By Rob Stein and Marc Kaufman
Washington Post Staff Writers
Friday, October 21, 2005
A diabetes medicine poised to win Food and Drug Administration approval sharply increases the risk of heart problems, strokes and death, researchers reported yesterday in an analysis that raises new questions about how the agency handles drug safety concerns.
The drug, Pargluva, the first in what had been considered a promising new class of drugs for millions of diabetics, more than doubles the risk of life-threatening cardiovascular complications, the researchers concluded after analyzing the same studies the drug's maker presented to the FDA. They said the agency should not approve the drug until additional research is conducted.
"I do not think it's wise to approve the drug or see the drug marketed until there's a dedicated cardiovascular safety trial," said Steven E. Nissen of the Cleveland Clinic Foundation, who led the analysis. "We have to put safety first."
The FDA notified Bristol-Myers Squibb on Tuesday that the drug was "approvable" pending additional information about the safety issue, but it did not specifically request a new study to examine the risk of cardiovascular problems.
The critique comes as the FDA is struggling to restore its credibility after a series of embarrassing revelations and controversies, including the withdrawal of the blockbuster painkiller Vioxx last year because of serious side effects, a protracted battle over the Plan B morning-after contraceptive, and the abrupt resignation of Commissioner Lester M. Crawford, which remains shrouded in mystery.
The latest controversy underscores the heightened concern over drug safety, with the scrutiny this time on an earlier stage in the FDA approval process. But to some, it also reflects a continuing FDA reluctance to fully embrace drug safety as a paramount concern.
"I think this shows again a system failure at the FDA, that the agency still doesn't give safety the full attention it needs," said Curt Furberg, a drug safety specialist from Wake Forest University and until recently a member of the FDA advisory panel on drug safety.
The reviewers and experts on the advisory panels "don't necessarily have a good understanding of safety," he said.
In a statement yesterday, the FDA said it was taking steps to better identify cardiac problems that might be caused by drugs. "While we cannot specifically discuss this particular drug, the FDA has made significant investments of resources and expertise in developing fundamentally better methods for identifying and monitoring cardiovascular safety issues with all drugs," the statement said.
Pargluva, also known as muraglitazar, is the first of a new class of drugs known as dual-action PPARs. Several similar drugs are already on the market for the nation's 16 million people with Type 2 diabetes, the most common form. The drugs lower levels of blood fats known as triglycerides, raise levels of "good cholesterol" and increase the effectiveness of insulin. Pargluva was designed to combine those effects in a single pill.
As part of its standard drug-review process, the FDA convened a panel of experts on Sept. 9 to examine Pargluva. The committee voted 8 to 1 to recommend approval.
Nissen and two colleagues, alarmed that the only heart specialist on the FDA panel had recused himself because of a conflict of interest, independently reviewed the data from five studies involving 2,374 patients that Bristol-Myers Squibb had submitted. The analysis found those taking the drug had more than twice the risk of death, heart attacks and strokes, and nearly triple the risk when all types of heart problems were included.
The Journal of the American Medical Association released the study yesterday, five weeks ahead of its Nov. 23 publication, because of its public health implications.
While the number of patients in the analysis was relatively small, the trends were so consistent that it would be surprising if the higher risk was because of chance, Nissen said, especially as patients in the studies tended to be healthier than the typical diabetic.
Catherine DeAngelis, the journal's editor, agreed with Nissen's call for more research on the drug's safety. "I'm somewhat perplexed that individuals who are supposed to be looking after the safety of the public would not have asked for the definitive tests before they approve this drug," DeAngelis said.
In a statement, Bristol-Myers Squibb said it is "eager to begin discussions with the FDA to address more fully the cardiovascular safety profile of the compound and to determine what additional information may be necessary." The firm will provide additional safety data to the FDA from ongoing studies and confer with the agency about whether additional studies are needed, a spokesman said.
Nissen, who until recently was chairman of the FDA's Cardiovascular and Renal Drugs Advisory Committee, criticized the FDA advisory panel that reviewed Pargluva for failing to scrutinize the drug more carefully.
"I think this particular FDA advisory panel completely missed the point and did not probe deeply enough," he said. "The company, as expected, presented the data in a way that was favorable to the drug. The FDA did identify these adverse trends but left it to the committee to decide their importance. The panel members did not probe to get beneath the surface. They did not connect the dots."
At the Sept. 9 hearing, David Orloff, the director of the FDA's Division of Metabolic and Endocrine Drugs, set the tone by saying the evidence that the drug may pose a risk was "based on very small numbers of events in individual studies and on small numbers overall."
Bristol-Myers Squibb's Rene Belder told the panel the company had collected safety data equal to 500 patient-years, and that "the results show that the incidence in cardiovascular events, when corrected for duration of exposure, is similar for [Pargluva] and placebo."
But Peter Lurie of Public Citizen's Health Research Group, a consumer group, advised against approval, citing safety concerns. Referring to the FDA's experience with Rezulin, a similar drug that was withdrawn in 2000 because it caused liver problems, Lurie said the "wisest course is to pay attention to the clinical data."
In the end, only Dean Follmann, head of biostatistics at the National Institute of Allergy and Infectious Diseases, voted against recommending approval, saying that because Pargluva is in a new class of drugs, it needed further study before being marketed to millions of people.
The single-action PPAR drugs on the market, Avandia and Actos, have been sold long enough to alleviate concerns they may pose similar risks, Nissen said.