Immune checkpoint inhibitors failed to demonstrate durable benefit in three large glioblastoma phase III trialsImmune Checkpoint Inhibitors Show Limited Success for Glioblastoma Patients

Glioblastoma (GBM) remains one of the most lethal primary brain tumors despite maximal surgical resection, radiotherapy, and temozolomide. Immune checkpoint inhibitors have failed to demonstrate durable benefit in three large phase III trials (CheckMate 143, 498, and 548), underscoring profound, multilayered immune resistance. This narrative review synthesizes the immunosuppressive GBM microenvironment, situating the microglia–myeloid-derived suppressor cell (MDSC)–regulatory T cell (Treg) axis within the broader immune landscape, including dendritic cells, natural killer cells, exhausted CD8+ T cells, neutrophils, and B cells. We distinguish ontogenetically distinct resident microglia from bone marrow–derived macrophages, separate monocytic (M-MDSC) from polymorphonuclear (PMN-MDSC) subsets, and examine how radiotherapy reshapes immunity. We extend the immunometabolic discussion beyond indoleamine 2,3-dioxygenase (IDO) to the adenosine (CD39/CD73/A2A), arginine, hypoxia, lactate, and glutamine pathways, and critically analyze why checkpoint blockade has failed. We summarize emerging strategies, including CSF1R, CCR2, CXCR2, CD47–SIRPα, STING, CD40, TGF-β, and IL-1β targeting, together with candidate biomarkers such as circulating MDSCs, CSF1, IL-1β, multiplex immunofluorescence, spatial transcriptomics, and single-cell RNA sequencing, for rational patient selection. We frame these mechanisms as a single, self-reinforcing circuit integrating myeloid-driven immune regulation, metabolic reprogramming, and treatment-induced immune remodeling. We argue that future progress depends on biomarker-driven combinations that reprogram the myeloid compartment, relieve metabolic suppression, and actively inflame the tumor, integrated with radiotherapy and tailored to molecular context (IDH and MGMT status).