Craniometadiaphyseal Osteosclerosis with Hip Dysplasia (CMDOH): A Comprehensive Disease Characterization

Disease: Craniometadiaphyseal Osteosclerosis with Hip Dysplasia (CMDOH) MONDO ID: MONDO:0957832 · OMIM: #620558 · Gene: AXIN1 (OMIM 603816; HGNC:903; chr16p13.3) Category:* Mendelian (autosomal recessive)

Summary

Craniometadiaphyseal osteosclerosis with hip dysplasia (CMDOH) is an ultra-rare, autosomal recessive sclerosing skeletal dysplasia caused by biallelic (homozygous, in the reported consanguineous families) C-terminally truncating variants in AXIN1. AXIN1 is the scaffold protein of the β-catenin destruction complex, the negative regulator of canonical Wnt signaling. The disease-causing variants truncate the C-terminal DIX domain, which mediates AXIN1 self-polymerization, thereby lowering AXIN1 protein levels and impairing destruction-complex assembly. The net functional consequence is enhanced basal canonical Wnt/β-catenin signaling — a "Wnt-gain" state — that shifts the balance of bone remodeling toward increased osteoblastic bone formation and reduced osteoclastic resorption, producing generalized osteosclerosis and hyperostosis.

The condition was defined in 2023 by Terhal et al. (PMID: 37582359), who described seven individuals from four families carrying three distinct homozygous truncating AXIN1 variants. The core phenotype comprises congenital/progressive macrocephaly, cranial hyperostosis (calvarial and skull-base sclerosis), and vertebral endplate sclerosis ("sandwich vertebrae"), together with frequent hip dysplasia, heart malformations, variable developmental delay, hematological anomalies, metaphyseal flaring, and dysmorphic facies. Bone biopsy demonstrated increased osteoblast activity and reduced osteoclast function at the growth-plate resorption zone, consistent with the Wnt-gain mechanism.

Mechanistically, CMDOH sits at the "hypomorphic" end of an AXIN1 dysfunction spectrum: germline biallelic DIX-domain truncations are partial loss-of-function (protein is reduced and polymerization is impaired but some Wnt-inhibitory function is retained), yielding a viable skeletal phenotype, whereas complete somatic loss of AXIN1 drives cancer (notably hepatocellular carcinoma) through full canonical Wnt activation. Crucially, the same study showed that a tankyrase inhibitor (XAV939) — which stabilizes AXIN1/AXIN2 — attenuates the Wnt overactivity in patient-derived and genome-edited cells, nominating a mechanistically rational (though not yet clinically tested) targeted therapy. Mouse Axin1 models corroborate the β-catenin–dependent, osteoblast-lineage mechanism. No disease-specific therapy currently exists; management is supportive and multidisciplinary.

Section 1 — Disease Information

Overview. CMDOH is a rare Mendelian sclerosing bone dysplasia combining features of two classic radiographic categories: osteopetrosis-like changes (sandwich vertebrae, metaphyseal flaring) and endosteal-hyperostosis-like changes (calvarial and cortical thickening). It is characterized by macrocephaly, cranial hyperostosis, vertebral endplate sclerosis, hip dysplasia, and additional extraskeletal involvement (cardiac, developmental, hematological). "Sclerosing skeletal dysplasias result from an imbalance between bone formation and resorption" (PMID: 37582359) — CMDOH exemplifies this, with the imbalance tilted toward net bone accrual.

Key identifiers.

Synonyms / alternative names. Craniometadiaphyseal osteosclerosis with hip dysplasia; CMDOH; AXIN1-related sclerosing skeletal dysplasia. The name captures the anatomic distribution (cranio- = skull; meta-/diaphyseal = metaphysis/diaphysis of long bones) plus the associated hip dysplasia.

Information source. Knowledge is derived from an aggregated case series — deep clinical, radiographic, and functional characterization of individual patients (7 individuals / 4 families) reported in a single landmark paper, plus mechanistic corroboration from cellular models and model organisms. It is not derived from large EHR/registry datasets, reflecting the disorder's extreme rarity.

Section 2 — Etiology

Primary cause (genetic). CMDOH is a monogenic autosomal recessive disorder caused by biallelic C-terminally truncating variants in AXIN1. In the reported families, affected individuals were homozygous, consistent with parental consanguinity. There is no known environmental, infectious, or acquired cause.

Genetic risk factors. The disease is fully genetically determined by the AXIN1 genotype. Consanguinity is the principal contextual risk factor, increasing the probability of homozygosity for a rare recessive allele. No modifier loci have been formally mapped; the paralog AXIN2 is a plausible biological modifier because it can partially compensate for AXIN1 (see Mechanism).

Environmental risk factors. None identified. Sex does not appear to influence occurrence (recessive Mendelian trait). Family history / consanguinity is the relevant contextual factor.

Protective factors. No genetic or environmental protective factors are established. Biologically, retained partial AXIN1 function and AXIN2 compensation are inferred to make the phenotype viable (compared with the lethality of complete Axin1 loss in mice), but this is a feature of the hypomorphic allele class rather than a modifiable protective factor.

Gene–environment interactions. None demonstrated. The disorder is essentially environment-independent.

Section 3 — Phenotypes

The phenotype spectrum below is drawn from the 7 patients / 4 families in Terhal et al. 2023 (PMID: 37582359). Frequencies are qualitative given the tiny cohort.

Onset and severity. Reported ages span 3 months to 15.5 years. Macrocephaly is congenital and progressive; cranial hyperostosis emerges and progresses from the first year of life. Severity is variable across families.

Quality-of-life impact. No formal QoL instruments (EQ-5D, SF-36, PROMIS) have been applied. Anticipated impacts, by analogy with other cranial hyperostoses, include potential cranial-nerve compression (from skull-base sclerosis), orthopedic morbidity from hip dysplasia (pain, mobility limitation), and developmental/educational impact where developmental delay is present. These are inferred rather than measured.

Section 4 — Genetic / Molecular Information

Causal gene. AXIN1 (axis inhibitor 1) — OMIM 603816, HGNC:903, chromosome 16p13.3*. AXIN1 is the central scaffold of the β-catenin destruction complex (with APC, GSK3β, CK1), which phosphorylates β-catenin to target it for degradation.

Pathogenic variants (Terhal et al. 2023). Three distinct C-terminal truncating variants across four families:

All variants disrupt the C-terminal DIX domain, which mediates AXIN1 homo-polymerization and destruction-complex assembly.

Variant classification. Per ACMG/AMP criteria these are pathogenic (loss-of-function truncating variants in a gene with an established gene–disease relationship, segregating with disease in multiple consanguineous families, supported by functional evidence).

Allele frequency. These are private/ultra-rare variants; not present at appreciable frequency in gnomAD. The recurrent p.Arg723* appearing in two families likely reflects an arginine CpG mutational hotspot rather than a broad founder effect.

Somatic vs germline. In CMDOH the variants are germline. By contrast, AXIN1 is recurrently altered somatically in cancer (see F003).

Functional consequence. Hypomorphic loss of function. The truncations reduce AXIN1 protein levels and impair DIX-domain–mediated polymerization, but overexpression studies showed partially retained Wnt-inhibitory function — explaining why the phenotype is a viable skeletal dysplasia rather than the embryonic lethality seen with complete Axin1 loss. Net pathway effect: enhanced basal canonical Wnt/β-catenin signaling (gain of Wnt output via loss of a negative regulator).

Modifier genes. AXIN2 (the paralog) is the strongest candidate modifier: XAV939 rescue of AXIN1-knockout cells implicates AXIN2 stabilization as a compensatory route (see F006).

Epigenetic / chromosomal. No disease-specific epigenetic signature or large-scale chromosomal abnormality has been reported for CMDOH.

Section 5 — Environmental Information

No environmental, lifestyle, or infectious contributors are implicated. CMDOH is a purely genetic, autosomal recessive Mendelian disorder. Not applicable: toxins/radiation/pollution; smoking/diet/exercise/alcohol; infectious agents.

Section 6 — Mechanism / Pathophysiology

Causal chain (initiating lesion → clinical manifestation)

1. Biallelic C-terminal truncating variant in AXIN1 (e.g., p.Arg723, p.Arg805, or a frameshift) → removes/disrupts the DIX domain.
2. Loss of the DIX domain impairs AXIN1 self-polymerization and lowers AXIN1 protein levels → results in a weakened, less abundant scaffold.
3. Reduced functional AXIN1 → destabilizes the β-catenin destruction complex (AXIN1–APC–GSK3β–CK1) → leads to reduced phosphorylation/degradation of β-catenin.
4. → Cytoplasmic β-catenin accumulates and translocates to the nucleus → results in enhanced basal canonical Wnt/β-catenin transcriptional output (demonstrated directly in patient and genome-edited cells).
5. In the osteoblast lineage, elevated β-catenin → increases OPG (osteoprotegerin) and the OPG:RANKL ratio (inferred from mouse Axin1^Osx^ model) → leads to suppressed osteoclastogenesis (fewer TRAP+/MMP9+/cathepsin-K+ osteoclasts).
6. Branch A — increased osteoblastic bone formation; Branch B — decreased osteoclastic bone resorption. Both branches converge to → net increase in bone mass / osteosclerosis and hyperostosis.
7. Anatomically distributed → cranial hyperostosis + macrocephaly, vertebral endplate ("sandwich") sclerosis, metaphyseal flaring; with associated developmental effects producing hip dysplasia, cardiac malformation, developmental delay, hematological anomalies (the extraskeletal links are consistent with the broad developmental role of Wnt signaling but are inferred, not individually demonstrated).
8. Therapeutic branch: Tankyrase inhibition (XAV939) raises AXIN1/AXIN2 levels → restores destruction-complex activity → attenuates Wnt overactivity in mutant cells (in vitro proof of concept).

Steps 1–4 and step 8 are experimentally demonstrated in the human/cell system; step 5 is inferred from mouse osteoblast-lineage models; step 7's extraskeletal links are inferred from Wnt biology.

Detail by category

• Molecular pathway: Canonical Wnt/β-catenin signaling (KEGG hsa04310; Reactome "Degradation of beta-catenin by the destruction complex," R-HSA-195253). AXIN1 is the rate-limiting scaffold of the destruction complex.
• Cellular processes: Osteoblast differentiation/activity (increased); osteoclastogenesis (suppressed via OPG/RANKL). GO:0060070 (canonical Wnt signaling pathway), GO:0001649 (osteoblast differentiation), GO:0030316 (osteoclast differentiation), GO:0045597/0045668 (regulation of osteoblast/osteoclast differentiation).
• Protein dysfunction: DIX-domain truncation → loss of polymerization, reduced protein stability → hypomorphic LOF of the destruction-complex scaffold.
• Biochemical abnormality: Failure to phosphorylate/degrade β-catenin; downstream Wnt target activation (e.g., c-Myc, CCND1 seen in AXIN1-loss cancer models, supporting pathway direction).
• Tissue remodeling mechanism: Dysregulated bone remodeling with osteoblast–osteoclast imbalance at the growth-plate resorption zone; coarse trabeculae on biopsy.
• Cell types (CL): osteoblast (CL:0000062), osteoclast (CL:0000092), osteoprogenitor/mesenchymal stromal cell (CL:0000134), osteocyte (CL:0000137).

Molecular profiling

• Direct functional readouts (Wnt reporter activity) in patient-derived and CRISPR-edited cells: enhanced basal canonical Wnt activity (PMID: 37582359).
• Mouse osteoblast-lineage models: increased β-catenin, increased OPG:RANKL, decreased osteoclast markers.
• No dedicated patient transcriptomic/proteomic/metabolomic/single-cell datasets are available for CMDOH.

Section 7 — Anatomical Structures Affected

Organ / system level (primary). Skeletal system — with a distinctive craniospinal and metadiaphyseal distribution: - Skull / calvarium & skull base (UBERON:0000209 cranial skeletal system; UBERON:0004339 calvaria) — hyperostosis, macrocephaly. - Vertebral column (UBERON:0001130) — endplate sclerosis ("sandwich vertebrae"). - Long bones — metaphysis/diaphysis (UBERON:0002515 metaphysis; UBERON:0004770 diaphysis) — flaring, sclerosis. - Hip joint / pelvis (UBERON:0001464 hip; UBERON:0001474 bone of the pelvis) — hip dysplasia.

Secondary organ involvement. Cardiovascular (heart malformations, UBERON:0000948); hematopoietic system (hematological anomalies, UBERON:0002390); central nervous system (developmental delay; potential cranial-nerve compromise from skull-base sclerosis).

Tissue / cell level. Bone (connective) tissue; growth-plate cartilage/resorption zone. Cell populations: osteoblasts (increased activity), osteoclasts (reduced number/function), osteoprogenitors.

Subcellular level. β-catenin destruction complex is cytoplasmic; downstream signaling acts in the nucleus (β-catenin/TCF transcription). GO cellular components: GO:0030877 (beta-catenin destruction complex), GO:0005737 (cytoplasm), GO:0005634 (nucleus).

Localization / lateralization. Skeletal involvement is generalized and bilateral/symmetric (calvarium, spine, long bones); hip dysplasia may be uni- or bilateral.

Section 8 — Temporal Development

• Onset: Congenital (macrocephaly present at/near birth); cranial hyperostosis develops and progresses from the first year of life. Onset pattern is chronic and insidious/progressive.
• Progression: Radiographic sclerosis is progressive through childhood (cases documented from 3 months to 15.5 years). Progression rate appears slow-to-moderate and variable between families.
• Disease course: Chronic, lifelong. No episodic or relapsing-remitting pattern.
• Remission: None spontaneous; no treatment-induced remission described (no disease-modifying therapy in clinical use).
• Critical periods: The first years of life (active bone modeling) represent the window in which hyperostosis emerges — mechanistically the most rational period for any future targeted (Wnt-normalizing) intervention.

Section 9 — Inheritance and Population

• Epidemiology: Ultra-rare. Prevalence/incidence are unknown and unquantified; total literature comprises 7 individuals from 4 families. No registry-based estimate exists.
• Inheritance: Autosomal recessive. Affected individuals were homozygous; parents are obligate heterozygous carriers.
• Penetrance: Appears complete in biallelic individuals (all reported homozygotes affected), though the sample is too small for precise estimation.
• Expressivity: Variable — e.g., variable developmental delay, cardiac involvement, and stature across families/individuals.
• Genetic anticipation: Not applicable (not a repeat-expansion disorder).
• Germline mosaicism: Not reported.
• Founder effect: No broad founder effect established; the recurrent p.Arg723* likely reflects a CpG mutational hotspot rather than a shared haplotype.
• Consanguinity: Central — homozygosity arose in consanguineous families.
• Carrier frequency: Unknown; expected to be very low given ultra-rarity.
• Population demographics: Reported families are consistent with consanguineous backgrounds; no established sex bias (recessive). Distribution is sporadic, not endemic.

Section 10 — Diagnostics

Imaging (cornerstone of diagnosis). Skeletal survey / radiographs demonstrating: - Calvarial and skull-base hyperostosis/sclerosis; macrocephaly. - Vertebral endplate sclerosis ("sandwich vertebrae"). - Metaphyseal flaring and long-bone sclerosis. - Hip dysplasia on pelvic imaging. CT is useful to characterize skull-base sclerosis and potential foraminal narrowing.

Laboratory / biomarkers. No specific diagnostic biomarker. Bone turnover markers are not established as diagnostic. Hematological anomalies warrant a CBC and hematologic workup. Bone biopsy (research setting) showed increased osteoblast and reduced osteoclast function with coarse trabeculae.

Genetic testing (confirmatory). The diagnosis is molecularly confirmed by identifying biallelic C-terminal truncating AXIN1 variants. - WES/WGS is the practical first-line approach given phenotypic overlap among sclerosing dysplasias. - Targeted AXIN1 sequencing or inclusion in a sclerosing bone dysplasia / skeletal dysplasia gene panel is appropriate once suspected. - CMA/karyotype/FISH/mtDNA/repeat testing are not indicated (single-gene, SNV/indel etiology).

Clinical criteria & differential diagnosis. No formal consensus criteria exist (recently delineated). Key differentials among sclerosing/craniotubular dysplasias:

Screening. In known families, cascade carrier testing and, where desired, prenatal/preimplantation genetic testing for the familial AXIN1 variants are options. No population newborn screening exists.

Section 11 — Outcome / Prognosis

• Survival/mortality: No systematic survival data. Not reported to be lethal in childhood; the oldest reported individual was 15.5 years. Life expectancy is undefined.
• Morbidity: Driven by skeletal complications — potential cranial-nerve compromise from skull-base hyperostosis, orthopedic morbidity from hip dysplasia, and functional impact of developmental delay; cardiac malformations and hematological anomalies contribute variably.
• Disease course: Chronic, progressive sclerosis. Recovery potential is limited; no disease-modifying therapy is in clinical use.
• Prognostic factors: Presumed to include severity of cranial/skull-base involvement (nerve compression risk), cardiac status, and degree of developmental delay. No validated prognostic biomarkers.
• QoL measures: Not formally assessed.

Given the small cohort, all prognostic statements carry substantial uncertainty.

Section 12 — Treatment

No disease-specific, approved therapy exists. Management is supportive and multidisciplinary.

• Supportive/rehabilitative: Orthopedic management of hip dysplasia (bracing/surgery as indicated); monitoring for and surgical decompression of skull-base/foraminal narrowing if cranial-nerve compromise develops; cardiology management of heart malformations; hematology follow-up; developmental/physical/occupational therapy as needed. NCIT concepts: Supportive Care (NCIT:C15277), Orthopedic Surgery (NCIT:C16536), Physical Therapy (NCIT:C15367).
• Pharmacotherapy: No established drug therapy. No pharmacogenomic guidance.
• Advanced/experimental (mechanistically rational, not clinical): Tankyrase inhibition (e.g., XAV939) is an in-vitro-validated concept — it stabilizes AXIN1/AXIN2 and normalizes Wnt overactivity in mutant cells (PMID: 37582359). This has not been tested in patients; systemic tankyrase inhibition carries known intestinal/skeletal toxicity concerns and would require careful therapeutic-window definition. No gene/cell/RNA therapies are in development for CMDOH.
• Treatment strategy: Individualized, symptom-directed. Any future targeted (Wnt-normalizing) approach would most plausibly target the early childhood window of active bone modeling.

Section 13 — Prevention

• Primary prevention: Not possible for a Mendelian recessive disorder beyond reproductive genetic counseling.
• Genetic counseling & reproductive options: Central preventive measure. For carrier couples (25% recurrence risk per pregnancy), options include carrier/cascade testing, prenatal diagnosis, and preimplantation genetic testing (PGT) for the familial AXIN1 variants. Counseling on consanguinity-associated recessive risk is relevant.
• Secondary/tertiary prevention: Early imaging surveillance to detect and manage skull-base sclerosis (prevent nerve compression), hip surveillance, and cardiac/hematologic monitoring to prevent complications.
• Immunization / public health / environmental: Not applicable.

Section 14 — Other Species / Natural Disease

• Orthologous gene: Axin1 is conserved across vertebrates. Mouse Axin1 (NCBI Gene ID 12005) was originally identified as the Fused locus (PMID: 9230313); "Axin is a novel inhibitor of Wnt signaling." Taxonomy: Mus musculus (NCBI:txid10090), Danio rerio, etc.
• Naturally occurring disease: No naturally occurring CMDOH-equivalent has been catalogued in companion animals or wildlife (no OMIA entry identified for this specific disorder).
• Comparative biology: The canonical Wnt/β-catenin pathway and AXIN scaffold function are deeply evolutionarily conserved, underpinning the validity of mouse models. Classic mouse Fused/Axin1 mutant alleles cause axial duplication, reflecting Wnt's role in body-axis patterning.
• Zoonotic potential / cross-species transmission: Not applicable (genetic disorder).

Section 15 — Model Organisms

• Mouse — constitutive knockout: Axin1-null is embryonic lethal, precluding study of the adult skeletal phenotype and underscoring why viable human disease requires hypomorphic alleles.
• Mouse — conditional osteoblast-lineage deletion (Axin1^Osx^, Osterix-Cre): Increased β-catenin, elevated OPG and OPG:RANKL ratio, decreased TRAP+/MMP9+/cathepsin-K+ osteoclasts, suppressed osteoclastogenesis, and delayed postnatal bone growth — directly recapitulating the human biopsy finding of increased osteoblast/reduced osteoclast activity (Bone Research 2020, PMC7424530).
• Mouse — limb-mesenchyme deletion (Prrx1-Cre): Elevated β-catenin producing a fibular-hemimelia/tarsal-coalition phenotype, partially rescued by deleting one β-catenin allele (bone volume 92%→71%), formally demonstrating β-catenin dependence (eLife 2023, PMC9815809).
• Paralog — Axin2 knockout: Craniosynostosis and increased trabecular bone mass, illustrating overlapping AXIN function in skeletal Wnt regulation and supporting AXIN2 as a compensatory modifier.
• Cellular models: Patient-derived primary cells and CRISPR genome-edited cells carrying p.Arg723 / p.Asp796Glufs6, plus HEK293T reporter systems, demonstrated enhanced basal Wnt activity and XAV939 rescue (PMID: 37582359).
• Phenotype recapitulation & limitations: Osteoblast-lineage models faithfully reproduce the cellular remodeling imbalance but not the full craniospinal radiographic pattern; constitutive models are lethal. No model reproduces the complete extraskeletal spectrum (cardiac, hematological).
• Resources: MGI (mouse), Alliance of Genome Resources, IMPC/IMSR for Axin1/Axin2 alleles.

Key Findings (with evidence)

F001 — CMDOH is caused by biallelic C-terminal-truncating AXIN1 variants (autosomal recessive)

Terhal et al. 2023 identified three homozygous, C-terminally truncating AXIN1 variants in seven individuals from four families. Verbatim: "We identified three homozygous, C-terminally truncating AXIN1 variants in seven individuals from four families affected by macrocephaly, cranial hyperostosis, and vertebral endplate sclerosis. Other frequent findings included hip dysplasia, heart malformations, variable developmental delay, and hematological anomalies." (PMID: 37582359). This establishes the gene, variant class, inheritance, cohort size, and the core phenotype.

F002 — Mechanism: DIX-domain truncation enhances canonical Wnt/β-catenin signaling; tankyrase inhibition rescues

In patient-derived and genome-edited cells, "analyses of primary and genome-edited cells harboring the truncating variants revealed enhanced basal canonical Wnt pathway activity." The truncations impair DIX-domain polymerization but partially retain Wnt-inhibitory function on overexpression, and "addition of a tankyrase inhibitor attenuated Wnt overactivity in the AXIN1-mutant model systems" (PMID: 37582359) — establishing both the causal Wnt-gain mechanism and a candidate targeted therapy.

F003 — AXIN1 is a Wnt-pathway tumor suppressor; complete somatic loss causes cancer, whereas germline hypomorphic truncations cause skeletal dysplasia

AXIN1 is a bona fide tumor suppressor: in hepatocellular carcinoma, "up to 50% of cases exhibit aberrant activation of the canonical Wnt/β-catenin pathway, driven by CTNNB1 mutations or inactivating alterations in AXIN1, adenomatous polyposis coli, or ZNRF3, which are mutually exclusive" (PMID: 42237887). Functional demonstration: "the WNT pathway and its target gene c-Myc were activated when AXIN1 was missing" (PMID: 39653061); AXIN1 mutation upregulates CCND1 (p=0.022) with β-catenin–CCND1 correlation r=0.43 (PMID: 40344393). The dosage/allele-class contrast explains why germline hypomorphic DIX truncations yield a viable Wnt-gain skeletal phenotype rather than malignancy.

F004 — Clinical spectrum and specific variants (OMIM #620558)

Core radiographic features: macrocephaly (+2.2 to +6.1 SD), progressive calvarial/skull-base sclerosis, vertebral endplate ("sandwich") sclerosis, metaphyseal flaring, hypertelorism with low nasal bridge; plus hip dysplasia, heart malformations, variable developmental delay, hematological anomalies, and short stature in a subset. Bone biopsy showed increased osteoblast and reduced osteoclast activity at the growth-plate resorption zone. "Sclerosing skeletal dysplasias result from an imbalance between bone formation and resorption" (PMID: 37582359) frames the underlying remodeling defect.

F005 — Mouse Axin1 models support the Wnt-gain, osteoblast-lineage mechanism

Axin1 was cloned from the mouse Fused locus (PMID: 9230313); constitutive knockout is embryonic-lethal. Osteoblast-lineage deletion increases β-catenin and OPG:RANKL and suppresses osteoclastogenesis (mirroring the human biopsy); limb-mesenchyme deletion elevates β-catenin and is partially rescued by removing one β-catenin allele, confirming β-catenin dependence.

F006 — XAV939 dose-dependently rescues Wnt overactivity (partly via AXIN2)

In HEK293T cells engineered with AXIN1 p.Arg723 and p.Asp796Glufs6, XAV939 at 100 nM and 1 µM produced dose-dependent suppression of Wnt activation; it also rescued AXIN1-knockout cells, implicating AXIN2 stabilization. Tankyrases PARylate AXIN1/AXIN2 for degradation, so their inhibition raises AXIN levels and restores destruction-complex activity (PMID: 37582359).

Mechanistic Model (synthesis)

  Biallelic AXIN1 C-terminal truncation (p.Arg723*, p.Arg805*, frameshift)
                     │  removes/disrupts DIX domain
                     ▼
  ↓ AXIN1 protein level  +  impaired AXIN1 polymerization
                     │
                     ▼
  Weakened β-catenin destruction complex (AXIN1·APC·GSK3β·CK1)
                     │  ↓ β-catenin phosphorylation/degradation
                     ▼
  ↑ Nuclear β-catenin  →  ENHANCED canonical Wnt/β-catenin output   ◄── XAV939
                     │        (demonstrated in patient/edited cells)      (tankyrase
        ┌────────────┴─────────────┐                                       inhibitor,
        ▼                          ▼                                       stabilizes
  ↑ Osteoblast formation     ↑ OPG : RANKL ratio                           AXIN1/AXIN2
                                   │  (mouse model)                        → rescues)
                                   ▼
                             ↓ Osteoclast resorption
        └────────────┬─────────────┘
                     ▼
     NET ↑ BONE MASS → osteosclerosis / hyperostosis
                     ▼
  Macrocephaly · cranial/skull-base hyperostosis · sandwich vertebrae ·
  metaphyseal flaring · hip dysplasia · (cardiac, developmental, hematologic — inferred)

Allele-dosage spectrum of AXIN1 dysfunction:

Evidence Base

Note on evidence strength: The human genotype–phenotype and Wnt-gain findings rest on a single, well-executed 7-patient study with strong functional validation. The osteoblast/osteoclast step is supported by model-organism data. The tumor-suppressor/dosage framing is inferential context from cancer literature, not direct CMDOH data.

Limitations and Knowledge Gaps

1. Tiny cohort (n=7/4 families): Frequencies, penetrance, expressivity, and prognosis are imprecise; the phenotype spectrum may broaden as more cases are found.
2. No epidemiology: Prevalence, incidence, carrier frequency, and geographic/ethnic distribution are unquantified.
3. Extraskeletal mechanism is inferred: How Wnt-gain produces the cardiac, hematological, and developmental features is not mechanistically dissected.
4. No natural-history/QoL/survival data: Long-term outcomes and QoL are undefined.
5. Therapy is preclinical only: XAV939 rescue is in-vitro; efficacy, safety, therapeutic window, and delivery in patients are unknown. Systemic tankyrase inhibition has known toxicity.
6. No patient omics datasets (transcriptomic/proteomic/single-cell) exist to refine cell-type-specific mechanisms.
7. Modifier genetics unproven: AXIN2 compensation is inferred from cell rescue, not demonstrated as a clinical modifier.

Proposed Follow-up Experiments / Actions

1. Establish an international CMDOH registry and GeneMatcher-driven case collection to define prevalence, natural history, penetrance, and expressivity.
2. Generate a knock-in mouse carrying a patient-equivalent C-terminal Axin1 truncation (e.g., R723*-analog) to recapitulate the craniospinal radiographic phenotype and test therapeutic windows.
3. In-vivo XAV939 / next-generation tankyrase-inhibitor studies in the knock-in model, defining dose, timing (early bone-modeling window), and skeletal/intestinal toxicity — a prerequisite before any human consideration.
4. Patient-derived iPSC → osteoblast/osteoclast organoid models to map cell-type-specific Wnt targets (RNA-seq/single-cell) and OPG:RANKL dynamics directly in human cells.
5. Dissect extraskeletal mechanisms (cardiac, hematopoietic) via conditional Axin1-truncation models in relevant lineages.
6. Systematic differential-diagnosis workflow: incorporate AXIN1 into sclerosing-dysplasia gene panels and publish radiographic criteria distinguishing CMDOH (Wnt) from TMEM53/CTDI (BMP-SMAD) and osteopetroses.
7. Biomarker development: evaluate serum bone-turnover markers and Wnt-target readouts as potential monitoring/prognostic biomarkers.

Report compiled from a 5-iteration autonomous investigation; 6 confirmed findings, 12 papers reviewed. Primary source: Terhal et al., 2023 (PMID: 37582359).