■ site map
BLOG: February 2010
Wakefield proceedings: an exception?
Formal proceedings against a medical study author, such as the fiery controversy spinning around Dr. Andrew Wakefield 1998 autism study, are extremely rare. Is the general absence of professional and ethical misconduct proceedings against those involved in authoring medical studies and presenting their data a proof that the vast majority of them are up to highest professional and ethical standards?
Not at all. As pharmaceutical industry gained enormous financial power and influence, the integrity of medical studies has become deeply compromised by commercial interests. Here are a few recent examples.
In September of 2000, JAMA (Journal of American Medical Association) published an article presenting results of the CLASS study of Pharmacia's (now Pfizer) new painkiller, celecoxib, sold as Celebrex (Silverstein FE, Faich G, Goldstein JL, et al, Gastrointestinal toxicity with celecoxib vs nonsteroidal antiinflammatory drugs for osteoarthritis and rheumatoid arthritis: The CLASS study: a randomized controlled trial).
The results were presented as complete and final, but they only represented the first half (6 months) of the study, in which serious gastro-intestinal complications in Celebrex' group were significantly lower than in the second half of the study.
Following year, JAMA editors learned about the hoax, and even published article raising concerns about a possible increased risk of cardiovascular complications with the new drug. But the Celebrex study report
was never retracted by the journal.
None of the study authors - all 16 of whom had financial ties to, or were employed by Pharmacia - was subjected to any type of proceeding for professional or ethical misconduct.
And neither were the authors of the article presenting the results of VIGOR trial for Merck's painkiller Vioxx, published in the NEJM (The New England Journal of Medicine) in November 2000 (Bombardier C, Laine L, Reicin A, Shapiro D, Burgos-Vargas R, et al, Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. The VIGOR study group.).
The article reported that those with history of cardiovascular disease were at greater risk of suffering myocardial infraction (heart attack) with Vioxx, than with naproxen (older class of painkillers). But it failed to report that Vioxx also caused
significantly more total serious cardiovascular complications,
not myocardial infarction alone, regardless of a previous history of cardiovascular problems.
It failed to report three heart attacks in group with normal cardiovascular risk, which allowed it to claim that Vioxx only increases the risk of heart attack in those that are already at high risk.
The article also failed to report that Vioxx group had developed significantly higher overall incidence of serious diseases - two and a half times more than with naproxen.
The article downplayed the fourfold increased risk of myocardial infraction by suggesting it "may reflect the play of chance" (probability of which, according to FDA's review at in January 2001 was between 1 in 625 and 1 in 10,000).
In addition, it suggested that the perceived nominally elevated risk of Vioxx actually results from supposed naproxen's protective effect on the hearth (notion not supported by any evidence; oddly enough, the reason cited for early termination of a large National Institute of Health's Alzheimer's prevention trial ADAPT in December 2004 was unacceptably high cardiovascular risk associated with both, naproxen and Celebrex).
Despite all this, the article passed the scrutiny of the 93%-rejection-rate NEJM editorial board.
When the FDA, in February 2001, posted VIGOR study data submitted by Merck, (which wanted to add Vioxx's beneficial effects on stomach to the label), the discrepancies were noted by the users of the FDA's Web site. In June 2001, Seattle pharmacist Dr. Hrachovec and a colleague sent a letter to the NEJM, bringing to its attention Vioxx data discrepancy making both doctors and patients unaware of the true risks associated with Vioxx use. The journal
refused to publish the letter,
citing "limited space".
But "limited space" didn't prevent it from publishing a "Drug therapy" review article in August 2001, which reinforced misrepresented safety picture of both, Vioxx and Celebrex (FitzGerald GA, Patrono C., The coxibs, selective inhibitors of cyclooxygenase-2).
The journal - as well as the regulatory agencies - had a number of other warnings about dangers of Vioxx, but there was no reaction, until Merck voluntarily pulled the drug off the market in September 2004, after an ongoing Vioxx study (APPROVe) was stopped for safety reasons. The NEJM editorial board decides it is the time to repudiate the Vioxx article in an "expression of concern" in December 2005,
more than a year after the drug has been withdrawn.
How such an unprofessional, irresponsible, and unethical conduct of the editorial of a highly regarded medical periodical can be explained? The fact that in this period it made an estimated $697,000 to $836,000 from the article reprints - sold mainly to Merck, which used them to "inform" doctors and consumers about Vioxx - could be as good a reason as any.
In that same period, an estimated 20 millions of Americans used Vioxx. In fact, $20 billion of Vioxx and Celebrex was sold after everyone who should care - including FDA - knew they are either no safer or more dangerous (Vioxx), no more effective, but costing 7-20 times more than generic alternatives.
It did cost consumers more than their money. According to David Graham, the FDA'a post-marketing drug safety researcher who blew the Vioxx whistle on his employer, for allowing this dangerous drug to be on the market, this drug
has caused at least 30,000-40,000 deaths.
Needless to say, none of the doctors and scientists involved in VIGOR study - all 13 of them with financial ties to, or employed by Merck - has ever been in danger of answering for ethical and/or professional misconduct.
And so weren't the four authors from the American College of Rheumatology who, despite the evidence of high risk from manufacturers' studies, have put the two COX-2-specific inhibitors, Vioxx and Celebrex, at the top of drugs recommended for pharmacologic therapy for patients with osteoarthritis. These guidelines remained on the government's guideline website for four years, until the time when Vioxx has been withdrawn from the market. All four authors had financial ties to the drugs' manufacturers.
Eight out of the nine authors of the July 2004 update of the National Cholesterol Education Programs (NCEP) all employees of the National Institute of Health (NIH) - recommending statins for lowering cholesterol, as disclosed 1 week after it was published in Circulation, had financial ties to statin makers. Bad publicity that followed prompted NIH to adopt policy preventing future conflict of interest among its employees, but the NCEP recommendations were never re-evaluated, neither by NIH nor by the journal (Abramson and Starfield, The Effect of Conflict of Interest on Biomedical Research and Clinical Practice Guidelines: Can We Trust the Evidence in Evidence-Based Medicine?, 2005).
And so on, and on...
The purpose of this rather longish digression was not only to illustrate how rare - indeed, nearly unheard of - is to have medical study author taken to ethical/professional misconduct proceedings, but also to picture how rather common these days are purposeful distortion of data in medical articles, and undeclared conflict of interests of their authors.
And that fair, accurate reporting is not necessarily the first priority of leading medical journals.
It is so in the United States, and so it is in the United Kingdom. These few examples throw better light on the scene of the circus, making it easier to see what is hidden in the background.
That is not to suggest that Dr.
Andrew Wakefield and the two senior co-authors of the 1998 autism
John Walker-Smith and
We also need to hear what the accused have to say, and try to understand what is really happening here.