CD&D Washington Editor

SILVER SPRING, Maryland — The cost of developing a drug-eluting stent (DES) has been pegged as somewhere north of $100 million, but that number has not deterred several device makers from wading into this end of the cardiovascular device pool.

However, a recent discussion of FDA's proposed new guidance for DES development indicates that the agency wants even more effort for a number of aspects of this device development, including pivotal trials and post-approval studies (PAS). This inevitably will add additional costs to the regulatory pathway for a device that generally measures no more than 4 mm in diameter and, at most, 30 mm in length.

FDA and the Advanced Medical Technology Association (AdvaMed; Washington) hosted a workshop at FDA's new White Oaks headquarters last month, and DES product developers probably were not encouraged about the added hurdles being established by the agency.

Hesha Duggirala, PhD, an epidemiologist at FDA's Center for Devices and Radiological Health (CDRH), outlined the agency's PAS expectations, saying that its experiences with previous PAS efforts have prompted more interest in rigorous studies. She recommended that sponsors be ready with a "pretty decently hashed-out [PAS] protocol at the time of PMA submission."

Duggirala said such studies should include analyses of thrombosis and the rates of death in connection with myocardial infarction and compare them to the pivotal study population.

However, sponsors cannot expect to be choosy about enrollment. Duggirala said that PAS efforts must "have consecutive enrollment," meaning a large ratio of off-label use. She said that until "we have found the best way to address" the subject of off-label use, the agency wants a full analysis of all patients.

And a simple registry study will not suffice. FDA wants a hypothesis-driven study, and sponsors may have to follow enrollees for five years.

However, Duggirala said that if the incidence of patients who exhibit elevated ST segments falls off in the first two to three years, a shorter follow-up may suffice.

Duggirala said that FDA is open to the notion of a peri-approval PAS. "In the past, we've found significant delays" in PAS execution, she said, so a sponsor may want to start assembling a PAS prior to device approval.

This is "not to get the device out there before it is approved," Duggirala said, but rather to allow the sponsor to sign up institutional review boards and other participating organizations.

FDA also is open to the possibility of using the infrastructure of a PAS to house a supplemental investigational device exemption that would expand the labeled indication of the stent.

Duggirala urged interested sponsors to float such a proposal when presenting the protocol for any proposed PAS.

However, the agency will not allow sponsors to double-dip patient data from the original pivotal trial. "We're not looking for previously enrolled patients" for a nested supplemental IDE trial, Duggirala said.

Another aspect of this situation is that since the nested enrollees are part of a PAS, adverse events would qualify as medical device reports (MDRs). Duggirala said she had conferred with other offices at FDA, and those discussions have led her to conclude: "If it's a marketed device, regardless of whether those patients are under an IDE study, all those events are reportable as MDRs."

On the subject of pivotal trials, Andrew Farb, MD, a medical officer at CDRH, said the agency wants to see the rationale behind the proposed size of the study population, as always. FDA also wants to see specifics for intended uses, such as the range of lesion lengths and vessel diameters for which the device will be indicated.

Farb said that the use of a composite primary clinical endpoint — such as a composite of target-lesion revascularization, cardiac death and myocardial infarction (MI) related to the target vessel — "allows for a reasonably-sized study that can be completed in a timely fashion."

However, the drawback to a composite primary endpoint is that such a study may end up under-powered to pick up "outcome differences between individual components" of the composite endpoint, he advised.

As for secondary endpoints, "the expectation is that these will be exploratory in nature" and "not lead to additional claims," Farb said.

"We're in the era of concerns of safety" for DES products, the FDAer said, so the agency will want the protocol to determine rates of elevated ST intervals as well as of death related to infarction at one year, and describe the impact of thrombosis on death/MI.

To nobody's surprise, Farb said a sponsor will need to establish the relationship between dual anti-platelet therapy (DAPT) and thrombosis, so data collection efforts will have to include data on patient compliance with therapy, in terms of both prescriptions filled and actual use. However, sponsors also will have to capture "whether invasive or surgical procedures are interrupted due to" DAPT.

The guidance, Farb pointed out, indicates a 12-month primary endpoint "with a substantial number of patients [followed to] two years," but "our thinking has evolved."

He said that FDA is now of the mindset that "after 12 months after which DAPT has changed to aspirin alone," the agency will want patient follow-up to 18 months for half the patients.

On the topic of trials designed to expand the indications for an existing DES, Farb said the agency is not dead-set on a randomized, controlled trial, and he extended the same line of thinking to studies of future iterations of existing stent designs. Such a trial could compare enrollees to those who participated in the pivotal trial for the stent's first iteration, but a sponsor could also opt for a study employing objective performance criteria.

However, Farb cautioned sponsors about "outcome drift after several non-inferiority studies" of consecutive device iterations.

In this scenario, a DES "could be found non-inferior to the latest non-inferior DES, but no longer superior" to a bare-metal stent.