Congenital Fibrosis of the Extraocular Muscles (CFEOM): Disease-Characteristics Report

Executive summary

Congenital fibrosis of the extraocular muscles (CFEOM) is a genetically heterogeneous congenital cranial dysinnervation disorder (CCDD). Despite its historical name, it is primarily a developmental neurologic disorder: abnormal specification or axon growth/guidance of ocular motor neurons produces deficient or aberrant innervation, followed secondarily by extraocular-muscle hypoplasia, contracture, and fibrosis. The defining manifestations are congenital, usually nonprogressive restrictive ophthalmoplegia, strabismus, and frequently ptosis. CFEOM1 is most often caused by dominant KIF21A variants, CFEOM2 by biallelic PHOX2A loss-of-function variants, and CFEOM3 by dominant TUBB3 variants; rarer CFEOM-like phenotypes involve TUBA1A and TUBB2B. Mechanisms converge on motor-neuron specification, neuronal microtubule dynamics, kinesin–microtubule interactions, and cranial-axon pathfinding. (fritzsch2023evolutionanddevelopment pages 16-18, whitman2021axonalgrowthabnormalities pages 6-8, puri2023tubb3andkif21a pages 20-21)

There is no disease-modifying pharmacotherapy. Current care consists of amblyopia prevention, refractive correction, ocular-surface protection, and individualized strabismus and ptosis surgery. Population prevalence, health-related quality-of-life scores, long-term prospective outcomes, and controlled treatment-response rates remain poorly characterized.

1. Disease information

Definition and classification

CFEOM is a Mendelian CCDD characterized by congenital restriction of eye movements, generally accompanied by incomitant strabismus and ptosis. Human autopsy, MRI, and model-organism data shifted the accepted interpretation from a primary extraocular-muscle disease to primary developmental dysinnervation with secondary muscle fibrosis. In genetically confirmed CFEOM1, human pathology showed absence of the superior division of cranial nerve III (CN III) and corresponding motor neurons; MRI showed profound hypoplasia of the superior rectus and levator palpebrae superioris and abnormalities of ocular motor nerves. (whitman2021axonalgrowthabnormalities pages 6-8)

Common names: congenital fibrosis of the extraocular muscles; CFEOM; congenital external ophthalmoplegia; congenital restrictive ophthalmoplegia; generalized fibrosis syndrome; congenital ophthalmoplegia; and, historically, congenital fibrosis syndrome. “CFEOM” should be preferred because congenital external ophthalmoplegia has broader differential diagnoses.

Disease-level versus patient-level evidence: This report synthesizes aggregated disease-level resources, cohorts, pedigrees, case series, neuroimaging, animal models, and biochemical experiments. It does not contain identifiable EHR-derived patient data. The 2022 Chinese study, for example, aggregated 122 affected individuals from 96 families, while the recent large oCCDD genomics study analyzed 467 unsolved pedigrees. (jia2022clinicalandgenetic pages 13-14, jurgens2025expandingthegenetics pages 38-41)

Identifiers

Identifiers should be represented at both umbrella and subtype levels because CFEOM is genetically heterogeneous.

• OMIM phenotype entries commonly used: CFEOM1, 135700; CFEOM2, 602078; CFEOM3A, 600638. Individual gene–disease records should also be linked for KIF21A, PHOX2A, and TUBB3.
• Orphanet: “Congenital fibrosis of the extraocular muscles” and subtype records are available through the Orphanet nomenclature portal: https://www.orpha.net/.
• MONDO: use the current MONDO record returned for “congenital fibrosis of extraocular muscles” in the release being ingested; subtype mappings should be preserved. A MONDO identifier was not independently recoverable from the retrieved primary-literature corpus and should therefore be database-validated rather than inferred.
• ICD-10-CM: no highly specific CFEOM code; coding usually uses congenital malformation of eye/ocular motor disorder, strabismus, ophthalmoplegia, or ptosis codes according to manifestation.
• ICD-11: use the applicable developmental anomaly/ocular-motility category, supplemented by an Orphanet or OMIM identifier where the implementation permits rare-disease extension codes.
• MeSH: no uniquely specific CFEOM descriptor was established from the retrieved literature; indexing commonly falls under ophthalmoplegia, strabismus, eye-movement disorders, and congenital abnormalities.

The following table provides the core genotype–phenotype structure.

Table: This table summarizes the main genetically defined CFEOM subtypes and rarer tubulin-associated presentations, highlighting inheritance, distinguishing ocular findings, systemic associations, and developmental mechanisms. It is useful as a compact knowledge-base scaffold when exact identifiers are uncertain or subtype boundaries overlap.

2. Etiology, risk, and protective factors

Causal factors

The principal cause is a germline pathogenic variant affecting ocular motor-neuron development:

• KIF21A: usually heterozygous missense variants causing autosomal-dominant CFEOM1 and occasionally CFEOM3-like or syndromic phenotypes. Variants cluster in motor/stalk regions and reduce normal autoinhibition, producing an altered or gain-of-function state. (fritzsch2023evolutionanddevelopment pages 16-18, puri2023tubb3andkif21a pages 20-21)
• PHOX2A: biallelic loss-of-function variants causing autosomal-recessive CFEOM2 through failure of oculomotor and trochlear motor-neuron specification. (fritzsch2023evolutionanddevelopment pages 16-18)
• TUBB3: heterozygous missense variants causing autosomal-dominant CFEOM3, ranging from isolated ocular disease to multisystem neurodevelopmental tubulinopathy. (fritzsch2023evolutionanddevelopment pages 16-18, puri2023tubb3andkif21a pages 20-21)
• TUBA1A and TUBB2B: rare dominant/de novo tubulinopathy-associated CFEOM phenotypes, sometimes with cortical malformations such as polymicrogyria. These should not be treated as equally common causes of classic isolated CFEOM. (jia2022clinicalandgenetic pages 13-14)

Genetic risk factors

A pathogenic familial allele, an affected parent, and consanguinity in recessive PHOX2A disease are the clinically relevant risk factors. Dominant cases may also arise de novo. Variable expressivity is particularly important in TUBB3-related and some KIF21A-related disease; absence of severe ophthalmoplegia in a parent does not automatically exclude familial transmission. (fritzsch2023evolutionanddevelopment pages 16-18, fritzsch2023evolutionanddevelopment pages 14-16)

A common allele is not evidence of causality: the KIF21A 3′-UTR deletion c.690del has a reported gnomAD allele frequency of 0.1377*, including 156 homozygotes, and is therefore incompatible with being a highly penetrant cause of rare dominant CFEOM. The frameshift c.4602_4606del, p.(Thr1535GlnfsTer3), was reported as a VUS rather than an established pathogenic allele. (puri2023tubb3andkif21a pages 20-21)

Environmental, lifestyle, infectious, and protective factors

No reproducible toxin, infection, radiation exposure, diet, smoking behavior, occupation, or other environmental exposure is established as a cause or modifier of genetically defined CFEOM. No validated protective allele, diet, medication, or lifestyle intervention prevents the developmental dysinnervation. Accordingly, conventional gene–environment interaction models are not currently supported. These are evidence gaps, not proof that modifiers cannot exist.

3. Phenotypes

Core ocular phenotype

CFEOM1 classically presents with bilateral ptosis, eyes fixed below the horizontal midline, absent vertical movement, and variably limited horizontal movement. CFEOM2 combines bilateral ptosis, exotropia, severe movement restriction, and absent CN III/CN IV on MRI. CFEOM3 is more variable and often asymmetric, with variable ptosis and limited upgaze. (fritzsch2023evolutionanddevelopment pages 16-18, fritzsch2023evolutionanddevelopment pages 14-16)

Syndromic manifestations

Variant-specific TUBB3 disease may add facial weakness, additional cranial neuropathies, peripheral neuropathy, developmental delay, intellectual disability, corpus-callosal or white-matter abnormalities, and other brain malformations. TUBA1A/TUBB2B disease can include cortical malformations. These findings are not obligatory in classic isolated CFEOM and should trigger broader neurologic evaluation. (fritzsch2023evolutionanddevelopment pages 16-18, jia2022clinicalandgenetic pages 13-14)

In the 2022 Chinese CCDD cohort, 46/96 families (47.9%) had multiple congenital malformations. Among 88 families with high-resolution MRI, 15/88 (17.0%) had additional craniocerebral malformations. These percentages concern a mixed CCDD cohort and must not be presented as CFEOM-specific population frequencies. (jia2022clinicalandgenetic pages 13-14)

Quality of life

No robust CFEOM-specific EQ-5D, SF-36, PROMIS, or utility-weight dataset was found. Likely burdens include restricted field of binocular single vision, abnormal head posture, cosmetic/social effects of ptosis and strabismus, repeated surgery, amblyopia risk, and—where syndromic—neurologic disability. These impacts are clinically credible but lack disease-specific population estimates.

4. Genetic and molecular information

Principal genes and variant classes

• KIF21A — dominant missense variants are the canonical mechanism. The recurrent p.Arg954Trp allele remains a representative pathogenic variant. Most disease alleles alter motor/stalk-domain autoinhibition rather than simply abolishing protein production. (whitman2021axonalgrowthabnormalities pages 6-8, jurgens2025expandingthegenetics pages 36-37)
• PHOX2A — recessive nonsense, frameshift, splice, or damaging missense variants causing loss of function.
• TUBB3 — dominant missense variants; genotype strongly influences whether disease remains ocular or includes broader cranial/peripheral nerve and cerebral involvement. Reported variants with syndromic phenotypes include p.Arg262His and p.Arg380Cys; the Chinese cohort also associated p.Glu410Lys with syndromic findings. (jia2022clinicalandgenetic pages 13-14)
• TUBA1A/TUBB2B — predominantly heterozygous missense alleles with combined CFEOM and malformation-of-cortical-development phenotypes. (jia2022clinicalandgenetic pages 13-14)

All established CFEOM variants are germline. Somatic mosaicism is not a recognized principal mechanism, although low-level parental germline or somatic mosaicism may theoretically explain recurrence after an apparently de novo case. Population databases should be checked using the exact transcript and genome build; highly penetrant causal alleles are expected to be absent or exceptionally rare.

Diagnostic yields and recent genomics

In the 2022 Chinese cohort, WES identified ten pathogenic variants in KIF21A, TUBB3, and CHN1 across 43 families; 42 of the 43 genetically solved probands had CFEOM. Novel reported variants included KIF21A c.1064T>C, p.Phe355Ser; TUBB3 c.232T>A, p.Ser78Thr; and CHN1 c.650A>G, p.His217Arg. The authors concluded that “KIF21A and TUBB3 were the common pathogenic genes in Chinese CFEOM” and that MRI plus WES supported diagnosis. (jia2022clinicalandgenetic pages 13-14)

A later analysis of 467 previously unsolved oCCDD pedigrees, including 198 CFEOM probands, found pathogenic/likely pathogenic variants in 43/467 (9.2%) and prioritized VUS in another 70/467 (15.0%). Candidate findings extended beyond established genes to MYH10, KIF21B, TUBB6, TUBA4A, KIF5C, and others, but these newer gene associations require independent replication and functional validation before routine designation as definitive CFEOM genes. (jurgens2025expandingthegenetics pages 8-12, jurgens2025expandingthegenetics pages 38-41)

Modifier genes, epigenetics, and chromosomal abnormalities

No replicated CFEOM modifier gene or disease-specific epigenetic signature is established. Structural variants and chromosomal disruptions can produce oCCDD phenocopies or syndromic disease, so genome sequencing or chromosomal microarray is appropriate in unresolved syndromic cases. The recent 467-pedigree study explicitly integrated structural-variant analysis and found extensive heterogeneity. (jurgens2025expandingthegenetics pages 8-12, jurgens2025expandingthegenetics pages 38-41)

5. Environmental information

CFEOM is not known to be infectious, toxic, occupational, nutritional, radiation-induced, inflammatory, or lifestyle-mediated. Smoking, alcohol, diet, and exercise do not have established effects on disease occurrence. Environmental interventions cannot reverse embryonic cranial dysinnervation, although ordinary eye safety, ocular-surface care, and adherence to amblyopia treatment can reduce secondary morbidity.

6. Mechanism and pathophysiology

Causal chain

1. Upstream genetic lesion: pathogenic variant in PHOX2A, KIF21A, TUBB3, or another microtubule/axon-development gene.
2. Developmental cellular defect: failure of ocular motor-neuron specification (PHOX2A) or abnormal microtubule dynamics, motor-protein regulation, axon elongation, and guidance (KIF21A/TUBB3).
3. Neuroanatomic consequence: absent, hypoplastic, stalled, or misrouted CN III/IV/VI axons and abnormal innervation of extraocular muscles.
4. Secondary tissue consequence: denervation-related extraocular-muscle hypoplasia, contracture, and fibrosis.
5. Clinical phenotype: congenital ophthalmoplegia, fixed strabismus, ptosis, compensatory head posture, and amblyopia risk. (fritzsch2023evolutionanddevelopment pages 16-18, whitman2021axonalgrowthabnormalities pages 6-8)

KIF21A mechanism

KIF21A is an anterograde kinesin. Normally, interaction between the motor domain and the third coiled-coil stalk domain maintains a closed, autoinhibited state. CFEOM-associated missense variants attenuate this autoinhibition, increase microtubule association, and dysregulate cortical microtubule growth. In knock-in mice, superior-division CN III axons stall proximally in bulb-like enlargements containing abnormal growth cones and degenerating axons; distal nerve and target muscles are hypoplastic. (whitman2021axonalgrowthabnormalities pages 6-8, puri2023tubb3andkif21a pages 20-21)

PHOX2A mechanism

PHOX2A is a transcription factor required for development/specification of oculomotor and trochlear motor neurons. Loss of function prevents proper formation of these nuclei, explaining the MRI absence of CN III and CN IV and the profound CFEOM2 phenotype. (fritzsch2023evolutionanddevelopment pages 16-18)

TUBB3 mechanism

TUBB3 encodes neuron-enriched βIII-tubulin. Pathogenic missense variants alter microtubule behavior and interactions with kinesin motors, producing variant-specific errors in cranial axon growth, guidance, maintenance, and—in some alleles—cortical neuronal migration. This explains the continuum from isolated CFEOM3 to multisystem tubulinopathy. (fritzsch2023evolutionanddevelopment pages 16-18, puri2023tubb3andkif21a pages 20-21)

Suggested ontology annotations

• GO biological process: axon guidance (GO:0007411); neuron projection development (GO:0031175); cranial nerve development; microtubule-based movement (GO:0007018); microtubule polymerization/depolymerization; motor-neuron differentiation; neuron migration.
• GO cellular component: microtubule (GO:0005874); neuronal growth cone (GO:0030426); axon (GO:0030424); kinesin complex (GO:0005871); cytoplasm and cytoskeleton.
• Cell Ontology: motor neuron (CL:0000100); cranial motor neuron where available; skeletal muscle cell/myocyte (CL:0000188); extraocular-muscle fiber as the most specific supported term.

There is no established primary metabolic, immune, inflammatory, apoptotic, or oxidative-stress pathway. The “fibrosis” is downstream of dysinnervation rather than evidence of a systemic fibrosing disorder.

Molecular profiling and advanced technologies

No validated diagnostic transcriptomic, proteomic, metabolomic, lipidomic, single-cell, spatial-transcriptomic, or multi-omic CFEOM signature was found. Current molecular evidence is dominated by pedigree sequencing, structural/biochemical assays, neuroimaging, and engineered animal models. The large recent genomics study demonstrates the value of combined exome/genome and structural-variant analysis but also shows that most previously unsolved pedigrees remain without a definitive molecular diagnosis. (jurgens2025expandingthegenetics pages 8-12, jurgens2025expandingthegenetics pages 38-41)

7. Anatomical structures affected

Primary nervous-system structures: oculomotor nucleus and nerve (CN III), especially its superior division in KIF21A-CFEOM1; trochlear nucleus/nerve (CN IV), particularly in PHOX2A disease; and variably abducens pathways (CN VI). The midbrain and rostral hindbrain are the critical developmental regions.

Primary orbital structures: superior, inferior, medial, and lateral rectus; superior and inferior oblique; and levator palpebrae superioris. In KIF21A-CFEOM1, superior rectus and levator hypoplasia are especially prominent. Human MRI of 14 genetically affected individuals from six families demonstrated muscle hypoplasia and motor-nerve abnormalities. (whitman2021axonalgrowthabnormalities pages 6-8)

Secondary structures: eyelids, visual pathways affected by amblyopia, and—depending on genotype—corpus callosum, cerebral white matter, cortex, basal ganglia, additional cranial nerves, and peripheral nerves.

Suggested UBERON terms: eye (UBERON:0000970); extraocular muscle (use the current specific UBERON EOM record); oculomotor nerve (UBERON:0001643); trochlear nerve; abducens nerve; midbrain (UBERON:0001891); hindbrain (UBERON:0002028); upper eyelid; superior rectus muscle; levator palpebrae superioris.

Disease is typically bilateral in CFEOM1/2, whereas CFEOM3 may be unilateral, bilateral, or markedly asymmetric. (fritzsch2023evolutionanddevelopment pages 16-18)

8. Temporal development

The initiating defect occurs during embryonic ocular motor-neuron development. Clinical signs are present at birth or recognized in early infancy. The dysinnervation is nonprogressive, but secondary consequences evolve: amblyopia develops during the sensitive period of visual maturation; abnormal head posture and contractures may become more apparent with growth; and surgical alignment can drift or require revision. There are no defined early/intermediate/end-stage categories, remissions, or relapsing episodes.

The principal intervention window is early childhood: clear the visual axis, correct refractive error, treat amblyopia, and establish the most functional head position possible. The underlying nerve-development defect does not spontaneously recover.

9. Inheritance and population

Inheritance

• CFEOM1/KIF21A: autosomal dominant; familial or de novo.
• CFEOM2/PHOX2A: autosomal recessive; recurrence risk is 25% for each pregnancy when both parents are carriers.
• CFEOM3/TUBB3: autosomal dominant; familial or de novo, with variable expressivity.
• TUBA1A/TUBB2B-associated phenotypes: usually dominant/de novo, although individual pedigrees require variant-specific assessment. (fritzsch2023evolutionanddevelopment pages 16-18, fritzsch2023evolutionanddevelopment pages 14-16)

Anticipation is not established. Penetrance is often high for classic KIF21A-CFEOM1 but is variant- and family-dependent; expressivity is particularly variable in CFEOM3. Germline mosaicism should be discussed after an apparently de novo result because recurrence risk is low but not zero.

Epidemiology

CFEOM is very rare, but no reliable population-based prevalence, incidence, carrier-frequency, sex-ratio, mortality, or geographic-distribution estimate was recovered. Published cohorts are referral- and ancestry-dependent and should not be used as population prevalence samples. The Chinese cohort demonstrates worldwide occurrence and genetic heterogeneity but does not establish higher risk in Chinese ancestry. (jia2022clinicalandgenetic pages 13-14)

No consistent sex bias is expected for autosomal disease. Founder effects may exist in individual PHOX2A families or populations, but no universal founder allele is established.

10. Diagnostics

Clinical assessment

Diagnosis begins with congenital onset, nonprogressive restrictive motility, ptosis, globe position, forced-duction findings, and family history. Examination should include visual acuity appropriate for age, cycloplegic refraction, amblyopia assessment, pupil examination, ocular alignment in multiple gaze positions, head posture, eyelid function, Bell phenomenon, corneal exposure, fundus/optic-nerve examination, and complete neurologic/dysmorphology review.

Imaging

Obtain thin-section, high-resolution MRI of the brainstem, cranial nerves, and orbits when feasible. MRI can identify absent/hypoplastic or misdirected ocular motor nerves, extraocular-muscle hypoplasia, and syndromic cerebral abnormalities. In one broad Chinese CCDD cohort, all MRI-assessed patients except those with horizontal-gaze-palsy/progressive-scoliosis had cranial-nerve hypoplasia; MRI plus WES was judged diagnostically supportive. (jia2022clinicalandgenetic pages 13-14)

Genetic-testing algorithm

1. Phenotype-directed multigene panel covering at least KIF21A, PHOX2A, TUBB3, TUBA1A, and TUBB2B, with relevant oCCDD differential genes such as CHN1, MAFB, SALL4, HOXA1, ROBO3, ACKR3, and ECEL1.
2. If a classic familial phenotype strongly suggests one gene, targeted sequencing may be efficient—KIF21A for classic CFEOM1 and PHOX2A for classic recessive CFEOM2—but a panel generally better addresses overlap.
3. Trio WES for negative or syndromic cases; the Chinese cohort used WES followed by Sanger validation and segregation/de novo analysis. (jia2022clinicalandgenetic pages 13-14)
4. WGS with CNV/structural-variant analysis for persistently unsolved cases, especially when syndromic. A large recent study pre-screened 403/467 probands for known oCCDD genes and then applied exome/genome sequencing to unresolved pedigrees. (jurgens2025expandingthegenetics pages 38-41)
5. Chromosomal microarray where developmental delay, multiple congenital anomalies, or a chromosomal syndrome is suspected.

Karyotyping/FISH are not first-line unless a specific rearrangement is suspected. Mitochondrial DNA and repeat-expansion testing are not routine CFEOM tests. RNA sequencing may help resolve splice variants but is not an established clinical standard. A VUS must not direct irreversible treatment or predictive testing without additional evidence.

Differential diagnosis

Important alternatives include Duane retraction syndrome, Möbius syndrome, isolated congenital CN III/IV/VI palsy, congenital myasthenic syndrome, congenital myopathy, mitochondrial external ophthalmoplegia, MYF5-related external ophthalmoplegia with rib/vertebral anomalies, HOXA1/SALL4/ROBO3-related CCDD, orbital fibrosis, thyroid eye disease, congenital ptosis without ophthalmoplegia, and mechanical restrictive strabismus. Congenital stability, characteristic nerve/MRI anatomy, associated anomalies, and molecular testing distinguish these entities.

Screening

CFEOM is not included in routine newborn biochemical screening. Appropriate strategies are clinical newborn/infant eye examination in known families, cascade testing of relatives after identification of a pathogenic variant, and prenatal or preimplantation genetic testing when the familial variant is known.

11. Outcome and prognosis

CFEOM itself is not expected to shorten life in isolated disease; no disease-specific survival or mortality statistics exist. Morbidity is primarily visual and functional. Without timely care, ptosis, strabismus, and anisometropia can produce irreversible amblyopia. Persistent ophthalmoplegia and limited binocular visual fields remain lifelong even after successful alignment surgery.

Prognosis depends on genotype, baseline visual acuity, amblyopia, severity and symmetry of restriction, head posture, Bell phenomenon/corneal protection, and syndromic neurologic involvement. Surgery can improve primary-position alignment, head posture, eyelid position, and appearance, but does not restore normal innervation or full motility. Controlled long-term response rates and validated CFEOM-specific prognostic biomarkers are unavailable.

12. Treatment

Current clinical strategy

1. Protect vision early: cycloplegic refraction, glasses, occlusion or atropine penalization for amblyopia when indicated, and management of corneal exposure.
2. Characterize restriction: repeated motility measurements, head posture, eyelid function, and forced ductions when surgery is planned.
3. Strabismus surgery: individualized recession of tight muscles, often large inferior-rectus recession for marked infraduction; horizontal rectus surgery, transposition procedures, adjustable sutures, or periosteal fixation may be considered by experienced surgeons. Multiple procedures are often necessary.
4. Ptosis surgery: levator or frontalis-sling procedures, timed cautiously because poor Bell phenomenon and limited upgaze increase exposure-keratopathy risk.
5. Neurologic/developmental care: indicated for TUBB3/TUBA1A/TUBB2B syndromic disease.

Suggested NCIt intervention mappings: strabismus surgery; extraocular-muscle recession; extraocular-muscle resection; tendon transposition; ptosis repair; frontalis suspension; amblyopia therapy; corrective-lens therapy. Exact NCIt codes should be validated against the current thesaurus release.

There is no approved CFEOM-specific drug, pharmacogenomic algorithm, gene therapy, cell therapy, ASO, siRNA, immune therapy, or CRISPR treatment. Experimental work showing correction of mutant tubulin–kinesin interaction in a mouse/biochemical system is mechanistic proof of principle, not a clinically available therapy. The broader mechanistic literature shows that altered TUBB3–kinesin interaction can be experimentally rescued, supporting future target discovery. (puri2023tubb3andkif21a pages 20-21)

A recruiting observational study, NCT03059420, “Genetic Studies of Strabismus, Congenital Cranial Dysinnervation Disorders (CCDDs), and Their Associated Anomalies,” is designed for genetic/phenotypic discovery rather than therapeutic efficacy: https://clinicaltrials.gov/study/NCT03059420.

13. Prevention

Primary lifestyle or vaccine prevention is not applicable to a congenital Mendelian dysinnervation disorder. Reproductive prevention options include genetic counseling, carrier testing for relatives in PHOX2A families, cascade testing in dominant families, prenatal diagnosis, and preimplantation genetic testing for a known familial pathogenic variant.

Secondary prevention consists of early ophthalmologic detection and prevention of amblyopia or corneal exposure. Tertiary prevention includes optimized alignment/head posture, low-vision or educational support where necessary, and surveillance for neurologic complications in syndromic tubulinopathies. Population screening is not justified by current prevalence and intervention evidence.

14. Other species and natural disease

No well-validated naturally occurring veterinary disease that is genetically and phenotypically equivalent to human CFEOM was identified. There is no zoonotic transmission or cross-species infectious susceptibility. The relevant genes and ocular motor-development programs are evolutionarily conserved across vertebrates, enabling engineered mouse and zebrafish studies; conservation should not be confused with naturally occurring animal disease.

Suggested taxonomy identifiers for experimental work include Mus musculus, NCBI Taxon 10090, and Danio rerio, NCBI Taxon 7955.

15. Model organisms and experimental systems

KIF21A mouse models

Knock-in mice carrying the orthologous human CFEOM1 mutation reproduce ptosis/globe retraction and selective superior-division CN III pathology. Axons form proximal bulbs with enlarged growth cones, stall, and degenerate; distal nerves and superior rectus/levator targets become hypoplastic. These models strongly recapitulate developmental dysinnervation but do not reproduce every aspect of human visual behavior or surgical disease. (whitman2021axonalgrowthabnormalities pages 6-8)

PHOX2A mice

Loss-of-function models fail to specify or maintain oculomotor and trochlear motor-neuron populations, directly supporting the upstream transcription-factor mechanism of CFEOM2. (fritzsch2023evolutionanddevelopment pages 16-18)

TUBB3 systems

TUBB3 knock-in mice and recombinant tubulin assays model variant-specific microtubule and kinesin defects. In vitro work shows that disease-associated substitutions can impair kinesin motility/ATPase function, while engineered compensatory kinesin changes can restore axonal growth in experimental systems. These are powerful mechanistic models but do not establish safety or feasibility of analogous human treatment. (puri2023tubb3andkif21a pages 20-21)

Zebrafish and other vertebrate systems are useful for rapid analysis of cranial-axon pathfinding and conserved ocular motor circuitry, but the best directly disease-relevant evidence presently comes from mouse knock-in and recombinant microtubule–motor assays.

Recent developments and expert interpretation

• August 2023: Puri, Barry, and Engle synthesized evidence that TUBB3 and KIF21A variants alter microtubule dynamics and microtubule–kinesin interactions. Their abstract states that neuronal migration and axon guidance require “precise control of microtubule dynamics and microtubule-based cargo transport,” succinctly capturing the current mechanistic model. DOI: https://doi.org/10.3389/fnins.2023.1226181. (puri2023tubb3andkif21a pages 20-21)
• March 2024: a four-generation Chinese family study reported genetic investigation of CFEOM with keratoconus (DOI: https://doi.org/10.1016/j.heliyon.2024.e28036). This is relevant to phenotypic expansion but does not establish keratoconus as a general CFEOM feature.
• Online 2024 / print July 2025: the 467-pedigree oCCDD study found pathogenic/likely pathogenic variants in only 9.2% of previously unsolved probands and VUS in 15.0%, emphasizing substantial remaining genetic heterogeneity. It nominated multiple candidate genes, but expert interpretation should remain conservative until replication and functional confirmation. DOI: https://doi.org/10.1016/j.gim.2024.101216. (jurgens2025expandingthegenetics pages 8-12, jurgens2025expandingthegenetics pages 38-41)

Evidence limitations

The strongest evidence consists of human pedigrees, genotype–phenotype cohorts, MRI/pathology, knock-in mice, and in-vitro microtubule–kinesin assays. Major limitations are referral bias, small subtype-specific cohorts, inconsistent historical classification, lack of population registries, sparse standardized surgical outcomes, and virtually no disease-specific quality-of-life or prospective natural-history data. Candidate-gene findings from unsolved oCCDD cohorts must not be conflated with definitively validated CFEOM genes. No claim of environmental protection, pharmacologic efficacy, or advanced-omics biomarker is currently justified.

References

1. (fritzsch2023evolutionanddevelopment pages 16-18): Bernd Fritzsch. Evolution and development of extra-ocular nerves and muscles in vertebrates. Unknown journal, Jun 2023. URL: https://doi.org/10.20944/preprints202306.0416.v1, doi:10.20944/preprints202306.0416.v1.

2. (whitman2021axonalgrowthabnormalities pages 6-8): Mary C. Whitman. Axonal growth abnormalities underlying ocular cranial nerve disorders. Sep 2021. URL: https://doi.org/10.1146/annurev-vision-093019-114307, doi:10.1146/annurev-vision-093019-114307. This article has 22 citations and is from a peer-reviewed journal.

3. (puri2023tubb3andkif21a pages 20-21): Dharmendra Puri, Brenda J. Barry, and Elizabeth C. Engle. Tubb3 and kif21a in neurodevelopment and disease. Frontiers in Neuroscience, Aug 2023. URL: https://doi.org/10.3389/fnins.2023.1226181, doi:10.3389/fnins.2023.1226181. This article has 43 citations and is from a peer-reviewed journal.

4. (jia2022clinicalandgenetic pages 13-14): Hongyan Jia, Qian Ma, Yi Liang, Dan Wang, Qinglin Chang, Bo Zhao, Zongrui Zhang, Jing Liang, Jing Song, Yidi Wang, Ranran Zhang, Zhanhan Tu, and Yonghong Jiao. Clinical and genetic characteristics of chinese patients with congenital cranial dysinnervation disorders. Orphanet Journal of Rare Diseases, Dec 2022. URL: https://doi.org/10.1186/s13023-022-02582-5, doi:10.1186/s13023-022-02582-5. This article has 12 citations and is from a peer-reviewed journal.

5. (jurgens2025expandingthegenetics pages 38-41): Julie A. Jurgens, Brenda J. Barry, Wai-Man Chan, Sarah E. Mackinnon, M. Whitman, Paola M. Matos Ruiz, Brandon M Pratt, E. England, Lynn Pais, G. Lemire, E. Groopman, Carmen Glaze, Kathryn A Russell, M. Singer-Berk, Silvio Alessandro Di Gioia, Arthur S. Lee, Caroline Andrews, Sherin Shaaban, Megan M Wirth, Sarah Bekele, Melissa Toffoloni, Victoria R Bradford, Emma E. Foster, Lindsay Berube, Cristina Rivera-Quiles, Fiona M. Mensching, Alba Sanchis-Juan, Jack M. Fu, Isaac Wong, Xuefang Zhao, M. Wilson, B. Weisburd, M. Lek, Hugo Abarca-Barriga, C. Al-Haddad, Jeffrey Berman, E. Bothun, J. Capasso, O. Chacón-Camacho, Lan-Yun Chang, Stephen P Christiansen, M. Ciccarelli, M. Cordonnier, G. F. Cox, Cynthia J. Curry, L. Dagi, Thomas Lee Dahm, Karen David, B. Davitt, T. de Berardinis, J. Demer, J. Desir, F. D’Esposito, A. Drack, Eric Eggenberger, J. Elder, A. Elliott, K. Epley, H. Feldman, Carlos R. Ferreira, Maree P. Flaherty, A. B. Fulton, C. Gerth-Kahlert, I. Gottlob, Stephen Grill, D. Halliday, F. Hanisch, Eleanor Hay, G. Heidary, C. Holder, Jonathan C. Horton, A. Iannaccone, Sherwin J. Isenberg, S. Johnston, A. Kahana, J. Katowitz, M. Kazlas, Natalie C Kerr, Virginia E. Kimonis, M. Ko, Feray Koç, D. Larsen, G. Lay-Son, D. Ledoux, Alex V Levin, Richard Levy, Christopher J. Lyons, D. Mackey, Adriano Magli, Iason S. Mantagos, Candice Marti, I. Maystadt, Fiona McKenzie, Manoj P Menezes, Claudia N. Mikail, David T. Miller, K. B. Miller, M. Mills, K. Miyana, H. U. Møller, L. Mullineaux, J. Nishimura, A. Noble, P. K. Pandey, Piero Pavone, Johann Penzien, R. Petersen, James A. Phalen, A. Poduri, C. R. Polo, L. Prasov, F. Ramos, Maria Ramos-Cáceres, Richard M. Robb, Béatrice Rossillion, Mustafa Sahin, Harvey S Singer, Lois E. H. Smith, J. A. Sorkin, J. Soul, S. Staffieri, Heather Stalker, S. Stasheff, Sonya Strassberg, Mitchell B. Strominger, D. Taranath, Ioan T. Thomas, Elias I. Traboulsi, M. C. Ugrin, Deborah K. Vanderveen, Andrea L. Vincent, Marlene C. Vogel G, B. Wabbels, A. Wong, C. Woods, Carolyn Wu, Edward Yang, A. Yeung, Terri L. Young, J. Zenteno, Alexandra A. Zubcov-Iwantscheff, Johan Zwaan, Harrison Brand, M. Talkowski, D. MacArthur, A. O’Donnell-Luria, C. Robson, David G. Hunter, and Elizabeth C. Engle. Expanding the genetics and phenotypes of ocular congenital cranial dysinnervation disorders. Genetics in medicine : official journal of the American College of Medical Genetics, 27:101216-101216, Jul 2025. URL: https://doi.org/10.1016/j.gim.2024.101216, doi:10.1016/j.gim.2024.101216. This article has 20 citations.

6. (fritzsch2023evolutionanddevelopment pages 14-16): Bernd Fritzsch. Evolution and development of extra-ocular nerves and muscles in vertebrates. Unknown journal, Jun 2023. URL: https://doi.org/10.20944/preprints202306.0416.v1, doi:10.20944/preprints202306.0416.v1.

7. (jurgens2025expandingthegenetics pages 36-37): Julie A. Jurgens, Brenda J. Barry, Wai-Man Chan, Sarah E. Mackinnon, M. Whitman, Paola M. Matos Ruiz, Brandon M Pratt, E. England, Lynn Pais, G. Lemire, E. Groopman, Carmen Glaze, Kathryn A Russell, M. Singer-Berk, Silvio Alessandro Di Gioia, Arthur S. Lee, Caroline Andrews, Sherin Shaaban, Megan M Wirth, Sarah Bekele, Melissa Toffoloni, Victoria R Bradford, Emma E. Foster, Lindsay Berube, Cristina Rivera-Quiles, Fiona M. Mensching, Alba Sanchis-Juan, Jack M. Fu, Isaac Wong, Xuefang Zhao, M. Wilson, B. Weisburd, M. Lek, Hugo Abarca-Barriga, C. Al-Haddad, Jeffrey Berman, E. Bothun, J. Capasso, O. Chacón-Camacho, Lan-Yun Chang, Stephen P Christiansen, M. Ciccarelli, M. Cordonnier, G. F. Cox, Cynthia J. Curry, L. Dagi, Thomas Lee Dahm, Karen David, B. Davitt, T. de Berardinis, J. Demer, J. Desir, F. D’Esposito, A. Drack, Eric Eggenberger, J. Elder, A. Elliott, K. Epley, H. Feldman, Carlos R. Ferreira, Maree P. Flaherty, A. B. Fulton, C. Gerth-Kahlert, I. Gottlob, Stephen Grill, D. Halliday, F. Hanisch, Eleanor Hay, G. Heidary, C. Holder, Jonathan C. Horton, A. Iannaccone, Sherwin J. Isenberg, S. Johnston, A. Kahana, J. Katowitz, M. Kazlas, Natalie C Kerr, Virginia E. Kimonis, M. Ko, Feray Koç, D. Larsen, G. Lay-Son, D. Ledoux, Alex V Levin, Richard Levy, Christopher J. Lyons, D. Mackey, Adriano Magli, Iason S. Mantagos, Candice Marti, I. Maystadt, Fiona McKenzie, Manoj P Menezes, Claudia N. Mikail, David T. Miller, K. B. Miller, M. Mills, K. Miyana, H. U. Møller, L. Mullineaux, J. Nishimura, A. Noble, P. K. Pandey, Piero Pavone, Johann Penzien, R. Petersen, James A. Phalen, A. Poduri, C. R. Polo, L. Prasov, F. Ramos, Maria Ramos-Cáceres, Richard M. Robb, Béatrice Rossillion, Mustafa Sahin, Harvey S Singer, Lois E. H. Smith, J. A. Sorkin, J. Soul, S. Staffieri, Heather Stalker, S. Stasheff, Sonya Strassberg, Mitchell B. Strominger, D. Taranath, Ioan T. Thomas, Elias I. Traboulsi, M. C. Ugrin, Deborah K. Vanderveen, Andrea L. Vincent, Marlene C. Vogel G, B. Wabbels, A. Wong, C. Woods, Carolyn Wu, Edward Yang, A. Yeung, Terri L. Young, J. Zenteno, Alexandra A. Zubcov-Iwantscheff, Johan Zwaan, Harrison Brand, M. Talkowski, D. MacArthur, A. O’Donnell-Luria, C. Robson, David G. Hunter, and Elizabeth C. Engle. Expanding the genetics and phenotypes of ocular congenital cranial dysinnervation disorders. Genetics in medicine : official journal of the American College of Medical Genetics, 27:101216-101216, Jul 2025. URL: https://doi.org/10.1016/j.gim.2024.101216, doi:10.1016/j.gim.2024.101216. This article has 20 citations.

8. (jurgens2025expandingthegenetics pages 8-12): Julie A. Jurgens, Brenda J. Barry, Wai-Man Chan, Sarah E. Mackinnon, M. Whitman, Paola M. Matos Ruiz, Brandon M Pratt, E. England, Lynn Pais, G. Lemire, E. Groopman, Carmen Glaze, Kathryn A Russell, M. Singer-Berk, Silvio Alessandro Di Gioia, Arthur S. Lee, Caroline Andrews, Sherin Shaaban, Megan M Wirth, Sarah Bekele, Melissa Toffoloni, Victoria R Bradford, Emma E. Foster, Lindsay Berube, Cristina Rivera-Quiles, Fiona M. Mensching, Alba Sanchis-Juan, Jack M. Fu, Isaac Wong, Xuefang Zhao, M. Wilson, B. Weisburd, M. Lek, Hugo Abarca-Barriga, C. Al-Haddad, Jeffrey Berman, E. Bothun, J. Capasso, O. Chacón-Camacho, Lan-Yun Chang, Stephen P Christiansen, M. Ciccarelli, M. Cordonnier, G. F. Cox, Cynthia J. Curry, L. Dagi, Thomas Lee Dahm, Karen David, B. Davitt, T. de Berardinis, J. Demer, J. Desir, F. D’Esposito, A. Drack, Eric Eggenberger, J. Elder, A. Elliott, K. Epley, H. Feldman, Carlos R. Ferreira, Maree P. Flaherty, A. B. Fulton, C. Gerth-Kahlert, I. Gottlob, Stephen Grill, D. Halliday, F. Hanisch, Eleanor Hay, G. Heidary, C. Holder, Jonathan C. Horton, A. Iannaccone, Sherwin J. Isenberg, S. Johnston, A. Kahana, J. Katowitz, M. Kazlas, Natalie C Kerr, Virginia E. Kimonis, M. Ko, Feray Koç, D. Larsen, G. Lay-Son, D. Ledoux, Alex V Levin, Richard Levy, Christopher J. Lyons, D. Mackey, Adriano Magli, Iason S. Mantagos, Candice Marti, I. Maystadt, Fiona McKenzie, Manoj P Menezes, Claudia N. Mikail, David T. Miller, K. B. Miller, M. Mills, K. Miyana, H. U. Møller, L. Mullineaux, J. Nishimura, A. Noble, P. K. Pandey, Piero Pavone, Johann Penzien, R. Petersen, James A. Phalen, A. Poduri, C. R. Polo, L. Prasov, F. Ramos, Maria Ramos-Cáceres, Richard M. Robb, Béatrice Rossillion, Mustafa Sahin, Harvey S Singer, Lois E. H. Smith, J. A. Sorkin, J. Soul, S. Staffieri, Heather Stalker, S. Stasheff, Sonya Strassberg, Mitchell B. Strominger, D. Taranath, Ioan T. Thomas, Elias I. Traboulsi, M. C. Ugrin, Deborah K. Vanderveen, Andrea L. Vincent, Marlene C. Vogel G, B. Wabbels, A. Wong, C. Woods, Carolyn Wu, Edward Yang, A. Yeung, Terri L. Young, J. Zenteno, Alexandra A. Zubcov-Iwantscheff, Johan Zwaan, Harrison Brand, M. Talkowski, D. MacArthur, A. O’Donnell-Luria, C. Robson, David G. Hunter, and Elizabeth C. Engle. Expanding the genetics and phenotypes of ocular congenital cranial dysinnervation disorders. Genetics in medicine : official journal of the American College of Medical Genetics, 27:101216-101216, Jul 2025. URL: https://doi.org/10.1016/j.gim.2024.101216, doi:10.1016/j.gim.2024.101216. This article has 20 citations.

Artifacts

• Edison artifact artifact-00

Reference Validation

Checked with linkml-reference-validator 0.2.1.

All extracted references resolved successfully.