Background: Cefotaxime is commonly used empirically in bacterial meningitis, sepsis and methicillin-sensitive Staphylococcus aureus (MSSA) bacteraemia, but optimal paediatric dosing is unclear, particularly with respect to achieving adequate cerebrospinal fluid (CSF) exposure. The population pharmacokinetics (popPK) of unbound cefotaxime in serum and CSF have not been previously reported. This study aimed to (i) develop a popPK model of unbound cefotaxime in the serum and CSF of children, and (ii) use Monte Carlo simulations (MCS) to assess probability of target attainment (PTA) for PK/PD targets of 50%fT>MIC in serum and 100%fT>MIC in CSF for clinically relevant regimens.
Methods: Serum and CSF samples for PK analysis were prospectively collected from 97 children aged <1 day to 18 years admitted to the Royal Children’s Hospital, Melbourne. Unbound cefotaxime concentrations were quantified following ultracentrifugation using high performance liquid chromatography with mass detection. PopPK modelling with covariate analysis and MCS were conducted in NONMEM to estimate CSF penetration and predict the PTA across dosing regimens.
Results: A model with three compartments – serum, CSF, and CSF peripheral compartments – best described the data. The model parameter estimates are provided in. Creatinine clearance was a covariate on renal clearance, while total body weight influenced non-renal clearance. The model well described the time-courses of serum and CSF concentrations across a wide age range, with prediction-corrected visual predictive checks demonstrating very good predictive performance (Figure 1). The median unbound CSF/serum AUC ratio was 9.2%. MCS indicated that in CSF, the 200mg/kg/day continuous infusion achieved 90% PTA at MIC breakpoint of 1mg/L, but not for S. aureus meningitis (epidemiological cutoff value 4mg/L). In serum, this regimen achieved 90% PTA for MSSA bacteraemia at 4mg/L.
Conclusions: This is the first popPK study to describe penetration of unbound cefotaxime into CSF in children. The developed model enables future MCS to optimise dosing strategies and support empirical therapy, particularly in optimising cefotaxime exposure in CSF for treatment of meningitis.
