Medical Student Northwestern University Feinberg School of Medicine San Ramon, California, United States
Introduction: Preoperative features of meningioma patients, such as male sex, non-skull base location, and larger tumor volume, are predictive of tumor recurrence. However, prognostication with these features has not been evaluated in the context of newer molecular classifications. This study investigated whether preoperative clinical variables remain predictive of recurrence when molecular data are also considered.
Methods: Data was collected for 1170 meningiomas from a multi-institutional cohort, including cases with DNA methylation profiling (n=773) and targeted gene expression profiling for calculation of a 34-gene expression risk score (n=394). Preoperative variables included tumor location, biological sex, and peritumoral brain edema. Molecular variables included DNA methylation group, gene expression risk group, and CNV profile. Fisher’s Exact Test was used to determine associations between preoperative and molecular variables. Univariate and multivariate Cox-regression was used to determine association of preoperative and molecular features with progression free survival (PFS).
Results: Male sex and non-skull base (NSB) location were significantly associated with aggressive molecular groups: hypermitotic methylation subgroup (q < 0.0001, male OR: 2.23, NSB OR: 1.89) and high RNA risk score group (q < 0.0001, male OR: 1.74, NSB OR: 1.57). Only male sex was associated with chromosome 1p loss (q = 0.0014, OR: 1.97). Preoperative clinical features were not associated with PFS within methylation subgroups or RNA risk score groups. On univariate analysis, there was no significant difference in PFS between Merlin-intact and immune-enriched methylation subgroups. All RNA risk score groups exhibited significantly different PFS (q < 0.0001, HR: 3.69 & 14.72 for intermediate and high-risk groups). In univariate format, male sex was associated with PFS (q < 0.0001, HR: 1.85). In multivariate format, no preoperative clinical features were associated with PFS.
Conclusion : Preoperative clinical variables are associated with aggressive molecular groups but are not independently prognostic for PFS. These results highlight the importance of molecular characterizations of meningiomas over clinical features for prognostication.