Induction of in vitro resistance to BMS-284756 by Streptococcus pneumoniae.

Sir, BMS-284756 (T-3811) is a novel quinolone that lacks a fluorine at the C-6 position, and is active in vitro against Streptococcus pneumoniae, including strains with elevated ciprofloxacin MICs. Because the various quinolones differ in both target affinity and their activation of efflux pumps, one can speculate that the phenotypic expression of quinolone resistance will also differ. Studies have shown that in vitro fluoroquinolone resistance can be selected for in pneumococci, a property that may have clinical significance. The purpose of the present study was to clarify whether a quinolone that lacks a fluorine at the C-6 position would have faster or slower development of resistance compared with other fluoroquinolones. Previous authors have demonstrated that quinolones with a methoxy moiety at the C-8 position (e.g. gatifloxacin and moxifloxacin) have an increased antibacterial activity, especially against first-step gyraseand topoisomerase IV-resistant mutants. Recently, Drlica and colleagues have reported that compounds with a C-8 methoxy moiety differ in their ability to restrict the selection of resistant mutants. Quinolones with a C-8 methoxy group were shown to have an enhanced ability to block mutant growth and to kill mutant cells in Escherichia coli, Staphylococcus aureus and Mycobacterium spp., and fewer resistant mutants are selected when bacteria are challenged with C-8 methoxy quinolones than with C-8 hydrogen derivatives. In this study, 50 distinct clinical isolates of S. pneumoniae were repeatedly exposed to BMS-284756, ciprofloxacin, gatifloxacin, levofloxacin and gemifloxacin during a 6 day period. Mean MICs for the original isolates were as follows (mg/L): gemifloxacin, 0.014; BMS-284756, 0.030; gatifloxacin, 0.122; levofloxacin, 0.566; and ciprofloxacin 0.590. Approximately 5 10 cfu of each of the 50 strains were added to tubes containing 9.9 mL brain–heart infusion (BHI) broth supplemented with 5% bovine serum and two-fold dilutions of the quinolones (range 0.001–64 mg/L) (in analogy to MIC determination). The tubes were then incubated for 24 h at 37 C. Aliquots from those tubes containing the highest drug concentration that still permitted visible growth (i.e. 0.5 MIC) were used, following a 1:10 dilution, to inoculate a second set of serial drug dilutions. After overnight incubation, the bacteria were transferred again. Finally, after six serial transfers, the bacteria with the highest MICs were collected, stored and subcultured on quinolone-free agar for 10 days to assess the stability of resistance. S. pneumoniae MICs increased from 0.030 to 0.142 mg/L in BMS-284756-containing medium, from 0.014 to 0.250 mg/L in gemifloxacin-containing medium, from 0.122 to 0.574 mg/L in gatifloxacin-containing medium, from 0.566 to 4.724 mg/L in levofloxacin-containing medium and from 0.590 to 15.562 mg/L in ciprofloxacin-containing medium. Based on a breakpoint of 1 mg/L ( 2 mg/L for levofloxacin), 100% of the 300 selected mutants (50 strains incubated over 6 days) were inhibited by BMS-284756, 100% were inhibited by gemifloxacin, 98% by gatifloxacin, 70% by levofloxacin and 22% by ciprofloxacin. In order to analyse the rate of resistance development, we converted the MIC values of all clinical strains and selected mutant isolates to a log scale (base 2). Initial values at time zero (original MIC values) were subtracted from subsequent MICs on days 1–6. The slope of the MIC increase over the 6 day period was computed by linear regression and compared using ANOVA with the Student–Newman–Keuls post hoc test; comparisons were regarded as statistically significant if P 0.05. Judging from the squared correlation coefficients R (mean S.E.M.: BMS-284756, 0.766 0.019; ciprofloxacin, 0.922 0.008; gatifloxacin, 0.764 0.019; gemifloxacin, 0.931 0.009; Correspondence Journal of Antimicrobial Chemotherapy (2001) 48, 587–595