Fluoroquinolone AUIC Break Points and the Link to Bacterial Killing Rates: In Vitro Models

Abstract

1 vigorously advocate the use of AUC 24 / minimum inhibitory concentration (MIC) >125 for fluoroquinolones in order to achieve bacteriologic eradication against all target organisms and in all hosts. Their argument faces opposition from other groups in the field of pharmacodynamics.2 However, we welcome debate in the scientific arena because such dialogue opens the door for better understanding of fluoroquinolone pharmacokinetics/pharmacodynamics. Schentag et al.'s arguments attract attention to some important issues that necessitate further investigation and discussion concerning the way we design our in vitro modeling experiments, set study endpoints, define antibiotic break points, interpret results, and consider how antibiotic-resistant mutants develop within a bacterial population. Examination of and reflection on pharmacodynamic concepts may provide better approaches to designing in vitro and animal studies, as well as clinical trials, so that fluoroquinolones can be modeled to achieve maximum bacterial eradication (hopefully with acceptable toxicity) and optimal prevention of the development of bacterial resistance.We all agree that for fluoroquinolones, AUC 24 /MIC, and maximum concentration (C max )/MIC are the best predictors of microbiologic and, less so, clinical outcome. However, some important issues need to be considered to arrive at an optimal approach and draw conclusions from our in vitro modeling studies.First, it is time to realize that different fluoroquinolone AUC 24 /MIC (not simply AUC 24 /MIC >125) ratios are required for eradication and prevention of resistance in different organisms (e.g., Streptococcus pneumoniae, Pseudomonas aeruginosa). This has been well described previously 3,4 and confirmed by MacGowan et al.5 As well, we need to realize that the immune system in an immunocompetent host may directly or indirectly lower the AUC 24 /MIC ratios required for bacterial eradication, potentially to varied degrees, depending on the organism.Second, since only the unbound fraction of the drug is biologically active, 6,7 only the unbound (non-protein bound) fraction of the drug should be used in pharmacodynamic calculations. Thus, we should consider only the unbound drug AUC 24 /MIC and unbound drug C max /MIC pharmacodynamic predictors. This issue was made clear by Mouton et al. 8 in their efforts to standardize pharmacokinetic/pharmacodynamic terminology for antiinfective drugs.Third, it is important to set and interpret adequate antibiotic break points. Mouton 9 elegantly discussed methods to determine clinical break points and recognized the fact that those break points should be used to predict the success of therapy; microbiologic break points should have the capacity to distinguish between antibiotic naïve populations and resistant subpopulations. In the case of fluoroquinolone-resistant S. pneumoniae, clinical break points may capture 2-step mutations (parC, gyrA), which are associated with clinical failure, while microbiologic break points would capture 1-step (e.g., parC or po...