P05.08.B EXTRACELLULAR VESICLE-DERIVED MICRORNA SIGNATURES AS POTENTIAL BIOMARKERS IN PEDIATRIC BRAIN TUMORS

Abstract BACKGROUND Medulloblastoma (MB), a malignant embryonal brain tumor arising in the cerebellum, accounts for nearly 20% of pediatric brain tumors and is associated with substantial mortality in childhood. Current genomic analyses permit risk stratification by prognostically relevant consensus molecular subgroups: Wingless (WNT), Sonic Hedgehog (SHH), group 3, and group 4. Conventional response monitoring by imaging remains challenging, and a marker for measurable residual disease (MRD) is lacking. Developing robust, minimally invasive biomarkers with clinically actionable information is necessary to improve outcomes of medulloblastoma patients. Liquid biopsy techniques employing extracellular vesicles (EVs) have emerged as promising diagnostic and monitoring tools, providing deeper molecular insights into tumor biology. This study analyzed microRNA (miRNA) expression signatures as potential EV-derived biomarkers in pediatric brain tumors. MATERIAL AND METHODS A total of 23 samples, including 14 MB (3 SHH, 9 G3, 2 G4), 2 atypical teratoid rhabdoid tumors (ATRT), 2 ependymomas (EPN), 1 BCOR-driven tumor (BCOR), 1 embryonal tumor with multilayered rosettes (ETMR), and 1 non-malignant control, were analyzed. EVs were isolated from primary cell line supernatants, pellets, and CSF using ultracentrifugation combined with size exclusion chromatography. EV characterization was performed through nanoparticle tracking analysis, electron microscopy, and Western blotting. RESULTS Small RNA sequencing revealed a clear overlap of microRNA expression profiles (47 microRNAs) between primary cell line pellets and EVs from corresponding cell line supernatants and CSF, suggesting that EV-based microRNA profiling can robustly reflect tumor biology. Furthermore, differential expression analysis revealed an overlap in deregulation of microRNAs such as miR-183-5p, miR-769-5p, miR-96-5p, miR-181b-5p, and miR-10a-5p in MB G3 compared to controls in the previously specified sample types. Of particular note is the consistent observation of upregulation of the miR-182-5p-96-183-5p cluster in all studied sample types. Furthermore, a clear distinction was observed between MB and ETMR EV signatures, underscoring their diagnostic utility. CONCLUSION Our findings illustrate the potential of EV-derived miRNA signatures as a diagnostic and monitoring tool in pediatric medulloblastomas, offering a non-invasive molecular approach to improve clinical management.