LONDON – Celecoxib was associated with very low cardiovascular event rates, and its use posed no more risk than other painkillers commonly used to treat elderly individuals with arthritic conditions but no heart disease in a large, pragmatic, family practice–based study.
Results of the Standard Care Versus Celecoxib Outcome Trial (SCOT) reported at the annual congress of the European Society of Cardiology also showed that celecoxib was no more likely than nonselective nonsteroidal anti-inflammatory drugs (nsNSAIDs) to cause ulcer-related upper gastrointestinal (GI) tract complications.
Dr. Tom MacDonald
In fact, the rates of both cardiovascular and GI events were so low overall that it made the trial difficult to complete, said study investigator Dr. Tom MacDonald, professor of clinical pharmacology and pharmacoepidemiology at the University of Dundee (Scotland), which sponsored the study.
The on-treatment and intention-to-treat (ITT) cardiovascular event rates were 0.9% and 1.1% per 100 patient-years, he observed, adding that he would have expected the event rate to be around 2%-3% in the population studied. GI complication rates were even lower, with just 12 on-treatment and 15 ITT events reported during the entire follow-up period, which was a maximum of 6.3 years and mean of about 3 years.
“You may remember the brouhaha surrounding the use of rofecoxib and other [cyclo-oxygenase-2 inhibitors],” said Dr. MacDonald. Both coxibs and nsNSAIDs have been associated with adverse cardiovascular outcomes such as myocardial infarction (BMJ 2005;330:1366), and rofecoxib was voluntarily withdrawn in 2004 by its manufacturer from the U.S. market. A recent meta-analysis (Lancet 2013;382:769-79) has suggested that coxibs increase the risk of major cardiovascular events by about 37%.
The SCOT study (BMJ Open 2013;3:e002295) was designed to assess if celecoxib was better, worse, or the same as the other available NSAIDs in terms of its cardiovascular and gastrointestinal safety. It was originally set up because of a requirement by the European Medicines Agency, Dr. McDonald explained.
More than 9,400 patients aged 60 years or older with osteoarthritis or rheumatoid arthritis who were prescribed chronic NSAID therapy and had no existing cardiovascular disease were screened at 706 family practices in Scotland, England, Denmark, and the Netherlands. A total of 7,297 patients were included in the prospective study and were randomized to switch to treatment with celecoxib or to continue their current nsNSAID.
General practice records were linked to hospital and mortality databases to derive the primary composite endpoint of the first occurrence of hospitalization for nonfatal MI, nonfatal stroke, or cardiovascular death, as well as secondary endpoints such as time to first hospitalization or death from upper GI complications and all-cause mortality.
Randomized patients were about 68 years old, and about 40% of patients were male. Dr. MacDonald noted that, although there was no known existing cardiovascular disease at enrollment, the baseline characteristics showed that around 44% of patients had high blood pressure; a third of patients had high cholesterol; and 20%, 12%, and 38% were taking a statin, aspirin, or ulcer-healing treatments, respectively. The most common nsNSAIDs being used were diclofenac (38.7%) and ibuprofen (31%).
There was no significant difference between celecoxib or nsNSAIDs for any of the cardiovascular endpoints studied, with hazard ratios (HR) for the primary composite cardiovascular endpoints of 1.12 (95% confidence interval, 0.81-1.55; P = .5) while on celecoxib treatment and 1.04 (95% CI, 0.81-1.33; P = .75) in the ITT analysis. Similar results were obtained for all-cause mortality (HR, 1.2 and 0.92, respectively).
Dr. MacDonald reported that 50% of patients randomized to celecoxib and 30% randomized to continue nsNSAIDs withdrew from the study. The main reasons for stopping celecoxib were a lack of efficacy (11.2% vs. 2% for nsNSAIDs), adverse events (8.3% vs. 4.4%), patient request (6% vs. 2.3%), not tolerated (3.9% vs. 1.2%), or a serious adverse event (2.6% vs. 1.9%). There was, however, a lot of adverse publicity about the coxibs, he noted, and patients who had been happy on an nsNSAID might not have been happy with the switch.
The rates of serious cardiovascular adverse events (31.7% vs. 32.4%) or reactions (5.2% vs. 5.8%) were similar with celecoxib and nsNSAIDs, but there were significantly fewer serious GI adverse reactions with celecoxib than with nsNSAIDs (38 vs. 66; P = .007). Overall, the adverse reaction rate was 22% vs. 16.1%, respectively (P <.001).
“In the study population, nsNSAIDs and celecoxib both appeared acceptably safe,” Dr. MacDonald concluded. “In patients who get significant symptomatic relief from these medicines, the benefit/risk balance appears positive.”
Although the findings are perhaps reassuring, they are unlikely to change clinical practice, observed Dr. José López-Sendon, who was invited to comment on the study results after their presentation at the conference.