The value of metabolic LncRNAs in predicting prognosis and immunotherapy efficacy of gastric cancer

INTRODUCTION: As a unique feature of malignant tumors, abnormal metabolism can regulate the immune microenvironment of tumors. However, the role of metabolic lncRNAs in predicting the prognosis and immunotherapy of gastric cancer (GC) has not been explored. METHODS: We downloaded the metabolism-related genes from the GSEA website and identified the metabolic lncRNAs. Co-expression analysis and Lasso Cox regression analysis were utilized to construct the risk model. To value the reliability and sensitivity of the model, Kaplan–Meier analysis and receiver operating characteristic curves were applied. The immune checkpoints, immune cell infiltration and tumor mutation burden of low- and high-risk groups were compared. Tumor Immune Dysfunction and Exclusion (TIDE) score was conducted to evaluate the response of GC patients to immunotherapy. RESULTS: Twenty-three metabolic lncRNAs related to the prognosis of GC were obtained. Three cluster patterns based on metabolic lncRNAs could distinguish GC patients with different overall survival time (OS) effectively (p<0.05). The risk score model established by seven metabolic lncRNAs was verified as an independent prognostic indicator for predicting the OS of GC. The AUC value of the risk model was higher than TNM staging. The high-risk patients were accompanied by significantly increased expression of immune checkpoint molecules (including PD-1, PD-L1 and CTLA4) and increased tumor tolerant immune cells, but significantly decreased tumor mutation burden (TMB). Consistently, TIDE values of low-risk patients were significantly lower than that of high-risk patients. DISCUSSION: The metabolic lncRNAs risk model can reliably and independently predict the prognosis of GC. The feature that simultaneously map the immune status of tumor microenvironment and TMB gives risk model great potential to serve as an indicator of immunotherapy.