Brain Invasion in Meningiomas: Incidence and Correlations with Clinical Variables and Prognosis

OBJECTIVE: In meningioma, correlation of brain invasion with prognosis and clinical variables remains controversial. METHODS: Correlation of brain invasion with clinical and histopathologic variables was investigated in 467 patients with primary intracranial meningioma. RESULTS: Diffuse (n=3; 10%), clusterlike (n=11; 34%) or fingerlike (n =18; 56%) invasion was detected in 32 patients (7%). Brain invasion was more common in males than in females (13% vs. 5%; odds ratio, 2.75; 95% confidence interval, 1.29-5.89; P =0.009) and pattern of invasion differed between genders (P =0.037). Brain invasion was absent in 401 benign meningiomas and present in 48% of 60 atypical (n =29) and 50% of 6 anaplastic (n =3) meningiomas (P < 0.001) but was independent of tumor location and extent of resection. Progression occurred in 11% and was more frequent (31% vs. 15%; P =0.036) in invasive than in noninvasive tumors, but only after gross total resection and in univariate analyses, and independent of invasion pattern. In atypical meningiomas, frequency of adjuvant irradiation was similar comparing invasive and noninvasive tumors and grading solely based on brain invasion (n =20; 33%), other World Health Organization (WHO) criteria (n =31; 52%) or a combination of both (n =9; 15%). Risk of recurrence was lower (hazard ratio, 0.258, 95% confidence interval, 0.09-0.734; P =0.011) when grading exclusively based on brain invasion than when further WHO criteria were in addition present and the progression-free interval among the first was similar to benign tumors. CONCLUSIONS: Brain invasion and its patterns are correlated to gender. In contrast to the current WHO classification, invasion was associated with recurrence only after gross total resection and not independent of further histopathologic criteria of atypia.