Big pharma is paying the price for pushing off-label applications and putting inappropriate pressure on medical education, but is it too little too late? Industry influence runs deeper than you might imagine.
Last July, federal prosecutors announced that British drugmaker GlaxoSmithKline agreed to a $3 billion settlement with the U.S. government. According to the U.S. Justice Department, this represented the largest case of healthcare fraud in U.S. history. The charges, which Glaxo plead guilty to, involved the drugs Paxil, Wellbutrin and Avandia. $1 billion were for criminal penalties and the remaining $2 billion was for civil liabilities owed the states and the federal government. What was their crime? Marketing the antidepressants for off-label use and withholding important safety data about Avandia.
Before we decide to accept another pen, free lunch or tote bag, consider the impact that big pharma has on the dissemination of information to providers and the influence it imposes on medical research. Such influence is pervasive in medicine, to the point that we sometimes miss the forest for the trees. The pharmaceutical industry, and its egregious behavior, has been woven into the fabric of our healthcare system to the point that we accept its intrusion with little objection.
According to the Wall Street Journal, for more than a decade, Glaxo lavished gifts on physicians while encouraging them to write for off-label, non-FDA-approved uses. I suppose that once off-label uses become common practice, FDA approval just doesn’t seem necessary. However, it is one thing if the medical community identifies a legitimate off-label use and something quite different if the drug company markets such uses to physicians. How many pens does it take to influence physicians to write for unfounded, off-label uses? I don’t know. However, the reported gifts included European pheasant hunting trips, ski trips in Colorado, free spa treatments and Madonna concert tickets. Call me naïve, but I really thought these types of perks disappeared in the late 1980s or early 1990s. Apparently, the practice continues, as does the list of egregious behavior from Glaxo.
Glaxo hired a publication company to “ghost write” an article about Paxil, reporting its efficacy in children, despite the clinical trial cited showing no such findings. The sham article was used to market the use of the drug in depressed children, despite the fact that the FDA has never approved this drug for patients less than 18 years of age.
An article was published in 2008 in the International Journal of Risk and Safety in Medicine that highlighted the deception in the manufacturer-sponsored 2001 study. Paxil was reported to be, “Well tolerated and effective for major depression in adolescents.” However, no benefit was shown in either of the two primary outcome measures or any of the six secondary measures. But, additional measures not in the original study design were shown, in post hoc analysis, to have statistical benefit. Thus, they were included in the final report. In addition, several measures that were included in the original design produced negative results and were then removed from the study. Finally, in an article published in the Journal of the American Academy of Child and Adolescent Psychiatry, the draft was edited to inflate the beneficial effects, while minimizing the negative effects of suicidal ideation and self-harm. (Jureidini, J.N., et al, Internat J Risk Safety Med 20(1-2):73, 2008.)
According to Federal investigators, Glaxo reps referred to Wellbutrin as, “The happy, horny, skinny pill,” covering many of its off-label uses, while two studies identifying the cardiovascular risks associated with Avandia were suppressed. Avandia now has a black box warning, mandated by the FDA and the drug has been mandatorily withdrawn from the market in Europe.
Although the $3 billion settlement is a great step, it must be taken in perspective. When you compare this penalty to the billions in annual revenue earned by each drug, this penalty could feasibly be seen as a mere speeding ticket. Will the practices continue? Glaxo has had three prior settlements with the U.S. government in recent years. They are still in business, and no one is going to prison. So, I suspect these practices will continue. And Glaxo is not alone; these practices are pervasive in the pharmaceutical industry. Other notable settlements include Pfizer’s $2.3 billion payout for Bextra and Eli-Lilly’s $1.4 billion settlement for Zyprexa.
Let’s take a detailed look at some examples of unethical – I dare say even immoral – behavior of some pharmaceutical companies, misleading the FDA, physicians and patients, purely for financial gain. Ghost writers and phantom researchers are often employed by pharmaceutical companies to write articles or conduct research that is later published under the name, and endorsement, of a prominent author or researcher. Authors have reported these issues for years, but the average physician has remained unaware and uniformed. In a 1999 Lancet article, Marilynn Larkin wrote about her personal experience of being pressured by industry forces. “I was asked to sign an agreement that I would not disclose anything about the project,” she wrote, adding “I was told exactly what the drug company expected and given explicit instructions about what to play up and what to play down . . . I was pressured to rework my drafts to position the product more favorably.” (Larkin, M., Lancet 354:136, July 10, 1999).
In 2005, a researcher was approached by a pharma-funded education company about authoring a completed paper on the untoward side effects of warfarin. She declined, only to find herself reviewing the exact same article for a medical journal, submitted under another author’s name. (Fugh-Berman, A., J Gen Intern Med 20(6):546, June 2005.)
How influential are these industry-funded educational companies? Pfizer has reportedly paid medical education companies to write 18% of papers on Zoloft. An alarming statistic is that 70% of funding spent for industry-sponsored trials is paid for the services of medical education companies. (Sismondo, S., PLoS Med 4(9):e286, September 2007.) Regarding Rofecoxib, Merck disclosed that 22 of 24 clinical trials and 36 of 72 review articles included the services of medical education companies. In other words, most of the published trials discussed were ghost authored or research performed by phantom researchers.
Beyond ghost writing, what other ways do manufacturers influence the practice of medicine? Some companies have been shown to suppress negative results, while promoting positive results. Many physicians take for granted that an unbiased approach is taken when researching new medications or treatments, but this is not always the case. An article published in the Archives of Internal Medicine details how Vioxx (rofecoxib) was released in 1999 and generated $2 billion in annual sales for Merck. In 2004, the drug was voluntarily withdrawn due to cardiovascular complications. This apparently responsible action looks less so when we consider that the manufacturer was well aware of the negative side effects as early as 2000. This article also reported that there were 30 manufacturer-sponsored trials, including 17,256 patients, evaluating this drug. Although the manufacturer claimed it was not initially aware of the risks, 70% of those trials were completed by December 2000, discovering a strong trend for cardiovascular thromboembolic events and death. The relative risk was 2.18. Thus, those taking the drug were twice as likely to experience a cardiovascular event than those not taking it. This data was available prior to completion of the APPROVE trial (for adenomatous polyps), which is the trial in which Merck identified the issue of cardiovascular risk. (Ross, J.R., et al, Arch Intern Med 169(21):1976, November 23, 2009.)
In another attempt to expand the indications for the use of rofecoxib, Merck conducted three clinical trials. Two of the three were published, and it was reported that the drug was “Well tolerated.” Two of the three trials showed excess mortality. However, no statistical analysis was performed on the mortality data in the published trials. Internal Merck data reflected a hazard ratio of 3.0 for rofecoxib in Alzheimer’s populations. This data was not reported to the FDA in a timely manner, and when it was reported, the risk was minimized. If that weren’t enough, there was no IRB or safety monitoring board involved. (Psaty, B.M., et al, JAMA 299(15):1813, April 16, 2008.)
When researchers preferentially publish positive results and/or data, it skews the database that clinicians use to make treatment decisions. In 2010, a Cochrane database review was performed to evaluate the number of trials completed on pharmacological treatments for ischemic stroke, but not published in full. They found 940 trials, but interestingly, 125 studies evaluating 89 different treatments were completed but not published in full. Not publishing this important data handicaps our ability to fully understand the efficacy and risks of many treatment options for ischemic stroke. (Gibson, L.M., et al, Trials 11:43, April 22, 2010.)
In 2007, the FDA Amendments Act was passed, resulting in the requirement that researchers register their trials with the Federal government. This was an excellent step toward transparency of the research process and serves as a repository of all trials, positive, negative, completed or discontinued. In addition, those failing to comply may be fined up to $10,000 a day. (Wager, E., et al, J Roy Soc Med 102:4, January 2009.)
Another example of publication bias was reported in the New England Journal in 2008. 74 trials were identified in the FDA registry, studying 12 different antidepressants. 38 of the 74 (51%) of the studies were positive, per the FDA, and 37 of the 38 were published. Of the 36 remaining, 24 were negative and 12 questionable per the FDA. 3 of the 36 were published with negative data, and 22 were not published. Disappointingly, 11 were published with positive conclusions, which conflicted with the FDA’s findings. (Turner, E.H., et al, N Engl J Med 358(3):252, January 17, 2008.)
So, who are the gatekeepers? Some have noted that even prominent peer reviewed journals have a propensity to publish positive data. One such author reported that two-thirds to three-fourths of the published articles are manufacturer sponsored and that the profit margin for reprints is 70%, which manufacturers purchase in large quantities. (Smith, R., PLoS Med 2(5):e:138, May 2005.)
Another source of journal revenue is from advertising. Again, manufacturers don’t buy ads to market drugs with negative results. Thus, there is an incentive for journals to publish studies with positive findings and to publish data from pharmaceutical manufacturers.
With journal publishers thus entwined, the only reliable gatekeeper is the individual clinician. Awareness of the ongoing problem with the pharmaceutical industry is the first step, which should lead each of us to scrutinize the data provided in manufacturer-sponsored trials. The unfortunate reality is that patients are the innocent victims of this egregious behavior. Pharmaceutical representatives deliver a sales pitch, not a fair and balanced review of the literature. There is no Hippocratic oath in the pharmaceutical sales industry.
Challenging the assertions made by pharmaceutical representatives and advertisements should be the standard. In addition, every physician should become familiar with Clinicaltrials.gov, noting what studies have not been published and even noting those that have undergone significant methodological changes from the time of initial reporting to the time of completion and publication. A healthy dose of skepticism is the most effective medicine on the market.
Kevin Klauer, DO, EJD is Editor-in-chief of Emergency Physicians Monthly, CMO of Emergency Medicine Physicians, Vice Speaker of the ACEP Council.