Construction of a Prognostic Model based on CSC-related Genes in Patients with Colorectal Cancer.

Abstract

Colorectal cancer (CRC) is one of the most common and deadly malignancies. Lack of efficient biomarkers for prognosis has limited the improvement of survival outcome in patients with CRC. Numerous studies have demonstrated the important roles of cancer stem cells (CSCs) in both treatment resistance and disease recurrence of CRC. Thus, the current study aims to construct a prognostic model based on expression level of CSC-related genes for precise molecular subtyping of CRC patients with different prognoses, TME infiltration patterns and therapeutic responses. The RNA sequencing data and clinical information were obtained from UCSC Xena database, followed by identification of differential expressed genes, univariate Cox regression, and LASSO regression to identify prognostic CSC-related genes and construct a novel prognostic risk scoring model consisting of 21 CSC-related genes. The patients in high-risk group suffered poor survival outcome (P<0.0001). Moreover, the performance of CSC-related prognostic model was validated in individual GEO datasets including GSE41258 and GSE39582 (P<0.05). Furthermore, patients with high-risk score exhibited lower response rate to immune checkpoint inhibitors as compared to those in low-risk group (17.4% vs. 28.2%), indicating the potential of CSC-related prognostic model to predict the immunotherapy response. Collectively, our findings provide an effective model to predict the immunotherapy response and survival outcome in patients with CRC.