PURPOSE

The oncologic outcomes for atypical meningiomas can be poor. Generally, patients that have had a prior recurrence have a substantially elevated risk of a future recurrence. Additionally, certain tumor genomic profiles have been shown as markers of poor prognosis. We sought to characterize the genomic differences between primary and recurrent tumors as well as assess if those differences had implications on recurrence.

METHODS

We identified primary and recurrent gross totally resected WHO grade II meningiomas with > 30 days of post-surgical follow-up at our institution. For genes with a prevalence of > 5% in the cohort, we compared the mutational prevalence in primary and recurrent tumors. For a gene of interest, we assessed the time to radiographic recurrence using adjusted cox-regression.

RESULTS

We identified 88 meningiomas (77 primary, 16 recurrent) with a median follow-up of 5.33 years. Mutations in ARID1A found in association with recurrent tumors (7/16 recurrent tumors vs 5/72 primary tumors, p < 0.001). In the whole cohort, mutations in ARID1A were not associated with alterations in time to recurrence after adjusting for recurrence status (p = 0.713). When restricted to primary tumors, ARID1A is associated with a 625% increase in the hazard of recurrence (HR = 7.26 [1.42-37.0]; p = 0.017).

CONCLUSION

We demonstrate mutations in ARID1A, a chromatin remodeling gene, in a higher prevalence in recurrent tumors. We further demonstrate that when mutations in ARID1A are present in primary atypical meningiomas, these tumors tend to have worse prognosis. Further prospective study may validate ARID1A as a prognostic marker.