CACNA1E-Related Developmental and Epileptic Encephalopathy

1. Disease Information

Overview

CACNA1E-related developmental and epileptic encephalopathy — formally developmental and epileptic encephalopathy 69 (DEE69) — is a severe, early-onset neurodevelopmental disorder caused by de novo heterozygous variants in CACNA1E, the gene for the pore-forming α1 subunit of the Ca_V2.3 R-type voltage-gated calcium channel.

The defining clinical picture, straight from the paper that named it:

"we identified de novo CACNA1E variants in 30 individuals with DEE, characterized by refractory infantile-onset seizures, severe hypotonia, and profound developmental impairment, often with congenital contractures, macrocephaly, hyperkinetic movement disorders, and early death." — Helbig et al., Am J Hum Genet 2018 (PMID:30343943) [cached & validated]

The mechanistic one-liner: most pathogenic variants sit at the cytoplasmic ends of the four S6 helices — the channel's activation gate — and they're gain-of-function. The gate gets sloppy about closing. Calcium that should trickle in on a tight schedule instead floods in early and leaves late.

Key identifiers

(MONDO xrefs and synonyms verified locally against sqlite:obo:mondo this session.)

Orphanet: I could not confirm a dedicated ORPHA code for the CACNA1E-specific entity. CACNA1E appears as a causal gene under Orphanet's broader early-infantile DEE grouping (ORPHA:1934), but no ORPHA_* cache file for it exists in this repo. Flag as unresolved rather than asserting a code.

ICD: no disease-specific code. Falls under ICD-10 G40.4 (other generalized epilepsy and epileptic syndromes) / ICD-11 8A61 (developmental and epileptic encephalopathies) — these are syndrome-class codes, not entity codes, so don't treat them as identifiers.

Synonyms

• DEE69 · EIEE69
• Epileptic encephalopathy, early infantile, 69
• CACNA1E-related developmental and epileptic encephalopathy
• Developmental and epileptic encephalopathy with contractures, macrocephaly, and dyskinesias
• CACNA1E encephalopathy · Ca_V2.3 channelopathy (informal)

Data provenance

All disease-level knowledge here is aggregated from published individual-patient reports — GeneMatcher-assembled international cohorts, single case reports, and one video-EEG case study. There is no EHR-derived cohort, no patient registry, and no natural-history study for this disease. That's a real gap, not an oversight of my searching.

2. Etiology

Primary cause

Monogenic, essentially always de novo heterozygous missense variants in CACNA1E. No environmental, infectious, or multifactorial contribution is described anywhere in the literature I reviewed.

The variant architecture is startlingly tidy — a small mutational hotspot rather than a scatter:

"Most of the 14, partially recurring, variants cluster within the cytoplasmic ends of all four S6 segments, which form the presumed CaV2.3 channel activation gate." — PMID:30343943 [cached & validated]

That S6-distal clustering isn't unique to this gene, either. It's a recurring architectural motif across the calcium-channel epilepsies — a shared soft spot, like how the same vertebra tends to go in several different back conditions:

"Disease-causing variants predominantly involve gain-of-function mechanisms, with missense variants clustering in the distal S6 segment — a mutational hotspot shared across other voltage-gated calcium channel genes." — Lauerer & Lerche, J Neurochem 2023 (PMID:37822150) [fetched this session]

Genetic risk factors

• Causal: de novo GOF missense in CACNA1E. Recurrent hotspots: p.Gly352Arg (DI-S6) and p.Ala702Thr (DII-S6).
• Susceptibility loci / modifier genes: none identified. No GWAS, no modifier screen exists.
• Constraint: CACNA1E is a large, highly brain-expressed gene under evident missense constraint; Royer-Bertrand et al. explicitly compared the pathogenic-variant landscape against gnomAD and "identified protein regions intolerant to substitutions" (PMID:34702355). I did not retrieve the numeric gnomAD pLI/LOEUF/missense-Z values in this session (the browser is a JS app that WebFetch can't render) — fetch those directly before curating any specific number.

Environmental risk factors

None known. De novo mutation rate rises modestly with advanced paternal age as a general genomic phenomenon, but no paternal-age effect has been specifically demonstrated for CACNA1E — do not assert one.

Protective factors

None described in humans. The nearest thing is inverse-mechanistic and comes from mice: losing Ca_V2.3 function is anticonvulsant (see §15). No protective human allele is known.

Gene–environment interactions

None reported. Not applicable in any documented sense.

3. Phenotypes

The quantitative core

Almost every frequency in this disease traces back to one tabulation of the founding 30-person cohort. Here it is verbatim, because it's the single densest sentence in the whole literature:

"The 30 patients with pathogenic CACNA1E variants presented with a variable DEE characterized by refractory epilepsy (87% of patients) with median seizure onset at 4.5 months, movement disorders (60%), spastic quadriplegia (53%), congenital joint contractures (43%), macrocephaly (43%), and profound developmental impairments (ie, nonverbal and nonambulatory) (88%)." — Carvill, Epilepsy Curr 2019 (PMID:31064215) [cached & validated]

Phenotype table with HPO suggestions

Additional phenotypes from the non-epilepsy cohort (Royer-Bertrand 2021, PMID:34702355):

Verbatim: they found seven unrelated individuals with "intellectual disability, developmental regression and ASD-like behavioral profile, and notably without epilepsy" carrying de novo pathogenic CACNA1E variants (PMID:34702355) [fetched this session].

The seizure/movement-disorder trap — important for curation

Here's a wrinkle worth its own paragraph, because it undermines the neatness of the 60%-movement-disorder figure. When someone finally put a CACNA1E patient (p.Gly352Arg) on long-term video-EEG, some of the "movement disorder" turned out to be seizures wearing a costume:

long-term video-EEG revealed "tonic asymmetric seizures during wakefulness and mild paroxysmal dyskinesias of the trunk out of sleep which were thought to be a movement disorder and instead turned out to be focal hyperkinetic seizures." — Di Micco et al., Epileptic Disord 2024 (PMID:38780451) [fetched this session]

The authors call this the first documented description of the EEG findings in this disorder and note "a possible overlap between cortical and subcortical phenomena." So: some fraction of the reported dyskinesia burden is probably ictal, not extrapyramidal. Curate the movement-disorder phenotype with that caveat attached — it's a genuine open question, not a settled observation.

Quality of life

No EQ-5D, SF-36, PROMIS, or disease-specific QoL instrument has been applied to this population. Nothing to cite. Qualitatively, the profile — nonverbal, nonambulatory, refractory seizures, contractures, spastic quadriplegia, movement disorder — implies total dependence for all activities of daily living and a very high caregiver burden, but that inference is mine, not a published measurement. Curate it as description, not evidence.

4. Genetic / Molecular Information

The gene and its product

CACNA1E (1q25.3) encodes the ~2,251–2,270 aa, ~255–257 kDa α1E subunit — the pore-forming, voltage-sensing core of Ca_V2.3. Canonical four-domain (I–IV) architecture, each domain six transmembrane helices (S1–S6): S4s are the voltage sensors, S5–S6 line the pore, and the cytoplasmic ends of the four S6 helices form the activation gate where the disease variants live.

The channel carries the R-type current — historically "R for Resistant," since it's the high-voltage-activated current left standing after you block L, N, P, and Q types. From the abstract:

"CACNA1E is highly expressed in the CNS and is the pore-forming subunit of the voltage-gated calcium channel CaV2.3, which conducts high voltage-activated R-type calcium currents that initiate synaptic transmission." — PMID:30343943 [cached & validated]

Variant classes

Class 1 — GOF missense at the S6 activation gate (the disease-defining class). 14 partially recurring variants across 30 individuals in the founding cohort; ACMG classification pathogenic/likely pathogenic; germline de novo; absent from gnomAD.

Verbatim genotype–phenotype quotes (all PMID:31064215, cached & validated):

"Ten individuals carried missense variants in the DI-S6 domain, including 9 individuals with the recurrent p.Gly352Arg variant; all 10 individuals presented with hyperkinetic movement disorders, compared with only (2/19) individuals with variants outside this domain."

"In contrast, patients with missense variants located in the DIII-S6 presented with a milder phenotype, as 2 patients never developed seizures and one has been seizure-free for 5 years, spoke single words, and walked independently."

Class 2 — truncating variants (rare, milder, mechanistically murky).

"an additional 3 individuals with truncating CACNA1E variants were identified, including 1 somatic mosaic (27% of cells), 1 inherited from an unaffected parent, and 1 with unknown inheritance. All 3 individuals presented with a much milder phenotype" — PMID:31064215 [cached & validated]

This class matters out of proportion to its size: an unaffected transmitting parent means truncating CACNA1E alleles are not straightforwardly pathogenic, and it means the disease is not simply "too much or too little Ca_V2.3" on a single dial. Curate these as a distinct, uncertain class.

Class 3 — the non-epilepsy neurodevelopmental variants. Seven de novo variants (six missense, one splice-donor c.3422+1G>A predicted to disrupt the exon 22 donor site) producing ID/regression/ASD without seizures. Functional validation was explicitly not performed — Royer-Bertrand et al. list "lacks functional validation" among their own limitations. So we don't know whether these are GOF, LOF, or something else.

Functional consequence — the biophysics

Two independent GOF flavors, both documented:

"Functional analysis of several S6 variants revealed consistent gain-of-function effects comprising facilitated voltage-dependent activation and slowed inactivation." — PMID:30343943 [cached & validated, IN_VITRO]

"Another variant located in the domain II S4-S5 linker results in facilitated activation and increased current density." — PMID:30343943 [cached & validated, IN_VITRO]

"Electrophysiological recordings showed a hyperpolarizing shift in voltage-dependent activation, slowed kinetics of inactivation, and increased current density." — PMID:31064215 [cached & validated, IN_VITRO]

Put plainly: the gate opens at voltages where it shouldn't, and then dawdles about shutting. Two independent leaks in the same faucet.

Modifier genes, epigenetics, chromosomal abnormalities

• Modifiers: none identified.
• Epigenetics: no methylation episignature has been reported for CACNA1E (unlike, say, several other NDD genes with published episignatures). Nothing to curate.
• Chromosomal abnormalities: DEE69 is not a CNV/microdeletion syndrome. ClinGen has a dosage-sensitivity curation page for HGNC:1392 (search.clinicalgenome.org/kb/gene-dosage/HGNC:1392), but the site refused connection during this session — fetch the haploinsufficiency/triplosensitivity scores directly before curating them.
• Somatic vs germline: germline de novo for the missense class; one documented somatic mosaic truncating case (27% of cells).

5. Environmental Information

Short section, honestly stated: no environmental, lifestyle, toxicological, or infectious contribution is described for this disorder. It's a de novo monogenic channelopathy. CTD/TOXNET-style exposure associations for CACNA1E as a gene exist in the general toxicogenomics literature but have no established relationship to DEE69 — don't import them.

The only environment-adjacent factor with any mechanistic footing is fever/illness as a generic seizure precipitant in DEEs, and I found no CACNA1E-specific fever-sensitivity data. (Contrast this with Timothy_Syndrome in this KB, where a fever-exacerbation hypothesis is explicitly modeled — resist the temptation to copy that pattern here without evidence.)

One genuinely interesting modulator, though, and it's ionic rather than environmental: Ca_V2.3 is the classic zinc- and copper-sensitive calcium channel, and trace-metal modulation shapes convulsive seizure behavior in mice (PMID:32068980). That's a real biological handle on the channel, but it is model-organism pharmacology, not a human risk factor.

6. Mechanism / Pathophysiology

The causal chain

Here's the pathograph as I'd model it — five nodes, molecular → organism:

CACNA1E S6-gate GOF variant          [MOLECULAR — trigger]
   ↓ DIRECT
Increased neuronal Ca²⁺ influx        [CELLULAR — central effector]
   ↓ DIRECT
Altered excitability + transmitter release   [CELLULAR — intermediate]
   ├─ DIRECT ──→ Neuronal hyperexcitability / E-I imbalance   [TISSUE — central effector]
   │                 ├─→ Developmental and epileptic encephalopathy
   │                 └─→ Refractory infantile-onset seizures
   └─ INDIRECT ─→ Impaired neurodevelopment                   [ORGANISM — outcome]
     ├─→ Profound developmental impairment
     ├─→ Hyperkinetic movement disorder
     └─→ Spastic quadriplegia

Node 1 — the leaky gate (MOLECULAR). GOF missense at the S6 activation gate: facilitated (hyperpolarized) activation + slowed inactivation, sometimes plus increased current density. GO: GO:0005245 voltage-gated calcium channel activity (INCREASED), GO:0008331 high voltage-gated calcium channel activity (INCREASED).

Node 2 — calcium overload (CELLULAR). Ca_V2.3 lives in both compartments — presynaptic terminals and the somatodendritic membrane — so a gating defect raises Ca²⁺ at both ends of the neuron simultaneously. GO: GO:0070588 calcium ion transmembrane transport (INCREASED), GO:0019722 calcium-mediated signaling. CHEBI: CHEBI:29108 calcium(2+).

Node 3 — deranged synaptic signaling (CELLULAR).

"CACNA1E encodes Cav2.3, an R-type VGCC implicated in both presynaptic neurotransmitter release and postsynaptic somatodendritic integration and long-term potentiation." — PMID:31064215 [cached & validated]

GO: GO:0007269 neurotransmitter secretion (DYSREGULATED), GO:0042391 regulation of membrane potential (ABNORMAL), GO:0060291 long-term synaptic potentiation.

Node 4 — network hyperexcitability (TISSUE). This node conforms_to the KB's existing epilepsy_excitation_inhibition_imbalance#Neuronal Hyperexcitability and Hypersynchrony module — a clean conformance target.

"We establish pathogenic variants in CACNA1E as a cause of DEEs and suggest facilitated R-type calcium currents as a disease mechanism for human epilepsy and developmental disorders." — PMID:30343943 [cached & validated]

The cell-biological gearing was worked out in rodent CA1: Ca_V2.3 "triggers epileptiform activity in specialized neurons via plateau potentials and afterdepolarizations" (PMID:16686648). That's the specific machinery — R-type calcium entry sustains a depolarizing plateau after the spike, so instead of one action potential you get a burst, and instead of a burst you get a seizure. A dripping tap that never quite shuts off, and eventually the whole basin overflows in rhythm.

Node 5 — developmental encephalopathy (ORGANISM). Critically: the developmental impairment is not merely post-ictal damage. Two independent lines say so. (a) DIII-S6 patients who never seized still had disease; (b) the entire Royer-Bertrand cohort had ID/regression/ASD without epilepsy. Calcium is a developmental signaling currency — activity-dependent transcription, synapse refinement, circuit sculpting — so chronic mis-set calcium during a critical window degrades development independently of seizures. Curate the "developmental" and "epileptic" arms as parallel consequences, not as a cascade.

The CDKL5 convergence — the best mechanism news in this field

This one is genuinely elegant. Working from the other end — trying to figure out what the kinase CDKL5 actually phosphorylates — Sampedro-Castañeda et al. landed squarely on Ca_V2.3:

"We identified the voltage-gated Ca2+ channel Cav2.3 (encoded by CACNA1E) as a physiological target of CDKL5 in mice and humans. Recombinant channel electrophysiology and interdisciplinary characterization of Cav2.3 phosphomutant mice revealed that loss of Cav2.3 phosphorylation leads to channel gain-of-function via slower inactivation and enhanced cholinergic stimulation, resulting in increased neuronal excitability. Our results thus show that CDD is partly a channelopathy. The properties of unphosphorylated Cav2.3 closely resemble those described for CACNA1E gain-of-function mutations causing DEE69, a disorder sharing clinical features with CDD. We show that these two single-gene diseases are mechanistically related and could be ameliorated with Cav2.3 inhibitors." — Nat Commun 2023 (PMID:38081835) [fetched this session]

Read that again, because it's a two-gene convergence onto one channel state. In DEE69 the channel is broken open by a variant in its own gate. In CDKL5 deficiency disorder the channel is structurally fine but nobody's phosphorylating it, and unphosphorylated Ca_V2.3 behaves like a DEE69 mutant. Same phenotypic destination, two different roads — like two mutations in different enzymes of one pathway causing the same metabolite to pile up.

For this KB that suggests something concrete: a shared mechanism node ("Ca_V2.3 gain-of-function state") that both CACNA1E-Related_DEE and the existing CDKL5 entry could point at. Worth raising as a module candidate.

Additional GO terms for the CDKL5 arm: GO:0006468 protein phosphorylation, GO:0004674 protein serine/threonine kinase activity, and note the enhanced cholinergic stimulation finding (GO:0007271 synaptic transmission, cholinergic).

Protein structure

Two cryo-EM structures now anchor the mechanism in atoms:

• Human Ca_V2.3 + α2δ-1 + β3, 3.1 Å — "structural and electrophysiological characterizations showed that the CH2II helix stabilizes the inactivated conformation of the channel by tightening the cytosolic juxtamembrane segments" (PMID:36446785, Nat Commun 2022; PDB entries deposited with that paper).
• Human Ca_V2.3 + α2δ1 + β1 gating-mechanism structures — Nat Commun 2023, DOI 10.1038/s41467-023-36260-2.

Both papers frame Ca_V2.3 as a drug target "for pain, seizures, epilepsy, and Parkinson's disease." These structures were not available when the 2018 disease paper was written, so nobody has yet published a structure-mapped analysis of the DEE69 variant set onto the resolved gate. That's a very tractable open question and a good KNOWLEDGE_GAP candidate.

Not applicable / no data

• Metabolic changes: no metabolomic or energy-metabolism data. Nothing.
• Immune involvement: none described. Not an inflammatory disease.
• Tissue damage mechanisms: the only relevant thread is excitotoxicity, and it comes from mice — Ca_V2.3-null mice are protected from kainate excitotoxic cell death (PMID:17376845). Whether excitotoxic neuronal loss contributes in human DEE69 is unknown; no neuropathology series exists.
• Transcriptomics / proteomics / metabolomics / lipidomics / single-cell / spatial / multi-omics in patients: none published. The only omics that touched this disease is the SILAC phosphoproteomic screen in PMID:38081835 — and that was mouse/human neuronal material studying CDKL5, not patient tissue.
• Functional genomics screens: no CRISPR/RNAi screen targeting CACNA1E in a disease-relevant readout.

7. Anatomical Structures Affected

Primary: the central nervous system, full stop. CACNA1E is "highly expressed in the CNS."

Cell types:

Subcellular (GO cellular component): - GO:0005891 voltage-gated calcium channel complex - GO:0005886 plasma membrane - GO:0043195 terminal bouton (presynaptic pool) - GO:0030425 dendrite / GO:0043197 dendritic spine (somatodendritic pool) - GO:0042995 neuron projection

Lateralization: bilateral/diffuse for the encephalopathy. Notably, Di Micco et al. recorded tonic asymmetric seizures — so ictal semiology can be asymmetric even on a diffuse substrate. No structural brain malformation is a consistent feature; no systematic neuroimaging series has been published, which is itself a notable gap given that macrocephaly is in 43% of patients and nobody has characterized what's under those large skulls.

8. Temporal Development

Onset. Congenital for the structural features (contractures are present at birth — HP:0002803); infantile for the epilepsy. Median seizure onset 4.5 months (PMID:31064215). Hypotonia is a neonatal/early-infantile finding. The onset pattern is best described as congenital-plus-early-infantile rather than acute: the machinery was mis-set before birth, and the seizures are when it becomes audible.

Progression. No staging system exists for this disease. What can be said from the cohort:

• Seizures are refractory from the start — this is not a disorder with a good honeymoon period.
• Development is profoundly impaired and most patients never acquire speech or ambulation (88%) — so it reads as severe static-to-slowly-worsening impairment, not a neurodegenerative course. But no longitudinal study exists, so "static vs. progressive" is genuinely unresolved. Curate with PROGRESSIVE only if the schema demands a value, and flag it.
• Tone evolves: severe early hypotonia in many patients coexists with or gives way to spastic quadriplegia (53%).
• Early death occurred in a subset — "often with congenital contractures, macrocephaly, hyperkinetic movement disorders, and early death" (PMID:30343943). The abstract does not give a mortality rate or age distribution. Do not invent one.

Milder trajectories exist and matter. The DIII-S6 patients — some never seizing, one seizure-free five years with single words and independent walking — plus the entire truncating class ("much milder phenotype") and the whole non-epilepsy cohort together tell you this is a spectrum, not a monolith. Prognostic counseling anchored only on the severe DI/DII-S6 picture would be wrong for a meaningful minority.

Remission. Spontaneous remission not described. Treatment-induced seizure freedom is documented in five topiramate-treated participants (see §12) — that's the one bright spot in the course data.

Critical periods. No intervention-window data exist. Mechanistically, the developmental arm implies an early window (calcium-dependent circuit assembly in infancy), which would argue for early diagnosis and early mechanism-matched treatment — but this is a hypothesis, not a demonstrated therapeutic window.

9. Inheritance and Population

Inheritance: autosomal dominant, essentially always de novo. HP:0000006 Autosomal dominant inheritance.

Penetrance: for the recurrent GOF missense class, apparently complete — every reported carrier is affected. For the truncating class, penetrance is clearly incomplete: one such variant was inherited from an unaffected parent.

Expressivity: highly variable, and — unusually for a rare disease — partly predictable from variant location. DI-S6 → movement disorder in 10/10. DIII-S6 → milder, sometimes seizure-free. That domain-level genotype–phenotype correlation is one of the most curation-valuable facts in this entry.

Germline mosaicism: not reported. Somatic mosaicism is documented once (truncating variant, 27% of cells). Parental mosaicism has not been reported but cannot be excluded — relevant to recurrence-risk counseling.

Anticipation: not applicable (not a repeat-expansion disorder).

Founder effects / carrier frequency / consanguinity: all not applicable. De novo dominant disorders don't have carriers, founders, or a consanguinity signal.

Epidemiology: - Prevalence: not established. No population estimate exists. - The only defensible curation is a CASES_IN_LITERATURE measure: 30 individuals in the founding cohort, plus 7 in the non-epilepsy cohort, plus scattered case reports — on the order of 40–50 published individuals as of mid-2026. Prevalence class: ULTRA_RARE. - Incidence: unknown. - CACNA1E is a recognized contributor to the DEE gene pool but a numerically small one; it sits on the Genomics England PanelApp epileptic-encephalopathy panel and is reported as ClinGen "Definitive" for gene–disease validity with DEE69 (per ClinGen/GenCC search results this session — the ClinGen site itself refused connection, so verify the assertion ID and date before citing a CGGV reference).

Demographics: the founding cohort was an international GeneMatcher-assembled series with no reported ethnic clustering — expected, since de novo mutation doesn't respect ancestry. No sex ratio has been reported, and there's no biological reason to expect skew (autosomal, dominant, de novo). Age distribution of living patients is unstudied.

10. Diagnostics

Genetic testing — the whole ballgame

Diagnosis is molecular. There is no biochemical marker, no imaging signature, no functional test that makes this diagnosis.

"Using next-generation sequencing techniques, we identified de novo CACNA1E variants in 30 individuals with DEE" — PMID:30343943 [cached & validated]

Recommended approach, in order of yield: 1. Epilepsy/DEE gene panel or exome (WES) as first-tier for infantile-onset refractory epilepsy with developmental impairment. CACNA1E is on major DEE panels (Genomics England PanelApp panel 67, green). 2. Genome (WGS) when panel/exome is negative — better for the splice-region and non-coding space (note the c.3422+1G>A splice-donor variant class). 3. Trio testing — parental samples are essentially required, because de novo status is a major ACMG evidence line (PS2) for a gene where most missense is otherwise hard to classify. 4. Single-gene testing — only for targeted confirmation/cascade, not for discovery. 5. CMA / karyotype / FISH: appropriate as general NDD workup but not informative for this diagnosis — DEE69 is not a CNV syndrome. 6. Mitochondrial DNA testing / repeat-expansion testing: not applicable.

Variant interpretation tips specific to this gene: location is evidence. A missense at the cytoplasmic end of an S6 helix — especially at Gly352 or Ala702 — in a de novo trio configuration is a very different animal from a truncating variant, which may be benign or mildly-acting (an unaffected parent carried one). Royer-Bertrand's substitution-intolerance mapping against gnomAD provides a useful PM1-style regional argument.

MAXO: MAXO:0000533 molecular genetic testing.

EEG

"abundant epileptiform activity on EEG" — PMID:30343943 [cached & validated]

And, crucially, long-term video-EEG is not optional in this disease. Di Micco et al.'s case is the argument: routine assessment called the trunk paroxysms a movement disorder; video-EEG called them focal hyperkinetic seizures. That changes treatment. Their paper is the first published EEG characterization of the disorder — meaning the electroclinical phenotype is, at this point, an n-of-1.

MAXO: MAXO:0000932 electroencephalography.

Other modalities — status honestly reported

• Laboratory / biochemical tests: no diagnostic lab abnormality. No LOINC-codeable analyte. Routine metabolic workup is normal (and is typically done to exclude mimics).
• Biomarkers: none — not diagnostic, not prognostic, not pharmacodynamic. This is a real gap; a Ca_V2.3-function biomarker would be transformative for trial design.
• Imaging (MRI): no consistent malformation reported; no published imaging series. Given 43% macrocephaly, the absence of a systematic neuroimaging study is conspicuous.
• Biopsy / histopathology: no neuropathology series published. Not applicable clinically.
• Functional/electrophysiology beyond EEG: patch-clamp of variant channels is a research assay (heterologous expression), not a clinical test — though it's how the GOF mechanism was established and how a future variant-function service could work.
• Omics-based diagnostics (RNA-seq for the splice variants, proteomics, metabolomics, methylation episignature, liquid biopsy): none validated. RNA-seq is the one with a plausible near-term role, for splice-region variants like c.3422+1G>A.

Clinical criteria and differential diagnosis

No society-issued diagnostic criteria exist for DEE69 specifically. The syndrome is diagnosed as an ILAE-framework DEE plus a CACNA1E molecular result.

Differential diagnosis — the phenotype (infantile refractory seizures + profound impairment + hypotonia + movement disorder ± contractures ± macrocephaly) overlaps heavily with:

Screening

Not applicable. No newborn screening (de novo, no treatable metabolic marker), no carrier screening (no carriers), no cascade screening (parents are unaffected non-carriers in essentially all cases). Prenatal/preimplantation testing is only relevant for the rare family with proven parental mosaicism.

11. Outcome / Prognosis

Thin evidence, plainly stated.

Mortality. "Early death" is named in the founding abstract as an outcome in a subset (PMID:30343943). No mortality rate, no median survival, no cause-of-death breakdown has been published. By analogy with other severe DEEs, SUDEP, status epilepticus, and respiratory infection are the expected contributors — but that's inference, not data. Do not curate a survival percentage.

Morbidity. This is where the picture is unambiguous and severe: 88% nonverbal and nonambulatory. Add 53% spastic quadriplegia, 43% contractures, 60% movement disorder, 87% refractory epilepsy. Functionally that's total dependence for feeding, mobility, communication, and personal care — GMFCS-V-equivalent territory, though nobody has formally scored this cohort.

Quality of life measures: none applied. No EQ-5D, PROMIS, or disease-specific instrument. Genuine gap.

Complications (expected, largely uncurated in the literature): status epilepticus, aspiration and respiratory infection, feeding difficulty/failure to thrive, orthopedic sequelae of contractures and spasticity (hip dislocation, scoliosis), and the medication burden of polytherapy.

Recovery potential: none for the developmental impairment. Seizure freedom is achievable in a minority (topiramate, n=5).

Prognostic factors — the one real signal: variant domain. - DI-S6 / p.Gly352Arg → predicts hyperkinetic movement disorder (10/10). - DII-S6 / p.Ala702Thr → predicts the full severe spectrum. - DIII-S6 → predicts a milder course, potentially seizure-free with speech and ambulation. - Truncating → "much milder phenotype."

That's an unusually clean structure-to-prognosis map for a disease this young, and it should be the headline prognostic content of the KB entry.

Prognostic biomarkers: none.

12. Treatment

Topiramate — the mechanism-matched option

This is the therapeutic centerpiece, and it's a lovely bit of pharmacological serendipity: the drug that worked is a drug that blocks the very current the mutation amplifies.

"Five participants achieved seizure freedom on the anti-epileptic drug topiramate, which blocks R-type calcium channels." — PMID:30343943 [cached & validated]

Corroborated independently:

"Topiramate, an antiseizure drug (ASM) acting on CaV2.3 channels, has been identified as an effective treatment option in some patients." — Lauerer & Lerche, PMID:37822150 [fetched this session]

And notably it isn't only a seizure drug here — in the non-epilepsy cohort, "one patient demonstrated clinical improvement with topiramate" (PMID:34702355), hinting that lowering R-type current might touch the developmental/behavioral arm too.

Curation pattern for this KB:

- name: Topiramate
  therapeutic_modality: SMALL_MOLECULE
  treatment_term:
    preferred_term: Pharmacotherapy
    term: {id: NCIT:C15986, label: Pharmacotherapy}
    therapeutic_agent:
    - preferred_term: topiramate
      term: {id: CHEBI:63631, label: topiramate}
  target_mechanisms:
  - target: Increased Neuronal Calcium Influx
    treatment_effect: INHIBITS

Caveats worth stating in the entry: n=5, uncontrolled, retrospective, in a cohort where 87% had refractory epilepsy — so this is a promising signal, not an established standard of care. Topiramate is also a dirty drug pharmacologically (sodium channels, GABA-A, AMPA/kainate, carbonic anhydrase, and R-type calcium), so the R-type attribution is mechanistically plausible rather than proven.

The rest of the pharmacological picture

Nothing else has published efficacy data in this disorder. Standard DEE polytherapy applies by default (levetiracetam, valproate, benzodiazepines, vigabatrin, ketogenic diet, etc.), but I found no CACNA1E-specific response data for any of them and won't invent it. Seizures are described as pharmacoresistant as a rule.

Pharmacogenomics: none specific to this disease. Standard ASM pharmacogenetics (e.g. HLA-B*15:02 for aromatic ASMs) applies as it does to any epilepsy patient, not as a disease feature.

Advanced therapeutics — where things could go

• Selective Ca_V2.3 blockade is the obvious precision-therapy target, and it now has three independent endorsements: the topiramate signal, the mouse KO anticonvulsant phenotype, and the CDKL5 paper's explicit conclusion that both diseases "could be ameliorated with Cav2.3 inhibitors" (PMID:38081835). The two cryo-EM structures (PMID:36446785 and Nat Commun 2023) provide the structural basis for rational design. But: the only well-characterized selective blocker is SNX-482, a tarantula-venom peptide gating modifier (IC₅₀ ~30 nM) — a research tool, not a drug. It's a peptide, it partitions into membranes, it isn't CNS-deliverable. There is no clinical-stage selective Ca_V2.3 antagonist.
• Gene therapy / gene editing: nothing. And note the design difficulty — this is a gain-of-function dominant disorder, so the useful strategy is allele-specific knockdown (ASO or RNAi against the mutant allele), not gene replacement. No program exists.
• ASO therapy: no program. Mechanistically it's a reasonable fit for the recurrent hotspot variants (p.Gly352Arg, p.Ala702Thr are shared across many patients — exactly the economics that make an allele-specific ASO feasible). Purely prospective.
• Cell therapy, immunotherapy, targeted small-molecule oncology-style therapy: not applicable.

Surgical, supportive, rehabilitative

No epilepsy-surgery data (the encephalopathy is generalized/diffuse, so resective surgery is unlikely to apply; VNS/palliative approaches unreported). Orthopedic management of contractures and spasticity is expected practice, unreported in the literature.

Multidisciplinary supportive and developmental care is the practical mainstay: physical/occupational/speech therapy, spasticity and dystonia management, feeding support, seizure management, orthopedic surveillance.

MAXO suggestions: - MAXO:0000533 molecular genetic testing (diagnostic) - MAXO:0000932 electroencephalography (diagnostic) - MAXO:0000079 genetic counseling - MAXO:0000011 physical therapy - MAXO:0000950 supportive care - MAXO:0000004 surgical procedure (contracture release — inferred practice, uncited) - MAXO:0000088 dietary intervention (ketogenic diet — inferred, no CACNA1E-specific data) - NCIT:C15747 Supportive Care, NCIT:C15986 Pharmacotherapy

Clinical trials

Search result: no interventional trial specific to CACNA1E-related DEE or DEE69 was identified. Patients may be eligible for general DEE natural-history or platform studies, but no NCT identifier can be attached to this disease with confidence. Verify against ClinicalTrials.gov directly before curating any clinical_trials block.

13. Prevention

Bluntly: this section is mostly "not applicable," and saying so is more useful than padding it.

• Primary prevention: not possible. De novo mutation isn't preventable by any known behavioral, nutritional, or environmental intervention.
• Secondary prevention (early detection): the meaningful action is early genetic diagnosis — rapid trio exome/genome in an infant with refractory seizures and profound impairment. Earlier diagnosis buys mechanism-matched ASM selection (topiramate), stops the diagnostic odyssey, and enables accurate counseling. Whether early treatment changes the developmental trajectory is unknown but mechanistically plausible.
• Tertiary prevention: seizure control, aspiration/respiratory-infection prevention, contracture and hip-dislocation surveillance, nutritional support, status-epilepticus rescue planning. Standard severe-DEE management; nothing disease-specific published.
• Immunization: routine childhood schedule; no disease-specific vaccine strategy. (Standard practice for severe epilepsy applies — vaccinate, with fever management.)
• Population screening: not applicable — no newborn or carrier screening rationale.
• Genetic counseling: the real deliverable. Message: de novo dominant, recurrence risk low but not zero because parental gonadal mosaicism can't be excluded (and somatic mosaicism is documented in this gene). Prenatal testing available for a known familial variant. Domain-specific prognostic information (DI-S6 → movement disorder; DIII-S6 → milder) belongs in the counseling conversation.
• Public health / environmental interventions: not applicable.
• Prophylaxis: none established.

14. Other Species / Natural Disease

Orthologs:

Naturally occurring disease in other species: I found no OMIA entry or veterinary case literature describing a spontaneous CACNA1E disorder in companion animals or livestock. Report as "not identified," not as "does not exist" — absence of a search hit in a rare-gene space isn't proof.

Breed (VBO): not applicable — no breed-associated natural disease known.

Comparative biology: Ca_V2.3 is deeply conserved across vertebrates in both sequence and function; the R-type current, its zinc/copper sensitivity, and its presynaptic/somatodendritic dual localization are conserved features. The knockout mouse's seizure-resistance phenotype (§15) is the mirror image of the human GOF disease — which is about as good a cross-species mechanistic validation as you get without a knock-in.

Zoonotic potential / cross-species transmission: not applicable (genetic disorder).

15. Model Organisms

Mouse — the workhorse, and the evidence is strong

Constitutive Cacna1e knockout (Ca_V2.3⁻/⁻). These mice are the inverse-mechanism proof that Ca_V2.3 activity is pro-epileptogenic:

"PTZ-induced seizure susceptibility was dramatically reduced in Ca(v)2.3-deficient mice, whereas 4-AP sensitivity remained unchanged." — Weiergräber et al., Epilepsia 2006 (PMID:16686648) [fetched this session]

"Administration of kainic acid (30 mg/kg ip) revealed clear alteration in behavioral seizure architecture and dramatic resistance to limbic seizures in Ca(v)2.3(−/−) mice compared with controls." … "excitotoxic effects after kainic acid administration are absent in Ca(v)2.3(−/−) mice, whereas Ca(v)2.3(+/+) animals exhibited clear and typical signs of excitotoxic cell death." — Weiergräber et al., J Neurophysiol 2007 (PMID:17376845) [fetched this session]

And as summarized in the disease commentary:

"Cav2.3 is also known to play a role in epileptogenesis in rodents, and its deletion reduces susceptibility to chemically induced seizures." — PMID:31064215 [cached & validated, MODEL_ORGANISM]

Note the convulsant-specific pattern: protection against PTZ and kainate, not 4-AP. That's a mechanistic fingerprint, not blanket seizure resistance — Ca_V2.3 matters for particular routes into hypersynchrony.

Knockout mice also show normal brain structure, no spontaneous seizures, and no compensatory shifts in other calcium-channel expression — so the phenotype is attributable to the channel itself.

Ca_V2.3 phosphomutant mouse (2023) — the most disease-relevant model that exists. Sampedro-Castañeda et al. made mice in which the CDKL5 phosphorylation sites on Ca_V2.3 are ablated. The result: "loss of Cav2.3 phosphorylation leads to channel gain-of-function via slower inactivation and enhanced cholinergic stimulation, resulting in increased neuronal excitability," with properties that "closely resemble those described for CACNA1E gain-of-function mutations causing DEE69" (PMID:38081835). This is currently the closest available in-vivo model of the DEE69 channel state — an indirect GOF model rather than a patient-variant knock-in.

Ca_V2.3 in nigral dopaminergic neurons. Benkert et al., Nat Commun 2019 (PMID:31704946): Ca_V2.3 transcripts are the most abundant VGCC in mouse nigral neurons and rise with age; Ca_V2.3 knockout "afforded full protection from degeneration in vivo" in a neurotoxin Parkinson's model. Relevant here for channel biology and the movement-disorder question, not as a DEE69 model — but it's the reason people care about Ca_V2.3 selective blockers beyond epilepsy.

Zinc/copper modulation: PMID:32068980 — convulsive seizures are modulated in part by zinc ions through the pharmacoresistant Ca_V2.3 channel. Useful mechanistic color; not a disease model.

The critical gap

No patient-variant knock-in mouse exists. Nobody has made a Ca_V2.3-p.Gly352Arg or -p.Ala702Thr animal. Every in-vivo statement about this disease is therefore either (a) inverse — what happens when you remove the channel — or (b) indirect, via the phosphomutant. That's the single biggest experimental gap in the field, and it's the natural KNOWLEDGE_GAP / HUMAN_MODEL_MISMATCH to record: a knockout that is protected from seizures is not a model of a disease caused by gain of the same channel, however satisfying the logic feels.

Cellular models

• Heterologous expression (HEK/tsA-201 + α2δ and β subunits) with patch-clamp — the workhorse for variant functional characterization; this is how the GOF mechanism was established (PMID:30343943).
• Patient-derived iPSC neurons: none published for CACNA1E. The commentary explicitly calls for them: "future studies in patient-derived iPSC are likely to shed light on pathogenic mechanisms and potential therapeutic targets" (PMID:31064215). Still true seven years on.
• Organoids / MorPhiC: CACNA1E is not among the MorPhiC anchor genes (ISL1, EOMES, GCM1, NKX2-1). No cellular-phenotype data to import.
• Zebrafish: ZFIN maintains a human-disease page for DOID:0112205, but I found no published cacna1e zebrafish disease model. Verify before asserting.

Model resources

MGI (MGI:106217), IMPC/KOMP (check for Cacna1e allele availability), IMSR for strain sourcing, ZFIN, Alliance of Genome Resources for ortholog/phenotype integration.

Curation notes — what I'd flag before this goes in the KB

Solid, cite-able, already validated in this repo: the founding cohort's clinical percentages, the S6-gate variant clustering, the GOF biophysics, the domain-level genotype–phenotype correlations, the truncating-variant caveat, and the topiramate signal. All of that has snippet-verified evidence in references_cache/PMID_30343943.md and PMID_31064215.md.

New material worth adding to the existing entry (all fetched this session — run just fetch-reference and re-verify each snippet before committing): 1. PMID:34702355 — the non-epilepsy neurodevelopmental phenotype. This meaningfully broadens the entry beyond "DEE" and supports an ASD/ID phenotype block plus a has_subtypes or spectrum note. 2. PMID:38081835 — the CDKL5–Ca_V2.3 convergence. Strong candidate for a cross-disease mechanism link and possibly a shared module node with the CDKL5 entry. 3. PMID:38780451 — first EEG characterization; forces a caveat on the movement-disorder phenotype (some of it is ictal). 4. PMID:37822150 — Lauerer & Lerche review; good general-mechanism citation and supports the shared distal-S6 hotspot claim across calcium-channel epilepsies. 5. PMID:16686648 / PMID:17376845 — the primary mouse KO papers behind the existing MODEL_ORGANISM evidence, currently cited only secondhand through the commentary. 6. PMID:36446785 — cryo-EM structure; enables a structural node and a variant-mapping knowledge gap.

Things to verify, not assert: the ClinGen CGGV assertion ID (site was unreachable), gnomAD constraint numbers (browser is JS-only), the Orphanet code, ClinGen dosage scores, zebrafish paralog names, and any ClinicalTrials.gov entry.

Knowledge gaps worth encoding as discussions: - No patient-variant knock-in animal exists — everything in vivo is inverse or indirect (HUMAN_MODEL_MISMATCH fits better than KNOWLEDGE_GAP here, since evidence exists but its fidelity to the human GOF disease is the open question). - No selective, CNS-deliverable Ca_V2.3 antagonist despite three converging lines of rationale (already partly captured in the existing entry's gap_cacna1e_rtype_blocker_precision_therapy). - Truncating variants: benign, hypomorphic, or a distinct milder mechanism? An unaffected transmitting parent says the simple dose model is wrong. - No structure-mapped analysis of the DEE69 variant set onto the 2022/2023 cryo-EM gate. - No natural history, no imaging series, no neuropathology, no QoL instrument, no biomarker.

Sources: - De Novo Pathogenic Variants in CACNA1E Cause Developmental and Epileptic Encephalopathy with Contractures, Macrocephaly, and Dyskinesias — PubMed (PMID:30343943) - Calcium Channel Dysfunction in Epilepsy: Gain of CACNA1E — PubMed (PMID:31064215) - De novo variants in CACNA1E found in patients with intellectual disability, developmental regression and social cognition deficit but no seizures — PubMed (PMID:34702355) - Epilepsy-linked kinase CDKL5 phosphorylates voltage-gated calcium channel Cav2.3 — PubMed (PMID:38081835) - Seizure and movement disorder in CACNA1E developmental and epileptic encephalopathy — PubMed (PMID:38780451) - Voltage-gated calcium channels in genetic epilepsies (Lauerer & Lerche) — PMC11591408 (PMID:37822150) - Altered seizure susceptibility in mice lacking the Ca(v)2.3 E-type Ca2+ channel — PubMed (PMID:16686648) - Hippocampal seizure resistance and reduced neuronal excitotoxicity in mice lacking Cav2.3 — PubMed (PMID:17376845) - Structures of the R-type human Cav2.3 channel reveal conformational crosstalk — PubMed (PMID:36446785) - Molecular insights into the gating mechanisms of voltage-gated calcium channel CaV2.3 — Nature Communications 2023 - Cav2.3 channels contribute to dopaminergic neuron loss in a model of Parkinson's disease — PubMed (PMID:31704946) - Experimentally Induced Convulsive Seizures Are Modulated in Part by Zinc Ions through the Pharmacoresistant Cav2.3 Calcium Channel — PubMed (PMID:32068980) - OMIM #618285 — Developmental and Epileptic Encephalopathy 69 (403 to automated fetch; identifiers cross-verified via MONDO xrefs) - OMIM *601013 — CACNA1E gene - Genomics England PanelApp — CACNA1E (Epileptic encephalopathy panel) - GenCC — CACNA1E (HGNC:1392) submissions by classification - ClinGen dosage sensitivity — CACNA1E (connection refused during this session — verify manually) - MGI:106217 — mouse Cacna1e - ZFIN human disease page — developmental and epileptic encephalopathy 69 (DOID:0112205) - GTR — Developmental and epileptic encephalopathy, 69 (C4748988) - Orphanet — CACNA1E gene page