Identification of Prognosis-Related Oxidative Stress Model with Immunosuppression in HCC.

For hepatocellular carcinoma (HCC) patients, we attempted to establish a new oxidative stress (OS)-related prognostic model for predicting prognosis, exploring immune microenvironment, and predicting the immunotherapy response. Significantly differently expressed oxidative stress-related genes (DEOSGs) between normal and HCC samples from the Cancer Genome Atlas (TCGA) were screened, and then based on weighted gene coexpression network analysis (WGCNA), HCC-related hub genes were discovered. Based on the least absolute shrinkage and selection operator (LASSO) and cox regression analysis, a prognostic model was developed. We validated the prognostic model's predictive power using an external validation cohort: the International Cancer Genome Consortium (ICGC).Then a nomogram was determined. Furthermore, we also examined the relationship of the risk model and clinical characteristics as well as immune microenvironment. 434 DEOSGs, comprising 62 downregulated and 372 upregulated genes ( < 0.05 and |log2FC| ≥ 1), and 257 HCC-related hub genes were recognized in HCC. Afterward, we built a five-DEOSG (, , , , and ) prognostic risk model. Using the nomogram, the risk model was shown to have good prognostic value. Compared to the low risk group, HCC patients with high risk had poorer outcomes, worse pathological grades, and advanced tumor stages ( < 0.05). There were significant increases in , , , and expression in HCC samples, while expression was decreased. Finally, Real-time PCR (RT-qPCR) confirmed the mRNA expressions of five genes (, , , , ) in HCC cell lines. Our study constructed a prognostic OS-related model with strong predictive power and potential as an immunosuppressive biomarker for HCC leading to improving prediction and providing new insights for HCC immunotherapy.