Resistance emerging after fluoroquinolone therapy was investigated in a murine model of Pseudomonas aeruginosa infection. Mice were infected intraperitoneally by one of six strains and treated with pefloxacin or ciprofloxacin. In mice challenged with a low inoculum (1.6 x 105 CFU), no resistance occurred. With a higher inoculum (1.5 x 108 CFU) and after a single dose of antibiotic, posttherapy (PT1) strains with decreased susceptibility to quinolones (4-to 32-fold less) were isolated at a variable rate. The presence of talcum (125 mg) in the peritoneal cavity increased the risk of resistance after therapy. Pefloxacin (25 or 200 mg/kg) and ciprofloxacin (25 mg/kg) yielded similar resistance rates (61 to 77%), but ciprofloxacin (10 mg/kg) produced more resistance (83%) than did ciprofloxacin (50 mg/kg) (44%) (P < 0.02). Combined with a quinolone, ceftazidime (P < 0.001) or amikacin (P < 0.01), but not piperacillin, reduced the emergence of resistance. After several doses of ciprofloxacin, it was found that 25-mg/kg doses every 12 h produced more resistance than did 25-mg/kg doses every 8 h or 50-mg/kg doses every 12 h. Compared with the preceding experiments using parent strains, ciprofloxacin and pefloxacin were less efficient in killing bacteria in mice infected with PT, strains.Moreover, in one of these mice, a highly resistant PT2 strain (64-fold MIC increase for the quinolones) emerged. Besides increased MICs of the quinolones, there was a two-to eightfold increase in imipenem MIC for all PT1 and PT2 strains without alteration of other ,-lactam and aminoglycoside susceptibility. Some PT, strains also showed a decreased susceptibility to trimethoprim and chloramphenicol. During therapy with a quinolone, resistance can emerge rapidly, especially when there is a large number of bacteria or a foreign body present. This risk may depend on the dosing schedule and may be reduced by combined therapy.During the past decade, the emergence of resistance after therapy with newer ,-lactam compounds has been an increasing concern (12). In our laboratories, we have reproduced this phenomenon by using an experimental model of Enterobacter cloacae infection (6a). We showed that the emergence of resistance varied with the P-lactam compound given to the animal. When the newer fluoroquinolones appeared, it was hoped that the development of resistance during therapy would be avoided (6). As a result of the improved antibacterial activity of drugs such as pefloxacin and ciprofloxacin, the resistant variants found in the bacterial populations, although less susceptible than the parent strains, might still be regarded as susceptible to these drugs (6). Unfortunately, this did not hold true, and several preliminary therapeutic trials have shown that resistant strains emerging after therapy could generate therapeutic difficulties, especially in Pseudomonas infections (11,16). In this paper, we compared the abilities of pefloxacin and ciprofloxacin to produce resistance in our murine model, with six strains of Pseudomonas aeruginosa ...

Resistance emerging after pefloxacin therapy was investigated in an experimental Enterobacter cloacae infection. Mice were inoculated intraperitoneally (mean inoculum, 0.9 x 108 CFU) with one of four strains initially susceptible to quinolones and treated with a single 25-mg/kg dose of pefloxacin. This therapy produced a net decrease of bacterial counts in the peritoneal fluid, but with three of the isolates, posttherapy (PT1) strains emerged with decreased susceptibilities to quinolones (4-to 1,024-fold), to the structurally unrelated antibiotics (4-to 16-fold) chloramphenicol and trimethoprim, and sometimes to tetracycline and ,-lactam compounds. In a second set of experiments, new mice were similarly infected with PT1 strains and treated with up to five 25-mg/kg doses of pefloxacin. Compared with parent isolates, PT1 strains produced similar disease and peritoneal bacterial counts in the control animals. In treated mice, posttherapy (PT2) strains emerged that showed 8-to 64-fold increases in quinolone MICs compared with the PT1 strains inoculated. All PT1 and PT2 strains showed altered outer membrane protein patterns, principally marked by a decreased 37,000-molecular-weight band generally accompanied by an increased 42,000-molecular-weight band. Whole cells from all PT1 and PT2 strains, exposed to ['4C]pefloxacin for 15 to 60 s, bound significantly less radioactivity than the corresponding parent strains. After partial purification, DNA gyrase extracted from the most resistant isolates (one PT1 and the PT2 strains) showed a 100-to 450-fold 50% inhibitory concentration increase for pefloxacin. Altogether, pefloxacin can select in vivo two types of resistant strain, one with only decreased permeability and another with decreased permeability combined with altered DNA gyrase.Compared with nalidixic acid, the more recently developed fluoroquinolones display wider antibacterial spectra, more potent activities, and decreased frequencies of spontaneous bacterial resistance (11,21). These improvements have led to substantial use of newer quinolones in clinical therapeutics. However, the clinician is increasingly concerned by the emergence of resistance during therapy with these drugs (1, 3). Resistance also occurs in various models of experimental infections treated with fluoroquinolones, such as pseudomonal (4) or staphylococcal (15) endocarditis and pseudomonal peritonitis (20). The possible mechanisms for quinolone resistance include decreased outer membrane permeability (10) and altered DNA gyrase (2,3,14), or a combination of both (1,3,12,23).We show here that quinolone resistance can emerge during pefloxacin therapy of murine peritonitis caused by Enterobacter cloacae. We investigated the possible mechanisms of this resistance. MATERIALS AND METHODSBacterial strains. Enterobacter cloacae 218, 219, and 895 were clinical isolates described previously by Marchou et al. (19), and strain 908 was kindly provided by R. Then and P. Angehrn (26). Escherichia coli HB101 (17) was the host strain for pBR322, the plasmid u...