Glioblastoma remains one of the most lethal primary malignant brain tumors, and immunotherapy has produced limited clinical benefit despite major success in other cancers. This resistance reflects spatial and molecular heterogeneity, myeloid-dominant immune suppression, restricted lymphocyte infiltration, antigenic instability, blood–brain barrier-related constraints, corticosteroid exposure, and treatment-induced immune remodeling. Conventional preclinical models only partially reproduce these features, limiting their ability to predict patient-specific immunotherapy responses. Patient-derived glioblastoma organoids (GBOs) have emerged as experimentally tractable platforms that can preserve key features of parental tumors while enabling functional therapeutic testing. In this review, we discuss the rationale, technical evolution, applications, and translational challenges of GBOs in immuno-oncology, including tumor-immune modeling, checkpoint blockade, cellular therapies, myeloid modulation, vaccines, oncolytic virotherapy, multi-omic readouts, and artificial intelligence-assisted analysis. We place particular emphasis on oncolytic herpes simplex virus-based strategies and on the ability of GBOs to evaluate viral entry, replication, spatial spread, innate antiviral restriction, immunogenic cell death, and rational combinations. Finally, we outline staged, clinically integrated workflows while emphasizing immune-cell attrition, incomplete microenvironmental fidelity, resource requirements, and the need for prospective validation in small exploratory trials.
更多