Objective: To compare the upper gastrointestinal (UGI) tolerability of celecoxib (a cyclooxygenase-2 specific inhibitor) and diclofenac using data from three randomised, double-blind clinical trials in osteoarthritis (OA) and rheumatoid arthritis (RA).
Methods: Patients in two OA studies received either celecoxib 100 mg BID (n = 545), diclofenac 50 mg BID or TID (n = 540), or placebo (n = 200) for 6 weeks. In the RA study, patients received celecoxib 200 mg BID (n = 326) or diclofenac 75 mg BID (n = 329) for 24 weeks. The cumulative incidence of abdominal pain, dyspepsia, nausea or any of these events (UGI tolerability composite endpoint) after the first 6 weeks was estimated using time-to-event analysis.
Results: In the pooled OA trials, the cumulative incidence of the composite endpoint was significantly higher with diclofenac (17.6%; 95% CI: 14.4-20.9%) than celecoxib (11.1%; 95% CI: 8.4-13.8%; p = 0.002) and comparable with placebo (13.3%; 95% CI: 8.1-18.4%; p = 0.157). In the PA trial, the cumulative incidence of the UGI tolerability composite endpoint was also significantly higher with diclofenac (20.7%; 95% CI: 16.3-25.1%) than celecoxib (15.9%; 95% CI: 11.9-20.0%; p = 0.013). Celecoxib was also better tolerated than diclofenac in this trial in terms of the cumulative incidences of abdominal pain (p = 0.031) and dyspepsia (p = 0.062). The results of the UGI tolerability composite endpoint analysis were confirmed using the Cox proportional hazards model to controlfor other predictors of UGI adverse events.
Conclusion: The UGI tolerability of therapeutic dosages of celecoxib was significantly better than diclofenac in patients with RA or OA.