Epco-20. Development of a Network-Derived Predictive Biomarker for Prmt5 Inhibitor Response in Idh1-Mutant High-Grade Glioma Patients

Despite advances in disease classification, the therapeutic landscape of High-grade gliomas (HGG) remains limited, with a critical need for new treatments that complement or enhance existing therapies. A deeper understanding of the molecular underpinnings of HGG is key to identifying patients that may exhibit sensitivities to novel therapies. We assembled a cohort of 450 IDH1-mutant HGG (IDHmHGG) patients, integrating DNA and RNA sequencing with clinical data and constructed a probabilistic network model that represents how pathways interact to influence disease progression and survival. We identified and prioritized subnetworks of interconnected genes strongly linked to patient outcomes. One of our top prioritized subnetworks revealed key processes tied to IDHmHGGs, including cell-cycle, DNA repair, and epigenetic regulation. Our models predicted PRMT5 as regulating this subnetwork, highlighting it as a therapeutic target. Using public data from PRMT5 inhibitor-treated cells, we validated that genes within this subnetwork were significantly downregulated after treatment. Patients with high subnetwork expression had worse survival, suggesting PRMT5 inhibition could reverse poor outcomes. We developed a complex biomarker capturing subnetwork biological activity to classify patients and predict likelihood of benefiting from PRMT5-targeted therapy. Based on these findings, we licensed P-500, a brain-penetrant PRMT5 inhibitor. In a Phase I trial (NCT04089449), with an investigator-reported 19% objective response rate in IDHmHGG patients and 0% in IDH-wildtype HGG. Using baseline RNAseq, we generated network-derived biomarker scores for each patient. Patients with responses exhibited the highest biomarker scores, providing direct support for the potential of this complex biomarker to predict response to treatment with a PRMT5 inhibitor. These results support the hypothesis that targeting PRMT5 may provide enhanced therapeutic benefit in IDHmHGG patients in which the PRMT5-regulated subnetwork is activated. We propose a clinical trial for P-500 in IDHmHGG to evaluate our predictive biomarker as a precision-medicine approach for clinical development and therapeutic decision-making.