FLASH Radiotherapy and Organelle-Targeted Radiosensitization in Glioblastoma: A Conceptual and Translational Review

Radiotherapy remains a central component of standard treatment for glioblastoma (GBM), yet recurrence is common because GBM radioresistance is reinforced by enhanced DNA damage repair, glioma stem cells (GSCs), hypoxia, extracellular matrix remodeling, and an immunosuppressive tumor microenvironment. FLASH radiotherapy (FLASH-RT), delivered at ultra-high dose rates, has shown reproducible normal-tissue-sparing effects in preclinical models, including the brain. In GBM models, however, available evidence indicates that FLASH-RT generally preserves tumor control at levels comparable to conventional radiotherapy rather than providing clearly superior eradication of hypoxic or stem-like tumor compartments. In parallel, endoplasmic reticulum (ER)-targeted interventions have emerged as a candidate strategy for disturbing tumor proteostasis, modulating unfolded protein response (UPR) signaling, impairing synthesis of repair-associated proteins, and promoting immunogenic cell death. This narrative review summarizes representative mechanisms of GBM radioresistance, appraises the opportunities and limitations of FLASH-RT in intracranial disease, and explains why ER targeting is discussed here as a lead but unproven biological axis for radiosensitization. We further compare ER-directed approaches with mitochondrial-, lysosomal-, and delivery-enabled radiosensitization strategies, and outline the translational variables that would determine clinical testability, including beam modality, blood-brain barrier heterogeneity, pharmacokinetics, treatment sequencing, and biomarker development. In this review, "physical precision" refers primarily to dose-rate-driven ultra-rapid delivery and the possibility of widening the normal-tissue therapeutic window under FLASH conditions, rather than to a universal depth-dose advantage shared by all FLASH platforms. Direct experimental evidence for combining FLASH-RT with ER-targeted therapy in GBM is currently lacking. We therefore present this model as a hypothesis-generating conceptual and translational framework for future preclinical testing rather than as an established therapeutic advance.