Objective: To compare the lower osteoarthritis (OA) dose of rofecoxib to the recommended dose of celecoxib in two identically designed studies.
Methods: Patients with knee OA were randomized (2:2:1 ratio: rofecoxib 12.5 mg once daily (qd), celecoxib 200 mg qd, or placebo, respectively). The primary endpoint was patient global assessment of response to therapy (PGART) averaged over 6 weeks on a five-point scale. Rofecoxib would be declared at least as effective as celecoxib if the lower bound of the 95% confidence interval (95% CI) for difference in means was no lower than -0.5. Additional endpoints included Pain and Physical Function subscales of the Western Ontario and McMaster (WOMAC) OA Index. Adverse experiences (AEs) were recorded and combined from the two studies for analysis.
Results: Study 1 enrolled 395 patients (rofecoxib, n = 160; celecoxib, n = 157; placebo, n = 78). Study 2 enrolled 413 patients (rofecoxib, n = 159; celecoxib, n = 169; placebo, n = 85). Rofecoxib 12.5 mg was at least as effective as celecoxib 200 mg by PGART (Study 1 difference -0.09 [95% CI: -0.32, 0.14] and Study 2 difference 0.02 [95% CI: -0.20, 0.24]), and both were significantly (p < 0.001) more effective than placebo. Comparable efficacy was also seen for WOMAC Pain and Physical Function subscales with the active treatments. There was a significantly higher (p < 0.05) incidence of serious AEs with celecoxib than rofecoxib or placebo, none of which was drug-related. There were no significant differences in the pre-specified measurements of safety including drug-related AEs or discontinuations due to AEs, and the medications demonstrated similar safety as assessed by spontaneous reporting.
Conclusions: Rofecoxib 12.5 mg and celecoxib 200 mg provided comparable efficacy over 6 weeks, and both were significantly more efficacious than placebo. The medications demonstrated similar safety compared to one another and placebo. The primary limitations of these studies were that they were only 6 weeks long and were powered for efficacy. Therefore, conclusions about long-term safety cannot be inferred.