Background: Cyclooxygenase-2 (COX-2) selective inhibitor drugs have been associated with fewer gastrointestinal complications than traditional nonsteroidal anti-inflammatory drugs (NSAIDs), but they have an increased risk of thrombotic cardiovascular events compared with placebo. Limited data from observational studies suggest that NSAIDs may also increase cardiovascular risk. The two classes of drugs have not been directly compared for their cardiovascular risks. Because both classes are taken over long periods by large numbers of patients with common chronic pain conditions, differences in the risks of cardiovascular and gastrointestinal events could have far-reaching effects. The Multinational Etoricoxib and Diclofenac Arthritis Long-term (MEDAL) study compared the relative risks of gastrointestinal and thrombotic cardiovascular events for the COX-2 inhibitor etoricoxib (Arcoxia) and the traditional NSAID diclofenac (Voltaren).

The Study: The study involved more than 34,000 patients who required long-term analgesic therapy for osteoarthritis or rheumatoid arthritis that was not adequately managed using acetaminophen. Patients with a history of myocardial infarction or interventions to reperfuse myocardium were included.

Patients were randomly allocated to treatment with etoricoxib (90 or 60 mg daily) or diclofenac (75 mg twice or 50 mg three times daily). Placebo pills were used to provide each patient with the same number of pills per day. Low-dose aspirin was advised for all patients with diabetes or cardiovascular or cerebrovascular disease. Antiulcer therapy with misoprostol (Cytotec) or proton pump inhibitors was recommended for patients at high risk of gastrointestinal hemorrhage. All patients also received treatment for hypertension based on the recommendations of the seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.

The primary outcome was the first of several selected cardiovascular events including myocardial infarction, unstable angina pectoris, cardiac arrest, stroke, transient ischemic attack, pulmonary embolism, arterial or venous thrombosis, or sudden death. Other end points included gastrointestinal adverse events, renal dysfunction, and treatment discontinuation because of hypertension. The gastrointestinal events of interest included bleeding from the upper gastrointestinal tract, perforation from a nonmalignant gastric or duodenal ulcer, and clinically documented peptic ulcer. These events were adjudicated by an independent committee whose members were blinded to patient treatment allocation. This committee also judged potential lower gastrointestinal events. Patients were reassessed at clinic visits every four months and had telephone contact between visits.

Results: The 17,412 patients allocated to etoricoxib were similar to the 17,289 patients allocated to diclofenac in all significant variables. About 78 percent of the participants were white, 74 percent were women, the average age was 63 years, and the average body mass index was 29.5 kg per m². About 12 percent were smokers, 11 percent reported diabetes, and 38 percent had established heart disease or two or more cardiovascular risk factors. The average treatment lasted 18 months, and overall compliance was 98 percent for etoricoxib and 96 percent for diclofenac.

Thrombotic cardiovascular events occurred in 320 patients (1.24 per 100 patient-years) in the etoricoxib group and 323 patients (1.30 per 100 patient-years) in the diclofenac group. The hazard ratio was 0.95 (95% confidence interval [CI], 0.81 to 1.11), indicating that the treatments were almost identical in this effect. For all upper gastrointestinal events, the rates were 0.67 per 100 patient-years for etoricoxib and 0.97 per 100 patient-years for diclofenac. The hazard ratio for all upper gastrointestinal events of 0.69 indicated a significantly lower rate for etoricoxib. Nevertheless, rates of complicated upper gastrointestinal events and all lower gastrointestinal events were similar in both groups.

Conclusion: The authors conclude that patients taking etoricoxib for long-term therapy and comparable patients taking diclofenac had almost identical rates of thrombotic cardiovascular events. They stress that the choice of long-term therapy for osteoarthritis or rheumatoid arthritis must be individualized based on the risk factors of each patient. They also stress that the findings from this study may not be applicable to other agents in these two classes of drugs.