Objective To evaluate the efficacy of omeprazole triple therapy versus omeprazole quadruple therapy for Helicobacter pylori infection. Design Prospective, randomized, single-centre, investigator-blind study. Settings Departments of Gastroenterology and Histopathology, Evangelismos Hospital, Athens, Greece. Methods One hundred and forty-nine consecutive patients with active duodenal ulcer were randomized to receive omeprazole (20 mg b.d.), amoxicillin (1 g b.d.) and clarithromycin (0.5 g b.d.) (OAC(10), n = 78), or omeprazole (20 mg b.d.), colloidal bismuth subcitrate (120 mg q.i.d.), metronidazole (0.5 g t.i.d.) and tetracycline hydrochloride (0.5 g q.i.d.) (OBMT10, n = 71) for 10 days. Patients' symptoms were scored, and compliance and treatment-related side effects were assessed. Endoscopy was performed before treatment and at 10-12 weeks and 12 months after treatment. H. pylori infection and its successful eradication were sought by histology, immunohistochemistry and campylobacter-like organisms (CLO) tests on multiple biopsies taken from the gastric antrum, corpus and fundus. Patients were re-evaluated clinically and underwent a C-13-urea breath test (UBT) at 21-24 months. Those with dyspepsia and/or recrudescence of H. pylori were re-endoscoped. Results Patient groups were comparable for age, sex, smoking, occasional use of nonsteroidal anti-inflammatory drugs (NSAIDs), and current or past bleeding episodes. Six and seven patients in the OAC(10) and OBMT10 treatment groups, respectively, were lost to follow-up. Eight patients were non-compliant. Two ulcers in the OAC(10) group and one in the OBMT10 group did not heal. By intention-to-treat (ITT) and per-protocol (PP) analyses, ulcer healing rates were 86% (67/78) and 97% (67/69), respectively, for the OAC(10) group, and 82% (58/71) and 98% (58/59), respectively, for the OBMT10 group. H. pylori eradication at 10-12 weeks after treatment was 78% (61/78) and 88% (61/69) for OAC(10), and 65% (46/71) and 78% (46/59) for OBMT10, by ITT and PP analyses, respectively (P > 0.1). Side effects were more common with OBMT10. Relapse rates of H. pylori were 3% and 2% for the first and second years, respectively. Four H. pylori-negative patients developed reflux symptoms, but only two developed erosive oesophagitis between 12 and 24 months. Conclusions OAC(10) and OBMT10 were equally effective in healing active duodenal ulcers and eradicating H. pylori, but OAC(10) should be used as a first-line treatment because of its better tolerance.
