Prognostic Value of Autophagy-related Genes Correlated With Metastasis in Uveal Melanoma Patients

Abstract Half of the patients with primary uveal melanoma will develop progressive metastasis, leading to high mortality rate. Autophagy has been demonstrated to engage in metastasis in multiple tumors. Detection, diagnosis and treatment at the early-stage of uveal melanoma may help prevent potential tumor progression and optimize the prognosis. The purpose of our study was to discover autophagy-related genes (ARGs) correlated with uveal melanoma metastasis and determine their prognostic values. We analyzed the gene expression profiles and the clinical data from the Gene Expression Omnibus (GEO) database in uveal melanoma. A total of 14 and 16 differentially expressed ARGs were identified to be related to uveal melanoma metastasis from GSE22138 and GSE27831 sets. The two datasets shared three common genes including RAF1, CDKN1A and WIPI1 that occupied the core positions in the Protein-Protein Interactions (PPI) Network of ARGs. Following that, TCGA was introduced for survival analysis of the three genes. The survival analysis showed that high expression of RAF1 was related to favorable prognosis of uveal melanoma, whereas high expression of CDKN1A and WIPI1 suggested poor prognosis. Then a three-ARG based prognostic risk score model was constructed to predict survival outcomes. Univariate and multivariate Cox regression analyses indicated that the risk score can be considered as an independent prognostic factor for uveal melanoma, exhibiting good accuracy and sensitivity. In summary, we established an autophagy-related prognostic model based on uveal melanoma metastasis, which may contribute to the detection of early metastasis and prediction of prognosis, thereby prolonging survival through early personalized intervention.