RESUMO
BACKGROUND: Different bismuth quadruple therapies containing proton-pump inhibitors, bismuth salts, metronidazole, and a tetracycline have been recommended as third-line Helicobacter pylori eradication treatment after failure with clarithromycin and levofloxacin. AIM: To evaluate the efficacy and safety of third-line treatments with bismuth, metronidazole, and either tetracycline or doxycycline. METHODS: Sub-study with Spanish data of the "European Registry on H pylori Management" (Hp-EuReg), international multicenter prospective non-interventional Registry of the routine clinical practice of gastroenterologists. After previous failure with clarithromycin- and levofloxacin-containing therapies, patients receiving a third-line regimen with 10/14-day bismuth salts, metronidazole, and either tetracycline (BQT-Tet) or doxycycline (BQT-Dox), or single capsule (BQT-three-in-one) were included. Data were registered at AEG-REDCap database. Univariate and multivariate analyses were performed. RESULTS: Four-hundred and fifty-four patients have been treated so far: 85 with BQT-Tet, 94 with BQT-Dox, and 275 with BQT-three-in-one. Average age was 53 years, 68% were women. Overall modified intention-to-treat and per-protocol eradication rates were 81% (BQT-Dox: 65%, BQT-Tet: 76%, BQT-three-in-one: 88%) and 82% (BQT-Dox: 66%, BQT-Tet: 77%, BQT-three-in-one: 88%), respectively. By logistic regression, higher eradication rates were associated with compliance (OR = 2.96; 95% CI = 1.01-8.84) and no prior metronidazole use (OR = 1.96; 95% CI = 1.15-3.33); BQT-three-in-one was superior to BQT-Dox (OR = 4.46; 95% CI = 2.51-8.27), and BQT-Tet was marginally superior to BQT-Dox (OR = 1.67; 95% CI = 0.85-3.29). CONCLUSION: Third-line H pylori eradication with bismuth quadruple treatment (after failure with clarithromycin and levofloxacin) offers acceptable efficacy and safety. Highest efficacy was found in compliant patients and those taking 10-day BQT-three-in-one or 14-day BQT-Tet. Doxycycline seems to be less effective and therefore should not be recommended.