RESUMO
BACKGROUND: The main antibiotics used against Helicobacter pylori have been chosen empirically over time, with few preclinical studies to provide support. The rise in resistance to some of these antibiotics is prompting a reassessment of their use. This work aimed to evaluate the in vitro efficacy of 2 × 2 combinations of the most widely used antibiotics against H. pylori. MATERIALS AND METHODS: J99 reference strains and 19 clinical isolates of H. pylori with various antibiotic resistance phenotypes were used. Minimum inhibitory concentrations were carried out using the microdilution method in 96-well plates. The activity of 15 possible combinations of two antibiotics including amoxicillin, clarithromycin (CLA), levofloxacin, rifampicin, tetracycline, and metronidazole was determined for all strains by the checkerboard method. A mean fractional inhibitory concentration index (FICmean) was calculated for each combination and strain and the type of pharmacodynamic interaction was considered as synergic if FICmean ≤ 0.5, additive if 0.5 < FICmean ≤ 1, indifferent if 1 < FICmean < 4 or antagonistic if FICmean ≥ 4. RESULTS: Most of the 285 pharmacodynamic interactions tested with clinical strains were close to additivity (average FICmean = 0.89 [0.38-1.28]). No interaction was found to be antagonistic. When two antibiotics to which a strain was resistant were combined, the concentrations required to inhibit bacterial growth were higher than their respective breakpoints. CONCLUSION: The present results have shown that in vitro, the different antibiotics used in therapeutics have additive effects. The addition of the effects of two antibiotics to which a strain was resistant was not sufficient to inhibit bacterial growth. In probabilistic treatment, the choice of antibiotics to combine should therefore be based on the local epidemiology of resistance, and on susceptibility testing in the case of CLA therapy, so that at least one antibiotic to which the strain is susceptible is used.