Drug Problems: Duke Reassessing Data From Trial Led By FDA NomineeTweet
November 20, 2015
By David Hilzenrath and Charles R. Babcock
Four years after the FDA approved the blood thinner Xarelto for cardiac patients based on a global clinical trial, the institute that coordinated the study is trying to determine whether the use of potentially malfunctioning medical devices affected the trial results.
Robert Califf, a former Duke University researcher who is now President Obama’s nominee to head the FDA, co-chaired the executive committee that designed and oversaw the trial.
Records show that the Xarelto trial used “INRatio” blood-testing devices and that the reliability of INRatio devices was the subject of two FDA warning letters before the trial began in late 2006. In 2014, four years after the trial ended, a recall notice warned that certain INRatio devices could deliver false test results and that use of those devices may “cause severe or life-threatening injuries, including death.”
Based on those records, over the past two weeks POGO asked Califf, the Duke Clinical Research Institute (DCRI), the FDA, European regulator EMA, the makers of Xarelto, and others involved how if at all the trial might have been affected by defects in blood-testing devices.
“The DCRI is conducting an independent analysis of the trial data to understand what effect, if any, the potential device malfunction might have had in subpopulations and specific patient groups,” DCRI said Thursday in a statement to POGO provided by spokeswoman Susan Landis.
“The results of the Institute’s independent analysis will be shared with the public, as are all study results from trials and studies conducted by the DCRI,” the statement added.
A member of an FDA advisory committee that reviewed Xarelto in 2011 said he was unaware of the 2005 and 2006 warning letters when the advisory committee recommended that the FDA approve the drug for patients with a heart condition called atrial fibrillation.
“This information should have been available to us,” Vasilios Papademetriou, a cardiologist at the VA Medical Center in Washington, said in an email.
Depending on how widespread any inaccurate readings were, the results of the trial could have been skewed, he added. “It all depends on how often it happened and in how many patients,” he told POGO.
The FDA advisory committee endorsed approval of Xarelto by a 9-to-2 vote with one abstention. Papademetriou voted that the drug should be approved.
The FDA in 2011 approved the use of Xarelto to prevent patients with atrial fibrillation from developing blood clots and strokes. The agency issued that approval over the objections of the primary FDA reviewers responsible for assessing the drug’s safety and effectiveness. FDA officials and advisory committee members criticized various aspects of the clinical trial, as POGO recently reported. Those criticisms did not focus on the blood testing devices used in the study.
Xarelto, promoted in television ads featuring comedian Kevin Nealon and golfer Arnold Palmer, among other celebrities, generated U.S. sales of $1.5 billion last year, according to a financial report by drug-maker Johnson & Johnson. The drug has been prescribed for more than 3 million patients in the United States, Kristina Chang, a spokeswoman for Johnson & Johnson subsidiary Janssen, has said.
Califf founded DCRI, which conducts clinical trials for drug companies, and since early this year has been a deputy commissioner of the FDA. He sailed through a November 17 confirmation hearing at which Senators from both parties said they supported his appointment as FDA commissioner.
Califf did not respond directly to questions POGO sent on November 10. Spokespeople for the FDA, its parent the Department of Health and Human Services, the EMA, and the makers of Xarelto—Bayer and Johnson & Johnson subsidiary Janssen—provided no comment on the blood testing.
HHS Assistant Secretary for Public Affairs Kevin Griffis referred POGO to a researcher at Duke, Manesh Patel, who was the first author listed on a New England Journal of Medicine article reporting the results of the Xarelto trial. Patel, who said by email that he was traveling, looped in the DCRI spokeswoman.
Later, in an email apparently sent to POGO in error, Patel advised the spokeswoman not to answer follow-up questions about the statement.
“I would suggest that we stop responding to him,” Patel wrote. “[T]hanks for the help today but I would suggest one statement emails and then no further engagement until we are ready,” Patel added.
The DCRI spokeswoman said the institute’s statement also served as a response for the trial’s executive committee.
According to an article Califf co-authored in a European medical journal, the executive committee he co-chaired “designed the trial and was responsible for oversight of study conduct, retained independent ability to analyse and present the data, and take[s] responsibility for the accuracy and completeness of data analyses.”
The Xarelto clinical trial, known as ROCKET AF, compared Xarelto to warfarin, which has been on the market since the 1950s.
The INRatio testing devices were used to monitor the blood clotting rates, known as INRs (international normalized ratios), of trial subjects on warfarin. The tests were meant to help doctors running the Xarelto trial determine whether and how much those patients’ warfarin doses should be adjusted to keep their clotting in the desired range. If their blood got too thick, they would be at increased risk of developing blood clots and suffering strokes caused by clots traveling to their brains. If their blood got too thin, they would be at increased risk of hemorrhages, including strokes caused by bleeding in the brain.
Among the advertised advantages of Xarelto are that, unlike warfarin, it does not require regular blood monitoring or frequent dose adjustments. In the clinical trial, subjects on Xarelto were also tested with INRatio devices, but for them the devices were programed to deliver sham readings; the purpose of their INRatio tests was to prevent doctors and patients involved in the trial from knowing which subjects were taking the experimental drug,
The December 2014 recall notice said certain INRatio devices could deliver falsely low INR test results.
A falsely low reading could indicate that more blood thinner should be administered than was needed. That could lead to increased bleeding in clinical trial subjects on warfarin. In comparison to warfarin, it could make Xarelto appear better than it was at reducing the risk of bleeding.
“Relative to warfarin, which would have more bleeding than it should, [Xarelto] would look better in terms of intracranial hemorrhage,” Sidney Wolfe of the consumer advocacy group Public Citizen said.
The FDA has said that Xarelto and other anticoagulants approved in recent years are substantially less likely than warfarin to cause a hemorrhagic stroke, which involves bleeding in the brain.
Among the questions Califf, the HHS spokesman, and the DCRI spokeswoman left unanswered:
●What if anything did Califf, the FDA, and DCRI do before, during, and after the clinical trial to determine the accuracy of the INR readings the devices produced?
●Did they examine whether there were any discrepancies between measurements taken from the devices and measurements based on lab samples at two points in the trial when, according to the clinical trial protocol, parallel tests were run?
●What reason, if any, is there to believe that the INRatio devices used in the Xarelto trial were free of the defects cited in the FDA warning letters and recall notice?
The first warning letter, dated October 4, 2005, said an FDA inspection found that HemoSense Corp., the manufacturer of INRatio devices, had information indicating the devices “were generating clinically significant erroneous values.” Test results that are too high or too low “have the potential to cause or contribute to a death or serious injury, because: they may result in erroneous dosing and thus improper control of coagulation,” the letter said. The firm’s initial response was inadequate, the letter said, giving the company 15 days to take corrective action.
The second warning letter to HemoSense, dated November 29, 2006, said a subsequent inspection found that INRatio devices were “adulterated,” that the company had not investigated complaints about the possible failure of the devices, and that it was slow reporting a complaint about a malfunction that resulted in a death.
The ROCKET-AF trial, which involved more than 14,000 test subjects in 45 countries, began in December 2006 and ran until the fall of 2010.
Jackie Lustig, a spokeswoman for Alere, a corporate successor to HemoSense, said by email that the FDA told HemoSense in 2005 and again in 2007 that the company had adequately addressed the concerns raised in the warning letters.
In December 2014, a recall was issued for INRatio devices made after April 1, 2008. The recall notice said Alere received 18,924 reports of malfunctions, including 14 serious injuries. “According to the firm, all affected devices may fail,” the notice said.
(Update, 11-23-2015: The Alere spokeswoman told POGO that the FDA later revised its recall notice at the company’s request to say “not all” of the 18,924 complaints were related to the recall. The spokeswoman said most of them were unrelated. She declined to say what the company has done to address those.)
The 2014 recall didn’t require that Alere retrieve the devices from patients, Lustig noted.
Instead, the recall notice said patients with specific medical conditions should stop using the recalled devices, and other patients should have their INR readings verified by lab tests.
The FDA has permitted the INRatio systems to remain on the market, Lustig said, and the company “is working diligently to develop a software upgrade” for the devices to “address the potential for inaccuracies.”
Read our report, “Drug Problems: Dangerous Decision-Making at the FDA”
Read more articles in the “Drug Problems” series:
David Hilzenrath is Editor-in-Chief for the Project On Government Oversight.
Topics: Public Health and Science
Authors: David S. Hilzenrath
- December 8, 2016
- December 1, 2016
- August 12, 2016
- July 13, 2016
- July 13, 2016
- April 22, 2016
- March 24, 2016
- March 15, 2016