Integrative Analysis Identifies Multi-Omics Signatures That Drive Molecular Classification of Uveal Melanoma

SIMPLE SUMMARY: Uveal melanoma (UM) is a heterogeneous disease driven by accumulative alterations at multi-omics levels including genomics, epigenomics and transcriptomics. It is of great clinical interest to identify UM molecular subtypes and subtype-specific biomarkers that could be used for prognosis and targeted therapy. Integrative clustering (iCluster) analysis is a new approach designed to identify cancer integrative subtypes (iSubtypes) and multi-omics signatures that drive molecular classification of cancer. In this study, we performed an iCluster analysis of UM multi-omics data and identified four UM iSubtypes, which formed two major iSubtypes (denoted by M3 and D3) with distinct multi-omics landscapes. We showed that the integrative molecular classification of UM was determined by concordant alterations at multi-omics levels including DNA copy number, DNA methylation, gene expression and somatic mutation. We further derived a gene panel that can be used to classify UM into high- or low-risk groups for metastasis. ABSTRACT: By iCluster analysis, we found that the integrative molecular classification of the UM was primarily driven by DNA copy number variation on chromosomes 3, 6 and 8, differential methylation and expression of genes involved in the immune system, cell morphogenesis, movement and migration, and differential mutation of genes including GNA11, BAP1, EIF1AX, SF3B1 and GNAQ. Integrative analysis revealed that pathways including IL6/JAK/STAT3 signaling, angiogenesis, allograft rejection, inflammatory response and interferon gamma response were hypomethylated and up-regulated in the M3 iSubtype, which was associated with a worse overall survival, compared to the D3 iSubtype. Using two independent gene expression datasets, we demonstrated that the subtype-driving genes had an excellent prognostic power in classifying UM into high- or low-risk groups for metastasis. Integrative analysis of UM multi-omics data provided a comprehensive view of UM biology for understanding the underlying mechanism leading to UM metastasis. The concordant molecular alterations at multi-omics levels revealed by our integrative analysis could be used for patient stratification towards personalized management and surveillance.