HELICOBACTER-PYLORI - IN-VITRO INDUCTION OF RESISTANCE TO AZITHROMYCIN

Helicobacter pylori resistance to macrolides is possibly an important factor for the failure of macrolide therapy for H. pylori infection. T he aim of this study was to assess the propensity of H. pylori to deve lop in vitro resistance to azithromycin. In 73 clinical isolates taken from patients before starting antimicrobial therapy of H. pylori infe ction, MIC was determined using an agar dilution method (Muller-Hinton agar with 7.5% unlysed horse blood, pH = 7.2, at 35 degrees C, during 72 h in a humid microaerobic atmosphere). Each strain was first culti vated at half minimal inhibitory concentration (MIC) then in doubling concentrations until growth arrest. All experiments for induction of r esistance were performed on the same media, incubation temperature, at mosphere and time of MIC determination. MIC interpretative standards f or sensitivity, intermediate sensitivity and resistance of H. pylori t o azithromycin were less than or equal to 2, 4 and greater than or equ al to 8 mg/l, respectively. Of 73 strains, 5 died during the experimen ts, and in the remaining 68 strains, serial passage with increasing az ithromycin concentrations resulted in the development of resistance in 19 (26.9%) strains. Two strains had an MIC of 16 mg/l azithromycin. T hirty-three (48.5%) strains kept the same MIC or doubled their MIC, 16 (23.5%) strains had 4- to 16-fold MIC but still remained sensitive, 2 resistant strains had 128-fold MICs and 17 resistant strains had incr eased their MICs more than 128 times. Seventeen highly resistant strai ns (MIC >128 mg/l) were kept frozen at -70 degrees C for 3 months in a brain-heart infusion broth with 15% glycerol. MIC was assessed again to determine the stability of resistance. Eleven strains kept MICs >12 8 mg/l, 2 became sensitive and 1 intermediate, but reverted easily, af ter only 2 passages, to an MIC of >128 mg/l azithromycin. Although mac rolides are very active against H. pylori, the propensity to develop r esistance in a high proportion of strains has a clear impact on the ch oice of the right combinations of macrolides with other agents as well as the dosage of the macrolide antibiotics.