Critics say flaws in Celebrex study undermine results
NEW ORLEANS — A long-awaited study on painkillers called nonsteroidal anti-inflammatory drugs, the most widely prescribed class of drugs in the world, has concluded that the three most commonly used carry a similar risk of cardiovascular complications. Yet critics say the study was too flawed to fairly compare them.
Concerns about a type of NSAID called COX-2 inhibitors peaked in 2004 when the drug Vioxx was withdrawn from the market — a decision steeped in scandal because manufacturer Merck & Co had initially hidden data that would reveal the drug’s cardiovascular risks. A second COX-2 inhibitor, Pfizer Inc.’s Celebrex, was allowed to remain on the market with the condition that Pfizer conduct a study to prove that Celebrex was no worse than two older NSAIDs, naproxen and ibuprofen.
The study lasted 10 years and enrolled more than 24,000 patients, but faced challenges. Doctors in European Union countries would not participate because they were worried about Celebrex’s safety. Also, scientists hoped arthritis patients at high risk of heart disease would volunteer. But the study was ultimately dominated by those at low risk. And before it was over, about two-thirds of participants dropped out because their assigned drug was not controlling their pain.
A final analysis was presented November 13 during the annual American Heart Association Scientific Sessions, and published online in the New England Journal of Medicine. After taking the drug at recommended doses for a minimum of 18 months, 4.2 percent of those on Celebrex had a major cardiac event — heart attack or stroke, for example — compared with 4.3 percent of those taking naproxen and 4.8 percent taking ibuprofen. The percentage of people who died from cardiovascular disease was highest for those taking naproxen, at 1.1 percent, although the differences among the three drugs were not statistically significant, said Steven Nissen, chair of cardiovascular medicine at the Cleveland Clinic who led the study.
Kidney problems, a known risk for taking NSAIDs, occurred in 0.7 percent of the Celebrex group, 0.9 percent of the naproxen group and 1.1 percent of those taking ibuprofen. Of the three NSAIDs, Celebrex had the lowest risk of gastrointestinal complications, not a surprise because avoiding those complications was a key reason why the drug was developed.
Given the dangers of Vioxx, many doctors had dismissed Celebrex even before the Pfizer-funded study began, said Nissen. “Everybody thought they knew the answer.” And even he was surprised when Celebrex fared better for cardiovascular risk than the other two, and said the results should be reassuring to patients taking the drug, which is now available as a generic.
But the study has its skeptics. “Does it give us answers? No, and that’s the tragedy,” said Garret FitzGerald of the Institute for Translational Medicine and Therapeutics at the University of Pennsylvania. Critical of the study design since its inception, FitzGerald wrote a commentary on the trial, posted online November 13 in Circulation.
FitzGerald anticipated that, given the concerns about cardiovascular side effects, doctors would hesitate to enroll their high-risk patients in the study, and that it would need to go on longer to collect enough data. “This is a trial in which most people dropped out before it was finished,” he said. Also, doctors were allowed to bump up the doses of naproxen and ibuprofen if patients needed it for pain control, but the amount of Celebrex remained fixed. “It wasn’t a fair comparison,” he said.
The data also don’t account for which patients were taking aspirin when the study began, said Elliott Antman, a cardiologist at Brigham and Women’s Hospital in Boston. Aspirin, which acts as a blood thinner, can lower the risk of heart attack. Because of how the drugs work, naproxen and ibuprofen — but not Celebrex — can interfere with the benefits of aspirin. Having more patients at high risk of heart attack would have increased the study’s ability to detect risk, he adds. “This was a trial that was supposed to compare the outcomes in high-risk cardiovascular patients, but we actually still have that question,” Antman said. And given the time and expense of this trial, there is unlikely to be another. New, less problematic pain drugs, he said, are “an urgent clinical need.”
S.E. Nissen et al. Cardiovascular Safety of Celecoxib, Naproxen, or Ibuprofen for Arthritis. New England Journal of Medicine. Published online November 13, 2016. doi: 10.1056/NEJMoa1611593
G.A. FitzGerald. ImPRECISION: Limitations to Interpretation of a Large Randomized Clinical Trial. Circulation. Published online November 13, 2016. doi: 10.1161/CIRCULATIONAHA.116.026324
Prospective Randomized Evaluation Of Celecoxib Integrated Safety Vs Ibuprofen Or Naproxen (PRECISION). ClinicalTrials.gov Identifier: NCT00346216