Date of Award

January 2015

Document Type


Degree Name

Medical Doctor (MD)



First Advisor

Murat Gunel

Subject Area(s)

Neurosciences, Genetics


Meningiomas are the most common central nervous system primary tumors. Mutations in the tumor suppressor gene Neurofibromin 2 (NF2), located on chromosome 22 (Chr22), are present in 40-50% of sporadic meningiomas. In a subset of non-NF2-mutant meningiomas, recent studies have identified tumorigenesis driver mutations in the genes TRAF7, AKT1, KLF4, and SMO. However, the genetic basis for the marked clinical and histological heterogeneity that exists among NF2-mutant/Chr22-loss meningiomas remains to be established. In this study, we utilized next-generation sequencing techniques to identify and screen a large cohort of NF2-mutant/Chr22-loss meningiomas for concurrent co-driver mutations in novel genes that may contribute to the observed clinical and histological heterogeneity of these tumors.

We identified 25 NF2-mutant/Chr22-loss meningiomas that harbored concurrent somatic mutations in the SMARCB1 gene. SMARCB1 codes for a component of the SWI/SNF complex and is involved in epigenetic modification via nucleosome modulation and chromatin remodeling. SMARCB1 is also known to interact with GLI1, an important effector of the Hedgehog pathway, and EZH2, a member of the Polycomb-group proteins that is capable of epigenetically silencing gene expression via histone methylation. SMARCB1 somatic mutations have previously been associated with the highly aggressive malignant rhabdoid tumors and germline variants have previously been identified as the driver mutation responsible for familial schwannomatosis. Notably, a subset of patients with familial schwannomatosis develop multiple meningiomas as part of their disease course. We performed RNA expression analysis to determine gene expression differences in SMARCB1-mutant meningiomas versus non-SMARCB1-mutant meningiomas. Unsupervised hierarchical clustering demonstrated that the expression profiles of SMARCB1-mutant meningiomas cluster similarly to other NF2-mutant/Chr22 loss meningiomas. However, within this superfamily, SMARCB1-mutant meningiomas form distinct subgroups. This differential clustering is due, at least in part, to increased GLI1 and EZH2 expression suggesting that SMARCB1-mutant tumors represent a distinct genetic subtype of NF2-mutant meningiomas.

Genomic-clinical correlates for SMARCB1-mutant tumors showed no significant differences in this cohort in age at time of surgery (median = 64 years), gender predilection (80% female), or low-grade histologic subtype when compared to the natural history literature. However, the SMARCB1-mutant meningiomas were predominantly midline tumors (71%) with increased propensity to being high- grade lesions (36%), despite remarkable chromosomal stability. These clinical findings strongly paralleled our genetic data suggesting that increased expression of GLI and EZH2 may, respectively, contribute to the midline tumor location and serve as a substitute for widespread chromosomal instability in high-grade lesions. These results demonstrate that SMARCB1-mutant meningiomas represent a genetically and phenotypically distinct sub-group of NF2-mutant meningiomas and partially contribute to the observed clinical heterogeneity of convexity lesions. This study also suggests potential targets for therapeutic interventions that warrant future investigation.