New York – With all the sturm and drang surrounding the release of new data from the study that caused Vioxx to be pulled from the market, you’d think that there would be some answers about the drug’s real public health risks.
But clear answers aren’t coming. For one thing, any scientific debate is colored by the more than 11,000 lawsuits mounting against the drug’s maker, Merck. For another, the APPROVe, the 2,600-patient study that led Merck to pull Vioxx in September 2004, was not designed to provide clear answers to questions about the drug’s potential to cause heart attack and stroke. So now, even as new data emerge from the study, clear answers seem to be slipping further away.
Instead, the Vioxx mess is signaling big changes for how clinical trials will be conducted. More information about studies will likely be published before they are started, so that scientists can debate the way they are designed before they are signed up. Moreover, drug firms are increasingly pushing to eventually make more results publicly available on Web sites. But those changes are coming too late for Merck, and now the company and its critics are left in a public battle over exactly what the results of the study mean.
“A lot of people are talking off the cuff here,” says Robert Bresalier, APPROVe’s first author and a cancer prevention expert at M.D. Anderson Cancer Center. “I’ve never seen anything that’s generated so much emotion and speculation and comments from people I normally think of as fairly reputable that I think are to some extent not based on hard data.”
APPROVe was designed to test whether Vioxx reduced the risk of colon polyps, but instead it showed that patients who took the drug for at least 18 months had double the risk of heart attack and stroke as those who took placebo. Bresalier says the authors had to rush to get the paper published in the New England Journal of Medicine ahead of a February 2005 meeting by U.S. Food and Drug Administration experts.
By showing that it took 18 months for Vioxx to cause problems, APPROVe gave Merck a built-in legal defense. But critics say a 108-page data package Merck sent to the FDA two weeks ago put the 18-month defense on shaky ground. The existence of this data package was first reported by NPR, and a copy was provided by Merck.
One of the shortcomings of the New England Journal of Medicine paper from last March was that it only followed patients for two weeks after they had stopped taking Vioxx. But Merck had collected data for longer than two weeks—and analysis with these data seem to cut the key 18-month period down to as little as three months.
Cleveland Clinic cardiologist Steven Nissen, a longtime Vioxx critic, says that every analysis shows an earlier risk except the one that was used in the New England Journal. He worries that Merck may have chosen to present the data in a way that would be best for its legal defense. “It makes you very uncomfortable,” he says. However, says, the studies don’t clearly show exactly when the risk from Vioxx begins. “I think that we don’t know,” he says.
Merck argues that using the graphs provided in its data package to pinpoint exactly when risk emerges is not scientifically valid. Bresalier says he does not see the risk emerging earlier but said he and his co-authors will be looking at the new data package in detail over the next week. He says that he only got it recently. However, he also says that too much may have been made of the 18-month cut-off. “Nothing fell out of the air magically on that date,” says Bresalier.
The New England Journal, which has published editorials expressing concern over another key Vioxx study, says it is examining APPROVe but that it will take considerable time to look over all this data. “This underwent extensive review,” says Bresalier. “They didn’t have to publish it if they didn’t think it was reasonable.” He also says that the authors, Bresalier and John Baron of Dartmouth Medical School, not Merck, wrote the paper.
This controversy that has besieged the major studies of Vioxx is likely to increase the pressure for airing more decisions about designing clinical trials in the open. It would be much more difficult for critics to cry foul if the APPROVe authors had disclosed ahead of time that they would be tracking heart attacks and strokes for only two weeks after patients stopped using Vioxx.
Already, the pressure is on for more disclosure. On Friday, the World Health Organization called for all clinical trials to be registered in a public database as soon as they begin. Such a move could even eventually save drug firms money by revealing why certain medicines fail. For instance, a half-dozen medicines designed to treat diabetes and high cholesterol at the same time have been torpedoed by side effects, and more access to data might help companies prevent history from repeating itself.
But issues relating to Vioxx have moved out of the realm of science. The 18-month defense hasn’t helped Merck; it has so far lost one of two long-term use cases, and two of four short-term cases. At the same time, the Vioxx legal battle may scuttle any hope of getting better data from already conducted studies. Bresalier says that a legal injunction prevents the APPROVe authors from doing new analysis on their data.
And just because scientists are arguing does not mean jurors will care.