Hereditary Hemorrhagic Telangiectasia Type 4

1. Disease Information

Overview

Hereditary hemorrhagic telangiectasia (HHT; Osler–Weber–Rendu disease) is an autosomal dominant multisystem vascular dysplasia in which arteriovenous malformations (AVMs) — direct artery-to-vein connections lacking an intervening capillary bed — form in mucocutaneous sites (as small telangiectases) and in viscera (as large AVMs in lung, liver, brain, spine). GeneReviews (PMID:20301525, updated 19 Feb 2026) states:

"Hereditary hemorrhagic telangiectasia (HHT) is characterized by the presence of multiple arteriovenous malformations (AVMs) that lack intervening capillaries and result in direct connections between arteries and veins. The most common clinical manifestation is recurrent nosebleeds (epistaxis) beginning on average at age 12 years."

HHT4 is the fourth mapped locus for this phenotype. Bayrak-Toydemir et al. (PMID:16969873) studied one family with classic HHT in which linkage to HHT1 (ENG), HHT2 (ACVRL1) and HHT3 (5q) had been excluded, and reported:

"Whole genome linkage analysis and fine mapping results suggested a 7 Mb region on the short arm of chromosome 7 (7p14) between STR markers D7S2252 and D7S510. We obtained a maximum two point LOD score of 3.60 with the STR marker D7S817. This region was further confirmed by haplotype analysis. These findings suggest the presence of another gene causing HHT (HHT4)."

Identifiers (verified against MONDO via EBI OLS4, 18 Aug 2026)

Synonyms

• HHT4 (abbreviation; MONDO related synonym, xref GARD)
• Telangiectasia, hereditary hemorrhagic, type 4 (OMIM title form)
• Inherited from the parent concept, applicable but not HHT4-specific: Osler–Weber–Rendu disease/syndrome, Rendu–Osler–Weber syndrome, Osler disease.

Nature of the evidence base

Aggregated disease-level resource + a single pedigree study. There is no HHT4 registry, no EHR-derived cohort, no biobank series. The 2006 paper is a family-based linkage study with clinical phenotyping of the pedigree. All population-level information in this report describes HHT sensu lato.

2. Etiology

Causal factors

• Primary cause: an unidentified, presumptively heterozygous germline variant in a gene within the 7p14 interval, segregating with HHT in one pedigree with a maximum two-point LOD of 3.60 at D7S817 (PMID:16969873). The interval spans ~7 Mb between D7S2252 and D7S510.
• Candidate genes explicitly excluded by the original authors' sequencing: BMPER, CCM2, RALA, INHBA (reported in the 2006 paper's candidate-gene analysis and reproduced in OMIM 610655; the abstract itself does not enumerate them, so this should be curated as a full-text/OMIM-sourced claim). All four are plausible on prior biology — BMPER modulates BMP signaling, CCM2 causes cerebral cavernous malformation, RALA is a Ras-family GTPase, INHBA encodes the inhibin/activin βA subunit in the TGF-β superfamily.
• Non-genetic causes: none. HHT is monogenic.
• Mechanistic prior: because all four known HHT genes (ENG, ACVRL1, SMAD4, GDF2) act in one pathway, the strong a priori expectation is that the HHT4 gene is a further BMP9-10/ENG/ALK1/SMAD4 pathway component or modifier. This is a hypothesis, not a finding — curate it as a KNOWLEDGE_GAP, never as pathophysiology.

Genetic risk factors

• The segregating 7p14 haplotype is the risk factor, in a Mendelian dominant sense.
• No susceptibility loci, GWAS signals, or modifier alleles have been reported for HHT4.
• For HHT generally, PTPN14 has been proposed as a modifier that "protects SMAD4 from ubiquitination and turnover to potentiate BMP9 signaling in endothelial cells" (bioRxiv preprint; treat as preliminary, not peer-reviewed evidence).

Environmental risk factors

None established for HHT4. For HHT generally, the phenotype is modulated by physiological states that drive angiogenesis or alter hemodynamics — puberty, pregnancy, and aging correlate with lesion appearance and progression; the JCI review (PMID:38357927) frames the disease as caused by "abnormal activation of angiogenesis," and animal work shows AVMs require an angiogenic trigger (wounding, VEGF) on top of the genetic lesion (Park et al., J Clin Invest 2009;119:3487–96, PMID:19805914).

Protective factors

• Genetic: none reported.
• Environmental: none established. Of note, a 20-year Danish follow-up (Kjeldsen et al., Orphanet J Rare Dis 2016, PMID:27876060) found lower cancer incidence in HHT patients than matched controls — "Cancer diagnoses had been registered in the follow-up period in 4 (5%) HHT patients and in 38 (17%) controls" — consistent with constitutively impaired angiogenesis, but this is an observation about HHT1/HHT2 patients, not a protective factor for HHT4.

Gene–environment interactions

The best-supported GxE concept in HHT is the "second hit / angiogenic trigger" model: heterozygous pathway loss is necessary but not sufficient, and a local angiogenic or injury stimulus precipitates AVM formation. Evidence is model-organism (PMID:19805914). Whether it applies to the HHT4 lesion is unknown and untestable until the gene is found.

3. Phenotypes

HHT4-specific phenotype

The only HHT4-specific phenotypic statement in the literature is in the 2006 abstract:

"Here we report on linkage results on a family with classic features of HHT, albeit a less severe phenotype with regards to epistaxis and telangiectases, in which linkage to HHT1, HHT2, and HHT3 is ruled out."

and

"The features in this family that strongly suggest the presence of a hereditary, multisystem vascular dysplasia would be easily missed during the typical evaluation and management of a patient with an AVM. This family helps emphasize the need to obtain a very detailed, targeted medical and family history for even mild, infrequent but recurring nosebleed, subtle telangiectases."

Curation implications, all directly supported: - HHT4 as described is a mild/attenuated mucocutaneous phenotype with preserved visceral AVM burden — epistaxis is infrequent, telangiectases subtle, but an AVM brought the family to attention. - The family is therefore at high risk of being under-ascertained by Curaçao criteria, which weight epistaxis and telangiectases heavily. - Per-member clinical detail (number of affected individuals, AVM organ distribution, ages of onset) is in the full text only and is paywalled (Wiley/Ovid; no PMC deposit). Do not assert per-member specifics without obtaining the full text.

Phenotype spectrum inherited from HHT (extrapolated — flag as such)

Frequencies below are for HHT overall (predominantly ENG/ACVRL1 cohorts) and should carry evidence_source: HUMAN_CLINICAL with an explicit note that they are parent-disease frequencies, not HHT4 frequencies. Per the dismech frequency SOP, prefer omitting frequency: on HHT4 phenotypes rather than importing a band that no HHT4 evidence supports.

(All HPO IDs above were verified against the repository's cache/hp/terms.csv label cache; each label is the canonical HPO label.)

Quality-of-life impact

HHT-wide, not HHT4-specific. The PATH-HHT trial used a validated HHT-specific QoL instrument alongside the Epistaxis Severity Score (ESS) and demonstrated measurable impairment and treatment-responsive improvement: the mean between-group difference in HHT-specific QoL change was −1.4 points (95% CI −2.6 to −0.3) on a 0–16 scale (PMID:39292928). Al-Samkari (Blood 2024, PMID:38864625) states that "HHT-associated bleeding results in substantial psychosocial morbidity and iron deficiency anemia that may be severe." An expert consensus call for composite hematologic outcome measures in HHT was published in 2026 (PMID:42253246).

Given the 2006 report describes attenuated epistaxis and telangiectases, QoL burden in HHT4 may be driven disproportionately by AVM complications rather than by bleeding — but this is inference, not data.

4. Genetic / Molecular Information

Causal gene: UNKNOWN

This is the single most important curated fact. Record it explicitly rather than leaving the gene slot silently empty.

McDonald et al. (Front Genet 2015, PMID:25674101) confirm the status: "Linkage analysis identified two additional HHT loci at chromosome 5q31 and chromosome 7p14, but the genes still remain unknown." The same review offers an important alternative hypothesis for gene-negative HHT — that "most of the ∼3% of patients with HHT according to Curaçao criteria who are not found to have a mutation … have an undetected deep intronic mutation in ACVRL1 or ENG." A rigorous entry should carry this as a competing interpretation of the HHT4 family alongside the novel-gene hypothesis.

The four established HHT/HHT-overlap genes (context, not HHT4 causes)

HGNC IDs below verified against cache/hgnc/terms.csv (lowercase hgnc: prefix is canonical in this repository):

Genes relevant to differential diagnosis: RASA1 (hgnc:9871, CM-AVM1), EPHB4 (hgnc:3395, CM-AVM2), CCM2 (hgnc:21708).

Detection rates: mutations in ENG and ACVRL1 "have been reported to cause up to 85% of HHT," and the Utah group reports "approximately 96% of individuals with HHT have a mutation in these two genes, when published (Curaçao) diagnostic criteria for HHT are strictly applied" (PMID:25674101). The 2026 ERJ review puts it at ">90%" across the three main genes (PMID:41713948). The residual few percent is the population from which HHT3/HHT4 families were drawn.

Pathogenic variants, classification, allele frequency, origin

• Variant type/class: unknown for HHT4. No variant has been reported, so no ACMG/AMP classification, no ClinVar record, no gnomAD allele frequency exists for this entity. Do not attribute any variant to HHT4.
• Origin: germline, inferred from dominant segregation across the pedigree (PMID:16969873). No somatic component described.
• Functional consequence: unknown. The prior — by analogy with ENG/ACVRL1/SMAD4/GDF2, in which HHT is "caused by loss-of-function mutations in the BMP9-10/ENG/ALK1/SMAD4 signaling pathway" (PMID:38357927) — is haploinsufficiency/loss of function, but this is untested for HHT4.
• Modifier genes, epigenetics, chromosomal abnormalities: nothing reported for HHT4. No DNA-methylation, histone-modification, aneuploidy, translocation, or inversion data exist. Report as "not available."

5. Environmental Information

• Environmental factors: none causal. HHT4 is monogenic.
• Lifestyle: no lifestyle factor causes HHT. Management-relevant behavioral factors (avoid vigorous nose blowing, heavy lifting, straining, nasal digital manipulation, NSAIDs/anticoagulants where bleeding is significant, scuba diving unless right-to-left shunt is excluded, liver biopsy) are listed in GeneReviews (PMID:20301525) as "agents/circumstances to avoid" — these are tertiary-prevention measures, not etiologic exposures.
• Infectious agents: not causal. Relevant only downstream: right-to-left pulmonary shunting permits bacteremic seeding, causing brain abscess (HP:0030049) — hence the guideline recommendation for antibiotic prophylaxis before dental and non-sterile invasive procedures, and air-filter precautions on IV lines (PMID:20301525).

No ECTO exposure term is appropriate for this entry; do not force one.

6. Mechanism / Pathophysiology

What is established for HHT4

Nothing at the molecular level. No protein, no pathway, no cell-biological assay, no expression study exists for HHT4. The pathophysiology block should therefore be modeled as a causal chain from an unidentified 7p14 lesion into the shared HHT effector chain, with the first edge flagged as inferred.

The shared HHT mechanism (inherited chain — the honest way to model HHT4 downstream biology)

The canonical, well-evidenced chain (PMID:38357927; PMID:41347972):

1. Loss-of-function lesion in the BMP9/BMP10–ENG–ALK1–SMAD4 axis (MOLECULAR) — endothelial BMP9/10 ligands bind the ALK1 type I receptor with endoglin as co-receptor; endoglin "serves as a reservoir of these ligands on the surface of ECs, enhancing ligand-induced responses." → GO:GO:0030509 BMP signaling pathway (DECREASED/LOSS_OF_FUNCTION); GO:GO:0007179 transforming growth factor beta receptor signaling pathway; GO:GO:0060395 SMAD protein signal transduction.
2. Failure of SMAD1/5/8-mediated transcriptional maintenance of endothelial quiescence (CELLULAR) — "The BMP9-10/ENG/ALK1/SMAD4 signaling pathway maintains vascular quiescence by repressing angiogenic pathways" (PMID:38357927).
3. De-repression of pro-angiogenic signaling, notably the VEGF axis (CELLULAR) — crosstalk with the VEGF pathway is the basis for repurposing anti-VEGF drugs. → GO:GO:0048010 vascular endothelial growth factor receptor signaling pathway (INCREASED).
4. Excessive endothelial proliferation and dysregulated angiogenesis (CELLULAR) — "abnormal activation of angiogenesis, a process causing excessive EC proliferation and hypervascularization." → GO:GO:0001525 angiogenesis (INCREASED); GO:GO:0001935 endothelial cell proliferation (INCREASED); GO:GO:0002040 sprouting angiogenesis. → CL:CL:0000071 blood vessel endothelial cell; CL:CL:0000115 endothelial cell; CL:CL:0002543 vein endothelial cell.
5. Loss of the capillary bed and arteriovenous shunt formation (AVM/telangiectasis) (TISSUE) — "multiple arteriovenous malformations (AVMs) that lack intervening capillaries and result in direct connections between arteries and veins" (PMID:20301525). Mural-cell recruitment is abnormal. → GO:GO:0001974 blood vessel remodeling; CL:CL:0000192 smooth muscle cell; CL:CL:0000669 pericyte. → UBERON:UBERON:0001982 capillary; UBERON:UBERON:0001637 artery.
6. Two divergent clinical outputs (ORGANISM):
7. Fragile superficial lesions → hemorrhage → chronic iron loss → iron-deficiency anemia (epistaxis, GI bleeding).
8. Large visceral AVMs → shunt physiology → hypoxemia and paradoxical embolism (pulmonary), high-output cardiac failure and portal hypertension (hepatic), hemorrhage (cerebral).

Modifier of the chain: the angiogenic "second hit"

Heterozygous pathway loss alone does not reliably produce AVMs in mice; an angiogenic stimulus (wounding, exogenous VEGF) is required for de novo AVM formation (Park et al. 2009, PMID:19805914). This explains the focal, progressive, age-dependent lesion distribution and is the mechanistic bridge to why puberty and pregnancy modulate the phenotype.

Protein dysfunction, metabolism, immunity, tissue damage, epigenetics

• Protein dysfunction: haploinsufficiency (reduced dosage of a receptor/co-receptor/ligand) rather than aggregation or gain of function, in the four known genes. Unknown for HHT4.
• Metabolic changes: no primary metabolic defect. Secondary: iron deficiency from chronic blood loss (CHEBI:18248 iron — verify with OAK before use); hepatic VM can produce a hyperdynamic circulatory state.
• Immune involvement: not an autoimmune or immunodeficiency disorder. Immune relevance is limited to loss of pulmonary capillary filtration permitting septic embolization.
• Tissue damage: hemorrhage, ischemia (paradoxical embolic stroke), and shunt-driven organ overload; not oxidative stress or fibrosis in the classic sense. Hepatic VM may progress to nodular change and cirrhosis (HP:0001394)/portal hypertension.
• Epigenetics: no data for HHT4 or for HHT more broadly at a level suitable for curation.

Molecular profiling and advanced technologies

No transcriptomic, proteomic, metabolomic, lipidomic, single-cell, spatial, or CRISPR-screen dataset exists for HHT4. There is no GEO/ArrayExpress/PRIDE/MetaboLights accession attributable to this entity. For HHT generally, endothelial-cell transcriptomic and BMP9-response studies exist (e.g., NCT05632484 studies EC genotype–phenotype under BMP9 stimulation in ACVRL1/ENG/SMAD4 carriers), and brain-AVM single-cell/genomic work is reviewed in PMID:42106885 — none of it is HHT4-specific.

Curation guidance: the datasets: block for HHT4 should be left empty rather than populated with parent-disease or gene-only accessions. Per the repository's dataset SOP, a GENE_ONLY or relaxed-name search here would necessarily return HHT1/HHT2 data — a textbook Named Entity Confusion route, since HHT4 has no gene to search on.

7. Anatomical Structures Affected

Inherited from HHT; the 2006 report establishes visceral AVM involvement in the index family but the organ distribution requires full-text confirmation.

Organ level

Body systems: cardiovascular (primary), respiratory, hepatobiliary/digestive, nervous, integumentary, hematologic (secondary).

Tissue and cell level

• Vascular endothelium is the primary affected tissue — the disease is endothelial-cell-autonomous in mouse models (endothelial-specific Acvrl1 depletion suffices; PMID:24896812).
• Cells: CL:0000115 endothelial cell, CL:0000071 blood vessel endothelial cell, CL:0002543 vein endothelial cell, CL:0000192 smooth muscle cell (mural), CL:0000669 pericyte.

Subcellular level

Plasma membrane receptor complex (ALK1/endoglin/BMPRII), cytoplasm→nucleus SMAD shuttling. GO cellular-component terms (GO:0005886 plasma membrane, GO:0005634 nucleus) are generic and add little; prefer the biological-process annotations above. No HHT4-specific subcellular data.

Localization and lateralization

Lesions are multifocal, bilateral, and asymmetric, not lateralized. Telangiectases favor the face, oral mucosa, and fingertips; PAVMs favor the lower lobes.

8. Temporal Development

For HHT4 specifically: the family was ascertained via AVM presentation in adulthood, with mild/infrequent epistaxis that "would be easily missed" (PMID:16969873). No age-of-onset data are published.

Inherited from HHT (PMID:20301525):

• Onset: insidious, age-dependent penetrance of individual manifestations. Epistaxis begins on average at age 12 years; telangiectases appear later, sometimes in childhood; GI bleeding "is rarely seen before age 50 years." Cerebral AVMs are congenital.
• Onset pattern: chronic and insidious, punctuated by acute hemorrhagic or embolic events — "complications from bleeding or shunting may be sudden and catastrophic."
• Progression: lesion burden accumulates lifelong; the disease is progressive in lesion count and severity but not staged in an oncologic sense. There is no accepted staging system; severity is graded instrumentally (Epistaxis Severity Score 0–10; a change of ≥0.71 points is clinically significant per PMID:39292928).
• Course pattern: chronic/lifelong with an episodic bleeding overlay.
• Remission: no spontaneous remission. Treatment-induced remission of bleeding is achievable and is now the therapeutic goal.
• Critical windows: childhood/adolescence (initial PAVM and CAVM screening), pregnancy (PAVM growth and rupture risk — "sizable pulmonary AVMs discovered during pregnancy are treated during the second trimester"), and the pre-procedural window (prophylaxis before dental/invasive procedures once shunt is known).

9. Inheritance and Population

Inheritance

• Autosomal dominant (HP:0000006), established by segregation and haplotype analysis in the index pedigree (PMID:16969873).
• Penetrance: near-complete for the disease but strongly age-dependent for individual manifestations — GeneReviews describes "considerable intrafamilial variability and age-related penetrance of individual manifestations" (PMID:20301525); Anzell et al. call HHT "a near-fully penetrant autosomal dominant disorder" (PMID:40964703). For HHT4 itself, penetrance is unquantified.
• Expressivity: highly variable; the JCI review describes HHT as "an inherited vascular disorder with highly variable expressivity" (PMID:38357927). The HHT4 family's mild mucocutaneous phenotype may itself be an expressivity phenomenon rather than a locus-specific signature — an important, curatable open question.
• Anticipation: none (not a repeat-expansion disorder).
• Germline mosaicism: not reported for HHT4; documented occasionally in HHT generally.
• Founder effects, consanguinity, carrier frequency: not applicable/unknown for HHT4. A single-family locus has, by definition, a private variant.
• Recurrence risk: 50% to offspring of an affected individual; 50% to sibs when a parent is affected (PMID:20301525). However, predictive genetic testing is impossible in HHT4 families because no variant has been identified — at-risk relatives must be managed by clinical screening. This is the single most consequential practical difference between HHT4 and HHT1/HHT2.

Epidemiology

HHT4 prevalence is effectively 1 known family worldwide. Curate as CASES_IN_LITERATURE / ULTRA_RARE, rate_per_100000 unset — not as an Orphanet band.

Parent-disease epidemiology (for context, explicitly labeled):

Note: the gnomAD-based estimate is ENG/ACVRL1-only and therefore says nothing about HHT4's contribution — but it does bound the gene-negative fraction indirectly.

Sex ratio: approximately 1:1. The Korean series reported 41 males and 71 females among 112 identified patients, but concluded "an almost equal prevalence among men and women" after accounting for ascertainment (PMID:31455059).

Geographic distribution: worldwide, no ethnic restriction. Regional high-prevalence pockets in HHT reflect founder variants in ENG/ACVRL1 (e.g., Netherlands Antilles). The HHT4 family's ancestry should be taken from the full text, not assumed.

10. Diagnostics

Clinical criteria — the operative diagnostic route for HHT4

Because no HHT4 gene test exists, HHT4 is diagnosed clinically, by the Curaçao criteria (Shovlin et al., Am J Med Genet 2000, PMID:10751092). GeneReviews states:

"The clinical diagnosis of HHT can be established in a proband with at least three of the following diagnostic criteria: recurrent epistaxis; mucocutaneous telangiectases in characteristic locations; visceral AVMs; a first-degree relative diagnosed with HHT on the basis of the preceding criteria."

Three or more criteria = definite; two = possible/suspected; fewer than two = unlikely.

The 2006 paper's central clinical warning is a criteria-sensitivity warning: in a family with attenuated epistaxis and subtle telangiectases, an AVM may be the presenting and near-only finding, and "would be easily missed during the typical evaluation and management of a patient with an AVM." Curate this as a diagnostic pitfall.

Genetic testing

• Recommended approach: sequential or panel-based testing of ACVRL1, ENG, SMAD4 (± GDF2, RASA1, EPHB4). GeneReviews: "The molecular diagnosis is established by identification of a heterozygous pathogenic variant in ACVRL1, ENG, or SMAD4 by molecular genetic testing." McDonald et al. propose "a five gene (ENG, ACVRL1, SMAD4, RASA1, and GDF2) NGS panel" for suspected hereditary telangiectasia that is not classic HHT (PMID:25674101).
• HHT4 is, operationally, a diagnosis of exclusion made after this panel is negative — plus exclusion of linkage to the known loci, which in the original study required family-based analysis.
• Deletion/duplication analysis of ENG/ACVRL1 must be included before concluding a family is gene-negative (whole-exon deletions are a recognized cause; e.g., two distinct ENG deletions in one family, Wooderchak et al. 2010).
• WES/WGS: the appropriate modern approach for a gene-negative family. McDonald et al.: "Recent availability of whole exome and genome testing has created new opportunities to facilitate gene discovery, identify genetic modifiers to explain clinical variability, and potentially define an increased spectrum of hereditary telangiectasia disorders." WGS is specifically indicated over WES for this population, because the leading competing hypothesis for gene-negative HHT is a deep intronic ACVRL1/ENG variant that exome capture cannot see (PMID:25674101).
• CMA, karyotype, FISH, mtDNA, repeat-expansion testing: not indicated; no chromosomal or repeat mechanism is implicated.
• A predictive test for HHT4 relatives does not exist. Do not curate one.

Clinical tests, imaging, and screening

Per the Second International Guidelines (PMID:32894695) and GeneReviews (PMID:20301525):

Contraindicated/avoid: liver biopsy (PMID:20301525).

Biopsy/histopathology: biopsy is not part of HHT diagnosis. Histology of a telangiectasis shows dilated post-capillary venules connecting directly to arterioles with loss of the intervening capillary bed and perivascular mononuclear infiltrate — reported for HHT generally, never for HHT4 tissue.

Omics-based diagnostics: none validated for HHT of any type. No liquid biopsy, no proteomic or metabolomic diagnostic. Biomarker research exists but is exploratory (see PMC4379940, "Research on potential biomarkers in hereditary hemorrhagic telangiectasia").

Differential diagnosis

Screening for asymptomatic individuals

Cascade screening in an HHT4 family must be clinical, not molecular: "If the pathogenic variant in the family is not known, at-risk family members should be evaluated for signs and symptoms of HHT, and screening should be offered to at-risk family members if the diagnosis cannot be ruled out" (PMID:20301525). No newborn or population carrier screening exists or is indicated.

11. Outcome / Prognosis

No HHT4-specific outcome data exist. Parent-disease data:

Survival and mortality

The best population-based evidence is the 20-year Danish follow-up of an unselected County of Fyn cohort (Kjeldsen et al. 2016, PMID:27876060): 73 HHT patients and 218 matched controls.

"A total of 32 (44%) HHT patients and 97 (44%) controls passed away during follow-up. The survival curves were evenly distributed showing similar survival rates in the two groups. … The mortality was not increased among Danish HHT patients compared to controls."

The authors emphasize this reflects "a clinical unselected series of HHT patients with the whole spectrum of severity" — i.e., HHT in a well-managed population is compatible with normal life expectancy, though referral-center series with severe visceral disease report excess early mortality. There is no 5-/10-year survival convention for HHT (not a malignancy).

Cancer: reduced incidence in HHT patients versus controls (5% vs 17%) in the same study — an intriguing, angiogenesis-consistent finding.

Morbidity, disability, quality of life

Driven by (a) chronic bleeding → transfusion-dependent iron-deficiency anemia and fatigue; (b) AVM complications → stroke, brain abscess, hemorrhage, high-output heart failure, PH. Al-Samkari (PMID:38864625): HHT "affects 1 in 5000 persons, making it the second most common inherited bleeding disorder worldwide," and HHT-associated bleeding "results in substantial psychosocial morbidity and iron deficiency anemia that may be severe."

Instruments: Epistaxis Severity Score (disease-specific, validated), HHT-specific QoL score (0–16), plus generic EQ-5D/SF-36. A 2026 expert group has called for composite hematologic endpoints (PMID:42253246).

Complications (curatable list)

Iron-deficiency anemia; transfusion dependence; paradoxical embolic stroke and TIA; brain abscess; intracerebral hemorrhage; hemothorax/hemoptysis from PAVM rupture; high-output cardiac failure; portal hypertension and biliary ischemia from hepatic VM; pulmonary hypertension — "PH is a recognised but heterogeneous complication of HHT, with reported prevalence ranging widely from 1.5% to 45%, depending on diagnostic methods and study populations" (PMID:41713948); pregnancy-associated PAVM hemorrhage; atrial fibrillation management dilemmas (anticoagulation vs bleeding — see the 2026 left-atrial-appendage-closure study, PMID:41506960).

Prognostic factors

Genotype (in the known genes), AVM organ distribution and size, baseline ESS and hemoglobin/ferritin, presence of PH or high-output failure, and access to an HHT Center of Excellence. HHT4 has no genotype-based prognostic information — its prognosis must be assessed lesion-by-lesion. No validated prognostic biomarker exists for any HHT type.

12. Treatment

Treatment of HHT4 is identical to treatment of HHT — it is organ- and symptom-directed, and none of it requires knowing the gene. This is the reassuring practical corollary of the unidentified locus, and worth stating explicitly in the entry.

Regulatory status: as of the 2024 JCI review, "this not-so-uncommon bleeding disorder still currently lacks any FDA- or European Medicines Agency-approved (EMA-approved) therapies" (PMID:38357927), reiterated in Blood 2024: "there remain no regulatory agency-approved therapies for HHT" (PMID:38864625). All pharmacotherapy is off-label repurposing.

Systemic pharmacotherapy

PATH-HHT, the practice-changing trial (Al-Samkari et al., N Engl J Med 2024;391:1015–1027; PMID:39292928; NCT03910244):

"The trial was closed to enrollment in June 2023 after a planned interim analysis met a prespecified threshold for efficacy. A total of 144 patients underwent randomization; 95 patients were assigned to receive pomalidomide and 49 to receive placebo. … At 24 weeks, the mean difference between the pomalidomide group and the placebo group in the change from baseline in the Epistaxis Severity Score was −0.94 points (95% confidence interval [CI], −1.57 to −0.31; P = 0.004). … Adverse events that were more common in the pomalidomide group than in the placebo group included neutropenia, constipation, and rash."

"Among patients with HHT, pomalidomide treatment resulted in a significant, clinically relevant reduction in epistaxis severity."

Note pomalidomide's teratogenicity (thalidomide analog) — REMS-equivalent contraception requirements apply.

Paradigm shift (curate as a treatment-strategy claim): "This has led to a recent paradigm shift away from repetitive temporizing procedural management toward effective systemic medical therapeutics to treat bleeding in HHT" (PMID:38864625).

Procedural and surgical

Supportive, preventive and counseling

• Humidification, topical nasal moisturizers, hemostatic products.
• Antibiotic prophylaxis before dental and non-sterile invasive procedures when pulmonary shunting is present; air filters on IV lines to prevent paradoxical air embolism.
• Genetic counseling (NCIT:C15240) — in HHT4 this must cover the absence of a testable variant and the consequent reliance on clinical screening for at-risk relatives.
• Pregnancy: pre-conception PAVM/CAVM screening; treat sizable PAVMs in the second trimester; prefer iron repletion over transfusion for anemia (PMID:20301525; PMID:32894695).

Advanced therapeutics and pipeline

No gene therapy, cell therapy, RNA therapy, or gene editing is available for HHT. Gene-directed therapy is by definition impossible for HHT4 until the gene is identified — a clean, citable rationale for gene discovery. Kasthuri (ASH Education Program 2025, PMID:41347972) describes the pipeline:

"The initial clinical studies evaluating medications for the treatment of HHT have involved repurposing drugs that were previously approved for other indications. In the wake of these efforts, several therapies specifically for HHT are currently being developed and are in preclinical studies and early phase human trials or may soon start pivotal phase III trials."

Pharmacogenomics

None established for HHT. No CPIC/PharmGKB HHT guideline exists.

13. Prevention

• Primary prevention: not possible — the disease is germline and monogenic. Only reproductive options (preimplantation or prenatal genetic testing) prevent transmission, and these are unavailable in HHT4 because no variant is known. This is a concrete, citable harm of the unresolved locus.
• Secondary prevention (the main lever): presymptomatic detection of PAVMs and CAVMs in clinically at-risk relatives, followed by preemptive embolization to prevent stroke and brain abscess. The full screening cadence is in §10.
• Tertiary prevention: antibiotic prophylaxis and IV air filters in shunt carriers; avoidance of anticoagulants/NSAIDs where bleeding is significant; avoidance of scuba diving "unless TCE within the last five years was negative for evidence of a right-to-left shunt"; avoidance of vigorous nose blowing, heavy lifting, straining, and nasal digital manipulation; avoidance of liver biopsy (all PMID:20301525).
• Immunization: no vaccine relevance beyond routine care.
• Genetic screening: cascade molecular screening is standard in HHT1/HHT2/JP-HHT and impossible in HHT4; cascade clinical screening substitutes.
• Public health / environmental interventions: not applicable. The relevant public-health issue is underdiagnosis — supported by the Korean ascertained prevalence of ~1/500,000 against an expected ~1/5,000 (PMID:31455059) and the gnomAD-derived 2–12× upward revision (PMID:40964703).

14. Other Species / Natural Disease

• Taxonomy: Homo sapiens, NCBITaxon:9606. HHT4 is described only in humans.
• Breed: not applicable; no VBO term.
• Orthologous genes: cannot be specified — the causal gene is unknown, so no ortholog can be named. Any orthology statement in an HHT4 entry would be fabrication. (For the known HHT genes, mouse Eng, Acvrl1, Smad4, Gdf2 and zebrafish acvrl1 are the orthologs of interest.)
• Natural disease in other species: no naturally occurring HHT-equivalent is recorded in OMIA for any species for any HHT type. Vascular malformations occur in companion animals but are not established as an HHT homolog.
• Comparative biology: the BMP9/10–ALK1–endoglin–SMAD axis is deeply conserved across vertebrates — zebrafish acvrl1 disruption (the violet beauregarde mutant) produces cranial vessel endothelial-cell excess and shunting (Roman et al., Development 2002;129:3009–19, PMID:12050147), demonstrating conservation of the arteriovenous-patterning function down to teleosts.
• Zoonotic potential / cross-species transmission: not applicable (non-infectious, germline genetic).

15. Model Organisms

The HHT4-specific position

There is no HHT4 model of any kind — no mouse, no zebrafish, no cell line, no iPSC, no organoid, no computational model. Building one requires the gene. Any model listed under an HHT4 entry must be curated as a model of the shared HHT mechanism, and — under the dismech ModelMechanismLink semantics — should attach only to the downstream, pathway-level nodes (dysregulated BMP signaling → excess angiogenesis → AVM), never to the "7p14 locus lesion" trigger node, which no model reproduces. A HUMAN_MODEL_MISMATCH discussion is warranted: these models carry ENG/ACVRL1 lesions, and their relevance to HHT4 rests entirely on the unproven assumption that the HHT4 gene lies in the same pathway.

Models available for HHT (shared mechanism)

What models cannot capture (curate as limitations): epistaxis and the nasal mucosal microenvironment; the age-dependent human accrual of telangiectases; GI bleeding after age 50; the human-specific organ distribution of visceral AVMs; and — critically for this entry — anything at all about the HHT4 lesion.

Resources: MGI (mouse Eng, Acvrl1, Smad4), ZFIN (acvrl1), IMPC/KOMP, Alliance of Genome Resources.

Curation recommendations specific to this entry

1. Model the gene slot as an explicit locus, not an absence. Record 7p14, flanking markers D7S2252/D7S510, peak marker D7S817, LOD 3.60 — and a first-class statement that no causal gene is known.
2. Add a KNOWLEDGE_GAP discussion on gene identity, with proposed experiments: WGS (not WES) of the original pedigree with deep-intronic and structural-variant analysis of ACVRL1/ENG; re-contact and re-phenotyping; RNA-seq of patient endothelial cells for aberrant splicing.
3. Add a competing-interpretation discussion citing McDonald et al. (PMID:25674101): a deep intronic ACVRL1/ENG variant is a live alternative to a novel 7p14 gene, and is not excluded by the 2006 linkage data alone.
4. Add a HUMAN_MODEL_MISMATCH discussion for any model-organism evidence imported: all existing HHT models carry known-gene lesions.
5. Add an INTERPRETATION note on the HHT3/HHT4 transposition hazard (5q31.3–q32 vs 7p14) — this is exactly the Named Entity Confusion class that passes every automated check.
6. Do not populate datasets:. With no gene and no cohort, every discoverable accession would be a GENE_ONLY or wrong-disease hit.
7. Frequencies: omit frequency: on phenotypes rather than importing HHT-wide bands; the one HHT4-specific phenotypic claim is reduced severity of epistaxis and telangiectases.
8. Prevalence: curate as CASES_IN_LITERATURE / ULTRA_RARE with no numeric rate; put the 1/5,000 parent figure on the parent entry.
9. Treatments can be inherited from HHT in full with high confidence, because HHT management is gene-agnostic — but gene therapy and reproductive genetic testing must be recorded as unavailable for HHT4 specifically.

Evidence quality summary

Sources

• A fourth locus for hereditary hemorrhagic telangiectasia maps to chromosome 7 — PubMed (PMID:16969873)
• OMIM 610655 — TELANGIECTASIA, HEREDITARY HEMORRHAGIC, TYPE 4
• MedGen 341824 — Hereditary hemorrhagic telangiectasia type 4
• Hereditary Hemorrhagic Telangiectasia — GeneReviews (PMID:20301525, updated 19 Feb 2026)
• McDonald et al. 2015, HHT: genetics and molecular diagnostics in a new era (PMID:25674101)
• Al Tabosh et al. 2024, JCI — HHT: from signaling insights to therapeutic advances (PMID:38357927)
• Al-Samkari et al. 2024, NEJM — Pomalidomide for Epistaxis in HHT (PMID:39292928)
• Al-Samkari 2024, Blood — How I treat bleeding in HHT (PMID:38864625)
• Kasthuri 2025, ASH Education Program — What's new in HHT (PMID:41347972)
• Faughnan et al. 2020, Second International Guidelines for HHT (PMID:32894695)
• Jutant et al. 2026, ERJ — Pulmonary hypertension associated with HHT (PMID:41713948)
• Anzell et al. 2025, Circ Genom Precis Med — HHT prevalence from gnomAD (PMID:40964703)
• Kjeldsen et al. 2016, 20-year follow-up of Danish HHT patients (PMID:27876060)
• Dakeishi et al. 2002, Genetic epidemiology of HHT in northern Japan (PMID:11793473)
• Kim et al. 2019, HHT in South Korea (PMID:31455059)
• Cole et al. 2005, A new locus for HHT (HHT3) maps to chromosome 5 (PMID:15879500)
• Govani & Shovlin 2010, Fine mapping of the HHT3 locus (PMID:20701797)
• Park et al. 2009, JCI — Real-time imaging of de novo AVM in an HHT mouse model (PMID:19805914)
• Tual-Chalot et al. 2014, Endothelial depletion of Acvrl1 in mice (PMID:24896812)
• Ruiz et al. 2016, Transmammary immunoblocking of BMP9/BMP10 mouse model (PMID:27874028)
• Roman et al. 2002, Disruption of acvrl1 in zebrafish cranial vessels (PMID:12050147)
• MONDO:0012532 via EBI OLS4
• Orphanet — Hereditary hemorrhagic telangiectasia (ORPHA:774, parent concept)

Note on ontology terms: all HPO, GO, CL, UBERON, NCIT, CHEBI, HGNC, and MONDO identifiers presented in tables above were checked against this repository's local term caches and carry their canonical labels, except where marked (verify with OAK) — those are suggestions requiring just validate-terms confirmation before use.

Reference Validation

Checked with linkml-reference-validator 0.2.1.

Quotes not found in the cited source

Searched the abstract, any retrieved full text, and the title. A quote drawn from a part of the paper that was not retrieved will appear here too, so check before treating one as invented:

Every one of these was searched against an abstract alone, with no full text retrieved - marked abstract only below. Where full text can be fetched, re-running with it will settle them; where the source publishes only a summary to PubMed, as GeneReviews chapters do, it will not, and the quote has to be checked by hand against the chapter itself.

• PMID:38357927 (abstract only): "The BMP9-10/ENG/ALK1/SMAD4 signaling pathway maintains vascular quiescence by repressing angiogenic pathways"
• closest text in source: "HHT is caused by loss-of-function mutations in the BMP9-10/ENG/ALK1/SMAD4 signaling pathway"