Factors influencing fluoroquinolone resistance - Letters - Letter to the Editor

Emerging Infectious Diseases, Dec, 2003 by Daniel F. Sahm, Clyde Thornsberry, Mark E. Jones, James A. Karlowsky

To the Editor: Recently, Scheld summarized factors that he considered to have an influence on the efficacy of fluoroquinolones (1). In the review, ciprofloxacin was presented as the most active fluoroquinolone against Pseudomonas aeruginosa with MICs typically two- to eightfold lower than those for levofloxacin, moxifloxacin, or gatifloxacin. However, because the National Committee for Clinical Laboratory Standards (NCCLS) MIC interpretative breakpoints are fluoroquinolone-specific, percent susceptibility is considered to be a better measure by which to compare fluoroquinolone activities. Our company has conducted annual investigations called TRUST (Tracking Resistance in the United States Today) since 1996. These surveillance studies have consistently shown similar susceptibility rates for levofloxacin (67.7% in 2002) and ciprofloxacin (67.4% in 2002) against P. aeruginosa (2,3). Both agents show higher in vitro activity against P. aeruginosa than gatifloxacin and moxifloxacin (2-4). A critique of antipseudomonal fluoroquinolone activity should also consider peak achievable fluoroquinolone levels at a site of infection, the area under the serum concentration curve in 24 hours (AU[C.sub.24h]), and the AU[C.sub.24h]/MIC ratio (5). At equivalent dosages for nosocomial pneumonia, levofloxacin (750 mg intravenously, once daily) has a threefold higher peak serum level ([C.sub.max]) and threefold higher AU[C.sub.24h] than ciprofloxacin (400 mg intravenously, every 8 hours) (package inserts for Levaquin and Cipro). While certain P. aeruginosa isolates have lower ciprofloxacin than levofloxacin MICs, the two fluoroquinolones have equivalent activity against P. aeruginosa because of their equivalent AU[C.sub.24h] /MIC ratios (6). We agree strongly with Scheld's suggestion that the fluoroquinolone used clinically should be the fluoroquinolone tested by the laboratory and reported; surrogate testing of fluoroquinolones may lead to major errors in reporting, particularly for Enterobacteriaceae (2,3,7).

The review also stated that levofloxacin-resistant strains of P. aeruginosa emerge at a significantly higher rate than with ciprofloxacin. However, a recent study of P. aeruginosa isolated from cystic fibrosis patients reported that fewer resistant mutants were isolated after exposure to levofloxacin (11 mutants) than to ciprofloxacin (28 mutants) (8).

With regards to S. pneumoniae, the review stated that in vitro studies have demonstrated that ciprofloxacin (1-4 mg/L) and levofloxacin (1-2 mg/L) are not as active as moxifloxacin (0.06-0.25 mg/L) and gatifloxacin (0.5-1 mg/L) against pneumococci. As with P. aeruginosa, fluoroquinolone comparisons against S. pneumoniae should not be limited to MICs alone because pharmacokinetic and pharmacodynamic characteristics differ for each fluoroquinolone. Pneumococcal time-kill studies with levofloxacin, gatifloxacin, and moxifloxacin in a pharmacodynamic model have demonstrated that these three agents possess equal bactericidal activity and are equally effective in preventing resistance development because the lower in vitro MICs for gatifloxacin and moxifloxacin were offset by the higher serum and tissue levels of levofloxacin (9). In the same study, ciprofloxacin did not exhibit rapid killing and selected for resistance faster than the other three agents (9). TRUST and other U.S. surveillance studies, using the NCCLS-recommended broth-dilution method, have shown that S. pneumoniae remain highly susceptible to levofloxacin with resistance rates in the United States of <1%; the MI[C.sub.90] for levofloxacin in these studies has remained at 1 mg/L from 1997 through 2002 (10-15). Further, levofloxacin, gatifloxacin, and moxifloxacin are equally effective in rates of clinical cure and microbiologic eradication of pneumococcal respiratory infections (16, and FDA website; available from: URL: http://www.fda. gov/cder/foi/nda/99/21061_Tequin.htm and http://www.fda.gov/cder/foi/ nda/2001/21277 Avelox.htm)

The review implied that, in general, higher AU[C.sub.24h]/MIC ratios were associated with better patient outcomes. For S. pneumoniae, several pharmacodynamic studies have demonstrated that a target AU[C.sub.24h]/MIC ratio of 30 to 35 for fluoroquinolones is the best correlate for successful bacteriologic eradication, clinical cure, and prevention of emergence of resistance during therapy (5,9,17-19). Levofloxacin, gatifloxacin, and moxifloxacin all achieve this AU[C.sub.24h]/MIC ratio (9). Zhanel et al. demonstrated that AU[C.sub.24h]/MIC ratios above the target value of 30 to 35 did not improve bacteriologic eradication or reduce the emergence of resistance (9). Moreover, no clinical data support the claim that higher AU[C.sub.24h]/MIC ratios correlate with better patient outcomes.

The review discusses the question of whether C-8-methoxyquinolones (moxifloxacin and gatifloxacin) have a lower propensity to select resistant mutants of S. pneumoniae compared with levofloxacin. Mutation prevention concentration is a theoretical laboratory concept based on agar dilution methodology, and no published data have shown any clinical correlation between this theory and clinical outcomes. NCCLS does not recommend agar dilution for susceptibility analysis of S. pneumoniae. Moreover, the extremely low levels of resistance in S. pneumoniae (<1%) after many years of fluoroquinolone use do not support the theory of mutation prevention concentration. The review did not reference an analysis of 16 penicillin-resistant S. pneumoniae strains by Kolhepp et al. (20). In that broth-dilution study, in vitro resistance developed in a greater proportion of strains exposed to gatifloxacin (11/16) and moxifloxacin (8/16) than to levofloxacin (2/16). Similarly, in a study by Klepser et al. that used an in vitro pharmacodynamic model, levofloxacin was less likely than moxifloxacin to select for resistant isolates of S. pneumoniae; moreover, after 24 hours of exposure, levofloxacin MICs remained unchanged while moxifloxacin MICs increased two- to eightfold (21).