Journal Slams Vioxx Study
Questions about 3 heart attacks not included in data analysis could hurt Merck's legal defenseDec 9, 2005 | Los Angeles Times
The New England Journal of Medicine on Thursday accused researchers funded by Merck & Co. of understating the health risks of the painkiller Vioxx by withholding adverse data from a 2000 report on the drug.
In a rare "Expression of Concern" published online, editors of the journal said the Merck scientists failed to report three nonfatal heart attacks among the Vioxx users. The total number of heart attacks among the drug users was 20, they said, not the 17 reported.
Experts disagreed on how medically significant the three heart attacks were but agreed that the decision not to include the cases could bolster claims in thousands of lawsuits against Merck that the company was deliberately withholding data about the drug's health risks. Merck voluntarily withdrew the drug from the market in September 2004.
In a statement released late Thursday, Merck said the three heart attacks were disclosed in information sent to the Food and Drug Administration in 2000 and in news releases issued at the time of the study.
The three heart attacks were not included in the report, the statement said, because they occurred "after the pre-specified cutoff date" for analysis of the study results.
The report "fairly and accurately described the results of the study," Merck said in the statement.
But the journal's executive editor, Dr. Gregory D. Curfman, rejected Merck's explanation.
"It would have been their responsibility to provide us with an update after the cutoff time," especially because the three heart attacks occurred shortly after the study's end, he said. "Those data could have easily been put in the article, but they never were."
The editors also said the study's authors withheld more relevant data about strokes and other heart problems linked to the drug, producing inaccuracies and deletions that "call into question the integrity of the data."
The editors said they had asked the Vioxx report's authors to submit a correction and that two of the lead authors had agreed.
The accusations by the prestigious New England Journal of Medicine could hurt Merck's ability to defend itself in the 6,500 lawsuits blaming Vioxx for heart attacks, strokes and deaths. Analysts have estimated that the cases could eventually cost the company $50 billion.
"This study was one of the major studies on which Merck premised its defense, and one of its leading witnesses is an author on the study," said Carl Tobias, a law professor at the University of Richmond. The data dispute "erodes the credibility of the witness and the company."
Dr. Alise Reicin, Merck's vice president for clinical research, testified this week in a Houston Vioxx trial that the company never misled doctors and the public about studies linking heart attacks to Vioxx.
Jurors began deliberations Thursday afternoon in that lawsuit, brought by the widow of 53-year-old Richard "Dicky" Irvin of St. Augustine, Fla.
The Houston trial will be the third to reach a jury. In the first, a jury in a small town near Houston awarded $253 million to the widow of a patient who died after taking the drug. Merck won the second trial last month in a case brought by an Idaho postal worker who had taken Vioxx for two months.
Before Merck withdrew Vioxx from the market, it was one of the most widely prescribed and expensive drugs being sold.
It is part of a class of drugs known as Cox-2 inhibitors, which are designed to reduce pain without triggering the gastrointestinal bleeding produced by aspirin and other nonsteroidal anti-inflammatory drugs, such as naproxen.
The 2000 study, involving more than 8,000 patients, did show less bleeding with Vioxx, but it also showed a 220% increase in the risk of heart attacks and other cardiovascular events for Vioxx compared with naproxen. The actual number of events in both groups was low, less than 1%.
Merck scientists attributed the difference in risk to a protective effect of naproxen, a contention that many experts disagreed with.
If the three heart attacks are included in the data, Vioxx produced a 300% increase in risk.
That increase in risk "does not materially change any of the conclusions of the article," the Merck statement said.
Dr. James F. Fries, a rheumatologist at the Stanford School of Medicine, agreed. "From the standpoint of scientific proof, [these extra cases don't] make any difference," he said. The cases "don't change the knowledge that the drug can cause" problems.
But, he added, "what it might raise in some people's minds was a suspicion that those cases had been suppressed."
But Curfman said the three extra cases would "definitely, materially change the conclusions" of the study.
All three of the heart attacks occurred among people who were not otherwise considered at risk for heart attacks, he said.
The authors "wanted to conclude" that the only hazards were among people who were already at risk, he said. "But that conclusion is no longer sustained" because the heart attacks were in people with no risk.
"Even the low-risk people were at risk now," said Mark Robinson, a Newport Beach lawyer representing Vioxx plaintiffs. "That is the key to this whole thing. And that is what they [the journal editors] are upset about. That is the significance."
Curfman said the journal editors had suspicions about the article more than a year ago, when they reviewed their files days after Merck removed Vioxx from the market.
At that time, managing editor Steven Morrissey discovered a diskette that included a pre-submission version of the manuscript. He determined that some heart attack data had been edited out two days before the manuscript was submitted to the journal. That included a table labeled "Cardiovascular Events" that was empty of data, except for a footnote.
"It made us suspicious, but there wasn't enough for us to take any action," Curfman said.