Abstract

Background and aims: A past history of peptic ulceration increases the risk of an ulcer developing during non-steroidal anti-inflammatory drug (NSAID) use. Whether this is due to Helicobacter pylori infection or to reactivation of the original lesion is unclear.

Methods: We used multivariate regression analyses of three large similar trials to identify factors that placed patients at high risk of ulcer development or relapse. We compared the efficacy of omeprazole 20 mg daily, misoprostol 200 μg twice daily, and ranitidine 150 mg twice daily in preventing ulcers and erosions at different sites and in patients who were H pylori positive and negative.

Results: Patients with endoscopic lesions (which healed) initially were significantly more likely than those without to develop further erosions or ulcers during treatment (rate ratio 2.12, 1.07–4.17). Risk mounted further with ulcers versus erosions, particularly those that had been slow to heal. There was a highly significant tendency for the relapse lesion to replicate the site and type of the original lesion (mean odds ratios ranging from 3 to 14). Treatment failure was significantly less likely with omeprazole than with placebo, misoprostol, or ranitidine. This advantage was especially evident in H pylori positive patients receiving acid suppression (5.7% v 16.6% for gastric ulcer with omeprazole).

Conclusions: Relapse of lesions in patients taking NSAIDs was highly site and type specific and not adversely affected by H pylori status. This strongly implies that local mucosal factors predispose to ulcer development in patients taking NSAIDs. Identification of the responsible mucosal changes would aid understanding and could promote better treatment.