Management of hydrocephalus associated with vestibular schwannoma and other cerebellopontine angle tumors

Abstract

Objective: Hydrocephalus (HCP) resulting from cerebellopontine angle (CPA) tumors is not rare. This retrospective study was designed to investigate the incidence of HCP and the clinical presentations, management options, and outcomes of HCP in 284 patients with CPA tumors.

Methods: A retrospective study of 284 consecutive patients with CPA tumors (mostly vestibular schwannomas) treated from 1985 to 1996 at Toronto Western Hospital managed by one surgical team consisting of a neurosurgeon and a neuro-otologist.

Results: Thirty-nine patients (13.7%) had radiographic and/or clinical evidence of HCP, 37 preoperatively and 2 postoperatively. Tumor type distribution was 33 vestibular schwannomas, 5 meningiomas, and 1 cavernous hemangioma. Only five patients (12%) had obvious obstruction at the fourth ventricular level. In 36 patients (92%), symptoms were mostly chronic and mild, consistent with normal pressure hydrocephalus. Multivariate analysis confirmed the strong association of tumor size and incidence of HCP (P < .0001). Four patients underwent permanent shunting before microsurgical tumor excision, mainly because of florid symptoms of HCP. Microsurgical tumor excision without preoperative shunting was performed in 23 patients, 5 of whom required postoperative shunting in the first 2 months after tumor excision. Eighteen patients (78%) did not need shunts after tumor resection. With regard to tumor size, the postoperatively shunted group did not differ from the patients who had surgery but did not require shunt treatment (P < 0.50). The remaining 10 patients with preoperative HCP received shunts as the only treatment (3 patients), stereotactic radiosurgery (3 patients), or expectant management (4 patients). Two other patients without preoperative HCP developed postoperative HCP and required shunts. Postoperatively, we observed a significant (P < 0.001) increase in the incidence of pseudomeningocele and a nonsignificant (P < 0.1) increase in cerebrospinal fluid leaks (rhinorrhea and/or otorrhea) in patients without shunts as compared with postoperative patients without HCP. The patients were followed after any treatment for a mean of 3.2 years (range, 6 mo-10 yr). Follow-up in the patients who had surgery but did not require a shunt revealed a 61% decrease in clinical symptoms related to HCP and a 75% decrease in radiographic signs of HCP.

Conclusion: In the presence of HCP, operative resection of CPA tumors can be performed without permanent cerebrospinal fluid shunting. Precautionary measures to decrease the incidence of postoperative complications related to cerebrospinal fluid leak in patients with preoperative HCP include meticulous obliteration of any exposed air cells, including those around the internal auditory canal, accurate restoration of the dural barrier, and temporary lowering of intracranial pressure with a ventricular or lumbar drain. Patients with persistent symptomatic HCP after tumor excision should be treated with a ventriculoperitoneal shunt. Delaying this decision until the postoperative period is safe and avoids unnecessary shunting in the majority of patients.