Identification of molecular subtypes and a risk model based on inflammation-related genes in patients with low grade glioma

Abstract

Lower grade gliomas (LGGs) exhibit invasiveness and heterogeneity as distinguishing features. The outcome of patients with LGG differs greatly. Recently, more and more studies have suggested that infiltrating inflammation cells and inflammation-related genes (IRGs) play an essential role in tumorigenesis, prognosis, and treatment responses. Nevertheless, the role of IRGs in LGG remains unclear. In The Cancer Genome Atlas (TCGA) cohort, we conducted a thorough examination of the predictive significance of IRGs and identified 245 IRGs that correlated with the clinical prognosis of individuals diagnosed with LGG. Based on unsupervised cluster analysis, we identified two inflammation-associated molecular clusters, which presented different tumor immune microenvironments, tumorigenesis scores, and tumor stemness indices. Furthermore, a prognostic risk model including ten prognostic IRGs (ADRB2, CD274, CXCL12, IL12B, NFE2L2, PRF1, SFTPC, TBX21, TNFRSF11B, and TTR) was constructed. The survival analysis indicated that the IRGs risk model independently predicted the prognosis of patients with LGG, which was validated in an independent LGG cohort. Moreover, the risk model significantly correlated with the infiltrative level of immune cells, tumor mutation burden, expression of HLA and immune checkpoint genes, tumorigenesis scores, and tumor stemness indices in LGG. Additionally, we found that our risk model could predict the chemotherapy response of some drugs in patients with LGG. This study may enhance the advancement of personalized therapy and improve outcomes of LGG.