PREVENTION OF DUODENAL-ULCER RECURRENCE WITH PENICILLIN

The relationship between Helicobacter pylori and duodenal ulcer has be en interpreted as causal, but existing evidence has been criticized fo r methodologic reasons. We conducted a randomized trial of phenoxymeth ylpenicillin, 2.4 g twice daily, and placebo to determine whether an a ntibiotic, which has no effect on the integrity of the epithelial cell s, prevents recurrence of duodenal ulcer. Patients with an active duod enal ulcer and positive H. pylori culture, assessed from antral biopsy specimens, were treated with omeprazole, 40 mg once daily, for a tota l of 4 weeks. After 2 weeks' treatment with omeprazole, these patients were randomized to receive phenoxymethylpenicillin (n = 85) or placeb o (n = 85). In patients whose ulcer had healed after 4 weeks' omeprazo le therapy, phenoxymethylpenicillin or placebo was continued for an ad ditional 12 weeks. Patients without ulcer recurrence during this treat ment period were followed up for a further 6 months. Endoscopy was per formed and H. pylori culture assessed at the end of the first treatmen t period (14 weeks), at the end of the follow-up period (a further 6 m onths), and immediately if ulcer symptoms recurred. Five out of 58 (9% ) patients receiving phenoxymethylpenicillin had duodenal ulcer recurr ence during the treatment period, compared with 34 out of 68 (50%) pat ients taking placebo (p < 0.0001; log-rank test) (Fig. 1). [GRAPHICS] The ulcer recurrence rate in the phenoxymethylpenicillin-treated group remained low for another 5 months after penicillin treatment but appr oached the rate seen in the placebo group during the sixth month of th e follow-up period. The prevalence of H. pylori-positive patients afte r 14 weeks of penicillin administration and at the end of the follow-u p period was 47% and 80%, respectively, compared with 86% and 90%, res pectively, in the placebo group. These data provide strong evidence th at duodenal ulcer has an underlying bacterial component, with H. pylor i being the likely cause. Pronounced inhibition of H. pylori growth wa s obtained with omeprazole alone. This inhibition was increased by the addition of phenoxymethylpenicillin, but eradication was generally no t obtained.