A Comprehensive Guide to Uveal Melanoma: Diagnosis, Molecular Classification and the Current Treatment Algorithm

A Comprehensive Clinical Synthesis: From Molecular Classification to the Era of Tebentafusp and PKC Inhibitors

Uveal melanoma arises from the iris, ciliary body and choroid. Posterior uveal melanoma is biologically distinct from cutaneous melanoma — with different driver mutations, a different metastatic pattern (liver tropism) and different treatment responses. The key to understanding this distinction is the disease's two-layer genetic architecture.

The first step in diagnosis is assessment by an experienced ocular oncologist, including a dilated fundus examination and slit-lamp biomicroscopy. Uveal melanoma is one of the rare solid tumours diagnosed primarily through expert clinical examination and multimodal ocular imaging — without a mandatory biopsy. Distinguishing a small melanoma from a naevus can be difficult; documenting growth on serial follow-up is therefore crucial.

Because the liver is the principal target of haematogenous spread in uveal melanoma, staging imaging should also focus on this organ. Systemic metastasis is detected in only ~2–3% of patients at diagnosis, and the first detected site is usually the liver.

The 8th edition of AJCC defines tumour size categories for choroidal and ciliary body melanoma using thickness and largest basal diameter (categories 1–4); the T classification also incorporates ciliary body involvement and extrascleral extension. In metastatic disease, the distinction between M1a–M1c is based on the diameter of the largest metastatic lesion. However, AJCC alone is insufficient — its prognostic power increases substantially when chromosome 3 and 8q status are added.

Translating these data into practical risk groups determines both surveillance intensity and referral to clinical trials:

Treatment selection weighs four factors together: tumour location, size/thickness, the eye's visual potential and metastatic status. The flowchart below summarises the main decision points from localised to metastatic disease.

5a. Local treatment for localised disease

Small / borderline naevus-melanoma lesions: In the absence of growth or high-risk clinical signs, short-interval observation is acceptable — because the management of small choroidal melanomas remains controversial, and immediate treatment of every small lesion has not been shown to reduce metastasis. Documented growth, subretinal fluid, orange pigment, proximity to the disc margin or increasing thickness warrants a transition to local treatment.

Medium-sized choroidal melanoma: An eye-preserving approach is fundamental. For most patients, the preferred treatment is plaque brachytherapy (the established COMS-based approach) or, at a specialist centre, proton/charged-particle radiotherapy. Local sclerouveal resection is an option in selected cases; TTT is generally considered an adjunct/ablative tool alongside brachytherapy rather than a standalone treatment.

An eye with a large or diffuse tumour, optic nerve invasion, severe secondary glaucoma or no remaining visual potential: Enucleation is often the most appropriate treatment. Extrascleral extension worsens prognosis; gross orbital involvement may require exenteration — but there is no evidence that this prolongs survival. Therefore, exenteration is a decision about local control and symptom management, rather than an "oncological reflex".

5b. Systemic treatment for metastatic disease

The first decision point in the metastatic algorithm is HLA-A*02:01 status. In positive patients, tebentafusp takes first place as the first systemic treatment to demonstrate a survival advantage in randomised phase 3 data. In negative patients, the landscape changed substantially in 2026.

5c. Liver-dominant metastasis and adjuvant treatment

In liver-dominant disease a multidisciplinary board decision is essential. The FDA has approved the HEPZATO KIT (percutaneous hepatic perfusion) system for adults with unresectable hepatic metastases involving less than 50% of the liver, preserved liver function and absent/limited extrahepatic disease. Other options in selected cases include hepatic resection, isolated/percutaneous hepatic perfusion, TACE and radioembolisation — but the evidence for most of these rests on selected patient series and specialist-centre experience. These treatments should not be used "automatically", but in the presence of hepatic-dominant biology + good performance status + preserved liver reserve .

Adjuvant systemic treatment (for patients without metastasis but at high risk) currently has no established standard. Strategies such as fotemustine, sunitinib/valproic acid, ipilimumab, interferon and crizotinib have been investigated, but no clear, practice-changing adjuvant standard has emerged. Thus, in molecularly high-risk M0 patients, the approach is systematic surveillance and, where possible, referral to clinical research.

Follow-up has two components: monitoring ocular local control and systemic (liver-focused) surveillance for metastasis. On the ocular side, fundus examination, photography and US/OCT are used to look for recurrence/radiation complications, more frequently during the first 2–3 years and then every 6–12 months (local recurrence increases metastatic risk). Systemic surveillance is individualised according to anatomical and molecular risk.

Over the past two decades, uveal melanoma has evolved from an anatomical diagnosis into a disease defined by molecular risk classification. Appropriate management today requires integrating tumour location, size and visual potential; the patient's molecular risk profile (BAP1/monosomy 3/8q/GEP/PRAME); and metastatic status (particularly the liver and HLA-A*02:01). Treatment options range from eye-preserving radiotherapy and selected observation in localised disease to tebentafusp, the rapidly strengthening darovasertib + crizotinib combination and PRAME-directed cellular therapies in metastatic disease. Yet the two faces of the disease remain unchanged: local control of the primary tumour is usually successful, while distant metastasis is still the greatest challenge. Consequently, the backbone of modern uveal melanoma management — from diagnosis to surveillance — is multidisciplinary decision-making, integration of molecular risk and priority for clinical research. The most tangible progress lies in moving from "one-size-fits-all treatment" to "the right treatment for the right patient at the right risk level".

This article is a medical review for informational purposes; it does not provide individual treatment recommendations and does not replace official clinical guidelines. All diagnostic and treatment decisions in uveal melanoma should be made by an experienced multidisciplinary oncology team, based on each patient's anatomical and molecular characteristics. The studies discussed do not have equal levels of evidence: tebentafusp is supported by randomised phase 3/registration data, whereas darovasertib + crizotinib (OptimUM-02), as of 2026, is an "emerging" option preceding full peer-reviewed final publication and mature overall survival data; IMA203 is in early-phase development, and most liver-directed treatments are supported by selected series/expert opinion. Drug availability varies between countries; FDA approval or clinical trial data do not mean that a treatment is licensed or accessible in Türkiye. All data were verified against NCI PDQ, Nature/Nature Genetics/Nature Reviews Clinical Oncology, Science, FDA approval announcements, OptimUM-02 results and the relevant current literature; the Turkish review provided by the user was used as the structural basis.