Background Dual inhibitors of cyclo-oxygenase (COX)-1 and COX-2 can potentially induce elevation of systolic and diastolic blood pressure in some patients. This is thought to be due to COX-2 mediated effects on the nephron, influencing natriuresis. We examined the relative effect of selective COX-2 inhibitors, rofecoxib and celecoxib on blood pressure (BP) during a 6-week, placebo controlled OA efficacy trial.

Methods 1082 patients with OA of the knee or hip, after withdrawal of previous therapy for OA, were randomised to treatment with rofecoxib 25 mg (QD (N = 471), celecoxib 200 mg (QD (N = 460) or placebo (N = 151). Paracetamol was allowed as rescue during the trial. Treatment groups were similar in terms of age, race, gender, baseline OA severity, history of hypertension, and baseline BP. The mean age was 62 years, the majority were female. Greater than 40% in each group had a history of hypertension. BP (systolic, SBP and diastolic, DBP) was assessed at office visits at baseline, 2, 4, and 6 weeks in the sitting position after patients rested 10 min. Relative OA efficacy was determined by% patients with good or excellent response to therapy (%PGART) and WOMAC questionnaire at scheduled office visits over 6 weeks.

Results The percent of patients with pre-defined changes in SBP (increase > 20 mm Hg and SBP >140) and DBP (increase > 15 mm Hg and DBP >90) was 9.6% for rofecoxib and 9.4% for celecoxib; 2.8% for rofecoxib and 2.0% for celecoxib, respectively compared to placebo (3.3% systolic change and 2.0% diastolic change). Differences between coxibs and placebo were significant for SBP (p = 0.015); rofecoxib and celecoxib did not differ significantly from one another in terms of either SBP or DBP. There was one discontinuation due to hypertension (overall incidence 0.2% for rofecoxib, 0% for celecoxib, and 0% for placebo). Mean changes from baseline in SBP were 1.9 mm Hg for rofecoxib. 0.2 mm Hg for celecoxib, and -4.3 mm Hg for placebo, and in DBP were 0.3 mm, -0.4 mm, and -2.6 mm Hg respectively. Relief of night pain and% PGART over six weeks were significantly greater for rofecoxib compared to celecoxib [night pain change from baseline: -36.2 mm for rofecoxib vs -32.7 mm for celecoxib (p = 0.023); PGART: 58.0% for rofecoxib vs 49.9% for celecoxib (p = 0.014)]. In addition, although not a prespecified endpoint, rofecoxib resulted in significantly greater improvement than celecoxib in scores for each of the WOMAC composite domains (pain, stiffness, and physical function: p < 0.01).

Conclusion Like dual COX-inhibiting NSAIDs, selective COX-2 inhibitors can induce elevations of systolic and diastolic blood pressure. Rofecoxib produced nearly identical small changes in SBP and DBP compared to celecoxib despite providing superior efficacy as determined by PGART and WOMAC.