Skip to Main Content

A key clinical trial used to win regulatory approval for the best-selling Xarelto blood thinner is being reviewed by the Food and Drug Administration over concerns that a device used to test patients was faulty and may have skewed results. The European Medicines Agency is also reviewing the trial data.

The reviews are gaining attention, in part, because the trial was overseen by Dr. Robert Califf, who was nominated to head the FDA. At the time the trial was conducted, Califf headed the Duke Clinical Research Institute, which is also undertaking its own analysis, and he was one of the principal investigators running the trial.

Known as ROCKET-AF, the trial results convinced the FDA four years ago to approve Xarelto for preventing strokes in people with atrial fibrillation, or abnormal heart rhythms. That was the second agency endorsement that year. Marketed by Johnson & Johnson and Bayer, the drug has since become the leading seller among a new type of blood thinners, generating nearly $3.7 billion in sales in 2014.


The trial compared Xarelto with warfarin, a decades-old standard of care, and relied on a device known as INRatio to measure the effect that warfarin had on clotting. Such devices are used by physicians and patients because the amount of warfarin needed must be closely tracked.

However, the FDA had sent two warning letters to the device maker about erroneous readings in 2005 and late 2006, just as the trial got under way. But the problem was not mentioned during a 2011 FDA advisory committee meeting, according to the meeting transcript and one committee member who asked not be named. And a recall occurred last year.


The issue, which was first reported by the Project on Government Oversight, is renewing controversy over the drug. At the time of the 2011 FDA advisory committee meeting, FDA medical staff recommended that Xarelto not be approved after determining the drug was not as effective as warfarin in preventing major bleeding.

Both J&J and Bayer wrote us saying they are “aware of this matter. We have conducted a number of sensitivity analyses,” which “confirm the results of the Rocket study and the positive benefit-risk profile of Xarelto.”

However, an EMA spokeswoman wrote us that “the manufacturer of Xarelto has recently informed the agency that the defect in the device could have an impact on the study results … Due to the defect, it is now thought that the [INRatio] device has impacted the clotting results measured for the warfarin [portion of the study], which might affect the overall results for Xarelto as compared with warfarin.”

The EMA review is expected to be completed in early 2016.

An FDA spokeswoman would only say that the agency “is aware of concerns regarding the INRatio device and its use in the ROCKET-AF trial and is reviewing relevant data.” She added that she it is not known when the review will be completed.

Similarly, a spokeswoman for the Duke Clinical Research Institute said an “independent analysis” is being conducted but did not indicate when it may be completed. [UPDATE: The DCRI just issued a statement saying a new analysis validated the findings of the trial].

To what, if any, extent the regulatory look backs will have an effect on Califf’s nomination is unclear. Despite his ties to industry — the DCRI runs numerous trials for drug and device makers — Califf has bipartisan support in Congress. Although two US senators recently placed a hold on his nomination, these were done for reasons unrelated to his work for the pharmaceutical industry.

A spokesman for the US Department of Health and Human Services did not respond to questions about whether Califf was aware of the warning letters or why the FDA did not provide that information to the advisory committee.

Given ongoing chatter about Califf’s ties to industry, the nomination might be delayed, but is unlikely to be derailed, according to Daniel Carpenter, a Harvard University political scientist who has studied the FDA but has not taken a position on Califf.

Another expert agreed.

“It’s difficult to know what, if anything, Califf knew about the device,” said Ira Loss, who tracks the FDA for Washington Analysis, a consulting firm. “He’s overseeing a trial but is not in the field with patients. It could be an issue if it had a significant effect on altering the results of the trial, but every major multinational clinical has issues. It’s hard to know how serious this is. But why didn’t somebody notice it five years ago?”