Introduction

The withdrawal of rofecoxib (Vioxx®) from the market due to increased risk of cardiovascular (CV) events has led to speculation about the future of the cyclooxygenase-2 (COX-2) inhibitors.1-3 In the first of a 3-part series, CIAOMed will examine possible underlying mechanisms that may be responsible for elevated CV risk in select COX-2 inhibitors and whether such explanations extend to all coxibs. Subsequent installments will examine how to manage patients in light of the new reports and the future of this frequently used drug class, including the next-generation COX-2 inhibitors in phase III of development.

Merck & Co, Inc, removed rofecoxib from the worldwide market on September 30, 2004, after finding that the agent, approved for the treatment of arthritis and acute pain, was associated with a 2-fold greater risk of myocardial infarction (MI) or stroke compared to placebo after 18 months of treatment (1.96 RR, 95% confidence interval 1.20 to 3.19; P < .05).4 This decision triggered a chain of events. Pfizer Inc soon announced its intention to further examine the safety of valdecoxib (Bextra®), given reports of increased CV events in 2 trials of patients who had recently undergone coronary artery bypass graft (CABG) surgery.5

On October 29, the Food and Drug Administration (FDA) gave Merck an "approvable letter" for etoricoxib (Arcoxia®), a next-generation COX-2 inhibitor, but the federal agency noted in the letter that more safety and efficacy data would be required as a condition of approval.6

Some opinion leaders have criticized both Merck and the FDA for not acting on earlier warnings about rofecoxib's potential thrombotic risk.7,2 The FDA and the European Medicines Evaluation Agency each plans to turn a sharp eye on cardiovascular safety data for the remaining COX-2s.8 The FDA will hold a public advisory committee meeting in early 2005 to discuss the cardiovascular safety profile of COX-2 inhibitors.9

Is CV Risk a Class Effect?

Experts disagree on this question. "Evidence to date does not indicate that celecoxib is associated with an increased risk for thromboembolic cardiovascular events," Lee S. Simon, MD, rheumatologist and Associate Clinical Professor of Medicine at Harvard Medical School in Boston, Massachusetts, and coauthor Vibeke Strand, MD, Adjunct Clinical Professor in the Division of Immunology and Rheumatology, Stanford University School of Medicine, Stanford, California, wrote in a commentary just published by the Cleveland Clinic Journal of Medicine.3

The same types of studies that revealed CV risk with rofecoxib show no such signal in celecoxib (Celebrex®), Steven B. Abramson, MD, rheumatologist and Professor in the Department of Rheumatology/Medicine at New York University School of Medicine, New York, told CIAOMed. For example, Dr. Abramson said that data regarding each drug's CV risk comes from a randomized, controlled clinical trial of gastrointestinal protection compared to one or more nonselective nonsteroidal anti-inflammatory drugs (NSAIDs), and from observational data. Incidence of MI was significantly higher with rofecoxib than in naproxen (0.1% vs 0.4%) in the Vioxx Gastrointestinal Outcomes Research (VIGOR) trial.10 In contrast, the relative risk of serious thromboembolic events (RR, 1.1) with celecoxib was similar to that of NSAIDs in the Celecoxib Long-term Arthritis Safety Study (CLASS).11 A Medicaid case-control study recently reported a significantly elevated relative risk of acute MI with rofecoxib compared to that of celecoxib (P = .011), though the risk was increased only for the first 90 days of treatment.12

"Line up the levels of proof," Dr. Abramson said in an interview with CIOAMed. "Wherever rofecoxib signaled, celecoxib did not, given the opportunity."

Leslie Crofford, MD, CIAOMed Board Member and Professor of Internal Medicine and Chief of Rheumatology at the University Of Kentucky College Of Medicine, Lexington, Kentucky, acknowledged that COX-2 inhibitors exhibit a "class effect," but said that it depends on the definition. "I think it's a class effect, but you must be careful how you define the ‘class,'" Dr. Crofford said. "Celecoxib barely makes it into the COX-2 category. It is much less selective than rofecoxib and a little more COX-2 selective than some drugs we don't call coxibs," such as diclofenac(Voltaren®).13

Furthermore, some signals may point to thrombotic issues with celecoxib, Dr. Crofford noted, adding that ischemic complications have occurred with celecoxib in patients at high risk for thrombosis due to conditions such as Raynaud's phenomenon, lupus anticoagulant, or elevated cardiolipin antibodies.14

Valdecoxib has been associated with CV events in patients at high thrombotic risk due to recent CABG, reported Dr. Crofford, an observation that contrasts with a recent data review by Pfizer Inc that revealed no increased CV thromboembolic risk with valdecoxib in osteoarthritis or rheumatoid arthritis patients.5

"Although randomized clinical trials do not indicate a cardiovascular risk (with valdecoxib), Dr. Abramson added, "there is not enough longitudinal data in large clinical populations to draw firm conclusions." The higher incidence of CV events in patients receiving parecoxib (Dynastat®, Pfizer)/valdecoxib for post-CABG pain "will lead to further studies of the drug," he said.

Mechanisms of CV Risk

One often quoted theory to explain the CV risk seen with COX-2s is that these agents block the good CV effects of prostaglandin I2 (a COX-2 product) but do not inhibit the negative CV effects of thromboxane A2 (a COX-1 product).1 Prostaglandin I2 reduces platelet aggregation and vascular smooth muscle proliferation, and increases vasodilation. Thromboxane A2 promotes the opposite of those benefits. Nonselective NSAIDs inhibit both substances. Blocking only prostaglandin I2 could produce a prothrombotic state in persons at CV risk,1 according to Garret A. FitzGerald, MD, Professor of Medicine and Pharmacology, Institute of Translational Medicine and Therapeutics, University of Pennsylvania, Philadelphia.

"I think that COX imbalance is only part of the story," said New York University's Dr. Abramson, who noted that celecoxib exhibited this imbalance in short-term studies of urinary secretion of prostacyclin and thromboxane in normal volunteers. However, he added that it does not appear to be disposed to CV adverse events in the clinic.

Other factors to consider include:

Effect on blood pressure and edema. Rofecoxib raises blood pressure and causes edema in a dose-related fashion.3 Celecoxib and rofecoxib both produce a roughly 2% incidence of hypertension and edema, similar to that observed with nonselective NSAIDs at those doses used for chronic analgesia. However, there does not appear to be a dose-response association with blood pressure; a hypertension dose-response is "not apparent with celecoxib at any dose," Drs. Simon and Strand reported in the Cleveland Clinic Journal of Medicine commentary. 3 Their comments were in response to a review of epidemiologic studies, clinical trials, and new drug application summaries of both agents.

"It's hard not to be concerned about the blood pressure signal from NSAIDs, whether they are COX-selective or nonselective," even without causal evidence linking elevated blood pressure to cardiovascular events in COX-2 users, Dr. Abramson said, adding that studies show that as many as 40% to 50% of osteoarthritis patients have hypertension, making this a high-risk group at the outset.

Degree of COX-2 selectivity. If COX-2 selectivity promotes GI protection, it may seem that more would be better. Recent published trials noted that the relatively lower selectivity of celecoxib may explain why it demonstrates less CV risk compared with rofecoxib, Dr. Crofford said. "The second generation [of coxibs] is more selective and will end up not being a good thing on balance," she said.

Dr. Abramson did not disagree with the possibility that higher selectivity may lead to greater CV risk.

Half-life, tissue deposition. Rofecoxib has a relatively long half-life (about 17 hours) compared to that of celecoxib (about 11 hours when fasting) or valdecoxib (8 to 11 hours). This should prolong the duration of the prothrombotic imbalance demonstrated by Dr. FitzGerald, Dr. Abramson pointed out. A drug with a short plasma half-life but prolonged effect at inflammation sites might combine efficacy with lower CV risk, he said.

The importance of half-life and cardiovascular risk is illustrated by apparent differences between 2 nonselective proprionic acids, ibuprofen and naproxen, Dr. Abramson said. Naproxen, with a longer half-life, provides sustained inhibition of platelet activation and may be cardioprotective. Conversely, ibuprofen (half-life 2 to 4 hours) not only does not appear to be cardioprotective but also may abrogate the effect of aspirin if given prior to low-dose aspirin. Thus, one must be wary of attributing adverse events to "class" effects without accounting for other physicochemical, pharmacokinetic, and pharmacodynamic properties, he concluded.

Dr. Crofford also commented on half-life as a factor in CV risk. She noted that celecoxib's lower COX-2 selectivity and short half-life compared to others in its class may have turned out to be assets. Even relatively low COX-2 specificity may help prevent GI side effects because COX-1 is dominant in the GI tract, she said.

Homeostasis in the cardiovascular system involves both COX-1 and COX-2, so that lower COX-2 selectivity may be an advantage. Additionally, celecoxib's relatively short half-life permits COX-2 to be active in vascular tissue at certain times over a 24-hour period. "Pfizer may turn out to have been fortunate with respect to the degree of selectivity and the pharmacologic properties of celecoxib," Dr. Crofford said, adding that CV risk likely is a COX-2 class effect "if you get enough selectivity and a long enough half life."

"This (CV risk) is not necessarily a "coxib" effect, per se," Dr. Abramson said. "However, it may be an effect of highly selective COX-2 inhibitors with prolonged plasma half-lives."

Selective COX-2 inhibition may turn out to be like tumor necrosis factor-α inhibition, Dr. Crofford said. "You don't want too much of it."

The series' next installment will focus on managing patients in light of rofecoxib's withdrawal.


References:

  1. FitzGerald GA. Coxibs and cardiovascular disease. N Engl J Med. 2004;351:1709-1711.
  2. Topol EJ. Failing the public health—rofecoxib, Merck, and the FDA. N Engl J Med. 2004;351:1707-1709.
  3. Simon LS, Strand V. A world without Vioxx: To COX-2 or not to COX-2? Cleve Clin J Med. 2004;71:849-856.
  4. Merck announces voluntary worldwide withdrawal of Vioxx. Available at: http://www.vioxx.com/rofecoxib/vioxx/hcp/hcp_notification_physicians.jsp. Accessed November 1, 2004.
  5. Pfizer Provides Information to Healthcare Professionals About Its Cox-2 Medicine Bextra R (Valdecoxib). Pfizer news release. Available at: http://www.pfizer.com/are/news_releases/2004pr/mn_2004_1015.html. Accessed November 1, 2004.
  6. Merck Receives "Approvable" Letter from FDA on New Drug Application for ARCOXIAâ„¢. October 29, 2004. Available at: http://www.merck.com/newsroom/press_releases/research_and_development/
    2004_1029.html    . Accessed November 4, 2004.
  7. Mukherjee D, Nissen SE, Topol EJ. Risk of cardiovascular events associated with selective COX-2 inhibitors. JAMA. 2001;286:954-959.
  8. EMEA to review COX-2 inhibitors. October 22, 2004. Available at: http://www.emea.eu.int/htms/hotpress/h11790804.htm. Accessed November 3, 2004.
  9. FDA to hold COX-2 advisory committee in early 2005. October 26, 2004. Available at: http://www.medcall.com/article.php?story=20041026061659407. Accessed November 4, 2004.
  10. Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med. 2000;343:1520-1528.
  11. White WB, Faich G, Whelton A, et al. Comparison of thromboembolic events in patients treated with celecoxib, a cyclooxygenase-2 specific inhibitor, versus ibuprofen or diclofenac. Am J Cardiol. 2002;89:425-430.
  12. Solomon DH, Schneeweiss S, Glynn RJ, et al. Relationship between selective cyclooxygenase-2 inhibitors and acute myocardial infarction in older adults. Circulation. 2004;109:2068-2073.
  13. FitzGerald GA , Patrono C. The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J Med. 2001;345:433-442.
  14. Crofford LJ, Oates JC, McCune WJ, et al. Thrombosis in patients with connective tissue diseases treated with specific cyclooxygenase 2 inhibitors. A report of four cases. Arthritis Rheum. 2000;43:1891-1896.