News

The New England Journal of Medicine publicly alleged Thursday at least two of the authors of a major Canadian-led study on the former blockbuster drug Vioxx withheld data on adverse events from the journal.

In a stinging editorial released online, the prestigious journal did not specify which of the study authors withheld the information.

But Dr. Gregory Curfman, executive editor of the journal, revealed in an interview that the two were employees of Merck & Co., the company which made the drug and funded the study.

Dr. Alise Reicin, Merck's vice-president for clinical research and Deborah Shapiro, were the only Merck employees among the 12 authors of the study, published in the Nov. 23, 2000 issue of the journal.

The lead author of the study was Dr. Claire Bombardier, a prominent rheumatologist at Toronto's Mount Sinai Hospital. She defended the data disclosure in an e-mail exchange Thursday.

"I believe that the VIGOR paper appropriately disclosed the data as per the pre-specified plan of analysis," Dr. Bombardier insisted.

Dr. Curfman said the journal hoped the editorial would serve as a warning of what happens when researchers ignore their responsibility to report honestly and fully.

"We hope by issuing the statement of concern that it will be a stark reminder to the science community, to the scholarly community, that if we don't have this principle of trust, that the whole scholarly process is just going to collapse into rubble," he said.

The discrepancy came to light when Dr. Curfman recently gave a deposition in a case brought against Merck by the widow of a man who died while taking Vioxx.

In the process, he recently received a memo written by one of the Merck authors to the other Merck author about the study. The memo indicated both knew months before publication of three additional heart attacks that hadn't been factored in to the study findings.

The memo was dated July 5, 2000 — two weeks before the authors submitted the first of two sets of revisions to the journal.

"So we know that both Merck authors had knowledge of the contents of this memo on July 5 and there was quite a lot of information in there about cardiovascular end points," Dr. Curfman said.

And what of Bombardier and the other authors?

"At the very least, it would give the appearance that there wasn't a really good process for how they dealt with the manuscript," Dr. Curfman said.

The editorial, which Dr. Curfman co-authored with the journal's editor-in-chief and managing editor, also charged that other relevant adverse cardiovascular events were deleted from the original manuscript two days before it was submitted for publication.

Analysis of a computer diskette containing the study revealed those changes were made on a Merck computer, Dr. Curfman said.

Withholding the data had the effect of downplaying the already significant difference in cardiovascular risks between Vioxx and naproxen, a generic pain relief medication to which Vioxx was being compared in the study, the journal said.

"Taken together, these inaccuracies and deletions call into question the integrity of the data on adverse cardiovascular events in this article," stated the editorial, titled an Expression of Concern.

"We have asked the authors to submit a correction to the journal."

The journal told the authors of its concerns last Monday, but did not offer to let them issue their response at the same time as the editorial, Dr. Curfman said.

Bombardier said the authors were only informed at noon on Thursday that the journal was releasing the editorial. It was placed on the journal's website three hours later.

"The authors of the paper will be preparing a response to the editorial but have not had the time to finalize it yet," she said.

Merck defended the research in a brief statement issued from its U.S. headquarters in Whitehouse Station, N.J. It reiterated Bombardier's point that the three heart attacks had occurred after a pre-specified cut-off date.

"Nevertheless, these additional events were disclosed to the FDA in 2000, presented publicly at the FDA's advisory committee in February 2001, and included in numerous press releases subsequently issued by Merck," the release said.

"We also note that these additional events did not materially change any of the conclusions in the article."

Researchers who know Bombardier came to her defence Thursday, including Dr. Andreas Laupacis, president of Toronto's Institute for Clinical Evaluative Sciences.

"I have a huge amount of respect for her. I have no reason whatsoever to think that she's anything but incredibly honest," said Laupacis, who said the allegations underscore the need for total transparency of drug studies.

His views on Bombardier were echoed by Dr. Gurkirpal Singh, an expert on cox-2 inhibitors — the drug class to which Vioxx belongs — who teaches at Stanford University in California.

"With what I know of her personally, I would have to believe she did not know about this," said Singh, who has been highly critical of Merck's pattern of disclosure of Vioxx data.

He was less willing to give the company the benefit of the doubt, dismissing its argument that the data in question weren't submitted because they arose after the cut-off date.

"This is semantics, isn't it?" Dr. Singh asked.

"Let's say that there's this pre-specified cut-off date. And after the pre-specified cut-off date, say 20 heart attacks occur in the naproxen group so that the difference between Vioxx and naproxen disappears. Do you think the authors wouldn't have reported that?"

An analysis contained in the New England Journal editorial showed the three additional heart attacks meant the risk of heart attack among people taking Vioxx was five-fold higher than those on naproxen, not four-fold higher, as the published paper said.

"This makes the villainy even blacker," Arthur Schafer, director of the University of Manitoba's ethics centre and a harsh critic of the marketing of Vioxx and other cox-2 drugs, said of the allegations in the journal editorial.

"But you have to ask: How come millions of prescriptions were written for Vioxx, given that the published results showed a 400-per-cent increase (in heart attacks)?"

Still, Mr. Schafer applauded the journal's action.

"There are massive career and financial incentives for researchers to come up with industry-favourable results," he said in an interview.

"So what the New England Journal is doing, and maybe what others will follow their lead in doing, is they are providing some counterbalancing disincentives: 'If you are going to be a scientist whose work lacks integrity, you are going publicly to be humiliated.'

"And I think that's all to the good."

The editorial said the journal learned of the undisclosed adverse events in 2001 when the U.S. Food and Drug Administration released additional data on Vioxx, a one-time giant in the pharmaceutical world that was pulled from global markets in September of 2004.

The journal said until last month, it had believed the events came to light too late to be included in the article. But it has since learned at least two of the authors knew of the events 41/2 months before the manuscript was published.

Sales of Vioxx were suspended when data showed it increased the risk of heart attacks and strokes in some users.

The drug is currently the subject of a number of lawsuits, including one which went to the jury Thursday. But expert panels in both the U.S. and Canada have advised drug regulators Vioxx should be allowed back on the market for use by certain low-risk patients.