Metabolomic and drug profiling of patient-derived glioblastoma models uncovers targetable vulnerabilities in IDH wild-type tumors

Background IDH (isocitrate dehydrogenase) wild-type glioblastomas exhibit rapid progression and poor prognosis compared to their IDH mutant glioma counterparts, highlighting an urgent need for novel, biology informed therapies. Methods We performed an integrative analysis combining tissue metabolomics and ex vivo drug screening in matched patient-derived tumor cells (PDCs) to uncover IDH-relevant vulnerabilities. Results IDH wild-type glioblastomas showed distinct metabolic reprogramming with upregulation of nucleotide metabolism, central carbon metabolism, fatty acid beta-oxidation, and redox mechanisms, supporting enhanced proliferation and tumor plasticity. Strikingly, microbiota-derived metabolites (p-cresyl sulphate and indoxyl sulphate) were enriched in tumor tissue over plasma, suggesting impaired blood-brain barrier integrity in IDH wild-type glioblastomas. A machine learning classifier trained on metabolic profiles accurately distinguished IDH status with 100% accuracy. Drug response profiles across 66 compounds revealed significant higher efficacy (p < 0.05) of HDAC inhibitors (panobinostat, vorinostat), MDM2 inhibitors (AMG232, R7112) and nuclear export/mTOR inhibitors (selinexor, temsirolimus) in IDH wild-type PDCs. Importantly drug responses aligned with metabolic signatures, indicating that metabolic context shapes therapeutic susceptibility. Conclusion Our multi omics approach reveals fundamental metabolic disfunctions between IDH wild-type and mutant gliomas, and links these to differential drug responses. These findings underscore the potential of metabolism informed precision oncology strategies for IDH wild-type glioblastomas.