Background: The incidence of lung squamous cell carcinoma (LUSC) increased substantially in recent years. Systematical investigation of the immunogenomic pattern is critical to improve the prognosis of LUSC.
Methods: Based on the TCGA and GEO dataset, we integrated the immune-related genes (IRGs) expression profile and the overall survival (OS) of 502 patients with LUSC. The survival-related and differentially-expressed IRGs in LUSC patients were evaluated by univariate cox regression and LASSO regression analysis. By applying multivariate cox analysis, a new prognostic indicator based on IRGs was established. We also used CIBERSORT algorithms and TIMER database to analyze immune infiltration of LUSC. Both gene set enrichment analysis (GSEA) and principal component analysis (PCA) was carried out for functional annotation. With the assist of computational biology, we also investigated the latent properties and molecular mechanisms of these LUSC-specific IRGs. We analyzed the correlation between immune checkpoints and risk score.
Results: A novel prognostic model was established based on 11 IRGS, including CXCL5, MMP12, PLAU, ELN, JUN, RNASE7, JAG1, SPP1, AGTR2, FGFR4, and TNFRSF18. This model performed well in the prognostic forecast, and was also related to the infiltration of immune cells. Besides, the high-risk groups and the low-risk groups exhibited distinct layout modes in PCA analysis, and GSEA results showed that different immune status among these groups.
Conclusions: In summary, our researches screened out clinically significant IRGs and proved the significance of IRG-based, individualized immune-related biomarkers in monitoring, prognosis, and discern of LUSC.
Keywords: Bioinformatics; Immunogenomic landscape; Lung squamous cell carcinoma; Prognostic index.
© The Author(s) 2020.