Spatiotemporal and metabolic heterogeneity of tumor-associated macrophages in glioblastoma: from single-cell insights to therapeutic targeting

The immunosuppressive and therapy-resistant nature of glioblastoma (GBM) is fundamentally driven by the profound spatiotemporal and metabolic heterogeneity of tumor-associated macrophages (TAMs). This review proposes a spatiotemporal-metabolic axis as an integrative framework to decipher the functional plasticity of TAMs and its therapeutic implications. Drawing on the latest single-cell and spatial multi-omics data, we first delineate the lineage competition landscape of TAMs. Within this landscape, brain-resident microglia, border-associated macrophages (BAMs), and peripherally recruited bone marrow-derived macrophages (BMDMs) engage in dynamic interplay during tumor evolution, culminating in a shifted ecosystem dominated by BMDMs at recurrence. These subsets are not randomly distributed but are spatially organized through niche-instructive signals—such as hypoxia, perivascular cues, and tumor-derived metabolites—leading to context-dependent enrichment: immunosuppressive TAMs accumulate in the tumor core, BAMs localize to perivascular zones and express pro-angiogenic factors, while hypoxic necrotic regions are populated by metabolically reprogrammed HMOX1 + TAMs. Metabolically, TAMs engage in symbiotic nutrient exchange with glioma cells via enhanced glycolysis, amino acid catabolism, and lipid accumulation, collectively reinforcing an immunosuppressive microenvironment. Building on this multidimensional understanding, we highlight emerging therapeutic strategies that move beyond broad depletion: metabolic-epigenetic interference (e.g., targeting lactate-driven histone lactylation), phagocytosis checkpoint blockade (e.g., CD47-SIRPα axis), and niche-precise targeting of hypoxic or perivascular TAM subsets. This review provides an integrative roadmap for developing next-generation immunotherapies that leverage the spatiotemporal and metabolic logic of TAMs to reprogram the GBM microenvironment.