Post-Marketing Cardiovascular Safety Signal

POST-MARKETING CARDIOVASCULAR SAFETY SIGNAL

LARGE-SCALE CLINICAL TRIALS

At the time of market approval of celecoxib and rofecoxib, upper gastrointestinal safety information of these two drugs was based on endoscopy studies. As the reduced frequency of mucosal injury in the upper gastrointestinal tract may not correlate well with incidence of serious gastrointestinal events that include ulcer, perforation, obstruction, or bleeding, manufacturers of celecoxib and rofecoxib sponsored large-scale clinical trials that were powered to provide definitive evidence on upper gastrointestinal safety for these drugs.

The Celecoxib Arthritis Safety Study (CLASS) was the first large-scale randomized trial of a COX-2 inhibitor (Silverstein et al., 2000). Patients with osteoarthritis or rheumatoid arthritis were randomly assigned to receive celecoxib, ibuprofen, or diclofenac for more than six months (Table 47.1). Approximately 20% of the study subjects took low dose aspirin (325 mg a day or less) for cardiovascular disease prevention during the study. There was no signifi-cant reduction of risk of upper gastrointestinal ulcer complications within the first 12 months of therapy (Hrachovec and Mora, 2001). Among patients who did not use aspirin, risk of upper gastrointestinal ulcer complication was reduced by approximately 50%. In the first published report of CLASS results, incidence of stroke, myocardial infarction, and angina was virtu-ally the same in the celecoxib group and in the ibupro-fen/diclofenac group during the first six months of therapy. No cardiovascular safety signal was observed in this study.

The first cardiovascular safety signal of a COX-2 inhibitor came from the Vioxx Gastrointestinal Outcomes Research Study (VIGOR) (Bombardier et al., 2000), which was another large trial that evaluated the risk of adverse upper gastrointestinal outcomes among patients on COX-2 inhibitor or non-selective NSAID. Patients with rheumatoid arthritis were randomly assigned to receive rofecoxib 50 mg once per day or naproxen 500 mg twice per day, with a median follow-up of 9 months. Unlike the enrollment criteria for CLASS, aspirin use was not allowed in VIGOR. Rofecoxib users had 60% lower risk of perfo-ration, obstruction, and severe upper gastrointestinal bleeding than naproxen users. However, reported inci-dence of myocardial infarction was higher among the rofecoxib group (relative risk, 5.0; 95% confidence interval [95% CI], 1.68–20.13) (Curfman, Morris-sey and Drazen, 2005). The VIGOR investigators hypothesized that the increased relative risk among the rofecoxib users could be a result of inherent cardio-vascular risk of rofecoxib, cardio-protective effect of naproxen, or both.