Maintenance therapy of duodenal ulcer with H2-receptor antagonists--a meta-analysis.

A theoretical basis for similar recurrence rates among H2-receptor antagonists exists based on recent concepts of ulcer recurrence, ulcer healing and suppression of nocturnal gastric acidity. In order to compare H2-receptor antagonists in the maintenance therapy of duodenal ulcer, a meta-analysis was carried out using 29 studies in the literature that met strict criteria. When the results of the placebo-controlled studies were expressed as odds ratios, a technique used to minimize differences in protocol design and patient populations among studies, cimetidine, ranitidine, famotidine and nizatidine were all found to be superior to placebo to approximately the same extent. Odds ratios (and 95% confidence limits) for the recurrences in the pooled studies were cimetidine 0.22 (0.18-0.28), ranitidine 0.23 (0.18-0.30), famotidine 0.28-0.31 and nizatidine 0.36. These reflected similar 1-year recurrence rates of 24.9% (n = 530) for 400 mg cimetidine nocte, 22.4 (n = 508) for 150 mg ranitidine nocte, 28.0% (n = 371) for 20 mg or 40 mg famotidine nocte, and 21.8% (n = 261) for 150 mg nizatidine nocte. In studies to compare cimetidine and ranitidine directly, the odds ratio (and 95% confidence limits) was 0.64 (0.48-0.86). However, for two studies done by a single protocol, the odds ratio of 0.51 (0.35-0.75) tended to differ from the odds ratio of 0.85 (0.54-1.33) for six other studies (P = 0.09). These reflected recurrence rates for cimetidine and ranitidine of 28.3% and 16.8% (two studies) and 23.3% and 20.6% (six studies) respectively.