The largest study to date of a controversial cardiac surgery drug shows it increases death rates and damages kidney function, according Duke University Medical Center researchers.
Aprotinin, a drug used to limit bleeding, was temporarily suspended from marketing in the U.S. in November 2007 after a small Canadian study was stopped because similar findings were discovered. The drug, Trasylol, is manufactured by Baylor AG.
"We're not surprised by the results,” says Dr. Andrew Shaw, an associate professor in Duke Medicine's department of anesthesiology and the lead author of the paper which appears in the February 21 edition of the New England Journal of Medicine. “It's what we expected to find.”
The Duke study is significant because “it is more than twice the size of the next largest study of aprotinin,” says Shaw. The prospective data was collected between 1996 and 2005. “Unlike the highly selected nature of randomized trial populations, our data represent the every day cardiac bypass surgery patient population. The data were collected at a time when aprotinin was thought to be safe.”
The Duke team started analyzing its database of patients after a 2006 NEJM study reported aprotinin use may increase the risk of heart attack, stroke and serious kidney injury.
“We were looking for an association between exposure to the drug and subsequent adverse outcomes,” Shaw says. “We found an increased incidence of death in patients who received aprotinin. That higher death rate seemed to persist even when we were able to control for the differences seen between the patient groups.”
Shaw and his Duke colleagues also linked aprotinin to impaired kidney function. “Kidney function is measured by serum creatinine levels which indicate how well the blood is filtering waste products,” he says. The study found aprotinin use increased serum creatinine levels, but they did not report an increase in patients needing dialysis. Shaw believes “that's because we probably didn't have enough patients who needed dialysis in our study to detect a significant statistical difference, although the incidence was numerically higher.”