San Bruno, CA - A new observational study has found that aprotinin (Trasylol, Bayer), used to reduce bleeding during cardiac surgery, is associated with serious end-organ damage, particularly renal failure, and the authors recommend that it no longer be used [1]. They say that the less expensive generic medications, aminocaproic acid and tranexamic acid, are safe alternatives.

Lead author Dr Dennis Mangano (Ischemia Research and Education Foundation, San Bruno, CA, a nonprofit biomedical research organization) says: "Our study provides compelling evidence of aprotinin's serious risks and strongly suggests discontinuation of use."

But one expert has urged caution in interpreting the results of the current study, saying it should be viewed alongside the randomized data on aprotinin, which have suggested it to be safe and effective, albeit in a narrower patient population. Mangano, however, disputes this.

Aprotinin has been marketed internationally for the past 13 years and given to an estimated one million surgery patients. 2006 sales are projected to be in excess of $600 million.

In a paper published in the January 26, 2006, issue of the New England Journal of Medicine, Mangano et al conclude: "Given our findings, especially with regard to serious renal events among patients undergoing either primary or complex surgery, and given the cost of aprotinin therapy, which is at least 10 times that of aminocaproic or tranexamic acid, we estimate that considerable global healthcare savings would accrue if aprotinin were replaced by either aminocaproic acid or tranexamic acid."

He adds: "Certainly, our findings—coming on the heels of the Vioxx experience—indicate that the problem of drug safety is not only ubiquitous but also much more elusive than previously thought. In fact, our findings raise even more troubling concerns, for aprotinin has been on the market for three times as long as Vioxx, yet few comprehensive safety studies have been conducted since approval; the life-threatening complications with aprotinin found here occurred far more frequently than those with Vioxx; and far less-expensive generic alternatives to aprotinin, which are equally effective in limiting bleeding, have been available but have been underused."

In the paper, Mangano et al explain that antifibrinolytic agents such as aprotinin, aminocaproic acid, or tranexamic acid are given to the majority of patients undergoing invasive cardiovascular procedures, but evidence supporting their safety is limited and is largely composed of several published secondary analyses from sponsor-supported studies. They note that no study has been adequately powered to assess relatively infrequent, but clinically serious, safety events and that the comparative safety of the three agents has not been assessed within one study, which would be desirable given the large cost differences between the agents.

They also point out that conducting placebo-controlled trials would now be impossible, given that antifibrinolytic practice currently is embedded and dictated by guidelines. So to address the safety issue, they conducted a non-sponsor-supported, prospective, observational study, which they say is large and detailed enough to "allow comparative safety assessment among the three agents by exacting propensity and multivariable analyses."

The study involved 4374 patients undergoing cardiac bypass surgery at 69 centers worldwide. Of these, 1374 patients received no antifibrinolytic and acted as the control group, 1295 patients received aprotinin, 883 patients received aminocaproic acid, and 822 patients received tranexamic acid.

Results showed that after propensity-adjusted, multivariable logistic regression, use of aprotinin was associated with about a doubling in the risk of renal failure. It was also associated with significant increases in the risk of MI and heart failure. Neither aminocaproic acid nor tranexamic acid was associated with an increased risk of renal, cardiac, or cerebral events. All the agents reduced blood loss.

Mangano et al estimate that replacing aprotinin with one of two safe generic drugs would prevent renal failure requiring dialysis in 11 050 patients per year, saving at least $1 billion in dialysis costs, and would reduce drug costs by at least $250 million.

"We estimate that as many as 10 000 patients may be unnecessarily on dialysis today due to aprotinin use. This serious impact on human lives underscores once again the necessity for meticulous, postapproval surveillance, as well as ongoing, unbiased analysis of drug safety—all conducted by entirely independent entities," said Mangano. "This is easier said than done, however, for the economic forces are—and will continue to be—substantial, with little corporate incentive to identify safety problems once drugs are approved and marketed," he added.

Mangano stated, "We believe that the independent observational approach is likely the only method practical for unbiased assessment of drug safety in high-risk populations once a drug is marketed and practice is embedded. Unfortunately, such studies also are very costly, and it is up to society to find creative ways to independently assess safety. Otherwise, the Vioxx—and now aprotinin—sagas will be only the first of a series of public-health drug-safety failures."

Mangano's study is accompanied by both a Perspective and an editorial, both of which praise the study. In the Perspective [2], Dr David Hunter (Harvard School of Public Health, Boston, MA) says the current study by Mangano et al "may be a model for the future." He emphasizes that the large amount of data (about 7500 data fields per patient) collected by Mangano et al permitted the use of propensity scores and conventional multivariate techniques to control confounding and ensured that the clinical outcomes of interest were not obscured by misclassification due to a lack of clinical detail or an inaccurate administrative assignment of diagnostic codes.

We need to ensure that skepticism about the value of observational studies does not engender nihilism.

"Studies such as the one by Mangano and colleagues point the way to the prospective design of studies to assess drug safety and to the collection of as much information as necessary to provide answers of the highest quality. Substantial questions remain about how to fund and administer these studies, but we need to ensure that skepticism about the value of observational studies does not engender nihilism. In the absence of evidence from randomized trials, the best-quality data must be made available to ensure the safety of medications," Hunter concludes.

The editorial by Dr Gus Vlahakes (Harvard Medical School) agrees that the large number of patients and information gathered about those patients gives this observational study credibility [3]. He adds that the role of independent clinical research in phase 4 testing should be encouraged and supported by the FDA.

But some experts are urging caution over the interpretation of the current study. One of these is Dr Artyom Sedrakyan (Agency for Healthcare Research and Quality, Rockville, MD), who has previously conducted a review of randomized trials of aprotinin that suggested that the agent reduced the risk of stroke by almost 50%, with no increase in mortality, MI, or renal failure in a CABG population.

Commenting on Mangano's study for heartwire, Sedrakyan said it has shown "interesting results," but he believes the authors should have been a little more careful about the conclusions they reached. "This is still an observational study, and observational studies have selection bias that, even with propensity scores, it is not possible to fully adjust for. You also can never be sure that absolutely every variable has been identified to adjust for," he told heartwire. This is particularly relevant for aprotinin as this agent is far more likely to be used in high-risk patients, he added.

Observational studies have selection bias that . . . it is not possible to fully adjust for. You also can never be sure that absolutely every variable has been identified to adjust for.

He says Mangano's study should be viewed alongside the many systematic reviews of randomized data on aprotinin that have suggested it to be safe and effective in the populations studied during the 1990s. "This study may better reflect real-life clinical use of the drug than the randomized trials, in that it will have included many groups excluded from the randomized trials, and it may therefore raise questions about use of aprotinin in some of these patient groups. But these results should not lead to the dismissal of the results of previous randomized trials and systematic reviews of those trials that have suggested the drug to be safe."

He believes cardiac surgeons need to look carefully at their population of patients when deciding how to interpret these results and whether or not to use aprotinin. Noting that the patient population has changed in recent years as sicker patients are referred to surgery, he suggested an appropriate reaction would be to think carefully about the use of aprotinin in types of patients excluded from the randomized trials. "If you have a population of younger, healthier male patients, similar to those included in the randomized trials, it looks as though aprotinin can be used safely. But if your population is substantially different from that included in the randomized trials, you may need to exercise caution when thinking about using this agent," he said.

Bayer issued a statement saying the results of the Mangano study are not consistent with the more than 15 years of clinical data and experience Bayer has amassed on aprotinin. "Bayer has studied aprotinin in a series of prospective, randomized, double-blind, placebo-controlled clinical trials encompassing almost 6500 open-heart-surgery patients worldwide. This study population is larger than the patient population in the study by Mangano et al. Bayer's randomized trials examined results from 3817 aprotinin patients in comparison with the 1295 aprotinin patients in the Mangano et al observational study," the company says. It notes that double-blind, randomized controlled trials are the accepted standard for the assessment of the efficacy and safety of drugs and do not suffer from the limitations that can exist in observational studies and that it believes that aprotinin is a safe and effective treatment when used in accordance with the product labeling.

Here are two alternatives that don't have this safety issue, are just as effective at reducing blood loss, and are much less expensive. So why would anyone use aprotinin?

Mangano told heartwire that he does not agree that the randomized trials showed aprotinin to be safe. "None of the randomized trials were powered to assess safety, and combining a whole load of underpowered trials is not the way to address the issue. One large trial is what is required," he commented. He also pointed out that the largest randomized trial with aprotinin did actually show that the drug was associated with graft thrombosis. "The onus surely is to do no harm. Our trial has shown a serious safety issue with aprotinin, but there are two alternatives that don't have this safety issue, are just as effective at reducing blood loss, and are much less expensive. So why would anyone use aprotinin?" he added.