Background: Colon adenocarcinoma (COAD) is a common gastrointestinal tumor and often occurs in the left colon with a poor prognosis. The progression of COAD is closely related to the tumor microenvironment, especially the hypoxia. Currently, few studies have reported the correlation between hypoxia-related genes and the prognosis of COAD patients. Furthermore, we constructed a prognostic model using four hypoxia-related genes to predict the prognosis of COAD patients.
Methods: The mRNA expression profiles and corresponding clinicopathological data were downloaded from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO). The string online analysis tool was used to construct a protein–protein interaction network (PPI) of hypoxia-related genes. Kaplan–Meier curve was used to analyze the relationship of hypoxia risk score and the overall survival of COAD patients, and the receiver operating characteristic (ROC) curve was used to assess the reliability.
Results: We screened out four hypoxia genes, including TKTL1 (transketolase like 1), SLC2A3 (solute carrier family 2 member 3), ALDOB (aldolase, fructose-bisphosphate B) and ENO3 (enolase 3), which were used to construct a hypoxia risk model to predict the overall survival of COAD patients. Besides, we also found that the hypoxia risk score was correlated with the immunosuppression of tumor microenvironment.
Conclusion: The model we constructed with four survival-related hypoxia genes, including TKTL1, SLC2A3, ALDOB and ENO3, could be used to predict the overall survival of COAD patients with high stability.
Keywords: bioinformatic analysis, hypoxia risk model, tumor microenvironment, colon adenocarcinoma, prognosis