Meningioma Genomics: Gene Discovery, Molecular Mechanisms, And Clinical Correlations

Meningiomas, the most common primary intracranial tumors, can cause significant morbidity and mortality, requiring novel targeted therapies. The genomic basis of approximately 80% of sporadic meningiomas has recently been established, however, the rest remain mutation-unknown. Identification of additional driver and/or co-driver genes could guide future targeted therapies. Known meningioma driver genes can be categorized into six mutually exclusive genomic subgroups: (1) NF2 with/without chromosome 22 loss, (2) SMARCB1 with/without NF2, (3) PI3K pathway (AKT1, PIK3CA, and PIK3R1) with/without TRAF7, (4) KLF4 with/without TRAF7, (5) Sonic Hedgehog pathway (SMO, SUFU, and PRKAR1A), and (6) POLR2A. Initial associations between subgroups and clinical features have been described. However, further investigation in a large cohort could identify novel relationships and validate previous observations, which could advance our understanding of tumorigenesis and guide clinical management. Recently established is the distinct subgroup of meningiomas harboring mutations in the SMARCB1 tumor suppressor gene, a core subunit of the SWI/SNF chromatin-remodeling complex. Notably, SWI/SNF mutations in other tumors have been associated with dysregulation of the PRC2 complex, an epigenetic regulator whose signature marker of activity is H3K27me3. However, the molecular mechanism of tumorigenesis in SWI/SNF-mutant meningiomas remains unknown, limiting options for targeted therapies. Thus, the aims of this study were three-fold: (1) genomically characterize mutation-unknown meningiomas to identify novel driver and/or co-driver genes, (2) perform gene expression and H3K27me3 methylation analyses to assess PRC2 activity in SWI/SNF-mutant meningiomas, and (3) perform genomic-clinical correlation analyses to identify novel associations and validate previous observations. Methods included targeted screening of known meningioma driver genes via molecular inversion probe sequencing, custom amplicon sequencing, and/or quantitative real-time PCR. Remaining mutation-unknown samples were analyzed via whole-exome and whole-genome sequencing. Moreover, the gene expression and H3K27me3 methylation profiles of SWI/SNF-mutant meningiomas was characterized via RNA gene expression microarrays and H3K27me3 chromatin immunoprecipitation followed by sequencing, respectively, and correlated to assess PRC2 activity. Lastly, a large cohort of meningiomas was analyzed for genomic-clinical correlations using Chi-squared and Fisher’s exact tests (for nominal variables) and Kruskal-Wallis tests followed by Dunn’s method (for continuous and ordinal variables); unbiased clustering was performed using clinical features. We identify novel candidate meningioma driver mutations in additional SWI/SNF subunits including PBRM1, ARID2, and SMARCD1. We demonstrate elevated H3K27me3 signal and concordantly decreased gene expression in SWI/SNF-mutant meningiomas, suggesting increased PRC2 activity either by decreased inhibition (loss-of-function) or potentiation (neomorphic function) by mutant SWI/SNF. Inhibitors of EZH2, the H3K27me3 catalytic subunit of PRC2, may thus prove efficacious in SWI/SNF-mutant meningiomas. Lastly, we identify novel genomic-clinical correlations, including associations between KLF4 and female sex, NF2 and male sex, NF2 and greater tumor volume, and POLR2A and lateral posterior cranial fossa locations, in addition to validating previous observations. In conclusion, this study identifies additional novel candidate meningioma driver mutations in SWI/SNF subunits, proposes a molecular mechanism underlying SWI/SNF-mutant meningiomas, and reveals novel genomic-clinical correlations in addition to validating previous associations. Overall, these results may guide future targeted therapies and clinical management to improve the outcomes of patients with meningiomas