CSF penetration of vancomycin in critical care patients with proven or suspected ventriculitis: a prospective observational study

Background Vancomycin is recommended for ventriculitis. However, penetration into the CNS is relatively poor. Objectives To investigate the population pharmacokinetics of vancomycin in serum and CSF in critical care patients with proven or suspected CNS infections from neurosurgical procedures. Patients and methods This was an observational pharmacokinetic study in critical care patients with proven or suspected CNS infections receiving intravenous vancomycin. Multiple blood and intraventricular CSF samples were collected. Population pharmacokinetic analysis and simulation were undertaken with ADAPT5 and Pmetrics. Results A total of 187 blood and CSF samples were collected from 21 patients. The median (range) C-max and C-min concentrations in serum were 25.67 (10.60-50.78) and 9.60 (4.46-23.56) mg/L, respectively, with a median daily dose of 2500 (500-4000) mg. The corresponding median concentrations in CSF were 0.65 (<0.24-3.83) mg/L and 0.58 (<0.24-3.95) mg/L, respectively. The median AUC(0-24) in serum and CSF was 455.09 and 14.10mgh/L, respectively. A three-compartment linear population pharmacokinetic model best fitted the observed data. Vancomycin demonstrated poor penetration into CSF, with a median CSF/serum ratio of 3% and high intersubject pharmacokinetic variability of its penetration. Conclusions Therapeutic drug monitoring in both serum and CSF and higher daily doses may be an option to ensure adequate trough levels and to optimize patient therapy. Novel dosing strategies designed to reduce renal toxicity, such as administration by continuous infusion, should be investigated in further clinical studies to avoid antibiotic underexposure in CSF.