Our objective was to determine the efficacy and safety of valdecoxib (a cyclo-oxygenase 2 inhibitor) in the treatment of arthritis. Randomised, controlled trials comparing 10 or 20 mg valdecoxib with placebo or non-steroidal anti-inflammatory drugs (NSAIDs) in patients with active osteoarthritis or rheumatoid arthritis. The manufacturer provided clinical trial reports. Data were combined through meta-analysis. Main outcomes were patient global rating of arthritis, arthritis pain, Western Ontario and McMaster Universities indices for osteoarthritis, American College of Rheumatology indices for rheumatoid arthritis, discontinuation, endoscopic ulcers, clinically significant upper gastrointestinal or renal events. Nine trials (five in osteoarthritis, four in rheumatoid arthritis) were included with 5726 patients. Overall, valdecoxib 10 and 20 mg were superior to placebo and equivalent in efficacy to maximum daily doses of NSAIDs. Significantly fewer discontinuations because of gastrointestinal adverse events (4% versus 8%), or endoscopic ulcers of 3 mm or more (5% versus 13%) occurred with valdecoxib compared with NSAIDs. Clinically significant upper gastrointestinal events occurred in 2/2733 (0.1%) with valdecoxib compared with 8/1846 (0.4%) with NSAIDs. Rates of clinically significant renal events were the same (2-3%) for valdecoxib and NSAIDs. At an appropriate dose valdecoxib was as effective as NSAIDs in osteoarthritis and rheumatoid arthritis. There were fewer gastrointestinal adverse event withdrawals and endoscopically detected ulcers. Convincing evidence of reduced major gastrointestinal adverse events could not be addressed by the trials. (C) 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.
