Share this article on:

Journal Scan: Pharma Bias Detected in Thrombolytic Stroke Trials

Lovato, Luis MD

doi: 10.1097/01.EEM.0000413158.15926.fb
Journal Scan


Though no statistics back up this claim, it's probably not an overstatement that the NINDS trial on tPA for stroke has generated more debate than most studies in recent memory. That famous — or infamous, depending on your point of view — study concluded that acute ischemic stroke patients who received intravenous tPA were 30 percent more likely to have minimal or no disability at three months compared with placebo.

The National Institute of Neurological Disorders (NINDS) trial on recombinant tissue plasminogen activator for acute ischemic stroke, published in 1995, prompted the FDA to approve tPA for use in acute ischemic stroke the following year. (N Engl J Med 1995;333[24]:1581.) Detractors noted that the trial's benefit came at the cost of increased symptomatic intracerebral hemorrhage in the treatment group (6.4%) vs. placebo (0.6%).

Since then, the American Heart Association and American Stroke Association have been unwavering supporters of tPA for acute stroke, giving it a Grade A recommendation (Circulation 1996;94[5]:1167), then subsequently a Class I recommendation (Stroke 2007;38[5]:1655), and then finally a Class I endorsement for extending the post-stroke thrombolytic window to 4.5 hours in select patients. (Stroke 2009;40[8]:2945.) Soon after each of these national stroke guidelines were published, the American Academy of Neurology released endorsement statements of its own.

A controversy has always lingered, however, about the efficacy, cost, safety, and use of tPA for acute stroke, especially when used outside the research setting. After NINDS, a large retrospective study from the Cleveland area found that only 1.8 percent of stroke patients received tPA. Fifty percent of those who received tPA in this study had deviations from therapy guidelines, and 15.7 percent of them developed symptomatic intracerebral hemorrhage. (JAMA 2000;283[9]:1151.) Mortality was also significantly higher in the tPA group than in the non-tPA group (15.7% vs. 5.1%, p<0.001). Even more concerning, over time the NINDS trial data themselves have been subjected to increased scrutiny, bringing into question the results of this fairly influential trial. (Ann Emerg Med 2009;54[3]:329.)

Meanwhile, back in the ED, emergency physicians are practicing between a rock and a hard place, damned if they administer tPA and damned if they don't, with patients stuck somewhere in the middle. EPs are blamed for debilitating outcomes and death when tPA results in hemorrhage, despite it being a well documented complication. Yet physicians who withhold tPA for safety and efficacy concerns are subject to negligence claims for not offering limb-saving therapy supported by “national guidelines.”

Fortunately, our professional organizations made some efforts to protect us from opportunistic litigation. The American College of Emergency Physicians in 2002 issued a policy that insufficient evidence existed to endorse tPA for acute ischemic stroke unless institution-wide commitment was in place to help guide acute therapy and manage complications. The American Academy of Emergency Medicine issued a similar statement with the following conclusion: Given the cited absence of definitive evidence ... it is inappropriate to claim that either use or non-use of intravenous thrombolytic therapy constitutes a standard of care issue in the treatment of stroke. And finally, the Society for Academic Emergency Medicine passed a policy statement in 2003 that similarly questioned the benefit of tPA for acute ischemic stroke, though they rescinded that statement in 2009.

Radecki's stated goal was to examine the role pharmaceutical financial interest has on influencing the medical literature regarding thrombolytics for acute ischemic stroke so this month's Journal Scan is sure to fuel the flames of this hot debate.

Pharmaceutical Sponsorship Bias Influences Thrombolytic Literature in Acute Ischemic Stroke

Radecki RP

West J Emerg Med


Radecki used the ISI Web Science website as a source for an extensive literature search on stroke and various thrombolytics. He found more than 2,500 articles, but excluded those that did not report on thrombolytic intervention for acute stroke. Remaining articles were then rank ordered by the number of citations received since publication. Finally, the top 20 cited articles were reviewed for any pharmaceutical financial interest.

Articles were judged to have pharma interest if the drug manufacturer directly or indirectly funded the study, if the drug manufacturer was involved in subject recruitment or site monitoring, or if any of the authors disclosed financial interests in the company manufacturing the study drug. Notably, articles were judged not to have any pharma interest if the company's only contribution was to provide the study drug.

Using these criteria, 85 percent (17/20) of the most cited articles were found to be tied to pharma interests. Only 15 percent (3/20) of articles had no significant ties to pharma.

The article mentions several potential limitations of this study. Even though one might infer that the more an article is cited, the more it influences practice, that isn't necessarily the case. An article might be frequently cited as a textbook example of a poorly designed study. Older articles also had more time to be cited, yet perhaps may no longer carry much influence on current practice. On the other hand, articles that have been recently published may be very influential on current practice but not yet had enough time to be widely referenced. Finally, judging pharma interest in each of the selected articles was completely dependent on author disclosures in print.

The article by Radecki concludes that an overwhelming percentage of the most influential literature on thrombolytics for acute ischemic stroke is susceptible to pharmaceutical bias, which might influence physician concern over the efficacy and safety of this therapy.

This article offers an interesting look at the potential influence pharma has had on the stroke literature. Unfortunately, it doesn't offer conclusive evidence on what's best for our patients. Until the thrombolytics-for-stroke debate is settled, we as emergency physicians should push for multidisciplinary, institution-wide support for the care of early stroke and its complications. We should clearly discuss and document risks and benefits with patients and families, and we should promote further nonbiased research on the subject. And perhaps most importantly, clearly determine and document the exact time your patient's stroke began, for the patient's protection (and ours). Establishing that your patient doesn't qualify for an AHA-endorsed therapy in this case might actually be a good thing.

Click and Connect! Access the links in EMN by reading this issue on or in EMN's app for the iPad, available in the Apple app store.



Dr. Lovato is an associate professor at the David Geffen School of Medicine at UCLA, the ED director of clinical practice at Olive View-UCLA Medical Center, and the co-chair for the Emergency Medicine Best Practices Committee for the Los Angeles County Department of Health Services.

Back to Top | Article Outline

EMN Quick Poll

Does your ED routinely administer tPA for stroke? Vote in our quick poll on or tell us your opinion by writing to

© 2012 Lippincott Williams & Wilkins, Inc.