To develop a reliable 2-compartment population pharmacokinetic (PK) model for unbound ceftriaxone in a critically ill population and determine an optimal dosing regimen. This was a prospective, single-center, observational study of critically ill patients treated with ceftriaxone. Unbound serum ceftriaxone concentrations were measured using validated ultrafiltration and ultra-performance liquid chromatography-tandem mass spectrometry. PK analysis and dosing simulations were performed using an iterative 2-stage Bayesian fitting procedure and Monte Carlo simulations. The PK/pharmacodynamics (PD) target was attained when unbound serum ceftriaxone concentrations exceeded 4 times the minimum inhibitory concentration (MIC) for ≥60% of the dosing interval (ƒT>4xMIC ≥ 60%). 91 patients were enrolled, and 173 unbound ceftriaxone concentrations were acquired. The population PK parameter estimates were hepatic clearance 5.2±0.9L/h/1.85m2, the unbound renal clearance of ceftriaxone divided by the creatinine clearance 0.61±0.24, lean body mass corrected volume of distribution of the central compartment 0.82±0.21L/kg, and intercompartmental distribution rate constant from central to peripheral compartment 0.18±0.08h-1. Dosing simulations predicted ƒT>4 mg/L of 88% (95% CI: 69-100%) for 2,000 mg ceftriaxone once daily and ƒT>4 mg/L of 100% (95% CI: 100-100%) both for 1,000 mg twice daily and continuous infusion of 2,000 mg daily. We developed a reliable population PK model for unbound ceftriaxone in a critically ill population. Dosing simulations revealed ƒT>4 mg/L ≥60% for 1,000mg twice daily and 2,000 mg once daily or by continuous infusion.
Read full abstract