Meta-analysis of randomized clinical trials comparing lansoprazole with ranitidine or famotidine in the treatment of acute duodenal ulcer.

OBJECTIVE The aim was to compare the clinical efficacy of lansoprazole with the efficacies of ranitidine and famotidine in order to rank this drug in the hierarchy of duodenal ulcer treatments. METHODS All randomized clinical trials in which lansoprazole was used to treat patients with duodenal ulceration were reviewed. The meta-analysis included a quality assessment for each trial. The main criterion chosen for the meta-analysis was the endoscopic healing rate at 4 weeks; other criteria were the healing rate at 2 weeks and the absence of pain at 2 and 4 weeks. Statistical methods used were the Der Simonian and Laird method and the method of Peto et al. A sensitivity analysis was performed according to the H2 blocker type (famotidine or ranitidine). Odds ratios against histamine-receptor blockers were used to compare lansoprazole indirectly with omeprazole and other drugs. RESULTS Five double-blind trials were identified including 848 patients. The mean 4-week healing rate in patients treated with lansoprazole was 85%, which was significantly higher than the healing rate in patients treated with H2-receptor blockers (75%). This difference of 10% was significant (P < 0.01) according to the Der Simonian method, and the corresponding odds ratio of 2.27 (95% confidence interval 1.5-3.2) was significant according to the Peto method (P < 0.01). There was also a significant difference in favour of lansoprazole for 2-week healing rates (mean difference 20%, P < 0.01) and for the percentage of patients without pain at 2 weeks (mean difference 8%; P < 0.02). Indirect comparisons of 4-week healing rates showed no difference between 30 mg lansoprazole and 20 mg omeprazole and confirmed that both drugs had a greater efficacy than ranitidine, famotidine, nizitidine, cimetidine or sucralfate. CONCLUSION This meta-analysis showed that 30 mg lansoprazole was more effective in producing healing at 2 and 4 weeks than ranitidine or famotidine. Lansoprazole also led to a greater reduction in the percentage of patients free of pain at 2 weeks. The efficacy of lansoprazole was not different from that of omeprazole.