Abstract
BACKGROUND: Glioma is a common invasive tumor of the central nervous system, and its pathological features significantly impair the quality of life of patients, with high mortality risk and easy recurrence. For glioma Lysine Oxidase 2(LOXL2), there are few reports in the scholarly literature. Based on the current situation of the insufficiency of current therapies, this study focuses on analyzing the biological function of LOXL2 in the occurrence and development of glioma by bioinformatics technology, and systematically evaluates the potential association between this molecular marker and the prognosis of patients. Through the integration of clinical data and molecular mechanism research, this study aims to provide a new theoretical basis for improving the diagnosis and treatment strategy of glioma. METHODS: Integrated CCGA (exploratory) and TCGA (validation) cohorts. Analyzed LOXL2 expression patterns, GO/KEGG pathways, immune infiltration, single-cell distribution (scRNA-seq), and survival associations. Prognostic models were established via KM survival, COX regression, nomogram, and DCA. RESULTS: LOXL2 overexpression correlated with higher glioma malignancy (P<0.001), particularly in IDH wild-type and 1p/19q non-codeleted subtypes (P<0.001). GO/KEGG revealed LOXL2 involvement in ECM remodeling. Immune analysis showed LOXL2 mediates macrophage-neutrophil immunosuppressive networks. scRNA-seq localized LOXL2 in tumor cells, stroma, and macrophages. High LOXL2 predicted worse overall survival (P<0.001). ROC-AUC for 1/3/5-year survival: CCGA: 0.817/0.897/0.925; TCGA: 0.793/0.776/0.730. Respectively, which proved that LOXL2 could be used as an independent prognostic indicator for glioma. Through the construction of nomogram and DCA model evaluation, the results indicate that LOXL2 has important translational value in the prognosis prediction of glioma. CONCLUSIONS: This study revealed that LOXL2 can be used as a potential biomarker in glioma and is correlated with clinical prognosis. LOXL2 may affect the dynamic balance of the tumor microenvironment by regulating the immune function of macrophages and neutrophils in the extracellular matrix (ECM). The prediction model was established based on CCGA clinical data and COX regression analysis of LOXL2 gene expression data, which provides a theoretical foundation for the development of LOXL2-targeted therapy and the construction of genomic integrated prognostic model.