Autosomal Dominant Epilepsy with Auditory Features (ADEAF) — Research Report

Prepared: 2026-08-05 · Target MONDO: MONDO:0010898 · Category: Mendelian (focal epilepsy syndrome)

A note before diving in: the evidence base here is unusually clean-edged for a rare disease. One syndrome, one very characteristic aura, three genes that all converge — from three quite different directions — on the same secreted-protein-holds-the-synapse-together story. Where a claim comes from a GeneReviews table rather than a PubMed abstract, I flag it, because those sentences will not survive just validate-references and need re-quoting from the primary paper.

1. Disease Information

Overview

ADEAF is a familial focal (lateral temporal lobe) epilepsy in which the defining ictal event is a sound that isn't there — a buzz, a ring, a voice, a song — sometimes accompanied or replaced by a sudden inability to understand speech while the person is otherwise fully alert. Brains are structurally normal, development is normal, and most people do well on standard antiseizure medication. Onset is typically adolescence/early adulthood.

GeneReviews defines it as a focal epilepsy syndrome characterized by auditory symptoms and/or receptive aphasia as "prominent ictal manifestations," with normal brain development (Michelucci R, Pasini E, Nobile C, Ottman R. Autosomal Dominant Epilepsy with Auditory Features. GeneReviews®, NCBI Bookshelf NBK1537, PMID:20301709).

The 2022 ILAE nosology retains ADEAF as a named syndrome under "epilepsy syndromes with onset at a variable age" (Riney K et al., Epilepsia 2022;63:1443–1474, doi:10.1111/epi.17240). Recent literature increasingly uses the etiology-agnostic term epilepsy with auditory features (EAF) to cover sporadic and non-Mendelian cases: the name shift happened "to acknowledge this entity also in a non-familial context/pattern" (Furia A et al., Front Neurol 2022;12:807939, PMID:35153984).

Key identifiers

MONDO parents: MONDO:0017704 familial partial epilepsy; MONDO:0800496 epilepsy with auditory features.

Synonyms (verbatim from the MONDO record)

ADEAF · ADLTE · ADPEAF · "autosomal dominant lateral temporal lobe epilepsy" · "autosomal dominant partial epilepsy with auditory features" · "partial epilepsy with auditory aura" · "partial epilepsy with auditory features" · "adolescent/adult onset autosomal dominant epilepsy with auditory features."

Level of evidence

Everything below is disease-level aggregated (family series, multi-center cohorts, registries, GeneReviews) plus mechanistic work in mice/cells. There is no EHR-derived individual-patient resource for ADEAF; the largest clinical series (Bisulli 2018, n=123) is a single-center tertiary referral cohort, which biases toward more severe/refractory cases.

2. Etiology

Causal factors

Monoallelic (heterozygous) pathogenic variants in one of three genes, all encoding secreted or secretion-dependent proteins rather than ion channels — which was the surprise in 2002:

"Most inherited forms of epilepsy result from mutations in ion channels. However, one form of epilepsy, autosomal dominant partial epilepsy with auditory features…" (Fukata Y et al., Science 2006;313:1792–5, PMID:16990550)

Verbatim (Dazzo E et al., Am J Hum Genet 2015;96:992–1000, PMID:26046367):

"Overall, RELN mutations occurred in 7/40 (17.5%) ADLTE-affected families." "ADLTE is genetically heterogeneous, and mutations in LGI1 account for fewer than 50% of affected families."

Critical caveat on yield. Those family-based percentages are not the diagnostic yield in an unselected clinic population. In the largest genotyped EAF cohort (Bisulli F et al., Seizure 2021;85:115–118, PMID:33453592; 112 unrelated probands, 29.5% familial):

"We identified a genetic diagnosis for 8% of our cohort, including pathogenic/likely pathogenic variants (4/8 novel) in LGI1 (2.7%, CI: 0.6-7.6); RELN (1.8%; CI: 0.2-6.3); SCN1A (2.7%; CI: 0.6-7.6) and DEPDC5 (0.9%; CI 0-4.9)." "This study shows that the contribution of each of the known genes to the overall disorder is limited and that the genetic background of EAF is still largely unknown."

Secondary / emerging genes in EAF pedigrees (not classical ADEAF genes; phenotypically heterogeneous families): DEPDC5 (hgnc:18423), SCN1A (hgnc:10585), CNTNAP2 (hgnc:13830) — Pippucci T et al., Neurol Genet 2015;1:e5, PMID:27066544. And newly, KCNQ2 (hgnc:6296): a family segregating c.2251T>G presented with "epilepsy with auditory features (EAFs), focal epilepsy, and generalized epilepsy, and none of them suffered from neonatal seizures," leading the authors to "unveil… the prospect of its inclusion in screening panels for EAFs" (Talarico M et al., Int J Mol Sci 2024;26:295, PMID:39796146).

Risk factors

• Genetic: carrying a heterozygous pathogenic LGI1/RELN/MICAL1 variant. Family history of epilepsy is the dominant risk descriptor. No validated common-variant susceptibility loci specific to ADEAF (no dedicated GWAS; the syndrome is too rare).
• Environmental: none established as causing the disease. Sound is a seizure trigger, not a cause — reflex seizures precipitated by sudden noises (telephone, doorbell) in 8–13% of affected individuals (GeneReviews Table 2); "seizures were precipitated by environmental noises in 8% of patients" in RELN families (Michelucci R et al., Epilepsy Behav 2017;68:103–107, PMID:28142128).
• Age/sex: no sex effect on penetrance — "Penetrance did not differ by gender" (Rosanoff MJ, Ottman R, Neurology 2008;71:567–71, PMID:18711109). Sex ratio in the largest EAF series was 58 M / 65 F (Bisulli 2018, PMID:29464704).

Protective factors

No genetic or environmental protective factors are documented. Not available for this disease. (Speculatively, residual protein secretion level is protective — see §4 dose-response — but that's a modifier of severity, not a population protective factor.)

Gene–environment interactions

The only well-described G×E is sound-triggered reflex seizure susceptibility in mutation carriers — an environmental stimulus (auditory input to a genetically hyperexcitable lateral temporal cortex) precipitating events. Sleep is a second permissive state: seizures "frequently occur during sleep" (Furia 2022, PMID:35153984). Beyond that: no data.

3. Phenotypes

Frequencies below are from GeneReviews Table 2 (ranges reflect different published series) unless a primary PMID is given. Note for curation: GeneReviews prose is not a PubMed abstract and will fail snippet validation — re-source each frequency from Ottman 2004 / Michelucci 2017 / Bisulli 2018 / Ho 2012 before writing evidence blocks.

Winawer's phenotype-defining paper is the cleanest primary quote for the aura spectrum (Winawer MR et al., Neurology 2000;54:2173–6, PMID:10851389):

"Auditory hallucinations were most common, but other sensory symptoms (visual, olfactory, vertiginous, and cephalic) were also reported. Autonomic, psychic, and motor symptoms were less common. The clinical semiology points to a lateral temporal seizure origin."

Auditory quality (Ottman R et al., Neurology 2004;62:1120–6, PMID:15079011):

"In families with mutations, the most common auditory symptom type was simple, unformed sounds (e.g., buzzing and ringing)."

Simple (humming/buzzing/ringing, tinnitus-like) vs complex (voices, music, specific songs) vs distortions (volume change) vs negative (sudden disappearance of ambient sound) — the simple/complex/distortion split matters prognostically (§11).

Phenotype characteristics

• Onset: adolescence/early adulthood, "age 10–30 years" typical, reported range 4–50 years (GeneReviews). Mean onset 20 years in RELN families (PMID:28142128).
• Severity: mild-to-moderate in most; a minority is refractory. Highly variable even within a family.
• Progression: episodic/non-progressive. No neurodegeneration, no intellectual decline (moderate/severe ID is an ILAE exclusion criterion).
• Frequency of events: FBTCS often only once or twice a year; auras can be much more frequent.
• Quality of life: no ADEAF-specific EQ-5D/SF-36/PROMIS study exists. Not available. Real-world impact is driven by driving restrictions, FBTCS unpredictability, and the socially isolating weirdness of auditory auras (often misread as psychiatric). Behavioral problems, depression with suicide attempts, and migraine have been reported in isolated pedigrees, but systematic study attributed depression to epilepsy/medication rather than shared genetic susceptibility (GeneReviews).

4. Genetic / Molecular Information

LGI1 (hgnc:6572, 10q23.33; protein O95970, "epitempin")

Discovery (Kalachikov S et al., Nat Genet 2002;30:335–41, PMID:11810107):

"Here we describe identification of the causative gene in autosomal-dominant partial epilepsy with auditory features (ADPEAF, MIM 600512), a rare form of idiopathic lateral temporal lobe epilepsy characterized by partial seizures with auditory disturbances." "…identifying presumptive mutations in one copy of the leucine-rich, glioma-inactivated 1 gene (LGI1) in each of five families with ADPEAF."

Prior linkage: 10q22-24 / 10q24 (Winawer MR et al., Epilepsia 2002;43:60–7, PMID:11879388 — "maximum multipoint LOD score of 2.93").

• Variant types: >40 reported pathogenic variants; roughly ⅓ truncating (nonsense/frameshift/splice → NMD), the rest missense. Sequence analysis detects ~95%; exon-level del/dup ~5% (GeneReviews).
• Domain architecture & clustering: N-terminal leucine-rich repeat (LRR) domain + C-terminal EPTP/epitempin seven-bladed β-propeller. Mutations cluster in the LRR (Ho YY, Ionita-Laza I, Ottman R, Neurology 2012;78:563–8, PMID:22323750): "ADPEAF-causing mutations clustered significantly in the LRR domain (exons 3-5) of LGI1 (p = 0.026)."
• Functional consequence: loss of function. Two flavors — secretion-defective (majority; misfolded protein retained and degraded by ER quality control) and secretion-positive but binding-defective (e.g. S473L, R474Q, E383A) which reach the extracellular space but fail to engage ADAM22/23.
• Constraint (gnomAD API, GRCh38): pLI ≈ 1.000, observed/expected LoF = 0.190 (11 observed vs 57.7 expected), LOEUF = 0.315, missense Z = 3.98. Textbook haploinsufficiency signature.
• Somatic vs germline: germline. (Biallelic somatic LGI1 loss is a glioma phenomenon — "loss of both copies of LGI1 promotes glial tumor progression," PMID:11810107 — but that is a separate, tumor-suppressor context and should not be conflated with ADEAF.)
• De novo: ~1% of ADEAF (GeneReviews); a de novo LGI1 variant was found in a Turkish LTLE-with-auditory-aura cohort (PMID:26773249).

Allelic spectrum extension (2025, important and new). Biallelic LGI1 variants cause a far more severe disease (Hirano Y et al., Brain 2025;148:3514–3522, PMID:40455867, doi:10.1093/brain/awaf202):

"Monoallelic pathogenic variants in LGI1 cause autosomal dominant epilepsy with auditory features with onset in childhood/adolescence." "Affected individuals presented DEE with neonatal/infantile-onset epilepsy (n = 6/6), global developmental delay/intellectual disability (n = 6/6) and infant/premature death (n = 5/6)." "Functional analyses revealed that all LGI1 variants result in reduced secretion and ADAM22-binding. Residual LGI1 function levels correlated with clinical severity, ranging from infantile lethality to intermediate phenotypes."

That paper establishes a clean dose–response allelic series: ~50% LGI1 (heterozygous ADEAF) → focal epilepsy, normal cognition; residual function ~6.7–40% (biallelic hypomorph) → DEE of graded severity; null → infantile lethality. A prior mouse study likewise found "approximately 50% of LGI1 and approximately 10% of ADAM22 protein levels are sufficient to prevent lethal epilepsy."

RELN (hgnc:9957, 7q22.1; protein P78509, Reelin)

From PMID:26046367 (verbatim abstract):

"We show that ADLTE-related mutations significantly decrease serum levels of Reelin, suggesting an inhibitory effect of mutations on protein secretion." "We also show that Reelin and LGI1 co-localize in a subset of rat brain neurons, supporting an involvement of both proteins in a common molecular pathway underlying ADLTE." "Homozygous RELN mutations are known to cause lissencephaly with cerebellar hypoplasia."

• Seven heterozygous missense variants across seven families; 3D modeling predicted structural effects on domain folding.
• Penetrance in RELN families: 60% (20/33 carriers affected) (GeneReviews).
• Constraint: pLI = 1, o/e LoF = 0.253 (108 obs / 428 exp), LOEUF = 0.296. Note the paradox worth flagging in curation: RELN is LoF-constrained, yet ADEAF alleles are missense secretion-impairing, while biallelic RELN LoF gives Norman-Roberts lissencephaly — a second allelic series with a dosage/mechanism split.

MICAL1 (hgnc:20619, 6q21; protein Q8TDZ2)

The odd one out — gain of function (Dazzo E et al., Ann Neurol 2018;83:483–493, PMID:29394500):

"We identified two ADLTE-causing variants in the MICAL-1 gene: a p.Gly150Ser substitution occurring in the enzymatically active monooxygenase (MO) domain and a p.Ala1065fs frameshift indel in the C-terminal domain, which inhibits the oxidoreductase activity of the MO domain." "In cell-based assays, both variants significantly increased MICAL-1 oxidoreductase activity and induced cell contraction in COS7 cells, which likely resulted from deregulation of F-actin dynamics." "This suggests that dysregulation of the actin cytoskeleton dynamics is a likely mechanism by which MICAL-1 pathogenic variants lead to ADLTE."

Constraint: pLI ≈ 0, LOEUF = 0.855 — i.e. MICAL1 is not LoF-constrained, which is exactly what you'd predict for a gain-of-function disease gene. Nice internal consistency check.

Variant classification, allele frequency, modifiers, epigenetics

• ACMG/AMP classification: ADEAF variants in ClinVar are typically P/LP for recurrent LGI1 alleles; many novel missense land as VUS pending functional data. The secretion/ADAM22-binding assays (HiBiT split-nanoluciferase, cell-surface binding) are the field's de facto PS3-grade functional evidence.
• Population frequency: pathogenic alleles are private/family-specific; absent or ultra-rare in gnomAD. No recurrent founder allele is established.
• Modifier genes: none validated. Reduced penetrance (54–85%) implies modifiers exist but they are uncharacterized — a genuine knowledge gap worth a KNOWLEDGE_GAP discussion in the KB entry.
• Epigenetics: no ADEAF-specific methylation/chromatin data. Not available.
• Chromosomal abnormalities: an intragenic CNTNAP2 deletion was found in one EAF family (PMID:27066544); LGI1 exon-level deletions account for ~5% of LGI1-positive cases. No recurrent cytogenetic syndrome.

5. Environmental Information

• Environmental factors: none causal. Auditory stimuli act as acute precipitants only (§2).
• Lifestyle: standard epilepsy precipitants (sleep deprivation, alcohol, medication non-adherence) apply by extrapolation; no ADEAF-specific study. Explicitly flag as extrapolated, not measured.
• Infectious agents: none. Not applicable.
• One genuinely relevant "acquired environment": anti-LGI1 autoimmune encephalitis targets the same protein. "Antibodies against epitempin disrupting the LGI1-ADAM22 complex cause a rare form of autoimmune encephalitis, characterized in some cases by peculiar faciobrachial dystonic seizures (FBDS)" (PMID:35153984). Same molecular node, different insult — a beautiful natural experiment, and a mandatory differential in adult-onset cases.

6. Mechanism / Pathophysiology

The core causal chain (LGI1 arm)

LGI1 loss of function → failure of the trans-synaptic LGI1–ADAM22/ADAM23 bridge → reduced AMPA-receptor-mediated transmission + loss of Kv1.1 at the axon initial segment and presynaptic terminals → intrinsic and synaptic hyperexcitability of lateral-temporal/hippocampal glutamatergic neurons → focal epileptiform discharge in auditory association cortex → auditory aura → propagation → focal impaired awareness / bilateral tonic-clonic seizure.

Step by step, with citations:

(1) Ligand–receptor assembly. ADAM22 is the receptor (PMID:16990550):

"ADAM22, a transmembrane protein that when mutated itself causes seizure, serves as a receptor for LGI1. LGI1 enhances AMPA receptor-mediated synaptic transmission in hippocampal slices. The mutated form of LGI1 fails to bind to ADAM22. ADAM22 is anchored to the postsynaptic density by cytoskeletal scaffolds containing stargazin."

(2) Trans-synaptic complex. (Fukata Y et al., PNAS 2010;107:3799–804, PMID:20133599):

"Extracellularly secreted LGI1 links two epilepsy-related receptors, ADAM22 and ADAM23, in the brain and organizes a transsynaptic protein complex that includes presynaptic potassium channels and postsynaptic AMPA receptor scaffolds. A lack of LGI1 disrupts this synaptic protein connection and selectively reduces AMPA receptor-mediated synaptic transmission in the hippocampus." "Thus, LGI1 may serve as a major determinant of brain excitation."

(3) Structure. (Yamagata A, Fukai S, Cell Mol Life Sci 2019, PMID:31432233 — review):

"LGI1 consists of the N-terminal LRR domain and the C-terminal epitempin-repeat (EPTP; also known as EAR) domain." "The crystal structure of the full-length LGI1 in complex with ADAM22 exhibits a 2:2 heterotetramer in a dimer-of-dimer assembly." "transsynaptic linkage through the tripartite complex of ADAM22–(LGI1)2–ADAM23 in synapses" "Through this synaptic protein network, LGI1 modulates AMPA receptor-mediated synaptic transmission." "LGI1 is enriched at the axon initial segment and colocalized with ADAM22/23 and the voltage-gated potassium (Kv1) channels."

(4) Two mutational routes into the same failure:

"Among them, 19 mutations result in secretion-defective proteins presumably due to the failure of protein folding." / "The E383A mutation disrupts the Ca2+ coordination inside of the β-propeller structure." / "The S473L mutation substantially reduces the binding to ADAM22." / "The R474Q mutation disables the assembly of the tripartite complex of ADAM22, ADAM23, and LGI1." (PMID:31432233)

(5) Kv1.1 and intrinsic excitability — and it's reversible (Extrémet J et al., J Neurosci 2023;43:8596–8606, PMID:37863654):

"We previously showed that LGI1 deficiency in a mouse model (i.e., knock-out for LGI1 or KO-Lgi1) decreased Kv1.1 channel density at the axon initial segment (AIS) and at presynaptic terminals, thus enhancing both intrinsic excitability and glutamate release." "…the selective expression of LGI1 in KO-Lgi1 neurons from mice of both sexes, using single-cell electroporation, reduces intrinsic excitability and restores both the Kv1.1-mediated D-type current and Kv1.1 channels at the AIS."

(6) Developmental arm — synaptic pruning (Zhou YD et al., Nat Med 2009;15:1208–14, PMID:19701204):

"We discovered that the normal postnatal maturation of presynaptic and postsynaptic functions was arrested by the 835delC mutant LGI1, and contrastingly, was magnified by excess wild-type LGI1. Concurrently, mutant LGI1 inhibited dendritic pruning and increased the spine density to markedly increase excitatory synaptic transmission. Inhibitory transmission, by contrast, was unaffected."

(7) Cell-type specificity — it's the excitatory neurons (Boillot M et al., Brain 2014;137:2984–96, PMID:25234641):

"Emx1-Lgi1cKO mice displayed early-onset and lethal seizures, whereas CaMKIIα-Lgi1cKO mice presented late-onset occasional seizures associated with variable reduced lifespan. In contrast, neither spontaneous seizures nor increased seizure susceptibility to convulsant were observed when Lgi1 was deleted in parvalbumin interneurons." "We suggest that LGI1 secreted from excitatory neurons, but not parvalbumin inhibitory neurons, makes a major contribution to the pathogenesis of LGI1-related epilepsies. Our data further indicate that LGI1 is required from embryogenesis to adulthood to achieve proper circuit functioning."

(8) Anatomical origin of discharge (PMID:40455867): "we observed epileptic discharges from the isolated whole hippocampus of Lgi1-/- knockout mice, experimentally modelling the hippocampal origin of LGI1-related epilepsy." Note the tension worth curating explicitly: the human syndrome is lateral-temporal/neocortical by semiology, while the mouse models discharge from hippocampus — a legitimate HUMAN_MODEL_MISMATCH candidate.

The RELN arm

Reelin is a large secreted glycoprotein signaling through VLDLR/ApoER2–DAB1; mutant alleles reduce serum reelin, i.e. loss of secreted ligand — the same category of failure as LGI1, in a partly overlapping cell population ("Reelin and LGI1 co-localize in a subset of rat brain neurons," PMID:26046367). Reelin has "important functions in both the developing and adult brain," so the plausible chain is impaired neuronal positioning/plasticity in temporal cortex → altered excitability. The precise adult-brain mechanism is not resolved — this is an honest knowledge gap, not something to over-narrate.

The MICAL1 arm

MICAL-1 is an actin-disassembling monooxygenase (it oxidizes methionine residues on F-actin). Gain of oxidoreductase activity → excess actin filament disassembly → deranged cytoskeletal dynamics in developing/adult neurons → circuit-level hyperexcitability (PMID:29394500). This is the least mechanistically nailed-down of the three arms, and MICAL1's low constraint plus only two families means it should be curated with appropriate epistemic hedging (see also Nobile & Dazzo, Genes 2022;13:715, PMC9141472).

Suggested ontology terms for pathophysiology nodes (all verified via local OAK)

Biological processes (GO): - GO:0098990 AMPA selective glutamate receptor signaling pathway - GO:0097113 AMPA glutamate receptor clustering - GO:0050804 modulation of chemical synaptic transmission - GO:0007268 chemical synaptic transmission - GO:0060291 long-term synaptic potentiation - GO:1905805 excitatory synapse pruning · GO:0098883 synapse pruning - GO:0005249 voltage-gated potassium channel activity (MF) - GO:0038026 reelin-mediated signaling pathway - GO:0030042 actin filament depolymerization (MICAL1 arm; modifier: INCREASED)

Cellular components: GO:0043194 axon initial segment; GO:0032281 AMPA glutamate receptor complex; GO:0110157 reelin complex.

Cell types (CL): CL:0000679 glutamatergic neuron; CL:0000598 pyramidal neuron; CL:1001571 hippocampal pyramidal neuron; CL:0000617 GABAergic neuron (for the negative result — PV interneuron deletion does not cause seizures); CL:0000127 astrocyte (LGI1 is also secreted by astrocytes).

Suggested module conformance: epilepsy_excitation_inhibition_imbalance#Excitation-Inhibition Imbalance — ADEAF is close to a textbook conformer (channel/synaptic dysfunction → E/I imbalance → hyperexcitability/hypersynchrony → seizures). Substitute the disorder-specific trigger (LGI1–ADAM22/23 trans-synaptic complex failure) at the top node.

Molecular profiling / advanced technologies

Transcriptomics, proteomics, metabolomics, lipidomics, single-cell, spatial, and CRISPR-screen data specific to ADEAF: not available. Mouse Lgi1 expression is "predominantly neuronal and is consistent with the anatomic regions involved in temporal lobe epilepsy" (PMID:11810107). Human protein-level reference: UniProt O95970 (LGI1), Q9P0K1 (ADAM22), P78509 (RELN), Q8TDZ2 (MICAL1); complex structures in the PDB from the Yamagata/Fukai work.

7. Anatomical Structures Affected

• Primary organ / system: brain; nervous system. UBERON:0000955 brain, UBERON:0000956 cerebral cortex.
• Lobe/region: lateral temporal lobe neocortex — "Seizures originate in the lateral temporal neocortex, distinguishing EAF from mesial temporal lobe epilepsy" (PMID:35153984). UBERON:0001871 temporal lobe.
• Auditory cortex: UBERON:0001393 auditory cortex; primary auditory cortex UBERON:0034751 (Heschl's gyrus ≈ UBERON:0002773 anterior transverse temporal gyrus); auditory association cortex ≈ UBERON:0034752 secondary auditory cortex; UBERON:0002769 superior temporal gyrus.
• Language cortex: posterior superior temporal (Wernicke) region for ictal receptive aphasia — no clean UBERON term; use UBERON:0002769 plus HP:0033848.
• Hippocampal formation (UBERON:0002421): the discharge origin in mouse models and in the biallelic human DEE phenotype; not the primary site in classic human ADEAF.
• Cell/tissue level: glutamatergic pyramidal neurons of neocortex and hippocampus; excitatory synapses; PV interneurons spared as a source of LGI1.
• Subcellular: axon initial segment (GO:0043194), presynaptic terminal, postsynaptic density, endoplasmic reticulum (site of misfolded-LGI1 retention/degradation — GO:0005783, verify before use).
• Lateralization: notable left predominance of EEG abnormalities, "with marked left predominance" in RELN families (PMID:28142128); GeneReviews notes left predominance in some series and a DTI finding of "increased fractional anisotropy in the left temporal lobe" in LGI1-positive families. Seizures themselves may be bilateral-independent across a pedigree.

8. Temporal Development

• Onset: adolescence/early adulthood; typical 10–30 years; range 4–50 (GeneReviews). RELN families mean 20 years (PMID:28142128). Onset <10 years is a bad-prognosis marker (PMID:29464704).
• Onset pattern: episodic from the start — first event is usually an aura or an FBTCS; no prodrome, no insidious decline.
• Stages: none formally defined. Practically: (i) isolated auras, often unrecognized for years; (ii) auras + intermittent FBTCS (usual presentation to care); (iii) either long-term remission or, in a minority, drug-resistant focal epilepsy considered for surgery.
• Course: non-progressive, chronic, lifelong susceptibility. FBTCS "occurred only once or twice per year" in one 34-person series (GeneReviews).
• Remission: both treatment-induced and spontaneous remission occur; withdrawal relapse is a real hazard — "Drug withdrawal often leads to clinical relapses, not always responding to reinitiation of treatment" (PMID:35153984). Cumulative terminal-remission rates 26.6% / 35.7% / 51.6% at 10 / 20 / 30 years (PMID:29464704).
• Critical periods: the mouse data argue LGI1 "is required from embryogenesis to adulthood" (PMID:25234641), and that embryonic-onset loss is catastrophic while late-postnatal loss is mild — relevant to any future gene-replacement timing question.

9. Inheritance and Population

Epidemiology

• Prevalence: unknown. GeneReviews: "The prevalence of ADEAF is unknown but likely very low." Orphanet lists it among rare disorders (<1/2,000) without a numeric class.
• Among Mendelian focal epilepsies it's not marginal: "9/48 (19%) of families…met criteria for ADEAF" (GeneReviews); ~3% of people with epilepsy have a significant family history, of whom a fraction meets ADEAF criteria.
• Incidence: not estimable — "a precise estimate of the incidence of EAF is currently not available" (PMID:35153984), largely because isolated auditory auras go unreported.
• Suggested dismech encoding: measure_type: POINT_PREVALENCE, prevalence_class: UNKNOWN (or ULTRA_RARE with a note), no rate_per_100000. Do not invent a number.

Genetic parameters

• Inheritance: autosomal dominant with reduced penetrance (HP:0000006 Autosomal dominant inheritance; HP:0003829 Incomplete penetrance).
• Penetrance: 54% across eight LGI1 families (PMID:15079011); best pooled estimate 67% (95% CI 55–77%) across all 24 published LGI1 families (PMID:18711109) —

  "Overall penetrance was 67% (95% CI 55-77%), and did not vary according to mutation type or location within the gene." "Our results suggest that about two-thirds of individuals who inherit a mutation in LGI1 will develop epilepsy. This probably overestimates the true penetrance in the population because it is based on data from families containing multiple affected individuals." RELN: 60% (20/33 carriers). GeneReviews quotes an overall 54–85% range for counseling.

• Expressivity: variable — auditory aura vs aphasic seizures vs FBTCS-only within the same pedigree; some LGI1 families even contain individuals with idiopathic generalized epilepsies (PMID:15079011).
• Anticipation: not a repeat-expansion disorder; apparent generational increase in penetrance in the Rosanoff analysis was attributed to ascertainment ("probably because of limited information about early generations").
• Germline mosaicism: not reported. De novo: ~1%.
• Founder effects / consanguinity: none for classic AD ADEAF. Consanguinity is relevant only for the new biallelic LGI1 DEE phenotype — all four families in PMID:40455867 were consanguineous.
• Carrier frequency: not applicable (dominant); pathogenic alleles are essentially absent from gnomAD.

Demographics

• Populations: described worldwide; the deepest family series are Italian (Nobile/Michelucci, 40 families) and North American (Ottman/Columbia). Reports from Turkey, Korea, China, Brazil. No population enrichment established.
• Sex ratio: ~1:1 (58 M / 65 F, PMID:29464704); penetrance sex-independent.
• Geographic variant distribution: private family variants; no geographic clustering.

10. Diagnostics

Clinical criteria (ILAE 2022, via GeneReviews summary)

Mandatory: focal sensory auditory seizures and/or focal cognitive seizures with receptive aphasia; normal brain MRI. Exclusionary/alerts: generalized-onset seizures or other focal-onset seizure types; moderate/severe intellectual disability; generalized epileptiform discharges; focal abnormalities on neurologic exam.

Electrophysiology (the workhorse)

EEG (LOINC — routine EEG 24708-6, verify): interictal EEG normal in a substantial minority; focal temporal sharp waves/spikes in 57–80%, often left-predominant. Sleep-deprived and prolonged video-EEG increase yield. Prognostically loaded: unremarkable EEG predicts remission (HR 3.5, PMID:29464704).

Imaging

MRI brain (3T, epilepsy protocol) — expected normal; its main job is excluding hippocampal sclerosis, focal cortical dysplasia (think DEPDC5), and tumor. Research-level findings: left temporal FA increase on DTI; one Brazilian family with a left lateral temporal malformation that did not fully co-segregate (GeneReviews). A completed NIH imaging study exists: NCT00072813 — "MRI in Autosomal Dominant Partial Epilepsy With Auditory Features" (status: Completed).

Laboratory / biomarkers

• No routine biochemical marker. Serum Reelin is reduced by ADLTE RELN mutations (PMID:26046367) — a research biomarker, not a validated clinical assay.
• Anti-LGI1 antibodies (serum + CSF) are the essential rule-out in adult-onset cases — a negative-defining test rather than a positive one.
• Biopsy/histopathology: not indicated; no characteristic pathology.

Genetic testing (the actual diagnostic)

Recommended approach: multigene epilepsy panel including at minimum LGI1, RELN, MICAL1, plus DEPDC5, SCN1A, CNTNAP2, and (emerging) KCNQ2; escalate to exome/genome if the panel is negative — which it will be most of the time (92% of unselected EAF probands, PMID:33453592). Single-gene LGI1 testing is reasonable only in a classic large AD pedigree. Include exon-level del/dup analysis (~5% of LGI1 findings). CMA, karyotype, FISH, mtDNA, and repeat-expansion testing are not indicated unless the phenotype is atypical.

Differential diagnosis (with distinguishing features)

Screening

No population or newborn screening (adult-onset, non-treatable-by-early-detection, moderate penetrance). Cascade predictive testing of at-risk relatives is available once the familial variant is known, but must be counseled around the 54–85% penetrance and the absence of any preventive intervention. Prenatal and preimplantation genetic testing are technically available (GeneReviews) — and are exactly the kind of decision that belongs with a genetic counselor, not a panel default.

11. Outcome / Prognosis

• Survival/mortality: no excess mortality documented for classic heterozygous ADEAF beyond the general epilepsy SUDEP risk; not quantified for this syndrome. (Contrast: biallelic LGI1 DEE, where 5/6 died between 9 months and 24 years — PMID:40455867.)
• Seizure outcome — the single most important recent number (Bisulli F et al., Epilepsia 2018;59:834–843, PMID:29464704, n=123, median follow-up 11 years):

  "At last assessment, 42 patients had achieved TR (34.1%)." "The cumulative rates of TR were 26.6%, 35.7%, and 51.6% at 10, 20, and 30 years from inclusion." "Our data show a wide prognostic spectrum of EAF, ranging from mild forms with spontaneous remission, to severely refractory epilepsy addressed to surgery. The outcome, less favorable than expected from previous studies, appears to be primarily a function of 3 prognostic negative risk factors: age at onset < 10 years, auditory aura characterized by complex auditory hallucinations, and focal epileptiform abnormalities on scalp EEG."

Positive predictors on multivariate analysis: age at onset >10 years (HR 3.2, p=.028); distortion-only auras vs simple/complex hallucinations (HR 2.9, p=.041); unremarkable EEG (HR 3.5, p=.041).

This directly contradicts the older "benign syndrome" framing ("The clinical course of ADEAF is usually benign," GeneReviews). Curate both, with the tension made explicit — the discrepancy is partly referral-center ascertainment (Bologna tertiary cohort) vs family-study ascertainment. Worth a DISCREPANCY/KNOWLEDGE_GAP discussion node. - Morbidity/disability: cognition and neurologic exam are normal by definition. Disability is seizure-driven (driving, employment, injury from FBTCS). - Quality-of-life instruments: no ADEAF-specific QOLIE-31/EQ-5D data. Not available. - Complications: status epilepticus is rare; aphasic status epilepticus (HP:0032849) is conceptually possible but not established as characteristic. - Prognostic biomarkers: none molecular. Genotype does not predict outcome — "no significant clinical differences were observed between families with an LGI1 pathogenic variant and families without an identified pathogenic variant," and RELN vs LGI1 families are clinically "indistinguishable" (PMID:28142128). The only genotype–phenotype signal is domain-level: "Auditory symptoms were less frequent in individuals with truncation mutations in the EPTP domain than in those with other mutation type/domain combinations (58% vs 80%, p = 0.018)" (PMID:22323750).

12. Treatment

There are no ADEAF-specific randomized trials. Everything below is standard focal-epilepsy practice applied to this syndrome, plus small-series experience. Say so plainly in the KB entry.

Pharmacotherapy

• First line: sodium-channel-blocking ASMs, carbamazepine as the prototype, usually as monotherapy and often at low dose. GeneReviews: "Seizure control is usually readily achieved with standard anti-seizure medications (ASM)" and "Traditionally sodium channel blockers such as carbamazepine have been more frequently used with clear benefit," while noting no formal trials. Furia 2022: "EAF is considered a syndrome with a good response to anti-seizure medications used for focal epilepsy, such as carbamazepine in monotherapy."
• Reasonable alternatives (focal-epilepsy standard): oxcarbazepine, lamotrigine, levetiracetam, lacosamide.
• Avoid the classic misstep: "When misdiagnosed as idiopathic generalized epilepsy, EAF might be treated with drugs that are not optimal (i.e., phenobarbital or valproate), leading to poor response" (PMID:35153984).
• Withdrawal caution: relapse after withdrawal may not re-respond (PMID:35153984) — argues for conservative, counseled taper decisions.
• Response: "completely or almost completely controlled by antiepileptic treatment in the vast majority of cases (96%)" in RELN families (PMID:28142128) — but note this is seizure control at last visit, not the stricter 5-year terminal remission of Bisulli 2018 (34.1%). Two different endpoints; don't blend them.

Pharmacogenomics

Not ADEAF-specific, but clinically mandatory for the first-line drug: HLA-B*15:02 screening before carbamazepine in Southeast Asian ancestry (SJS/TEN risk) and HLA-A*31:01 as a risk allele for carbamazepine hypersensitivity in European/Japanese populations — see CPIC carbamazepine guideline / FDA labeling. This is a real and actionable pharmacogenomic link for this syndrome's first-line agent.

Advanced / experimental therapeutics

• Chemical chaperone (preclinical, promising): 4-phenylbutyrate rescues secretion-defective LGI1 — "The 4PBA treatment significantly improved the secretion of LGI1 mutants and their binding to ADAM22" and "small molecules serving as chemical correctors might be new therapeutic options for LGI1-mediated epilepsy" (PMID:31432233). In 2025, 4PBA "selectively enhanced Cys48Phe secretion" (PMID:40455867). Mutation-class-specific: only helps foldable-but-retained alleles, useless for nulls. CHEBI:41500 (4-phenylbutyric acid).
• Gene/protein restoration (preclinical): single-neuron LGI1 re-expression normalizes excitability (PMID:37863654); Thy1-LGI1 transgene rescues the lethal Lgi1−/− phenotype. No AAV-LGI1 clinical program exists as of this writing. Given the ~50%-protein-sufficiency threshold, dose control would matter.
• Gene therapy / ASO / cell therapy / immunotherapy: none in trials for ADEAF. Not available. (Immunotherapy is for the autoimmune LGI1 disease, a different entity — don't cross-wire them in the KB.)

Surgical / interventional

Resective surgery is reserved for the drug-resistant minority — "Surgery might be employed instead in resistant cases" (PMID:35153984); the Bologna cohort included patients "addressed to surgery" (PMID:29464704). Outcome data for surgery specifically in ADEAF are not published in any systematic form; note that a genetic, potentially bilateral-network epilepsy is a less favorable surgical substrate than unilateral hippocampal sclerosis.

Supportive

Standard epilepsy self-management, sleep hygiene, driving counseling, seizure-safety education, and — specific to this syndrome — counseling that unexplained sounds are seizures, not psychosis. Genetic counseling is a core deliverable (§13).

Suggested NCIT annotations (verified via local sqlite:obo:ncit unless noted)

Clinical trials

NCT00072813 — MRI in Autosomal Dominant Partial Epilepsy With Auditory Features (observational imaging; Completed). No interventional ADEAF trial found on ClinicalTrials.gov as of 2026-08-05.

13. Prevention

• Primary prevention: none — the disease is germline. The only "primary prevention" lever is reproductive: genetic counseling, prenatal diagnosis, or PGT once the familial variant is known (GeneReviews), each carrying the heavy caveat of 54–85% penetrance and a generally treatable, non-degenerative phenotype.
• Secondary prevention: early recognition of auditory auras in known families → earlier diagnosis and treatment. Whether earlier treatment changes long-term remission is unstudied.
• Tertiary prevention: adherence, avoidance of sleep deprivation and (for the reflex-sensitive minority) unpredictable loud stimuli; correct drug selection (avoid valproate/phenobarbital chosen on a mistaken IGE diagnosis); cautious ASM withdrawal given relapse risk; driving/occupational safety.
• Immunization / public health / environmental interventions: not applicable.
• Prophylaxis: no pre-symptomatic ASM prophylaxis is indicated or studied in unaffected carriers. Given ~⅓ of carriers never develop seizures, treating carriers would be net harm.
• Counseling specifics: offspring of an affected individual have a 50% chance of inheriting the variant, of whom 54–85% will manifest (GeneReviews); siblings' risk depends on parental carrier status, and a clinically unaffected parent may still be a carrier because of reduced penetrance.

14. Other Species / Natural Disease

This is one of the better comparative stories in epilepsy genetics — dogs got there partly first.

• Taxonomy: Homo sapiens NCBITaxon:9606; Mus musculus NCBITaxon:10090; Rattus norvegicus NCBITaxon:10116; Canis lupus familiaris NCBITaxon:9615.
• Naturally occurring canine disease — LGI2 in Lagotto Romagnolo (Seppälä EH et al., PLoS Genet 2011;7:e1002194, PMID:21829378):

  "Using genome-wide association in 11 discordant sib-pairs from this pedigree, we mapped the disease locus to a 1.7 Mb region of homozygosity in chromosome 3 where we identified a protein-truncating mutation in the Lgi2 gene, a homologue of the human epilepsy gene LGI1." "We show that LGI2, like LGI1, is neuronally secreted and acts on metalloproteinase-lacking members of the ADAM family of neuronal receptors, which function in synapse remodeling, and that LGI2 truncation, like LGI1 truncations, prevents secretion and ADAM interaction." "LGI2 acts at least in part through the same ADAM receptors as LGI1, but earlier, ensuring electrical stability (absence of epilepsy) during pruning years, preceding this same function performed by LGI1 in later years."

Onset ~7 weeks, remitting by ~4 months — a developmentally shifted version of the same pathway, and a proposed model for remitting childhood epilepsy. Breed: Lagotto Romagnolo (VBO term available — look up before curating). Note this is autosomal recessive in dogs and involves the paralog LGI2 (hgnc:18710), so it is a pathway homolog, not a direct ADEAF ortholog model. - Canine ADAM23 risk haplotype (Koskinen LL et al., BMC Genomics 2015;16:465, PMID:26084559): a 28-kb ADAM23 risk haplotype replicated across Belgian Shepherd, Schipperke, Finnish Spitz, and Beagle (p_raw = 2.76e-15), frequency 0.49–0.70 — "ADAM23 plays a role in synaptic transmission and interacts with known epilepsy genes, LGI1 and LGI2, and should be considered as a candidate gene for human epilepsies." Low-penetrance risk gene, causative variant still unidentified. This is real veterinary relevance: idiopathic epilepsy is a common canine neurological disease and a DNA test exists for the Lagotto LGI2 allele. - Orthologs: mouse Lgi1 (MGI), Reln (the classic reeler mouse), Adam22, Adam23, Mical1; dog LGI2, ADAM23. Register in OMIA for the canine entries. - Evolutionary conservation: the LGI–ADAM synaptic module is conserved across mammals, with paralog subfunctionalization by developmental window (LGI2 early, LGI1 later — PMID:21829378). Genuinely elegant. - Zoonotic potential / cross-species transmission: not applicable.

15. Model Organisms

Model limitations to record explicitly: 1. No model reproduces the defining human feature — the auditory aura and ictal receptive aphasia are subjective, language/percept-dependent phenomena with no rodent readout. Any mouse-derived phenotype claim must be tagged MODEL_ORGANISM and must not be the sole support for a human phenotype. 2. Dosage mismatch: mouse homozygous nulls are lethal-epileptic while human heterozygotes have a mild focal epilepsy; the human-equivalent het mouse shows only threshold changes. 3. Anatomical mismatch: mouse discharges originate in hippocampus; human ADEAF is lateral-temporal neocortical. Strong candidate for a HUMAN_MODEL_MISMATCH discussion with a proposed experiment (region-restricted lateral-temporal Lgi1 deletion; auditory-cortex-targeted recordings). 4. RELN and MICAL1 arms are under-modeled — no published heterozygous-missense Reln ADLTE mouse, no Mical1 GoF knock-in.

Resources: MGI (mouse alleles for Lgi1, Adam22, Adam23, Reln), Alliance of Genome Resources, IMPC/KOMP, IMSR/MMRRC for strain requests, OMIA for the canine LGI2/ADAM23 entries, Cellosaurus for cell lines.

Appendix A — Suggested dismech pathophysiology skeleton

A causal chain that stays honest about scale and evidence:

1. LGI1 Loss of Function — biological_scale: MOLECULAR; gene hgnc:6572; secretion-defective vs binding-defective as two mechanistic_hypotheses-free routes (both established, so model as two upstream nodes converging, not competing hypotheses) → downstream to (2)
2. Failure of the LGI1–ADAM22/ADAM23 Trans-Synaptic Complex — MOLECULAR; GO:0050804; cell types CL:0000679 → (3a) and (3b)
3. a. Reduced AMPA Receptor-Mediated Synaptic Transmission — CELLULAR; GO:0098990, GO:0097113, modifier DECREASED b. Loss of Kv1.1 at the Axon Initial Segment — CELLULAR; GO:0005249 + GO:0043194, modifier DECREASED
4. Glutamatergic Neuron Hyperexcitability — CELLULAR; CL:0000598 / CL:1001571
5. Excitation-Inhibition Imbalance — CELLULAR; conforms_to: epilepsy_excitation_inhibition_imbalance#Excitation-Inhibition Imbalance
6. Lateral Temporal Neocortical Epileptogenesis — TISSUE; UBERON:0001393 / UBERON:0002769 / UBERON:0002773
7. Focal Sensory Auditory Seizure ± Receptive Aphasia → Focal to Bilateral Tonic-Clonic Seizure — ORGANISM; HP:0011158, HP:0032696, HP:0007334

Parallel entry nodes: RELN Secretion Deficiency (GO:0038026, gene hgnc:9957) and MICAL1 Oxidoreductase Gain of Function (GO:0030042, modifier INCREASED, gene hgnc:20619), both feeding node (5). A separate Developmental Arm: Arrested Excitatory Synapse Pruning (GO:1905805, modifier DECREASED, PMID:19701204) feeding node (4) captures the Zhou/Anderson mechanism without overwriting the acute-excitability chain.

Suggested discussions entries: - KNOWLEDGE_GAP — ~50% of ADEAF families and 92% of unselected EAF probands remain genetically unsolved (PMID:33453592). - KNOWLEDGE_GAP — modifiers explaining 54–85% penetrance are entirely uncharacterized. - KNOWLEDGE_GAP — "benign" (GeneReviews) vs 34.1% terminal remission (PMID:29464704); ascertainment-driven discrepancy, unresolved. - HUMAN_MODEL_MISMATCH — mouse discharges are hippocampal; human semiology is lateral-temporal neocortical; and no model can express an auditory aura.

Appendix B — Curation warnings

1. GeneReviews quotes will fail just validate-references. Every frequency I sourced to GeneReviews Table 2 (57–71% auditory, 17–20% aphasic, 88–92% FBTCS, 8–13% sound-triggered) needs re-sourcing to a PubMed-indexed primary paper before it becomes an evidence snippet. The RELN paper (PMID:28142128) supplies verifiable substitutes: 71% auditory, 88% FBTCS, 8% noise-precipitated, 80% EEG abnormalities, 96% controlled.
2. Do not map MICAL1-ADLTE to OMIM ETL8 (616461) — that entry is GAL-related. Verified via MedGen this session.
3. ADEAF ≠ anti-LGI1 encephalitis. Same protein, opposite direction of arrival (germline LoF vs acquired autoantibody), different age, different treatment. Keep them as distinct entries with a cross-reference.
4. Seed the MONDO caches for MONDO:0010898 in both cache/enums/*.csv and cache/mondo/terms.csv before pushing, and remember the term must be seeded from the worktree, not just the primary checkout.
5. Every mouse-derived claim gets evidence_source: MODEL_ORGANISM; every COS7/HEK/neuron-culture claim gets IN_VITRO; family series and cohorts get HUMAN_CLINICAL.

Sources

• GeneReviews: Autosomal Dominant Epilepsy with Auditory Features (NBK1537, PMID:20301709)
• Kalachikov et al., Nat Genet 2002, PMID:11810107
• Winawer et al., Neurology 2000, PMID:10851389 · Winawer et al., Epilepsia 2002, PMID:11879388
• Ottman et al., Neurology 2004, PMID:15079011 · Rosanoff & Ottman, Neurology 2008, PMID:18711109 · Ho et al., Neurology 2012, PMID:22323750
• Dazzo et al., Am J Hum Genet 2015, PMID:26046367 · Michelucci et al., Epilepsy Behav 2017, PMID:28142128 · Dazzo et al., Ann Neurol 2018, PMID:29394500 · Nobile & Dazzo, Genes 2022
• Pippucci et al., Neurol Genet 2015, PMID:27066544 · Bisulli et al., Seizure 2021, PMID:33453592 · Talarico et al., Int J Mol Sci 2024, PMID:39796146
• Bisulli et al., Epilepsia 2018, PMID:29464704 · Furia et al., Front Neurol 2022, PMID:35153984 · Riney et al., Epilepsia 2022
• Fukata et al., Science 2006, PMID:16990550 · Fukata et al., PNAS 2010, PMID:20133599 · Zhou et al., Nat Med 2009, PMID:19701204 · Boillot et al., Brain 2014, PMID:25234641 · Yamagata & Fukai, CMLS 2019, PMID:31432233 · Extrémet et al., J Neurosci 2023, PMID:37863654 · Hirano et al., Brain 2025, PMID:40455867
• Seppälä et al., PLoS Genet 2011, PMID:21829378 · Koskinen et al., BMC Genomics 2015, PMID:26084559
• gnomAD (constraint metrics via API) · OMIM 616436 / ETL7 · ClinicalTrials.gov NCT00072813