Precision medicine has transformed cancer treatment by tailoring therapies to the specific molecular aberrations of an individual patient’s tumor. The integration of high-resolution multi-omics with high-throughput functional profiling holds the potential to further refine patient stratification, advancing personalized medicine and improving treatment outcomes. While (epi)genetic profiling has drastically improved classification and biomarker discovery for molecular diagnostics in diffuse adult gliomas, these advances have yet to translate into effective therapeutic strategies. In high-grade gliomas, precision medicine approaches remain to be fully established.

We investigated over 45 patient-derived tumor organoid and orthotopic xenograft models of high-grade gliomas. Omics profiling included genomics (targeted DNA-seq), epigenomics (DNA methylation arrays) and transcriptomics (bulk RNA-seq). Functional screens were conducted on 27 organoid models using personalized 203-compound libraries targeting cancer pathways and epigenetic modifiers, and on 16 models with 1,280 FDA-approved drugs. Unsupervised multi-omics factor analyses were employed to identify patient-specific therapeutic vulnerabilities. Further validation involved time- and dose-dependent drug efficacy assessments, as well as biomarker validation in patient tumor tissues across molecular subgroups.

Multi-omics analysis revealed a spectrum of molecular profiles encompassing genetic, epigenetic, and transcriptomic subtypes of high-grade gliomas. Multi-omics factor analysis (MOFA), integrating multi-omics and drug response profiling, identified distinct subgroups with differential drug responses associated with IDH1 mutation status and pediatric-like DNA methylation signatures. IDH1 mutant high-grade astrocytomas were sensitive to histone deacetylase inhibitors, whereas IDH wild-type MYC-amplified glioblastoma was sensitive to histone methyltransferase inhibitors. The observed differential efficacy of these drugs to patient subgroups were linked to specific (epi)genetic and transcriptomic biomarkers. Drug responses were validated in time- and dose-dependent fashion. While IDH wild-type glioblastomas exhibited heterogeneous treatment responses, no robust biomarker-defined responder subgroups were identified.

Our findings underscore the potential of integrating multi-omics with functional profiling to define precision medicine strategies. This approach enables the identification of distinct patient subgroups in preclinical models, paving the way for tailored therapeutic interventions in high-grade gliomas. While IDH1-mutant and pediatric-like gliomas exhibit distinct drug response profiles, implementing precision medicine in glioblastomas remains a significant challenge.

LIH, FNR GLASS-LUX/DIOMEDES/MITOFIT, TRANSCAN3 PLASTIG