Treatment of patients with osteoarthritis with rofecoxib compared with nabumetone.

Background: Rofecoxib and nabumetone were developed to provide gastrointestinal benefits over traditional nonsteroidal antiinflammatory drugs (NSAIDs). However, there is limited comparative information relating to these 2 drugs.

Objective: The objective of this study was to compare rofecoxib and nabumetone, at their lower, recommended doses, in patients with osteoarthritis (OA).

Methods: Nine hundred seventy-eight patients with knee OA and a positive history of NSAID response were randomized to 12.5 mg rofecoxib per day (N=390), nabumetone 500 mg twice a day (N=392), or placebo (N=196) for 6 weeks. The primary efficacy end point was percent of patients with a "good" or "excellent" Patient Global Assessment of Response to Therapy (PGART) at week 6; PGART was also evaluated over days 1 to 6. Additional end points included investigator assessment of response, pain walking over 6 days and 6 weeks, joint tenderness, discontinuation as a result of lack of efficacy, and quality of life. Adverse experiences (AEs) were collected.

Results: Significantly more rofecoxib (50.4%) than nabumetone (43.3%, P=0.043) or placebo (29.5%, P<0.001) patients had a good or excellent PGART at week 6. Median time to a good or excellent PGART was significantly shorter with rofecoxib (52 hours) than nabumetone (100 hours, P=0.001) or placebo (>124 hours, P<0.001). Results for rofecoxib and nabumetone were similar in all additional end points except pain in walking over 6 days and 6 weeks, in both of which the rofecoxib treatment group demonstrated better results. There were significantly (P<0.050) more overall and serious AEs and discontinuations resulting from AEs with rofecoxib than nabumetone. Five rofecoxib and one nabumetone patients had confirmed thrombotic cardiovascular events (P=0.123). Information on thrombotic cardiovascular events from this study was included in a published, prespecified pooled analysis and is included here for completeness.

Conclusions: At their recommended starting doses for OA, both agents were more effective than placebo. Rofecoxib at a dosage of 12.5 mg demonstrated significantly better efficacy in PGART than 1000 mg nabumetone in these patients known to be NSAID responders. Significantly more AEs occurred with rofecoxib than nabumetone. Considering these data and other recent safety information regarding cyclooxygenase-2 selective and nonselective NSAIDS, physicians must make risk/benefit assessments for each individual patient when considering the use of these agents, as recommended by the U.S. Food and Drug Administration.