Heart Surgery Drug Pulled From Market

By GARDINER HARRIS
WASHINGTON, Nov. 5 — Pressured by regulators, the German pharmaceutical giant Bayer AG announced Monday that it had agreed to withdraw the controversial heart surgery drug Trasylol after a Canadian study suggested that it increased death rates.
Dr. John K. Jenkins, a leading official of the Food and Drug Administration, said, “F.D.A. could not identify a specific patient population where the benefits of using Trasylol could outweigh the risks.”
Because of concerns that supplies of alternative medicines could run short, the agency has asked for a phased withdrawal of the drug, which is given before heart surgery to reduce the risks of excessive bleeding.
Bayer said in a statement that it still believed Trasylol was beneficial when used as directed. The F.D.A. said it would try to accommodate doctors who still wanted to use the drug by allowing Bayer to provide it as part of a study.
About 200,000 patients around the world received Trasylol last year, 110,000 of them in the United States, Bayer said. The drug, known generically as aprotinin, has had F.D.A. approval since 1993.
Two weeks ago, researchers from the Ottawa Health Research Institute stopped a study of Trasylol in 3,000 heart bypass and valve surgery patients because although those given the drug bled less, they were half again as likely to die as those given two older alternatives. With the results so preliminary, the circumstances and timing of the deaths are among factors still unclear.
The researchers said the overall numbers were not large enough to be statistically significant, but that may have been because the study was not given enough time.
“The fact that this trial was stopped early weighed heavily in our decision” to request withdrawal of the drug, Dr. Jenkins said.
The F.D.A. had asked the researchers and Canadian health authorities for complete results of the study before making a decision. But when the researchers said they would need at least six more weeks to comply, the agency decided to act.
The study was the latest in a series of worrisome signs that Trasylol could be dangerous. Last year, a study published in The New England Journal of Medicine found that the drug increased the risks of kidney failure, heart attack and stroke. That research concluded that halting the drug’s use would prevent 10,000 to 11,000 cases of kidney failure a year and save more than $1 billion a year in dialysis costs.
The F.D.A. convened a panel of experts in September of last year to review the safety of the drug, and the panel concluded that Trasylol should remain on the market. But within days of the panel’s meeting, the agency discovered that Bayer had sponsored yet another study of Trasylol suggesting that the drug increased the risks of death and stroke.
The company had failed to disclose the results of its study to the agency or the advisory panel. Indeed, Bayer scientists had defended Trasylol at the panel’s hearing but had not mentioned their own study or its worrisome findings. A company investigation later concluded that the findings had been withheld as a result of “regrettable human error.”
This September the agency reconvened the expert panel, and it again voted to recommend that Trasylol continue to be sold. The agency itself subsequently determined that the drug should be used only in patients at great risk of excessive bleeding.
Trasylol’s history demonstrates that once a drug is approved, halting its sales is exceedingly difficult. Experts on advisory panels are often loath to take widely used medicines out of doctors’ hands, even when their safety is uncertain.
This history also shows how vulnerable the F.D.A.’s drug approval system can be to unwelcome surprises. In the interest of speed, the agency often approves drugs that have not been shown to extend life but have proved effective in lowering cholesterol, shrinking tumors, moderating blood sugar or causing other effects. But these effects may not result in any lasting benefit. Cancer patients, for instance, may die just as quickly even if their tumors shrink.
In the case of Trasylol, which works by blocking enzymes that dissolve blood clots, the agency had assumed that any drug that reduced excessive bleeding during heart surgery would benefit patients.
The first hints that this assumption might have been unwise came from researchers who undertook so-called retrospective studies, simply comparing the histories of patients given Trasylol with those of patients given other drugs. Those researchers found that the Trasylol patients had fared poorly.
Such studies, however, are fraught with statistical problems. Patients given Trasylol may have been sicker than those given other medicines, so their greater problems may have been due to their underlying illness, not Trasylol. The only way to solve such statistical riddles is to undertake a study in which patients are randomized to receive one of two or three different medicines and are then followed closely.
That is how the Canadian trial was undertaken.
As a result of it, the F.D.A. suggested withdrawal of Trasylol, while regulators in Germany, where sales were a fraction of those in the United States, demanded it.
“I would say that the F.D.A. decision to request a suspension was one we reached independent of German authorities or any other regulatory agency,” Dr. Jenkins said.
Trasylol sales amounted to $135 million, most of it in the United States, through the first three quarters of this year, Bayer said. The company added that the drug was not one of its top 10 sellers.