A biochemical basis for the gastrointestinal toxicity of non-steroid antirheumatoid drugs.

The non-steroid anti-inflammatory drugs inhibit cyclo-oxygenase and therefore the biosynthesis of prostaglandins, and it has been proposed that this biochemical intervention is the basis not only of their therapeutic activity but also contributes to their gastrointestinal side-effects. Since endogenous prostaglandins may exert mucosal protective actions, inhibition of cyclo-oxygenase by aspirin-like drugs would be expected to lead to gastric damage. The clinically-used anti-inflammatory compounds reduce prostaglandin levels in the experimental inflammatory exudate, but also inhibit prostacyclin formation in gastric tissue and induce gastric damage. In contrast, anti-inflammatory doses of sodium salicylate and an experimental drug BW755C (3-amino-1m-(trifluoromethyl)-phenyl-2-pyrazoline) fail to inhibit gastric mucosal prostacyclin formation (but do reduce prostaglandin levels in the inflammatory exudate) and cause only minimal gastric damage, supporting the relationship between production of gastric erosions and the inhibition of mucosal prostacyclin production. Anti-inflammatory doses of BW755C also fail to inhibit prostaglandin formation in the small intestine. Furthermore, studies on homogenates of gastric mucosa and ileum show that sodium salicylate, paracetamol and BW755C are only weak inhibitors of gastric mucosal cyclo-oxygenase in vitro. The development of nonsteroid anti-rheumatic drugs which have minimal action on prostaglandin production by the gastro-intestinal tract and exhibit reduced gastrointestinal toxicity, thus appears to be a feasible proposition.