Boston, MA - In a letter published online February 22, 2006 in the New England Journal of Medicine, Dr Claire Bombardier (University Health Network, Toronto, ON) and the other non-Merck investigators from the Vioxx Gastrointestinal Outcomes Research (VIGOR) clinical trial have stoutly refused the request by editors of the journal for a correction to the published VIGOR study [1,2]. In making that request [3], Drs Gregory D Curfman, Stephen Morrissey, and Jeffrey M Drazen charged that:

• Three additional MIs were not included in the paper but were known to at least two of the authors in time to add the during the review process.
• Excluding these three MIs resulted in understatement of the difference in rofecoxib (Vioxx, Merck) and naproxen risks and downplayed the risk associated with rofecoxib.
• Relevant cardiovascular AE data were dropped from the manuscript before submission.

Bombardier et al say that the three MIs occurred after a predetermined cutoff date for reporting cardiovascular events, that they were unknown to the investigators drafting the manuscript, and that "changing the analysis post hoc and after unblinding would not have been appropriate." They further write that adding in the three MIs would not have changed the paper's conclusions, based on "assessment of the relative risks and widely overlapping 95% confidence intervals." Finally, they say that the table Curfman et al charge was dropped from the manuscript was indeed dropped—because its data were included in the text—and that the table never included the three extra MIs. "Thus we stand by our original article, which was written in line with basic clinical trial principles, specifying that data must be analyzed according to plans that are determined before unblinding," Bombardier et al conclude.

Not so fast, respond Curfman et al. They charge that at least two of the authors were aware of "critical data on an array of adverse cardiovascular events that were not included in the VIGOR article" more than four months before publication. The editors also say that the "prespecified cutoff date" Bombardier et al cite was one that "the sponsor selected shortly before the trial ended [and] was one month earlier than the cutoff date for the reporting of adverse gastrointestinal events." The editors write, "This untenable feature of trial design, which inevitably skewed the results, was not disclosed to the editors or the academic authors of the study."

The three additional MIs occurred during the 12 days after the cardiovascular cutoff date and within the period when GI adverse events were still being counted.