Abstract
BACKGROUND: Growing evidence has shown that genetic or epigenetic alterations are highly involved in the initiation and progression of renal cell carcinoma (RCC). This study aimed to find prognostic methylation markers in clear cell RCC (ccRCC). METHODS: In this study, we developed and confirmed an integrated and comprehensive methylation signature by integrating DNA methylation, gene expression, and The Cancer Genome Atlas (TCGA) survival data. First, the methylation signature was found and checked based on data analysis of published datasets. Then, independent predictive factors were selected using the Cox proportional model and incorporated into the nomogram. Finally, the predictive nomogram was derived and validated using a concordance index and calibration plots. RESULTS: A series of differentially expressed and methylated genes were identified. After intersection analysis, Gene Ontology (GO) analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, protein-protein interaction (PPI) analysis, and correlation analysis, FCGR1A, F2, and NOD2 were established as a predictive signature. According to the Kaplan-Meier survival analysis, the risk score system based on the predictive signature was able to stratify the patients into high- and low-risk groups with significantly different overall survival. The receiver operating characteristic (ROC) analysis further showed that the predictive signature yielded high sensitivity and specificity in predicting the prognosis outcome of ccRCC patients. Moreover, univariate and multivariate Cox regression analysis confirmed that the three-gene methylation signature was an independent prognostic factor in ccRCC. Finally, a nomogram comprising the predictive signature and several independent variables were constructed and proved to effectively predict ccRCC patient survival. CONCLUSIONS: The three-gene methylation signature was revealed to be a potential novel and independent adverse predictor of prognosis for ccRCC patients and may serve as a promising marker for treatment management and survival outcome improvement. However, substantial validation experiments are required to characterize the molecular background of the predictive signature.