July 14, 2010
by Rita Rubin
A large clinical trial of Avandia, sponsored by its maker, “was inadequately designed and conducted to provide any reassurance” that the controversial diabetes drug does not increase cardiovascular risk, a Food and Drug Administration scientist wrote in a memo released Friday.
The lengthy memo by Thomas Marciniak, a medical team leader in the Division of Cardiovascular and Renal Products, is part of a 765-page briefing document prepared by FDA scientists in advance of next week’s advisory committee meeting on Avandia’s fate.
The RECORD trial, a study ordered by the European Medications Agency, enrolled 4,447 patients. It compares Avandia, the trade name for rosiglitazone, combined with metformin or a sulfonylurea, which are two other diabetes drugs, to metformin combined with sulfonylurea.
It is an “open label” trial, which means that the patients and researchers are aware of who’s getting which drugs, knowledge that could bias the findings, Marciniak wrote.
He cited a number of other problems with the study, including a lack of complete information about which study participants had died, information that could have made Avandia look riskier.
Next Tuesday and Wednesday’s advisory committee meeting will be the second in three years to review Avandia’s risk/benefit profile.
At a July 2007 meeting, the panelists voted 20-3 that Avandia did raise heart attack risk. Yet, the panel voted 22-1 to recommend keeping the drug on the market. The FDA usually, but not always, follows its advisory committee recommendations.
Although their terms have expired, the FDA has taken the unusual step of inviting the 2007 advisory committee members to vote alongside their successors at next week’s meeting.
Concerns surfaced in ’07
Concerns about Avandia’s safety were raised in May 2007, when Steven Nissen, chair of cardiovascular medicine at the Cleveland Clinic, and coauthor Kathy Wolski published a report in The New England Journal of Medicine suggesting Avandia increased users’ risk of heart attacks. That term refers to problems related to an inadequate blood supply to the heart, including angina and heart attacks.
In February of this year, an investigation by the Senate Finance Committee concluded that maker GlaxoSmithKline, or GSK, knew of possible Avandia heart risks for several years before publication of Nissen’s study.
In a statement Friday, Murray Stewart, vice president for clinical development at GSK, said the company stands by its conviction that Avandia does not have unique cardiovascular risks.
“Since 2007 we have seen results from six controlled clinical trials looking at the cardiovascular safety of Avandia,” Stewart said, “and together they show that this medicine does not increase the overall risk of heart attack, stroke or death.”
The FDA has compiled a list of eight questions for the advisory panel members to consider. The next-to-last question asks them to recommend a specific regulatory action, ranging from keeping Avandia on the market and removing all warnings on its label about heart attack risk to pulling the drug off the market.
A split decision?
Committee members have their choice of five answers to that question, reducing the odds “of an overwhelming/unanimous vote for any one option,” according to a report released Friday by Concept Capital’s Washington Research Group, which advises institutional investors.
The Concept Capital report speculated that the FDA advisory panel might end up evenly split between allowing Avandia to stay on the market, with additional label revisions and restrictions on prescribing, or taking it off the market.
However, the Concept Capital report said, the most important question might be the one before that: “Based on the available data, please discuss the benefit-to-risk profile of rosiglitazone in the context of other available anti-diabetic therapies.”
As Concept Capital’s Cole Werble, Michael McCaughan and Ramsey Baghdadi write, “this is the critical question FDA decision-makers ask of a therapy that might have serious safety risks when other therapies exist on the market.” The agency did not ask that question of advisory committee members in 2007.
David Graham, the FDA scientist who made headlines in 2004 when he testified before a Senate committee that the agency was not equipped to prevent another Vioxx — the pain-reliever pulled from the market after a study found it increased heart attack risk — said in an interview Thursday that studies consistently have shown that Actos, the only other drug in the same class as Avandia, protects against heart attacks.
On the other hand, Graham says, Avandia, which has never been shown to be more effective than Actos in controlling blood sugar in diabetes, consistently appears to be riskier to the heart in comparisons with other drugs. Graham will be presenting an overview of the research at the advisory committee meeting.
“Really, what is most clinically relevant is the comparison of Avandia vs. Actos in the same study,” Graham said.
In a paper published online June 28 by The Journal of the American Medical Association, Graham and his coauthors analyzed data from 227,571 Medicare beneficiaries who’d been prescribed Avandia or Actos. They concluded that Avandia was associated with a higher risk of stroke, heart failure and death from any cause than Actos.
At the advisory meeting, Graham will report on eight other studies comparing patients prescribed Avandia to those prescribed Actos. All eight, he says, suggest Actos is safer.
To definitively answer whether Avandia increases cardiovascular risk, the FDA asked GSK to conduct the TIDE trial. The trial, which aims to study 16,000 patients in countries around the world, began enrolling participants in May. It is designed to compare the risk of heart attacks, stroke and cardiovascular death in diabetes patients randomly assigned to take either Avandia or Actos. It is also designed to test the impact of vitamin D supplement use on risk of death and cancer.
He called the TIDE trial unethical because “it’s treating humans as if they are laboratory rats. Why on Earth would anyone want to be randomized to Avandia in a clinical trial the purpose of which is to prove with absolute certainty that Avandia increases risk?”
In April 2010, the FDA asked the Institute of Medicine, or IOM, to answer five questions about ethical and scientific issues in studying the safety of approved drugs. The IOM is part of the National Academies, created to advise the government and the public. In light of the Avandia advisory committee meeting, the FDA asked the IOM to answer one question first: “What are the ethical and informed consent issues that must be considered when designing randomized clinical trials to evaluate potential safety risks?”
In a letter released today, the IOM said, “It is appropriate for FDA to require that a properly designed trial be conducted to provide additional evidence about an approved drug’s efficacy and safey” when there is too much uncertainty about risks vs. benefits to make “a responsible policy decision.”
In addition, the IOM said, the FDA should ensure that the trial includes an ongoing, “comprehensive and meaningful” informed consent process. But Graham says the TIDE informed consent process is more like “misinformation.” For example, he says, it does not mention the 2007 advisory committee’s overwhelming vote that Avandia raises heart attack risk or that the American Diabetes Association says Avandia shouldn?t be prescribed.
Graham and colleague Kate Gelperin?s critique of both the TIDE and RECORD trials is among the briefing documents for the advisory committee.
At a news conference Thursday, Deputy FDA Commissioner Joshua Sharfstein said he couldn’t predict when the agency would make a decision based on the advisory committee’s recommendations.
“Obviously, we’re going to have to look at a lot of information,” Sharfstein said. “We’re going to try to make a decision as quickly as we can under the circumstances.”