Untitled Document

Fun with statistics:
Do celecoxib (Celebrex) and rofecoxib (Vioxx) increase risk of cardiovascular events?
Volume XI, Issue 9
Harold J. DeMonaco, M.S., Director of Drug Therapy Management

A recent special communication published in JAMA (http://jama.ama-assn.org/issues/v286n8/rfull/jsc10193.html) raised questions about the safety of the Cox-2 specific NSAIDs, rofecoxib (Vioxx) and celecoxib (Celebrex), relative to cardiovascular risk. The authors of the JAMA paper examined the results of two major randomized trials; VIGOR (Vioxx Gastrointestinal Outcomes Research Study) and CLASS (Celecoxib Long-Term Arthritis Safety Study) along with two smaller studies and compared the cardiovascular event rate to that identified in a meta-analysis of 4 previously-published aspirin primary prevention trials.

VIGOR

The VIGOR trial was an industry-sponsored double-blind, randomized trial in 8076 subjects with rheumatoid arthritis (RA) comparing the incidence of GI toxicity with rofecoxib (Vioxx) 50mg daily or naproxen 1000mg daily. Aspirin use was not permitted. VIGOR was not intended to examine cardiovascular events nor was the sample size calculated based on cardiovascular risk.

The data and safety monitoring board for the study identified what appeared to be an increased risk of cardiovascular events in this blinded study. One hundred eleven of the 4047 study subjects on rofecoxib and 50 of the 4029 subjects treated with naproxen suffered cardiovascular events. The risk of suffering a serious cardiovascular event was 2.2 times higher in those treated with rofecoxib than in those treated with naproxen.

CLASS

The CLASS trial was an industry-sponsored double-blind randomized trial comparing celecoxib (Celebrex) to ibuprofen and diclofenac in 8059 subjects with osteoarthritis (OA) or rheumatoid arthritis. The trial was designed to determine if a difference existed between the study drugs with respect to gastrointestinal toxicity over a 6-month treatment period. Low-dose aspirin was allowed. No difference was noted in the incidence of cardiovascular events among the subjects treated with celecoxib, ibuprofen and diclofenac.

The authors of the JAMA paper rightly point out the methodological problems with their analysis:

The incidence of cardiovascular events was not a specified outcome measure in either study
Post hoc analysis is fraught with a bit of error
The patient populations studied were heterogeneous and cardiac risk varied greatly.

Despite these serious methodological issues, the authors proceeded to compare the results of these studies with a meta-analysis of the US Physicians' Health Study, the UK Doctors Study, the Thrombosis Prevention Trial and the Hypertension Optimal Treatment trial. The annualized rate of myocardial infarction with placebo was 0.52% based on a meta-analysis of these 4 studies. The MI rate in VIGOR and CLASS was 0.74% (p=0.04 compared to placebo group in the meta-analysis) with rofecoxib (Vioxx) and 0.8% (p=0.02 compared with placebo group in the meta-analysis with celecoxib (Celebrex). Based on these findings there appears to be an increased risk of a cardiovascular event associated with the use of a Cox-2 specific NSAID.

The FDA has been reviewing the data; reviewers' comments are available at http://www.fda.gov/ohrms/dockets/ac/01/briefing/3677b2_06_cardio.pdf . There are several interesting pieces of information in the FDA data that are not available in the JAMA report.

The event rate in the VIGOR trial for naproxen was more than twice that of the placebo subjects in the studies' baseline meta-analysis (1.19 vs. 0.52). No one is suggesting that naproxen increases cardiovascular risk, however.
The increased risk of a cardiovascular event during treatment with rofecoxib extended into those subjects without a prior history of cardiovascular disease or risk factors. This would suggest that rofecoxib (Vioxx) not only is not protective, but increases baseline risk.

At the moment it is unclear whether there is an increased cardiovascular risk with rofecoxib (Vioxx) as compared to naproxen. Despite claims to the contrary, it is not clear that celecoxib (Celebrex) is safer than Vioxx in this regard. Only a well-designed randomized trial will answer the myriad of questions raised by the JAMA report.

Recommendations:

Since Vioxx and Celebrex have not been shown to be more effective than older and non-specific NSAIDs, their use in patients without risk factors for GI bleeding or for treatment of acute musculoskeletal pain should be discouraged
Vioxx and Celebrex are reasonable choices for patients with RA or OA who are at high risk of GI bleeding based on their ARAMIS score.
For those patients with concomitant cardiovascular risk factors in whom low-dose aspirin prophylaxis is indicated, the risk of GI bleeding should be considered. The American Heart Association currently recommends the use of 75mg to 325mg aspirin daily. Presumably, the lower the aspirin dose, the lower the risk of GI bleeding.