FDA Tap Dances on Pradaxa Decision

FDA officers clearly don't practice medicine.* And as a result, they can make some really boneheaded approval decisions. Case in point is the agency's October nod to the 150-mg pill of the anticoagulant dabigatran (Pradaxa), but not the 110-mg pill. (For background starters on this topic and my unalloyed opinions, go to last year's No. 4 "Top Ten" story.)

The government's tap-dancing rationale for such an otherwise-inexplicative move is now available in the NEJM. And it's not terribly convincing. The FDA's argument can be distilled to this:

• Because the 150-mg BID regimen was superior to the 110-mg BID regimen in the pivotal trial (RE-LY), we only approved the higher-dose pill.
• We couldn't find a subpopulation in the pivotal trial (like older patients) for whom the 110-mg BID dosage was clearly superior to the higher dosage.

So it appears that the FDA concentrated on the treatment benchmark of high-dose Pradaxa, when considering approval of the 110-mg pill, rather than the traditional standard of care: warfarin. This really doesn't make a whole lot of sense, especially when the 110-mg BID regimen—which unlike warfarin, doesn't require laboratory monitoring—was noninferior (ie, equivalent) to warfarin therapy. By its own logic, the FDA should then recall the use of warfarin to prevent stroke in patients with atrial fibrillation.

Nevertheless in the same breath, the FDA officers acknowledge that there "were certainly reasons why we might have approved both doses [of dabigatran]." Among these is the recognition that "patients and doctors value choices" and the fact that some patients refuse to take warfarin because of fear of bleeding. In the end, however, it appears that the agency was skittish about patients or doctors opting for the warfarin-comparable 110-mg BID regimen, instead of the warfarin-superior 150-mg BID regimen.

But that's really a clinical decision that should be left up to practicing clinicians—not government officers pouring over statistical trial data in a windowless room, and certainly not government officers for whom warfarin treatment, with all its headache-inducing monitoring and pesky dose fiddling, is a whitewashed abstraction. The FDA's restriction on treatment options for patients with atrial fibrillation simply means that fewer of these patients will receive effective and relatively convenient stroke prevention.

Aaagh. I can't shake my head enough.

At heartwire, RE-LY investigator Stuart J. Connolly, MD, FRCPC, essentially concurs and responds to the FDA editorial by calling it an "apology" for "a bad decision." He declares in no uncertain terms: 

I just think that they're wrong; they should have approved the lower dose. I don't think what they're saying is incorrect—they couldn't find a subgroup in the RE-LY study—but I do think that there are patients where the 110-mg dose makes perfect sense.

And adds:

What's needed is [a reduced dose for] those patients in whom the bleeding risk is considered to be sufficiently high that they wouldn't want to use the 150-mg dose. Without the 110-mg dose available, there are, I think, a lot of patients who now will not receive dabigatran at all—these are patients for whom warfarin was not being used because the bleeding risk was considered to be high, or patients who have some bleeding on the dabigatran 150-mg dose for whom there is no alternative available.

What's clear is that they don't trust physicians to make rational choices about the use of the two doses. The point that they're missing is that physicians have a lot more information at their disposal when they're encountering a patient than what is in the RE-LY database, including an appreciation of the patient's own values, a lot more information about the patient's bleeding risk, and other information about the patient's history that just isn't in the RE-LY database.

Heartwire also reports that Boehringer Ingelheim continues to "discuss" the approval of the 110-mg pill with the FDA.

It's clearly time (overtime) that the FDA admit its mistake and approve the 110-mg pill of dabigatran.

N.B. The 110-mg pill is approved in Canada (along with the 75- and 150-mg pills) for the prevention of stroke and systemic thromboembolism in patients with atrial fibrillation.

RE-LY = Randomized Evaluation of Long-term Anticoagulant Therapy.

* At least that's my only explanation for the boneheaded approval decision on Pradaxa.