Bad News About the COX-2 Inhibitors

Special Report

Comment by Jonathan Abrams, MD, Professor of Medicine, Division of Cardiology; University of New Mexico, Albuquerque

There has been enormous controversy regarding the benefit/safety relationship of the COX-2 inhibitors, which were initially introduced in 1999. A recent FDA panel garnered considerable national publicity and “rendered an ambivalent verdict on the future of the cyclo-oxygenase-2 (COX-2) inhibitors.” It initially appeared that the Merck drug, rofecoxib (ROF), was the only 1 of the 3 marketed COXIBS, including celecoxib (CEL) and valdecoxib (VAL), whose safety was questioned. Recent issues of the New England Journal of Medicine contain numerous articles (3 original reports, 2 editorials) that deal with this complex and unhappy story.

Is Rofecoxib Safe? One of the first hints suggesting increased cardiovascular risk with COX-2 inhibitors came from the VIGOR trial, published in 2000, demonstrating more cardiovascular (CV) events in individuals randomized to ROF compared to naproxen. This study generated enormous controversy regarding ROF, ultimately culminating in Merck’s withdrawal of the drug from the market. APPROVe, a recent study of 2600 individuals with a history of colorectal adenoma, randomized subjects to 25 mg ROF or placebo, was terminated prematurely because of increased cardiovascular risk in the ROF cohorts, with more confirmed serious atherothrombotic events during follow-up of 2.5 years vs placebo, relative risk of 1.9, P = 0.008. The event curves separated after 18 months of treatment. There also was a nonadjudicated 5-fold increase in congestive heart failure. APPROVe is the first placebo-controlled study to document a substantial risk for ROF; many other reports had not shown any increase in CV events vs conventional NSAID. Okie and colleagues discuss the conflicting data in the literature, including contradictory meta-analyses, and conclude that there is “an increased risk of confirmed thrombotic events associated with the long-term use of rofecoxib.”

Is Celecoxib Associated with CV Risk? The answer apparently is yes, and it is strongly supported by the APC (Adenoma Prevention with Celecoxib) study of 2000 patients with a history of colorectal cancer; a trial comparing 2 doses of CEL (200 mg or 400 mg BID) or placebo for prevention of colorectal adenoma. After 3 years, the composite CV end point was 1% in the placebo group, compared to 2.3% in the 200 mg BID cohort and 3.4% in the 400 mg BID cohort. Because of these findings, early discontinuation of the trial was recommended. This study appears to be a smoking gun, indicating that ROF is not the only culprit in the COX-2 armamentarium. Okie et al suggest “there may be a real increase in cardiovascular risk associated with the use of celecoxib in particular, and the class of selective COX-2 inhibitors in general.”

Are valdecoib and parecoxib safe? A late-breaking trial was presented at the American College of Cardiology meeting in Orlando, which employed 2 COXIBS to treat post operative pain following coronary bypass surgery. Seventeen hundred patients were randomly assigned intravenous parecoxib (PAR) for at least 3 days, followed by oral valdecoxib (VAL) through day 10. Three cohorts, placebo-placebo, PAR-placebo, and VAL-placebo, were studied. The primary end points included a broad composite of CV events.

Results: Both groups using active coxibs were associated with increased CV risk, 7.4% in each vs 4.0% in the placebo group, resulting in a risk ratio of 1.9 for all comparisons P = 0.002. Okie et al conclude “short term COX-2 inhibitors are associated with a significant risk of thromboembolic events in patients with high risk of such events.” They note that prior studies have suggested possible adverse effects of these 2 COXIBs. They state “selective COX-2 inhibitors should be avoided in patients undergoing CABG. . . and probably. . . in patients undergoing any vascular procedure for atherosclerotic disease.”

Two editorials discuss the FDA COX-2 meeting and the question as to whether other NSAIDs may be harmful; much discussion is on the role of naproxen in this controversy.1,5 “The members (advisory committee) sense that the cardiovascular risk with celecoxib appeared smaller than that with rofecoxib and valdecoxib.”1 The vote was 31 to 1 in favor of keeping CEL on the market and a split vote for ROF 17-15 and VAL 17-13, staying on the market. Criticism of both Merck and Pfizer came for some committee members who urged that manufacturers be prohibited from direct to consumer advertising (the FDA apparently can not ban this as such, but can require a black box warning on COX-2 labeling). This is a sobering report about the special FDA hearing, with little good news for the COX-2 agents.

A second editorial by a law professor is particularly critical of the Merck decision not to act on the early warnings that ROF might be hazardous.5 There is emphasis on the financial profits from the sales of these drugs that complicates this unhappy saga and fuels the controversy and publicity attending the removal of Vioxx from the market. [On April 7, 2005, Bextra (VAL) was withdrawn from the market by Pfizer].


The remarkably intense focus on the COX-2 drugs over the past several years is well documented in this series of articles. There have now been several COX-2 meta-analyses which are in conflict. It appears clear, through the retrospectroscope, that these drugs do impart hazard via putative interactions regarding prostacycline, thromboxane inhibition, etc. These agents may have been allowed to stay on the market for too long after early warning systems were activated. It is likely, but not proven, that patients with increased CV risk have a higher likelihood of having an adverse event with a COXIB. We do not know whether short duration of therapy is safer than long duration. We do not know whether high doses are more hazardous than low doses, although recent trials with both ROF and CEL would suggest that is the case. Reports that some patients only appear to respond to COX-2 agents for relief of pain are subjective. Helping our patients is one of our primary objectives! If there is an increased risk, it should be clearly stated and understood by all. This issue is, of course, heightened by the enormous success of COXIB sales, attributed by critics to the intense widespread direct consumer advertising campaigns on television and magazines. Patient choice of a COX-2 inhibitor may also be related to their long duration of action, requiring once or twice daily therapy when compared to some traditional NSAIDs. While there has been no evidence that short term use of these drugs, eg, a few days to 2 weeks, imparts risk, the CABG trial suggests this may be the case.4

This is a cautionary tale. Use of COX-2 drugs for arthritis, inhibition of potentially precancerous colonic polyps, or reducing the risk of GI bleeding compared to other NSAIDs, cannot be supported by current data. Millions of dollars in lawsuits against Merck and possibly Pfizer are likely, relating to presumed drug related major cardiovascular events that some believe have been prevented by earlier cessation of marketing. There does not appear to be any winner in this evolving story.


1. Okie S. N Engl J Med. 2005;352:1283-1285.

2. Bresalier RS, et al. N Engl J Med. 2005;352: 1092-1102.

3. Solomon SD, et al. N Engl J Med. 2005;352:1071-1080.

4. Nussmeier NA, et al. N Engl J Med. 2005;352: 1081-1091.

5. Eisenberg RS. N Engl J Med. 2005;352:1285-1287.