Abstract

Highly selective COX-2 inhibitors ('coxibs') were developed in an attempt to minimise severe gastrointestinal toxicity associated with conventional NSAIDs, without loss of anti-inflammatory efficacy. Five years ago, we concluded that the evidence then available indicated, at most, a small advantage for rofecoxib and celecoxib (the first coxibs to be marketed) over conventional NSAIDs in terms of reduced risk of severe gastrointestinal complications, such as bleeding or perforation of gastric or duodenal ulcers.1 In September 2004, newly released evidence about cardiovascular risk with rofecoxib led Merck Sharp & Dohme to voluntarily withdraw the drug worldwide.2 This has raised questions about the balance of benefit and risk with coxibs in general. Here we discuss the evidence relating to the gastrointestinal and cardiovascular safety of coxibs in patients with osteoarthritis (OA) or rheumatoid arthritis (RA).