Arthritis-drug maker withheld risk data, UW study finds

The maker of the controversial arthritis drug Vioxx knew of significant risks associated with the medicine at least three years before a massive nationwide recall, but withheld the information from federal officials and later downplayed the death rates, according to an analysis by University of Washington professors.

Their article, which appeared Tuesday in the prominent Journal of the American Medical Association (JAMA), contends that Merck, which pulled Vioxx from the market in 2004, had concluded internally as early as 2001 that people at risk of Alzheimer's disease who were enrolled in two Vioxx clinical trials were dying at three times the rate of those taking a placebo.

The UW researchers have used court documents from a flood of lawsuits against Merck to contend that the company switched calculations to mask the drug's potential dangers.

A Merck attorney, Jim Fitzpatrick of New York, countered Tuesday that the company disclosed all its clinical results thoroughly. He said "we completely disagree" with allegations that Merck misrepresented its findings.

The article is one of a pair of studies in this week's issue of JAMA to take an unusually detailed look at the clash of corporate interests and scientific integrity over Vioxx, the subject of the most expensive drug recall in history. Vioxx is the brand name for rofecoxib, an anti-inflammatory drug.

A companion article by researchers at New York's Mount Sinai School of Medicine found that people on Merck's payroll ghost-wrote dozens of studies on Vioxx and recruited prominent academic investigators with little involvement with the studies or financial ties to Merck to attach their names to the work.

In an accompanying editorial, JAMA's top editors said such manipulation of research calls for "drastic actions." The editorial argues that researchers, regulators and editors of medical journals such as JAMA are complicit, but it singled out doctors who lend their names on scientific papers as "unprofessional and demeaning to the medical profession."

Merck is facing about 27,000 Vioxx lawsuits. In November, the New Jersey company agreed to settle a majority of the cases for $4.85 billion without admitting guilt.

The co-author of the UW paper, Richard Kronmal, a professor of statistics at the UW, has served as a paid expert witness for Vioxx plaintiffs. The lead author is Bruce Psaty, a UW professor of medicine who has not been involved in the suits.

Vioxx is a prescription arthritis drug and painkiller that was promoted as a gentler alternative to drugs such as aspirin. Merck withdrew it in September 2004 after disclosing that a study found it doubled the risk of heart attacks and strokes.

But information was available well before the recall that showed that Vioxx posed serious potential harm to subjects in clinical trials.

As early as 2000, the Food and Drug Administration (FDA) was given information that showed Vioxx posed far greater risk of heart attacks and strokes than Merck let on in published academic papers, according to subsequent analysis by independent researchers.

In 2005, the New England Journal of Medicine, another leading medical journal, publicly accused Merck of undercounting Vioxx deaths in a 2000 report. Six months later, the journal retracted part of a 2004 paper that claimed Vioxx caused heart damage only when taken for more than 18 months.

Some of the early evidence prompted pharmacists from Seattle's Group Health Cooperative to withhold Vioxx from the list of drugs it covers.

In the latest JAMA article, the UW researchers compared internal Merck memos about two Vioxx studies involving patients at risk of dementia against information the company submitted to the FDA and published later.

In April 2001, a Merck statistician concluded that 34 people in the Vioxx groups had died, compared with 12 in the two placebo groups. But in its submissions to the FDA three months later, Merck used different methods to figure the mortality rate, coming up with 29 deaths among Vioxx takers and 17 in the placebo groups.

Merck apparently did this, in part, by omitting some of the deaths that had occurred after the subjects quit taking the drug, said Psaty, who also is an investigator at Group Health.

Merck's internal memo included all the people enrolled in the studies. But it gave the FDA results only for subjects who stuck with the trials, who generally tend to be healthier and differ in other ways from those who dropped out.

"There was great disparity between what [Merck] knew and what they made public in a timely fashion," Psaty said. "They were not being forthright."

Kyung Song: 206-464-2423 or ksong@seattletimes.com