Résumé
Background Cerebral ventriculitis might be caused by Gram-negative bacteria, including ESBL producers. Temocillin may be a useful treatment option in this scenario; however, no consistent data are available regarding its penetration into the CSF.Objectives To describe the population pharmacokinetics of temocillin in plasma and CSF and to determine the probability for different simulated dosing regimens to achieve pharmacokinetic/pharmacodynamic (PK/PD) targets in the CSF.Methods Ten post-neurosurgical critically ill adult patients requiring continuous drainage of CSF were included in this monocentric, prospective, open-label, non-randomized study. They received 2 g loading dose temocillin over 30 min IV infusion, followed by a 6 g continuous infusion over 24 h. Total and unbound concentrations were measured in plasma (n = 88 and 86) and CSF (n = 88 and 88) samples and used to build a population PK model. Monte Carlo simulations were performed to estimate the PTA at 100% Css>MIC (steady state concentration above the MIC) in CSF.Results All patients were infected with Enterobacterales with temocillin MICs <= 8 mg/L. The median (min-max) temocillin penetration in CSF was 12.1% (4.3-25.5) at steady state. Temocillin unbound plasma pharmacokinetics were best described by a one-compartment model. PTA for the applied dosing regimen was >90% for bacteria with MIC <= 4 mg/L.Conclusions The currently approved dose of 6 g by continuous infusion may be adequate for the treatment of ventriculitis by Enterobacterales with MIC <= 4 mg/L if considering 100% Css>MIC as the PK/PD target to reach. Higher maintenance doses could help covering higher MICs, but their safety would need to be assessed.