Perspective on the Vioxx Recall

Perspective on the Vioxx Recall

Several authors review the evidence that led to the withdrawal of rofecoxib.

Most physicians -- like their patients -- got information about the withdrawal of rofecoxib (Vioxx) from the media. Two essays in the October 21, 2004, issue of the New England Journal of Medicine (published online on October 6) provide additional perspective.

In the APPROVe (Adenomatous Polyp Prevention on Vioxx) study, which had been in progress for about 3 years, 2600 people with previously removed colorectal polyps were assigned randomly to receive rofecoxib or placebo. The trial was launched because nonsteroidal anti-inflammatory drugs inhibit growth of colorectal adenomas in some patients. According to Cleveland Clinic's Eric Topol, 3.5% of rofecoxib recipients and 1.9% of placebo recipients suffered myocardial infarctions or strokes during the trial (an excess of 16 events per 1000 patients). Because the trial included only people without known cardiovascular disease, Topol implies that risk could be even greater among patients with preexisting cardiovascular disease.

Topol has expressed concern about rofecoxib for several years. For example, following publication of the VIGOR trial, which was a comparison between rofecoxib and naproxen in arthritis patients (Journal Watch Dec 8 2000), Topol and colleagues argued that a trial designed specifically to assess cardiovascular risks of cyclooxygenase (COX)-2 inhibitors was needed (Journal Watch Aug 28 2001). In VIGOR, adverse cardiovascular events occurred significantly more often in the rofecoxib group than in the naproxen group (46 vs. 20 events among about 4000 patients in each group, treated for a median of 9 months). In his current discussion, Topol criticizes both the drug company and the FDA for "failing the public health."

In a second essay, the University of Pennsylvania's Garret FitzGerald explains that the COX-2 enzyme is the main source of prostaglandin I₂, which inhibits platelet aggregation and causes vasodilation. Thus, inhibition of COX-2 theoretically could increase cardiovascular risk by tipping the balance toward platelet aggregation and vasoconstriction. FitzGerald also briefly presents his view of worrisome trends in clinical trials of other COX-2 inhibitors:

In the CLASS trial (a comparison of celecoxib [Celebrex], ibuprofen, and diclofenac), no increase in cardiovascular risk was evident at 6 months (Journal Watch Sep 22 2000); however, FitzGerald asserts that a retrospective analysis of longer-term unpublished data "also reveals signs of increased cardiovascular risk" with celecoxib.
Among nonaspirin users in the TARGET trial, a nonsignificant trend was noted toward more myocardial infarctions at 1 year with lumiracoxib (a COX-2 inhibitor not yet FDA approved) than with naproxen and ibuprofen (Journal Watch Sep 17 2004).

In addition, parecoxib (which was rejected by the FDA) is a prodrug that is converted in vivo to valdecoxib (Bextra). In a trial of intravenous parecoxib followed by oral valdecoxib for postoperative pain relief in patients who underwent coronary bypass surgery, a nonsignificant trend toward increased risk for MI and stroke was noted in parecoxib/valdecoxib recipients, compared with placebo recipients (J Thorac Cardiovasc Surg 2003; 125:1481).

FitzGerald asserts that the burden of proof now has shifted to those who claim that cardiovascular risk does not extend to other COX-2 inhibitors. We surely can expect additional heated debate about whether the rofecoxib experience represents a class effect for all COX-2 inhibitors. In the meantime, I believe that a prudent approach is to avoid the use of COX-2 inhibitors in patients with cardiovascular disease and to consider carefully the potential risks before using them in other patients.

— Allan S. Brett, MD

Published in Journal Watch General Medicine October 15, 2004