Article says diabetes pill would increase coronary risks
By Stephanie Saul
An article in a leading medical journal said yesterday that a proposed diabetes pill, Pargluva, seemed to significantly increase heart attack and stroke risks.
An accompanying editorial said the drug's sponsors, Bristol-Myers Squibb and Merck, had been "disingenuous" in submitting data to the Food and Drug Administration that "may have fostered an illusion of safety."
The authors - two of them Cleveland Clinic cardiologists who sounded alarms about Vioxx in 2001 - called for a major safety study of Pargluva before it goes to market.
The companies had been counting on selling the drug by late this year. The timetable was delayed earlier this week when the F.D.A. told them that Pargluva was approvable but only after the agency reviewed more safety data from completed studies. The type of safety study called for by the journal would mean that Pargluva would not reach consumers for at least two years, if at all.
The article by authors including Dr. Steven E. Nissen and Dr. Eric J. Topol and the editorial appeared on the Web site of The Journal of the American Medical Association (jama.ama-assn.org), which put the information out before its print publication in what it called the interest of public health.
The editorial, by Dr. James M. Brophy of McGill University in Montreal, also raised concerns about a cancer risk with Pargluva and listed eight ways that the companies' methodology for patient trials might have - perhaps unintentionally, he said - skewed results in favor of Pargluva's approval.
Among them, he said, studies of the drug included low-risk patients who were unlikely to have serious side effects and excluded other likely users of the drug with high cardiovascular risks, like elderly patients who often have Type 2 diabetes.
Responding to the article and editorial, a spokesman for Bristol-Myers Squibb, Tony Plohoros, said the company had not yet thoroughly reviewed the analysis in the journal, but added, "It's widely known that there are many different and appropriate ways to analyze data."
Mr. Plohoros, who said he was also speaking on behalf of the company's marketing partner, Merck, denied the editorial's suggestions that the clinical trials were intended to achieve positive results. "The trials were designed like any other clinical trial for a diabetes drug," he said. They were designed by outside experts and "thought leaders" in diabetes treatment, Mr. Plohoros added.
Bristol-Myers had hoped that Pargluva would reinvigorate its franchise in diabetes, a $20 billion market. Merck is eager for new drugs after its withdrawal last year of the painkiller Vioxx and related suits.
Several months ago, analysts at Friedman Billings Ramsey estimated 2008 sales for Pargluva at $950 million but recently lowered that estimate to $600 million, citing emerging questions about the drug.
Bristol-Myers stock was down 1.7 percent yesterday, closing at $21.67. Merck's stock closed unchanged.
Besides raising questions about Pargluva, the journal article appeared to challenge the F.D.A. advisory panel process. In an interview, Dr. Nissen criticized the panel that reviewed the drug last month and voted to recommend its approval.
Even though the F.D.A. staff was worried about Pargluva's cardiovascular risks, the panel included no cardiologists.
Like the article, an analysis by an F.D.A. safety officer, submitted before the panel's vote, said there was a doubling of cardiovascular risks associated with the drug. But the safety officer, unlike the article's authors, concluded that there was no clear pattern to the problems.
In disclosures with the journal article, Dr. Nissen said he had consulted for Takeda, Eli Lilly and GlaxoSmithKline as well as other companies. All three companies have diabetes products that would compete with Pargluva. But Dr. Nissen said he accepted no money for his consulting activities and instead asked that it be contributed to charity.
In a statement yesterday, the F.D.A. said its officials were "fully aware of the results of the Pargluva clinical trials, both with regard to safety and efficacy, and likewise appreciate the need for careful assessment of risk versus benefit for all drugs, particularly those indicated for long-term, preventive therapy." Because diabetes is a chronic disease, drugs used to treat it are frequently taken long term.
If approved, Pargluva would be the first of a new class of drugs called dual PPAR-agonists. The drugs are aimed at receptors in cells that regulate fat and sugar. Stimulating those receptors can control blood sugar while improving levels of cholesterol and other fats, or lipids.
The concept is beguiling because many of the 18 million Americans with diabetes also have excessive lipid levels. Still, attempts to develop dual PPAR-agonists have run into trouble in clinical trials, with some companies abandoning efforts.
The analysis found that in trials involving 2,374 patients taking Pargluva, 35 receiving Pargluva - or 1.47 percent - experienced death, heart attacks or strokes. That compared with 9 of the 1,352 patients in a control group, or 0.67 percent. When other cardiovascular problems were included in the analysis, like mini- strokes and congestive heart failure, 2.11 percent of Pargluva patients had adverse events compared with 0.81 percent for control subjects.
