Treatment of staphylococcal ventriculitis associated with external cerebrospinal fluid drains: a prospective randomized trial of intravenous compared with intraventricular vancomycin therapy

Object. Staphylococcal ventriculitis may be a complication in temporary external ventricular drains (EVDs). The limited penetration of vancomycin into the cerebrospinal fluid (CSF) is well known; the pharmacodynamics and efficacy of systemically compared with intraventricularly administered vancomycin is examined in this prospective study. Methods. Ten patients in whom EVDs were implanted to treat intracranial hemorrhage and who were suffering from drain-associated ventriculitis were randomized into two treatment groups. Five of these patients (median age 47 years) were treated with 2 g/day vancomycin administered intravenously (four infusions/day, Group 1), and the other five (median age 49 years) received 10 mg vancomycin intraventricularly once daily (Group 2). Vancomycin levels were measured in serum and CSF six times a day. The maximum vancomycin level in CSF was 1.73 +/- 0.4 mug/ml in Group 1 and 565.58 +/- 168.71 mug/ml 1 hour after vancomycin application in Group 2 (mean +/- standard deviation). Vancomycin levels above the recommended trough level of 5 mug/ml in CSF were never reached in Group 1, whereas in Group 2 they were below the trough level (3.74 +/- 0.66 mug/ml) only at 21 hours after intraventricular vancomycin application. The vancomycin level in the serum was constant within therapeutic levels in Group 1, whereas in Group 2 in most instances vancomycin was almost below a measurable concentration. In both groups bacteriologically and laboratory-confirmed CSF clearance could be obtained. Conclusions. Intraventricular vancomycin application is a safe and efficacious treatment modality in drain-associated ventriculitis, with much higher vancomycin levels being achieved in the ventricular CSF than by intravenous administration.