When Vioxx and Celebrex were approved about two years ago, we were quite skeptical about the claims that they were much safer in the gastrointestinal (GI) tract than other, older nonsteroidal anti-inflammatory drugs (NSAIDs). At the time of approval, neither Merck (Vioxx) nor Pharmacia (Celebrex) had done the comparative long term, higher dose randomized trials in which the newer COX-2 inhibitor drug would be compared to the least dangerous of these older drugs such as ibuprofen (Motrin)...
When Vioxx and Celebrex were approved about two years ago, we were quite skeptical about the claims that they were much safer in the gastrointestinal (GI) tract than other, older nonsteroidal anti-inflammatory drugs (NSAIDs). At the time of approval, neither Merck (Vioxx) nor Pharmacia (Celebrex) had done the comparative long term, higher dose randomized trials in which the newer COX-2 inhibitor drug would be compared to the least dangerous of these older drugs such as ibuprofen (Motrin) to find out if there is a statistically significantly lower amount of serious GI complications such as perforations, ulcers or bleeding with the COX-2 inhibitor drug. In addition, we were worried, as we are about any new drug that does not have an important advantage over existing drugs, about hidden dangers. This is especially true for the COX-2 inhibitors since they affect so many different parts of the body.
Prior to recent FDA advisory committee meetings, the two companies submitted new studies which they hoped would convince the FDA to allow the labels of their drugs to state that they were safer, as far as GI toxicity, than the older NSAIDs. Pharmacia played more by the rules and conducted randomized studies in which Celebrex was compared with either ibuprofen or diclofenac, the two NSAIDs with the least GI toxicity. The study failed to show a significantly lower amount of these serious GI adverse effects for Celebrex, compared to the other drugs. Cheating, somewhat, Merck decided that Vioxx would probably appear to have a more favorable safety profile if it was compared to a more-dangerous-than-ibuprofen/diclofenac-NSAID and they chose naproxen as the comparator drug for their study. Although the Merck study did find a statistically significantly lower amount of serious GI complications with Vioxx compared to naproxen, there were several other problems, beyond the wrong choice of naproxen for comparison, which make the findings not really applicable to the general population of people using Vioxx. First, the study was limited to people who had rheumatoid arthritis, a disease for which Vioxx is not even approved. Second, almost 60 percent of the people in the study were simultaneously using steroids such as prednisone as adjunctive treatment for their rheumatoid arthritis. Since steroids themselves can cause ulcers, this distorts the findings.
We agree with the conclusions of FDA's Office of Postmarketing Drug Risk Assessment that the 73 deaths seen with celecoxib [Celebrex](36) or rofecoxib [Vioxx](37) from gastrointestinal bleeding, obstruction, perforation or stenosis show that the current labeling for the two drugs "reflect[s] the risk of fatal gastrointestinal bleeding, obstruction, perforation or stenosis."
In a recent study published in the Proceedings of the National Academy of Sciences, the ability of rabbits to withstand temporary experimental coronary artery occlusion was significantly impaired by treatment with celecoxib which completely blocked the cardioprotective effects of the COX-2 enzyme. The authors of that study concluded that COX-2 enzyme is a Òcardioprotective protein.Ó Therefore, it is implied, drugs which block this cardioprotective enzyme, such as COX-2 inhibitors, may neutralize its protective effects.
In Merck's study comparing Vioxx to naproxen, there was a highly statistically significant five-fold increase in heart attacks in the overall Vioxx group (0.5 percent) compared to the naproxen group (0.1 percent). This amounted to approximately 20 heart attacks with Vioxx (out of 4047 patients) compared with 4 with naproxen (out of 4029 patients). This increased number of heart attacks was also accompanied by an increase in other thrombotic (blood clotting) adverse effects such as strokes and blood clots in the legs as well as problems with hypertension in the Vioxx group compared with the naproxen group.
Although the Celebrex study did not find a significantly elevated number of heart attacks in those using that drug compared with those using the older NSAIDs (ibuprofen or diclofenac), there was also cause for concern about heart toxicity with that drug. FDA Cardio-Renal division reviewer Dr. Throckmorton found that "the incidence of adverse events related to cardiac ischemia (decreased blood flow to the heart) was higher in the celecoxib [Celebrex] group...and was most pronounced in the group of patients not taking ASA (aspirin)" as a cardiovascular protective drug. In these patients the rate of myocardial infarction was also highest in the celecoxib group (0.2 percent) compared with users of the other two drugs (0.1 percent). For all patients, on and off aspirin, there was a higher incidence of atrial fibrillation, a cardiac arrhythmia, in the celecoxib group than in either of the other two groups, again more pronounced in the group not taking aspirin. The author concluded by stating that "the data do not exclude a less apparent pro-thrombotic [blood clot forming] effect of celecoxib, reflected in the relative rates of cardiac adverse events related to ischemia."
Once again, a seemingly magic bullet appears to have self-destructed as research reveals the larger context in which it operatesÑthe risks as well as the benefits. The benefits of COX-2 inhibitors as far as reducing GI toxicity appear to have been grossly exaggerated and oversold. Years after the research on these benefits was done, a rapid accumulation of evidence on risks is occurring. For an important enzyme which is close to ubiquitous in the body, it is less than surprising that blocking its activity in one part, the gastrointestinal tract, must be balanced against the apparently harmful effects of blocking its critical functions in other parts of the body.
We strongly urge the retention of the NSAID class-warning label for these drugs, possibly adding that there is no evidence of statistically significant reduction in serious GI toxicity for celecoxib. This should take the form of a box warning (for all the drugs) which should be placed at the beginning of the label.
A second box warning about cardiovascular toxicity needs to be added. It should warn of the lack of platelet aggregation inhibition of the drugs which protects those at risk from an increased occurrence of heart attacks. In addition, the evidence which is rapidly accumulating about the heart damage caused by these drugs must be mentioned in this cardiovascular box warning.
What You Can Do
In light of the above discussion, we continue to advise the patient-protective five-year-rule for these drugs, as we do for all other new drugs that are not breakthroughs. Do Not Use.