–10, 14, 15, 20, 29, 32, 50]. The 2016 patient had Caucasian and Filipino ancestry [26]. Because most variants are de novo and private, no founder variants or populations with increased frequency have been described. The sources consulted do not include patients or data from Portugal or Brazil, which does not demonstrate their absence.

9.3. Special populations

Series that recruited through epilepsy networks overestimate the frequency of seizures, and paediatric series underestimate adult manifestations [9, 29]. In the adult, the diagnosis may arise from a movement disorder or a psychiatric disorder, and the phenotype may be mild enough to allow employment and family life [4, 24, 29]. Late-onset dystonia without evident neurodevelopmental disorder is a possible expression that is still poorly characterised [30]. Fetuses with growth restriction and increased nuchal translucency constitute another population in which the diagnosis can be made [8].

9.4. Reference networks and patient organisations

In Europe, the relevant reference network is ERN ITHACA (rare congenital malformations and rare ID), with a neurodevelopmental disorders working group and the ILIAD registry; for patients with epilepsy, ERN EpiCARE (rare and complex epilepsies), which maintains lists of patient organisations for rare epilepsies and leaflets for families [18, 19]. Unique (United Kingdom) publishes a guide for families on ZMYND11-related syndromic intellectual disability (2020) and a guide on 10p deletions, and mentions a family group on a social network [17]. NETRE (European Network for Therapies in Rare Epilepsy) contributed to recruitment of the 2021 series [9]. The NORD (National Organization for Rare Disorders) directory lists a ZMYND11 Treatment Foundation, whose page was not accessible for this edition.

10. Illustrative Clinical Case

10.1. Japanese boy with p.(Arg600Trp) (Yoshimatsu and colleagues, 2026)

A Japanese boy, studied in the Japanese national programme for rare and undiagnosed diseases (IRUD), was born at 39 weeks and 1 day after fetal growth restriction identified during pregnancy, with a weight of 2220 g (−2.64 SD), length of 44.5 cm (−2.2 SD) and HC of 30 cm (−2.4 SD), that is, small for gestational age, and Apgar scores of 8 and 9 [10]. From birth he had hypotonia, respiratory difficulty and poor feeding, was admitted to the neonatal intensive care unit on the second day, required high-flow nasal cannula until day 22, achieved full oral feeding on day 19 and was discharged on day 35 weighing 2700 g [10].

In the neonatal period he had hypopigmented hair, low-set ears, hypertelorism, a broad nasal root, small nares with broad alae, a small mouth and prognathism with a narrow mandible, and right cryptorchidism; later, a right inguinal hernia, exotropia and hypermetropic astigmatism, corrected with glasses, appeared [10]. Brain MRI at 6 months showed no structural abnormalities. He achieved head control at 14 months and rolled over at 16; at 2 years he could not sit unaided, and the Kyoto Scale of Psychological Development gave a total developmental quotient of 25 (postural-motor 24, cognitive-adaptive 24, language-social 40), corresponding to moderate to severe global delay [10]. He had self-injurious behaviour (hitting his own face), which the authors attributed to the absence of expressive language.

The karyotype on day 11 was 46,XY, chromosome 4 FISH at 11 months was negative and the microarray at 13 months was normal. The IRUD programme exome, with a mean coverage of 60.88× and the whole of ZMYND11 covered by at least 20 reads, identified NM_001370100.5:c.1798C>T p.(Arg600Trp) in the heterozygous state, de novo by trio Sanger sequencing [10]. The variant, already recorded in ClinVar as pathogenic or likely pathogenic, was classified as pathogenic by the criteria PS1, PM2, PS2 and PP3 [10, 47]. At 2 years and 9 months he maintained the sitting position, without independent standing or walking [10].

The authors compared the case with the two published carriers of the same variant, Cobben's boy and the Austrian woman described at 34 years, and with 47 other cases from the literature, concluding that p.(Arg600Trp) is associated with microcephaly, birth small for gestational age, short stature, cryptorchidism and nipple abnormalities, more frequent than with loss-of-function variants [10, 25, 29]. Figure 1 of the article shows the growth chart and the craniofacial and thoracic features (https://pmc.ncbi.nlm.nih.gov/articles/PMC13000240/figure/Fig1/), and Figure 2 the diagram of the gene with the variant in exon 15 and the family chromatograms (https://pmc.ncbi.nlm.nih.gov/articles/PMC13000240/figure/Fig2/); the content of the figures was consulted through the article's legends [10].

The case illustrates three clinical points: a normal microarray does not exclude the disorder, the recurrent missense variant p.(Arg600Trp) defines a severe extreme of the spectrum with growth restriction, and management (neonatal care, optical correction, developmental support) was symptomatic, as throughout the disorder. The severity of this patient does not represent most carriers of loss-of-function variants, whose ID is mostly mild [10].

10.2. Girl with ABPE, ASD and ADHD and average IQ (Oates and colleagues, 2021)

By contrast, patient #7 of the 2021 series, the daughter of non-consanguineous European parents, had her first focal seizure at 4 years, 30 minutes after waking, with staring, stiffening and falling to the side for 30 minutes, and postictal aphasia in the second seizure; morning seizures followed with head drops, perioral jerks and arm extension, and a tonic-clonic seizure, probably with secondary generalisation, at 9 years [9]. The initial EEG showed spike-wave activation in sleep, which did not persist, and video-telemetry at 8 years and 3 months showed irregular multifocal spikes over the right hemisphere, activated in drowsiness and sleep, with normal background activity; MRI showed no structural cause [9]. She was treated with sodium valproate, to which levetiracetam was added and, after the spike-wave activation in sleep, ethosuximide [9].

Behaviour was difficult from 18 months. At 5 years she read well but had social communication difficulties; at 7 years and 7 months, the ADOS-2, ADI-R and Conners-III scales confirmed ASD and combined ADHD, and at 10 years the full-scale IQ on the WISC-V was 87, with extremely low adaptive functioning (0.2nd percentile on the ABAS-III) in the conceptual, social and practical domains, anxiety, tantrums and prolonged low mood; she had no dysmorphism and growth was average [9]. Her younger brother had autism and ADHD but was not tested. The case shows that the ZMYND11 disorder can present as focal epilepsy with ASD and ADHD without ID or dysmorphism, and that IQ does not measure the degree of disability when adaptive functioning is severely impaired [9]. Figure 1 of the article shows this patient's EEG; the full text consulted was the accepted manuscript, without access to the images.

11. Prognosis and Long-Term Follow-up

11.1. Prognosis

The disorder is neither progressive nor degenerative in the published sources, and most children acquire walking and some language: in the 2020 series, independent walking occurred at a median of 24 months and first words at a median of 2.5 years, and only 2/14 were not walking or speaking at 2.5 and 4 years [2, 4]. ID is mild in about three quarters of the patients with loss-of-function variants assessed (13/17), and functional prognosis depends as much on behavioural abnormalities as on cognition; behavioural abnormalities may constitute the main burden when ID is mild [4, 10]. Longitudinal cognitive data are limited to one patient, whose IQ fell from 63 at 9 years to 55 at 25, without the source allowing real decline to be distinguished from differences between assessment instruments [11]. Some adults are employed, and in one family four affected people in two generations had special educational needs, with two siblings later employed [4].

The prognosis of epilepsy ranges from spontaneous remission to drug resistance; episodes of spike-wave activation in sleep and seizure clusters are accompanied by transient regressions [9]. Seizures may reappear or worsen in adulthood, as in the patient with p.(Arg600Trp), whose seizures recurred from the age of 20 and became drug-resistant [29]. Movement disorders (dystonia, ataxia, tremor) may appear or worsen after childhood [29]. The recurrent missense variants p.(Arg600Trp) and p.(Ser421Asn) have a worse motor and cognitive prognosis, with absent or very late walking and absent expressive language [10, 20, 26].

Published adults are up to 47 years old in the 2021 series, and men aged 45 and 48 years and women aged 32 and 34 years have been described [6, 9, 24, 29, 51]. The sources provide no life expectancy data; in severe ID in general, life expectancy is reduced by the combination of health problems and by delayed treatment of acute illness [52].

11.2. Surveillance

There is no published surveillance protocol for the ZMYND11 disorder. Table 11.1 brings together the follow-up elements supported by the sources; intervals are defined only for the genetic review of ID and for drug monitoring, and in the other cases frequency is decided on clinical grounds.

Table 11.1. Surveillance

11.3. Surveillance in the deletion form

In limited 10p15.3 deletions, surveillance is the same. In deletions extending into 10p14 or forming part of an unbalanced rearrangement, the additional genes determine their own surveillance, such as that of parathyroid function, hearing and the kidneys when the deletion includes GATA3 [6, 14]. In HDR syndrome, deafness is the most constant manifestation and is usually early, moderate to severe and bilateral, and PTH may be low or merely inappropriately normal for the degree of hypocalcaemia, so assessment includes serum calcium with simultaneous PTH, audiological assessment and renal ultrasound and function; the sources consulted do not define repeat intervals [21]. In deletions reaching the DGS2 region, cardiac and cellular immunity assessment follows the same reasoning [22].

12. Summary and Final Considerations

ZMYND11-related disorder results, in most patients, from haploinsufficiency of a chromatin reader that recognises the H3.3K36me3 mark and modulates transcriptional elongation in highly expressed genes. The same loss of dosage explains the 10p15.3 microdeletion syndrome, of which ZMYND11 is the critical gene, and the two forms are brought together under the same Orphanet entity. The gene–disease relationship is definitive and haploinsufficiency is established, which allows null variants and deletions to be classified; missense variants require trio segregation, and two of them, p.(Arg600Trp) and p.(Ser421Asn), are associated with more severe phenotypes through a mechanism not yet elucidated.

The phenotype combines developmental delay predominantly affecting speech, mostly mild to moderate ID, behavioural abnormalities in about 85–88% (attention deficit, impulsivity, aggressiveness, autistic traits), hypotonia, early feeding difficulties and subtle dysmorphism with thick eyebrows, prominent eyelashes, a bulbous nasal tip and a wide mouth, present in about two thirds. Epilepsy, in 31–47% depending on the series, includes focal epilepsies of the atypical SeLECTS spectrum with a risk of spike-wave activation in sleep and generalised epilepsies, with a course ranging from spontaneous remission to drug resistance. The spectrum extends to people with normal IQ and ASD or ADHD, and in adulthood to dystonia, ataxia and tremor and to seizure recurrence.

The diagnosis is made by exome or genome, preferably as a trio, or by microarray in the deletion form; a normal microarray does not exclude the disorder, and small intragenic deletions require CNV calling with adequate resolution. Parental testing is part of diagnosis and counselling, because there is transmission from parents with a mild phenotype and from mosaic mothers, and the recurrence risk ranges from less than 1% to 50%.

There is no specific treatment. Management is multidisciplinary and symptomatic, with antiseizure medications chosen by seizure type, attention to the behavioural effects of levetiracetam and to the restrictions on valproate, intervention for behaviour and language, nutritional support and neurological surveillance that continues into adulthood. Therapeutic evidence is limited to isolated cases, and recent developments are preclinical. The real frequency is unknown: Orphanet records 40 cases or families and the 2026 review gathered 50 cases with a sequence variant, figures that measure published cases and not prevalence.

13. References

1. OMIM® Online Mendelian Inheritance in Man. #616083 INTELLECTUAL DEVELOPMENTAL DISORDER, AUTOSOMAL DOMINANT 30, WITH SPEECH DELAY AND BEHAVIORAL ABNORMALITIES; MRD30. Johns Hopkins University; last edited 08/09/2022. https://omim.org/entry/616083. Accessed 28 October 2026.
2. GeneGuidance. ZMYND11 — Zinc finger MYND-type containing 11 (gene record and clinical curation). Version of 2026-09-26.
3. ClinGen. ZMYND11 (HGNC:16966) — gene-disease validity (Definitive, 2024-03-28) and dosage sensitivity (HI 3, TS 0, 2018-03-28) [Internet]. Clinical Genome Resource; [cited 2026-10-28]. Available from: https://search.clinicalgenome.org/kb/genes/HGNC:16966
4. Yates TM, Drucker M, Barnicoat A, Low K, Gerkes EH, Fry AE, et al. ZMYND11-related syndromic intellectual disability: 16 patients delineating and expanding the phenotypic spectrum. Hum Mutat. 2020;41(5):1042-50. doi:10.1002/humu.24001. PMID: 32097528.
5. Orphanet. ZMYND11-related developmental delay-speech delay-seizures-behavioral abnormalities-craniofacial dysmorphism syndrome due to 10p15.3 microdeletion. ORPHA:687424 [Internet]. Paris: INSERM; 2026 [cited 2026-10-28]. Available from: https://www.orpha.net/en/disease/detail/687424
6. DeScipio C, Conlin L, Rosenfeld J, Tepperberg J, Pasion R, Patel A, et al. Subtelomeric deletion of chromosome 10p15.3: clinical findings and molecular cytogenetic characterization. Am J Med Genet A. 2012;158A(9):2152-61. doi:10.1002/ajmg.a.35574. PMID: 22847950; PMCID: PMC3429713.
7. Huynh MT, Tran CT, Joubert M, Bénéteau C. Intragenic deletion of the ZMYND11 gene in 10p15.3 is associated with developmental delay phenotype: a case report. Cytogenet Genome Res. 2021;161(8-9):445-8. doi:10.1159/000518689. PMID: 34818214.
8. Jing XY, Zhen L, Li DZ. Prenatal phenotype of 10p15.3 microdeletion syndrome: a series of five cases. Taiwan J Obstet Gynecol. 2026;65(4):808-10. doi:10.1016/j.tjog.2024.01.040. PMID: 42362275.
9. Oates S, Absoud M, Goyal S, Bayley S, Baulcomb J, Sims A, et al. ZMYND11 variants are a novel cause of centrotemporal and generalised epilepsies with neurodevelopmental disorder. Clin Genet. 2021;100(4):412-29. doi:10.1111/cge.14023. PMID: 34216016.
10. Yoshimatsu H, Kido J, Sawada T, Sugawara K, Misumi Y, Hayashi Y, et al. ZMYND11 p.Arg600Trp variant associated with a distinctive neurodevelopmental phenotype. Hum Genome Var. 2026;13(1). doi:10.1038/s41439-026-00339-1. PMID: 41820311; PMCID: PMC13000240.
11. OMIM® Online Mendelian Inheritance in Man. *608668 ZINC FINGER MYND DOMAIN-CONTAINING PROTEIN 11; ZMYND11. Johns Hopkins University; last edited 08/09/2022. https://omim.org/entry/608668. Accessed 28 October 2026.
12. Orphanet. ZMYND11-related developmental delay-speech delay-seizures-behavioral abnormalities-craniofacial dysmorphism syndrome. ORPHA:694304 [Internet]. Paris: INSERM; 2026 [cited 2026-10-28]. Available from: https://www.orpha.net/en/disease/detail/694304
13. Orphanet. ZMYND11-related developmental delay-speech delay-seizures-behavioral abnormalities-craniofacial dysmorphism syndrome due to a point mutation. ORPHA:694308 [Internet]. Paris: INSERM; 2026 [cited 2026-10-28]. Available from: https://www.orpha.net/en/disease/detail/694308
14. Minale EMP, Martone S, Criscuolo C, Marra R, Lasorsa VA, Ruggiero R, et al. A complex neurodevelopmental phenotype resembling a chromatinopathy with concurrent 7p duplication and 10p deletion involving ZMYND11: a case report and literature review. Mol Genet Genomic Med. 2026;14(4):e70164. doi:10.1002/mgg3.70164. PMID: 42003802; PMCID: PMC13093532.
15. Tumiene B, Čiuladaitė Ž, Preikšaitienė E, Mameniškienė R, Utkus A, Kučinskas V. Phenotype comparison confirms ZMYND11 as a critical gene for 10p15.3 microdeletion syndrome. J Appl Genet. 2017;58(4):467-74. doi:10.1007/s13353-017-0408-3. PMID: 28933030.
16. Manickam K, McClain MR, Demmer LA, Biswas S, Kearney HM, Malinowski J, et al. Exome and genome sequencing for pediatric patients with congenital anomalies or intellectual disability: an evidence-based clinical guideline of the American College of Medical Genetics and Genomics (ACMG). Genet Med. 2021;23(11):2029-37. doi:10.1038/s41436-021-01242-6. PMID: 34211152.
17. Unique – Rare Chromosome Disorder Support Group. ZMYND11-related syndromic intellectual disability [guide for families]. Version 1. Oxted: Unique; 2020. Available from: https://rarechromo.org/media/singlegeneinfo/Single%20Gene%20Disorder%20Guides/ZMYND11%20related%20syndromic%20intellectual%20disability%20QFN.pdf
18. European Reference Network ITHACA [Internet]. [cited 2026-10-28]. Available from: https://ern-ithaca.eu/
19. European Reference Network EpiCARE [Internet]. [cited 2026-10-28]. Available from: https://epi-care.eu/
20. Bodetko A, Chrzanowska J, Rydzanicz M, Borys-Iwanicka A, Karpinski P, Bladowska J, et al. Further delineation of clinical phenotype of ZMYND11 variants in patients with neurodevelopmental dysmorphic syndrome. Genes (Basel). 2024;15(2):256. doi:10.3390/genes15020256. PMID: 38397245; PMCID: PMC10888010.
21. OMIM® Online Mendelian Inheritance in Man. #146255 HYPOPARATHYROIDISM, SENSORINEURAL DEAFNESS, AND RENAL DYSPLASIA SYNDROME; HDRS. Johns Hopkins University; last edited 21/11/2023. https://omim.org/entry/146255. Accessed 28 October 2026.
22. OMIM® Online Mendelian Inheritance in Man. %601362 DIGEORGE SYNDROME/VELOCARDIOFACIAL SYNDROME COMPLEX 2. Johns Hopkins University; last edited 13/08/2009. https://omim.org/entry/601362. Accessed 28 October 2026.
23. Iossifov I, Ronemus M, Levy D, Wang Z, Hakker I, Rosenbaum J, et al. De novo gene disruptions in children on the autistic spectrum. Neuron. 2012;74(2):285-99. doi:10.1016/j.neuron.2012.04.009. PMID: 22542183; PMCID: PMC3619976.
24. Coe BP, Witherspoon K, Rosenfeld JA, van Bon BW, Vulto-van Silfhout AT, Bosco P, et al. Refining analyses of copy number variation identifies specific genes associated with developmental delay. Nat Genet. 2014;46(10):1063-71. doi:10.1038/ng.3092. PMID: 25217958; PMCID: PMC4177294.
25. Cobben JM, Weiss MM, van Dijk FS, De Reuver R, de Kruiff C, Pondaag W, et al. A de novo mutation in ZMYND11, a candidate gene for 10p15.3 deletion syndrome, is associated with syndromic intellectual disability. Eur J Med Genet. 2014;57(11-12):636-8. doi:10.1016/j.ejmg.2014.09.002. PMID: 25281490.
26. Moskowitz AM, Belnap N, Siniard AL, Szelinger S, Claasen AM, Richholt RF, et al. A de novo missense mutation in ZMYND11 is associated with global developmental delay, seizures, and hypotonia. Cold Spring Harb Mol Case Stud. 2016;2(5):a000851. doi:10.1101/mcs.a000851. PMID: 27626064; PMCID: PMC5002929.
27. Popp B, Ekici AB, Thiel CT, Hoyer J, Wiesener A, Kraus C, et al. Exome Pool-Seq in neurodevelopmental disorders. Eur J Hum Genet. 2017;25(12):1364-76. doi:10.1038/s41431-017-0022-1. PMID: 29158550; PMCID: PMC5865117.
28. Aoi H, Mizuguchi T, Ceroni JR, Kim VEH, Furquim I, Honjo RS, et al. Comprehensive genetic analysis of 57 families with clinically suspected Cornelia de Lange syndrome. J Hum Genet. 2019;64(10):967-78. doi:10.1038/s10038-019-0643-z. PMID: 31337854.
29. Indelicato E, Zech M, Amprosi M, Boesch S. Untangling neurodevelopmental disorders in the adulthood: a movement disorder is the clue. Orphanet J Rare Dis. 2022;17(1):55. doi:10.1186/s13023-022-02218-8. PMID: 35172867; PMCID: PMC8848801.
30. Thomsen M, Ott F, Loens S, Kilic-Berkmen G, Tan AH, Lim SY, et al. Genetic diversity and expanded phenotypes in dystonia: insights from large-scale exome sequencing. Ann Clin Transl Neurol. 2025;12(8):1648-59. doi:10.1002/acn3.70100. PMID: 40533913; PMCID: PMC12343315.
31. van der Veen S, Eggink H, Elting JWJ, Sival D, Verschuuren-Bemelmans CC, de Koning TJ, et al. The natural history of progressive myoclonus ataxia. Neurobiol Dis. 2024;199:106555. doi:10.1016/j.nbd.2024.106555. PMID: 38844245.
32. Dale RC, Mohammad S, Han VX, Nishida H, Goel H, Tangye SG, et al. Pathogenic variants in chromatin-related genes: linking immune dysregulation to neuroregression and acute neuropsychiatric disorders. Dev Med Child Neurol. 2025;67(8):1095-102. doi:10.1111/dmcn.16276. PMID: 39985218; PMCID: PMC12237219.
33. Chick SL, Holmans P, Cameron D, Grozeva D, Sims R, Williams J, et al. Whole-exome sequencing analysis identifies risk genes for schizophrenia. Nat Commun. 2025;16(1):7102. doi:10.1038/s41467-025-62429-y. PMID: 40753099; PMCID: PMC12318047.
34. Bai X, Zhang J, Wang S, Huang K, Pan Z, Liang A, et al. Novel intermolecular zinc fingers and redox-driven conformational changes dictate tumor suppressor ZMYND11's role in cooperative recognition of diverse targets. Nucleic Acids Res. 2026;54(3). doi:10.1093/nar/gkag048. PMID: 41591843; PMCID: PMC12839526.
35. Wen H, Li Y, Xi Y, Jiang S, Stratton S, Peng D, et al. ZMYND11 links histone H3.3K36me3 to transcription elongation and tumour suppression. Nature. 2014;508(7495):263-8. doi:10.1038/nature13045. PMID: 24590075; PMCID: PMC4142212.
36. Guo R, Zheng L, Park JW, Lv R, Chen H, Jiao F, et al. BS69/ZMYND11 reads and connects histone H3.3 lysine 36 trimethylation-decorated chromatin to regulated pre-mRNA processing. Mol Cell. 2014;56(2):298-310. doi:10.1016/j.molcel.2014.08.022. PMID: 25263594; PMCID: PMC4363072.
37. Li J, Galbo PM Jr, Gong W, Storey AJ, Tsai YH, Yu X, et al. ZMYND11-MBTD1 induces leukemogenesis through hijacking NuA4/TIP60 acetyltransferase complex and a PWWP-mediated chromatin association mechanism. Nat Commun. 2021;12(1):1045. doi:10.1038/s41467-021-21357-3. PMID: 33594072; PMCID: PMC7886901.
38. Yang X, Li L, Qu W, Cheng X, Zhang J, Sun Y, et al. Zmynd11 is essential for neurogenesis by coordinating H3K36me3 modification of Epha2 and PI3K signaling pathway. Cell Biosci. 2025;15(1):55. doi:10.1186/s13578-025-01392-z. PMID: 40281637; PMCID: PMC12032794.
39. Chang X, Li W, Matsui S, Huynh C, Erickson C, Guo F, et al. ZMYND11 functions in bimodal regulation of latent genes and brain-like splicing to safeguard corticogenesis. Nat Commun. 2025;16(1):9010. doi:10.1038/s41467-025-64063-0. PMID: 41068108; PMCID: PMC12511308.
40. Greben AW, Wu XS, Robb JE, Vogel IR, et al. ZMYND11 restrains KMT2A to enable a neuronal developmental program [preprint]. bioRxiv. 2025. doi:10.1101/2025.10.31.685692.
41. Chang X, McKinley M, Li W, Yang I, et al. Selective vulnerability of excitatory cortical neurons to ZMYND11 loss reveals cell-type-specific mechanisms [preprint]. bioRxiv. 2026. doi:10.64898/2026.06.16.732737.
42. Jones KL, Jones MC, del Campo M. Sotos syndrome (cerebral gigantism syndrome). In: Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 202–205.
43. Ocansey S, Cole TRP, Rahman N, Tatton-Brown K. Sotos syndrome. 2004 Dec 17 [updated 2025 Jun 5]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1479/
44. Ostrowski PJ, Tatton-Brown K. Tatton-Brown-Rahman syndrome. 2022 Jun 30. In: Adam MP, Feldman J, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK581652/
45. Pappas J, Rabin R. SETD2 neurodevelopmental disorders. 2021 Dec 30 [revised 2022 Sep 22]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK575927/
46. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-8. doi:10.1002/humu.23348.
47. National Center for Biotechnology Information. ClinVar: VCV000208648, VCV000225254, VCV001209429 [Internet]. Bethesda: NCBI; [cited 2026-10-28]. Available from: https://www.ncbi.nlm.nih.gov/clinvar/variation/208648/
48. Firth HV, Hurst JA. Autism and autism spectrum disorders. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 350–353.
49. Specchio N, Wirrell EC, Scheffer IE, Nabbout R, Riney K, Samia P, et al. International League Against Epilepsy classification and definition of epilepsy syndromes with onset in childhood: position paper by the ILAE Task Force on Nosology and Definitions. Epilepsia. 2022;63(6):1398-442. doi:10.1111/epi.17241. PMID: 35503717.
50. Juanes M, Touzon MS, Loos M, Caraballo R. Monozygotic twins with ZMYND11-related developmental and epileptic encephalopathy with spike-and-wave activation in sleep. Epileptic Disord. 2026;28(3):897-9. doi:10.1002/epd2.70191. PMID: 41652659.
51. Li J, Bachu V, Shahrvini B, Baniqued M, Haghighat S, Mallepally NR. Expanding the gastrointestinal phenotype of 10p15.3 microdeletion syndrome: refractory atypical gastroparesis in an adult. Cureus. 2025;17(1):e77555. doi:10.7759/cureus.77555. PMID: 39958070; PMCID: PMC11829799.
52. Firth HV, Hurst JA. Intellectual disability. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 188–193.
53. Firth HV, Hurst JA. Seizures with developmental delay/intellectual disability. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 308–311.
54. Firth HV, Hurst JA. Deletions and duplications (including microdeletions and microduplications). In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 660–665.
55. Dagli AI, Mathews J, Williams CA. Angelman syndrome. 1998 Sep 15 [updated 2021 Apr 22; revised 2025 May 1]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301323. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1144/
56. Deardorff MA, Raible SE, Krantz ID. Cornelia de Lange syndrome. 2005 Sep 16 [updated 2026 Jun 16]. 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/NBK1104/
57. Stevens CA. Rubinstein-Taybi syndrome. 2002 Aug 30 [updated 2023 Nov 9]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301699. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1526/
58. Schrier Vergano S, Santen G, Wieczorek D, Matsumoto N. Coffin-Siris syndrome. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 23556151. Available from: https://www.ncbi.nlm.nih.gov/books/NBK131811/
59. Shuman C, Kalish JM, Weksberg R. Beckwith-Wiedemann syndrome. 2000 Mar 3 [updated 2023 Jun 8; revised 2023 Sep 21]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. PMID: 20301568. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1394/
60. Strehlow V, Myers KA, Morgan AT, Scheffer IE, Lemke JR. GRIN2A-related disorders. 2016 Sep 29 [updated 2024 Jul 4]. In: Adam MP, Bick S, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK385627/
61. Firth HV, Hurst JA. Epilepsy in infants and children. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 410–415.
62. OMIM® Online Mendelian Inheritance in Man. #616708 DESANTO-SHINAWI SYNDROME; DESSH. Johns Hopkins University; last edited 07/08/2020. https://omim.org/entry/616708. Accessed 28 October 2026.
63. U.S. Food and Drug Administration. Levetiracetam tablets [prescribing information; SPL set ID 01aef7f7-6092-4e39-8451-dc02d8db9254, A-S Medication Solutions]. Revised 2025-11-21. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=01aef7f7-6092-4e39-8451-dc02d8db9254
64. U.S. Food and Drug Administration. Valproic acid oral solution [prescribing information; SPL set ID 1286ab49-fa34-488e-e063-6394a90ad582, NorthStar Rx]. Revised 2024-05-27. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=1286ab49-fa34-488e-e063-6394a90ad582
65. Parke-Davis (Pfizer). Zarontin (ethosuximide) capsules [prescribing information; SPL set ID 0e008f33-70a1-4bc6-b3a0-d45214418ab6]. Revised 2026-05-25. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=0e008f33-70a1-4bc6-b3a0-d45214418ab6
66. U.S. Food and Drug Administration. Risperidone tablets [prescribing information; SPL set ID 03d3b52d-3a7d-4595-84c9-332da6ca9171]. Revised 2026-06-25. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=03d3b52d-3a7d-4595-84c9-332da6ca9171
67. U.S. Food and Drug Administration. Methylphenidate hydrochloride extended-release capsules [prescribing information; SPL set ID 10aa0dc0-bd0e-4927-a19a-8722d78ad67d, Amneal]. Revised 2026-05-27. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=10aa0dc0-bd0e-4927-a19a-8722d78ad67d
68. European Medicines Agency. Risperdal — Article 30 referral: Annex I–IV (harmonised summary of product characteristics, sections 4.1 and 4.2); CHMP opinion of 24 Jul 2008 and European Commission decision of 7 Oct 2008. Available from: https://www.ema.europa.eu/en/medicines/human/referrals/risperdal
69. ClinicalTrials.gov. Search “ZMYND11” and “10p15.3” (API v2) [Internet]. Bethesda: NLM; [cited 2026-10-28]. Available from: https://clinicaltrials.gov/search?term=ZMYND11
70. Zheng R, Liu ZQ, Shen Y, Wang XT. Decoding ZMYND11-related syndromic intellectual disability: from epigenetic mysteries to therapeutic horizons. Neurosci Bull. 2026;42(8):1923-8. doi:10.1007/s12264-026-01664-9. PMID: 42334725; PMCID: PMC13424048.