ABCC9-Related Intellectual Disability and Myopathy Syndrome

1. Disease Information

Overview

ABCC9-related Intellectual disability Myopathy Syndrome (AIMS) is an ultra-rare autosomal recessive K_ATP channelopathy caused by biallelic loss-of-function (LoF) variants in ABCC9, which encodes the sulfonylurea receptor 2 (SUR2) regulatory subunit of ATP-sensitive potassium channels. The core triad — established by the original 2019 description and confirmed across two subsequent cohort expansions — is (i) developmental delay with intellectual disability, (ii) muscle weakness and fatigability, and (iii) cerebral white matter abnormality resembling periventricular leukomalacia or small-vessel disease. Variable additional features include seizures, spasticity, microcephaly, corpus callosum abnormality, short stature, a shared facial gestalt, sensorineural hearing loss, and age-dependent cardiac systolic dysfunction.

AIMS is the mechanistic mirror image of Cantú syndrome: the same gene, opposite channel consequence. This LoF-vs-GoF duality is the single most important framing fact for the entry.

"We term this channelopathy resulting from loss-of-function of SUR2-containing K_ATP channels ABCC9-related Intellectual disability Myopathy Syndrome (AIMS). The phenotype differs from Cantú syndrome, which is caused by gain-of-function ABCC9 mutations, reflecting the opposing consequences of K_ATP loss- versus gain-of-function." — Smeland et al. 2019, PMID:31575858 [ABSTRACT-VERIFIED]

Key identifiers

Synonyms

• AIMS (the field-standard acronym; used in all three primary papers)
• ABCC9-related Intellectual disability Myopathy Syndrome
• Intellectual disability and myopathy syndrome (IDMYS — the OMIM-preferred name)
• SUR2 loss-of-function channelopathy

The draft's synonym list is accurate; consider adding IDMYS explicitly, since that is the OMIM label a curator or user is most likely to search.

Information provenance

All knowledge is aggregated case-level literature, not EHR- or registry-derived. The entire published world experience is ~20 individuals from ~13 families across three papers. There is no natural-history cohort, no registry, no patient-reported-outcome dataset, and no ICEES/COHD-style EHR signal available for this disease. Every frequency figure below is a small-n case-series proportion and should be curated with that caveat.

2. Etiology

Primary causal factor

Biallelic (homozygous or compound heterozygous) loss-of-function variants in ABCC9 → non-functional SUR2-containing K_ATP channels. This is a monogenic, fully genetic etiology; no infectious, toxic, or environmental cause contributes.

"Loss-of-function mutation of ABCC9, the gene encoding the SUR2 subunit of ATP sensitive-potassium (KATP) channels, was recently associated with autosomal recessive ABCC9-related intellectual disability and myopathy syndrome (AIMS)." — Efthymiou et al. 2024, PMID:38217872 [ABSTRACT-VERIFIED]

Genetic risk factors

• Causal: biallelic ABCC9 LoF (see §4 for the full allelic series).
• Consanguinity is a major risk factor for occurrence. Of the 2024 cohort's seven families, consanguinity was documented in the Pakistani, two Egyptian, British-Pakistani, and Saudi Arabian families [FULL-TEXT — RE-VERIFY]; the 2026 cohort included a first-cousin union [FULL-TEXT — RE-VERIFY].
• Founder-like enrichment: the Norwegian c.1320+1G>A allele is disproportionately frequent in Finns (see §9).
• Modifier genes: none identified. The 2019 authors explicitly kept the door open — "Notably, gene interactions with the homozygous ABCC9 c.1320+1G>A variant cannot be excluded as participating in the syndrome." (PMID:31575858, full-text body) [ABSTRACT-VERIFIED — full-text body line] This is a legitimate KNOWLEDGE_GAP discussion item.

Environmental risk factors

None are causal. However, AIMS has a genuinely important gene–environment / gene–exposure interaction at the level of decompensation triggers, which is arguably the most clinically actionable content in this entry:

Protective factors

No genetic or environmental protective factors are described. Heterozygosity is effectively protective/neutral for AIMS: parents are clinically unaffected carriers, and functional assays show no dominant-negative effect (§4, §6).

3. Phenotypes

Frequencies from the pooled published cohorts

The 2024 Brain paper provides the only published pooled frequency table (their cohort n=9, plus the 2019 cohort n=6, total n=15) [FULL-TEXT — RE-VERIFY all counts before curating frequency:]:

Curation caution. Note the discordance between cohorts — corpus callosum abnormality was 0/6 in 2019 but 6/9 in 2024, and dysmorphism was 6/6 in 2019 but 4/9 in 2024. This is ascertainment/expertise variation across small series, not a real biological gradient. Per docs/frequency-evidence-guidelines.md, the safest course for several of these rows is to omit frequency: and record the counts in notes: instead.

Officially curated HPO annotations (OMIM:619719)

The HPO project has annotated IDMYS from the 2019 cohort only. This is the authoritative, ready-to-use term set (IDs and labels retrieved from the HPO API and exact as given):

Additional terms needed for features from the 2024/2026 expansions

All verified during this research:

Ontology gap worth flagging: HPO has no single term capturing "periventricular-leukomalacia-like white matter change of vascular origin in a genetic channelopathy". The pragmatic composition is HP:0006970 + HP:0030891, with preferred_term carrying the nuance (permitted by the repo's preferred_term vs term.label rule).

Phenotype characteristics

• Onset: infantile in 5/6 of the original cohort (HP:0003593), congenital in 1/6 (HP:0003577). The 2026 series extends onset to the neonatal period (seizures on day 3 of life). One 2024-cohort participant had developmental regression at 15 months [FULL-TEXT — RE-VERIFY].
• Severity: intellectual disability is mild in most (5/6 mild, 1/6 moderate in the 2019 cohort) but "of variable severity" across the expanded cohort. Muscle weakness is mild-to-moderate (MRC grade 4 to 4+ in affected groups); one 2019 patient had entirely normal strength despite reporting exhaustibility.
• Progression: classically static/non-progressive in childhood, but two progressive dimensions are now documented: (a) white matter lesions accumulate — "Lesions have increased significantly since the first investigation" in the oldest 2019 patient (PMID:31575858 full text), and (b) adult motor and cognitive decline from the fifth decade in the 2026 sexagenarian.
• Episodic overlay: discrete decompensations (fever/vomiting-triggered coma episode; post-operative cardiogenic shock) punctuate an otherwise stable course. temporality: RECURRENT / EPISODIC is appropriate for the spasm and decompensation phenotypes.
• Quality-of-life impact: no formal instrument (EQ-5D, SF-36, PROMIS) has ever been applied in AIMS — a genuine data gap. Functional proxies from the 2019 cohort are quantitative and citable: all six had reduced 6-minute walk distance (e.g. 512.5 m vs 765 m reference) and moderate-to-severe balance deficits on the miniBESTest (scores 14–21 / 28). Supported-living need is documented: patients live in own homes or sheltered housing "with daily help and supervision," and one attends a sheltered work program but "is easily exhausted, and needs rest during the day."

4. Genetic/Molecular Information

Causal gene

Gene architecture matters mechanistically: ABCC9 has ≥39 exons and yields two major, tissue-divergent splice isoforms differing only in the terminal exon — SUR2A (striated muscle; pairs with Kir6.2/KCNJ11) and SUR2B (smooth muscle; pairs with Kir6.1/KCNJ8). Any variant proximal to the terminal exon — as all AIMS variants are — hits both isoforms, which is precisely why AIMS is multisystem while the exon-38-specific heterozygous DCM variants are cardiac/skeletal-limited.

Pathogenic variant spectrum (complete published allelic series)

Cohort 1 — Smeland et al. 2019 (PMID:31575858), 6 patients / 2 Norwegian families:

Cohort 2 — Efthymiou et al. 2024 (PMID:38217872), 9 patients / 7 families + 1 IUFD family [FULL-TEXT — RE-VERIFY]:

Cohort 3 — Nagaraj et al. 2026 (PMID:42290677), 5 patients [FULL-TEXT — RE-VERIFY]:

Spectrum summary for curation: - Variant classes: nonsense (4), frameshift (5+), canonical splice-site (3), whole-exon in-frame deletion (1), multi-exon genomic deletion (1). No pathogenic missense variant has ever been reported for AIMS — a sharp genotype contrast with Cantú syndrome, which is overwhelmingly missense GoF. - Molecular consequence: NMD of the transcript and/or drastic SUR2 truncation. > "All variants are predicted to lead to nonsense mediated decay of ABCC9 transcripts and/or drastic truncation of SUR2." — PMID:42290677 [ABSTRACT-VERIFIED] - Origin: exclusively germline. No somatic ABCC9 involvement in AIMS (somatic ABCC9 appears in COSMIC only as incidental cancer variation, unrelated to this disease). - Functional consequence: complete loss of function, not dominant-negative — see §6. The correct dismech modifier on the ABCC9 gene descriptor is DECREASED, and the relevant GENO framing is biallelic LoF. - ACMG/AMP: these are PVS1-anchored (null variant in a gene where LoF is the established mechanism) plus PM3 (recessive, in trans), PS3 (functional assays), PP1 (segregation) — expected Pathogenic. Note that ClinGen has published no variant pathogenicity assertions for ABCC9 (0 records), so no CGGV:/ClinVar expert classification is citable for AIMS variants.

Allele frequencies

• c.1320+1G>A: Finnish gnomAD AF 0.0007 (18/24,850); non-Finnish European 5/128,232 (AF 0.00004); absent in Asian and African populations; homozygous state absent from gnomAD (PMID:31575858 full text). Earlier ExAC counts: Finnish 3/6,586; non-Finnish European 4/66,386.
• Across the 2024 allelic series: "All the reported ABCC9 variants are…rare in the general population (max. allele frequency 0.000088) and absent in homozygous state in the gnomAD database." [FULL-TEXT — RE-VERIFY]
• Not retrieved: gnomAD gene-level constraint metrics for ABCC9 (pLI / LOEUF / o-e LoF). The gnomAD browser is client-rendered and could not be scraped in this session. If constraint is wanted for the entry, pull it from the gnomAD API or a local download rather than asserting a value.

Allelic series — other ABCC9 diseases (essential differential context)

The paralogous pancreatic/neuronal pair ABCC8/KCNJ11 completes the family logic: GoF → neonatal diabetes; LoF → congenital hyperinsulinism.

Modifier genes, epigenetics, chromosomal abnormalities

• Modifiers: none identified. The 2024 paper's observation that two unrelated families share c.3269delG with differing severity (P2 had seizures and regression; P3 had isolated ID with a normal MRI) is direct evidence of modifier or stochastic effects and is an excellent KNOWLEDGE_GAP/mechanistic_hypotheses candidate.
• Epigenetics: no data. No methylation episignature, no DiseaseMeth/ENCODE-based analysis exists for AIMS. Do not populate this section.
• Chromosomal abnormalities: none causal. Notably, karyotype/G-banding, high-resolution SNP array, FMR1 CGG repeat analysis, DMPK PCR, mtDNA screening, and metabolic screening were all normal in the 2019 patients before exome analysis — clinically useful negative information for the diagnostics section. The one exception is P4's 73.3 kb intragenic deletion of exons 13–33, which is a CNV detectable by array/CNV-aware sequencing.

5. Environmental Information

• Environmental factors: none contributory to causation. CTD/TOXNET yield no AIMS-relevant exposure.
• Lifestyle factors: none causal. Exercise is a symptom trigger (spasms, myalgia, fatigability), not a risk factor. Two 2019 patients became overweight (>97.5th percentile) in adulthood after early failure-to-thrive, and obesity contributed to obstructive sleep apnea in one — a modifiable comorbidity, not an etiologic factor.
• Infectious agents: not applicable. One febrile illness preceded a coma episode, but as a decompensation trigger rather than a pathogen-mediated cause.

The substantive content for this section is iatrogenic exposure (§12 drugs-to-avoid), which is where the real "environmental" risk in AIMS lies.

6. Mechanism / Pathophysiology

Molecular foundation

K_ATP channels are octameric complexes of four pore-forming Kir6.x subunits and four regulatory SURx subunits that couple cellular metabolic state to membrane excitability. ATP binding to Kir6.x inhibits the channel; Mg-ADP/Mg-ATP binding to the SURx nucleotide-binding domains stimulates it. SUR2 is therefore the metabolic sensor arm of the channel — losing it does not merely reduce current, it abolishes the metabolism-to-excitability coupling entirely in SUR2-dependent tissues.

Tissue pairing (PMID:31575858 full text): "Pancreatic and neuronal K_ATP channels are predominantly formed by Kir6.2 and SUR1, smooth muscle K_ATP channels are comprised of Kir6.1 and SUR2B, and the predominant combination in striated muscle is Kir6.2 and SUR2A."

Proximal molecular defect

The exon-8 in-frame deletion removes SUR2 residues Ala389–Gln440, disrupting multiple transmembrane helices in TMD1. Consequences measured in Cosm6 cells:

1. ~50% reduction in SUR2 protein on Western blot.
2. Complete abolition of ⁸⁶Rb⁺ efflux — "cells expressing Kir6.2/SUR2AΔ8 showed no rubidium efflux above the background levels observed in GFP-transfected cells."
3. Complete absence of K_ATP current in inside-out patch clamp.
4. Same result for SUR2BΔ8 — so both isoforms are dead.
5. No dominant-negative effect: a 1:1 WT:Δ8 co-expression gave WT-like efflux rates and unchanged ATP IC₅₀ / pinacidil activation — "in the heterozygous context, the c.1320+1G>A mutation is without significant effect." This is the molecular explanation for healthy carrier parents.

The 2026 paper reproduced complete LoF for three further truncating variants (p.Glu961LysfsTer10, p.Gly1090AspfsTer2, p.Ser601ArgfsTer9) using a landing-pad HEK293 / FLIPR membrane-potential assay (no oligomycin-A-induced hyperpolarization) plus patch clamp, again with no major dominant-negative effect [FULL-TEXT — RE-VERIFY].

"Functional tests of recombinant channels confirm that disease-associated SUR2 truncations cause a complete loss-of-function." — PMID:42290677 [ABSTRACT-VERIFIED]

Causal chains — three mechanistically distinct downstream arms

The pathophysiology graph should be modeled as one molecular trigger fanning into three tissue-specific arms, because the arms have genuinely different intermediate biology.

Arm A — Skeletal muscle (myopathy). This arm is causally proven to be muscle-intrinsic.

ABCC9 biallelic LoF (MOLECULAR)
  → loss of SUR2A/Kir6.2 KATP in skeletal myocytes (CELLULAR)
  → failure of fatigue-induced action-potential shortening and
    resting-membrane-potential stabilization (CELLULAR)
  → excessive Ca2+ entry, elevated resting tension,
    abnormal unstimulated force generation (CELLULAR)
  → myofiber stress/degeneration, impaired fatigue recovery (TISSUE)
  → fatigability, muscle weakness, painful exercise-induced spasms (ORGANISM)

The 2019 mechanistic rationale, verbatim from the full text: "In skeletal muscle, K_ATP channels are typically closed at rest but open in response to metabolic stress or fatigue. Channel activation results in action potential shortening and stabilization of the resting membrane potential during the development of fatigue, which serves to reduce intracellular calcium, decrease resting tension and protect myocytes from damage. Therefore, loss of K_ATP function might be expected to result in failure to recover from fatigue, myofiber degeneration, and excessive calcium influx."

McClenaghan et al. 2023 then localized this arm by tissue-selective knockdown and, importantly, falsified the obvious Ca²⁺-influx hypothesis:

"Given the roles of KATP channels in all muscles, we sought to determine how myopathy arises using tissue-selective suppression of KATP and found that LoF in skeletal muscle, specifically, underlies myopathy. In isolated muscle, SUR2 LoF results in abnormal generation of unstimulated forces, potentially explaining painful spasms in AIMS. We sought to determine whether excessive Ca2+ influx through CaV 1.1 channels was responsible for myopathology but found that the Ca2+ channel blocker verapamil unexpectedly resulted in premature death of AIMS mice and that rendering CaV 1.1 channels nonpermeable by mutation failed to reverse pathology; results which caution against the use of calcium channel blockers in AIMS." — PMID:37154692 [ABSTRACT-VERIFIED]

That negative result is scientifically important and should be curated as a supports: REFUTE evidence item against a CaV1.1-influx sub-hypothesis — it is exactly the kind of falsified mechanism dismech is designed to record rather than discard.

Arm B — Cerebral vasculature → white matter injury (the leading explanation for the CNS phenotype).

ABCC9 biallelic LoF (MOLECULAR)
  → loss of SUR2B/Kir6.1 KATP in vascular smooth muscle (CELLULAR)
  → increased vascular tone / hypertension; loss of metabolic
    vasodilatory reserve (TISSUE)
  → impaired dynamic coupling of cerebral blood flow to
    neuronal metabolic demand; chronic hypoperfusion (TISSUE)
  → ischemic periventricular white matter injury, small-vessel
    disease, lacunar-like cavitation, mineral deposition (TISSUE)
  → intellectual disability, spasticity, hyperreflexia,
    seizures, later cognitive decline (ORGANISM)

The 2019 authors' framing: "recent data demonstrate that SUR2-containing K_ATP channels play a critical role in regulation of cerebral vascular architecture… Since K_ATP channel GoF results in chronic vasodilation and altered neuro–vascular coupling, it is conceivable that K_ATP loss of function may impact the cerebral vasculature in a way that results in impaired dynamic coupling of blood flow to match neuronal metabolic demand." They add that the white matter hyperintensities seen in both Cantú syndrome and AIMS "could result from ischemic events due to dysregulated cerebral blood flow, although the cognitive phenotype seems to be more definite in AIMS than in Cantú syndrome."

Supporting evidence: hypertension was significantly increased in SUR2-STOP mice; cardinal vein red-cell velocity was increased in SUR2-STOP zebrafish; and the 2026 sexagenarian's imaging evolved into frank severe vascular leukoencephalopathy with perivascular space dilation and lacunar-like cavitations [FULL-TEXT — RE-VERIFY]. Note also that MR angiography was normal with normal cerebral vessel calibers in the 2019 patients — so this is a functional microvascular/autoregulatory defect, not a macrovascular malformation. That distinction is a genuine contrast with Cantú syndrome (dilated, tortuous vessels) and should be stated explicitly.

Mark this arm as status: EMERGING in mechanistic_hypotheses — the 2019 authors themselves say "it is conceivable," and no direct human CBF/neurovascular-coupling measurement in AIMS exists.

Arm C — Cardiac (age-dependent).

ABCC9 biallelic LoF (MOLECULAR)
  → loss of SUR2A/Kir6.2 KATP in cardiomyocytes; loss of
    metabolic stress protection (CELLULAR)
  → reduced contractile reserve, cardiomyocyte apoptosis
    (TUNEL+ in zebrafish) (CELLULAR/TISSUE)
  → biventricular systolic dysfunction, LV dilation (TISSUE)
  → early-stage dilated cardiomyopathy, heart failure;
    acute cardiogenic shock under stress (ORGANISM)

Human anchor: the two oldest 2019 patients had biventricular systolic dysfunction (EF 35–40% and 48%), raised NT-proBNP, and cardiac-MRI LV dilation "compatible with early-stage dilated cardiomyopathy," while a teenage echocardiogram in the same patient had been normal — establishing age-dependent emergence. The authors compare with the heterozygous exon-38 DCM patients who were older with far worse EFs (15–23%) and conclude: "This may suggest a progressive cardiomyopathy which will require longitudinal analysis in AIMS."

Notably, zebrafish hearts showed no fibrosis (AFOG) and normal myofiber/tropomyosin structure despite marked ventricular enlargement — so this is a functional/apoptotic remodeling process rather than a primary fibrotic cardiomyopathy. That argues against naive conformance to fibrotic_response and for conformance to cardiomyopathy_maladaptive_remodeling.

Unexplained: the intellectual disability itself

The 2019 authors are admirably candid, and this should be curated as a first-class knowledge gap:

"It is not obvious how myocyte K_ATP dysfunction could explain the intellectual disability or anxiety, and there are no reports of cognitive impairment in previous SUR2 mutant animal models." Neuronal K_ATP is predominantly Kir6.2/SUR1, though SUR2 transcripts are reported in central and peripheral neurons and implicated in hippocampal sclerosis of aging and ALS.

This is a textbook HUMAN_MODEL_MISMATCH case per the repo's CLAUDE.md guidance, not a plain KNOWLEDGE_GAP: the SUR2-STOP mouse was formally tested for cognition and anxiety and was negative on all of it (locomotor activity, sensorimotor battery, Morris water maze spatial learning, elevated plus maze), while the defining human feature is intellectual disability plus anxiety in 4/4 assessed. The authors' own caveat — "mild cognitive impairment, as seen in humans, can be difficult to recognize in animals" — plus their suggestion of working-memory/fear-conditioning testing maps directly onto proposed_experiments.

Cellular processes, cell types, and compartments

Metabolic, immune, biochemical, and profiling data

• Biochemical abnormality: the defect is an ion channel defect. Routine biochemistry is characteristically normal: "S-CK, B-lactate, nerve conduction velocities, and electromyography, including repetitive nerve stimulation, are normal in all individuals, except for discrete myopathic discharges in patient 2–2." Abnormal values appear only during crises (lactate 6, CK 738) or incidentally (hyperprolactinemia 720 mIU/L in one patient, with normal pituitary MRI).
• Metabolic changes: no primary metabolic derangement; inborn-errors-of-metabolism screening was normal. Mouse SUR2 nulls show lower serum glucose and enhanced insulin action (MGI) — an interesting, unexplored translational thread with no human counterpart reported.
• Immune involvement: none mechanistically. Atopic dermatitis (2/6), food allergies with anaphylaxis, and cyclosporine treatment in one patient are recorded features whose relationship to SUR2 is entirely unexplained — curate as associations, not mechanism.
• Transcriptomics / proteomics / metabolomics / lipidomics / single-cell / spatial / multi-omics / CRISPR screens: no AIMS-specific datasets exist. The only quantitative expression data is zebrafish qPCR showing ~4-fold reduced abcc9 mRNA in SUR2-STOP larvae, consistent with NMD. Leave these subsections empty rather than importing generic ABCC9 GTEx data as if it were disease data.

Suggested dismech module conformance

AIMS is a strong multi-module conformer candidate: - cardiomyopathy_maladaptive_remodeling → Arm C (#Ventricular Remodeling). Good fit; note the absence of fibrosis in the fish model. - epilepsy_excitation_inhibition_imbalance → the seizure phenotype (#Excitation-Inhibition Imbalance), with the zebrafish PTZ hypersensitivity as supporting MODEL_ORGANISM evidence. - A new K_ATP channelopathy grouping is the more valuable structural contribution: AIMS + Cantú syndrome + DCM1O + neonatal diabetes + congenital hyperinsulinism form a textbook GoF/LoF antagonistic pair set across two paralogous gene pairs. This parallels the existing cellular_senescence / senescence_tumor_suppression two-module treatment of opposing arms, and would fit kb/groupings/ with grouping_basis: [SHARED_MECHANISM, SHARED_GENE_FAMILY].

7. Anatomical Structures Affected

Primary organs / systems

* commonly-used IDs stated from knowledge; verify with just validate-terms / OAK before committing. Only UBERON:0002437 and UBERON:0008967 were verified live in this session.

Secondary involvement: skeleton (lumbar hyperlordosis, scoliosis, congenital hip dislocation, Achilles contractures), eye/oculomotor (nystagmus, strabismus/esotropia, heterotropia), inner ear (bilateral high-frequency sensorineural hearing loss), skin (cutis marmorata in 6/6, spotty hypopigmentation, café-au-lait macule, atopic dermatitis, facial telangiectasia), craniofacial/dental (macrodontia, malocclusion, widely spaced incisors), GI (feeding difficulties, ileal atresia in the neonate), upper airway (obstructive sleep apnea).

Tissue and cell level: striated muscle fibers (CL:0008002), cardiac myocytes (CL:0000746), vascular smooth muscle cells (CL:0000192). Muscle histology is strikingly bland — a single biopsy showed only "unspecific changes of mitochondrial aggregation and muscle fiber caliber variation" — consistent with mouse data showing histological abnormality only after chronic exercise.

Subcellular level: the plasma membrane K_ATP complex (GO:0008282) is the site of the defect; the 2019 paper specifically frames the loss as of "plasmalemmal K_ATP function." A mitochondrial angle exists in the literature (a prior mouse allele retains a "mitochondria-limited short form" of SUR2, and one patient's biopsy showed mitochondrial aggregation) but mitochondrial SUR2 is not established as an AIMS mechanism — do not assert it.

Lateralization: predominantly bilateral and symmetric (white matter changes, hearing loss, hyperreflexia, ventricular dysfunction). Documented asymmetric exceptions: unilateral left lower-extremity weakness with limping in one patient, unilateral Achilles contracture in another, and a single left-anterior-horn periventricular lesion.

8. Temporal Development

Onset - Neonatal: recurrent emesis, feeding difficulty, ileal atresia, multifocal seizures on day 3 (2026 case) — the earliest documented presentation. - Infantile: HP:0003593 in 5/6 of the original cohort — hypotonia, delayed psychomotor development, feeding difficulty, low weight, typically noticed in toddler years. - Congenital: HP:0003577 in 1/6 (congenital hip dislocation). - Prenatal / lethal extreme: recurrent intrauterine fetal death in heterozygous couples, including two IUFDs of fetuses confirmed homozygous for p.(Arg620Ter) [FULL-TEXT — RE-VERIFY]. - Pattern: insidious and chronic, with superimposed acute episodes.

Progression / stages — a useful three-stage model emerges from the pooled natural history: 1. Early (infancy–childhood): hypotonia, developmental delay, feeding difficulty, failure to thrive, toe-walking/in-toeing, lumbar lordosis. Cardiac and cognitive testing may be normal. 2. Intermediate (adolescence–early adulthood): mild-to-moderate ID confirmed on neuropsychological testing, fatigability, exercise-induced myalgia/spasms, hyperreflexia, balance impairment, accumulating white matter lesions, anxiety; weight often shifts from low to overweight. 3. Late (fourth decade onward): biventricular systolic dysfunction/early DCM; progression of leukoencephalopathy toward vascular small-vessel disease with cavitation and calcification; motor and cognitive decline with global muscle atrophy and dementia-like presentation (2026 sexagenarian, died in her early sixties of aspiration-related asphyxiation) [FULL-TEXT — RE-VERIFY].

Rate: slow. Static-appearing through childhood, which is why AIMS was long mistaken for a static encephalopathy — but explicitly not static on imaging ("Lesions have increased significantly since the first investigation") or in late adulthood.

Course: chronic, lifelong, progressive with an episodic overlay. Not relapsing-remitting.

Remission: none. The only documented reversibility is imaging: the fever-triggered multifocal lesions in patient 1–4 "were normalized a few weeks later, except for a white matter lesion," and he "regained the same psychomotor level as before" — evidence that acute decompensations can be survivable and partly reversible.

Critical periods / windows for intervention - Perioperative period — the highest-acuity documented window (two post-operative decompensations in one neonate). - Antiseizure and anesthetic drug selection at any age. - Adulthood cardiac surveillance — the 2019 patient had a normal teenage echocardiogram and dysfunction by age 33, so a single normal childhood echo does not discharge the risk. - Early developmental years for habilitation, feeding, and educational support.

9. Inheritance and Population

Epidemiology. No prevalence or incidence estimate has ever been published, and there is no Orphanet prevalence class (no ORPHA code exists). The literature total is ~20 individuals from ~13 families (2019: 6; 2024: 9 + IUFDs; 2026: 5). For dismech Prevalence, the honest structured record is:

prevalence:
- population: Worldwide
  measure_type: CASES_IN_LITERATURE
  prevalence_class: ULTRA_RARE
  notes: >-
    Approximately 20 affected individuals from ~13 families reported across
    three publications (2019-2026). No prevalence or incidence estimate has
    been published; no Orphanet epidemiology record exists.

Do not derive a rate_per_100000 from carrier frequency — that inference is not in the literature.

Inheritance pattern. Autosomal recessive (HP:0000007). Verified by homozygosity in all affected individuals with heterozygous unaffected parents and one heterozygous unaffected sibling, plus one compound heterozygote with variants confirmed in trans (paternal/maternal).

Penetrance. Appears complete for the core triad in biallelic carriers, though the very small n and the 2026 participant with isolated ID and a normal MRI show variable expressivity of individual features. Heterozygotes have no AIMS phenotype: "Heterozygous parents do not show any conserved clinical pathology but report multiple incidences of intra-uterine fetal death" (PMID:38217872) [ABSTRACT-VERIFIED].

Expressivity. Variable — most sharply demonstrated by the two unrelated homozygotes for the identical c.3269delG allele with markedly different severity (regression + seizures vs. IQ 69 with normal imaging). Cardiac involvement is age-dependent rather than variant-dependent.

Genetic anticipation. Not applicable (no repeat expansion).

Germline mosaicism. Not reported.

Founder effect. Strongly suggested for c.1320+1G>A: both 2019 families came from the same area of Northern Norway with probable Finnish ancestry, yet whole-genome kinship analysis excluded relatedness (kinship coefficient 0.0403, versus ~0.05 expected for unrelated samples), and the allele sits in a shared 3.8 Mb homozygous block (chr12:18,326,590–22,176,010, hg19). Combined with the Finnish gnomAD AF of 0.0007, the authors concluded: "Considering the probable Finnish ancestry of all patients, the syndrome might be more prevalent in the Finnish population than others." A second Norwegian allele (c.284+1G>A) appeared in the 2024 cohort.

Consanguinity. A major contributor outside Northern Europe — documented in Pakistani, Egyptian (×2), Saudi Arabian, and British-Pakistani families, and in the 2026 first-cousin family.

Carrier frequency. Best estimate is the Finnish c.1320+1G>A allele frequency of 0.0007 (≈1 in 690 carriers for that single allele in Finns) — but note this is a per-allele figure, not a pan-allelic carrier rate, and no carrier-screening study exists.

Population demographics. Reported ancestries: Norwegian (with probable Finnish ancestry), Dutch, Pakistani, British Pakistani, Egyptian, Saudi Arabian, Brazilian and Italian contributions via the 2026 author network. Absent from gnomAD Asian and African populations for the founder allele. Geographic distribution reflects where exome sequencing and expert networks are concentrated, not true disease distribution — this is ascertainment bias and should be stated as such.

Sex ratio. Roughly balanced with no evidence of sex bias: 2019 cohort 2F/4M; 2026 cohort 4F/1M. Consistent with autosomal inheritance. (Note: the PMID:31575858 MeSH set anomalously includes Genetic Diseases, X-Linked and Intellectual Disability/parasitology — clear MEDLINE indexing artifacts, not claims about the disease. Do not propagate these.)

Age distribution of reported individuals. Newborn to early sixties; the 2019 cohort spanned 11–33 years at report.

10. Diagnostics

The diagnostic gestalt

The 2026 authors give the single most useful practical recommendation: the constellation of periventricular leukomalacia + developmental delay/intellectual disability + muscle weakness and fatigability should prompt investigation for an ABCC9-related disorder [FULL-TEXT — RE-VERIFY]. This matters because the imaging pattern mimics acquired perinatal hypoxic-ischemic injury, so AIMS is easily dismissed as non-genetic.

Laboratory tests — mostly normal, which is itself diagnostic information

• Creatine kinase: normal in all 2019 patients and in the 2026 participant tested; elevated (738) only during acute crisis. A normal CK does not exclude AIMS — a critical negative, since it steers clinicians away from a myopathy workup. LOINC: CK 2157-6.
• Lactate: normal at baseline; elevated in crisis (6 mmol/L). LOINC 2524-7.
• Prolactin: hyperprolactinemia in one patient with normal pituitary MRI (incidental). NT-proBNP: raised in the patient with heart failure, normal in his sister with milder dysfunction.
• Normal: metabolic/IEM screening, mtDNA sequence and deletion screening.

Biomarkers: there is no diagnostic or prognostic biomarker for AIMS. NT-proBNP serves only as a generic heart-failure marker. This is a real gap worth recording.

Imaging — the highest-yield test

Brain MRI is the cornerstone. Findings: periventricular and juxtacortical white matter T2/FLAIR hyperintensities (centrum semiovale, frontal and parietal, anterior temporal lobes, external capsules); reduced periventricular white matter volume; corpus callosum hypoplasia/agenesis; punctate lesions in basal ganglia, pons, brainstem and cerebellum suggesting mineral deposition/calcification; in advanced disease, dilated perivascular spaces and lacunar-infarct-like cavitations. MR angiography is normal with normal vessel calibers — a key discriminator from Cantú syndrome. MRS was normal in the one patient tested. Serial imaging is recommended, since lesions accumulate.

Cardiac imaging: echocardiography ± cardiac MRI. Cardiac MRI detected LV dilation that echo missed (LV diameter was within normal range on echo in the same patient) — so echo alone may under-call early cardiomyopathy.

Electrophysiology — characteristically normal

EMG, nerve conduction velocity, repetitive nerve stimulation, and EEG were normal in all 2019 patients (bar discrete myopathic discharges in one). Notably, one patient with clinical tonic–clonic episodes had repeatedly normal EEGs — clinically important, because normal EEG does not exclude the paroxysmal events, and it feeds the hypothesis that some episodes are circulatory rather than epileptic.

Biopsy / pathology

Muscle biopsy is low-yield and non-specific: "caliber changes, mitochondrial aggregations." Not recommended as a diagnostic route; its main value is exclusionary.

Genetic testing — the definitive route

• Recommended first-line: exome or genome sequencing (trio-based where possible). Both cohorts were solved this way, and the 2019 patients had already had extensive normal targeted testing (G-banding, high-resolution SNP array, FMR1 CGG, DMPK PCR, multiple neuromuscular gene panels, mtDNA, IEM screening) before ES/GS succeeded.
• Gene panels: ABCC9 is included in large neurodevelopmental/neuromuscular panels (the 2019 study used a >4,800-gene panel). No AIMS-specific panel exists.
• WGS adds value for excluding alternatives, resolving relatedness (kinship coefficient), and for CNV detection — essential given the 73.3 kb exon 13–33 deletion in the 2026 cohort. A sequencing pipeline blind to intragenic CNVs can miss AIMS.
• Single-gene testing is reasonable only for targeted familial testing or in a Finnish/Northern Norwegian patient with the classic triad (founder allele).
• CMA/karyotype/FISH: normal in AIMS; CMA would only detect the large intragenic deletion if probe coverage is adequate.
• mtDNA testing: normal — but frequently performed, because the phenotype (myopathy + white matter disease + lactate rise in crisis) mimics mitochondrial disease.
• Repeat expansion testing: not applicable.
• RNA/functional confirmation: cDNA analysis from fibroblasts confirmed exon 8 skipping in the 2019 family — a good template for splice-variant validation. Recombinant channel assays (⁸⁶Rb⁺ efflux, patch clamp, landing-pad/FLIPR membrane potential) are research-grade PS3 evidence, not clinical tests.
• Omics diagnostics (proteomics/metabolomics/epigenomics/liquid biopsy): no role established.

Clinical criteria and differential diagnosis

There are no formal consensus diagnostic criteria — diagnosis is molecular plus phenotype fit.

Differential, with discriminators (several of these are documented misdiagnoses in the published families, which makes them high-value KB content):

Screening

• Newborn screening: not included in any program; no biochemical marker exists to enable it.
• Carrier screening: not offered. Mechanistically justifiable in Finnish/Northern Norwegian populations for c.1320+1G>A given AF 0.0007, but no such program or recommendation is published — flag as inference, not practice.
• Cascade testing: appropriate for at-risk relatives once a familial variant is known; ACMG-standard.
• Prenatal/PGT: technically available for known familial variants (§13).

11. Outcome / Prognosis

Survival and mortality. No survival statistics, life-expectancy estimate, or mortality rate exists — n is far too small. Documented outcomes at the extremes: recurrent intrauterine fetal death in homozygous fetuses (the most severe reported outcome), and death in the early sixties from aspiration-related asphyxiation in the oldest reported patient after progressive motor/cognitive decline [FULL-TEXT — RE-VERIFY]. Most reported individuals were alive at report, aged 11–33. Preliminary read: survival into at least the sixth decade is possible, with heart failure, aspiration, perioperative decompensation, and refractory neonatal seizures as the identified threats.

Morbidity and function. Lifelong disability across motor, cognitive, and stamina domains. Concrete functional data: reduced 6-minute walk distance in 6/6; miniBESTest balance scores 14–21/28 in 6/6; supported or sheltered living with daily supervision in the adults; sheltered employment limited by exhaustion. Global proximal muscle atrophy with inability to walk in the advanced case.

Quality-of-life measures. None applied. No EQ-5D, SF-36, PROMIS, or disease-specific instrument has been used in AIMS. Explicit gap.

Complications. Heart failure / early dilated cardiomyopathy; refractory seizures including infantile spasms; obstructive sleep apnea; scoliosis and contractures; aspiration; feeding failure requiring gastrostomy; anxiety requiring pharmacotherapy; perioperative cardiogenic shock; progressive vascular leukoencephalopathy; sensorineural hearing loss.

Recovery potential. No recovery of the neurodevelopmental deficit. Acute decompensations can resolve substantially — the coma/tetraplegia episode resolved with return to baseline psychomotor level and near-complete lesion resolution, and the neonate's cardiac function normalized with VSD closure by 2 months.

Prognostic factors. Provisional, low-confidence given n: older age (cardiac dysfunction and cognitive decline both age-associated); surgical/anesthetic exposure; neonatal presentation (the most acutely unstable course); presence of seizures. No molecular genotype–phenotype correlation is established — the two unrelated c.3269delG homozygotes with divergent severity argue directly against simple variant-based prognostication, and all variants tested are complete LoF regardless of position.

Prognostic biomarkers. None. NT-proBNP tracks heart failure only.

12. Treatment

There is no disease-specific, curative, or channel-directed therapy for AIMS. Management is entirely supportive and surveillance-based. The most valuable content in this section is the drugs-to-avoid list, which is unusually well-evidenced for a 20-patient disease.

Contraindicated / high-caution agents — the actionable core

Reported general principle: loss of SUR2-dependent K_ATP channels is expected to alter cardiac resilience to stressors and may have complex, unpredictable effects on cardiac-acting drugs [FULL-TEXT — RE-VERIFY]. The 2026 authors call for systematic screening of antiseizure drugs for cardiovascular toxicity in SUR2-deficient animal models.

Contrast worth curating explicitly. Cantú syndrome's GeneReviews "agents to avoid" list is minoxidil, diazoxide, and ACE inhibitors — i.e. K_ATP openers (PMID:25275207) [ABSTRACT-VERIFIED]. AIMS' avoid-list is calcium channel blockers and Ca-channel-blocking anesthetics/antiseizure drugs. The two diseases have opposite pharmacological hazards, exactly as their opposite channel defects predict. This is a compelling, non-obvious pairing for the KB.

Supportive and symptomatic management, with NCIT terms

Also documented: cyclosporine for severe atopic eczema (one patient), steroids given during the ADEM-like episode, and hearing aids implied by bilateral high-frequency SNHL (note: there is no reliable NCIT clinical-action term for hearing-aid use, per CLAUDE.md).

Advanced therapeutics

• Gene therapy, gene editing, cell therapy, RNA therapy, targeted therapy, immunotherapy: none developed, none in trial. Conceptually, an autosomal-recessive LoF channelopathy is a rational gene-replacement target, but ABCC9 is a large gene, the relevant tissues (skeletal muscle, vascular smooth muscle, brain vasculature, heart) are broadly distributed, and no preclinical program is published.
• Pharmacogenomics: no PharmGKB/CPIC entry for ABCC9 in AIMS. The clinically relevant pharmacology is the mechanism-based hazard above, not metabolizer status. (Incidentally, the 2026 neonate also had G6PD deficiency, which carries its own independent drug-avoidance list — a confounder in that case, not an AIMS feature.)

Clinical trials

No clinical trial has ever been registered for AIMS. A ClinicalTrials.gov search yields no AIMS/IDMYS interventional or observational study. Leave clinical_trials: empty.

A mechanistic caution for curators

It is tempting to reason that if AIMS is loss of K_ATP function, a K_ATP opener (diazoxide, pinacidil, minoxidil, nicorandil) should help. The functional data argue this cannot work: SUR2Δ8 and the truncating variants produce channels with no measurable current at all, and pinacidil activation was tested and required WT SUR2 to be present. A pharmacological opener needs a functional channel to open. I state this explicitly because it is an inviting inference that the literature does not support — it should not be curated as a candidate therapy, and if recorded at all it belongs in a discussions KNOWLEDGE_GAP entry, clearly labeled as unsupported reasoning rather than as a treatment.

13. Prevention

Primary prevention. Not possible for the genotype. Prevention is reproductive, via genetic counseling, carrier testing of at-risk relatives, prenatal diagnosis, and preimplantation genetic testing for known familial variants. The recurrent IUFDs in heterozygous couples make reproductive counseling unusually consequential here — couples may present with pregnancy loss before any living affected child is diagnosed.

Secondary prevention (early detection). No newborn or population screening exists. Realistic secondary prevention is early molecular diagnosis via exome/genome sequencing in a child with the triad, which unlocks the drug-avoidance list and cardiac surveillance before harm occurs.

Tertiary prevention (preventing complications) — the most actionable level, and where the entry should concentrate: 1. Perioperative planning — anesthetic and antiseizure drug review before any procedure; avoid propofol; monitor cardiac function and lactate post-operatively. 2. Avoid calcium channel blockers and Ca-channel-blocking antiseizure drugs; prefer levetiracetam. 3. Longitudinal cardiac surveillance — periodic echocardiography (consider cardiac MRI, which detected dilation echo missed) with NT-proBNP, continuing into adulthood even after normal childhood studies. 4. Serial brain MRI to track lesion accumulation. 5. Audiological monitoring for progressive high-frequency SNHL. 6. Orthopedic surveillance for scoliosis and contractures; physiotherapy to preserve function. 7. Weight management — the low-weight-to-overweight trajectory contributed to OSA. 8. Aspiration risk assessment in advanced disease (cause of death in the oldest patient). 9. Activity pacing — exertion triggers spasms and myalgia; graded rather than maximal exercise (note the mouse data showing muscle histopathology emerges specifically with chronic heavy exercise).

Immunization, public health, environmental, and prophylactic interventions: not applicable — no infectious, environmental, or population-level dimension to this disease.

14. Other Species / Natural Disease

Taxonomy of species with relevant ABCC9 biology:

Naturally occurring disease in another species — a genuine find for this entry. OMIA:002710-9615, "Cardiomyopathy, dilated, ABCC9-related" in Canis lupus familiaris. An ABCC9 p.R1186Q variant, homozygous in all affected dogs, underlies sudden cardiac death / dilated cardiomyopathy (SCDY/DCM) in Manchester Terriers, with death typically before 2 years of age. VBO term: Manchester Terrier (look up the VBO identifier before curating).

This is comparatively informative in a specific way: the canine disease is a recessive, cardiac-predominant ABCC9 disorder — mechanistically adjacent to AIMS' cardiac arm and to human DCM1O, but without a reported neurodevelopmental or myopathic phenotype. Whether that reflects true species divergence, a milder/different variant effect (missense vs. truncating), or simply the limits of assessing cognition in dogs is unresolved — a legitimate comparative-biology knowledge gap.

Comparative pathology and conservation. K_ATP subunit architecture and the SUR2/Kir6.x tissue pairing are deeply conserved across vertebrates, which is precisely why zebrafish and mouse SUR2 nulls both lose ventricular myocyte K_ATP current and both reproduce the muscle and cardiac phenotypes. The conservation breaks down for cognition: no SUR2-mutant animal recapitulates the intellectual disability (§6, §15).

Zoonotic potential / cross-species transmission: not applicable — a germline genetic disease.

15. Model Organisms

Mouse

SUR2-STOP (CRISPR/Cas9; the purpose-built AIMS model), Smeland et al. 2019 - Allele: c.3446_3450delACTTCinsGA → premature stop after K1148, p.Y1149Stop, in TM15. - Channel validation: functional K_ATP essentially absent in both ventricular myocytes and aortic smooth muscle cells on inside-out patch clamp (18 WT vs 10 mutant patches, p<0.0001) — confirming it reproduces the human channel defect even though the genetic lesion differs. - Recapitulates: fatigability (multiple-trial inverted screen — comparable first-trial performance, progressive decline across trials, significant genotype and genotype×trial/session effects; n=9 per group); reduced LV fractional shortening (p<0.01); increased LVIDd normalized to body length; increased blood pressure; increased LV mass. - Fails to recapitulate: cognition and anxiety. Negative across 1-hour locomotor activity, center/periphery distance, a four-test sensorimotor battery (ledge, platform, pole, inclined screen), Morris water maze cued and place trials and probe trials, and elevated plus maze. The authors' verbatim conclusion: "Collectively, the behavioral findings suggest that the SUR2-STOP mice do not exhibit marked deficits in learning and memory nor show any obvious anxiety-like behaviors."

Tissue-selective and CaV1.1 models, McClenaghan et al. 2023 (PMID:37154692) - Tissue-selective K_ATP suppression localized myopathy to skeletal muscle specifically. - Isolated muscle showed abnormal unstimulated force generation — a mechanistic correlate of painful spasms. - Verapamil caused premature death in AIMS mice. - A CaV1.1-nonpermeable genetic model failed to reverse pathology — falsifying the excess-Ca²⁺-influx-through-CaV1.1 hypothesis.

Pre-existing SUR2 alleles (MGI:1352630) — background/comparative: null alleles show spontaneous death from episodic coronary artery vasospasm, hypertension, growth retardation, lower serum glucose with enhanced insulin action; a homozygous exon 5 deletion causes cardiac mitochondrial defects, cardiomyopathy and early postnatal death. An earlier internal-deletion allele retains a mitochondria-limited short SUR2 form and shows impaired exercise capacity and myofiber damage.

Explicit model caveat from the authors: "The animal models used in this study do not recapitulate the genetic defect identified in the AIMS patients, but were chosen as the functional effects of the frameshift mutations introduced into SUR2-STOP mice and fish mirror the functional effect of the SUR2 Δ8 mutation." They call for knock-in models carrying the actual human AIMS variants — a clean proposed_experiments item.

Zebrafish

SUR2-STOP zebrafish (CRISPR/Cas9), Smeland 2019 + Efthymiou 2024 - Allele: c.2944_2957del13 → frameshift, stop after S984, p.S985Stop, in TM12. - ~4-fold reduction in abcc9 mRNA (qPCR), consistent with NMD; complete absence of functional K_ATP in ventricular myocytes. - Craniofacial: significantly decreased normalized interorbital distance — recapitulating human hypotelorism (4/6 patients). Notably the mouse did not show this, making the fish the better dysmorphology model. No other striking dysmorphism. - Motor: reduced total movement, reduced total swimming distance, reduced duration of high-speed movements, but similar total time moving (n=62 per genotype) — a fatigability-like rather than paralysis-like signature. Hatching (which requires muscle contraction) was normal. - Cardiac: fractional shortening reduced 29%, ejection fraction 25%, cardiac output 28%, with proportionally reduced stroke volume; unchanged end-diastolic and end-systolic volumes in larvae; markedly enlarged ventricle with abnormal morphology in 5/6 adults; no fibrosis (AFOG); normal tropomyosin myofiber structure; numerous TUNEL-positive cells in both chambers versus very few in WT. No cardiac abnormality in heterozygotes — matching healthy human carriers. - Vascular: increased cardinal vein red-cell velocity, concordant with the hypertension in SUR2-STOP mice. - Seizure susceptibility (2024): exaggerated motor response to pentylenetetrazole. > "In vivo studies of abcc9 loss-of-function in zebrafish revealed an exacerbated motor response to pentylenetetrazole, a pro-convulsive drug, consistent with impaired neurodevelopment associated with an increased seizure susceptibility." — PMID:38217872 [ABSTRACT-VERIFIED]

In vitro / cellular systems

• Cosm6 cells transiently transfected with Kir6.2 + FLAG-SUR2A/SUR2B (WT, Δ8, 1:1 mix): Western blot, ⁸⁶Rb⁺ efflux, inside-out patch clamp, ATP dose–response (IC₅₀), pinacidil activation.
• Landing-pad HEK293 with FLIPR Blue membrane-potential dye and oligomycin A / glibenclamide challenge — the 2026 platform, notable as a scalable variant-classification assay suitable for future ABCC9 VUS interpretation.
• Patient fibroblasts for cDNA/splicing analysis (exon 8 skipping).
• No iPSC, organoid, or myotube model of AIMS has been reported — a clear gap, especially given that the ID phenotype is the one thing animal models miss and a human neural or cerebrovascular model could address it.

Recapitulation summary

Model limitations to record. No model reproduces the two defining CNS features — intellectual disability and cerebral white matter abnormality. Combined with the fact that neuronal K_ATP is predominantly Kir6.2/SUR1 rather than SUR2, this makes the CNS arm of AIMS pathophysiology the least secure part of the mechanism and the highest-value research target. Per CLAUDE.md this is a HUMAN_MODEL_MISMATCH, not a KNOWLEDGE_GAP: the model evidence exists and is negative, so the open question is translational validity rather than absence of data.

Recommended dismech curation actions

1. Keep MONDO:0859224 — verified correct, with OMIM:619719 / DOID:0070600 / UMLS C5676904 as mappings. Add IDMYS to synonyms.
2. Keep hgnc:60 for ABCC9 — verified against genenames.org.
3. Fix the GO label if used: GO:0008282 canonical label is "inward rectifying potassium channel", not "ATP-sensitive potassium channel complex" (that is a synonym). Put the K_ATP phrasing in preferred_term.
4. Model the pathophysiology as one molecular trigger with three arms (skeletal muscle / cerebrovascular-white matter / cardiac), tagging biological_scale per node. Give the cerebrovascular arm a mechanistic_hypotheses entry with status: EMERGING.
5. Add a HUMAN_MODEL_MISMATCH discussion for the intellectual-disability/anxiety phenotype vs. the formally negative SUR2-STOP mouse behavioral battery, with the authors' proposed working-memory/fear-conditioning experiments as proposed_experiments.
6. Add a supports: REFUTE evidence item for the CaV1.1-Ca²⁺-influx sub-hypothesis (PMID:37154692) — a rare, valuable falsified-mechanism record.
7. Curate the drugs-to-avoid content prominently (verapamil/CCBs, phenytoin/fosphenytoin, propofol, lacosamide maintenance), and consider the opposite-hazard pairing with Cantú syndrome (K_ATP openers) as an explicit cross-reference.
8. Consider a kb/groupings/ entry for K_ATP channelopathies — AIMS/Cantú/DCM1O/neonatal diabetes/congenital hyperinsulinism, grouping_basis: [SHARED_MECHANISM, SHARED_GENE_FAMILY], with GoF-vs-LoF as the differentiating axis.
9. Use measure_type: CASES_IN_LITERATURE + prevalence_class: ULTRA_RARE; do not fabricate a rate.
10. Omit frequency: on the discordant phenotypes (corpus callosum, dysmorphism) per the repo's frequency-evidence guidelines; record raw counts in notes:.
11. Before committing: run just fetch-reference for any PMID not already cached, then just validate-references, just validate-terms, and just validate — and specifically re-verify every claim marked [FULL-TEXT — RE-VERIFY] above, since those came through a summarizing intermediary rather than from a cached abstract I read directly.

Primary sources

Databases consulted: OMIM (#619719, 601439), MONDO/OLS4 (MONDO:0859224), HPO annotation API (OMIM:619719), Disease Ontology (DOID:0070600), HGNC (HGNC:60), UniProt (O60706), Ensembl (ENSG00000069431), NCBI Gene (10060, 20928), MGI (MGI:1352630), OMIA (OMIA:002710-9615), ClinGen (search.clinicalgenome.org/kb/genes/HGNC:60), gnomAD (allele-level data via publications; gene constraint not retrieved), ClinicalTrials.gov (no AIMS studies), Orphanet (no AIMS code found).

Sources: - Smeland et al. 2019, Nat Commun (PMC6773855) - Efthymiou et al. 2024, Brain (PMC11068106) - Nagaraj et al. 2026, Neurology Genetics (PMC13262668) - McClenaghan et al. 2023, EMBO Mol Med (PMC10245035) - OMIM #619719 — Intellectual Disability and Myopathy Syndrome - OMIM #601439 — ABCC9 - MONDO:0859224 via EBI OLS4 - HPO annotations for OMIM:619719 - HGNC ABCC9 (genenames.org REST) - MGI:1352630 — mouse Abcc9 - OMIA:002710-9615 — ABCC9-related dilated cardiomyopathy in dog - ClinGen gene curation for ABCC9 (HGNC:60) - Disease Ontology DOID:0070600