Ocular Melanoma: Comprehensive Disease-Characteristics Report

Scope and category. “Ocular melanoma” is an umbrella term, not one homogeneous disease. It principally includes uveal melanoma (UM)—melanoma of the choroid, ciliary body, or iris—and conjunctival melanoma (Co-M), an ocular-surface mucosal melanoma with different embryology, drivers, metastatic routes, and treatment. This report therefore treats them separately, emphasizing UM because it accounts for approximately 80% of ocular melanomas and has the larger evidence base. The evidence is aggregated disease-level literature and registry/trial data, not individual electronic-health-record data. (butt2024conjunctivalmelanomaa pages 1-2, pasalic2023geneticandepigenetic pages 1-2)

Executive summary

UM is the commonest primary intraocular malignancy in adults, with incidence around 5–6 per million/year in the United States and Europe and marked enrichment in fair-skinned populations. Approximately 90% arise in the choroid. GNAQ/GNA11-pathway activation initiates most tumors; later BAP1, SF3B1, or EIF1AX alterations and chromosome 3/8q status largely determine metastatic risk. Local radiotherapy or surgery controls the ocular tumor, but roughly half of patients eventually develop hematogenous metastases, usually in the liver. Tebentafusp is the first systemic therapy to produce a randomized overall-survival benefit, but it applies only to HLA-A*02:01-positive unresectable/metastatic UM. Co-M is much rarer—approximately 0.46 cases per million/year—but its incidence is increasing; it often arises from conjunctival melanocytic intraepithelial lesions and has BRAF, NRAS, NF1, UV-related, and PD-L1 biology closer to cutaneous melanoma. (kastelan2024biologicalcharacteristicsand pages 2-3, fuentesrodriguez2024recentadvancesin pages 2-3, butt2024conjunctivalmelanomaa pages 1-2, hassel2023threeyearoverallsurvival pages 1-3)

1. Disease information

Definition and identifiers

• Uveal melanoma: malignant melanocytic neoplasm arising in the choroid, ciliary body, or iris. Confirmed identifier: MONDO:0006486. Open Targets identifies BAP1, MBD4, GNA11, GNAQ, and SF3B1 among its strongest disease-associated targets. (OpenTargets Search: uveal melanoma,ocular melanoma,conjunctival melanoma, lissak2024whatsetsuveal pages 1-3)
• Conjunctival melanoma: invasive melanoma arising from basal melanocytes of conjunctival epithelium. It is a mucosal/ocular-surface melanoma, not a subtype of UM. (butt2024conjunctivalmelanomaa pages 1-2)
• Synonyms: ocular melanoma, eye melanoma, intraocular melanoma, uveal malignant melanoma, choroidal melanoma, ciliary-body melanoma, iris melanoma, and conjunctival melanoma. “Ocular melanoma” should remain a parent term in a knowledge base.
• Coding: ICD-10-CM generally uses the site-specific malignant-neoplasm-of-eye family C69.x; ICD-O morphology is melanoma-specific and topography depends on choroid, ciliary body, iris, or conjunctiva. Exact ICD-10/ICD-11, MeSH, OMIM, and Orphanet cross-references should be curator-verified because no single code covers all ocular melanoma subtypes. UM is usually sporadic and therefore does not have one Mendelian OMIM disease entry equivalent to BAP1 tumor-predisposition syndrome.

A concise structured mapping is provided below.

Table: Compact knowledge-base mapping table contrasting uveal melanoma with conjunctival melanoma across disease identity, genes, anatomy, phenotypes, mechanisms, and treatments. It highlights confirmed and suggested ontology mappings while avoiding invention of uncertain IDs.

2. Etiology

Causal and susceptibility factors

UM is usually an acquired clonal cancer. Early activating mutations in GNAQ, GNA11, CYSLTR2, or PLCB4 initiate melanocytic proliferation; subsequent chromosomal and tumor-suppressor/splicing alterations govern malignant progression. Germline BAP1 pathogenic variants cause autosomal-dominant BAP1 tumor-predisposition syndrome, which increases risks of UM, cutaneous melanoma, mesothelioma, and clear-cell renal carcinoma. Rare inherited MBD4 and other DNA-repair predispositions are also reported, but most UM is not inherited. (OpenTargets Search: uveal melanoma,ocular melanoma,conjunctival melanoma, kulbay2024uvealmelanomacomprehensive pages 2-5, fuentesrodriguez2024recentadvancesin pages 2-3)

Established host associations include older age, fair skin, light iris color, poor tanning/sunburn sensitivity, iris or choroidal nevus, oculodermal melanocytosis/nevus of Ota, dysplastic-nevus phenotype, and family history of melanoma. Occupational associations with welding or irritant exposure have been reported, but causality is less certain. Direct solar causation remains debated and is substantially weaker than in cutaneous melanoma because most UM arises in the sun-shielded choroid and has low mutational burden. (sorrentino2024geneticfeaturesof pages 1-2, kulbay2024uvealmelanomacomprehensive pages 2-5, pasalic2023geneticandepigenetic pages 1-2)

For Co-M, fair skin, older age, UV exposure/signatures, and precursor conjunctival melanocytic intraepithelial lesion are important. Approximately 70% arise from C-MIN/PAM with atypia; the remainder arise from nevi or de novo. BRAF occurs in approximately 30% and NRAS in approximately 14–25%; NRAS-mutant disease may have greater metastatic risk. (butt2024conjunctivalmelanomaa pages 1-2)

Protective factors and gene–environment interaction

No genetic variant, diet, medication, or lifestyle intervention is proven to prevent UM. UV-protective eyewear is sensible for general ocular health and may be more biologically relevant to conjunctival/iris disease, but evidence that it prevents posterior UM is insufficient. A plausible gene–environment distinction is that UV exposure contributes more strongly to Co-M’s BRAF/NRAS/NF1-like landscape, whereas inherited pigmentation phenotype and rare BAP1 susceptibility interact with largely non-UV initiating events in UM. Smoking, alcohol, infection, exercise, and diet are not established causal or protective determinants.

3. Phenotypes

UM is often insidious and unilateral. Up to approximately 30% of patients are asymptomatic and diagnosed on routine ophthalmic examination. Symptoms depend on size and location: blurred or reduced vision, photopsias, floaters, visual-field loss, metamorphopsia, pain, and occasionally a visible iris lesion. Exudation, macular involvement, vitreous hemorrhage, or retinal detachment can produce severe or progressive vision loss. Iris tumors may present 10–20 years earlier than posterior tumors and cause heterochromia, corectopia, secondary glaucoma, or a growing pigmented lesion. (kastelan2024biologicalcharacteristicsand pages 1-2, kastelan2024biologicalcharacteristicsand pages 2-3)

Suggested HPO annotations include decreased visual acuity, visual-field defect, photopsia, vitreous floaters, retinal detachment, ocular pain, heterochromia iridis, corectopia, secondary glaucoma, and unilateral ocular abnormality. Frequencies beyond the approximately 30% asymptomatic estimate vary substantially by tumor site and referral population.

Co-M usually presents as a growing amelanotic, brown, or black conjunctival lesion—most often bulbar and near the limbus—sometimes with feeder vessels, irritation, or invasion of eyelid/orbit. It can cause loss of vision or eye, facial disfigurement, and death. Recurrence occurs in approximately 33–45%, regional nodal metastasis in approximately 25%, and reported five-year disease-specific mortality is approximately 27%. Suggested HPO terms include conjunctival pigmentation, conjunctival mass, decreased visual acuity, recurrent neoplasm, and lymph-node metastasis. (butt2024conjunctivalmelanomaa pages 1-2)

Quality-of-life burdens include visual disability, monocular depth-perception loss, treatment-related retinopathy/optic neuropathy, cosmetic change, anxiety, depression, and fear of recurrence. A 2024 French prospective protocol is specifically measuring HADS, FCRI, EORTC QLQ-C30, QLQ-OPT30, information satisfaction, and communication in 250 UM survivors, illustrating that psychological surveillance is now a recognized component of care. (kastelan2024biologicalcharacteristicsand pages 1-2)

4. Genetic and molecular information

Somatic drivers and prognostic alterations

• GNAQ/GNA11: mutually exclusive gain-of-function mutations, usually at Q209 or R183, occur collectively in approximately 85–94% of UM; individual summaries report GNA11 around 55% and GNAQ around 40%. These are early events found even in benign uveal nevi and are not by themselves strong metastatic predictors. (kastelan2024biologicalcharacteristicsand pages 2-3, fuentesrodriguez2024recentadvancesin pages 2-3)
• CYSLTR2 p.Leu129 occurs in approximately 2–4%, usually in GNAQ/GNA11-wild-type tumors. PLCB4 p.Asp630 occurs in approximately 2.5%. Both are activating initiating events. (fuentesrodriguez2024recentadvancesin pages 2-3)
• BAP1: somatic loss-of-function—nonsense, frameshift, splice, missense, deletion, or loss of chromosome 3—is associated with class-2 phenotype, epithelioid morphology, early metastasis, and poor survival. One 2024 synthesis reports BAP1 alteration in approximately 38% of primary and 84% of metastatic samples. Germline pathogenic variants are rare and absent from or extremely rare in general-population databases. (kastelan2024biologicalcharacteristicsand pages 2-3, lissak2024whatsetsuveal pages 3-7)
• SF3B1: recurrent hotspot missense variants occur in approximately 20–25%, alter RNA splicing, and confer intermediate/late metastatic risk. EIF1AX variants occur in approximately 8–13% and usually mark lower-risk, disomy-3 tumors. These alterations are generally mutually exclusive with BAP1 loss. (kastelan2024biologicalcharacteristicsand pages 2-3, lissak2024whatsetsuveal pages 3-7)
• MBD4: biallelic DNA-glycosylase loss creates a hypermutated subset and may increase immunogenicity; Open Targets ranks MBD4 strongly among UM disease associations. (OpenTargets Search: uveal melanoma,ocular melanoma,conjunctival melanoma)

Somatic-driver allele frequencies are tumor frequencies, not population frequencies. Germline variant classification must use ClinVar/ClinGen and ACMG/AMP criteria; tumor-only sequencing cannot establish germline origin. VUS should not direct prophylactic surgery or family testing without validated reclassification.

Chromosomal and epigenetic abnormalities

Monosomy 3, particularly with 8q gain/amplification, is the canonical high-risk cytogenetic pattern. 6p gain is generally associated with a more favorable disomy-3 class; 6q loss and complex 8q alterations occur in intermediate-risk/SF3B1 tumors. BAP1 loss reshapes chromatin and DNA methylation; class-1 versus class-2 UM has distinct methylation, transcriptomic, miRNA, and histone-regulatory programs. Altered miRNAs are promising diagnostic/prognostic biomarkers but are not yet standard standalone tests. (fuentesrodriguez2024recentadvancesin pages 2-3, pasalic2023geneticandepigenetic pages 1-2)

DecisionDx-UM’s 12-gene expression profile stratifies tumors into class 1A, 1B, and 2, with reported five-year metastatic risks of approximately 2%, 21%, and 72%, respectively. TCGA-style integration further divides UM into four molecular groups spanning disomy-3/EIF1AX through monosomy-3/BAP1/8q-amplified disease. These are prognostic—not proof that adjuvant systemic therapy improves survival. (fuentesrodriguez2024recentadvancesin pages 2-3)

5. Environmental information

No infectious agent is known to cause UM or Co-M; vaccination and antimicrobial prophylaxis are therefore not applicable. UV radiation has uncertain relevance to posterior UM but a more convincing relationship to conjunctival melanoma. Welding and selected occupational exposures are epidemiological signals rather than established sufficient causes. There is no reproducible evidence that tobacco, alcohol, obesity, diet, or physical inactivity materially changes UM risk. (kulbay2024uvealmelanomacomprehensive pages 2-5, butt2024conjunctivalmelanomaa pages 1-2)

6. Mechanism and pathophysiology

Causal chain

1. A uveal melanocyte acquires activating GNAQ/GNA11, CYSLTR2, or PLCB4 alteration.
2. Constitutive Gαq/11–TRIO–Rho and PLCβ signaling activates PKC, RASGRP3–RAF–MEK–ERK, PI3K–AKT–mTOR, and YAP/TAZ programs, promoting proliferation, survival, motility, calcium signaling, and metabolic adaptation.
3. A later BAP1, SF3B1, or EIF1AX event plus chromosome 3/6/8 evolution establishes prognostic phenotype. BAP1 loss disrupts deubiquitination, chromatin regulation, DNA repair, calcium homeostasis, differentiation, and metabolism.
4. Tumor cells undergo extracellular-matrix remodeling, transendothelial migration, and hematogenous dissemination; absence of ocular lymphatics helps explain UM’s blood-borne, liver-dominant route.
5. Dormant hepatic micrometastases may remain clinically occult for years before angiogenic and immune escape produces detectable disease. (kulbay2024uvealmelanomacomprehensive pages 2-5, fuentesrodriguez2024recentadvancesin pages 2-3, pasalic2023geneticandepigenetic pages 1-2)

Suggested GO terms include MAPK cascade, protein kinase C signaling, phosphatidylinositol-mediated signaling, TOR signaling, cell proliferation, negative regulation of apoptosis, chromatin organization, DNA repair, RNA splicing, angiogenesis, cell migration, extracellular-matrix organization, and immune-response regulation.

Immune, tissue, and metabolic biology

The eye is immune privileged, and UM generally has low tumor mutational burden—approximately 0.5 mutations/Mb and a median of about 32 coding mutations in one clinical synthesis. Infiltration by lymphocytes and macrophages, HLA-I/II upregulation, NF-κB activation, LAG-3, and galectin-3 can paradoxically mark aggressive disease. Tumor-derived extracellular vesicles promote proliferation, migration, and invasion; circulating hybrid cells and ctDNA are emerging markers. (carvajal2022clinicalandmolecular pages 1-2, kulbay2024uvealmelanomacomprehensive pages 2-5)

A 2024 single-cell/bulk study analyzed 37,660 malignant cells from 17 tumors, identifying substantial intratumoral transcriptional heterogeneity and two states with different prognosis and immune context. A separate 2024 study combined single-cell RNA/TCR sequencing with metastatic PDX and coculture experiments, finding tumor-reactive T cells among activated, exhausted, and cytotoxic-effector populations—supporting rational TIL/TCR selection. These are human-tissue plus experimental-model findings, not yet validated clinical diagnostics. (karlsson2024patientderivedxenograftsand pages 1-2)

7. Anatomical structures affected

Primary UM affects the uvea: choroid approximately 90%, ciliary body approximately 7%, and iris approximately 2–3%. Secondary local structures include retina, macula, optic disc/nerve, vitreous, sclera, anterior chamber, and orbit. UM is usually unilateral. Dissemination most often affects liver (approximately 89% or more), followed by lung and bone; one review reports lung 29% and bone 17%. (kulbay2024uvealmelanomacomprehensive pages 2-5, pasalic2023geneticandepigenetic pages 1-2)

Co-M begins in conjunctival epithelium, commonly bulbar conjunctiva/limbus, and may invade cornea, eyelid, lacrimal drainage structures, orbit, regional nodes, lung, liver, or brain. (butt2024conjunctivalmelanomaa pages 1-2)

Suggested UBERON terms: eye, uvea, choroid, ciliary body, iris, retina, sclera, conjunctiva, bulbar conjunctiva, orbit, and liver. Suggested CL terms: melanocyte, endothelial cell, fibroblast, macrophage, CD8-positive T cell, and hepatic stellate cell. Relevant subcellular GO compartments include plasma membrane, nucleus/chromatin, spliceosomal complex, mitochondrion, and extracellular vesicle.

8. Temporal development

UM is predominantly adult/late-adult onset; median diagnosis is approximately 58–62 years, whereas iris melanoma tends to present 10–20 years earlier. Onset is usually chronic and clinically silent. AJCC eighth-edition staging incorporates tumor size/category, ciliary-body involvement, and extraocular extension; metastatic disease is stage IV. (kastelan2024biologicalcharacteristicsand pages 1-2, kastelan2024biologicalcharacteristicsand pages 2-3)

At primary diagnosis, fewer than 2% have radiologically detectable metastases, yet 32–45% may develop them within 15 years and some recur more than 30 years later. This supports an early-dissemination/dormancy model and lifelong risk-adapted follow-up. Spontaneous durable remission is exceptional; treatment-induced local control is common, but eradication of occult micrometastases is not assured. (NCT05502900 chunk 1, pasalic2023geneticandepigenetic pages 1-2)

9. Inheritance and population

UM incidence is approximately 5–6 per million/year in the United States/Europe, about 7 per million in Australia, and only 0.2–0.3 per million/year in much of Asia and Africa. A 2024 synthesis gives a range of 4.9–7.4 per million in high-incidence populations. Most patients are White/Caucasian; sex differences are small and inconsistent. (kastelan2024biologicalcharacteristicsand pages 2-3, pasalic2023geneticandepigenetic pages 1-2)

Most UM is sporadic and multifactorial. BAP1 tumor-predisposition syndrome is autosomal dominant, incompletely penetrant, age-dependent, and variably expressive; anticipation and consanguinity are not characteristic. Founder variants may exist in particular families/populations, but there is no population-wide “carrier frequency” suitable for general screening. Germline testing is most appropriate for young onset, bilateral/multifocal UM, strong family history, or personal/family histories of mesothelioma, renal-cell carcinoma, cutaneous melanoma, or BAP1-inactivated melanocytic tumors.

Co-M incidence is approximately 0.46 per million/year, with an increasing rate ratio around 1.4 and a particularly sharp rise after age 65. It represents approximately 0.25% of all melanomas and 5% of ocular melanomas. (butt2024conjunctivalmelanomaa pages 1-2)

10. Diagnostics

Clinical diagnosis

UM is often diagnosed clinically by an ocular oncologist using dilated fundus examination, slit-lamp examination for anterior tumors, color fundus photography, optical coherence tomography, fundus autofluorescence, and A-/B-scan ultrasonography. MRI can characterize selected lesions; systemic CT/MRI/ultrasound evaluates metastases, especially liver. Biopsy is not always required for a classic lesion but is used for uncertain diagnosis and molecular prognostication.

Histology shows spindle, mixed, or epithelioid melanoma; epithelioid morphology, high mitotic activity, closed vascular loops, ciliary-body involvement, and extrascleral extension are adverse features. Immunohistochemistry includes melanocytic markers such as SOX10, S100, Melan-A/MART1, HMB45, and nuclear BAP1. Co-M requires excisional biopsy when feasible, careful margin/orientation assessment, and evaluation for pagetoid intraepithelial spread. (butt2024conjunctivalmelanomaa pages 1-2, pasalic2023geneticandepigenetic pages 1-2)

Differential diagnosis includes choroidal nevus, congenital hypertrophy of retinal pigment epithelium, melanocytoma, hemangioma, metastasis, lymphoma, retinal-pigment-epithelium lesions, inflammatory granuloma, and hemorrhagic retinal detachment. Co-M differentials include conjunctival nevus, C-MIN/PAM, complexion-associated melanosis, foreign-body pigmentation, and ocular-surface squamous neoplasia.

Genetic and omics testing

Fine-needle aspiration or resection tissue may undergo chromosome 3/8/6 testing by FISH, SNP array, MLPA, or NGS; BAP1/SF3B1/EIF1AX sequencing; and validated GEP. Broad WES/WGS is useful for atypical cases, research, or metastatic precision oncology but is not required for every primary tumor. CMA can detect copy-number changes; conventional karyotyping has limited sensitivity; mitochondrial and repeat-expansion testing are not relevant. (fuentesrodriguez2024recentadvancesin pages 2-3)

Blood ctDNA, circulating tumor cells, extracellular vesicles, miRNA, and circulating hybrid cells are promising for disease monitoring. Early ctDNA decline during tebentafusp correlates with survival, but liquid biopsy does not yet replace imaging or tissue-based risk classification. Normal liver-function tests cannot exclude hepatic metastasis. (carvajal2022clinicalandmolecular pages 1-2, pasalic2023geneticandepigenetic pages 1-2)

There is no population screening program. High-risk BAP1 families merit genetic counseling, cascade testing for a confirmed pathogenic familial variant, dermatologic/ophthalmic surveillance, and syndrome-specific renal/mesothelioma surveillance.

11. Outcome and prognosis

Overall five-year survival for UM is often reported at 50–70%, with localized-disease estimates around 70–80%. Approximately half ultimately metastasize. Historical median survival after metastatic diagnosis is approximately 6–12 months, with liver involvement driving mortality; older series report nearly 90% mortality by two years after hepatic metastasis. (kastelan2024biologicalcharacteristicsand pages 2-3, pasalic2023geneticandepigenetic pages 1-2)

Adverse prognostic factors are large basal diameter/thickness, ciliary-body involvement, extraocular extension, epithelioid morphology, high mitotic rate, monosomy 3, 8q gain, BAP1 loss, class-2 GEP, elevated LDH, high hepatic tumor burden, and poor performance status. Favorable factors include small iris-confined disease, disomy 3, 6p gain, EIF1AX mutation, and class-1A GEP. SF3B1 generally denotes intermediate and sometimes late relapse. (lissak2024whatsetsuveal pages 3-7, fuentesrodriguez2024recentadvancesin pages 2-3, pasalic2023geneticandepigenetic pages 1-2)

Morbidity includes irreversible visual-field loss, radiation retinopathy, maculopathy, optic neuropathy, cataract, glaucoma, dry eye, enucleation-related monocular disability, and psychological distress. Co-M adds repeated surface surgery, limbal-stem-cell injury, scarring, and possible orbital exenteration. (kastelan2024biologicalcharacteristicsand pages 1-2, butt2024conjunctivalmelanomaa pages 1-2)

12. Treatment

Localized UM

Management is individualized by size, location, visual potential, extraocular extension, and patient preference. Options include observation of selected indeterminate/small lesions; plaque brachytherapy; proton-beam or stereotactic radiotherapy; transpupillary thermotherapy as an adjunct in selected small lesions; local resection; and enucleation for very large, painful, blind, or extensively invasive tumors. Local control does not eliminate pre-existing micrometastases. Suggested NCIT concepts include Plaque Brachytherapy, Proton Radiation Therapy, Local Excision, and Enucleation. (kastelan2024biologicalcharacteristicsand pages 2-3, pasalic2023geneticandepigenetic pages 1-2)

Conjunctival melanoma

Preferred treatment is “no-touch” complete excision with margin control, often with adjuvant cryotherapy. Topical mitomycin-C or interferon, plaque/proton/photon radiotherapy, and exenteration are selected by intraepithelial spread, margins, multifocality, and invasion. Lifelong ocular and nodal surveillance is warranted. BRAF/MEK inhibition or PD-1-based immunotherapy may be considered for molecularly appropriate unresectable/metastatic disease, but evidence remains mainly case reports and small series. (butt2024conjunctivalmelanomaa pages 1-2)

Metastatic UM

Tebentafusp is an engineered gp100–HLA-A02:01 T-cell-receptor/CD3 bispecific and the preferred evidence-based systemic option for eligible HLA-A02:01-positive, unresectable/metastatic UM. In the phase III long-term analysis, median overall survival was 21.6 versus 16.9 months with control (HR 0.68, 95% CI 0.54–0.87), and three-year survival was 27% versus 18%. Common adverse events were rash 83%, pyrexia 76%, pruritus 70%, and hypotension 38%; only 2% discontinued for toxicity, and there were no treatment-related deaths. Publication: 14 December 2023; DOI/URL: https://doi.org/10.1056/NEJMoa2304753; NCT03070392. The abstract concludes that the analysis “supported a continued long-term benefit of tebentafusp for overall survival.” (hassel2023threeyearoverallsurvival pages 1-3)

In a 127-patient phase II refractory cohort, objective response was only 5%, but one-year survival was 62% and median survival 16.8 months, illustrating that RECIST response underestimates benefit. Early ctDNA reduction correlated with survival. Publication: October 2022; DOI/URL: https://doi.org/10.1038/s41591-022-02015-7; NCT02570308. (carvajal2022clinicalandmolecular pages 1-2)

Checkpoint inhibitors—pembrolizumab/nivolumab, ipilimumab, or combinations—have substantially less activity than in cutaneous melanoma because UM is low-TMB and immunosuppressive, but they remain options when tebentafusp is unavailable/inapplicable or in trials. Cytotoxic chemotherapy has low response rates. Liver-dominant disease may be treated with resection/ablation in selected oligometastatic cases, embolization/radioembolization, immunoembolization, isolated or percutaneous hepatic perfusion with melphalan, or other center-specific liver-directed approaches. Multidisciplinary sequencing is essential. (kulbay2024uvealmelanomacomprehensive pages 2-5, pasalic2023geneticandepigenetic pages 1-2, hassel2023threeyearoverallsurvival pages 1-3)

Active/late-phase development

• Darovasertib, an oral selective PKC inhibitor: phase II neoadjuvant/adjuvant trial in 160 localized-UM patients, testing tumor shrinkage, conversion from enucleation to radiation, reduction of radiation dose to critical structures, and long-term metastasis outcomes; NCT05907954. (NCT05907954 chunk 1)
• Belzupacap sarotalocan/AU-011: randomized, double-masked phase III suprachoroidal drug-plus-laser photoactivation study in 108 patients with indeterminate lesions or small choroidal melanoma; NCT06007690. (NCT06007690 chunk 1)
• Adjuvant melatonin: randomized open-label phase III trial of 20 mg nightly for five years in 100 high-risk patients, with metastasis incidence as the primary endpoint; NCT05502900. This is experimental, not preventive standard care. (NCT05502900 chunk 1)

No validated CPIC-style pharmacogenomic dosing guideline, approved gene therapy, CAR-T product, RNA therapy, or stem-cell therapy currently exists for ocular melanoma.

13. Prevention

Primary prevention: no intervention is proven to prevent UM. Sun-safe behavior and UV-blocking eyewear are reasonable, particularly for conjunctival/iris health, but should not be represented as proven posterior-UM prevention. There is no vaccine or chemoprophylaxis.

Secondary prevention: no population screening is recommended because the disease is rare. Routine eye examinations can detect asymptomatic tumors; targeted surveillance is appropriate for choroidal nevi with suspicious growth features, oculodermal melanocytosis, and BAP1 families. Prompt referral of suspicious conjunctival pigmentation prevents diagnostic delay.

Tertiary prevention: preserve vision through timely local treatment and manage radiation retinopathy, glaucoma, cataract, and psychosocial morbidity. Molecular risk stratification guides hepatic imaging every 3–12 months depending on risk; the 2024 GEP review suggests annual imaging for class 1A, every 6–12 months for class 1B, and every 3–6 months for class 2, although schedules vary by guideline and country. (fuentesrodriguez2024recentadvancesin pages 2-3)

14. Other species and natural disease

Naturally occurring ocular melanocytic neoplasms occur in dogs, cats, and horses. Canine anterior-uveal melanoma is often locally invasive but biologically less predictably metastatic than human posterior UM; feline diffuse iris melanoma can cause glaucoma and metastasis; equine ocular melanocytic disease has breed/color associations. These conditions are veterinary diseases and useful comparative pathology, but they are not exact orthologous models of human GNAQ/GNA11-driven, liver-tropic UM. No infectious transmission or zoonotic potential exists. Suggested taxa include Homo sapiens (NCBI:9606), Canis lupus familiaris (9615), Felis catus (9685), Equus caballus (9796), Mus musculus (10090), and Danio rerio (7955). Breed-specific VBO mappings require veterinary-database curation.

15. Model organisms

Model systems include established UM cell lines, three-dimensional spheroids/organoids, primary cultures, chicken chorioallantoic membrane assays, zebrafish xenografts/transgenics, mouse subcutaneous and orthotopic xenografts, PDX, syngeneic models, and genetically engineered GNAQ/GNA11-pathway models.

A 2023 zebrafish PDX platform generated spheroids from primary human UM within 24 hours, retained melanocytic markers, and produced a reproducible metastatic phenotype after intravenous implantation. Experiments used at least two biological replicates with more than 20 fish each; navitoclax and everolimus demonstrated utility for rapid drug-response screening. Publication: 15 April 2023; DOI/URL: https://doi.org/10.3390/ph16040598. The abstract states that the model “recapitulated molecular features of the disseminating UM.” (yin2023zebrafishpatientderivedxenograft pages 1-2)

Mouse PDX preserves patient-specific architecture/genomics and supports pharmacology, while orthotopic models reproduce ocular growth. Limitations include immunodeficiency, variable engraftment, cost, long latency, and incomplete hepatic tropism. Syngeneic models retain immunity but frequently use cutaneous melanoma cells whose genetics differ from UM. Zebrafish enables live imaging, small sample requirements, and high-throughput screening but differs in temperature, pharmacokinetics, adaptive immunity, and ocular/liver physiology. No single model reproduces the complete human genetic, histologic, immune, dormancy, and metastatic phenotype; convergent validation across organoid, zebrafish, PDX, and immune-competent systems is preferable. (yin2023zebrafishpatientderivedxenograft pages 1-2, karlsson2024patientderivedxenograftsand pages 1-2)

Evidence-quality and curation notes

The strongest treatment evidence is the randomized phase III tebentafusp trial. Epidemiology, genetics, and natural history are supported by large aggregated cohorts and recent 2023–2024 reviews, whereas protective factors, Co-M systemic therapy, adjuvant prevention, liquid-biopsy surveillance, and many multi-omics signatures remain investigational. Direct abstract quotations were limited to short passages to preserve context. DOI URLs and publication dates are provided where retrieved; PMIDs were not consistently present in the retrieved metadata and should be added through PubMed cross-linking rather than inferred. Exact HPO, UBERON, CL, GO, NCIT, ICD-11, OMIM, Orphanet, and MeSH identifiers flagged as “suggested” should undergo ontology-service validation before production ingestion.

References

1. (butt2024conjunctivalmelanomaa pages 1-2): Karam Butt, Rumana Hussain, Sarah Ellen Coupland, and Yamini Krishna. Conjunctival melanoma: a clinical review and update. Cancers, Sep 2024. URL: https://doi.org/10.3390/cancers16183121, doi:10.3390/cancers16183121. This article has 20 citations.

2. (pasalic2023geneticandepigenetic pages 1-2): Daria Pašalić, Tamara Nikuševa-Martić, Ankica Sekovanić, and Snježana Kaštelan. Genetic and epigenetic features of uveal melanoma—an overview and clinical implications. International Journal of Molecular Sciences, 24:12807, Aug 2023. URL: https://doi.org/10.3390/ijms241612807, doi:10.3390/ijms241612807. This article has 23 citations.

3. (kastelan2024biologicalcharacteristicsand pages 2-3): SNJEŽANA KAŠTELAN, ANA DIDOVIĆ PAVIČIĆ, DARIA PAŠALIĆ, TAMARA NIKUŠEVA-MARTIĆ, SAMIR ČANOVIĆ, PETRA KOVAČEVIĆ, and SUZANA KONJEVODA. Biological characteristics and clinical management of uveal and conjunctival melanoma. Oncology Research, 32:1265-1285, Jul 2024. URL: https://doi.org/10.32604/or.2024.048437, doi:10.32604/or.2024.048437. This article has 9 citations and is from a peer-reviewed journal.

4. (fuentesrodriguez2024recentadvancesin pages 2-3): Aurélie Fuentes-Rodriguez, Andrew Mitchell, Sylvain L. Guérin, and Solange Landreville. Recent advances in molecular and genetic research on uveal melanoma. Cells, 13:1023, Jun 2024. URL: https://doi.org/10.3390/cells13121023, doi:10.3390/cells13121023. This article has 27 citations.

5. (hassel2023threeyearoverallsurvival pages 1-3): Jessica C. Hassel, Sophie Piperno-Neumann, Piotr Rutkowski, Jean-Francois Baurain, Max Schlaak, Marcus O. Butler, Ryan J. Sullivan, Reinhard Dummer, John M. Kirkwood, Marlana Orloff, Joseph J. Sacco, Sebastian Ochsenreither, Anthony M. Joshua, Lauris Gastaud, Brendan Curti, Josep M. Piulats, April K.S. Salama, Alexander N. Shoushtari, Lev Demidov, Mohammed Milhem, Bartosz Chmielowski, Kevin B. Kim, Richard D. Carvajal, Omid Hamid, Laura Collins, Koustubh Ranade, Chris Holland, Constance Pfeiffer, and Paul Nathan. Three-year overall survival with tebentafusp in metastatic uveal melanoma. New England Journal of Medicine, 389:2256-2266, Dec 2023. URL: https://doi.org/10.1056/nejmoa2304753, doi:10.1056/nejmoa2304753. This article has 341 citations and is from a highest quality peer-reviewed journal.

6. (OpenTargets Search: uveal melanoma,ocular melanoma,conjunctival melanoma): Open Targets Query (uveal melanoma,ocular melanoma,conjunctival melanoma, 31 results). Buniello, A. et al. (2025). Open Targets Platform: facilitating therapeutic hypotheses building in drug discovery. Nucleic Acids Research.

7. (lissak2024whatsetsuveal pages 1-3): Karina Lissak, Zuzanna Szczepaniak, Agata Konopka, Natalia Wdowiak, Martyna Choinka, Małgorzata Komarów, Dominika Karasińska, Jakub Kalisiak, Mateusz Koralewicz, Milena Orzeł, and Bartłomiej Orzeł. What sets uveal melanoma apart, and how can we address it? a comprehensive review of pathophysiology, diagnosis and treatment. Quality in Sport, 23:55123, Sep 2024. URL: https://doi.org/10.12775/qs.2024.23.55123, doi:10.12775/qs.2024.23.55123. This article has 1 citations.

8. (kulbay2024uvealmelanomacomprehensive pages 2-5): Merve Kulbay, Emily Marcotte, Raheem Remtulla, Tsz Hin Alexander Lau, Manuel Paez-Escamilla, Kevin Y. Wu, and Miguel N. Burnier. Uveal melanoma: comprehensive review of its pathophysiology, diagnosis, treatment, and future perspectives. Biomedicines, 12:1758, Aug 2024. URL: https://doi.org/10.3390/biomedicines12081758, doi:10.3390/biomedicines12081758. This article has 48 citations.

9. (lissak2024whatsetsuveal pages 3-7): Karina Lissak, Zuzanna Szczepaniak, Agata Konopka, Natalia Wdowiak, Martyna Choinka, Małgorzata Komarów, Dominika Karasińska, Jakub Kalisiak, Mateusz Koralewicz, Milena Orzeł, and Bartłomiej Orzeł. What sets uveal melanoma apart, and how can we address it? a comprehensive review of pathophysiology, diagnosis and treatment. Quality in Sport, 23:55123, Sep 2024. URL: https://doi.org/10.12775/qs.2024.23.55123, doi:10.12775/qs.2024.23.55123. This article has 1 citations.

10. (karlsson2024patientderivedxenograftsand pages 1-2): Joakim Karlsson, Vasu R. Sah, Roger Olofsson Bagge, Munir Iqbal, Samuel Alsén, Sofia Stenqvist, Alka Saxena, Lars Ny, Lisa M. Nilsson, and Jonas A. Nilsson. Patient-derived xenografts and single-cell sequencing identifies three subtypes of tumor-reactive lymphocytes in uveal melanoma metastases. eLife, May 2024. URL: https://doi.org/10.7554/elife.91705, doi:10.7554/elife.91705. This article has 6 citations and is from a domain leading peer-reviewed journal.

11. (kastelan2024biologicalcharacteristicsand pages 1-2): SNJEŽANA KAŠTELAN, ANA DIDOVIĆ PAVIČIĆ, DARIA PAŠALIĆ, TAMARA NIKUŠEVA-MARTIĆ, SAMIR ČANOVIĆ, PETRA KOVAČEVIĆ, and SUZANA KONJEVODA. Biological characteristics and clinical management of uveal and conjunctival melanoma. Oncology Research, 32:1265-1285, Jul 2024. URL: https://doi.org/10.32604/or.2024.048437, doi:10.32604/or.2024.048437. This article has 9 citations and is from a peer-reviewed journal.

12. (NCT05907954 chunk 1): (Neo)Adjuvant IDE196 (Darovasertib) in Patients With Localized Ocular Melanoma. IDEAYA Biosciences. 2023. ClinicalTrials.gov Identifier: NCT05907954

13. (NCT06007690 chunk 1): A Phase 3 Randomized, Masked, Controlled Trial to Evaluate Efficacy and Safety of Belzupacap Sarotalocan (AU-011) Treatment Compared to Sham Control in Subjects With Primary Indeterminate Lesions or Small Choroidal Melanoma. Aura Biosciences. 2023. ClinicalTrials.gov Identifier: NCT06007690

14. (NCT05502900 chunk 1): Gustav Stalhammar. Adjuvant Melatonin for Uveal Melanoma. Gustav Stalhammar. 2022. ClinicalTrials.gov Identifier: NCT05502900

15. (yin2023zebrafishpatientderivedxenograft pages 1-2): Jie Yin, Gangyin Zhao, Helen Kalirai, Sarah E. Coupland, Aart G. Jochemsen, Gabriel Forn-Cuní, Annemijn P. A. Wierenga, Martine J. Jager, B. Ewa Snaar-Jagalska, and Arwin Groenewoud. Zebrafish patient-derived xenograft model as a preclinical platform for uveal melanoma drug discovery. Pharmaceuticals, 16:598, Apr 2023. URL: https://doi.org/10.3390/ph16040598, doi:10.3390/ph16040598. This article has 22 citations.

16. (sorrentino2024geneticfeaturesof pages 1-2): Francesco Saverio Sorrentino, Carola Culiersi, Antonio Florido, Katia De Nadai, Ginevra Giovanna Adamo, Francesco Nasini, Chiara Vivarelli, Marco Mura, and Francesco Parmeggiani. Genetic features of uveal melanoma. Genes, 15:1356, Oct 2024. URL: https://doi.org/10.3390/genes15111356, doi:10.3390/genes15111356. This article has 12 citations.

17. (carvajal2022clinicalandmolecular pages 1-2): Richard D. Carvajal, Marcus O. Butler, Alexander N. Shoushtari, Jessica C. Hassel, Alexandra Ikeguchi, Leonel Hernandez-Aya, Paul Nathan, Omid Hamid, Josep M. Piulats, Matthew Rioth, Douglas B. Johnson, Jason J. Luke, Enrique Espinosa, Serge Leyvraz, Laura Collins, Howard M. Goodall, Koustubh Ranade, Chris Holland, Shaad E. Abdullah, Joseph J. Sacco, and Takami Sato. Clinical and molecular response to tebentafusp in previously treated patients with metastatic uveal melanoma: a phase 2 trial. Nature Medicine, 28:2364-2373, Oct 2022. URL: https://doi.org/10.1038/s41591-022-02015-7, doi:10.1038/s41591-022-02015-7. This article has 196 citations and is from a highest quality peer-reviewed journal.

Artifacts

• Edison artifact artifact-00