Does treatment of Helicobacter pylori with antibiotics alone heal duodenal ulcer? A randomised double blind placebo controlled study.

BACKGROUND Treatment of Helicobacter pylori infection prevents duodenal ulcer relapse. It has not been established if treatment of the infection heals duodenal ulcer. AIM To test the hypothesis that treatment of the infection was associated with healing of duodenal ulcer. METHODS A randomised, double blind placebo controlled trial was performed to study the efficacy of an antibiotic only regimen consisting of 300 mg metronidazole, 500 mg amoxycillin, and 250 mg clarithromycin, each given four times daily for two weeks, in the healing of duodenal ulcer as assessed by endoscopy. Symptoms were controlled with acetaminophen and antacids. RESULTS Of 100 consecutive patients with endoscopically established duodenal ulcer, 97 with positive rapid urease test on antral biopsy specimens were admitted into the study and 81 completed the trial. Of these, 40 were randomised to receive antibiotics and 41 to receive placebo. Treatment with antibiotics resulted in 92.5% (95% confidence interval (95% CI) 84.3-100) healing at four weeks and 100% at eight and 12 weeks; the corresponding healing rates for placebo treatment were respectively, 36.6%, 61%, and 63.4% (95% CIs 21.8-51.3, 46.0-75.9, and 48.7-78.2 respectively). The differences between the two treatment groups were significant at p < 0.001 at each time point and by life table analysis. Clearance of H pylori as assessed by urease test on antral biopsy specimens at four weeks and eradication of the organism as determined by 13C-urea breath test at eight weeks were achieved in 85% and 62.5% of patients respectively. Duodenal ulcer healed at four weeks in 87.2% and 86.2% (95% CIs 76.7-97.7 and 73.7-98.8) of patients in whom H pylori clearance or eradication, was achieved, versus 42.9% and 51.9% (95% CIs 27.9-57.8 and 38.3-65.5; p < 0.001 and < 0.003 respectively) in whom these processes failed. Stepwise discriminant analysis on 32 clinical, personal, and endoscopic characteristics as well as H pylori clearance and eradication identified H pylori clearance as the most discriminative variable for the healing of duodenal ulcer at four weeks, followed by ulcer depth and eradication of the organism. CONCLUSIONS Treatment with an antibiotic only regimen was effective for the healing of duodenal ulcer, and clearance as well as eradication of H pylori contributed significantly to the healing. The results constituted the strongest evidence to date that H pylori infection was aetiologically related to duodenal ulceration, and support the concept of treating duodenal ulcer associated with H pylori as an infection and relieving its symptoms with acid reducing agents such as antacids.