Concentration-response relationships as a basis for choice of the optimal endpoints of the antimicrobial effect: daptomycin and vancomycin pharmacodynamics with staphylococci in an in vitro dynamic model

The abilities of different indices of bacterial killing to ensure reasonable concentration-response relationships have been compared only in studies in vitro with fluoroquinolones. To ascertain the relevance of conclusions drawn in these studies to other antibiotic classes, five widely used indices that reflect the rate of initial killing (time to reduce the initial inoculum 10- and 100-fold-T-90% and T-99%, respectively), the extent of killing (minimal number of surviving organisms-N-min), and the entire antimicrobial effect (number of surviving organisms (N-t) at the time t close to the end of the observation period (48 h in most experiments), and area between the control growth curve and the time-kill curve from zero point to t-ABBC) were examined with daptomycin (DAP)- and vancomycin (VAN)-exposed Staphylococcus aureus. To compare the pharmacodynamics of DAP and VAN and examine different parameters, killing kinetics of differentially susceptible S. aureus were studied over a wide range of ratios of area under the curve (AUC) to MIC. Killing kinetics of two clinical isolates, S. aureus 866 (MICDAP 0.35 mg/L and MICVAN 0.70 mg/L) and S. aureus 10 (MICDAP 1.1 mg/L and MICVAN 1.3 mg/L), were studied in an in vitro dynamic model that simulates human pharmacokinetics of DAP (as a single dose) and VAN (as two 12-h doses). Mono-exponential concentration decays were mimicked with half-lives of 9 h (DAP) and 6 h (VAN) at AUC/MIC ratios varying from 33 to 1150 h. T-90%, T-99% and N-t (at t = 48 h) exhibited loose, if any, correlations with log AUC/MIC. Both ABBC (a direct measure of the antimicrobial effect) and N-min (an inverse measure of the effect) correlated well with log AUC/MIC (r(2) = 0.8-0.9). Based on the ABBC-log AUC/MIC relationships, the effects of DAP on S. aureus were predicted to be slightly greater than those of VAN at a given AUC/MIC ratio. The better abilities of ABBC and N-min, and the inability of T-90% and T-99% to provide reasonable AUC/MIC relationships with DAP and VAN support earlier findings reported with fluoroquinolones. (c) 2006 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.