Construction and Validation of Prognostic Signature Model Based on Metastatic Features for Colorectal Cancer

Abstract Background Colorectal cancer (CRC) is a common malignant cancer with a poor prognosis. Liver metastasis is the dominant cause of death in CRC patients, and it often involves changes in various gene expression profiling. This study proposed to construct and validate a risk model based on differentially expressed genes between primary and liver metastatic tumors from CRC for prognostic prediction. Methods Transcriptomic and clinical data of CRC were downloaded from The Cancer Genome Atlas database (TCGA) and Gene Expression Omnibus database (GEO). Identification and screening of candidate differentially expressed genes (DEGs) between liver metastatic tissues and corresponding primary tumors were conducted by R package “limma” and univariate Cox analysis in the GSE50760 and TCGA cohort. Last, absolute shrinkage and selection operator (LASSO) Cox regression was carried out to shrink DEGs and develop the risk model. CRC patients from the GSE161158 cohort were utilized for validation. Functional enrichment, CIBERSORT algorithm, and ESTIMATE algorithm for further analysis. Results An 8-gene signature risk model, including HPD, C8G, CDO1, FGL1, SLC2A2, ALDOB, SPINK4, and ITLN1, was developed and classified the CRC patients from TCGA and GEO cohorts into high and low-risk groups. The high-risk group has a worse prognosis compared with the low-risk group. The model was verified as an independent indicator for prognosis. Moreover, tumor immune infiltration analyses demonstrated that monocytes (P = 0.006), macrophage M0 (P < 0.001), and macrophage M1 (P < 0.001) were enriched in the high-risk group, while plasma cells (P = 0.010), T cells CD4 memory resting (P < 0.001) and dendritic cells activated (P = 0.006) were increased in the low-risk group. Conclusions We developed and validated a risk predictive model for the DEGs between liver metastases and primary tumor of CRC, which can be utilized for the clinical prognostic indicator in CRC.