Distinguishable Prognostic Signatures of Left- and Right-Sided Colon Cancer: a Study Based on Sequencing Data

Abstract

Background/Aims: Left- and right-sided colon cancers are considered to be two different diseases and have altered outcomes. However, specific molecules to predict the prognosis of left- and right-sided colon cancers are currently lacking. Methods: Expression profiling of colon cancer were downloaded from The Cancer Genome Atlas (TCGA). Differentially expressed genes (DEGs) of left- and right-sided colon cancers were compared by DESeq analysis. The prognostic values of DEGs were assessed by univariate and multivariate Cox regression. Prognostic index models of two side colon cancers were conducted with prognostic values genes, respectively. Interaction of DEGs was then analyzed by the protein-protein interaction (PPI). Different biology function of two sides of colon cancer was assessed by Gene Set Enrichment Analysis (GSEA). Results: A total of 167 DEGs were identified between left- and right-sided colon cancers based on TCGA data. Using univariate COX regression analysis, five genes (PHACTR3, CKMT2, CYP2W1, ERFE, HOXC4) were related to overall survival in left-sided, and eight distinguishable genes (EREG, ERFE, HOXC6, SLC22A31, TFF1, GFI1, ZG16, RASL10B) in right-sided. Further, left-sided prognostic model was established with PHACTR3 and CKMT2 (HR=2.040; 95%CI=1.004-4.145; P=0.049). Distinguishable prognostic signature for right-sided colon cancer was established based on EREG, ERFE, GFI1, and RASL10B (HR=3.530; 95%CI: 1.934-6.444; P< 0.001) in multivariate analysis. PPI analysis of 167 DEGs showed that CCL5, GNG4, GNLY, GZMH, DRD2, and FASLG genes were at the core of interaction network. In GSEA function analysis, four pathways, including antigen processing and presentation, natural killer cell mediated cytotoxicity, intestinal immune network for Iga production, and type I diabetes mellitus, were significantly enriched in the DEGs of the right-sided colon cancer. Conclusions: This study constructs a panel of potential prognostic model of left- and right-sided colon cancers, respectively. We also provide molecular biological alterations between left- and right-sided colon cancers.