The earliest and most extensively studied genetic event in meningioma pathogenesis involves alterations in NF2, located on chromosome 22q (Figure 1 illustrates the genomic driver landscape, distinguishing NF2-mutant and NF2-independent molecular pathways). NF2 mutations or deletions occur in approximately 40% to 60% of sporadic meningiomas and represent the dominant driver alteration in this disease (4, 5). The NF2 gene encodes Merlin, a tumor suppressor protein that regulates contact inhibition and cytoskeletal organization. Loss of Merlin function disrupts these regulatory processes and leads to activation of multiple oncogenic pathways, including Hippo and focal adhesion kinase (FAK) signaling, promoting tumor growth and proliferation.

Clinically, NF2-mutant meningiomas tend to arise along the convexities or parasagittal regions of the brain. They frequently exhibit widespread chromosomal instability, often characterized by monosomy 22 as well as losses in chromosomes 1p, 6q, 10q, 14q, and 18q (6). These copy-number alterations contribute substantially to the aggressive potential of NF2-mutant tumors, with high copy-number burden correlating strongly with recurrence and poor outcomes.

Advances in sequencing technologies have revealed a diverse array of NF2-independent genetic alterations that define biologically and clinically distinct molecular subgroups of meningiomas. Rather than constituting a single homogeneous category, these non-NF2 alterations encompass genes with diverse molecular functions and are associated with preferential anatomical locations, characteristic histologic subtypes, and differing prognostic implications. In general, these mutations are mutually exclusive with NF2 loss and are particularly enriched in skull-base meningiomas, which often display more stable genomic profiles and indolent clinical behavior.

Mutations in TRAF7, an E3 ubiquitin ligase, occur in approximately one-quarter of all meningiomas (5) (Figure 1). TRAF7-mutant tumors frequently co-occur with the KLF4 K409Q mutation, a combination that is nearly pathognomonic for secretory meningioma (5), a distinct histologic subtype characterized by eosinophilic pseudopsammoma bodies. These TRAF7/KLF4-mutant tumors typically arise from the anterior skull base, are most often classified as WHO grade 1, and generally demonstrate a benign clinical course.

Alterations in the PI3K/AKT/mTOR signaling pathway represent another major NF2-independent subgroup. Mutations in AKT1, particularly the E17K variant, result in constitutive pathway activation and are most commonly observed in midline skull-base meningiomas, including those of the planum sphenoidale and olfactory groove (7). Although these tumors are frequently WHO grade 1 by histology, they exhibit distinct radiographic and biological characteristics and have emerged as rational candidates for targeted pathway inhibition. Similarly, PIK3CA mutations occur in approximately 7% of meningiomas and are enriched in non-NF2 skull-base tumors (7). These alterations further support the biological distinction of this subgroup and provide a therapeutic rationale for PI3K or mTOR inhibition in selected patients.

Activating mutations in SMO, a key component of the Hedgehog signaling pathway, are identified in approximately 5% of meningiomas and show a strong predilection for the olfactory groove and anterior midline skull base (7). SMO-mutant tumors form a molecularly defined subgroup with characteristic transcriptional profiles and represent one of the clearest examples of a genotype-directed therapeutic opportunity in meningioma.

Beyond skull-base–predominant alterations, several additional NF2-independent mutations are associated with specific histopathologic phenotypes and prognostic implications. POLR2A mutations define a transcriptionally unique subset of largely benign meningiomas with low recurrence risk (8). SUFU mutations, though rare, implicate upstream dysregulation of Hedgehog signaling and may have implications for pathway-directed therapies (9). In contrast, loss of BAP1 function is strongly associated with rhabdoid meningioma, a high-grade variant characterized by aggressive clinical behavior and poor prognosis (10). Germline or somatic loss of SMARCE1 is characteristic of clear cell meningioma, a subtype frequently occurring in younger patients and associated with a higher risk of recurrence, particularly in spinal locations (11). Finally, mutations in ARID1A highlight the role of SWI/SNF chromatin-remodeling dysfunction in a subset of biologically aggressive and higher-grade meningiomas (12).

Together, these observations underscore that NF2-independent meningiomas represent a heterogeneous collection of molecularly, anatomically, and clinically distinct entities, rather than a single category. Recognition of this heterogeneity allows refinement of meningioma classification beyond traditional histopathology, improves prognostic stratification, and facilitates the development of personalized, genotype-informed therapeutic strategies (Table 1 integrates driver mutations with clinical, anatomical, and prognostic associations, serving as a translational reference.).

One of the most transformative advances in meningioma research has been the incorporation of DNA methylation profiling into tumor classification (Figure 2 depicts DNA methylation-based classification and its correlation with WHO grading and prognosis). Sahm et al. (13) pioneered a methylation-based grading system that stratifies meningiomas into biologically distinct classes with far superior prognostic accuracy compared to WHO histologic grade alone. This approach integrates genome-wide methylation signatures, copy-number patterns, and transcriptional programs to identify tumors with similar biological behaviors, regardless of histologic appearance.

Further refinement occurred through the work of Nassiri et al. (14), who conducted a large multi-institutional analysis combining methylation, transcriptomic, copy-number, and mutational data. Their study identified three major consensus molecular groups (CMGs) (Figure 2): a Merlin-intact group associated with favorable outcomes, an immune-enriched group characterized by inflammatory microenvironments and intermediate prognosis, and a proliferative group marked by high cell-cycle activity and poor clinical outcomes. These CMGs predict progression-free survival with greater accuracy than histologic grade and reveal biological features that may have therapeutic implications, such as immune signatures or proliferative pathway activation.

Methylation profiling increasingly informs real-world clinical decision-making. It aids in predicting recurrence risk following gross total resection, identifies histologically benign-appearing tumors with aggressive molecular features, guides the need for adjuvant radiotherapy, and assists in selecting patients for molecularly tailored clinical trials. For recurrent, atypical, and morphologically ambiguous meningiomas, methylation profiling is rapidly becoming an essential diagnostic complement.

Recent work has further refined the biological understanding of grade 3 meningiomas, highlighting the central role of TERT promoter mutations, CDKN2A/B homozygous deletions, high copy-number burden, and proliferative methylation-defined subgroups in driving aggressive behavior (24). These tumors exhibit marked genomic instability, dysregulated cell-cycle control, and enrichment of DNA repair and proliferation pathways. Recognition of these molecular characteristics supports incorporation of molecular risk factors into treatment algorithms, including consideration of early adjuvant radiotherapy, intensified surveillance schedules, and prioritization for enrollment in molecularly targeted clinical trials.

Liquid biopsy approaches are being explored as adjuncts to tissue-based diagnostics in meningioma. Circulating tumor DNA (ctDNA) and tumor-specific methylation signatures detectable in plasma offer a non-invasive means of tumor classification and disease monitoring (16). Although sensitivity remains limited in low-volume disease, ongoing improvements in assay sensitivity raise the possibility that blood-based molecular profiling may assist in preoperative risk stratification, particularly in tumors located in surgically challenging regions. Integration of plasma-derived molecular data with imaging and clinical parameters may eventually help guide the extent of surgical resection, inform adjuvant therapy decisions, and tailor postoperative surveillance strategies (17).

Emerging artificial intelligence (AI)–driven modeling approaches are increasingly being integrated with molecular profiling in meningioma research. Machine learning frameworks combining radiomic features, genomic alterations, DNA methylation patterns, and clinical variables have demonstrated potential for improving risk stratification and recurrence prediction. Integrative AI platforms can identify imaging–molecular correlations that are not readily apparent through conventional analysis, enabling non-invasive inference of tumor biology and facilitating longitudinal disease tracking (25). As these tools mature, AI-assisted systems may support personalized surveillance strategies and treatment planning by dynamically integrating molecular, radiologic, and clinical data.

A key limitation in advancing meningioma molecular research is fragmentation of datasets across institutions. Establishment of centralized, multi-institutional repositories integrating genomic, epigenetic, radiologic, and clinical outcome data would substantially accelerate validation of molecular classifiers and therapeutic targets. Such initiatives would require standardized biospecimen processing protocols, harmonized bioinformatics pipelines, secure data-sharing infrastructure, and robust ethical governance frameworks. Centralized repositories could facilitate large-scale validation studies, improve reproducibility, and enable AI-driven predictive modeling, thereby enhancing the translational impact of molecular profiling.