Background Patient outcomes during ventriculitis may be improved by antibiotic dose optimisation strategies that increase the achievement of therapeutic concentrations at the infection site. We performed a population pharmacokinetic (PK) study in neurocritical care patients to define ceftriaxone dosing regimens required to achieve effective cerebrospinal fluid (CSF) exposures.

Methods Patients receiving ceftriaxone for treatment of ventriculitis or extracerebral infections or for prophylaxis following external ventricular drain insertion were recruited and subject to serial plasma and CSF sampling. Population PK modeling and dosing simulations to achieve the following plasma targets: (a) unbound ceftriaxone concentration above pathogen minimum inhibitory concentration over the dosing interval (100% fT>MIC) and (b) unbound ceftriaxone concentration at least fourfold above pathogen minimum inhibitory concentration over the dosing interval (100% fT>4×MIC), were performed.

Results Ten patients were recruited; median age, weight, and creatinine clearance were 57 years, 60 kg, and 107 mL/min/1.73m2, respectively. Ceftriaxone PK displayed considerable variability, especially in CSF, with between subject variability ranging from 21% to 794%. Median total ceftriaxone CSF penetration was 1.43% (range 0.33–8.42). Intermittent infusions of 2 g every 8 hours achieved 99.5% and 82% probability of attaining 100% fT>MIC and fT>4×MIC in plasma for an MIC of 1 mg/L, respectively. The model was unable to accurately predict ceftriaxone concentrations in CSF, precluding CSF dosing simulations.

Conclusions High attainment of plasma target exposures was achieved with higher than standard dosing. Dosing recommendations to optimise targeted CSF ceftriaxone exposures for treatment of ventriculitis could not be made given inadequate model predictability.