The New England Journal of Medicine said it has discovered that data about three Vioxx patients who suffered heart attacks were excised from a crucial study sponsored by Merck & Co., making the painkiller look safer than it should have.
In a highly unusual “Expression of Concern” published yesterday afternoon, the medical journal cited “inaccuracies and deletions” in the 2000 study that it said “call into question the integrity of the data.” Editors said the study’s main author was working on a correction.
The medical journal’s allegation could prove damaging to Merck’s credibility in the estimated 6,500 lawsuits it is facing on behalf of people who claim they suffered heart attacks as a result of taking Vioxx. Merck has vowed to fight each case individually. The company withdrew the widely used painkiller from the market in September 2004 after a study showed an increased risk of heart attacks after 18 months of use.
In a statement issued late yesterday, Merck said the study “fairly and accurately” described the results of the trial. The company said the three omitted heart attacks weren’t included in the article submitted for publication because they occurred “after a pre-specified cutoff date.”
The company said data about the heart attacks were provided to the Food and Drug Administration, presented to an FDA advisory panel, disclosed to physicians and included in news releases later issued by the company. I t also said including the three additional heart attacks “did not materially change” the conclusions of the study, dubbed Vigor, for Vioxx Gastrointestinal Outcomes Research study.
Merck said it learned of the journal’s editorial “recently” and hasn’t “had an opportunity to formally respond in the New England Journal of Medicine given the timing of its publication.” It said it looks forward to providing a more complete response to the editorial.
Last night, Merck declined additional comment on the journal’s allegation, which came shortly after the third Vioxx trial, in Houston, went to the jury. “We have our statement that you read,” said senior vice president and general counsel Kenneth C. Frazier. “We have a jury that’s deliberating. And the judge has reiterated his request that the parties not comment. And we are going to honor that request.”
The disclosure doesn’t change what’s now known about Vioxx’s risks. Researchers and lawyers involved in Vioxx litigation already have been using the complete data provided to the FDA.
But plaintiffs attorneys said they would use the disclosure about the withheld data to attack Merck’s credibility. “We will tell jurors they must consider the fact that in the most pivotal cardiovascular studies, there was a deliberate attempt to mislead the scientific community,” said Samuel L. Davis, partner at Davis, Saperstein & Salomon in Teaneck, N.J., who represents plaintiffs in many Vioxx cases. Merck lost the first case in Texas state court and won the second in New Jersey state court.
VIGOR Study Data
Study Group Person-Years of Exposure No. of Myocardial Infarctions Relative Risk As Published
Rofecoxib (Vioxx) 2315 17 4.25
Naproxen* 2316 4
Including Three Missing Infractions
Rofecoxib (Vioxx) 2698 20 5.00
Naproxen 2699 4
Naproxen is commonly marketed as Aleve.
Source: New England Journal of Medicine
The disclosure also drew some strong reactions in the research community. “I think this is a remarkable series of events and proves deception,” said Drummond Rennie, a deputy editor of the Journal of the American Medical Association and professor at University of California San Francisco.
Merck shares, which had shed 61 cents, or 2O/0, to $29.68 in 4 p.m. New York Stock Exchange composite trading fell an additional 86 cents, or 2.9%, in after-hours trading to $28.82 following the late-afternoon disclosure.
Published in November 2000, the Vigor study has been a central piece of evidence debated in the three Vioxx trials conducted so far.
The study, which began in early 1999 and was paid for by Merck, wasn’t specifically designed to look at heart-attack risk, but at gastrointestinal problems. Its main goal was to determine whether regular use of Vioxx would cause fewer ulcers and gastrointestinal bleeding in rheumatoid arthritis patients than another widely used painkiller, naproxen, and thus give Vioxx a marketing advantage. The study did conclude that Vioxx performed better.
In its editorial, the medical journal said it learned late last month that the data had been withheld when its executive editor was deposed by plaintiff’s attorneys. It said it had learned that at least two authors of the study knew about the three omitted heart attacks in July 2000, more than four months before the article was published and before the first of two revisions of the article were submitted to the journal. It didn’t identify the two authors.
The medical journal said it had discovered from a computer diskette submitted along with the article that some data were deleted from the manuscript just two days before it was initially submitted for publication in May 2000.
In an interview, Gregory Curfman, executive editor of the New England Journal, said that while the overall findings of the study weren’t invalidated, one of its conclusions, that the cardiac risk from Vioxx was seen only in the higher-risk subgroup of patients, did become invalid. He said all three new cases came from a different group of patients who were at lower cardiac risk.
Adding the omitted heart attacks, the journal said in its editorial, means that 20 of 2,315 patients taking Vioxx in the study suffered heart attacks compared with four taking naproxen. The increase means that Vioxx patients were five times as likely to suffer a heart attack as those on naproxen. The study as published reported the relative risk as 4.25 times as high.
“It turns out to be very important,” Dr. Curfman said, adding that the study wasn’t crucial for Merck in winning FDA approval of the drug but “was a very important trial for them to promote Vioxx.”
Dr. Curfman said that editors at the journal pulled out the file on Vigor a few days after the Vioxx recall last year. Stephen Morrissey, the journal’s managing editor, “went through the file and found” a diskette that was sealed inside a brown envelope. The editors accessed the diskette and found three versions of the study; two that the editors has seen before on paper and one that they hadn’t.
Editors also noticed that some data had been deleted from the text of the manuscript and from a table titled “cardiovascular events,” Dr. Curfman said. Using word-processing software to track down when changes had been made and by whom, the editors found out the data had been taken out two days before the article was submitted to the journal, he said. He said they determined the deletions were made by someone working from a Merck computer.
However, the editors didn’t contact the study’s authors to ask for an explanation, believing they didn’t have enough information to make any conclusions, Dr. Curfman said.
“All we had was a skeleton of a table,” he said of the deleted information. “We knew the FDA already had the data and that was part of the public record,” he said, adding that the editors assumed the three additional heart attacks happened among patients after the article was submitted. “We believed they were late events,” he said.
It wasn’t until Dr. Curfman gave a deposition on Nov. 21 to plaintiffs attorneys, who had subpoenaed the documents, that editors learned the matter was far more serious. “We became aware of an internal Merck memo that confirmed that at least some of the authors knew [of the extra heart attacks] as early as July 5, 2000,” before the study was published by the journal, Dr. Curfman said.
From the Trial
- Jury Instructions: See the 12 questions facing the nine-member jury
- Painful Fallout: Tracking Merck’s share price since Vioxx’s launch
- Timeline: Vioxx’s move from the market to the courthouse
- Glossary: Explanations of the key terms used in reporting on Vioxx
The deposition was conducted by Donald C. Arbitblit, a partner in Lieff Cabraser Heimann & Bernstein, LLP, according to a copy reviewed by The Wall Street Journal.
Asked about Merck’s statement about the cutoff date, Dr. Curfman said, “They never told us about an arbitrary cutoff.”
The editors at the medical journal said they have been in touch with the study’s main author, Claire Bombardier, a researcher at the Institute for Work and Health at Mount Sinai Hospital and the University Health Network in Toronto, and that the authors are working on a correction to be submitted to the journal soon. Efforts to reach Dr. Bombardier by telephone and email weren’t successful.
Two of the study’s co-authors are Merck employees. They are Alise Reicin, who is vice president of clinical research, and Deborah Shapiro, a biostatistician. Merck said Dr. Reicin and Dr. Shapiro weren’t available to comment.
The disclosure by the New England Journal is just the latest sign of the increasingly tough position of medical journals, which stand at a pressure point between drug-company promotion and scientific research. Increasingly, journal editors say, they struggle with evidence that the industry-backed studies they publish may not always tell the whole story — and are often crafted to serve as part of sophisticated marketing campaigns.
The controversy over the highly promoted class of drugs called the Cox-2 inhibitors, which includes Vioxx and Pfizer Inc. rival Celebrex, has been a catalyst for such concerns. In an earlier case, editors at JAMA toughened their review policies after evidence surfaced that a study they had published in 2000, showing positive results based on six months of use of
Pfizer’s Celebrex, had left out results that in some cases extended to 12 months. When all of the data were included, the findings were less positive. Pfizer has said that many patients discontinued the treatment after six months, making it hard to analyze the additional data.
Journals are becoming more aggressive in trying to police the studies they receive. But journal editors say they lack the power and resources to serve as auditors of the studies they publish. “We have to be alert to anything that might be part of a marketing campaign,” said Fiona Godlee, editor of BMJ, formerly called the British Medical Journal, speaking before
the New England Journal editorial. “It’s an uncomfortable reality we have to live with and do our best to mitigate the worse offenses.” But, she said, “The whole enterprise is based largely on trust.”
Yesterday, in closing arguments at the third Merck trial in Houston, attorneys painted opposing pictures of the drug maker’s handling of the painkiller, what it knew about Vioxx’s cardiovascular dangers and whether Vioxx contributed to the death of a man who took it for less than a month.
Evelyn Irvin Plunkett alleges that Vioxx caused the 2001 fatal heart attack of her husband, Richard “Dicky” Irvin who took the drug for less than a month to treat back pain. In considering the case, the jury of five men and four women will have to answer 12 questions about the roles of Merck, the doctor who prescribed Vioxx to Mr. Irvin, Mr. Irvin himself and the dangers posed by Vioxx. I f Merck is found liable; jurors will have to consider what would be fair compensation to Mr. Irvin’s children and widow for suffering and the loss of his income and companionship.
The plaintiffs attorney, Andy Birchfield, argued that a verdict against Merck would send the pharmaceutical industry a critical message about its conduct. He told jurors the case will have “far-reaching implications” on how drug companies act.
Risks of Vioxx, he said, “were known and known early,” he said. Merck “could take the high road to patient safety or the low road to sales.”
In his closing remarks, Merck’s attorney, Philip S. Beck, argued that Vioxx didn’t play a big role in Mr. Irvin’s heart attack. He told jurors that a key question in the case was whether Mr. Irvin had a plaque rupture, or a breaking open of plaque in his artery that led to his heart attack. Mr. Beck said two witnesses for the plaintiffs testified that if Mr. Irvin’s heart attack was caused by a plaque rupture, that would rule out a role for Vioxx. He said that one of the experts, a pathologist, was inconsistent in his testimony about whether or not Mr. Irvin had a plaque rupture.
Mr. Beck also told jurors that Merck acted responsibly and appropriately and asked the jurors to find in the company’s favor, saying, “Our good name is at stake.”
Merck lost the first Vioxx trial earlier this year when a Texas jury awarded $253.4 million to the widow of a man who died while taking Vioxx. Under state damage caps, that award will likely be reduced to $26 million. The Whitehouse Station, N.J., drug company came back to win a case tried in New Jersey Superior Court last month. All three cases have involved people who took the drugs for less than 18 months.
— Anna Wilde Mathews and Scott Hensley contributed to this article.