In vivo pharmacodynamic efficacy of gatifloxacin against Streptococcus pneumoniae in an experimental model of pneumonia: impact of the low levels of fluoroquinolone resistance on the enrichment of resistant mutants.

Abstract

To investigate the impact of low levels of fluoroquinolone resistance on the emergence of resistant mutants, we examined the mutant selection window (MSW) hypothesis in experimental pneumonia in rabbits infected with pneumococci with various susceptibility levels to fluoroquinolones and treated with gatifloxacin using a human-like regimen (equivalent to 400 mg once daily). The MSW corresponds to the range of concentrations between the minimal inhibitory concentration (MIC) and the mutant prevention concentration (MPC), which is the antibiotic concentration that prevents selection of resistant mutants. Five pneumococcal strains were tested and were defined as follows [MIC of ciprofloxacin (mg/L)/MIC of gatifloxacin (mg/L)/MPC of gatifloxacin (mg/L)/involved quinolone resistance mechanisms]: strain 16089=0.5/0.25/0.25/wild-type; strain MS1A=2/0.5/1/efflux; strain MS2A=8/1/8/parC S79F; strain MR3B4=10/1/8/parC S79T; strain Gyr-1207=6/4/4/gyrA S81F. A 48 h human-like treatment with gatifloxacin was significantly bactericidal on pneumonia induced by strain 16089 ( > 6 log(10) killing) as well as the efflux derivative strain MS1A ( > 5 log(10) killing). However, a small number of parC-gyrA mutants were recovered in 26% of the animals infected with this efflux strain. As expected, no decrease in viable bacteria counts was observed when pneumonia was induced by the gyrA resistant strain. In contrast, because of the enrichment of highly resistant mutants in 100% of the animals, no significant bacterial reduction was observed after treatment of pneumonia induced by the two susceptible parC mutated strains. A classification and regression tree (CART) analysis identified T(MSW) (percentage of the time during which gatifloxacin serum concentrations are inside the MSW) and AUC(MSW) (area under curve between MIC and MPC values) as the best parameters associated with the enrichment of resistant pneumococci. This study shows that the acquisition of a low level of fluoroquinolone resistance (especially a parC mutation and to a lesser extent an efflux mechanism) is associated with a clearly lower potential for preventing resistance development. These data support the concept that resistant mutants are selectively enriched when antibiotic concentrations fall inside the mutant selection window and suggest that in vivo dynamic models have to be used to predict the relative abilities of quinolones to prevent mutant selection.