Population pharmacokinetics of vancomycin in patients with diabetic foot infection: a comparison of five models

Abstract

Purpose: This study aimed to compare individual pharmacokinetic (PK) parameters of vancomycin with predicted values from five population PK models in patients with diabetic foot infections (DFIs).

Methods: Patients with a diagnosis of DFI and an estimated glomerular filtration rate (eGFR) ≥ 30 mL/min were included in the study. Individual PK data was carried on by collecting three vancomycin serum concentrations in a steady-state condition. Five published population-based nomograms were assumed to predict PK parameters. Optimal vancomycin exposure was considered as a trough level of 15-20 mg/L or the area under the curve over 24 h/minimum inhibitory concentration (AUC₂₄/MIC) ≥ 400.

Results: A total of 48 samples from 16 patients were analyzed. There was a statistically significant difference between the volume of distribution (V_d) obtained from population methods and the individual estimations (P ≤ 0.001 in Ambrose and Burton, P = 0.010 and 0.006 in Bauer and Burton revised models, respectively). AUC/MIC ≥ 400 was achieved in 68.7% of patients while 50% had a trough level of less than 15 mg/L.

Conclusions: Vancomycin PK parameters, particularly individualized V_d, may not be predictable by population nomograms in patients with DFI and stable renal function. Moreover, the weak correlation between AUC₂₄ values and trough concentrations underlines the starting practice of vancomycin AUC₂₄-based monitoring and dosing in the clinical setting.

Keywords: Area under the curve (AUC); Diabetic Foot infection; Methicillin-resistant Staphylococcus aureus (MRSA); Pharmacokinetics; Therapeutic drug monitoring (TDM); Vancomycin.

Fig. 1

The distribution of mean C_t correlated with AUC₂₄. The linear regression line defined by the following equation: AUC = 10.37 C_t + 429.1