Uveal melanoma (UM) is the most common form of intraocular malignancy and the second most common site of primary melanoma after the skin. UMs account for about 80% of all the neoplasms in the eye and originate from melanocytes situated in the uvea. The uveal tract is the middle vascular layer of the eye contributing the blood supply to the sclera. The choroid, ciliary body and iris are all continuous with each other and make up the uvea (see Figure). More than 90% of UM arise from the choroid, while iris and ciliary body melanomas are far less frequent. 

 (click to enlarge)

Malignant melanomas can also develop, although less commonly, from the conjunctival membrane and the orbit. Metastatic deposits to the ocular region from a primary cutaneous site can also occur, but these are rare and account for less than 5% of all metastases to the eye and orbit. Overall, around 3.7% of all melanoma cases are of primary intraocular origin. In this review we will focus on primary melanomas from the uvea as they represent a distinct entity with its own peculiar clinical and pathological characteristics. 

UM is a rare cancer and its incidence is reported at between two and eight cases per million per year in Caucasians. Both sexes are almost equally affected, although a slight prevalence in males has been described with a male to female ratio of approximately 1.3. 

In Europe UM incidence shows a north to south gradient, decreasing from over eight cases per million in northern to less than two per million in southern countries. The age of presentation peaks between the sixth and the seventh decade, but it is remarkably lower for anterior UM which is more common among patients aged less than 20 years. 

Several risk factors for UM have been identified and include light-coloured irides, congenital ocular melanocytosis, melanocytoma and neurofibromatosis. Familial cases of UM are very rare and the gene BAP1 on chromosome 3 has been implicated in the development of UM, mesothelioma and other cancers. It has been theorised that exposure to ultraviolet radiation can increase the risk of neoplasia although this remains unproven.¹

Clinical presentation

Most patients with UM present with painless visual disturbance or sudden loss of vision. Presentation depends upon the size and location of the lesion (iris versus choroid or ciliary body). The most common symptoms are of blurred vision, visual field defect, photopsia, irritation and pain. 

Tumour growth can lead to secondary retinal detachment leading to visual loss. UM usually presents as a dome or mushroom-shaped subretinal mass. The diagnosis is mostly established by ophthalmic examination and the accuracy of diagnosis is nowadays very high, over 99%. 

Interestingly, UM is one of the very few cancers that are treated even in the absence of a definite histopathological diagnosis. Approximately 30% of patients are completely asymptomatic and are incidentally diagnosed on routine examination of the eye while undergoing eye testing. Staging for UM follows the TNM staging system. 

Staging and primary therapy

Staging

Prior to planning primary ocular therapy, a systemic work-up is mandatory to rule out metastatic disease. Less than 5% of patients are usually found to have distant metastases at the time of first diagnosis. 

In view of the peculiar pattern of metastatic diffusion, systemic staging should always include an adequate imaging of the liver. Usually a four-phase intravenous (IV) contrast-enhanced computed tomography (CT) of the liver, or MRI of the liver with IV Primovist, are the methods of choice to exclude liver metastases. 

As the risk of extra-hepatic metastases at first presentation is extremely low, a chest x-ray is considered adequate to complete staging in the absence of metastatic disease in the liver. 

Surgery 

Based on their thickness, UMs are grouped as small-sized (< 3mm), medium-sized (3-5mm) and large-sized (> 5mm). Historically, orbital exenteration (or enucleation) was the treatment of choice for UM. It still is the preferred option for large tumours with very limited or no chance of salvaging vision. However, over the past 15 years there has been a significant shift towards vision-preserving management after a pivotal randomised trial (COMS trial) showed no difference in survivorship between enucleation and brachytherapy for small and medium-sized UMs.²

Local (endoscleral) resection can be an alternative treatment option for a tumour that allows for removal of the tumour with retention of the eye. It also provides tissue for histopathological and cytogenetic analysis. Adjuvant brachytherapy can then be used to reduce the likelihood of local recurrence.

Plaque brachytherapy

This involves the application of gold-shielded episcleral plaques using a variety of different isotopes, mostly Ruthenium-106 or Iodine-125. The plaque is placed on the external surface of the eye covering the base of the intraocular tumour and is usually removed after two to four days. 

Proton-beam external radiotherapy has also been used and there is a limited experience in using stereotactic radiotherapy, although the radiation dose distribution is not as precise as with some other techniques.

Prognosis

One of the most challenging aspects of managing patients with UM is to assess their risk of systemic recurrence after therapy of the primary tumour and consequently to consult in relation to long-term outcome. 

Patients with small tumours, especially if located in the iris, tend to have an excellent outcome and their metastatic rate at 10 years is between 4-7%. Conversely, patients with large tumours located in the choroid have a 10-year metastatic rate up to 25-30%. Cytogenetic alterations, especially monosomy of chromosome 3 and aberrations of chromosome 8 (loss of chromosome 8p), have been strongly associated with an increased risk of metastases and shorter overall survival. 

More recently, gene expression profiling provides the best prediction of the metastatic potential of UM, although it is not routinely used yet, mostly due to its high cost. 

Surveillance

As there are no lymphatics in the uvea, UM spreads haematologically and it is well known that the liver is the first and only site at the time of systemic relapse in up to 80-85% of patients. Therefore, several strategies have been proposed to monitor patients after primary therapy in order to allow the early identification of metastatic disease that would be amenable of radical resection. 

There is no consensus as to the frequency of surveillance and the optimal imaging modality. It also remains controversial whether earlier detection of asymptomatic metastatic disease would lead to a survival benefit. 

At St Vincent’s University Hospital in 2009, we established a regular surveillance imaging plan for patients with UM after primary therapy. Out of 154 patients followed up for a median of 31 months, we diagnosed 80% of asymptomatic metastatic relapses on surveillance imaging and almost 50% of them had resectable disease and underwent liver metastasectomy. 

Interestingly, we observed that progression-free survival (PFS) and overall survival (OS) were superior for patients who had resectable liver metastases compared to those who had not (PFS = 9.6 vs 5.5 months; OS 27.9 vs 13.4 months, respectively). 

Overall, most relapses were observed in patients with medium and large-sized tumours. These data suggest that an active surveillance strategy, especially in patients with higher-risk disease, could lead to anticipation in the diagnosis of hepatic metastases and of a higher chance of radical resection and prolonged remission.³

Unfortunately, most patients would develop intra-hepatic relapse after liver metastasectomy, so larger groups of patients and longer follow-up are necessary to draw any final conclusions. 

Metastatic disease

Historical data indicate that the median survival after liver metastases develop is 2-12 months and the one-year survival rate is 10-15%. Hepatic failure due to metastases is the most common cause of death in patients with metastatic UM. 

Numerous different techniques have been used to treat metastatic disease but given the rarity of this disease, there is a substantial lack of data from prospective and adequately dimensioned clinical trials. In most cases, systemic therapies are extrapolated from the cutaneous counterpart, although in view of the remarkable differences between these two entities this is not an entirely satisfactory and justified approach.  

Surgery of metastases

Liver metastases of UM are often multifocal; however, active radiological surveillance could increase the rate of asymptomatic, solitary hepatic metastases, which allow for surgical resection. There are no randomised studies investigating the role of surgery in the management of liver metastases of UM, although single institution data have shown improvement in survival with liver resection. 

Given the quite unique natural history of metastatic UM, all patients with newly diagnosed liver metastases should be considered for upfront surgical resection before any other therapeutic intervention. It is crucial that these patients are discussed and treated by a multidisciplinary team. 

Percutaneous intra-hepatic chemosaturation 

This treatment has emerged over the past decade as a promising option for patients with metastatic UM confined to the liver but who are not candidates for a surgical resection. In this treatment a hepatic arterial catheter is surgically sited and a chemotherapy agent – usually melphalan – is delivered locally and then captured and filtered extracorporeally prior to return to the systemic circulation via a veno-venous bypass circuit. This allows for maximum drug exposure within the liver and rapid systemic clearance. 

A recently published randomised controlled trial on 93 patients with metastatic melanoma (including primary UM) to the liver showed a significant improvement in objective response and hepatic progression-free survival compared to the best alternative care. No overall survival advantage was observed, possibly due to cross-over between the two treatment arms. 

Patients are reassessed for response after each treatment and the procedure is repeated after six weeks in case of response. Usually each patient would receive between four and five treatments. Intra-hepatic chemosaturation is an invasive procedure and patients are usually hospitalised for two or three days. It is costly and can have complications and side-effects, such as neutropaenia, thrombocytopaenia and transaminitis.⁴ Therefore, patient selection is crucial in order to achieve the best results and minimise the risk of serious adverse events. 

Other local therapies

Hepatic trans-arterial chemoembolisation (TACE) combines arterial embolisation with infusion of cytotoxic drugs and has been shown to be effective in the treatment of primary hepatocellular carcinomas and hepatic metastases. However, it is not considered as the preferred treatment option for patients with metastatic UM unless surgery and intra-hepatic chemosaturation have already been excluded or are not available. 

Stereotactic radiotherapy has been proposed for patients with oligometastatic disease who are not candidates for surgery or intra-hepatic chemosaturation. The experience is quite limited in this setting although it could be a reasonable option when a non-invasive procedure is preferred. 

Systemic therapy

Chemotherapy

Metastatic UM is considered as a chemo-refractory disease as the objective response rate to cytotoxic agents has historically been reported to be less than 10%. Moreover, no randomised trial of chemotherapy has demonstrated an improvement in survival over supportive care. 

As a result, there are no standard chemotherapeutic regimes and the most common approach has been to treat patients with single-agent dacarbazine or with combination chemotherapy (less often) as done in cutaneous melanoma. 

Immunotherapy

Treatment of metastatic melanoma changed dramatically about five years ago when the first immune checkpoint inhibitor – the anti-CTLA4 monoclonal antibody ipilimumab – was approved following demonstration of improved overall survival in a large phase 3 study on cutaneous melanoma. More recently, two programmed death-1 protein (PD-1) inhibitors – pembrolizumab and nivolumab – have been found to be more active and less toxic than ipilimumab. It is crucial to note that patients with UM have been systematically excluded from all the immunotherapy trials, so there is no prospective randomised evidence that these novel agents can have the same clinically significant impact on metastatic UMs as demonstrated for cutaneous melanomas. 

The molecular pathology of UM is remarkably different from its cutaneous counterpart, therefore it is not surprising that the current data seem to indicate an inferior efficacy of the immune checkpoint inhibitors on UMs. 

Targeted therapies

The vast majority of UMs do not harbour BRAF mutations, but about 80% of UMs harbour mutations in GNAQ or GNA11 which may potentially act as drivers of mitogen-activated protein kinase kinase (MAPK/ERK kinase). The MAPK/ERK kinase 1 and 2 inhibitor selumetinib has shown promising results in a randomised phase 2 trial and a larger phase 3 trial is under way.⁵

Conclusion

UM is a rare cancer but it still accounts for significant morbidity and mortality in a proportion of affected patients. As UM is uncommon it is important that patients are treated at institutions where there is a sufficient number of cases to ensure the highest standard of care. 

Despite the remarkable improvements in the local therapy of primary UMs, overall survival of metastatic UM has remained the same for many decades and access to clinical trials of novel therapies should be implemented.  

References

1. Jovanovic P, Mihajlovic M, Djordjevic-Jocic J, et al. Ocular melanoma: an overview of the current status. Int J Clin Exp Pathol 2013; 6: 1230-1244
2. Collaborative Ocular Melanoma Study Group. The COMS randomized trial of iodine 125 brachytherapy for choroidal melanoma: V. Twelve-year mortality rates and prognostic factors: COMS report No. 28. Arch Ophthalmol 2006; 124(12): 1684-1693 
3. Lee CL, Gullo G, Horgan N, et al. Outcome of active surveillance and primary hepatic metastasectomy on the outcome of uveal melanoma (UM): analysis from a population-based tertiary centre database. Submitted 
4. Hughes MS, Zager J, Faries M, et al. Results of a Randomized Controlled Multicenter Phase III Trial of Percutaneous Hepatic Perfusion Compared with Best Available Care for Patients with Melanoma Liver Metastases. Ann Surg Oncol. 2016; 23: 1309-1319
5. Shoushtari AN, Carvajal RD. Treatment of Uveal Melanoma. Cancer Treat Res. 2016; 167: 281-293