Meningiomas represent the most prevalent primary intracranial neoplasms encountered in adult neurosurgical and neuro-oncological practice. Although clinicians traditionally categorize these lesions using histopathological grading systems established by the World Health Organization, significant biological and clinical heterogeneity persists within each grade. Consequently, histological features alone frequently fail to predict clinical trajectories, leading to unexpected recurrences in benign tumors or prolonged dormancy in higher-grade lesions. To overcome these diagnostic hurdles, researchers developed a machine learning-based molecular classifier of meningiomas utilizing genome-wide DNA methylation data. This novel diagnostic platform delivers reproducible, objective subtype assignments that substantially enhance risk stratification and guide patient-tailored therapeutic interventions.

Overcoming the Limitations of Histopathological Grading

Historically, neuro-oncologists have relied on cellular architecture, mitotic counts, and brain invasion to grade meningiomas. However, these morphological criteria exhibit considerable interobserver variability among pathologists. Furthermore, standard microscopy fails to capture the intricate epigenetic and transcriptomic landscapes driving tumor recurrence. Although multi-omic discovery studies previously delineated four distinct molecular subgroups—immunogenic, NF2-wildtype, hypermetabolic, and proliferative—the stochastic nature of multi-omic clustering algorithms restricted their routine prospective deployment. Clinicians require a single-platform, standardized assay capable of reliably classifying individual patient tumors. Therefore, developing a robust epigenetic classifier bridges a vital gap between translational discovery and routine neuropathological assessment, ultimately providing oncologists with actionable prognostic insights.

Development of the Molecular Classifier of Meningiomas

To establish a scalable diagnostic tool, investigators constructed a dedicated machine learning algorithm trained solely on DNA methylation arrays. By analyzing an extensive international multi-institutional cohort of 1,698 meningioma cases, the researchers validated the performance of this dedicated molecular classifier of meningiomas across independent patient series. DNA methylation serves as an exceptionally stable epigenetic signature that reliably reflects cellular lineage and oncogenic pathway activation. Importantly, the newly developed machine learning model extracts critical CpG methylation features to assign each tumor into one of the four established biological groups without requiring complex multi-omics sequencing pipelines. As a result, this single-modality testing strategy drastically simplifies laboratory workflows while preserving high diagnostic fidelity.

Biological Concordance Across Multi-Omic Platforms

Crucially, the validation study demonstrated that DNA methylation-derived classifications robustly mirror underlying genomic and transcriptomic phenotypes. Comprehensive multi-platform analyses, including whole-exome sequencing, RNA sequencing, and copy number profiling, corroborated the biological authenticity of predicted subgroups. For instance, tumors classified as NF2-wildtype harbored no NF2 alterations; instead, over half displayed canonical non-NF2 mutations such as AKT1, TRAF7, or SMO. In addition, RNA pathway analyses revealed significant upregulation of immune response genes in the immunogenic cohort, metabolic reprogramming pathways in the hypermetabolic group, and accelerated cell-cycle programs within proliferative tumors. Cellular deconvolution further confirmed dense macrophage infiltration in immunogenic lesions and elevated neoplastic cell purity in aggressive subtypes.

Clinical Outcomes and Progression-Free Survival Stratification

The clinical validity of this epigenetic classification tool is strongly supported by longitudinal outcome data. In the large validation cohort, group-specific progression-free survival aligned almost perfectly with historical multi-omic benchmarks. Patients with hypermetabolic meningiomas exhibited an intermediate clinical course with a median progression-free survival of 7.4 years. In contrast, patients harboring proliferative meningiomas experienced aggressive disease biology, showing a poor median progression-free survival of only 2.5 years. Meanwhile, tumors within the immunogenic and NF2-wildtype categories demonstrated favorable clinical behavior, with median progression-free survival remaining unreached across extended follow-up intervals. Consequently, the tool identifies high-risk individuals who require vigilant surveillance or early adjuvant radiation therapy.

Translating Epigenetic Classification into Neurosurgical Workflows

Integrating machine learning-driven DNA methylation profiling into routine surgical neuro-oncology holds immense therapeutic potential. Because this classifier is publicly accessible for immediate clinical translation, multidisciplinary tumor boards can readily leverage its prognostic output. For example, identifying a proliferative molecular profile in a gross-totally resected grade 1 lesion could prompt early adjuvant radiotherapy rather than passive surveillance. Conversely, confirming an NF2-wildtype or immunogenic signature in an atypical tumor could spare the patient from unnecessary radiation toxicity. Furthermore, characterizing subgroup-specific biology paves the way for targeted clinical trials, such as metabolic inhibitors for hypermetabolic tumors or checkpoint immunotherapy for immunogenic lesions.

Frequently Asked Questions

What is the molecular classifier of meningiomas?

The molecular classifier is a validated machine learning model that uses DNA methylation data to stratify meningiomas into four biologically distinct groups: immunogenic, NF2-wildtype, hypermetabolic, and proliferative. This tool overcomes the limitations of histopathological grading to deliver highly accurate, reproducible predictions of tumor progression and clinical outcomes.

How does the classifier predict progression-free survival in patients?

The classifier stratifies recurrence risk based on tumor biology. In validation studies, proliferative meningiomas exhibited a short median progression-free survival of 2.5 years, while hypermetabolic tumors had a median progression-free survival of 7.4 years. In contrast, immunogenic and NF2-wildtype tumors demonstrated indolent clinical behavior with prolonged disease-free survival.

Why is DNA methylation preferred over multi-omic testing?

DNA methylation arrays provide a highly stable, reproducible, single-platform molecular measurement that captures comprehensive genomic and epigenetic information. Relying exclusively on methylation profiling simplifies laboratory diagnostic workflows, reduces diagnostic costs, avoids algorithm stochasticity, and facilitates rapid, practical translation into routine clinical neuropathology and neurosurgical practice worldwide.

Disclaimer: This content is for informational and educational purposes only and does not constitute medical advice, diagnosis, or treatment. Healthcare professionals should exercise their independent clinical judgment when interpreting research findings. Treatment decisions should always be individualized based on specific patient presentations, institutional protocols, and multidisciplinary consensus. Refer to the latest local and national guidelines for clinical practice.

References

1. Landry AP et al. Development and validation of a molecular classifier of meningiomas. Neuro Oncol. 2025 Jun 21. doi: 10.1093/neuonc/noae242. PMID: 39775867.
2. Nassiri F, Liu J, Patil V, Mamatjan Y, Wang JZ, Hugh-White R, et al. A clinically applicable integrative molecular classification of meningiomas. Nature. 2021;597(7874):119-125.
3. Sahm F, Schrimpf D, Stichel D, Jones DTW, Hielscher T, Schefzyk S, et al. DNA methylation-based classification and grading system for meningioma: a multicentre, retrospective analysis. Lancet Oncol. 2017;18(5):682-694.
4. Wang JZ, Patil V, Landry AP, et al. Molecular classification to refine surgical and radiotherapeutic decision-making in meningioma. Nat Med. 2024;30(11):3173-3183.