Identification of Long Non-Coding RNA Signatures for Specific Disease-Free Prognosis in Clear Cell Renal Carcinoma

Abstract

Patients with clear cell renal carcinoma (ccRCC) often relapse after nephrectomy, even for those with localized or locally advanced diseases. However, ideal prognostic biomarkers specifically for localized, locally advanced, and metastatic ccRCC are inadequate. In this paper, we systematically identified potential biomarkers for disease-free prognosis of patients with ccRCC. The expression and clinical data of The Cancer Genome Atlas (TCGA, n = 603) and international cancer genomics consortium (ICGC, n = 392) were leveraged to identify long non-coding RNA (lncRNA)-based prognostic markers. The expression data of GSE53757 (n = 144) and GSE66270 (n = 28) from Gene Expression Omnibus (GEO) were used to screen for ccRCC-related differentially expressed (DE) genes. 893 DE lncRNAs in common of the three datasets were screened out [Benjamini-Hochberg (BH) adjusted P <; 0.05]. Univariate and multivariate Cox analysis revealed that the 2-lncRNA signature (LINC00176 and CTD-2145C24.5) could be an independent prognostic marker for DFS of patients with localized ccRCC (log-rank P = 0.006). Another 5-lncRNA signature (DSCR9, RP11-271C24.2, RP11-424M24.5, CTD-2171N6.1, and CTC-499B15.8) have great potential in the disease-free prognosis not only of patients with locally advanced (log-rank P <; 0.001) disease but also for patients with metastatic disease (log-rank P <; 0.001). Functional analysis shown that the 2-lncRNA signature was associated with protein deacetylation and RNA splicing (BH-adjusted P <; 0.05), while 5-lncRNA signature was associated with deoxyribonuclease activity, RNA-dependent ATPase activity, and helicase activity (BH-adjusted P <; 0.05). We identified two lncRNA signatures for DFS prognosis of patients with ccRCC, which may be valuable clinical tools at molecular-level.

Keywords

• Biomarker
• biomedical computing
• bioinformatics