Abstract
BACKGROUND: Medulloblastoma (MB) is one of the most common malignant embryonal brain tumor types, subdivided into four major molecular groups: WNT, SHH, Group 3 (G3), and Group 4 (G4). Current therapy includes resection, chemotherapy, and cranio-spinal irradiation. MB prognosis is amongst others dependent on molecular group, subtype, and molecular alterations. High-risk MB subgroups, including TP53-mutant SHH, MYC-amplified G3 and relapsed MB, all require more effective treatments. Therefore, we aim to generate more advanced preclinical models, recapitulating these very high-risk MB subtypes, to identify novel therapeutic strategies. MATERIAL AND METHODS: Long-term in vitro MB tumoroid models have been established and molecularly characterized from patient- and PDX-derived tumor material. High-throughput drug screening with ~225 compounds was used to identify (MB type-specific) drug sensitivities and synergistic combinations. RESULTS: Thus far, we established 19 MB tumoroid models from SHH TP53-mutant (n=6), G3 MYC-amplified (n=8), and G3 non MYC-amplified (n=4) tumors, including five relapse models. DNA methylation and transcriptome analyses demonstrates that these tumoroids retain their parental tumor identity after long-term culture (>5 passages). Analyses of high-throughput screens show MB-wide and molecular group-specific drug sensitivities. Notably, inhibitors of BCL2, CDK7, CDK12, HDAC, and CHECK1 demonstrate efficacy in all tumors, while MCL1 inhibitors appear to be more effective in TP53-mutant SHH compared to G3 MB. Combinatory screens in TP53-mutant SHH MB reveal synergistic drug combinations, including MCL1 with BCL2 inhibitors and CDK12/13 with MTOR inhibitors. CONCLUSION: We have established a large cohort of high-risk MB tumoroid models from patient and PDX tumors that reflect the parental tumor. These in vitro models are highly relevant and provide the basis for future preclinical studies, accelerating the discovery of new treatment modalities in MB. Our current data reveal novel drug combinations for high-risk TP53-mutant SHH MB that we are further validating in vitro before testing in vivo using PDX models.