Systematic review: ulcer definition in NSAID ulcer prevention trials

BACKGROUND: In clinical trials of peptic ulcer prevention, the most appropriate definition of an ulcer remains challenging. AIMS: To evaluate the ulcer definitions used in clinical trials of ulcer prevention among non-steroidal anti-inflammatory drug users and to determine whether any specific definition is preferred. METHODS: A systematic literature search of the PubMed, Medline and EMBASE databases was conducted. Results were limited to full papers published in English from June 1987 to June 2007 that met the following criteria: randomized, controlled non-steroidal anti-inflammatory drug trials of > or =8 weeks' duration, with a primary end point of ulcer upon endoscopy. RESULTS: Forty five publications met the inclusion criteria and were reviewed. Overall, an ulcer diameter of > or =3 mm was used in 25 publications and most included a description of ulcer depth. Of the remainder, ulcer was defined as any lesion with unequivocal/observable depth (with no lower limit for ulcer diameter; five publications) or an excavated mucosal break >3 mm (one publication), whereas nine defined a minimum ulcer size of > or =5 or >5 mm. Ulcer definition was unclear in the remaining five publications. CONCLUSION: In clinical trials of ulcer prevention among non-steroidal anti-inflammatory drug users, a gastric or duodenal lesion > or =3 mm in diameter with significant depth is the preferred definition.