Abstract
Purpose :
Survival of patients affected by uveal melanoma (UVM), a major intraocular cancer, is greatly reduced by the development of metastasis, with one year survival being less than 20%. The molecular tumor characteristics associated with metastasis are not clear. The purpose of this study was to discover the gene expression changes between non-metastatic and metastatic UVM cases and evaluate their prognostic significance.
Methods :
We utilized the UVM RNA-seq dataset from The Cancer Genome Atlas (TCGA) to discover the genes associated with UVM metastasis and survival. Differential expression analyses between metastatic and non-metastatic tumors were performed using the “limma” R package. The hazard ratios (HR) were computed using Cox proportional hazards model to correlate differentially expressed genes with survival. Bioinformatics analyses were conducted to identify associated biological functions and pathways.
Results :
A total of 646 genes (>2-fold) were differentially expressed between metastatic and non-metastatic tumors (p<0.01) and 328 genes were significantly correlated with patient survival. The top five genes upregulated in metastasis and associated with reduced survival include, HTR2B (FC=24.12; HR=5.85), RIMS2 (FC=9.91; HR=6.79), VGF (FC=8.58; HR=8.71), MYEOV (FC=8.38; HR=7.33), and ISM1 (FC=7.22; HR=10.07). The top 5 genes downregulated in metastasis and associated with enhanced survival include, GSAT3 (FC= -10.81; HR=0.07), GATA4 (FC= -10.71; HR=0.04), MYO7B (FC= -7.40; HR=0.18), COL11A1 (FC= -6.56, HR=0.17), and SYNPR (FC= -6.42; HR=0.13). Functional annotation of the differentially expressed genes revealed a number of molecular and cellular functions including cell movement, growth, proliferation, and development.
Conclusions :
We identified several differentially expressed genes associated with metastasis in UVM patients. The genes significantly and highly correlated with patient survival may serve as potential prognostic factors or therapeutic targets. Several genes belonged to functional categories associated with cell movement and homeostasis, indicating their significance in metastasis. The findings from this study would aid in the development of prognostic and predictive biomarkers for metastatic UVM.
This is a 2020 ARVO Annual Meeting abstract.