A randomized trial of triple therapy for pediatric Helicobacter pylori infection and risk factors for treatment failure in a population with a high prevalence of infection.

BACKGROUND Few trials of treatment for Helicobacter pylori infection have been conducted in high-prevalence or pediatric populations, and risk factors for treatment failure are poorly understood. METHODS As part of a study evaluating the effect of H. pylori therapy on iron deficiency, we conducted a household-randomized, open-label treatment trial involving children aged 7-11 years in 10 villages in western Alaska. We screened 690 children, of whom 219 with iron deficiency and H. pylori infection (determined on the basis of positive results of the 13C urea breath test) were enrolled in the treatment phase of the study. These 219 children received treatment with iron sulfate alone (the control group) or with iron sulfate combined with a 2-week course of lansoprazole, clarithromycin, and amoxicillin (the intervention group). Children in the intervention group who were allergic to amoxicillin or macrolides received metronidazole. Children in the intervention group who did not respond to treatment were re-treated with a 2-week course of metronidazole-based quadruple therapy. RESULTS Two months after initiating therapy, 34% of 104 children in the intervention group and 0.90% of 111 children in the control group tested negative for H. pylori. Among children in the intervention group, risk factors for treatment failure were lack of metronidazole (adjusted odds ratio [aOR], 145), fewer treatment doses (aOR, 0.74), larger household population (aOR, 1.5), and lower body mass index (aOR, 0.69). These 4 variables predicted most of the variation in H. pylori infection status. Among 50 children who were re-treated, 84% tested negative for H. pylori at the 8-month follow-up visit, including those with poor treatment compliance. CONCLUSIONS Among disadvantaged populations with a high prevalence of H. pylori infection, the response to standard treatment regimens may be low. Treatment compliance, household crowding, and re-treatment may influence treatment success. Metronidazole may be appropriate first-line therapy.