Foramen magnum meningiomas: surgical results and risks predicting poor outcomes based on a modified classification

OBJECTIVE This study aimed to evaluate neurological function and progression/recurrence (P/R) outcome of foramen magnum meningioma (FMM) based on a modified classification. METHODS This study included 185 consecutive patients harboring FMMs (mean age 49.4 years; 124 females). The authors classified the FMMs into 4 types according to the previous classification of Bruneau and George as follows: Type A (n = 49, 26.5%), the dural attachment of the lesion grows below the vertebral artery (VA); Type B (n = 39, 21.1%), the dural attachment of the lesion grows above the VA; Type C1 (n = 84, 45.4%), the VA courses across the lesion with or without VA encasement or large lesions grow both above and below the bilateral VA; and Type C2 (n = 13, 7.0%), Type C1 plus partial/total encasement of the VA and extradural growth. RESULTS The median preoperative Karnofsky Performance Scale (KPS) score was 80. Gross-total resection (GTR) was achieved in 154 patients (83.2%). Lower cranial nerve morbidity was lowest in Type A lesions (16.3%). Type C2 lesions were inherently larger (p = 0.001), had a greater percentage of ventrolateral location (p = 0.009) and VA encasement (p < 0.001), lower GTR rate (p < 0.001), longer surgical duration (p = 0.015), higher morbidity (38.5%), higher P/R rate (30.8%, p = 0.009), and poorer recent KPS score compared with other types. After a mean follow-up duration of 110.3 months, the most recent follow-up data were obtained in 163 patients (88.1%). P/R was observed in 13 patients (7.2%). The median follow-up KPS score was 90. Compared with preoperative status, recent neurological status was improved in 91 (49.2%), stabilized in 76 (41.1%), and worsened in 18 (9.7%) patients. The multivariate Cox proportional hazard regression model demonstrated Type C2 (HR 3.94, 95% CI 1.04-15.0, p = 0.044), nontotal resection (HR 6.30, 95% CI 1.91-20.8, p = 0.003), and pathological mitosis (HR 7.11, 95% CI 1.96-25.8, p = 0.003) as independent adverse predictors for tumor P/R. Multivariate logistic regression analysis identified nontotal resection (OR 4.06, 95% CI 1.16-14.2, p = 0.029) and pathological mitosis (OR 6.29, 95% CI 1.47-27.0, p = 0.013) as independent risks for poor outcome (KPS score < 80). CONCLUSIONS The modified classification helped to predict surgical outcome and P/R in addition to the position of the lower cranial nerves. Preoperative imaging studies and neurological function should be reviewed carefully to establish an individualized management strategy to improve long-term outcome.