Outcome trials of COX-2 selective inhibitors: global safety evaluation does not promise benefits

Gastrointestinal toxicity is the most frequent adverse effect associated with nonsteroidal anti-inflammatory drug use. The most clinically relevant side effects of this toxicity are ulcer complications, including perforation, obstruction, or bleeding. Selective cyclooxygenase (COX-2) inhibitors (coxibs) have been proposed as a safer alternative to traditional, nonsteroidal anti-inflammatory drugs and they are currently widely used in clinical practice. The aim of this review was to analyze the available evidence and then critically evaluate the outcome trials supporting the use of coxibs in terms of their clinical gastrointestinal benefits and global safety.

All published clinical trials on selective COX-2 inhibitors were identified by searching Medline, the World Wide Web (WWW), and abstracts in Congress proceedings. From these, we selected randomized trials that clinically evaluated relevant safety outcome measures. Papers only describing endoscopic evaluation were excluded.

Our search yielded three outcome trials and two pooled safety analyses. The outcome studies supporting the gastrointestinal and global safety of coxibs were found to be biased in their design, analysis, and dissemination, and interpretation of a clinical benefit. Cost considerations would make the use of coxibs acceptable only in patients at high gastrointestinal risk.

The association of the reduced gastroerosive potential of coxibs with improved meaningful outcomes is debatable. Bias in the design of the trials, selection of outcome measures, post-hoc changes in analysis and the variables used, as well as flaws in the publication and reporting of trial results cast serious doubts on the gastrointestinal and global safety profile of coxibs. In addition, their high cost and the lack of clear identification of patients that would benefit most from treatment means the effectiveness of these drugs is uncertain at the moment.