Pharmacodynamics of vancomycin at simulated epithelial lining fluid concentrations against methicillin-resistant Staphylococcus aureus (MRSA): implications for dosing in MRSA pneumonia

Abstract

Little is known regarding killing activity of vancomycin against methicillin (meticillin)-resistant Staphylococcus aureus (MRSA) in pneumonia since the extent of vancomycin penetration into epithelial lining fluid (ELF) has not been definitively established. We evaluated the impact of the extent of ELF penetration on bacterial killing and resistance by simulating a range of vancomycin exposures (24-h free drug area under the concentration-time curve [fAUC24]/MIC) using an in vitro pharmacodynamic model and population-based mathematical modeling. A high-dose, 1.5-g-every-12-h vancomycin regimen according to American Thoracic Society/Infectious Diseases Society of America guidelines (trough concentration, 15 mg/liter) with simulated ELF/plasma penetration of 0, 20, 40, 60, 80, or 100% (fAUC24/MIC of 0, 70, 140, 210, 280, or 350) was evaluated against two agr-functional, group II MRSA clinical isolates obtained from patients with a bloodstream infection (MIC = 1.0 mg/liter) at a high inoculum of 10(8) CFU/ml. Despite high vancomycin exposures and 100% penetration, all regimens up to a fAUC24/MIC of 350 did not achieve bactericidal activity. At regimens of < or = 60% penetration (fAUC24/MIC < or = 210), stasis and regrowth occurred, amplifying the development of intermediately resistant subpopulations. Regimens simulating > or = 80% penetration (fAUC24/MIC > or = 280) suppressed development of resistance. Resistant mutants amplified by suboptimal vancomycin exposure displayed reduced rates of autolysis (Triton X-100) at 72 h. Bacterial growth and death were well characterized by a Hill-type model (r2 > or = 0.984) and a population pharmacodynamic model with a resistant and susceptible subpopulation (r2 > or = 0.965). Due to the emergence of vancomycin-intermediate resistance at a fAUC24/MIC of < or = 210, exceeding this exposure breakpoint in ELF may help to guide optimal dosage regimens in the treatment of MRSA pneumonia.

Figures

FIG. 1.

The killing activities of simulated vancomycin regimens in an in vitro infection model profiling the total population over time versus S203 or S204 (A and B) or a resistant population growing on vancomycin 2-mg/liter BHI agar (C and D). The pharmacodynamic relationship between the log ratio area and the ƒAUC/MIC (R2 represents the coefficient of determination) is shown in the bottom panels (E and F).

FIG. 2.

Observed and fitted bacterial counts for simulated vancomycin exposures in ELF against S203 (A) or S204 (B).

FIG. 3.

Structure of the final model with a susceptible and a resistant subpopulation describing the effect of vancomycin as inhibition (IC50,S and IC50,R) of the saturable growth function for each subpopulation. The lag compartments are linked to the compartments for replicating cells via a first-order rate constant (klag = 1/MTTlag). Arrows for loss of bacteria due to washout from the unfiltered in vitro model are not shown. See Materials and Methods for further explanation of parameters.

FIG. 4.

Postexposure isolates of MRSA S203 (A) or MRSA S204 (B) for each vancomycin regimen simulated (as shown in the key to the symbols) in the IVPM were obtained at 72 h and subsequently evaluated for differences in autolysis profiles over 8 h. The pharmacodynamic relationship between the extent of autolysis and the ƒAUC/MIC for S203 (C) or for S204 (D) is shown.