Establishment and verification of prognostic model and ceRNA network analysis for colorectal cancer liver metastasis
Xuan Zhang1, Tao Wu1, Jinmei Zhou2
1Department of Colorectal Surgery, Yunnan Cancer Hospital, The Third Affiliated Hospital of Kunming Medical University, Kunming, China.
Objects:
Colorectal cancer (CRC) is one of the most common cancers in the world. Approximately two-thirds of patients with CRC will develop colorectal cancer liver metastases (CRLM) at some point in time. In this study, we aimed to construct a prognostic model of CRLM and its competing endogenous RNA (ceRNA) network.
Methods:
RNA-seq of CRC, CRLM and normal samples were obtained from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus database. Limma was used to obtain differential expression genes (DEGs) between CRLM and CRC from sequencing data and GSE22834, and Gene Ontology and Kyoto Encyclopedia of Genes and Genomes functional enrichment analyses were performed, respectively. Univariate Cox regression analysis and lasso Cox regression models were performed to screen prognostic gene features and construct prognostic models. Functional enrichment, estimation of stromal and immune cells in malignant tumor tissues using expression data (ESTIMATE) algorithm, single-sample gene set enrichment analysis, and ceRNA network construction were applied to explore potential mechanisms.
Results:
An 8-gene prognostic model was constructed by screening 112 DEGs from TCGA and GSE22834. CRC patients in the TCGA and GSE29621 cohorts were stratified into either a high-risk group or a low-risk group. Patients with CRC in the high-risk group had a significantly poorer prognosis compared to in the low-risk group. The risk score was identified as an independent predictor of prognosis. Functional analysis revealed that the risk score was closly correlated with various immune cells and immune-related signaling pathways. And a prognostic gene-associated ceRNA network was constructed that obtained 3 prognosis gene, 14 microRNAs (miRNAs) and 7 long noncoding RNAs (lncRNAs).
Conclusions:
In conclusion, a prognostic model for CRLM identification was proposed, which could independently identify high-risk patients with low survival, suggesting a relationship between local immune status and prognosis of CRLM. Moreover, the key prognostic genes-related ceRNA network were established for the CRC investigation. Based on the differentially expressed genes between CRLM and CRC, the prognosis model of CRC patients was constructed.
