Uveal melanoma is an aggressive malignancy that originates from melanocytes in the eye. Even if the primary tumor has been successfully treated with radiation or surgery, up to half of all UM patients will eventually develop metastatic disease. Despite the common origin from neural crest-derived cells, uveal and cutaneous melanoma have few overlapping genetic signatures and uveal melanoma has been shown to have a lower mutational burden. As a consequence, many therapies that have proven effective in cutaneous melanoma -such as immunotherapy-have little or no success in uveal melanoma. Several independent studies have recently identified the underlying genetic aberrancies in uveal melanoma, which allow improved tumor classification and prognostication of metastatic disease. In most cases, activating mutations in the Gα11/Q pathway drive uveal melanoma oncogenesis, whereas mutations in the BAP1, SF3B1 or EIF1AX genes predict progression towards metastasis. Intriguingly, the composition of chromosomal anomalies of chromosome 3, 6 and 8, shown to correlate with an adverse outcome, are distinctive in the BAP1 mut , SF3B1 mut and EIF1AX mut uveal melanoma subtypes. Expression profiling and epigenetic studies underline this subdivision in high-, intermediate-, or low-metastatic risk subgroups and suggest a different approach in the future towards prevention and/or treatment based on the specific mutation present in the tumor of the patients. In this review we discuss the current knowledge of the underlying genetic events that lead to uveal melanoma, their implication for the disease course and prognosis, as well as the therapeutic possibilities that arise from targeting these different aberrant pathways.

Uveal melanoma is a highly aggressive cancer of the eye, in which nearly 50% of the patients die from metastasis. It is the most common type of primary eye cancer in adults. Chromosome and mutation status have been shown to correlate with the disease-free survival. Loss of chromosome 3 and inactivating mutations in BAP1, which is located on chromosome 3, are strongly associated with 'high-risk' tumors that metastasize early. Other genes often involved in uveal melanoma are SF3B1 and EIF1AX, which are found to be mutated in intermediate- and low-risk tumors, respectively. To obtain genetic information of all genes in one test, we developed a targeted sequencing method that can detect mutations in uveal melanoma genes and chromosomal anomalies in chromosome 1, 3, and 8. With as little as 10 ng DNA, we obtained enough coverage on all genes to detect mutations, such as substitutions, deletions, and insertions. These results were validated with Sanger sequencing in 28 samples. In >90% of the cases, the BAP1 mutation status corresponded to the BAP1 immunohistochemistry. The results obtained in the Ion Torrent single-nucleotide polymorphism assay were confirmed with several other techniques, such as fluorescence in situ hybridization, multiplex ligation-dependent probe amplification, and Illumina SNP array. By validating our assay in 27 formalin-fixed paraffin-embedded and 43 fresh uveal melanomas, we show that mutations and chromosome status can reliably be obtained using targeted next-generation sequencing. Implementing this technique as a diagnostic pathology application for uveal melanoma will allow prediction of the patients' metastatic risk and potentially assess eligibility for new therapies.

No abstract

Uveal melanoma (UM) is the most frequently found primary intra-ocular tumor in adults. It is a highly aggressive cancer that causes metastasis-related mortality in up to half of the patients. Many independent studies have reported somatic genetic changes associated with high metastatic risk, such as monosomy of chromosome 3 and mutations in BAP1. Still, the mechanisms that drive metastatic spread are largely unknown. This study aimed to elucidate the potential role of microRNAs in the metastasis of UM. Using a next-generation sequencing approach in 26 UM samples we identified thirteen differentially expressed microRNAs between high-risk UM and low/intermediate-risk UM, including the known oncomirs microRNA-17-5p, microRNA-21-5p, and miR-151a-3p. Integration of the differentially expressed microRNAs with expression data of predicted target genes revealed 106 genes likely to be affected by aberrant microRNA expression. These genes were involved in pathways such as cell cycle regulation, EGF signaling and EIF2 signaling. Our findings demonstrate that aberrant microRNA expression in UM may affect the expression of genes in a variety of cancer-related pathways. This implies that some microRNAs can be responsible for UM metastasis and are promising potential targets for future treatment.

Background: Uveal melanoma (UM) is the most common primary ocular malignancy in adults in the Western world. UM with a mutation in SF3B1, a spliceosome gene, is characterized by three or more structural changes of chromosome 1, 6, 8, 9, or 11. Also UM without a mutation in SF3B1 harbors similar chromosomal aberrations. Since, in addition to SF3B1, mutations in U2AF1 and SRSF2 have also been observed in hematological malignancies, UM without a SF3B1 mutation—but with the characteristic chromosomal pattern—might harbor mutations in one of these genes. Methods: 42 UMs were selected based on their chromosomal profile and wildtype SF3B1 status. Sanger sequencing covering the U2AF1 (exon 2 and 7) hotspots and SRSF2 (exon 1 and 2) was performed on DNA extracted from tumor tissue. Data of three UM with an SRSF2 mutation was extracted from the The Cancer Genome Atlas (TCGA). Results: Heterozygous in-frame SRSF2 deletions affecting amino acids 92–100 were detected in two UMs (5%) of 42 selected tumors and in three TGCA UM specimens. Both the UM with an SRSF2 mutation from our cohort and the UM samples from the TCGA showed more than four structural chromosomal aberrations including (partial) gain of chromosome 6 and 8, although in two TCGA UMs monosomy 3 was observed. Conclusions: Whereas in myelodysplastic syndrome predominantly missense SRSF2 mutations are described, the observed SRSF2 mutations in UM are all in-frame deletions of 8–9 amino acids. This suggests that the R625 missense SF3B1 mutations and SRSF2 mutations in UM are different compared to the spliceosome gene mutations in hematological cancers, and probably target a different, as yet unknown, set of genes involved in uveal melanoma etiology.

The prevalence of mutations in the UM related genes is not different in polyploid UM compared with diploid UM. Moreover, similar to patients with diploid UM, BAP1 mutation is the most significant prognostic predictor of metastasis in patients with polyploid UM.

Uveal melanoma (UM) is an aggressive intra-ocular cancer with a strong tendency to metastasize. Metastatic UM is associated with mutations in BAP1 and SF3B1, however only little is known about the epigenetic modifications that arise in metastatic UM. In this study we aim to unravel epigenetic changes contributing to UM metastasis using a new genome-wide methylation analysis technique that covers over 50% of all CpG’s. We identified aberrant methylation contributing to BAP1 and SF3B1-mediated UM metastasis. The methylation data was integrated with expression data and surveyed in matched UM metastases from the liver, skin and bone. UM metastases showed no commonly shared novel epigenetic modifications, implying that epigenetic changes contributing to metastatic spreading and colonization in distant tissues occur early in the development of UM and epigenetic changes that occur after metastasis are mainly patient-specific. Our findings reveal a plethora of epigenetic modifications in metastatic UM and its metastases, which could subsequently result in aberrant repression or activation of many tumor-related genes. This observation points towards additional layers of complexity at the level of gene expression regulation, which may explain the low mutational burden of UM.

WD and SY contributed equally to the work presented here and should therefore be regarded as equivalent first authors. AO and JvR contributed equally to the work presented here and should therefore be regarded as equivalent second authors. See the appendix for the members of the Rotterdam Ocular Melanoma Study Group.