Current issues of ACP Journal Club are published in Annals of Internal Medicine


Amoxicillin plus metronidazole eradicated Helicobacter pylori and reduced recurrence of duodenal ulcer

ACP J Club. 1993 May-June;118:69. doi:10.7326/ACPJC-1993-118-3-069

Source Citation

Hentschel E, Brandstätter G, Dragosics B, et al. Effect of ranitidine and amoxicillin plus metronidazole on the eradication of Helicobacter pylori and the recurrence of duodenal ulcer. N Engl J Med. 1993 Feb 4;328:308-12.



To compare duodenal ulcer recurrence in patients with Helicobacter pylori infection who were treated either with ranitidine and an antimicrobial regimen without bismuth or with ranitidine alone.


1-year, randomized, double-blind, placebo-controlled trial.


Hospital outpatient clinics in Austria.


104 patients (mean age 49 y, 74% men) who had ≥ 2 symptomatic duodenal ulcer recurrences separated by at least 3 months, a current ulcer ≥ 5 mm in diameter, and biopsy-proven H. pylori infection. Exclusion criteria were bleeding ulcers; previous gastric surgery; other serious disease; use of corticosteroids, anti-inflammatory drugs, or antibiotics in the month before the study; lactation; or possibility of pregnancy. Follow-up was 98%.


All patients took ranitidine, 300 mg before bedtime for 6 weeks, and, if the ulcers had not healed, for another 4 weeks. 52 patients were randomized to oral amoxicillin, 750 mg 3 times daily, and oral metronidazole, 500 mg 3 times daily, for 12 days. 52 patients were randomized to placebo.

Main outcome measures

Ulcer recurrence and presence of H. pylori infection (2 positive culture, histologic, or rapid urease tests). Patients were examined and had endoscopies after treatment and 2, 6, and 12 months later.

Main results

At the end of treatment (6 weeks), patients who took antibiotics compared with those who took placebo had a higher rate of ulcer healing (P = 0.01) and a higher rate of eradication of H. pylori (P < 0.001) (Table). The groups did not differ for ulcer healing at 10 weeks (98% vs 94%). During follow-up, patients who took antibiotics had 4 ulcer recurrences (8%) compared with 42 recurrences (86%) in the placebo group (P < 0.001) (Table). 1 patient became positive for H. pylori during follow-up. Side effects occurred in 8 patients (15%) who took antibiotics and in 1 patient (2%) who took placebo {P < 0.05}*.


Patients with recurrent duodenal ulcers and Helicobacter pylori infection, who were treated with ranitidine and an antimicrobial regimen without bismuth and ranitidine, had a lower rate of ulcer recurrence than patients who were treated with ranitidine alone. Recurrence was associated with H. pylori infection.

Source of funding: Biochemie Ltd.

For article reprint: Dr. E. Hentschel, Wiener Gebietskrankenkasse, Hanusch Hospital, 1140 Wien, Heinrich-Collinstrasse 30, A-1140 Vienna, Austria. FAX 43-1-91021-2653.

*Numbers calculated from data in article.

Table. Amoxicillin plus metronidazole vs placebo for duodenal ulcer and H. pylori infection†

Outcomes at 6 wk Amoxicillin plus metronidazole Placebo RBI (95% CI) NNT (CI)
Ulcer healing at 6 wk 92% 75% 23% (4 to 51) 6 (3 to 30)
Eradication of H. pylori 89% 2% 4500% (772 to 25941) 1 (1 to 1)
Outcome at 1 y Amoxicillin plus metronidazole Placebo RRR (CI) NNT (CI)
Ulcer recurrence 8% 86% 91% ( 78 to 96) 1 (1 to 2)

†Abbreviations defined in Glossary; RRR, RBI, NNT, and CI calculated from data in article.


Those skeptical about the causal role of H. pylori in duodenal ulcer disease have argued that the mucosal protective effects of bismuth compounds used in H. pylori eradication regimens are probably responsible for the observed decrease in ulcer recurrence and not the eradication of the organism. The study by Hentschel and colleagues shows that eradication of H. pylori by an antimicrobial regimen that does not include a mucosal protective agent still results in a marked reduction of ulcer recurrences. An equally well-designed study from a second group of investigators confirming the results of this study would lay to rest the argument about the mucosal protective effect.

Because Hentschel and colleagues did not include a bismuth agent in the antimicrobial regimen, a double-blind study design was possible. Studies that have included bismuth compounds in the treatment regimen have typically been single blind and thus have been subject to methodologic criticisms (1). Although the investigation was done in Austria, the study population demographics were similar to those that might have been encountered in the United States; therefore, the results should possess portability.

This study further shows that duodenal ulcer disease can be cured if the index ulcer is caused by H. pylori and the initial healing regimen includes antimicrobial therapy directed against H. pylori. The practicing physician is obligated to determine if H. pylori is present when duodenal ulcer disease is diagnosed and to treat the patient accordingly. Long-term acid suppression therapy for infectious ulcer disease (in the absence of a trial of antimicrobial therapy) can no longer be justified.

James S. Barthel, MD
University of Missouri School of Medicine Columbia, Missouri, USA