Appearance of a metronidazole-resistant Helicobacter pylori strain in an infected-ICR-mouse model and difference in eradication of metronidazole-resistant and -sensitive strains.

We tested whether antibiotic-resistant strains appeared in vivo after the failure of treatment using the Helicobacter pylori-infected euthymic mouse model. The numbers of colonies isolated from 56 ICR mice 2 weeks after 4 days of treatment with metronidazole (3.2, 10, or 32 mg/kg of body weight) or amoxicillin (1, 3.2 or 10 mg/kg), with treatment started 4 days after H. pylori CPY2052 inoculation, were counted, and the isolated strains were tested for their sensitivities to two antibiotics to rule out the presence of antibiotic-resistant strains. One metronidazole-resistant strain was detected in a mouse treated with 10 mg of metronidazole per kg, and the MIC of metronidazole for this strain was 25 microg/ml, compared to a MIC of 1.56 microg/ml for the original strain. However, no resistant strain was detected in the amoxicillin treatment group. After the examination described above, mice challenged with a metronidazole-resistant or -sensitive strain isolated from the stomach of a mouse were treated with metronidazole or amoxicillin. The metronidazole-resistant strain was more difficult to eradicate in vivo than the sensitive strain after treatment with metronidazole but not after treatment with amoxicillin. Thus, a metronidazole-resistant H. pylori strain was selected by insufficient treatment, but no resistant strain was selected with amoxicillin. Eradication of a metronidazole-resistant H. pylori strain in vivo required a higher dosage than eradication of a metronidazole-sensitive H. pylori strain. These results may explain one of the reasons for H. pylori treatment failure.