Identification of a Novel Oxidative Stress-Based Molecular Classification and Treatment Vulnerabilities in WHO Grade 2/3 Meningiomas.

Objective: The World Health Organization (WHO) grading based on histopathology cannot always accurately predict tumor behavior of meningiomas. To overcome the limitations of the WHO grading, the study aims to propose a novel oxidative stress-based molecular classification for WHO grade 2/3 meningiomas. Methods: Differentially expressed oxidative stress-related genes were analyzed between 86 WHO grade 1 (low grade) meningiomas and 99 grade 2/3 (high grade) meningiomas. An oxidative stress-based molecular classification was developed in high-grade meningiomas through consensus clustering analysis. Immune microenvironment features, responses to immunotherapy and chemotherapy, and targeted drugs were evaluated. Three machine learning models: logistic regression, support vector machine, and random forest, were built for differentiating the classification. Key oxidative stress-related genes were verified in human meningeal cells (HMC) and two meningioma cells (CH-157MN and IOMM-Lee) via reverse transcription quantitative polymerase chain reaction (RT-qPCR) and western blot. After knockdown of Forkhead Box M1 (FOXM1) or Prion Protein (PRNP), cell growth, migration, and reactive oxygen species (ROS) levels were measured through cell counting kit-8 (CCK-8), transwell, and immunofluorescence, respectively. Results: We classified high-grade meningiomas into two oxidative stress-based clusters, termed cluster 1 and cluster 2. Cluster 1 exhibited higher infiltrations of immune and stromal cells and higher expression of classic immune checkpoints: Cluster of Differentiation 86 (CD86), Programmed Cell Death 1 (PDCD1), and Leukocyte-Associated Immunoglobulin-Like Receptor 1 (LAIR1), indicating that cluster 1 meningiomas might respond to immunotherapy. Drug sensitivity was heterogeneous between the two clusters. Three classifiers were established, which could accurately differentiate this molecular classification. FOXM1 and PRNP were experimentally evidenced to be highly expressed in meningioma cells, and their knockdown hindered cell growth and migration and triggered ROS accumulation. Conclusion: In summary, our findings established a novel oxidative stress-based molecular classification and identified potential treatment vulnerabilities in high-grade meningiomas, which might assist personalized clinical management.