VIOXX

THE CLAIM

That Merck knew Vioxx increased cardiovascular risk years before withdrawing it, presented the trial data misleadingly, and marketed the drug aggressively to physicians in the interval.

THE THEORY, AS ITS PROPONENTS TELL IT

This file turns on a specific documented omission, which is why it is graded confirmed rather than contested.

The VIGOR trial, published in the New England Journal of Medicine in 2000, compared Vioxx with naproxen for gastrointestinal safety. It found a large relative increase in heart attacks in the Vioxx arm. The company's interpretation was that naproxen was protective rather than that Vioxx was harmful, which was a hypothesis with little support.

Three additional myocardial infarctions in the Vioxx group were known before publication and did not appear in the paper. The journal published an Expression of Concern in 2005 stating that data had been deleted from the manuscript before submission.

Internal documents produced in litigation showed a training programme for sales representatives on deflecting physicians' cardiovascular questions, and a list of doctors identified as problematic for raising the issue publicly.

The drug was withdrawn in 2004 when the APPROVe trial, run for a different purpose, showed the cardiovascular risk unambiguously. FDA scientist David Graham testified to the Senate that he estimated tens of thousands of excess heart attacks and sudden deaths, and described pressure within the agency over his findings.

WHAT IS KNOWN

The VIGOR data, the omitted events and the journal's Expression of Concern are all matters of public record.

Merck settled personal injury litigation for $4.85 billion in 2007 without admitting liability, and reached a $950 million criminal and civil resolution with the Justice Department in 2011 including a misdemeanour plea over off-label promotion.

Graham's Senate testimony and his published Lancet analysis are public. His estimate of excess events has been disputed in magnitude, not in direction.

The episode drove significant reform, including mandatory trial registration, which exists substantially because of this case.

EVIDENCE FOR

• Three myocardial infarctions known before publication and absent from the published paper.
• The New England Journal of Medicine's Expression of Concern about deleted data.
• Internal sales training materials on deflecting cardiovascular questions, produced in discovery.
• A $950 million criminal and civil resolution including a guilty plea.

EVIDENCE AGAINST (THE STRONGEST READING)

• The naproxen-protective hypothesis, while wrong, was a real scientific position held by some researchers at the time and was not invented purely as a defence.
• The excess-event estimates vary widely depending on assumptions, and the highest figures are not the consensus.
• The FDA approved and monitored the drug, so the regulatory failure is shared and is not solely the company's.

ASSESSMENT

Confirmed, and the cleanest pharmaceutical instance of the pattern in OSA-017 and OSA-127: the producer generates the evidence, controls how it is published, and the publication shapes what every prescribing physician believes. The mechanism was not a conspiracy of silence but the ordinary editorial control a sponsor has over its own trial write-up. Mandatory trial registration exists because of this file, which is the clearest evidence that the problem was structural rather than exceptional.

WHAT WOULD CHANGE THE GRADE

• CONFIRMED by a specific documented omission: three myocardial infarctions in the Vioxx arm were known before the VIGOR paper was published and did not appear in it. The New England Journal of Medicine issued an Expression of Concern in 2005 stating data had been deleted before submission.
• Established in discovery: sales training on deflecting physicians' cardiovascular questions, and a list of doctors identified as problematic for raising it.
• Disputed in magnitude, not direction: David Graham's estimate of tens of thousands of excess cardiac events. Estimates vary widely with assumptions.
• Mandatory clinical trial registration exists substantially because of this case, which indicates the problem was structural rather than exceptional.

FURTHER READING

• Rofecoxib (Vioxx) · Wikipedia
• New England Journal of Medicine, Expression of Concern regarding VIGOR (2005)
• David Graham et al., Risk of acute myocardial infarction, The Lancet (2005)