Pharmacodynamic modeling of carbapenems and fluoroquinolones against bacteria that produce extended-spectrum beta-lactamases

Background: Bacteria that produce extended-spectrum beta-lactamases (ESBLs) are resistant to penicillins, cephalosporins, and monobactams. The results of clinical studies suggest that the carbapenems imipenem and meropenem may be effective against bacteria that produce ESBLs, although it is not known whether the new once-daily carbapenem ertapenem or the fluoroquinolones are useful against infections caused by ESBL-producing bacteria. Objective: The present study compared the simulated pharmacodynamics of the carbapenems imipenem, meropenem, and ertapenem; the simulated pharmacodynamics of the fluoroquinolones levofloxacin, gatifloxacin, and ciprofloxacin with those of the carbapenems-1 and the simulated pharmacodynamics of levofloxacin 750 mg with those of levofloxacin 500 mg, all against gram-negative isolates that did and did not produce ESBLs. Methods: Pharmacokinetic data were obtained from studies in healthy humans. Minimum inhibitory concentrations (MICs) for bacteria that did and did not produce ESBLs were determined in triplicate using broth-microdilution techniques as recommended by National Committee for Clinical Laboratory Standards guidelines. Monte Carlo simulation was used to construct pharmacodynamic models for imipenem, meropenem, ertapenem, levofloxacin, gatifloxacin, and ciprofloxacin. Pharmacodynamic measures of interest were the probability of the free concentration remaining above the MIC greater than or equal to40% of the time (T>MIC greater than or equal to40%) for carbapenems and the likelihood of achieving a free AUC:MIC ratio greater than or equal to125 for fluoroquinolones. Results: MICs were determined for 39 isolates that produced ESBLs and 45 isolates that did not. Bacteria that did not produce ESBLs were greater than or equal to93% susceptible to all carbapenems and fluoroquinolones tested. Among bacteria that produced ESBLs, rates of susceptibility to the specific agents were as follows: imipenem, 100%; meropenem, 97%; ertapenem, 87%; levofloxacin, 54%; gatifloxacin, 44%; and ciprofloxacin, 36%. In the pharmacodynamic models, imipenem and meropenern had an equal likelihood of achieving a free T>MIC greater than or equal to40% against bacteria that produced ESBLs (greater than or equal to97%) and bacteria that did not produce ESBLs (greater than or equal to98%). In contrast, the likelihood of ertapenem achieving a free T>MIC greater than or equal to40% was lower against bacteria that produced ESBLs (78%) than against bacteria that did not produce ESBLs (94%). Similarly, the fluoroquinolones were less likely to achieve a free AUC:MIC ratio greater than or equal to125 against bacteria that produced ESBLs (2%-13%) than against bacteria that did not produce ESBLs (85%-91%). Conclusions: Carbapenems had superior in vitro activity against bacteria that produced ESBLs compared with fluoroquinolones. Pharmacodynamic modeling based on local ESBL-producing isolates and pharmacokinetic data from healthy humans indicated that imipenem and meropenem may have a greater likelihood of achieving pharmacodynamic targets against bacteria that produce ESBLs than ertapenem or fluoroquinolones. Copyright (C) 2004 Excerpta Medica, Inc.