Hydrocephalus in Patients with Vestibular Schwannoma: Institutional Experience with Temporary and Permanent CSF Diversion

Background Hydrocephalus (HCP) associated with vestibular schwannoma (VS) can be communicating (C-HCP) or obstructive (O-HCP). O-HCP occurs preoperatively, while C-HCP occurs pre- and postoperatively as well as after Gamma Knife radiosurgery (GKRS). We report our experience managing hydrocephalus in patients with vestibular schwannoma.

Methods Electronic medical records were searched from January 2000 to July 2017 for patients with a diagnosis of VS and HCP or ventriculomegaly undergoing GKRS or surgery managed with an external ventricular (EVD) or lumbar drain (LD) or ventriculoperitoneal (VPS) or lumboperitoneal shunt (LPS). Sixty-two patients were found; 2 were excluded as they had VPS placed at other institutions after primary tumor management. Collected data included patient characteristics; CSF diversion type; intervention timing; tumor size; surgical approach; symptomatic hydrocephalus categorized as asymptomatic, possibly symptomatic, or symptomatic; hydrocephalus symptoms including normal pressure hydrocephalus (NPH)-type symptoms—gait and mental status changes and urinary incontinence—and obstructive-type symptoms—nausea, vomiting, headache, eye motility, and vision changes; and reason for shunt placement. O-HCP was defined as enlarged ventricles with tumor occlusion of the cerebral aqueduct or fourth ventricle by loss of T2 signal on preoperative MRI. C-HCP was defined as enlarged ventricles with continuity of the ventricular system on T2 preoperative MRI.

Results Over the study time period, 583 patients have undergone GKRS and 734 surgery for VS. Nine (9/583,1.5%) patients had VPS after GKRS (median age and tumor size 69 [57−81] years and 1.7 [1.1−2.5] cm, respectively); all had C-HCP and NPH-type symptoms at a median of 1.6 (0.5−11) years and improved with VPS. Fifty-one (51/734,6.9%) patients had a retrosigmoid or translabyrinthine approach and CSF diversion before and/or after surgery (median age and tumor size: 55 [18−86] years and 4 [1.2−5.5] cm, respectively). Forty-five (88%) patients had CSF diversion prior to tumor removal. Thirty-seven (82%) with O-HCP and 8 (18%) with C-HCP; 22(49%) were asymptomatic, 15 (33%) possibly symptomatic, and 8 (18%) symptomatic. Diversion included 31 (69%) EVD, 8 (18%) VPS, and 6 (13%) LD. All LD and 12 EVDs were placed immediately prior to surgery; the remaining 19 EVDs at a median of 1 (1−3) days prior to surgery. Patients with VPS had placement at a median of 30.5 (11−47) days prior to surgery. Eight (18%) patients who had temporary CSF (6 EVD, 2 LD) eventually underwent VPS at a median time of 14 (2−239) days. Reasons included three patients with recurrent CSF leak, three for failed EVD weaning, and two for NPH-type symptoms. Six (6/51, 12%) patients had permanent CSF diversion after initial surgery (5 VPS, 1 LPS) at a median time of 28 (9−276) days—5 for recurrent CSF leak, 2 of which also had NPH-type symptoms, and 1 for NPH-type symptoms alone.

Conclusion Regardless of either GKRS or surgery as treatment for VS, rates of permanent CSF diversion were low at 1.5 and 3.0%, respectively. Including only surgical patients who underwent permanent diversion postoperatively decreases the rate to 1.9%. Patients who underwent temporary CSF diversion had a conversion rate to permanent of 18%.

Publication History

Publication Date:
02 February 2018 (online)