Population pharmacokinetic and pharmacodynamic modeling of norvancomycin

In this paper, population pharmacokinetics (PPK) and pharmacodynamics of norvancomycin in patients were investigated. The studied dataset was derived from 146 patients with confirmed or suspected gram-positive bacterial infections, as well as 20 healthy volunteers. A PPK model was developed and validated by the nonlinear mixed effect model (NONMEM) software. The norvancomycin minimum inhibitory concentrations (MICs) for the isolates from patients were determined by the agar dilution method. The best model was a two-compartment pharmacokinetic model with exponential inter-individual error and an additive residual error statistic model. The findings of the present study indicated that the change in CLcr values had different effects on drug clearance (CL). In patients with renal dysfunction (CLcr <= 85 /min), CL (L/h)=2.54 center dot(CLcr /50)(1.20), while in patients with normal renal function (CLcr > 85 ml/min), CL=6.0 center dot(WT/60)(0.52). An increased volume of peripheral distribution (V-2) was observed when norvancomycin was co-administered with diuretics. Inter-individual variability in CL, V-1, Q, and V-2 was 35.92%, 11.40%, 0, and 79.75%, respectively. Residual variability was 3.05 mg/L. The logistic stepwise analyses revealed that only the ratio of AUC(24) /MIC was a major factor which could significantly predict the clinical outcome and bacterial eradication in patients. As the AUC(24)/MIC ratio was > 579.90 for staphylococcal infection and > 637.67 for enterococcal infections, approximately 95% of patients would be predicted to achieve a cured clinical outcome. In conclusion, AUC(24)/MIC should be a major pharmacokinetics/pharmacodynamics (PK/PD) parameter to predict the clinical efficacy of norvancomycin. An optimized regimen of norvancomycin can be simulated and developed for different subgroups of patients who have special physiologic and pathologic conditions.