Abstract

Background and aims: Factors predisposing to endoscopic ulcer formation or healing with non-steroidal anti-inflammatory drugs (NSAIDs) have not been well defined.

Methods: We used multivariate analysis of data from three large similar trials to identify factors associated with endoscopic lesions and healing. We compared the effectiveness of omeprazole 20 mg and 40 mg daily, misoprostol 200 μg four times daily, and ranitidine 150 mg twice daily in healing ulcers and erosions at different sites and in patients who were Helicobacter pylori positive and negative.

Results: Older age, past ulcer history, rheumatoid arthritis, and H pylori infection were significantly associated with ulcers. Duodenal ulcer was significantly more likely than gastric ulcer with a past ulcer history (odds ratio 1.59, 1.16–2.17), H pylori infection (1.4, 1.04–1.92), and male sex (2.35, 1.75–3.16) while female sex, older age (≥60 years: 1.39, 1.03–1.88), and higher NSAID dose (>1 defined daily dose: 1.57, 1.16–2.14) were associated with gastric ulceration. Sex differences were seen in both H pylori positive and negative patients. Gastric and duodenal ulcer healing was significantly faster with omeprazole 20 mg than with misoprostol 200 μg four times daily or ranitidine 150 mg twice daily although misoprostol was more effective at healing erosions. Gastric ulcer healing was slower with large ulcers (0.37, 0.25–0.54 for >10 mm v 5–10 mm) or a past ulcer history (0.51, 0.34–0.76), and faster with H pylori infection (1.55, 1.06–2.29), especially with acid suppression (72% v 37% at four weeks with ranitidine).

Conclusions: Among NSAID users, H pylori and male sex independently increase the likelihood of duodenal ulceration. H pylori infection does not affect duodenal ulcer healing and enhances gastric ulcer healing by ranitidine and possibly other acid suppressing treatments.