RESUMO
Background: Patients with low-grade glioma (LGG) have wildly varying average lifespans. However, no effective way exists for identifying LGG patients at high risk. Cuproptosis is a recently described form of cell death associated with the abnormal aggregation of lipid acylated proteins. Few investigations have been conducted on cuproptosis-associated genes and LGG thus far. The purpose of this research is to establish a predictive model for cuproptosis-related genes in order to recognise LGG populations at high risk. Methods: We analyzed 926 LGGs from 2 public datasets, all of which were RNA sequencing datasets. On the basis of immune scores, the LGG population was split into different risk categories with X-tile. LASSO and Cox regressions were employed to filter cuproptosis-associated genes and construct prediction models. The accuracy of the predictive models was measured by using TCGA internal validation set and the CGGA external validation set. In addition, LGG immune cell infiltration was viewed using CIBERSORT and ssGSEA algorithms and correlation analysis was done with cuproptosis-related genes. Finally, immune escape capacity in LGG low- and high-risk groups was evaluated using the TIDE method. Results: The prediction model constructed by four cuproptosis-related genes was used to identify high-risk populations in LGG. It performed well in training and all validation sets (AUC values: 0.915, 0.894, and 0.774). Meanwhile, we found that FDX1 and ATP7A in the four cuproptosis-related genes were positively correlated with immune response, while GCSH and ATP7B were opposite. In addition, the high immune score group had a lower TIDE score, indicating that their immune escape capacity was weak. Conclusion: High-risk individuals in LGG can be reliably identified by the model based on cuproptosis-related genes. Furthermore, cuproptosis is closely related to tumor immune microenvironment, which gives a novel approach to treating LGG.
