Beyond Maastricht IV: are standard empiric triple therapies for Helicobacter pylori still useful in a South-European country?

BACKGROUND Empiric triple treatments for Helicobacter pylori (H. pylori) are increasingly unsuccessful. We evaluated factors associated with failure of these treatments in the central region of Portugal. METHODS This single-center, prospective study included 154 patients with positive (13)C-urea breath test (UBT). Patients with no previous H. pylori treatments (Group A, n = 103) received pantoprazole 40 mg 2×/day, amoxicillin 1000 mg 12/12 h and clarithromycin (CLARI) 500 mg 12/12 h, for 14 days. Patients with previous failed treatments (Group B, n = 51) and no history of levofloxacin (LVX) consumption were prescribed pantoprazole 40 mg 2×/day, amoxicillin 1000 mg 12/12 h and LVX 250 mg 12/12 h, for 10 days. H. pylori eradication was assessed by UBT 6-10 weeks after treatment. Compliance and adverse events were assessed by verbal and written questionnaires. Risk factors for eradication failure were determined by multivariate analysis. RESULTS Intention-to-treat and per-protocol eradication rates were Group A: 68.9% (95% CI: 59.4-77.1%) and 68.8% (95% CI: 58.9-77.2%); Group B: 52.9% (95% CI: 39.5-66%) and 55.1% (95% CI: 41.3-68.2%), with 43.7% of Group A and 31.4% of Group B reporting adverse events. Main risk factors for failure were H. pylori resistance to CLARI and LVX in Groups A and B, respectively. Another independent risk factor in Group A was history of frequent infections (OR = 4.24; 95% CI 1.04-17.24). For patients with no H. pylori resistance to CLARI, a history of frequent infections (OR = 4.76; 95% CI 1.24-18.27) and active tobacco consumption (OR = 5.25; 95% CI 1.22-22.69) were also associated with eradication failure. CONCLUSIONS Empiric first and second-line triple treatments have unacceptable eradication rates in the central region of Portugal and cannot be used, according to Maastricht recommendations. Even for cases with no H. pylori resistance to the used antibiotics, results were unacceptable and, at least for CLARI, are influenced by history of frequent infections and tobacco consumption.