Abstract BACKGROUND Meningiomas represent the most common primary intracranial tumors, accounting for approximately one-third of adult CNS tumors. Despite their frequency, model systems that accurately replicate the complex tumor-microenvironment interactions across meningioma grades remain underdeveloped, limiting translational progress. This study aimed to establish and characterize human meningioma tumor models to interrogate grade-specific cellular behaviors and microenvironmental dynamics MATERIAL AND METHODS Patient-derived meningioma cell lines were established through mechanical and enzymatic dissociation of surgically resected tumors, followed by expansion in specialized culture media. GFP transfection was performed to enable live-cell tracking. For ex vivo modeling, human neocortical slices — obtained from access tissue from non-meningioma surgeries, either 2 cm distal to the glioma core or from epilepsy surgeries under neuronavigation guidance — were injected with GFP-labeled meningioma cells. Tumor progression, cluster formation, and tissue integration were assessed over six days using longitudinal live imaging and immunofluorescence staining. Single-cell RNA sequencing (scRNA-seq) was employed to resolve cellular heterogeneity and tumor-microenvironmental interactions. Machine learning-based image segmentation combined with manual curation enabled precise quantification of tumor cell distribution and interaction patterns within the host tissue. RESULTS Grade 3 meningioma cells demonstrated significantly enhanced cluster formation, tissue invasion, and microenvironmental remodeling compared to lower-grade counterparts. Ex vivo models recapitulated patient tumor behaviors, with scRNA-seq revealing upregulation of genes associated with mesenchymal transition, extracellular matrix remodeling, and immune modulation in Grade 3 tumors. Microenvironmental analysis indicated that higher-grade tumors established more extensive cell-cell interaction networks and induced greater astrocytic and microglial activation relative to Grade 1 and 2 models. Furthermore, immune profiling revealed a more immunosuppressive environment, with myeloid-derived suppressor cells, reducing immune surveillance and facilitating tumor evasion. Additionally, there was increased activation of astrocytes and microglia, contributing to a supportive tumor microenvironment. CONCLUSION Our patient-derived meningioma models reveal grade-specific differences in tumor behavior and TME interactions. Grade 3 meningiomas show aggressive clustering and microenvironmental remodeling, offering insights into malignant progression. These models provide a high-fidelity platform for studying tumor biology and preclinical evaluation of therapies targeting tumor-stroma interactions
Mohammed et al. (Wed,) studied this question.