Abstract 6759: High-throughput drug screening and single-cell network analysis identify rational combination therapies in IDH-mutant glioma

Abstract IDH-mutant gliomas are lethal brain tumors marked by multiple coexisting malignant cell states, likely to elicit heterogeneous drug sensitivities thus limiting the effectiveness of monotherapy. To address this challenge, we integrated multimodal single-cell analysis and functional high-throughput drug screening to identify and pharmacologically target state-specific vulnerabilities, thus supporting rational combination therapy development. Single-nucleus RNA sequencing (snRNA-seq) of 20 treatment-naïve, low-grade IDH-mutant gliomas, followed by Master Regulator (MR) analysis, identified key proteins representing mechanistic determinants of three established malignant states: astrocyte-like (AC), oligodendrocyte-like (OC), and neural progenitor cell-like (NPC). The analysis revealed that the AC state is transcriptionally orthogonal to OC, whereas the OC and NPC share substantial regulatory architecture. OncoMatch analysis identified patient-derived models (SF10417, SU-A03) that optimally recapitulate the MRs of these malignant subpopulations, supporting screening of a 374 oncology-focused compounds (FDA-approved and investigational) followed by transcriptional profile analysis at 24h, using the PLATE-seq technology, resulting in a comprehensive drug perturbational compendium for IDH-mutant gliomas. OncoTreat analysis prioritized compounds based on their ability to invert the aberrant MR activity signatures of each malignant state, revealing highly cell-state-specific drug sensitivities, supporting the design of rational combination strategies targeting tumor plasticity. For example, we found that AC-like cells may first be primed with the mTOR inhibitor temsirolimus to reprogram them toward a more drug-sensitive OC-like state, thereby converting a resistant lineage into one that is more vulnerable, including to the new class of IDH1 inhibitors. A second-line agent such as the HDAC inhibitor romidepsin or the topoisomerase inhibitor irinotecan can then be used, possibly in combination with IDH1 inhibitors, to target the resulting OC-like and pre-existing OC/NPC cells, thus implementing a two-step, sequential treatment strategy. In parallel, barcode-based lineage tracing in SF10417 and SU-A03 is being used to monitor cell-state stability and plasticity under drug treatment, while top candidate agents and combinations are being validated in patient-derived ex vivo glioma slice cultures that preserve the native microenvironment. Single-cell and spatial transcriptomic profiling (10x Genomics Xenium) of treated slices will map the reprogramming and elimination of malignant subpopulations in situ. Together, this framework provides a blueprint for discovering state-specific dependencies in IDH-mutant glioma and for guiding rational, combination-based strategies to overcome intratumoral heterogeneity. Citation Format: Patrick Kerwin, Luca Zanella, Andrea Califano, . High-throughput drug screening and single-cell network analysis identify rational combination therapies in IDH-mutant glioma abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 6759.