Predicting postoperative hydrocephalus in 227 patients with skull base meningioma.

OBJECT: Some patients develop communicating hydrocephalus after meningioma surgery, and this can develop into a serious clinical condition. However, this has rarely been addressed in the literature. Therefore, the authors sought to determine predictive patient variables for the occurrence of postoperative hydrocephalus following skull base meningioma surgery.

METHODS: For this purpose, the authors retrospectively analyzed all patients who underwent resection of intracranial meningiomas between 1998 and 2009 at the Department of Neurosurgery, University Hospital Zurich, Switzerland. Of 594 patients with meningioma, 227 (38%) had a lesion located at the skull base, and thus were included for analysis. The following patient variables were examined: demographic data (age and sex); tumor number (solitary vs multiple); tumor side and localization within the skull base region (anterior, medial, posterior); infiltration of the cavernous sinus; compression of the optic channel/optic nerve; tumor volume; preoperative embolization (yes/no); duration of surgery; Simpson grade of resection; histopathological features (WHO grade); number of surgeries (single vs multiple); preoperative embolization; duration of hospital stay; tumor recurrence; use of an artificial dural substitute; postoperative infection rate; and clinical outcome (Glasgow Outcome Scale score at discharge and at 3 months, and vital status at last follow-up). Hierarchical clustering, factor analysis, and stepwise regression models revealed a ranking list for the top predictive variables for the occurrence of postoperative hydrocephalus.

RESULTS: A total of 35 patients (5.9%) of the cohort of 594 developed communicating postoperative hydrocephalus, with no patient manifesting obstructive hydrocephalus. Of these 35 patients, 18 had a meningioma located at the skull base (18 [7.9%] of 227), in contrast to 17 patients with meningiomas in other locations (17 [4.6%] of 367). The following patient variables correlated with the occurrence of hydrocephalus, as defined by factor analysis: age, duration of surgery, duration of hospital stay, tumor volume, postoperative infection, and preoperative embolization. A stepwise regression analysis of the latter variables identified 2 variables as significantly predictive: age (p = 0.0012) and duration of surgery (p = 0.0013).

CONCLUSIONS: In this study, the incidence of communicating postoperative hydrocephalus was almost twice as high in patients with skull base lesions as in patients with meningiomas in other locations. Patient age, duration of surgery, duration of hospital stay, tumor volume, postoperative infection, and preoperative embolization were associated with the occurrence of hydrocephalus. In the statistical prediction model, patient age and duration of surgery were the most significant predictors of postoperative hydrocephalus after skull base meningioma surgery.