Review: acute human models of gastric mucosal injury

The use of acute human models of gastric mucosal injury has been stimulated by a need to understand more fully the problems of non-steroidal anti-inflammatory drugs but such models have other applications. None is ideal and they all share certain drawbacks. For none of them has a precise relationship to clinical events been established and they have all tended to be employed on a population of young healthy subjects who are not those at greatest clinical risk. Of individual methods mucosal potential difference is an indirect measure which is too often affected by other influences to be acceptable as a measure of mucosal injury when used alone, although it has some value as an adjunct to other measurements. Assay of DNA in gastric washings is a suitable technique for quantifying desquamation of gastric epithelial cells occurring in response to acute injury; on present evidence its significance is much more difficult to assess in the context of continuing challenge over several days. By contrast, measurement of microbleeding is more suitable for quantifying injury over several days of NSAID ingestion; little bleeding is recorded with a single acute challenge. Endoscopy can demonstrate macroscopic lesions which result from mucosal injury--injury which is quantified more easily and sensitively by measurements of cellular exfoliation or bleeding. Paradoxically, endoscopy's strength has been to underline the scientific weakness of acute models because it shows that it is rare for ulcers, which are the lesions of clinical concern to develop in these studies.