Prevalence and demographic determinants of metronidazole resistance by Helicobacter pylori in a large cosmopolitan cohort of Australian dyspeptic patients

Background: The pre-treatment sensitivity of Helicobacter pylori to metroni dazole is a key determinant of successful eradication therapy and should in fluence local choice of therapy. However, there are few data defining the p revalence of metronidazole resistance (MR) in Australia. Aim: To determine prospectively the prevalence and demographic determinants of MR in H. pylori isolates from a large and cosmopolitan cohort of dyspep tic patients in Sydney. Methods: Consecutive dyspeptic patients undergoing endoscopy had gastric bi opsies for histology, urease test and culture. Metronidazole resistance was determined by E-test after subculture. An MIC >8 mu g/mL defined MR. Patie nt age, gender, birthplace and history of previous nitroimidazole use were recorded. Results: In 732 patients, H. pylori was present in 46.4%. Culture was succe ssful in 81% and subculture for MR in 88% of these. In 237 evaluable patien ts the overall MR rate was 59.1%. Five patients had had prior triple therap y for H. pylori (of which four of five had MR). Therefore, the primary MR r ate in the study population was 58.6% (136/232). MR was more prevalent in y ounger patients (p=0.0002). The MR rate was 70.4% in patients 18-39 years, 66.7% in those aged 40-59 years and lowest (38.9%) in those 60 years or old er (p=0.002). The MR rate was highest in patients born in Southeast Asia (7 2.8%, 59/81) and significantly higher than in Australian born (48.1%, 26/54 ), or Southern European (46.2%, 24/52) born patients (p=0.002). There was n o gender difference. Logistic regression to determine the impact of each va riable (birthplace, age and gender) on MR identified Southeast Asia birthpl ace as a factor associated with greater likelihood of harbouring an MR isol ate (OR 1.88, p=0.02). Southern European born patients had the lowest risk of MR (OR 0.70, p=0.02) as did patients older than 60 years (OR 0.56, p=0.0 4). A definite history of prior metronidazole use was infrequent and not pr edictive of MR. Conclusions: While a high rate of MR is not unexpected in patients born in developing countries, the high rate in Australian born patients is surprisi ng and of concern. This may relate to the high local usage of nitroimidazol es as monotherapy and has important implications for the effectiveness of m etronidazole containing triple therapies.