The biology of uveal melanoma

Uveal melanoma (UM), a rare cancer of the eye, is distinct from cutaneous melanoma by its etiology, the mutation frequency and profile, and its clinical behavior including resistance to targeted therapy and immune checkpoint blockers. Primary disease is efficiently controlled by surgery or radiation therapy, but about half of UMs develop distant metastasis mostly to the liver. Survival of patients with metastasis is below 1 year and has not improved in decades. Recent years have brought a deep understanding of UM biology characterized by initiating mutations in the G proteins GNAQ and GNA11. Cytogenetic alterations, in particular monosomy of chromosome 3 and amplification of the long arm of chromosome 8, and mutation of the BRCA1-associated protein 1, BAP1, a tumor suppressor gene, or the splicing factor SF3B1 determine UM metastasis. Cytogenetic and molecular profiling allow for a very precise prognostication that is still not matched by efficacious adjuvant therapies. G protein signaling has been shown to activate the YAP/TAZ pathway independent of HIPPO, and conventional signaling via the mitogen-activated kinase pathway probably also contributes to UM development and progression. Several lines of evidence indicate that inflammation and macrophages play a pro-tumor role in UM and in its hepatic metastases. UM cells benefit from the immune privilege in the eye and may adopt several mechanisms involved in this privilege for tumor escape that act even after leaving the niche. Here, we review the current knowledge of the biology of UM and discuss recent approaches to UM treatment.     

Metastatic uveal melanoma: biology and emerging treatments

Uveal melanoma is the most common primary intraocular cancer in adults. Nearly half of primary uveal melanoma tumors metastasize, but there are currently no effective therapies for metastatic uveal melanoma. The recent discovery of mutations that underlie uveal melanoma metastasis, growth, and survival provide a key to the molecular understanding of this disease. Much work is now underway to leverage this knowledge to develop effective therapies. This review summarizes recently discovered molecular features of uveal melanoma and therapies being explored to capitalize on this knowledge.     

Proteomics in uveal melanoma

A new tool in the search for tumour markers is proteomic technology. Proteomics (or protein profiling) is the study of the proteome, the protein complement of the genome. The advantage of this technique in comparison with genomics is that the actual protein production can be measured. Gene microarrays determine levels of mRNA but do not necessarily predict the level of the corresponding proteins in a cell. In this study, we evaluated the use of proteomics in the aqueous humour of uveal melanoma patients compared with control patients using the surface-enhanced laser desorption ionization time-of-flight technique. The protein mass spectra of aqueous humours from 24 uveal melanoma eyes were compared with 24 control eyes using a strong anion exchange surface protein chip array. On the basis of two proteins (4543.43 and 6853.30 kDa), the aqueous humour of melanoma eyes and control eyes could be distinguished in 89% of cases. Therefore, proteomic evaluation might be helpful in finding diagnostic markers for uveal melanoma patients.     

Immunogenetic heterogeneity of uveal melanoma

Investigations have been performed to identify genetic markers in uveal melanoma (UM) patients. The immunogenetic heterogeneity of the histologically different forms of UM until now has been little analyzed. We subdivided our UM patients, all typed for class I and II HLA antigens and for Bf polymorphism, into two groups: 1) those with a high degree of malignancy (with nonspindle cells) and 2) those with a low degree of malignancy (with spindle cells). The deviated frequencies of class I HLA antigens (A32, B27) seem to be involved in the predisposition to spindle cell melanoma, while HLA class II (DR3, DR7) and class III (Bf F) strongly mark the worst form of UM. Different Gm allotype distributions between the two histological types of UM were also found.     

Animal models of uveal melanoma

Many attempts have been made to develop a suitable animal model to study more effectively the aetiology, pathogenesis, diagnosis and therapy of intraocular (uveal) melanoma. Uveal melanoma may spontaneously occur in some animals, including dogs, cats, horses, rats, mice, birds and fish. The histological features, metastatic behaviour and unpredictable nature of occurrence of these uncommon spontaneous tumours detract from their suitability as a model. Several methods have been developed to induce intraocular melanoma chemically or by radiation in laboratory animals. Some of these induced tumours resemble human uveal melanoma, although the majority originate from the retinal pigment epithelium. Uveal proliferations have been biologically induced by feline leukaemia/sarcoma virus and simian virus 40, although the presence of virus in tumour cells and extraocular tumours resulting from shed virus detract from the utility of this model. Inoculation of tissue culture hamster, murine or human melanoma cells into animal eyes has the advantage that the inoculation site and size of inoculum can be controlled. Disadvantages include the immune suppression necessary for tumour growth in some models as well as the fact that many of the melanoma cell lines are of cutaneous origin. Transgenic murine models have been developed using the promoter region of the tyrosinase gene to target expression of oncogenes in melanin-producing cells. Spontaneous intraocular pigmented tumours and distant metastases may occur, although many, if not all, of the intraocular tumours arise in the retinal pigment epithelium.     

The biology of melanoma metastasis

The process of cancer metastasis is sequential and selective and contains stochastic elements. The growth of melanoma metastases represents the endpoint of many lethal events that few tumor cells can survive. Primary tumors consist of multiple subpopulations of cells with heterogeneous metastatic properties, and the outcome of metastasis depends on the interplay of metastatic tumor cells with various host factors. This viewpoint is more optimistic than that of metastasis as a random process. A selective biological process is regulated by the interaction of tumor cells with their host, and these complex interactions can now be studied and manipulated.     

Targeted therapy for uveal melanoma

Uveal melanoma is the most common primary intra-ocular malignancy in adults. Overall mortality rate remains high because of the development of metastatic disease, which is highly resistant to systemic therapy. Improved understanding of the molecular pathogenesis of cancers has led to a new generation of therapeutic agents that interfere with a specific pathway critical in tumor development or progression. Although no specific genes have been linked to the pathogenesis of uveal melanoma, which differs from that of cutaneous melanoma, progress has been made in identifying potential targets involved in uveal melanoma apoptosis, proliferation, invasion, metastasis, and angiogenesis. This review focuses on the prospects for improving the systemic therapy of uveal melanoma using molecularly targeted agents that are currently in clinical use as well as agents being tested in clinical trials. Preclinical studies suggest potential benefit of inhibitors of Bcl-2, ubiquitin-proteasome, histone deactylase, mitogen-activated protein kinase and phosphatidylinositol-3-kinase-AKT pathways, and receptor tyrosine kinases. Modifiers of adhesion molecules, matrix metalloproteinase, and angiogenic factors also have demonstrated potential benefit. Clinical trials of some of these approaches have been initiated in patients with metastatic uveal melanoma as well as in the adjuvant setting after primary therapy.     

Prognostic factors in uveal melanoma

Uveal melanoma is the most common primary intraocular malignant tumour, with an annual incidence of approximately six cases per million per year. Approximately 40% of patients with posterior uveal melanoma develop metastatic melanoma to the liver within 10 years after initial diagnosis. Despite high accuracy of diagnosis and availability of various methods of treatment; the mortality due to uveal melanoma has remained unchanged. The prognosis in uveal melanoma depends on clinical, histopathological and cytological factors. Clinical factors that relate to prognosis include location, size, and configuration of the tumour. Uveal melanoma can arise in the iris, the ciliary body or the choroid. Iris melanomas have the best prognosis and ciliary body melanomas have the worst prognosis. Based on retrospective studies, the mortality rates for uveal melanoma for comparable sized tumours treated by enucleation or other globe conserving methods such as radiotherapy appear to be similar. Histopathological factors such as cell type, mitotic activity, microcirculation architecture, tumour-infiltrating lymphocytes and the presence of extrascleral extension are also significant predictors of survival. More recently, cytological factors such as cell proliferation, cytogenic, and molecular genetic prognostic markers have been identified with the hope of detecting high risk cases for adjuvant systemic immune therapy or chemotherapy. At present, the role of these therapeutic methods is not clearly established.     

Hormonal exposures and the risk of uveal melanoma

Objectives

Several studies suggest that hormonal mechanisms may be associated with the development of uveal melanoma. Therefore, the association between the risk of uveal melanoma and exposure to hormonal exposures was investigated in a case–control study from nine European countries.

Methods

Incident cases of uveal melanoma were frequency-matched to population and hospital controls by country, age, and sex. Female subjects were asked about their reproductive history, use of menopausal hormone replacement therapy and oral contraceptives. Among men, occupational handling of oils while working with transformers or capacitors which contain polychlorinated biphenyls (PCB) was solicited. Unconditional logistic regression analyses were calculated, adjusting for several potential confounders. Analyses were stratified by sex.

Results

Two hundred and ninety-three cases (165 men, 128 women) and 3,198 control subjects (2,121 men, 1,077 women) were interviewed. Among women, no associations were observed with hormonal status variables, intake of hormonal therapy or intake of oral contraceptives. Men showed an increased risk with occupational exposure to transformer/capacitor oils (OR = 2.74; Bonferroni-corrected 99.3% CI 1.07–7.02). However, these results were based on few exposed subjects only.

Conclusion

The results of this study do not support the hypothesis of a hormonal influence in the carcinogenesis of uveal melanoma. Our finding of a potentially increased risk with PCB-containing oils requires further research.     

Circulating tumor cells in uveal melanoma

Despite advances in the diagnosis and local tumor control, the overall mortality rate for uveal melanoma remains high because of the development of metastatic disease. The clinical and histopathological systems currently being used to classify patients are not sufficiently accurate to predict metastasis. Tumor genotyping has demonstrated significant promise but obtaining tumor tissue can be problematic. Furthermore, assessment of tumor tissue does not indicate whether tumor cells have actually been shed and cannot indicate whether treatment is reducing metastasis. The detection of circulating tumor cells in blood has been shown to be a prognostic biomarker that can be used to monitor the effectiveness of therapy in patients with metastatic carcinoma. Uveal melanoma disseminates hematogenously, and the detection of circulating melanoma cells may potentially be useful for diagnosis, risk stratification, and the monitoring of disease progression and treatment efficacy. PCR-based and immunomagnetic cell isolation techniques, derived from studies in patients with cutaneous melanoma, have been tested. For various biological and technical reasons, they have not demonstrated the accuracy and reproducibility required for an effective prognostic assay in patients with uveal melanoma. Assessments have been confounded by false positives and negatives and thus, correlations between circulating melanoma cells and survival have not yet been established. Circulating melanoma cell detection is a valuable tool for investigating metastasis in uveal melanoma and also has the potential to become a standard part of uveal melanoma management. However, more research on the biology of uveal melanoma as well as improvements upon the current technologies are needed.     

Tumor cell markers in uveal melanoma

Monoclonal antibodies to the melanoma-associated antigens HMB-45 and NKI/C3, and for S-100 protein were applied to archival tissue of 43 intraocular melanomas. Tn addition, the expression of the oncoproteins ras-p21 (ras 10) and mutated Ha-ras (E 184) as well as neu/erb-B2 (p185) were immunohistochemically evaluated. Incubation with antibodies to HMB-45 and NKI/C3 revealed consistently moderate to strong staining in all cases. Comparable ras-p21 immunostaining with normal epithelium observed in infiltrating components with a pronounced heterogeneous pattern, was particulary evident in epitheloid tumor cells. In melanomas of the spindle cell type B there was a tendency for patients with neu/erb-B2 positivity to have a worse prognosis. Using the chi-squared test for trend a significant correlation was found between S-100 reactivity, neu/erb-B2 amplification and the clinical outcome.     

New therapeutic agents in uveal melanoma

Uveal melanoma is the most common primary intraocular malignant tumour in adults. Five, ten and fifteen years after primary tumour treatment, up to 25%, 34% and 50% of patients may develop metastases, respectively. There are only a few systemic therapies that have been approved for uveal melanoma, all with doubtful efficacy. As the molecular knowledge over cancer has improved, new therapies are being developed. Several drugs, such as bortezomib, celecoxib, dacarbazine, anti-angiogenic agents (such as bevacizumab, sorafenib and sunitinib), temsirolimus, mitogen-activated protein kinase kinase (MEK) inhibitors, ipilimumab and AEB071 are candidate drugs, and studies are underway to determine the therapeutic effects of these drugs in uveal melanoma.     

Lymphocytic infiltration in uveal malignant melanoma

A study relating the intralesional infiltration of lymphocytes and plasma cells to patient survival was performed on cases of uveal malignant melanoma accessed at the Armed Forces Institute of Pathology, Washington, DC (AFIP) between 1954 and 1971. The authors examined 1193 cases using light microscopy. Of the 1078 cases with technically acceptable histologic sections, 134 tumors contained 100 or more lymphocytes per 20 high-power (X400) microscopic fields (20 HPF). The prevalence was 12.4%. This was designated the "high lymphocytic" group. An equivalent number of cases with fewer lymphocytes comprised the "low lymphocytic" group. The survival rate at 15 years was 36.7% for patients in the high lymphocytic group and 69.6% for patients in the low lymphocytic group. Using the Cox model, the authors found that an increased number of lymphocytes per 20 HPF was significantly associated with decreased survival (chi-square = 21.2, P = less than 0.0001). A significant association was observed even when we controlled for other risk factors (chi-square = 6.98, P = 0.008).     

Upcoming translational challenges for uveal melanoma

The past few years have witnessed major advances in the understanding of the molecular landscape of uveal melanoma (UM). The discovery of a mutational background that is completely different from the one of skin melanoma has granted to UM a stand-alone status. The absence of effective therapy for metastatic disease offers now a chessboard for targeted therapy but at the same time urges preclinical science to develop accordingly, to guide the use of economical resources to the best profit of patients. This review describes the current knowledge on the biology of this disease and discusses the challenges that must be undertaken to translate this knowledge into real benefit for patients.     

Lack of BRAF mutations in uveal melanoma

RAF proteins are serine/threonine kinases that mediate cellular responses to growth signals by activating the mitogen-activated protein kinase pathway. Mutations in the BRAF gene causing a V599E amino acid substitution that enhance the kinase activity have been described in >60% of cutaneous melanomas and premalignant melanocytic lesions. We have investigated the frequency of BRAF mutations at the expression level in melanomas of the uveal tract. None of the 30 metastases and 10 primary uveal melanomas tested expressed the V599E mutation. In contrast, this mutation was expressed by 65% of cutaneous melanoma samples, confirming previous results. In addition, a double mutation resulting in V599K substitution was detected in two suspect ocular metastases of cutaneous melanoma. Analysis of exon 11, the second common site of BRAF mutations, revealed only wild-type sequences in uveal melanomas. Analysis of tumor lysates showed the presence of phosphorylated mitogen-activated protein kinase, kinase, and mitogen-activated protein kinase in 50% of uveal and 100% of cutaneous melanoma metastases. Taken together, these results suggest that although the common BRAF mutations found in cutaneous melanoma do not play a role in tumorigenesis of uveal tract melanocytes, activation of the RAF/mitogen-activated protein kinase pathway may nevertheless play an important role in uveal melanoma.     

Immunohistochemical expression of phospho-Akt in uveal melanoma

The aim of this study was to evaluate the immunohistochemical expression of phospho-Akt and its possible association with clinicopathological features in uveal melanoma. Thirty-four enucleated eyes from 34 patients with choroidal melanoma were included in the study. Patients were divided into two groups based on the treatment received: (1) primary enucleation (n=18); (2) radiotherapy, either external beam or brachytherapy, and enucleation (n=16). Clinicopathological data were obtained. The minimum follow-up time was 72 months. Immunohistochemistry for phospho-Akt was performed using an anti-phospho-Akt (Ser 473) rabbit antibody. The association of phospho-Akt with clinicopathological parameters was investigated in each patient group separately. Phospho-Akt immunostaining was cytoplasmic in both groups. In the primary enucleation group, 10 tumours were phospho-Akt positive (55.5%). Patients with phospho-Akt-positive tumours were older (average 70.8 years versus 59 years, P=0.01) and phospho-Akt immunoreactivity was significantly associated with a higher risk of metastatic disease (Kaplan-Meier analysis, P=0.02). In the radiotherapy and enucleation group, nine tumours were phospho-Akt positive (56.2%). The absence of phospho-Akt expression was correlated with male gender (P=0.02). The following conclusions can be drawn from this study: (1) phospho-Akt immunoexpression was detected in 55.5% of uveal melanomas treated with primary enucleation and in 56.2% of uveal melanomas treated with radiotherapy and enucleation; (2) the association of phospho-Akt immunoexpression with clinicopathological features, including prognosis, merits further study.