Differences in the pharmacodynamics of ceftaroline against different species of Enterobacteriaceae studied in an in vitro pharmacokinetic model of infection.

Abstract

Dose-ranging experiments were performed to study the pharmacodynamics of ceftaroline against Enterobacteriaceae. A range of fT>MIC values (0%-100%) were simulated over 96 h using a single-compartment dilutional in vitro pharmacokinetic model using Escherichia coli, Klebsiella pneumoniae, Proteus mirabilis, Citrobacter koseri and Serratia marcescens (n = 16). Antibacterial effect was assessed by change in viable count and population profiles by growth on ceftaroline MIC ×2, ×4 and ×8 agar plates. The fT>MIC (%) was related to antibacterial effect using a sigmoid Emax model. The 24 h bacteriostatic effect fT>MIC was 39.7% ± 15.7% and 43.2% ± 15.6% for a -1 log drop for all strains. E. coli required lower exposures than K. pneumoniae, i.e. 24 h fT>MIC for a -3 log drop in viable count was 40.0% ± 9.6% and 84.8% ± 15.2% for K. pneumoniae. Similarly at 96 h, fT>MIC was >100% for K. pneumoniae (for four of five strains), 27.2%-66.2% for E. coli and 16.2%-86.6% for P. mirabilis. Strain-to-strain variation within species in the fT>MIC for static and cidal effect was marked; the 24 h bacteriostatic range was 14.1%-73.4% for P. mirabilis, 34.2%-44.6% for E. coli and 42.2%-62.5% for K. pneumoniae. Changes in ceftaroline population analysis profiles were observed with E. coli, K. pneumoniae and C. koseri, especially at fT>MIC values just below the bacteriostatic effect exposures. The pharmacodynamics of ceftaroline against the species within the Enterobacteriaceae group are different. K. pneumoniae requires higher drug exposures than E. coli, and P. mirabilis strains are highly variable, which may have important clinical correlates. Translational extrapolations from preclinical observations using E. coli to other Enterobacteriaceae species may not be optimal.