Choroidal and Ciliary Body Melanoma

All content on Eyewiki is protected by copyright law and the Terms of Service. This content may not be reproduced, copied, or put into any artificial intelligence program, including large language and generative AI models, without permission from the Academy.

Article initiated by:

Carol L. Shields, M.D.

All contributors:

Adam R. Sweeney, MD, Divakar Gupta, MD, Fabiola V. Monsivais-Rodríguez, MD, Grant M. Comer M.D., Jennifer I Lim MD, Jessica Randolph, M.D., Kelly Miller, Michael T Yen, MD, Neelakshi Bhagat, MD, FACS, Vamsi Gullapalli, MBBS, PhD, Jamie.Choi, Adam R. Sweeney, MD, Joobin Khadamy, MD, FEBO, FEBOS-CR, Jill Wells, MD, Constance Fry, MD, Ryan T. Yanagihara, Andrew W. Stacey, Carol L. Shields, M.D., Deanna Ingrassia Miano DO, MS, Benjamin King, MD

Assigned editor:

Adam R. Sweeney, MD, Benjamin King, MD

Review:

Assigned status Update Pending

 by Benjamin King, MD on October 6, 2025.

add
Contributing Editors:	add

Choroidal and Ciliary Body Melanoma

Choroidal melanoma. Courtesy of Paul Griggs, MD. © 2019 American Academy of Ophthalmology

DiseasesDB

2614

ICD-10

[1]69.3

OMIM

155720

MedlinePlus

001022

MeSH

D014604

A: Mosaic fundus image displaying a peripapillary choroidal melanoma measuring 6x5 disc diameters with a thickness of 2 mm. The image reveals suspected orange pigment, drusen, and the presence of subretinal fluid. B: Widefield OCT B-scan showing a choroidal elevation with notable choroidal shadowing. Subretinal fluid (SRF) is visible on the sides of the elevation, along with subretinal hyperreflective material (SRHM) and drusen. C: Ultrasonography image depicting the choroidal elevation with low internal echogenicity, characteristic of the choroidal melanoma. (Courtesy of J. Khadamy)

Disease Entity

Disease

Uveal melanoma is a malignant tumor arising from melanocytes in the uveal tract (iris, ciliary body, or choroid). It is the most common primary intraocular tumor in adults with an age-adjusted incidence of 5.1 per million. There is a strong tendency for metastasis, particularly to the liver, and prognosis is poor when the tumor has disseminated. The choroid is the most common site involved and gives rise to 90% of uveal melanomas. Melanomas arising within the ciliary body comprise 7% and melanomas of the iris account for 2%. This article focuses on the treatment of posterior (ciliary body and choroidal) uveal melanoma. The treatment of anterior (iris) melanoma or conjunctival melanoma is discussed elsewhere.

Risk Factors

Uveal melanoma is largely considered a sporadic event, though certain risk factors including light iris color, light skin color, ability to tan, northern European ancestry, and rarely a family history of uveal melanoma may predispose individuals to the disease. Those with pre-existing choroidal nevi are also at risk where the incidence of malignant transformation for each lesion is 1/5000 to 1/8845.

BAP1-Tumor Predisposition Syndrome

While uveal melanoma is not typically considered a hereditary malignancy, it is the most common malignancy associated with BAP1 tumor predisposition syndrome (BAP1-TPDS), an autosomal dominant familial cancer syndrome. BAP1-TPDS is characterized by an increased risk of uveal and cutaneous melanoma, mesothelioma, renal cell carcinoma, basal cell carcinoma, cholangiocarcinoma, meningioma, and sex cord tumors. The syndrome should be considered when an individual has two or more of these associated malignancies or when an individual with a single associated malignancy has a family history of confirmed BAP1-TPDS. The prevalence of BAP1-TPDS has been estimated at approximately 1 in 1,299 to 1 in 26,837 individuals, suggesting that the syndrome may be underrecognized world wide.

Diagnosis of BAP1-TPDS is established by identification of a germline pathogenic BAP1 variant through molecular genetic testing. Uveal melanoma associated with BAP1-TPDS tends to present at a younger age, with a reported mean age of approximately 50.5 years compared with 65 years in the general population. These tumors may also demonstrate more aggressive behavior and have been reported to occur bilaterally in some patients.

An emerging association of BAP1-TPDS are nail abnormalities. Lebensohn et al. recently reported polydactylous onychopapilloma in 97% of patients with BAP1-TPDS. Subsequently, Miano et al. reported a case of polydactylous onychopapilloma in a patient with uveal melanoma and BAP1-TPDS. Examination of the nail beds is a noninvasive assessment which may further aid ocular oncologists in recognizing patients who warrant evaluation for BAP1-TPDS. Ophthalmologists should consider germline testing and genetic counseling when a diagnosis of uveal melanoma is made with presence of either polydactylous onychopapillomas or positive family history of BAP1-TPDS tumors. These efforts can facilitate early surveillance and detection of associated malignancies in patients and their families.

Pathophysiology

Uveal melanomas develop from the unregulated clonal proliferation of uveal melanocytes. As with other malignancies, this results from a complex sequence of molecular events involving multiple driver mutations in different genes. Five genes are most commonly implicated in the tumorigenesis of uveal melanoma: GNAQ, GNA11, BAP1, SF3B1, and EIF1AX. Somatic mutations in either GNAQ or GNA11 are found in 83% to 89% of uveal melanomas and occur exclusive to one another. Mutations involving these genes are thought to occur early in tumorigenesis as neither are significantly associated with tumor size or risk for metastatic disease. Subsequent driver mutations in BAP1, SF3B1, or EIF1AX have been detected in 45%, 23% and 17% of uveal melanomas respectively and occur with almost complete exclusion to one another. Driver mutations in BAP1, SF3B1, or EIF1AX are thought to represent three separate molecular pathways - each promoting tumorigenesis but conferring different levels of metastatic potential with BAP1 associated with the highest risk and EIF1AX the lowest.

Diagnosis

Symptoms

The clinical presentation of malignant uveal melanoma is characterized by nonspecific findings associated with the location of the tumor. Approximately 30% of patients are asymptomatic at presentation. 38% complain of decreased vision, 9% report photopsias, 7% floaters and 6% peripheral vision loss. Only 2% of patients complain of eye pain. Very rarely an advanced case of iris melanoma may present with a secondary glaucoma due to tumor extension into the angle and pigment-laden macrophages causing blockage of the trabecular meshwork or neovascularization.

Clinical Diagnosis

Choroidal Melanoma

Collar button mushroom configuration of melanoma. (Image courtesy of Constance L. Fry, M.D.)

Posterior uveal melanomas typically present as a unilateral elevated domed-shaped gray-brown colored mass of the choroid with irregular margins. Less commonly, a melanoma may be amelanotic. About 20% of tumors will invade through Brüch's membrane resulting in a mushroom-shaped (or collar button) configuration. Melanomas should be differentiated from benign melanocytic lesions such as choroidal nevi or congenital hypertrophy of the retinal pigment epithelium (CHRPE) as well as choroidal hemorrhage.

Some choroidal nevi are difficult to distinguish from small choroidal melanomas. Shields et al. have identified high risk features predictive of growth in concerning choroidal nevi that can be remembered using the mnemonic TFSOM-UHHD: “To Find Small Ocular Melanoma-Using Helpful Hints Daily”:

• Thickness > 2 mm
• Fluid subretinal
• Symptoms of flashes or floaters
• Orange pigment
• Margin within 3mm on the optic nerve head
• Ultrasound Hollow
• Halo absent: A halo of depigmentation surrounding a pigmented choroidal lesion is more commonly seen in nevi and indicates stability.
• Drusen absent: Drusen overlying choroidal lesions suggest chronicity.

Choroidal melanocytic tumors that display no factors have a 3% chance for growth at 5 years and most likely represent choroidal nevi. Tumors that display one factor have a 38% chance for growth, and those with two or more factors show growth in over 50% of cases at 5 years.

The size of the choroidal melanomas is the most important clinical feature in determining disease prognosis. Size is classified into three categories according to a modification of the criteria of the Collaborative Ocular Melanoma Study (COMS): small, medium and large melanomas (Table 1). Treatment strategies may vary depending on this classification.

Table1: COMS classification of choroidal melanoma by size

	Apical height	Largest basal diameter
Small	1.0 - 2.5 mm	5.0 - 16.0 mm
Medium	2.5 - 10 mm	less than 16 mm
Large	more than 10 mm	more than 16 mm

Cilary Body Melanoma

Melanomas of the ciliary body are often relatively large when they present. They have the same characteristics as choroidal melanoma but more frequently have a sentinel vessel (dilated tortuous episcleral vessel overlying the tumor). Additionally, these tumors are more likely to present with anterior displacement of the lens-iris diaphragm and a secondary angle closure glaucoma.

Diagnostic Imaging

Fundus Photography

Serial fundus photos are critical in the follow up of choroidal nevi and melanoma. The widefield retinal (WFR) imaging such as Optos creates an easily replicated map of the fundus. Certain WFR imaging systems distort the color of the tumor and fundus.

Fundus Autoflorescence

Bscan ultrasonography showing a dome shaped choroidal mass. Courtesy of Constance L. Fry, MD

Orange lipofuscin pigmentation possesses autofluorescent properties. Liposfuscin is present in suspicious nevi and many melanomas. Lipofuscin fluorescence is brighter than that of drusen (drusen are common in choroidal nevi).

UV-B scan of choroidal melanoma.

Ultrasound

This is the primary diagnostic test that confirms the diagnosis of melanoma. Ultrasound also is useful in determining the size, e.g. thickness (apical height) and basal dimension, extraocular extension (such as scleral nodules), and document growth during the follow up of a suspicious nevus/small melanoma.

• Standardized A-mode: Posterior uveal melanomas classically show medium to low internal reflectivity (88%), often in a decrescendo fashion, a.k.a, positive angle kappa, with regularity of structure. The internal blood flow (vascularity) of the tumor can be seen as a fast movement of the internal spikes. Uncommonly, a large melanoma may be more irregular, particularly if there is necrosis.
• B-mode: Shapes: Dome-shaped is most common, mushroom-shaped, i.e. collar button is the most classic shape, and an irregular shape is uncommon. Other features: acoustically hollow zone within the tumor, choroidal excavation, and subretinal fluid.
• Ultrasound biomicroscopy is utilized to better delineate ciliary body and iris melanomas.

Color Doppler ultrasound

Choroidal melanomas show pulsatile blood flow at the tumor base. This finding is not found in nevi.

Fluorescein and indocyanin green angiography (ICGA)

• Hypofluorescence: Due to blockage of the choroidal blood flow by the pigmentation inherent to the tumor.
• Hyperfluorescence: Small hyperfluorescent spots may be seen due to lipofuscin deposition at the RPE level. Pinpoint hyperfluorescence is a high risk characteristic seen in choroidal nevi and if present with other high risk characteristics, one should consider that the tumor may be undergoing malignant transformation.
• Circulation:“Double circulation” pattern consisting of an internal circulation within the lesion and the normal vascularity of the overlying retina. This characteristic is more evident in ICGA
• Neovascularization: is not typical of melanoma and its presence suggests another diagnosis should be sought.

Standard spectral domain OCT (SD-OCT)

Standard spectral domain OCT does not penetrate deeply enough for detecting the internal characteristics of choroidal neoplasms. However, it is useful in visualizing changes in the neurosensory retina and the retinal pigment epithelium (RPE). As it was mentioned earlier, posterior uveal melanomas may show serous retinal detachment in the areas adjacent to the tumor. i.e. subretinal fluid. Lipofuscin deposition is also seen at the level of the RPE.

Enhanced Depth Imaging Spectral Domain OCT (EDI-OCT)

Enhanced Depth Imaging Spectral Domain OCT is a relatively new technology that is now commercially available. It is a method that allows for the evaluation of deeper structures such as the choroid and the internal portion of the sclera. In a study of 37 eyes by Shields et al. it was found that the distinguishing characteristics of choroidal melanomas readily seen in EDI-OCT are:

• Optical choroidal shadowing (100%): Due to the dense pigmentation of the melanoma, there is often shadowing beneath the tumor, which can obscure deeper structures in the choroid.
• Choriocapillaris Compression and Thinning (100%): The melanoma may compress the choriocapillaris, leading to thinning over the tumor.
• Subretinal fluid (92%): Frequently associated with choroidal melanoma, the presence of subretinal fluid is an important indicator, often detected by OCT even when clinical examination might miss it.
• Bacillary Layer Detachment (BALAD): A unique feature of choroidal melanoma observed in a significant number of cases, particularly in medium-sized tumors. It is associated with more extensive subretinal fluid.
• Subretinal lipofuscin deposits (95%)
• Shaggy photoreceptors (49%): The "shaggy" appearance of photoreceptors overlying the tumor is often seen in choroidal melanoma, especially in small melanomas. This finding helps differentiate melanoma from benign nevi.
• Ellipsoid Zone (EZ) Loss/Disruption: This is one of the most common findings in choroidal melanoma, indicating damage to the photoreceptors and outer retinal layers. It is more prevalent in larger tumors.
• Subretinal Hyperreflective Material (SRHM): This material can appear as either homogeneous or heterogeneous on OCT and is a common finding. It is often associated with the presence of subretinal fluid and more advanced melanoma.
• Retinal and RPE Changes: OCT often shows retinal thinning, retinal pigment epithelium (RPE) atrophy, or thickening over the tumor. RPE hyperplasia or disruption may also be observed.
• Other retinal changes included: Loss of photoreceptors, loss of external limiting membrane, irregularity of inner plexiform layer, irregularity of ganglion cell layer, intraretinal edema.

These OCT findings are essential for assessing tumor size, monitoring changes over time, and distinguishing choroidal melanoma from benign lesions such as choroidal nevi.

Management

Evaluation and Management

Management of uveal melanoma depends on several clinical factors, including tumor size, thickness, location, and activity; the status of the fellow eye; and the patient's age, general health, life expectancy, and psychological status. Treatment options include observation, radiation, laser therapy, surgery, medical therapy, or combinations of these approaches. A detailed ophthalmic evaluation should be performed, with tumor size and extent documented using fundus drawings, color photography, and ultrasonography. Known risk factors for tumor growth and metastasis should be considered, and treatment options discussed in the context of the patient's life prognosis, quality of life, anticipated visual outcome, and ability to tolerate treatment.

When clinical suspicion for uveal melanoma is present, the decision to observe or treat should incorporate both subjective and objective features. Observation may be appropriate for selected small lesions when ancillary testing has not established a definitive diagnosis, particularly when the lesion demonstrates slow growth, the patient is a poor surgical candidate, or the patient prefers a conservative approach. Observation may also be reasonable in patients with a life expectancy of less than 5 years or in very small tumors involving the patient's only functional eye, with treatment deferred until high-risk features or documented growth develop. Although the average tumor doubling time has been reported to be approximately 180 days, growth rates vary among melanomas and this variability should be considered when observation is selected. For lesions less than 3 mm in thickness, baseline fundus photography, fluorescein angiography, and A- and B-scan ultrasonography should be conducted, with repeat examination and imaging at approximately 3–4 months. If no growth is identified, longer-term surveillance with fundus photography every 6–12 months may be considered.

Mutli-modal imaging and examination is typically sufficient when clinically diagnosing uveal melanoma; however, biopsy may be considered when the diagnosis remains uncertain. Fine-needle aspiration biopsy (FNAB) can provide cytologic or molecular confirmation and may be performed when clinically indicated. FNAB is more commonly performed at the time of plaque brachytherapy to obtain tissue for cytologic and molecular prognostic testing, however can be conducted as a stand-alone surgery before treatment, or after enucleation (see Biopsy section).

In practice, clinicians integrate multiple features to estimate the likelihood that a melanocytic lesion will demonstrate growth and to distinguish a nevus from melanoma. Shields et al. demonstrated how increased tumor features correlates with increased risk of tumor growth. The presence of multiple high-risk clinical indicators should prompt further diagnostic evaluation and consideration of treatment; in selected cases, biopsy may be offered when the diagnosis remains uncertain.

General treatment

The goals of treating uveal melanoma is to achieve the following:

• Preservation of the life - minimizing the risk of metastasis and disease-related death
• Preservation of the vision - salvaging visual potential with careful and individualized treatment planning
• Preservation of the eye - while successful treatment of some tumors may inevitably compromise vision (ie: juxtapapillary and macular tumors), enucleation & exenteration are avoided unless the globe itself becomes compromised.

Periodic Observation

A choroidal nevus is a well-circumscribed, benign melanocytic tumor that is managed by periodic observation. They typically appear as small (<2 mm thick), asymptomatic, and with overlying retinal pigment epithelial atrophy and drusen, signifying a chronic condition. Approximately 4.6% of the United States population manifests a choroidal nevus. Prevelance is higher in Caucasian populations (5.6%), and lower in Hispanic (2.7%) populations and African American (0.6%). Approximately 1 in 5000-8800 choroidal nevi evolve into choroidal melanoma.

Patients with suspicious nevi should be reexamined and imaged in 3 months and monitored for changes in size. If no growth is observed, then reexamination can occur every 6 months thereafter. It is important to re-image all suspicious nevi with appropriate modalities at consecutive visits for accurate comparison of clinical features.

Documented growth of a lesion < 2 years is strongly suggestive of melanoma, whereas growth over >10 years is still consistent with a benign lesion. Since documented growth may be associated with worse systemic prognosis, some patients with small tumors that show three or more risk factors are treated promptly, without waiting for documentation of growth. Increasing tumor size both in base and thickness leads to increased risk for metastasis. Based on the few patients with medium size choroidal melanoma who refuse treatment and are followed, natural history studies have found that there is greater mortality and higher risk of death. Open communication with patients is imperative during the observation process as to emphasize potential risks and treatment goals should the time come.

Historical perspective

The primary goal of treatment for uveal melanoma is to prevent metastasis while preserving the eye and vision. Historically, enucleation was the standard treatment for uveal melanoma, providing both definitive local control and histopathologic diagnosis. Enucleation remained the gold standard for centuries until the late 1970s, when concerns regarding its potential to facilitate metastasis emerged. Advances in radiation oncology prompted the development of eye-sparing treatment strategies.

Plaque brachytherapy in choroidal tumors was pioneered by Packer et al. in the 1980's, where he described the placement of a customized plaque containing iodine-125 (I-125) overlying the tumor. Concerns regarding local recurrence and metastasis initially limited acceptance of eye-conserving therapy. The benefit to eye-salvaging therapy was highlighted in the 1990's-2000's by the Collaborative Ocular Melanoma Study (COMS) who demonstrated similar outcomes in enucleated and irradiated groups (See COMS section). Subsequent COMS publications reinforced these findings, establishing radiotherapy as a standard treatment for uncomplicated uveal melanoma.

Proton beam treatment for uveal melanoma was founded by Gragoudas et al. in the 1970's-1980's. Rather than plaque placement, one-time implantation of tantalum clips allows for targeted treatment of ophthalmic tumors. While there sufficient reportings of successful proton beam treatment outcomes and the potential for reduced ophthalmic side effects, limited availability due to geographical accessibility has prevented large-scale prospective analysis to-date.

Treatment - Eye

Surgical Resection

There are several methods for surgical management of posterior uveal melanoma including enucleation, exenteration, and local resection.

Enucleation

Enucleation is generally indicated for advanced melanomas that occupy most of the intraocular structures, have caused severe secondary glaucoma or total retinal detachment that is not reversible to alternative treatments, or tumors invading the optic nerve. Enucleation with a long section of the optic nerve is appropriate in such cases. Sometimes patient preference dictates this treatment as well. However, many juxtapapillary melanomas that abut the optic nerve and show no evidence of invasion can be managed by custom-designed notched radioactive plaques or proton beam therapy rather than enucleation.

There is no role for preenucleation radiation. The large sized tumor COMS trial indicated that external beam radiotherapy prior to enucleation did not provide a survival benefit compared to enucleation alone.

Radioactive Plaque (Image courtesy of Constance L. Fry, M.D.)

The "no touch enucleation" was introduced to minimize the amount of surgical trauma and theoretically to lessen the chance of tumor dissemination at the time of surgery. An essential aspect of this technique was to freeze the venous drainage from the tumor prior to cutting the optic nerve. The "no touch" technique has recently fallen into disuse at most centers because it is cumbersome and its benefits are only theoretical. However, a gentle tissue handling during a standard enucleation should be employed, without clamping the optic nerve prior to cutting it. Care should be taken not to pass needles through the sclera for traction either; rather the globe is distracted by stumps of the medial and lateral recti.

After removal of the eye, either fine needle aspiration biopsy or fixated tissue may be submitted for genomic expression profiling.

There have been advances in the types of orbital implants used following enucleation. The hydroxyapatite implant, designed to improve the ocular motility in patients undergoing enucleation, is used widely. Other implants include polyethylene and polymer coated hydroxyapatite. Furthermore, porous polyethylene implants have been used successfully.

Orbital Exenteration

Ciliary body melanoma with extrascleral extension. (Image courtesy of Constance L. Fry, M.D.)

The subject of orbital exenteration for uveal melanomas with extrascleral extension is controversial. Complete orbital exenteration should not be done in cases of mild degrees of extrascleral extension. However, in the rare instance of massive orbital extension in a blind, uncomfortable eye, primary orbital exenteration is probably justified. In most instances of orbital extension for uveal melanoma, it is not necessary to sacrifice the skin of the eyelid. The eyelid-sparing exenteration provides a better cosmetic appearance.

Local Resection

Local resection of melanomas involving the ciliary body and choroid are rarely conducted due to the invasiveness, risk of tumor spread, and proven effectivity of radiotherapy. Resection can be performed using a partial lamellar sclerouvectomy technique. This surgical technique is a modification of the one popularized by Foulds, and later Shields and Damato, in which the tumor is removed with the aim of leaving the retina and vitreous intact. It is best performed in growing small size ciliary body melanomas or choroidal melanomas near the equator. There is no current evidence that local resection of posterior uveal melanoma is different from enucleation or radiotherapy with regard to patient survival. More complications can be expected when larger post-equatorial tumors are managed in this manner. In addition, incomplete resection of posterior melanomas is a concern.

Radiotherapy

Plaque brachytherapy

The most commonly employed form of radiotherapy is brachytherapy using a episcleral radioactive plaque that is temporarily sutured to the scleral surface. Iodine-125, Ruthenium-106, and Palladium-103 plaques have largely replaced Cobalt-60 at most institutions.

Following the COMS medium size tumor trial, plaque brachytherapy became the standard of care for small and medium-sized melanomas located outside the macular region and posterior to the ora serrata. Shields and associates found that plaque brachytherapy can also be custom fit to treat small, medium, and even large uveal melanoma up to approximately 12 mm in thickness. Although treating large melanoma was effective, complications of radiation maculopathy and papillopathy were higher and often led to poor long term vision.

Innovations in radiotherapeutic planning have allowed plaque brachytherapy to be custom designed to treat melanomas at any location within the eye including the macula using a round or notched plaque, juxtapapillary region using a notched plaque, ciliary body using a round or curvilinear plaque, and iris using a curvilinear plaque.

Plaque brachytherapy involves careful pre-operative measurement of tumors. With these measurements and in conjunction with radiation oncology, a plaque is designed which will deliver appropriate radiation dose to the entire tumor, including the distal apex. The treatment involves two surgeries: plaque placement and plaque removal after treatment time is achieved. These times vary by institution and may range from 4-7 days.

Plaque types

Plaque design has transformed since the initial COMS study era, such as that seen with Eye Physics Plaques (Table 2). To date, both types of plaques are used within institutions across the United States.

Table 2: Comparison of COMS Plaques and Eye Physics Plaques

Information adapted from Marwaha G, Macklis R, Singh AD, Wilkinson A. Brachytherapy. Developments in Ophthalmology Published online 2013

Feature	COMS Plaque	Eye Physics (EP) Plaque
Thickness	~2.75 mm	~1.5–2.0 mm, depending on plaque design
Seed positioning	I-125 seeds embedded in a Silastic carrier, approximately 1 mm from the sclera	I-125 seeds positioned closer to the sclera within individually collimated slots
Dose distribution	Relatively uniform radiation dose with treatment distributed lateral to tumor borders	More focused radiation with steeper dose falloff outside the desired tumor borders
Clinical advantage	Standardized design with extensive clinical experience and long-term outcome data	Thinner profile and more individualized dose distribution, potentially reducing radiation exposure to adjacent ocular structures

Proton Beam Radiotherapy

Charged particle therapy applies a focused beam of collimated protons or helium ions to deliver a high dose of radiation to a targeted area. Damage to surrounding tissues is limited due to the Bragg peak effect, where the most destructive ionizing radiation occurs immediately before the location that the particles stop traveling. It is used in medium- to large-sized tumors or for tumors in regions inaccessible to plaque brachytherapy (such as those involving >180 degrees of the optic nerve margin). Proton beam therapy is a well-established treatment option but is cost prohibitive and limited to only a few dozen centers around the world. Though believed to be less severe, complications are similar to that of plaque brachytherapy.

Proton beam radiothearpy involves the following steps: First, tantalum marker rings are surgically placed on the sclera around the tumor to provide radiopaque landmarks for precise localization of tumor during proton delivery. After healing, the patient undergoes multiple rounds of proton beam radiation at a specialized center, which is generally delivered over several consecutive outpatient sessions. During each session, the patient is positioned and immobilized at the proton beam facility, and the tantalum rings are used to accurately align the eye with the planned treatment field. The proton beam is directed toward the tumor, delivering a highly localized dose while minimizing radiation exposure to surrounding ocular structures.

Stereotactic Radiosurgery

Stereotactic radiosurgery is a non-invasive form of therapeutic radiation that is delivered over a small, well-defined 3-dimensional area. GammaKnife and CyberKnife, are examples of stereotactic radiosurgery that have been used against choroidal melanomas unsuited for plaque brachytherapy. As with plaque brachytherapy, radiation injuries are common. However side effects from stereotactic radiation tend to cause robustly reduced vision in the majority of patients.

Adjuvants to radiotherapy

Plaque brachytherapy has proven superior to other stand-alone treatment modalities. However adjuvant utilization of transpupillary thermotherapy, laser photocoagulation and intravitreal anti-vascuar endothelial growth factor (anti-VEGF) have demonstrated promise in reducing risk of radiation complications such as radiation retinopathy, vascular pathology and radiation papillopathy. Laser therapies are rarely used as primary intervention due to the unacceptably high rate of incomplete control, tumor recurrence and extrascleral extension. Of note, laser therapies can only be utilized on choroidal tumors, not ciliary body tumors which lack visualization on posterior examination.

Transpupillary Thermotherapy

Transpupillary thermotherapy (TTT) supplements plaque brachytherapy through direct tissue destruction, usually ensuing at the time of plaque removal in the operating room, or during subsequent follow-up in the clnic setting. TTT alone is insufficient in tumor destruction, allowing for an estimated 33% recurrence rate by 10-years.

Photodynamic Therapy

Photodynamic therapy (PDT) involves intravenous administration of a photosensitizer that, when activated by a specific wavelength of light, releases reactive oxidative species within tumor vasculature and induces endothelial changes and thrombosis. Standalone PDT with verteporfin for choroidal melanoma is not recommended due to concern of recurrence long-term.

Argon Laser

Sector scatter photocoagulation has been extensively studied by Shields et al. and others. Laser therapy can be done at time of plaque removal, and at subsequent visits, with the goal of reducing rates of radiation retinopathy.

Complications to ophthalmic treatment

Presently, radiotherapy is the most widely employed intervention for posterior uveal melanoma. However, between 5% and 10% of patients treated with radiotherapy ultimately require enucleation of the affected eye because of tumor recurrence or radiation complications (e.g. chronic pain, neovascular glaucoma, radiation retinopathy, radiation optic neuropathy). Radiation tumor vasculopathy is also common as the irradiated tumor can become ischemic, resulting in macular edema, serous retinal detachment, retinal ischemia, and neovascular glaucoma.The complications associated with brachytherapy are associated with the necessity of 2 invasive operations requiring anesthesia and radiotoxic effects on healthy ocular tissue. As posterior choroidal tumors are the most common uveal melanomas, extraocular muscle repositioning is typically required. Radiation induced complications after brachytherapy are numerous and may depend on the tumor size, tumor location, dose of radiation use, and rate of use. Radiation complications are often delayed. Complications to the adnexa and anterior segment early on include diplopia and ptosis and later cataract, symptomatic dry eye, iris neovascularization, and secondary glaucoma. Posterior segment complications of brachytherapy include retinal detachment, cystoid macular edema, optic neuropathy, and hemorrhage of the vitreous, retina, choroid, or a combination of these structures. The most common posterior segment complication, however, is radiation retinopathy, with greater than 75% of individuals experiencing this complication following brachytherapy in some reports. The mechanism of injury in radiation retinopathy is hypothesized to be secondary to free radical damage to the vascular endothelial cells leading to occlusion of the capillary beds, microaneurysm formation and retinal ischemia. This can further lead to areas of retinal nonperfusion causing macular edema, neovascularization, and tractional retinal detachment. This may be managed with injections of vascular endothelial growth factor inhibitor, steroids and/or laser.

Proton beam radiotherapy has several recognized complications including cataract, chronic uveitis, hyphema, macula edema, retinal detachment, corneal melting, scleral atrophy and secondary glaucoma. Indeed the incidence of glaucoma following proton beam radiotherapy has been reported to be as high as 53%. Secondary glaucoma is also the leading complication necessitating subsequent enucleation.

As radiotherapy is the most commonly used treatment option, the main ocular complication following radiotherapy is retinopathy with resultant decreased visual acuity. Gunduz and associates studied retinopathy following plaque radiotherapy and noted nonproliferative retinopathy in 42% at 5 years and proliferative retinopathy in 8% at 5 years. Other radiation complications included cataract (39%), papillopathy (8%), vitreous hemorrhage (7%), neovascular glaucoma (1%), and scleral necrosis (1%). The treatment of these complications involves laser photocoagulation, anti-vascular endothelial growth factor medications, and anti-inflammatory medications.

Treatment - Metastatic Disease

Nearly half of all patients with uveal melanoma will develop metastatic disease. Uveal melanoma spreads exclusively hematogenously, unless extraocular extension ensues and lymphatics become infiltrated. In the Collaborative Ocular Melanoma Study (COMS), the most frequent sites of metastases at time of death were the liver (93%), lung (24%), and bone (16%). More than 80% had multiple sites of metastasis. The 5 and 10-year cumulative metastasis rates are 25% and 34%, respectively. Prognosis is very poor after metastasis has occurred.

Regional Treatment of liver metastasis

The liver is the most frequent site of metastasis. About 90% of patients with large choroidal melanomas in the COMS had liver metastases at the time of death, and 57% had exclusive liver metastases. Treating hepatic metastases from uveal melanoma requires a multidisciplinary approach. Treatment choice depends on the number of hepatic metastases, tumor size/volume, and time to systemic metastasis.

Direct surgical procedures on the liver can be used to manage metastatic liver nodules. Other regional therapies involve the hepatic artery, which predominantly supplies metastatic tissue.

Surgical resection of metastatic nodules

While good outcomes have been reported in patients with uveal melanoma with metastasis to the liver, surgical intervention is rarely indicated as most patients present with multiple liver metastases involving both liver lobes.

Transcatheter intraarterial therapies

Surgical or local ablative therapies are rarely used because hepatic metastases usually involve multiple regions of the liver. Transcatheter intra-arterial therapies that target the entire liver are more commonly used. These treatments may prolong survival for a few months. However, it is unlikely that it will be curative as most patients will develop progression of liver metastases despite these treatments.

Hepatic intra-arterial chemotherapy

Liver metastases are perfused by the hepatic arteries whereas normal hepatic tissue is primarily supplied by the portal vein. Direct delivery of chemotherapy to the hepatic artery via an indwelling catheter allows maximal exposure of hepatic metastases to cytotoxic chemotherapy while minimizing systemic toxicity. Hepatic intra-arterial chemotherapy with fotemustine, melphalan, cisplatin, vinblastine, and dacarbazine on patients with metastatic uveal melanoma have been performed but no significant increase in overall survival has been observed.

Transarterial chemoembolization

In hepatic transarterial chemoembolization (TACE), chemotherapy is infused through the hepatic artery followed by selective obstruction of a distal hepatic artery branch to “starve” a metastatic tumor and induce tumor necrosis. Conventional TACE involves the infusion of chemotherapy mixed with lipiodol (ethiodized oil) followed by occlusion of a prespecified hepatic artery branch that feeds the tumor with an embolic agent (e.g. polyvinyl sponge). The use of conventional TACE using a wide variety of chemotherapeutic medications and protocols have been used, but the best chemotherapy choice or embolic therapy remains unknown. Different embolization procedures have emerged. More recently, drug-eluting beads have been used as embolic materials that are infused with anti-cancer medications, such as doxorubicin, irinotecan, and epirubicin.

Patients undergoing TACE must have disease limited to the liver or liver-dominant disease without signs of hepatic failure, such as encephalopathy, or decreased portal vein blood flow (e.g. acute or chronic portal vein thrombosis).

Postembolization syndrome, characterized by fever, right upper quadrant abdominal pain, elevated liver enzymes, nausea, and/or vomiting, is a common complication. Other important but rare complications include cholecystitis, pulmonary embolism, hepatic abscess, bile duct injury, gastric mucosal injury, and acute pancreatitis. Hematologic or vascular complications have also been reported.

Transarterial hepatic immunoembolization

Immunoembolization involves hepatic artery embolization using granulocyte-macrophage colony-stimulating factor with lipiodol (instead of chemotherapy as in TACE) to induce tumor ischemia and stimulate antigen presenting cells to facilitate antigen uptake and enhance systemic immunity against tumor cells. One study with 34 participants with metastatic uveal melanoma showed positive outcomes, with an overall survival of 14.4 months. Common complications include transaminitis, abdominal pain, and nausea/vomiting.

Selective Internal Radiation Therapy

Selective internal radiation therapy, also referred to as transarterial radioembolization (TARE), involves inserting radioactive yttrium-90 microspheres into the hepatic artery that occlude the tumor microvasculature and deliver localized radiation to hepatic tumors. Studies have shown variable results with overall survival ranging from 9 to 24 months in patients with unresectable liver metastases from uveal melanoma.

Isolated hepatic perfusion

Isolated hepatic perfusion (IHP) is a complex open surgical procedure where hepatic blood flow is segregated from the systemic vasculature through cannulation and clamping of the hepatic artery and inferior vena cava and joining this circulation to an extracorporeal circuit. This distinct circulation allows administration of highly potent chemotherapies to the entire liver while bypassing systemic circulation and reducing undesirable systemic effects. Melphalan is the most commonly used chemotherapy agent. However, this surgical procedure is complex and extensive, limiting its availability. IHP is also associated with considerable morbidity and mortality due to veno-occlusive disease and hepatotoxicity. In addition, because of the open surgical nature of this procedure, IHP generally is not repeated.

Percutaneous isolated hepatic perfusion (PHP) is a minimally invasive alternative to IHP. A major advantage of PHP compared to IHP is the potential for multiple perfusion treatments. As with IHP, melphalan is the most commonly investigated chemotherapy agent in metastatic uveal melanoma. A randomized phase 3 clinical trial (NCT02678572) demonstrated superior efficacy of PHP compared with best alternative treatment with respect to disease control and overall survival. PHP received FDA approval in August 2023.

Systemic therapy

Chemotherapy is reserved for metastatic uveal melanoma. However, effective chemotherapy agents used against metastatic cutaneous melanoma are generally ineffective in metastatic uveal melanoma. Various antineoplastic agents, including standard chemotherapeutic agents, targeted cancer therapy, and immunotherapy have been investigated, but no significant improvement in overall survival or progression free survival has been observed. However, some recent studies have shown encouraging results.

Conventional chemotherapy

Chemotherapeutic agents have been evaluated as monotherapy or combination therapy, including dacarbazine, temozolomide, treosulfan, and fotemustine, with response rates of <10%.

Targeted Therapy

Uveal melanoma and cutaneous melanoma are clinically and biologically distinct. In uveal melanoma, mutations in GNAQ and GNA11 are the most common along with mutations to SF3B1, EIF1AX, and BAP1. In contrast, cutaneous melanoma most often involves mutations in BRAF, NRAS, and KIT. While effective immune checkpoint inhibitors have been identified in the treatment against metastatic cutaneous melanoma, these medications have not demonstrated the same effect in metastatic uveal melanoma. However, novel and promising treatments are under investigation.

One promising treatment option is tebentafusp, a bispecific fusion protein (gp100 peptide) that is composed of a monoclonal high-affinity T cell receptor, which presents uveal melanoma antigens, fused to an anti-CD3 single-chain antibody fragment that activates cytotoxic T cells. A phase 3 trial of tebentafusp was the first to show reduced risk of death from metastatic uveal melanoma compared to investigator’s choice (dacarbazine, pembrolizumab, or ipilimumab) in treatment-naive HLAA*02:01-positive patients with metastatic uveal melanoma; FDA approval occurred in 2022.

Adjuvant Therapy

At present, no effective adjuvant therapy decreases the risk of metastasis or improves overall survival.

Prognosis

The prognosis for uveal melanoma should be distinctly considered in terms of life, the globe, and visual acuity. With regards to life prognosis, uveal melanoma has been shown to be dependent on several clinical factors including tumor location in the ciliary body, large tumor size, diffuse (flat) configuration, and extraocular extension as well as cytology genetics, and molecular markers. In several articles, tumor size has been identified as one of the key clinical features predictive of metastasis. Shields et al. found that increasing millimeter thickness of uveal melanoma was associated with increasing risk for metastasis (Table 3). These findings highlight the importance of close monitoring, consistent documentation, and early intervention.

Table 3: Probability for systemic metastasis in posterior uveal melanoma based on tumor thickness

Information adapted from Shields CL, Furuta M, Thangappan A, Nagori S, Mashayekhi A, Lally DR, Kelly CC, Rudich DS, Nagori AV, Wakade OA, Mehta S, Forte L, Long A, Dellacava EF, Kaplan B, Shields JA. Metastasis of uveal melanoma millimeter-by-millimeter in 8033 consecutive eyes. Arch Ophthalmol 2009;127:989-998.

Tumor thickness, mm	Probability for Systemic Metastasis, %
	3 years	5 years	10 years	20 years
Using 1-mm increments
0.0-1.0	2.4	5.7	5.7
1.1-2.0	2.3	7.9	12.0
2.1-3.0	2.0	4.6	11.8
3.1-4.0	3.1	8.1	16.3
4.1-5.0	7.5	15.2	26.8
5.1-6.0	8.6	17.3	27.9
6.1-7.0	9.5	15.2	28.8
7.1-8.0	13.0	21.3	40.8
8.1-9.0	17.6	31.1	50.2
9.1-10.0	18.2	30.7	43.7
> 10.0	27.8	40.2	51.0
Using small, medium, and large COMS classifications
Small (0.0-3.0)	2.1	5.6	11.5	19.7
Medium (3.1-8.0)	7.2	13.9	25.5	37.3
Large (> 8.0)	22.3	35.0	49.2	66.9

Prognosis through tumor sampling

Fine needle aspiration biopsy of uveal melanoma (see Biopsy section for details) is used for prognostication and is typically performed immediately prior to radioactive plaque placement, however can be conducted independently prior to intervention, or after enucleation ensues. Cytogenetic analysis, gene-expression-profile (GEP) testing, next generation sequencing (NGS), and tumor cytology can aid in treatment decisions.

Cytogenetic analysis

Genetic evaluation of uveal melanoma demonstrated how cytogenetic combinations of tumors could correlate with varying survival outcomes (Table 4). More specifically, chromosome 3 loss (monosomy 3), 1p loss, 6q loss and 8q gain, have been associated with lower prognosis.

Table 4: Chromosomal alterations contributing to prognostication of uveal mealnoma.

	Associated tumor phenotypes	Prognostication
Monosomy 3	Larger diameter Ciliary body location High mitotic rate Extraocular tumor extension Higher influx of tumor-infiltrating lymphocytes and macrophages Epithelioid cell type Vascular loops	Poor prognosis Occurring in 57% of metastasizing tumors Worse prognosis when combined with 1p loss or 8q gain Associated with BAP1
1p loss		Poor prognosis Worse prognosis when combined with monosomy 3
6q		Gain: good prognosis Loss: poor prognosis (6q loss occurring in 40% of metastasizing tumors) Rarely occurs with monosomy 3 (<4% cases)
8q gain	Larger diameter Ciliary body location High mitotic rate Epithelioid cell type Vascular loops	Poor prognosis Worse prognosis when combined with monosomy 3

Molecular prognostication through biopsy

Combined efforts have built upon the above chromosomal analyses and allowed for subclassification of uveal melanomas. A clinically validated gene expression profile (GEP) test for uveal melanoma prognostication (DecisionDx-UM - Friendswood, TX) is currently the standard of care in most ocular oncology centers in the United States.

Background

Harbour et al described the influential role of the gene BAP1 in regulating the metastasis of uveal melanoma as it does in other cancers involving the lung, breast, and kidney. The BAP1 gene codes for a deubiquitinating enzyme that binds to BRCA1 and BARD1 to create a heterodimeric tumor suppressor complex. Inactivation of this gene has been described in up to 84% of class 2 uveal melanomas. As BAP1 is mapped to chromosome 3p21.31-p21.2, a loss of this region, as in monosomy 3, could predispose an individual to uveal melanoma via the classic mechanism underlying other tumor suppressor genes. While multiple genes are likely involved in the development of metastatic uveal melanoma, Onken et al has been the leader in translating this research into an applicable clinical test. Onken observed that monosomy 3 had limited ability in prognosis for uveal melanoma and thereafter identified 12 genes found to be predictive of metastatic disease. Using a 15 gene microarray assay, including 3 control genes, Onken et al found this test to accurately discriminate between low-grade (class 1A and class 1B) and high-grade (class 2) uveal melanoma. Both fine needle aspirate biopsies and formalin fixed, paraffin embedded tumor tissue provide adequate supply of tumor cells while being safe and technically simple. This technique, termed multi-gene expression profiling, has been compared against the other various prognostic indicators for accuracy in predicting risk of metastasis and was found to be superior to all other clinical, histopathologic and cytogenetic biomarkers. The multi-gene expression profile (GEP) assay was initially offered in 2009 and remains the only prognostic test for uveal melanoma validated in a multi-center prospective study. Overexpression of the cancer-testis antigen PRAME has been identified as an additional molecular risk factor independent of GEP status. The presence of PRAME mRNA can be detected from the same fine needle aspirate biopsy obtained for GEP. A hybrid prognostic model incorporating both PRAME and GEP status demonstrated greater accuracy in predicting metastatic disease and is now included as part of the interpretation of most prognostic biopsies.

Collaborative Ocular Oncology Group (COOG)

Prognostication through GEP was first validated by the Collaborative Ocular Oncology Group Report No. 1 (Table 5). Of note, the study also determined monosomy 3 outside of GEP was not an independent risk factor for metastasis or survival.

Table 5: Prognostication with gene expression profiling alone by COOG Report No. 1

Information adapted from Field MG, Harbour JW. Recent developments in prognostic and predictive testing in uveal melanoma. Curr Opin Ophthalmol. 2014;25(3):234-239.

RNA GEP Class	Metastatic Risk	5-year metastatic risk	Associated genetic mutation
Class 1A	Low	2%	SF3B1, EIF1AX
Class 1B	Intermediate	21%	SF3B1, EIF1AX
Class 2	High	72%	BAP1

Further analysis of GEP classification yielded speculation on Class 1A vs 1B differences in metastatic rates. While these subclassifications are still available on DescisionDx-UM reports, Classes 1A and 1B are often combined, which has been widely accepted following the validation of independent variable PRAME (PReferentially expressed Antigen in MElanoma).

Another report by COOG (COOG2.1) prospectively validated PRAME as an additional molecular biomarker, independent to GEP risk stratification. This work, lead by Harbour et al., established combined GEP and PRAME classification (Table 6) is superior to that of GEP alone (Table 5).  PRAME can be detected from the same fine needle aspirate biopsy sample obtained for GEP (requiring merely 6-10 tumor cells for successful analysis). DescisionDx-UM now offers this combination assay for interpretation of tumor genetics.

Table 6: Prognostication with gene expression profiling and PRAME status. To date, this is most commonly utilized throughout the United States.

Information adapted from Harbour et al. 15-Gene Expression Profile and PRAME as Integrated Prognostic Test for Uveal Melanoma: First Report of Collaborative Ocular Oncology Group Study No. 2 (COOG2.1). Journal of Clinical Oncology 2024.

GEP Class	PRAME	Probability metastasis free
		3 years	5 years
Class 1	Negative	97.4%	95.6%
	Positive	92.4%	80.6%
Class 2	Negative	74.2%	58.3%
	Positive	51.0%	44.8%

Molecular prognostication through fluid

Blood:

Callejo et al. demonstrate how uveal melanoma cells can be found circulating in the blood both before diagnosis and after treatment, including enucleation. This supports the hypothesis of micrometastasis, or subclinical spread, of uveal tumors which may not be recognized on systemic imaging. Commercially available blood tests such as Natera/Signatera (Austin, TX) allow for detection of circulating tumor DNA (ctDNA).  For example, a blood draw is conducted prior to treatment which will allow future assays to recognize pre-existing melanoma ctDNA in the blood.  Following enucleation, levels of ctDNA are expected to reduce. Should levels maintain or rise, suspicion for systemic disease may be warranted.

Aqueous:

Berry et al. demonstrate how uveal melanoma, including posterior choroidal tumors, can be identified with aqueous humor sampling. Uveal melanoma ctDNA, regardless of iris, ciliary body or choroidal location, were validated through this relatively minimally invasive test (when compared to direct tumor biopsy).  The potential of this novel technique continues to be explored. Significant literature has been published on aqueous humor biopsy for other posterior tumors such as retinoblastoma.

Globe prognosis

With regards to the globe, nearly 95% of patients treated with conservative measures such as radiotherapy or resection, maintain their globe on follow up. The greater the tumor thickness, the greater the risk for enucleation.

Visual prognosis

With regards to visual acuity, a study on 1106 consecutive plaque-irradiated patients with uveal melanoma found poor visual acuity (20/200 or worse) in 34% at 5 years and 68% at 10 years of follow-up. Radiation related damage (ie: cataract, radiation retinopathy, radiation papillopathy, radiation maculopathy, neovascular glaucoma and radiation induced dry eye) are key reasons for these gradual defects.^k Extent of visual change may be related to overall location of tumor, size of tumor, radiation dose and pre-existing ocular pathology.

Follow-up

Patients with uveal melanoma should have regular systemic follow up examinations by both an ocular oncologist and a medical oncologist. The ocular oncologist should monitor the uveal scar for tumor regression and complications of therapy. The medical oncologist should survey for metastatic disease. Particular evaluation of the liver, lung, and skin should be made as this malignancy most often metastasizes to these sites. Physical examination and liver function testing is recommended twice yearly, as well as annual liver magnetic resonance imaging and chest radiography for monitoring or CT chest/abdomen/pelvis or PET CT(less commonly used due to amount of radiation exposure).

Historical Perspectives

In 1809, the first known complete natural history of uveal melanoma was documented by Scottish surgeons Allan Burns and James Wardrop. In 1882, Ernst Fuchs was the first to describe sarcom des uvealtractus (“uveal sarcoma,” now known as uveal melanoma) as “one of the most malignant of diseases” and recommended enucleation as the treatment of choice to prevent metastasis and death.

The Zimmerman-McLean-Foster Hypothesis

Following enucleation, metastatic disease and mortality remained high. In the 1970s, a landmark paper by a group of expert ophthalmic pathologists proposed that an elevated intraocular pressure while cutting the optic nerve at the time of enucleation could facilitate the passage of tumor cells through the vortex veins into systemic circulation thereby resulting in liver metastases. This became known as the “Zimmerman-McLean-Foster Hypothesis.” Although this hypothesis sparked controversy regarding the safety of enucleation, it also inspired new surgical techniques, such as the “no touch enucleation,” and encouraged the use of globe-salvaging therapies such as radiation and laser therapies. Further studies have refuted this hypothesis, and have attributed high frequencies of liver metastases to micrometastases that occur years prior to the initial clinical diagnosis of uveal melanoma.

The Collaborative Ocular Melanoma Study

The Collaborative Ocular Melanoma Study (COMS) is a prospective multicenter trial that was funded and organized by the National Eye Institute in 1985 to evaluate the role of different interventions for patients with uveal melanoma. The first patient was enrolled in 1987 and accrual was completed in 1998. To date, the COMS is the largest study performed in ocular oncology. There were three major prospective multicenter COMS trials (Table 3). The results of the COMS confirmed numerous previous publications regarding management of choroidal melanoma.

Large-size tumor trial

The “Zimmerman-McLean-Foster Hypothesis” inspired the use of additional treatments with enucleation, such as pre- or post-enucleation radiation, particularly in large choroidal melanomas. This trial sought to evaluate if pre-enucleation radiotherapy offered a survival benefit compared to enucleation alone.

Large choroidal melanomas were defined as 2.0 mm or more in apical height and greater than 16.0 mm in longest basal diameter, or more than 10.0 mm apical height regardless of basal diameter, or greater than 8.0 mm apical height regardless of basal diameter if less than 2.0 mm to the optic disc.

The randomized controlled prospective large-sized tumor trial involved 1003 treatment-naive patients with large choroidal melanomas who were treated with enucleation alone (n=506) or pretreatment external beam radiotherapy (20 Gy) followed by enucleation (n=497). In the pretreatment group, radiotherapy was delivered as five daily fractions of 4.0 Gy and eyes were enucleated within 80 hours after the final fraction. All eyes were enucleated within 4 weeks after randomization in both groups. Patients were monitored 6 and 12 months after enrollment, then annually. The primary outcome was survival rates in each treatment group. The 5-year survival rates were 57% in the enucleation alone group and 62% with pre-enucleation radiation. The authors concluded that there were no survival differences between pre-enucleation radiotherapy and enucleation alone in patients with large choroidal melanomas.

The 10-year survival rates were 40% in the enucleation alone group and 45% in the pretreatment radiation group. The authors concluded that no survival differences were observed.

Following these trials, pre-enucleation radiotherapy was abandoned.

Medium-size tumor trial

Following the concerns of the “Zimmermann-McLean-Foster Hypothesis,” radiotherapy was considered the best alternative to enucleation for medium-sized choroidal melanomas. This trial aimed to evaluate differences in survival rates in patients treated with iodine-125 plaque brachytherapy alone compared to treatment with enucleation alone.

Medium-sized choroidal melanomas were defined as 2.5 to 10.0 mm in apical height and no more than 16.0 mm in longest basal diameter.

The randomized controlled prospective medium-sized tumor trial involved 1317 treatment-naive patients with choroidal melanomas who were treated with enucleation (n=660) or iodine-125 plaque brachytherapy (n=657). Plaque brachytherapy treatment or enucleation was performed within 4 weeks after enrollment. The primary outcome was survival rates in each treatment group. The 5-year survival rates were 81% for the enucleation group and 82% for the iodine 125-plaque brachytherapy group (p=0.48). The authors concluded that there were no survival differences in patients treated with enucleation and patients treated with iodine-125 plaque brachytherapy.

The 12-year survival rates were 41% in the enucleation group and 43% in the iodine-125 plaque brachytherapy group.

The results of these trials encouraged the use of iodine-125 plaque brachytherapy for medium-sized choroidal melanomas.

Small-size tumor trial

The treatment of small melanomas has been controversial and many were managed with only observation. Treatment was often deferred in small melanomas because of the potential for severe visual impairment after radiation.

Small choroidal melanomas were defined as 1.0 to 3.0 mm in apical height and 5.0 to 16.0 mm in largest basal diameter.

The prospective observational small-size tumor trial involved 188 patients who were managed by observation. The primary outcome was tumor growth, which was defined as an increase to a medium-size or large-size choroidal melanoma based on COMS tumor classification criteria. In untreated small choroidal melanomas, 11% grew by 1 year, 21% by 2 years, and 31% by 5 years.

Table 7: Overview of Collaborative Ocular Melanoma Study for small, medium and large tumors

*Large-size choroidal melanomas were defined as 2.0 mm or greater in apical height and greater than 16.0 mm in longest basal diameter, or greater than 10.0 mm apical height regardless of basal diameter, or greater than 8.0 mm apical height regardless of basal diameter if less than 2.0 mm to the optic disc. **Medium-sized tumor was defined as 3.1 to 10.0 mm in apical height and 16.0 mm or less in longest basal diameter. ***Small-sized tumors were defined as 1.0 to 3.0 mm in apical height and 5.0 to 16.0 mm in longest basal diameter.

Trial	Design	Interventions	Findings
Large-size* tumor trial	Randomized controlled	Enucleation alone vs external beam radiotherapy followed by enucleation	No difference between 5-year survival rates between enucleation alone (57%) and enucleation with pretreatment external beam radiotherapy (62%) in large choroidal melanomas.
Medium-size** tumor trial	Randomized controlled	Enucleation vs iodine-125 plaque brachytherapy	No difference in 5-year survival rates between enucleation (81%) and iodine-125 plaque brachytherapy (82%) in medium-size choroidal melanomas.
Small-size*** tumor trial	Observational	Observation	Tumor growth in untreated small choroidal melanomas were 11% by 1 year, 21% by 2 years, and 31% by 5 years.

Additional Resources

1. Shields JA, Shields CL. Intraocular Tumors: A Text and Atlas. Philadelphia: WB Saunders, 1992.
2. Shields JA, Shields CL. Intraocular Tumors. An Atlas and Textbook. 2nd edition. Philadelphia, Lippincott Williams and Wilkins, 2008.
3. www.fighteyecancer.com
4. www.choroidalmelanoma.com
5. www.melanomaeye.com
6. www.malignantmelanomainfo.com