Lipid metabolism-related gene signature predicts prognosis and depicts tumor microenvironment immune landscape in gliomas

Abstract

Background: Glioma is the most common primary brain tumor in adults and accounts for more than 70% of brain malignancies. Lipids are crucial components of biological membranes and other structures in cells. Accumulating evidence has supported the role of lipid metabolism in reshaping the tumor immune microenvironment (TME). However, the relationship between the immune TME of glioma and lipid metabolism remain poorly described.

Materials and methods: The RNA-seq data and clinicopathological information of primary glioma patients were downloaded from The Cancer Genome Atlas (TCGA) and Chinese Glioma Genome Atlas (CGGA). An independent RNA-seq dataset from the West China Hospital (WCH) also included in the study. Univariate Cox regression and LASSO Cox regression model was first to determine the prognostic gene signature from lipid metabolism-related genes (LMRGs). Then a risk score named LMRGs-related risk score (LRS) was established and patients were stratified into high and low risk groups according to LRS. The prognostic value of the LRS was further demonstrated by construction of a glioma risk nomogram. ESTIMATE and CIBERSORTx were used to depicted the TME immune landscape. Tumor Immune Dysfunction and Exclusion (TIDE) was utilized to predict the therapeutic response of immune checkpoint blockades (ICB) among glioma patients.

Results: A total of 144 LMRGs were differentially expressed between gliomas and brain tissue. Finally, 11 prognostic LMRGs were included in the construction of LRS. The LRS was demonstrated to be an independent prognostic predictor for glioma patients, and a nomogram consisting of the LRS, IDH mutational status, WHO grade, and radiotherapy showed a C-index of 0.852. LRS values were significantly associated with stromal score, immune score, and ESTIMATE score. CIBERSORTx indicated remarkable differences in the abundance of TME immune cells between patients with high and low LRS risk levels. Based on the results of TIDE algorithm, we speculated that the high-risk group had a greater chance of benefiting from immunotherapy.

Conclusion: The risk model based upon LMRGs could effectively predict prognosis in patients with glioma. Risk score also divided glioma patients into different groups with distinct TME immune characteristics. Immunotherapy is potentially beneficial to glioma patients with certain lipid metabolism profiles.

Keywords: glioma; immune; lipid metabolism; prognosis; tumor microenvironment.