Potential Prognostic Markers for Glioblastoma Associated with the Glioma Immune Microenvironment.

Glioblastoma (GBM) is a highly aggressive and lethal brain tumor with very poor prognosis despite recent progress in multimodal treatments. Within glioma tissue, various niche cells such as macrophages and neutrophils form a unique glioma immune microenvironment (GIME) by interacting with heterogenous cancer cells, and this has been implicated in disease progression and responsiveness to immunomodulatory therapies. This study explores novel potential prognostic markers associated with the GIME using integrated bioinformatics analyses, including single-cell RNA-sequencing (scRNA-seq), and spatial transcriptome (ST) datasets of clinical GBM specimens. We first identified 42 genes as being associated with poor prognosis in GBM from 5 different cohorts, GBM vs. nontumor tissue, grade IV vs. grade II gliomas, isocitrate dehydrogenase (IDH)-wild-type vs. IDH-mutant variants, mesenchymal vs. proneural and classical subtypes, and hazard ratio for overall survival. Among these, 32 genes were positively correlated with ESTIMATEScore, infiltration of various immune cell types, expression of known immune-related genes, and representative immune-associated biological signals. On scRNA-seq analysis, 7 genes were relatively concentrated in tumor-associated macrophages rather than in malignant cells. ST analysis revealed that Collagen beta(1-O)galactosyltransferase 1 (COLGALT1), Integrin subunit beta 2 (ITGB2), and Myosin light chain 12A (MYL12A) were distributed in the interface between the tumor and the peritumoral area, overlapping with the expression of representative immune-related genes. These findings support the potential of COLGALT1, ITGB2 and MYL12A as biomarkers for predicting the prognosis and immune responses of GBM, which can help in the development of potential immunotherapeutic strategies for GBM.