This article is a collaboration between MedPage Today and:
An FDA panel voted 9-2 with one abstention in favor of a narrow antithrombotic indication for cangrelor in percutaneous coronary intervention (PCI).
While that was a reversal from the 2-7 vote the drug got at an initial panel meeting in February 2014, the current panel members were nearly unanimous in expressing reservations and reluctance based on many of the same concerns as before.
Those issues centered on the single trial proffered as evidence (ignoring two prior negative trials) and on how the drug might fit into (or negatively alter) current clinical practice.
The Medicines Company had initially gone for a broad indication for its drug as antithrombotic therapy in PCI along with a bridging indication for patients whose antiplatelet therapy is interrupted for surgery, but was roundly rejected on both counts.
This time around, the proposed indication was narrowed to use as an adjunct to PCI for reducing risk of periprocedural ischemic complications, including myocardial infarction (MI) and stent thrombosis, in patients in whom treatment with an oral P2Y12 platelet inhibitor prior to PCI is not feasible and when glycoprotein IIb/IIIa receptor antagonists are not anticipated to be used.
While a more restrictive indication, it's still one for which there is a need, argued , who presented at the discussion as past president of the and director of the cardiac cath lab at Northeast Georgia Medical Center in Gainesville.
The vast majority of PCI procedures done in the country are on an ad hoc basis, with patients first seen by the interventional cardiologist in the holding room when they come in for diagnostic catheterization and then going straight on to PCI if the anatomy is right, he explained.
Added to that, there is a lack of clear guidance on clopidogrel (Plavix) timing and a desire to avoid unnecessarily preloading patients who might go on to surgery or medical treatment if the coronary anatomy doesn't favor PCI, Marshall noted.
So it's no surprise that many patients end up in that group without a P2Y12 platelet inhibitor or glycoprotein IIb/IIIa receptor antagonist on board at the time of PCI, he said.
"We're not that good at predicting which patients need angioplasty," he told the panel.
But even if it's typical practice, it's still substandard, a representative of the watchdog group Public Citizen argued at the public hearing portion of the meeting.
About a quarter of the clopidogrel arm in the CHAMPION PHOENIX trial got a 300 mg rather than 600 mg loading dose and more than a third got the drug during or after the procedure, delayed compared with cangrelor, he noted.
As a result, "one cannot conclude that cangrelor is superior to, or even noninferior, to clopidogrel appropriately administered," he said in explaining his group's strong opposition to approval, noting that use is likely to creep beyond the labeled population.
The antiplatelet agent is faster-on and faster-off than fellow P2Y12 inhibitor clopidogrel (Plavix), which would make it useful, according to , chief of cardiothoracic surgery at Brown Medical School in Providence, R.I, who spoke at the public hearing portion.
For patients found to have left main disease or other anatomy more suitable to surgery, they have to wait about a week for P2Y12 inhibitors to wear off or risk bleeding and transfusion that is linked to poorer outcomes, he noted.
And in ad hoc PCI, there are a variety of good reasons why physicians might elect not to start P2Y12 inhibitors before hand as well, such as nausea, cardiogenic shock, or use of sedating medication, added , chief of medicine at Stanford University in Stanford, Calif., and co-PI on the CHAMPION trials.
He agreed with the FDA staff review that concluded there was greater benefit than risk with cangrelor in the trial.
The Medicines Company, which did not seek a bridging indication as in the first panel meeting, did a number of sensitivity analyses of the CHAMPION PHOENIX data to address FDA concerns including only MIs and intraprocedural stent thrombosis meeting the SCAI definitions of clinically relevant events.
The number needed to treat was 156 to prevent one MI, stent thrombosis, or revascularization, while the number needed to harm was 461 to cause one additional GUSTO severe or moderate bleed or 198 to cause one additional TIMI major or minor bleed.
The panelists consistently described this as a narrow margin of benefit but several who voted "yes" said they were swayed by the FDA staff review description of a "small benefit but smaller risk."
In explaining their votes, though, several reiterated the concern expressed throughout the meeting that approval might change practice by giving clinicians a convenient, although no better and possibly less effective, alternative to preloading.
They called for the FDA to reflect such concerns in the label if it does follow their recommendation for approval.
From the American Heart Association: