Pharmacodynamic analysis of the activity of quinupristin-dalfopristin against vancomycin-resistant Enterococcus faecium with differing MBCs via time-kill-curve and postantibiotic effect methods.

Quinupristin-dalfopristin (Q-D) is a new water-soluble, semisynthetic antibiotic that is derived from natural streptogramins and that is combined in a 30:70 ratio. A number of studies have described the pharmacodynamic properties of this drug, but most have investigated only staphylococci or streptococci. We evaluated the relationship between Q-D, quinupristin (Q), and/or dalfopristin (D) susceptibility parameters and antibacterial activities against 22 clinical isolates of vancomycin-resistant Enterococcus faecium (VREF) by using the concentration-time-kill-curve method and by measuring postantibiotic effects. Q-D, Q, and D MICs and minimum bactericidal concentrations (MBCs) ranged from 0.125 to 1 and 0.25 to 64, 8 to 512 and >512, and 2 to 8 and 8 to 512 microgram/ml, respectively. There were no significant relationships between susceptibilities to the individual components and the susceptibilities to the Q-D combination product. In the time-kill-curves studies, Q-D at a concentration of 6 microgram/ml was at least bacteriostatic against all VREF tested. There was increased activity against more susceptible isolates when the isolates were grouped either by Q-D MBCs or by Q MICs. By multivariate regression analyses, the percent change in the inoculum from that at the baseline was significantly correlated with the Q MIC (R = 0.74; P = 0.008) and the Q-D concentration-to-MBC ratio (R = 0.58; P = 0.02) and was inversely correlated with the Q-D MBC-to-MIC ratio (R = 0.68; P = 0.003). A strong correlation existed between the killing rate and the Q-D concentration-to-MBC ratio (R = 0.99; P < 0.0001). Time to 99.9% killing was best correlated with the Q-D MBC (R = 0.96; P < 0.0001). The postantibiotic effect ranged from 0.2 to 3.2 h and was highly correlated with the Q-D concentration-to-MBC ratio (R = 0.96; P < 0.0001) and was less highly correlated with the Q MIC (R = 0.42; P = 0.04). Further study of these relationships with in vitro or in vivo infection models that simulate Q-D pharmacokinetics should further define the utility of these pharmacodynamic parameters in the prediction of Q-D activity for the treatment of VREF infections in humans.