An Optimized NGS Workflow Defines Genetically Based Prognostic Categories for Patients with Uveal Melanoma.

Abstract

Background: Despite advances in uveal melanoma (UM) diagnosis and treatment, about 50% of patients develop distant metastases, thereby displaying poor overall survival. Molecular profiling has identified several genetic alterations that can stratify patients with UM into different risk categories. However, these genetic alterations are currently dispersed over multiple studies and several methodologies, emphasizing the need for a defined workflow that will allow standardized and reproducible molecular analyses.

Methods: Following the findings published by "The Cancer Genome Atlas-UM" (TCGA-UM) study, we developed an NGS-based gene panel (called the UMpanel) that classifies mutation sets in four categories: initiating alterations (CYSLTR2, GNA11, GNAQ and PLCB4), prognostic alterations (BAP1, EIF1AX, SF3B1 and SRSF2), emergent biomarkers (CDKN2A, CENPE, FOXO1, HIF1A, RPL5 and TP53) and chromosomal abnormalities (imbalances in chromosomes 1, 3 and 8).

Results: Employing commercial gene panels, reference mutated DNAs and Sanger sequencing, we performed a comparative analysis and found that our methodological approach successfully predicted survival with great specificity and sensitivity compared to the TCGA-UM cohort that was used as a validation group.

Conclusions: Our results demonstrate that a reproducible NGS-based workflow translates into a reliable tool for the clinical stratification of patients with UM.