Prediction of the antimicrobial effects of trovafloxacin and ciprofloxacin on staphylococci using an in-vitro dynamic model.

To compare the pharmacodynamics of trovafloxacin and ciprofloxacin, three clinical isolates of Staphylococcus aureus with different MICs (0.03, 0.15, 0.6 and 0.1, 0.25, 1.25 mg/L, respectively) were exposed to decreasing concentrations of the quinolones according to their half-lives of 9.25 and 4 h, respectively. With each organism, single doses of trovafloxacin and twice-daily doses of ciprofloxacin were designed to provide 8-fold ranges of the ratio of area under the concentration-time curve (AUC) to the MIC, 58-466 and 116-932 (mg x h/L)/(mg/L), respectively. The antimicrobial effect was expressed by its intensity: the area between the control growth in the absence of antibiotics and the antibiotic-induced time-kill/regrowth curves (I(E)). Linear relationships established between I(E) and log AUC/MIC were bacterial strain-independent but specific for the quinolones (r2 = 0.99 in both cases). At a given AUC/MIC ratio, the I(E)s of trovafloxacin were greater than those of ciprofloxacin, suggesting that the antimicrobial effect of trovafloxacin compared with ciprofloxacin against staphylococci may be even greater than might be expected from the difference in their MICs. These data were combined with previous results obtained with three Gram-negative bacteria. Again, I(E) correlated well with the log AUC/MIC of trovafloxacin and ciprofloxacin in a strain- and species-independent fashion (r2 = 0.94 and 0.96, respectively). On this basis, a value of the AUC/MIC of trovafloxacin which might be equivalent to Schentag's AUC/MIC = 125 (mg x h/L)/(mg/L) reported as the breakpoint value for ciprofloxacin was estimated at 71 (mg x h/L)/(mg/L) with the respective MIC breakpoint of 0.27 mg/L. Based on the I(E)-log AUC/MIC relationships, the I(E)s were plotted against the logarithm of trovafloxacin and ciprofloxacin dose (D) for hypothetical representatives of S. aureus, Escherichia coli, Klebsiella pneumoniae and Pseudomonas aeruginosa with MICs corresponding to the MIC50s. These I(E)-log D relationships allow prediction of the effect of a given quinolone on a representative strain of the bacterial species.