Two prominent medical researchers are taking issue with Merck's conclusion that a follow-up study of patients who took the painkiller Vioxx shows that the drug posed no "statistically significant" risk to the heart once people stopped taking it.
Merck announced its conclusion on Thursday in reporting on the health of patients in a Vioxx study a year after that study was suspended in September 2004.
The company ended that study because a disproportionate number of people taking the drug were having strokes or heart attacks. At the same time, Merck took Vioxx off the market. Since then, lawyers have filed thousands of lawsuits on behalf of former Vioxx patients or their families.
The two independent medical researchers said in interviews yesterday that Merck might have been premature in reaching the conclusion announced Thursday, in part because the patients in the study needed to be monitored for longer than one year.
Merck acknowledged Thursday that the number of strokes or heart attacks among patients during the one-year follow-up indicated a 64 percent higher risk of cardiac problems among those who had taken Vioxx, compared with a placebo. But the company said the overall numbers of problems were too small to give them statistical significance.
The independent researchers, though, said it was too soon to draw any conclusions about whether Vioxx's dangers to coronary arteries ended once a person stopped taking it.
"It would be very important to continue to follow these patients in Year 2 and 3, to see if the risk continues," said Dr. Steven E. Nissen, chief of cardiology at the Cleveland Clinic.
The 64 percent higher risk that Merck acknowledged in the follow-up study, Dr. Nissen said, "has rather important scientific implications, because it suggests there was some kind of permanent or longstanding injury to the artery that makes it susceptible to these kinds of continuing events."
Dr. Bruce M. Psaty, a professor of medicine and epidemiology at the University of Washington, said yesterday that Merck should follow the patients until it knew for certain whether Vioxx caused long-term heart damage.
"From the public health perspective and the point of view of duty to human subjects, following them would be appropriate," Dr. Psaty said. "That would be an honorable thing to do."
In a statement yesterday, Merck said it had not decided whether to continue tracking the patients. "We have questions about whether scientifically rigorous data can be obtained in such a setting," the company said. "We will discuss it with regulatory agencies and the study's steering committee before making a final decision."
The issue of whether Vioxx, Merck's once-popular painkiller, can cause heart attacks and strokes even after patients stop taking it is crucial to public health and the company's future. Twenty million Americans took Vioxx from 1999 to 2004, when Merck stopped selling it. Epidemiologists estimate that the drug may have caused 100,000 heart attacks during the five years it was on the market.
Merck already faces at least 11,500 lawsuits, covering 23,000 patients, from people who say Vioxx caused their heart attacks and strokes. The company has vowed to fight each case but has lost three of the five to reach trial so far, with juries awarding verdicts of more than $10 million each time.
The data Merck released Thursday involved patients who had taken Vioxx as part of a clinical trial called Approve, which was designed to determine whether Vioxx — a painkiller approved for arthritis relief — could prevent colon polyps. In the Approve trial, 2,600 patients took either Vioxx or a placebo for three years. After researchers noted that patients taking Vioxx had nearly double the risk of heart attacks, other heart problems or stroke, the trial was ended.
But the risk persisted even after patients stopped taking Vioxx, according to the results Merck released Thursday. In the year after they stopped taking Vioxx, 28 patients had heart attacks, strokes or other cardiovascular problems, compared with 16 patients who had taken a placebo. Slightly more patients took Vioxx than the placebo, but even after adjusting for that, previous use of Vioxx was associated with a 64 percent increase in risk.
When Dr. Peter Kim, Merck's chief scientist, described those results on Thursday, he said they were "statistically insignificant."
Merck said the study had shown with 89 percent certainty that Vioxx caused heart problems even after patients stopped taking it. Scientists do not consider a study definitive unless it demonstrates a result with 95 percent certainty, a figure they consider statistically significant, or unlikely to be produced by chance.
Merck's interpretation of the data was narrowly correct but misleading, Dr. Nissen said.
Because the study followed patients for only one year, the number of patients who suffered heart problems was relatively small in both groups. As a result, the 64 percent higher rate in the Vioxx group could have been a result of chance — although it probably was not, Dr. Psaty said. "This is closer to a persistent finding than not."
If people who took Vioxx continue to have more heart problems than those who took a placebo — even for just a few more months — the difference between the groups will almost certainly reach statistical significance, Dr. Psaty and Dr. Nissen said.
In that case, Merck and former users of Vioxx would be faced with the possibility that Vioxx can cause heart problems for years after patients stop taking it. So far, very few patients have asserted that they have suffered heart problems after they stopped taking Vioxx. If that changes, the company could face a much higher number of lawsuits.
