Introduction: While WHO Grade I and Grade III meningiomas are clinically and molecularly well-described,
Grade II tumors, comprising 20% of all meningiomas, are less understood. This class
exhibits features of a distinct oncologic entity, such as varied treatment responsivity,
natural history, and increased malignant potential. However, incomplete characterization
of Grade II meningiomas limits the ability to tailor therapy to this intermediate
grade. Therefore, the goal of this study was to determine if genetic features distinguish
Grade II from Grade I tumors.
Methods: Through a retrospective chart review, 95 patients were identified who underwent surgical
resection of intracranial meningiomas at our institution between 2008 and 2018. Following
exclusion criteria application, formalin-preserved samples, which were collected intraoperatively,
from nine WHO Grade I and ten WHO Grade II meningiomas underwent bulk mRNA sequencing.
Notably, these primary analyses were performed in two separate experiments, but transcriptomic
results were normalized to internal control genes to address the resulting batch effects.
Differential expression analysis yielded fold-change (Grade II over Grade I) and multiple-test-corrected
significance (q-value < 0.05) values for each of the 45,000 genes assessed. Gene-set enrichment analysis
(GSEA), with each gene weighted by its differential expression, was then performed
using the Molecular Signatures Database (MSiG) and each of its nine human reference
gene-sets. Normalized enrichment scores (positive equals enriched in Grade II over
Grade I) and q-values were produced for each of the reference ontology terms.
Results: A total of 28,869 genes had sufficient expression in all samples for further analysis.
Compared to Grade I, Grade II meningiomas displayed an increased average expression
of 3,512 genes and decreased average expression of 5,615 genes ([Fig. 1A]). To better understand the biological relevance of these transcriptomic differences,
GSEA uncovered 11,433 unique ontology terms significantly altered in Grade II tumors
([Fig. 1B]). Interestingly, 10,057 of these pathways were enriched in Grade II (Normalized
Enrichment Score > 1.23; q-value < 0.05). In terms of degree of enrichment, immunologic functions and cell-type
determinants represented 65% and 31% of the top 100 terms, respectively. We next focused
on the Hallmark library of MSiG, which provides a more general picture of pathway
enrichment ([Fig. 2]). Among the top 10 enriched pathways, several metabolic functions, including oxidative
phosphorylation and glycolysis, were upregulated in Grade II meningiomas. Similarly,
there was a significant increase in the cellular stress pathways involving p53, hypoxia
responses, reactive oxygen species production, and the epithelial-mesenchymal transition.
Immunologic regulation, including MTORC1 signaling and TNF-alpha signaling via NFKB,
was also meaningfully elevated.
Conclusion: Through bulk transcriptomic analysis, the present study detected functionally relevant
alterations in Grade II meningiomas which distinguish this class from Grade I tumors.
Specifically, enhanced expression of metabolic, cellular stress, and immunologic pathways
potentially underly the higher clinical severity of Grade II tumors. Future studies
will seek to further uncover the specific mediators of these hallmarks and ascertain
their mechanistic impact and potential role in treating Grade II meningiomas.
