DataSheet1_Molecular Patterns Based on Immunogenomic Signatures Stratify the Prognosis of Colon Cancer.CSV (32.39 kB)

DataSheet1_Molecular Patterns Based on Immunogenomic Signatures Stratify the Prognosis of Colon Cancer.CSV

Download (32.39 kB)
posted on 2022-02-14, 05:05 authored by Cong Shen, Cong Luo, Zhijie Xu, Qiuju Liang, Yuan Cai, Bi Peng, Yuanliang Yan, Fada Xia

Background: Colon cancer is an aggressive and heterogeneous disease associated with high morbidity and mortality. The immune system is intimately involved in tumorigenesis and can influence malignant properties at the protein, epigenetic, and even genomic levels by shaping the tumor immune microenvironment (TIM). However, immune-related molecules that can effectively predict the prognosis of colon cancer remain under exploration.

Methods: A total of 606 patients from TCGA and GEO databases were employed in our study, in which 429 cases were set as the training cohort and 177 were defined as the validation cohort. The immune infiltration was evaluated by ESTIMATE, TIMER, and CIBERSORT algorithms. The risk signature was constructed by LASSO Cox regression analysis. A nomogram model was generated subsequent to the multivariate Cox proportional hazards analysis to predict 1-, 3-, and 5-year survival of patients with colon cancer.

Results: Infiltrating immune cell profiling identified two colon cancer clusters (Immunity_L group and Immunity_H group). The abundances of immune cells were higher in the Immunity_H group, which indicated a better prognosis. Through further statistical analysis, we identified four genes which were highly correlated with prognosis and representative of this gene set, namely ARL4C, SERPINE1, BST2, and AXIN2. When the patients were divided into low- and high-risk groups based on their risk scores, we found that patients in the high-risk group had shorter overall survival time. Moreover, a nomogram including clinicopathologic features and the established risk signature could robustly predict 1-, 3-, and 5-year survival in patients with colon cancer.

Conclusion: We identified two distinct immune patterns by analyzing clinical and transcriptomic information from colon cancer patients. A subsequently constructed immune-related gene-based prognostic model as well as a nomogram model can be used to predict the prognosis of colon cancer, thereby guiding risk stratification and treatment regimen development for colon patients.