Influence of pharmacokinetic model on vancomycin peak concentration targets

The aim of this study was to adapt the vancomycin therapeutic range to the kinetic models usually employed in clinical settings (one- and two-compartment models). Estimates of vancomycin pharmacokinetic parameters were obtained for both models in 22 hematologically malignant patients on vancomycin treatment using two serum concentrations and a bayesian algorithm. From these individually estimated pharmacokinetic parameters, an estimation of the maximum (Cssmax), 2 h postinfusion (Css2), and minimum (Cssmin) steady-state vancomycin serum concentrations for the one- and two-compartment models was made for a fixed 30 mg/kg/day dose. The linear regression equations between the predicted Css2 and Cssmin for the one- and two-compartment models do not differ significantly from the identity line, whereas the corresponding equation for Cssmax points to a 61% underestimation of Cssmax when the one-compartment model is used. From this latter regression equation, it is possible to define 20 mg/L (range of 18-21 mg/L) as a target Cssmax vancomycin serum concentration when a one-compartment model is used to monitor vancomycin therapy. Another practical approach would be to define the target concentration by a desired range at 2 h, which corresponds to a Cssmax value of 30-40 mg/L.