SETD1A-Related Early-Onset Epilepsy: Comprehensive Research Report

1. Disease Information

Overview. SETD1A-related early-onset epilepsy is a rare, monogenic neurodevelopmental disorder caused by heterozygous (typically de novo) loss-of-function or missense variants in SETD1A (SET Domain Containing 1A), a gene encoding the catalytic subunit of the human COMPASS (Complex Proteins Associated with Set1) histone H3 lysine-4 (H3K4) methyltransferase complex. SETD1A haploinsufficiency produces a clinically heterogeneous, age-dependent spectrum: infants and young children most often present with global developmental delay, hypotonia, dysmorphic facial features, and early-onset epilepsy, while older children, adolescents, and adults are more likely to present with intellectual disability, autism-spectrum features, obsessive-compulsive symptoms, and — in a genome-wide-significant fraction of cases — schizophrenia and other psychotic disorders (PMID:26974950, PMID:32346159).

Two overlapping, allelic OMIM phenotype entries currently describe the SETD1A clinical spectrum: - EPEO2 — Epilepsy, Early-Onset, 2, With or Without Developmental Delay (OMIM #618832): an autosomal dominant disorder with generalized tonic-clonic seizure onset in the first days, months, or years of life, with highly variable severity — from normal psychomotor development and normal neuroimaging to developmental delay with brain-imaging abnormalities. Caused by heterozygous missense mutations, first delineated by Yu et al. (PMID:31197650). - NEDSID — Neurodevelopmental Disorder with Speech Impairment and Dysmorphic Facies (OMIM #619056): developmental delay with mild-to-moderate intellectual disability or learning difficulties, behavioral/psychiatric abnormalities, delayed speech/language, dysmorphic facies, distal limb anomalies, GI/feeding difficulties, and hypotonia. Delineated in 15 patients by Kummeling et al. (PMID:32346159).

The disorder is also frequently discussed under the umbrella name "SETD1A-associated neurodevelopmental disorder" (or "SETD1A syndrome") spanning both entries, since the epilepsy and dysmorphic-ID phenotypes overlap substantially and are caused by variants in the same gene, sometimes in the same family.

Key identifiers: - Gene (OMIM): 611052 – SET Domain-Containing Protein 1A; SETD1A - Phenotype (OMIM): #618832 (EPEO2); #619056 (NEDSID) - HGNC: hgnc:17284 (SETD1A) — NCBI Gene ID 9739 - Orphanet: Listed under gene page for SETD1A; disease-causing mutations associated with "Non-specific syndromic intellectual disability" (ORPHA:528084); a dedicated ORPHA number specific to the epilepsy phenotype was not identified in this search - MeSH: Epilepsy (D004827); Intellectual Disability (D008607) — no SETD1A-specific MeSH descriptor exists - MONDO: not directly resolved in this search; likely maps to a SETD1A-related neurodevelopmental disorder term cross-referencing OMIM 618832/619056 - ICD-10/11: G40.- (Epilepsy) with Q87.8 (other specified congenital malformation syndromes) as adjunct — no disease-specific code - Gene aliases: KMT2F, Set1, Set1A, EPEDD, NEDSID

Data provenance. Nearly all published knowledge derives from aggregated case series and case reports (individual clinical/genetic case reports plus pooled cohort analyses from exome-sequencing studies of schizophrenia/developmental-disorder cohorts — e.g., the DDD study, Finnish SISu/birth-cohort exomes), not from a disease registry or large natural-history study. This is an ultra-rare, only recently delineated (2016 onward) gene-disease association.

2. Etiology

Disease causal factor: Purely genetic/monogenic. Heterozygous de novo (occasionally inherited, autosomal dominant) loss-of-function or missense variants in SETD1A are sufficient to cause disease; there is no known environmental, infectious, or multifactorial trigger required.

Genetic risk factors: - SETD1A is exceptionally constrained against loss-of-function variation in the general population — in the founding study, only 2 LoF variants were found among 45,376 non-schizophrenia ExAC exomes, placing SETD1A "among the 3% most constrained genes in the human genome" (PMID:26974950). - Loss-of-function variant classes: nonsense, frameshift, canonical splice-site variants (most causing NEDSID/severe phenotype), and a small number of missense variants (largely reported in the EPEO2/epilepsy phenotype). - A recurrent splice-acceptor variant (c.4582-2delAG, NM_014712.3, exon 16) has arisen independently multiple times (≥7 occurrences across cohorts, including de novo events), consistent with a mutational hotspot (PMID:26974950). - No common susceptibility loci or polygenic risk modifiers specific to the SETD1A-driven phenotype have been reported; however, an intronic SNP (rs11150601) has been separately associated with female schizophrenia risk in the UK Biobank — this is a common-variant association distinct from the rare monogenic LoF mechanism and should not be conflated with it.

Environmental risk factors: None established. Case reports occasionally note nonspecific perinatal factors (e.g., one infantile-spasms case followed mild birth asphyxia and low birth weight; PMID for case: PMC10620521/Frontiers in Neurology 2023, DOI 10.3389/fneur.2023.1278035) but these appear coincidental rather than causal, given the driving de novo genetic lesion.

Protective factors: None identified in the literature for humans. In mice, pharmacological rescue (see Mechanism, below) demonstrates that circuit/cognitive deficits are reversible in adulthood, implying the deficit is a dynamic functional state rather than irreversible structural damage — a form of "protective" pharmacological intervention rather than a naturally occurring protective genetic/environmental factor.

Gene-environment interactions: Not established; no CTD/PheGenI data specific to SETD1A gene-environment interaction were identified.

3. Phenotypes

Phenotype frequency data are drawn primarily from two pooled cohorts: the original Nature Neuroscience schizophrenia/DDD meta-analysis (PMID:26974950, n=14 LoF carriers), and the Kummeling et al. 2021 NEDSID cohort (PMID:32346159, n=15). Because case ascertainment differs (psychiatric-genetics cohort vs. pediatric neurodevelopmental cohort), frequencies below should be read as convergent signal rather than a single denominator.

Neurological / Epilepsy Phenotypes

Psychiatric / Behavioral Phenotypes (age-dependent, emerging later)

• Schizophrenia with hallucinations/delusions (genome-wide-significant association, PMID:26974950); frequently requiring long-term psychiatric hospitalization (8/10 carriers in the founding cohort)
• Autism-spectrum features, aggressive behavior/autoaggression, ADHD-type behavior
• Obsessive-compulsive symptoms/personality disorder
• Bipolar I disorder — reported as a possible expansion of the clinical spectrum (PMID:38646907)
• Psychosis more broadly, reviewed systematically in Colijn et al. 2024 (Prog Neuropsychopharmacol Biol Psychiatry 129:110888), which concludes LoF SETD1A variants "may underlie the development of psychosis in a small percentage of individuals with schizophrenia"

Suggested HPO: HP:0000708 (Behavioral abnormality); HP:0000717 (Autism); HP:0000709 (Psychosis, historically used)/HP:0031466 (or nearest available psychosis term); HP:0000722 (Obsessive-compulsive behavior)

Craniofacial / Dysmorphic and Other Systemic Phenotypes

• Subtle facial dysmorphism (core, mild): high forehead, macrocephaly (in some), low nasal bridge, broad/short nose, thick lips, hypertelorism, upward-slanting palpebral fissures, thin face — HP:0001999 (Abnormal facial shape), HP:0000238 (Frontal bossing), HP:0000342 (Narrow nasal bridge), HP:0000316 (Hypertelorism)
• Distal limb anomalies (broad/short hands, single palmar crease, toe deformity) — HP:0001167 (Abnormality of finger), HP:0100258 (Preaxial polydactyly, N/A generically) / HP:0001760 (Abnormality of the foot)
• GI/feeding difficulties — HP:0011968 (Feeding difficulties)
• Capillary hemangioma — HP:0005605 (Capillary hemangioma)
• Rarely, congenital structural anomalies: one case reported congenital bronchomalacia, laryngeal stridor, tracheal stenosis, atrial septal defect, patent ductus arteriosus, and pulmonary hypertension without any seizures or developmental delay at 2-year follow-up (PMC10063285) — illustrating markedly variable expressivity even among truncating variants.

Quality of life impact: No disease-specific EQ-5D/SF-36 data exist. Impact is inferred from the underlying phenotypes: epilepsy (seizure-related risk/burden), intellectual disability (educational/functional impact), and psychiatric illness (frequently requiring long-term hospitalization) — collectively substantial but unquantified in validated QOL instruments.

4. Genetic/Molecular Information

Causal gene: SETD1A (HGNC:17284; NCBI Gene 9739; OMIM *611052), located at cytoband 16p11.2, chr16:30,957,294–30,984,664 (GRCh38), spanning 23 exons and encoding a 1,707-amino-acid, ~186 kDa protein. Note: this locus lies just telomeric to (and is distinct from) the classic recurrent 16p11.2 BP4-BP5 microdeletion/duplication CNV region associated with autism/schizophrenia risk; SETD1A disease is caused by intragenic point/small indel variants, not the recurrent CNV.

Pathogenic variant spectrum: - NEDSID cohort (Kummeling et al. 2021, PMID:32346159): 14 distinct de novo heterozygous variants in 15 unrelated patients — 5 nonsense, 6 frameshift, 2 splice-site, 1 missense; all predicted to disrupt or delete the C-terminal SET catalytic domain (truncating variants correlate with more severe phenotype). - EPEO2 cohort (Yu et al. 2019, PMID:31197650): Four missense variants — p.Arg913Cys (R913C, inherited in a 4-generation family), p.Gln269Arg (Q269R, de novo), p.Gly1369Arg (G1369R, de novo), p.Arg1392His (R1392H, de novo) — all previously implicated in schizophrenia/developmental-disorder cohorts, now shown to cause early-onset epilepsy. - Additional individual case variants: p.Ser356Phe (missense, de novo; isolated epilepsy phenotype, Frontiers 2026 case, DOI 10.3389/fnins.2026.1864983); p.Glu1002Glyfs20 (frameshift, de novo; infantile epileptic spasms, PMC10620521); p.Gly708Argfs117 (frameshift, de novo; NEDSID-type with seizures, PMC9300109); p.Leu699Ter (nonsense, de novo; congenital cardiac/airway phenotype without epilepsy, PMC10063285). - Recurrent hotspot: c.4582-2delAG (splice acceptor, exon 16) — found ≥7 times independently across schizophrenia/developmental-disorder cohorts (PMID:26974950).

Variant classification: Almost uniformly classified pathogenic/likely pathogenic per ACMG/AMP criteria (typically PVS1+PS2+PM2 for truncating de novo variants; PS2+PM2+PP3 for missense de novo variants). ClinVar entries exist per-variant (e.g., VCV000533549) but no gene-wide systematic ClinVar tabulation of counts by variant class was retrievable in this search.

Population frequency: SETD1A LoF variants are essentially absent from the general population (2/45,376 non-schizophrenia ExAC exomes; PMID:26974950); gnomAD constraint metrics were not directly retrievable in this search but the gene's extreme depletion of LoF variants in population data is well established and consistent with a highly LoF-intolerant gene (historically near-ceiling pLI).

Somatic vs. germline: All reported disease-causing variants are germline (constitutional), predominantly de novo.

Functional consequence: Loss of function / haploinsufficiency is the dominant mechanism for both the epilepsy (missense, partial LOF) and NEDSID (truncating, more complete LOF) ends of the spectrum — i.e., use functional_impact_category: LOSS_OF_FUNCTION (with PARTIAL_LOSS_OF_FUNCTION plausible for hypomorphic missense alleles) rather than gain-of-function or dominant-negative, though formal functional classification per-variant (e.g., dominant-negative testing) has not been systematically reported.

Modifier genes: None established.

Epigenetic information: Targeted episignature profiling (>2 million CpGs) in 6 SETD1A patients found no strong disease-specific DNA methylation episignature (only 7 significant differentially methylated positions, indistinguishable from controls) — a notable negative finding that contrasts with its paralog SETD2, which shows strong, subgroup-specific episignatures (PMID:37166351). This suggests SETD1A haploinsufficiency's downstream chromatin effects are not readily captured by peripheral-blood methylation arrays, despite robust effects on H3K4 methylation and, in mouse cortex, ribosomal-gene-associated DNA methylation (see Mechanism).

Chromosomal abnormalities: No recurrent CNV/translocation mechanism is described; disease arises from intragenic SNVs/small indels. One case report described an 852 kb CNV deletion encompassing SETD1A as a cause of the episignature-cohort phenotype (PMID:37166351).

5. Environmental Information

No environmental toxins, occupational exposures, dietary factors, or infectious triggers have been implicated in SETD1A-related disease causation or modification. This is consistent with a fully penetrant, single-gene dominant mechanism. Isolated perinatal complications noted in case reports (mild birth asphyxia, low birth weight) appear incidental rather than causally linked.

6. Mechanism / Pathophysiology

Causal chain (ordered, from molecular lesion to seizure phenotype)

1. A heterozygous de novo (or rarely inherited) loss-of-function or missense variant in SETD1A truncates or destabilizes the SET/n-SET catalytic domain, or (for missense alleles) partially impairs catalytic/complex-assembly function → leads to ~50% reduction in functional SETD1A protein/complex activity (haploinsufficiency; demonstrated directly as a ~50% drop in RNA and protein in Setd1a+/- mice, PMID:32937141/PMC9476630).
2. Reduced SETD1A-COMPASS activity → results in globally and locally altered histone H3 lysine-4 mono-, di-, and trimethylation (H3K4me1/2/3) at enhancers, gene bodies, and promoters genome-wide (this step is directly demonstrated in cell/mouse models; PMC8595121 review).
3. Altered H3K4 methylation at developmental and synaptic gene loci → branches into at least three parallel downstream pathway disruptions (demonstrated with varying degrees of directness across model systems):
4. 3a. Neurogenesis/neural progenitor branch: SETD1A cooperates with histone chaperone HIRA to stabilize β-catenin, hyperactivating canonical Wnt/β-catenin signaling (GO:0060070) → promotes neural stem/progenitor proliferation at the expense of differentiation, and SETD1A loss independently impairs cell-cycle progression (increased G1, reduced S phase) in ESCs/NSCs/iPSCs (PMC8595121) → contributes to altered neuronal migration timing (a missense variant, p.R913C, accelerated migration of embryonic cortical neurons to superficial layers; PMC8595121) → disrupts normal cortical lamination/circuit assembly (largely inferred from mouse/embryonic-electroporation data, not directly shown in human tissue).
5. 3b. Synaptic/dendritic branch: Dysregulated expression of synaptic genes (downregulated Homer1, PTPRO, ABI1; upregulated SLITRK4) in mouse medial prefrontal cortex → results in reduced dendritic spine density (especially mushroom spines), reduced excitatory drive onto L2/3 pyramidal neurons (fewer functional synapses and weaker per-synapse strength), altered NMDA-receptor subunit composition, and increased short-term synaptic depression (PMID:32937141) → produces working-memory deficits and aberrant sensory-cortex ensemble oscillations in adult mice (PMID:31606247). In human iPSC-neuron models, the analogous perturbation instead increases dendritic complexity and synaptic connectivity (glutamatergic-neuron-driven), indicating that the net synaptic phenotype (hyper- vs. hypo-connectivity) is model- and developmental-stage-dependent (PMID:35508131; PMID:40962831).
6. 3c. Mitochondrial/metabolic branch: Consistent, developmental-stage-independent downregulation of genes enriched for mitochondrial function/respiratory complex I assembly in mouse cortex, plus age-specific (peaking at E18) disruption of synaptosomal proteins including delayed Synaptotagmin-2 (Syt2) upregulation (PMID for Clifton et al. 2022: DOI 10.1093/hmg/ddac105, PMC9476630) → parallels reduced metabolic capacity and reduced neurite outgrowth/spontaneous activity in human iPSC-derived SETD1A+/− neurons, which is rescued by exogenous metabolic-intermediate supplementation (DOI 10.1038/s41537-022-00326-9, PMC9800576) → suggests bioenergetic insufficiency as a contributing, and separately targetable, downstream mechanism.
7. Convergent network-level dysfunction in glutamatergic circuits: In human iPSC-derived excitatory/inhibitory neuronal networks, SETD1A+/− haploinsufficiency results in transcriptional upregulation of adenylyl cyclase genes (ADCY2/3/8) and downregulation of phosphodiesterase genes (PDE12/PDE7A/PDE1A) → leads to a hyperactive cAMP/PKA pathway (elevated baseline cAMP, elevated phospho-CREB) → causes increased network burst rate, shortened inter-burst intervals, and elevated miniature EPSC frequency/amplitude, driven predominantly by glutamatergic (not GABAergic) neurons (PMID:35508131). This hyperexcitable, hypersynchronized network state is directly analogous to an ictogenic circuit phenotype and is pharmacologically reversible: acute PKA inhibition (H89, KT5720) or adenylyl-cyclase inhibition (SQ22536) normalizes network bursting, as does the H3K4-demethylase-pathway inhibitor ORY-1001 — the latter also rescuing cognitive/circuit deficits in adult Setd1a+/- mice (PMID:31606247), demonstrating the deficit is a dynamic, reversible physiological state rather than fixed structural damage.
8. A separate, parallel epigenetic-metabolic loop (mouse cortex): SETD1A loss-of-function causes widespread hypomethylation at ribosomal-protein and RNA-processing gene loci (e.g., Rps17, Rpl3; 356 hypomethylated vs. 105 hypermethylated sites; enriched for snoRNA-associated genes), a subset of which are independently enriched for common schizophrenia GWAS signal — proposed to impair local translation capacity at synapses, complementing rather than explaining the seizure mechanism directly (Schizophrenia Bulletin 2024/2025, DOI 10.1093/schbul/sbaf091).
9. Net clinical output: the combination of (a) altered progenitor proliferation/migration timing during corticogenesis, (b) synaptic/dendritic remodeling favoring excitatory network instability, and (c) a hyperactive cAMP/PKA-driven hypersynchronous network state converges on a cortical excitatory/inhibitory imbalance that manifests, depending on developmental timing and residual SETD1A dosage, as early-onset seizures (infancy), global developmental delay/intellectual disability (childhood), and/or later-onset psychiatric illness including schizophrenia (adolescence/adulthood) — the "single-gene, age-dependent, multi-phenotype" pattern characteristic of this locus.

Upstream vs. downstream: The variant itself (upstream) → chromatin/H3K4 methylation change (proximal) → transcriptional dysregulation of neurodevelopmental, synaptic, mitochondrial and cAMP-pathway genes (intermediate) → circuit-level hyperexcitability/hypersynchrony and progenitor/migration timing defects (downstream, most proximal to phenotype).

Cell types involved: glutamatergic (excitatory) cortical pyramidal neurons (primary driver of the network phenotype; CL:0000679 glutamatergic neuron / CL:0000598 pyramidal neuron), GABAergic interneurons (secondary/less penetrant contribution; CL:0000617 GABAergic neuron), neural stem/progenitor cells (CL:0000047 neuronal stem cell / CL:0002608 embryonic neuroepithelial progenitor cell), induced pluripotent stem cells (CL:0002248 iPSC) and embryonic stem cells (CL:0002322 ESC) used as model systems.

Suggested GO terms: GO:0051568 (histone H3-K4 methylation), GO:0035097 (histone methyltransferase complex), GO:0000993 (RNA polymerase II complex binding — WDR82 recruitment), GO:0060070 (canonical Wnt signaling pathway), GO:0050808 (synapse organization), GO:0007268 (chemical synaptic transmission), GO:0007612 (learning), GO:0007613 (memory), GO:0007190 (activation of adenylate cyclase activity), GO:0006119 (oxidative phosphorylation / mitochondrial respiratory chain complex I assembly).

Suggested UBERON/CL for anatomy: UBERON:0000956 (cerebral cortex), UBERON:0001950 (neocortex), UBERON:0002021/UBERON:0002616 (hippocampal formation, implicated via memory studies), CL:0000540 (neuron).

7. Anatomical Structures Affected

• Organ level: Central nervous system (brain) is the primary organ affected; secondary/associated involvement in some patients includes cardiac (atrial septal defect, patent ductus arteriosus), respiratory/airway (bronchomalacia, laryngeal stridor, tracheal stenosis), and craniofacial/skeletal structures (dysmorphic facies, distal limb anomalies) — UBERON:0000955 (brain), UBERON:0000948 (heart), UBERON:0001004 (respiratory system), UBERON:0002204 (skeletal system).
• Tissue/cell level: Cerebral cortex (frontal, prefrontal, visual cortex regions studied), hippocampus (memory-related H3K4me3 dynamics), affecting glutamatergic pyramidal neurons and GABAergic interneurons predominantly, plus neural progenitor populations during corticogenesis.
• Subcellular level: Nucleus/chromatin (site of SETD1A-COMPASS catalytic activity; GO:0000785 chromatin), dendritic spines/postsynaptic density (structural and functional synaptic changes), mitochondria (respiratory complex I and mitochondrial assembly factor disruption; GO:0005739 mitochondrion).
• Localization: Bilateral, diffuse cortical involvement rather than a focal lesion (consistent with normal structural MRI in many cases, e.g., the isolated-epilepsy missense case with normal brain MRI); no consistent lateralization reported, though one EEG case showed bilateral frontal/right temporal predominance.

8. Temporal Development

• Onset: Markedly bimodal/age-dependent. Seizure onset ranges from the first days of life to early childhood (infantile epileptic spasms as early as 6 months; focal seizures from 3 months; generalized tonic-clonic seizures up to age 3 in some reports); developmental delay and dysmorphic features are typically noted in infancy; psychiatric phenotypes (schizophrenia, OCD, bipolar disorder) emerge later, typically adolescence to adulthood.
• Progression: Highly variable disease course — some patients achieve seizure freedom with treatment and normal-range cognitive outcomes (DQ 108/IQ 105 in one missense case), while others have persistent developmental delay, ongoing seizures, and evolving psychiatric illness. No formal staging system exists.
• Course pattern: Can be episodic (seizures responding to treatment, e.g., ACTH-responsive infantile spasms with sustained remission at 10-month follow-up) or more static/stable (fixed developmental delay/dysmorphism without progressive decline); no evidence of a neurodegenerative or regressive course.
• Critical periods: Embryonic/early postnatal cortical development (progenitor proliferation, neuronal migration, and synaptogenesis, corresponding to mouse E14.5–P70 developmental window studied) appears to be the critical window during which SETD1A haploinsufficiency has its most durable structural impact, though functional/circuit-level deficits (working memory, network hyperexcitability) remain pharmacologically reversible into adulthood in mouse models — an important translational implication for postnatal therapeutic windows.

9. Inheritance and Population

• Epidemiology: No formal prevalence/incidence estimates exist. This is an ultra-rare, only recently delineated (2016 onward) gene-disease association known almost entirely from case reports and pooled sequencing-cohort ascertainment (schizophrenia exome studies, DDD study, Finnish population cohorts). Fewer than several dozen individuals with epilepsy/NEDSID-spectrum phenotypes have been reported in the literature to date.
• Inheritance pattern: Autosomal dominant. The great majority of reported variants are de novo; one four-generation family with an inherited missense variant (R913C) segregating with early-onset epilepsy has been reported (PMID:31197650), demonstrating that inherited transmission with variable expressivity is possible.
• Penetrance: Appears high for some core features (developmental delay/ID) but variable expressivity is substantial — the same or similar variant classes (even truncating LoF) can produce isolated congenital cardiac/airway anomalies without any neurodevelopmental phenotype (PMC10063285) or isolated epilepsy with normal cognition (case reports), versus severe NEDSID-spectrum disease in others.
• Expressivity: Markedly variable, as above; genotype-phenotype correlation is partial — truncating variants disrupting the SET domain trend toward the more severe NEDSID phenotype, while some missense variants are associated with a relatively "pure" epilepsy phenotype (EPEO2), but exceptions exist in both directions.
• Genetic anticipation: Not reported/applicable (not a repeat-expansion disorder).
• Germline mosaicism: Not specifically documented in the retrieved literature, though as a de novo dominant disorder it is a theoretical possibility relevant to recurrence-risk counseling (not empirically quantified here).
• Founder effects: None reported; the recurrent c.4582-2delAG splice variant reflects a mutational hotspot (independent recurrent mutation), not a founder haplotype.
• Consanguinity: Not implicated (autosomal dominant, de novo mechanism).
• Carrier frequency: Not applicable in the classical AR sense; population LoF variant frequency is extremely low (near-absent in ExAC/gnomAD-scale reference populations).
• Population demographics: Cases reported from diverse populations (UK/European schizophrenia and DDD cohorts, Finnish cohorts, Chinese case series). No specific ethnic enrichment reported. Sex distribution not systematically reported as skewed, though a mouse study specifically examined sex-dependent effects of Setd1a haploinsufficiency on development and adult behavior (PMC11324134), suggesting possible sex-modulated expressivity worth noting as an open question.

10. Diagnostics

• Laboratory tests: No disease-specific biochemical or enzymatic biomarker exists; standard epilepsy workup (metabolic screen) is typically normal/noncontributory (as in the reported cases with normal DQ/IQ and normal MRI).
• Genetic testing (primary diagnostic modality):
• Whole-exome sequencing (WES) or whole-genome sequencing (WGS) as trio analysis is the standard approach used in essentially every reported case (de novo variant confirmation), and is the recommended first-tier test given the gene's inclusion in developmental-and-epileptic-encephalopathy and intellectual-disability gene panels.
• Targeted epilepsy/ID gene panels including SETD1A are increasingly available (GTR-listed).
• Sanger sequencing used routinely for segregation/de novo confirmation in trios.
• Chromosomal microarray (CMA) may detect the rare CNV-deletion mechanism (e.g., the 852 kb deletion reported in the episignature cohort) but will miss the majority (SNV/indel) of pathogenic variants.
• Methylation/episignature testing is not currently diagnostically useful for SETD1A, in contrast to many other chromatinopathies, since no robust disease-specific episignature was identified (PMID:37166351) — an important negative finding for diagnostic-test selection.
• Imaging: Brain MRI is frequently normal; when abnormal, findings can include white matter dysplasia and ventricular dilation (as in one NEDSID case, PMC9300109). No pathognomonic imaging finding exists.
• Electrophysiology: EEG is central to phenotyping — findings range from nonspecific background slowing/sharp waves to hypsarrhythmia (in the infantile-spasms case) to focal/multifocal sharp-slow complexes; no SETD1A-specific EEG signature exists.
• Clinical criteria: No formal consensus diagnostic criteria exist for "SETD1A-related epilepsy" per se; diagnosis rests on (1) a compatible phenotype (early-onset seizures ± developmental delay ± dysmorphism) plus (2) molecular confirmation of a heterozygous pathogenic/likely pathogenic SETD1A variant (ACMG/AMP-classified).
• Differential diagnosis: Other monogenic developmental and epileptic encephalopathies (e.g., SCN1A/Dravet spectrum, STXBP1, SETD1B-related disorder, other chromatinopathies such as KMT2A/Kabuki-spectrum, SETD2-related Luscan-Lumish/Rabin-Pappas syndromes) and 16p11.2 microdeletion/duplication syndrome (given genomic proximity, though mechanistically distinct) should be considered and excluded by targeted or genome-wide sequencing.
• Screening: No population or newborn screening program exists; case-finding is exclusively through diagnostic genetic testing in symptomatic individuals.

11. Outcome/Prognosis

• Survival/mortality: No mortality data specific to SETD1A-related epilepsy were identified; the disorder is not reported to be associated with increased mortality in the available literature (in contrast to many severe developmental and epileptic encephalopathies).
• Morbidity/function: Morbidity is driven by the combination of seizure burden, intellectual disability/developmental delay (ranging from normal-range cognition to moderate ID), and — in a subset — subsequent psychiatric illness (schizophrenia, bipolar disorder, OCD) that in the founding cohort led to long-term psychiatric hospitalization in the majority of affected schizophrenia-cohort carriers.
• Complications: Recurrent/breakthrough seizures after initial treatment response (documented in one case), psychiatric comorbidity, and (in a minority) associated congenital structural anomalies (cardiac, airway).
• Recovery potential: Several reported cases achieved good seizure control (phenobarbital response, ACTH-induced remission of infantile spasms sustained at follow-up) and age-appropriate or near-normal developmental trajectories, indicating a generally favorable seizure-outcome potential for at least a subset of patients, particularly those with missense (vs. truncating) variants and isolated epilepsy phenotypes.
• Prognostic factors: Variant class appears to be the most consistent prognostic signal reported — truncating/SET-domain-disrupting variants trend toward more severe, multisystem NEDSID-spectrum phenotypes, while some missense variants associate with milder, more isolated epilepsy phenotypes — though exceptions exist and this correlation is not absolute.
• Prognostic biomarkers: None validated.

12. Treatment

Pharmacotherapy (symptomatic, seizure control): - Standard antiseizure medications have been used empirically, with reported efficacy in individual cases: phenobarbital (significant reduction in seizure frequency in one focal-seizure case), valproic acid (used in combination with ACTH for infantile spasms), levetiracetam (initial seizure control in a generalized tonic-clonic seizure case, later recurrence). No SETD1A-specific antiseizure drug trial or guideline exists; treatment is empirical, following standard pediatric epilepsy/infantile-spasms protocols. - ACTH (adrenocorticotropic hormone) — used per standard infantile-epileptic-spasms-syndrome protocol; achieved rapid (within 3 days) and durable (10-month) seizure freedom in the single reported SETD1A-IESS case, with prednisone substituted after 4 weeks. - Vitamin B6 (pyridoxine, CHEBI:16709) — used empirically alongside ACTH in the infantile spasms case (standard practice to exclude/treat pyridoxine-dependent epilepsy pending genetic confirmation). - NCIT term: NCIT:C15986 (Pharmacotherapy) as the general treatment_term, with therapeutic_agent bound to specific agents (e.g., valproic acid CHEBI:39867, levetiracetam CHEBI:6437).

Ketogenic diet: Not specifically reported as trialed or validated for SETD1A-related epilepsy in the retrieved literature; SETD1A is not among the genetic etiologies with an established ketogenic-diet response signature (unlike SCN1A/Dravet, TSC1/TSC2, or UBE3A/Angelman).

Gene/precision therapy: No gene therapy, ASO, or targeted molecular therapy currently exists or is in clinical trials for SETD1A-related disease. However, preclinical mouse work provides proof-of-concept for a mechanistically targeted approach: the LSD1/KDM1A demethylase inhibitor ORY-1001 rescued cognitive and synaptic-circuit deficits when administered to adult Setd1a+/- mice (PMID:31606247), and pharmacological cAMP/PKA-pathway inhibitors (adenylyl cyclase inhibitor SQ22536; PKA inhibitors H89, KT5720) normalized hyperactive network activity in human SETD1A+/− iPSC-neuron models (PMID:35508131) — both representing candidate repurposable small-molecule strategies, though neither has reached clinical testing.

Supportive/rehabilitative care: Developmental/early-intervention services, speech-language therapy (NCIT:C159273), occupational/physical therapy (NCIT:C15302, NCIT:C121351), and psychiatric management (for later-emerging schizophrenia/bipolar/OCD features) are used as clinically indicated, following standard neurodevelopmental-disorder supportive-care pathways rather than a disease-specific protocol.

Genetic counseling: Recommended given autosomal dominant inheritance with possible germline mosaicism and documented instances of familial transmission (NCIT:C15240, Genetic Counseling).

Experimental treatments: No SETD1A-specific registered clinical trials (ClinicalTrials.gov) were identified in this search.

Treatment outcomes / algorithm: No systematic treatment-response data or standardized treatment algorithm exists; management is individualized, symptom-driven, and extrapolated from general pediatric epilepsy and neurodevelopmental-disorder practice.

13. Prevention

• Primary prevention: Not applicable — the disorder arises from de novo or dominantly inherited germline variants with no known preventable environmental trigger.
• Secondary prevention / screening: No population or targeted screening program exists. Prenatal testing (if a familial variant is known, as in the reported 4-generation family) and preimplantation genetic diagnosis would be theoretically available through standard clinical genetics pathways for at-risk families, though this was not specifically documented in the retrieved literature.
• Tertiary prevention: Early diagnosis via genetic testing enables early initiation of developmental support services and prompt, protocol-driven seizure treatment (e.g., ACTH for infantile spasms), which may improve developmental outcomes given the demonstrated reversibility of circuit-level deficits in animal models.
• Genetic counseling: The principal prevention-adjacent intervention currently available, given autosomal dominant inheritance and the possibility of parental germline mosaicism affecting recurrence risk.

14. Other Species / Natural Disease

No naturally occurring SETD1A-associated disease has been reported in non-human species (companion animals, livestock, or wildlife) in the retrieved literature; no OMIA entries or veterinary case series were identified. This is consistent with the disorder being a very recently characterized human monogenic condition studied primarily through engineered (not spontaneous) animal and cellular models.

• Orthology: The mouse ortholog is Setd1a (MGI:2446244); the Drosophila melanogaster ortholog is Set1 (dSet1/kmt2f; FlyBase, studied in Kummeling et al. 2021, PMID:32346159). Both are engineered (knockout/knockdown), not natural disease models.
• Zoonotic potential / cross-species susceptibility: Not applicable.

15. Model Organisms

Overall model-system summary: No single model currently recapitulates the full human triad of early-onset seizures + developmental delay + later psychiatric illness. Mouse germline knockouts best capture the synaptic/cognitive and mitochondrial/epigenetic axes; human iPSC-neuron networks best capture the network-hyperexcitability axis most directly relevant to seizure genesis; Drosophila establishes a conserved postmitotic neuronal requirement. A model with directly recorded electrographic seizures (EEG-monitored in vivo) attributable to Setd1a haploinsufficiency was not identified in this search and represents a notable gap for future work — this is a HUMAN_MODEL_MISMATCH-type gap worth flagging explicitly in a dismech entry (mouse/iPSC evidence strongly supports a hyperexcitable-network mechanism, but translational confirmation via a genuine in vivo seizure/EEG phenotype in the mouse model is not yet reported).

Summary Table of Suggested Ontology Bindings

Key Evidence Citations (PMID, with supporting quotes)

• PMID:26974950 — Singh et al., Nature Neuroscience 2016. "Rare loss-of-function variants in SETD1A are associated with schizophrenia and developmental disorders." Founding statistical association (Fisher's combined P = 3.3×10⁻⁹); establishes gene constraint and initial developmental-disorder/epilepsy signal.
• PMID:31197650 — Yu et al., Neuroscience Bulletin 2019. "De Novo and Inherited SETD1A Variants in Early-onset Epilepsy." Defines the EPEO2 missense-variant phenotype (R913C/Q269R/G1369R/R1392H) and underlying OMIM #618832.
• PMID:32346159 — Kummeling et al., Molecular Psychiatry 2021. "Characterization of SETD1A haploinsufficiency in humans and Drosophila defines a novel neurodevelopmental syndrome." Defines NEDSID (OMIM #619056) in 15 patients with 14 distinct truncating/splice/missense variants; Drosophila memory phenotype.
• PMID:37166351 — Lee et al., Human Molecular Genetics 2023. Episignature study showing no strong SETD1A-specific methylation signature (contrast with SETD2).
• PMID:31606247 — Mukai et al., Neuron 2019. Reversible cognitive/synaptic deficits in Setd1a-deficient mice; ORY-1001 rescue.
• PMID:32937141 — Nagahama et al., Cell Reports 2020. Attenuated excitatory synaptic function and schizophrenia-related behaviors in Setd1a+/- mice.
• PMID:35508131 — Wang et al., Cell Reports 2022. cAMP/PKA-pathway hyperactivity and pharmacological network-activity rescue in human SETD1A+/− iPSC-neurons.
• PMID:40962831 — Su et al., Molecular Psychiatry 2025. Synaptic transmission/plasticity dysregulation in isogenic human iPSC-neurons with exon-16 SETD1A truncating variants.
• PMID:38646907 — Case report/meta-analysis expanding the SETD1A clinical spectrum to bipolar I disorder.
• Case reports: Frontiers 2026 (DOI 10.3389/fnins.2026.1864983, isolated epilepsy, p.S356F); PMC10620521/Frontiers in Neurology 2023 (DOI 10.3389/fneur.2023.1278035, first infantile-spasms case, p.Glu1002Glyfs20); PMC9300109 (NEDSID with seizures, p.Gly708Argfs117); PMC10063285 (congenital cardiac/airway phenotype without epilepsy, p.Leu699Ter).

Data gaps flagged for curation: No formal prevalence/incidence figures; no gene-wide ClinVar variant-count tabulation retrieved; no in vivo (EEG-confirmed) mouse seizure model; no SETD1A-specific clinical trial or ketogenic-diet response data; MONDO ID not directly confirmed in this search and should be verified against the MONDO release before entry creation.

Reference Validation

Checked with linkml-reference-validator 0.2.1.

Quotes not found in the cited source

Searched the abstract, any retrieved full text, and the title. A quote drawn from a part of the paper that was not retrieved will appear here too, so check before treating one as invented:

Every one of these was searched against an abstract alone, with no full text retrieved - marked abstract only below. Where full text can be fetched, re-running with it will settle them; where the source publishes only a summary to PubMed, as GeneReviews chapters do, it will not, and the quote has to be checked by hand against the chapter itself.

• PMID:26974950 (abstract only): "among the 3% most constrained genes in the human genome"
• closest text in source: "By analyzing the whole-exome sequences of 4,264 schizophrenia cases, 9,343 controls and 1,077 trios, we identified a genome-wide significant association between rare loss-of-function (LoF) variants in SETD1A and risk for schizophrenia (P = 3.3 × 10(-9))"

Term Validation

Checked with linkml-term-validator 0.4.5, through the ols: adapter.

Terms the report names something else

These identifiers resolve, so nothing about them looks wrong, and the ontology calls them something unrelated to what the report calls them. That usually means the identifier is not the one the sentence needs:

• HP:0000238 (1 mention) - the report calls it "Frontal bossing"; HP calls it Hydrocephalus
• HP:0005605 (1 mention) - the report calls it "Capillary hemangioma"; HP calls it Large cafe-au-lait macules with irregular margins
• GO:0006119 (1 mention) - the report calls it "oxidative phosphorylation / mitochondrial respiratory chain complex I assembly"; GO calls it oxidative phosphorylation
• UBERON:0002616 (2 mentions) - the report calls it "hippocampal formation, implicated via memory studies"; UBERON calls it regional part of brain

Unresolved terms

These identifiers do not exist in an ontology that resolved other terms from the same prefix, so they were most likely invented:

• HP:0000342 (1 mention), reported as "Narrow nasal bridge" - HP does not contain this term

Obsolete terms

These terms are real but deprecated. Citing one is not a fabrication; it does mean the report is naming something the ontology has retired:

• GO:0051568 (obsolete histone H3-K4 methylation) (2 mentions)

Terms whose name is worth a second look

The report's name for these is recognisably related to the term's own name without being one of them. A loose paraphrase reads the same way as a citation of the wrong sibling term - and so does a related synonym, which the ontology records precisely because it names something adjacent rather than the same thing - so these are listed rather than judged:

• HP:0002384 (1 mention) - the report calls it "Focal-onset seizure"; HP calls it Focal impaired awareness seizure, and lists "Focal dyscognitive seizure" among its other names
• HP:0000709 (1 mention) - the report calls it "Psychosis, historically used"; HP calls it Psychosis
• HP:0100258 (1 mention) - the report calls it "Preaxial polydactyly, N/A generically"; HP calls it Preaxial polydactyly
• GO:0051568 (2 mentions) - the report calls it "histone H3-K4 methylation"; GO calls it obsolete histone H3-K4 methylation, and lists "histone H3 K4 methylation" among its other names
• GO:0000993 (1 mention) - the report calls it "RNA polymerase II complex binding — WDR82 recruitment"; GO calls it RNA polymerase II complex binding
• UBERON:0002204 (1 mention) - the report calls it "skeletal system"; UBERON calls it musculoskeletal system

Terms named inconsistently

The report gives these identifiers more than one name of its own:

• MGI:2446244 - called "Setd1a", "Orthology:* The mouse ortholog is Setd1a"

Prefixes with no resolver

Terms carrying these prefixes were not checked either way, because no configured ontology covers them. An unrecognised prefix may name an ontology this run could not reach as easily as one that does not exist, so nothing here is evidence of fabrication: ORPHA, MGI.