Gastrointestinal safety of cyclooxygenase-2 inhibitors: a Cochrane Collaboration systematic review

Nonselective non-steroidal anti-inflammatory drugs (NSAIDs) and cyclooxygenase-2 inhibitors (COX-2s) are used to treat a variety of arthritic and inflammatory conditions. The aim of this study was to assess the upper gastrointestinal (GI) harms of the long-term use of COX-2s, compared with nonselective NSAIDs and placebo, in arthritis sufferers.

A systematic review of randomized controlled trials (RCTs) was conducted. Searches were conducted in (1) Cochrane Central Register of Controlled Trials (CENTRAL), (2) the Cochrane Collaboration Library (2005), (3) MEDLINE (to December 2006), and (4) Excerpta Medica Database (EMBASE) (to June 2005). Reference lists from trials and abstracts of conference proceedings were searched by hand, and experts were contacted to identify further relevant trials. RCTs of celecoxib, rofecoxib, etoricoxib, valdecoxib, and lumiracoxib were included if they reported on endoscopic ulcers, clinically important ulcer complications, or adverse gastrointestinal (GI) symptoms with the use of these COX-2s, compared with placebo or with nonselective NSAIDs. Study selection and data extraction were performed in duplicate by independent reviewers. Data were analyzed by using Review Manager 4.2 in accordance with accepted meta-analysis techniques.

Compared with nonselective NSAIDs, COX-2s produced significantly fewer gastroduodenal ulcers (relative risk, 0.26; 95% confidence interval, 0.23-0.30) and clinically important ulcer complications (relative risk, 0.39; 95% confidence interval, 0.31-0.50), as well as fewer treatment withdrawals caused by GI symptoms. The co-administration of acetylsalicylic acid appears to reduce the GI safety of COX-2s in subgroup analyses.

COX-2s appear to offer greater upper GI safety and are better tolerated than nonselective NSAIDs. The co-administration of acetylsalicylic acid might reduce the safety advantage of COX-2s over that of nonselective NSAIDs.