Risks related to lack of standardization of tests to detect in vitro metronidazole resistance inHelicobacter pylori

Abstract

The correlation between metronidazole MIC readings obtained in vitro on two different media for two different inoculum dilutions was studied after two, three, and five days of incubation using 20 clinical isolates ofHelicobacter pylori. The PDM epsilometer (E test) was used. After two days of incubation, log MIC values on chocolate PDM and Vestfold charcoal medium, a new charcoal medium, showed good correlation. Charcoal media can thus be used for metronidazole sensitivity tests. In terms of double-dilution gradients, prolongation of incubation time from two to three days and change of inoculum dilution from 0.5 to 4 McFarland had little impact when studied separately. However, the combined effect of the two variables was marked. The matching of a sparse inoculum incubated for two days against a heavy inoculum incubated for three days resulted in readings which, for 30% of the pairs, differed by two double-dilution gradients or more. On the other hand, the separate influence of both variables was found to affect the correlation coefficients significantly. Five days of incubation was associated with a poor correlation when the variables were studied both separately and combined. Thus, the real influence of separate variables on MIC readings was not exposed in a scale of double-dilution gradients because the intervals were too large. For this reason, scales without thresholds are recommended.