Medical Device Daily Washington Editor

GAITHERSBURG, Maryland — Sun Tzu, the legendary military mind of China’s Zhou Dynasty, probably never recommended a strategy of burying the opponent in bits and bytes, but Medtronic (Minneapolis), essentially used that tactic in providing supporting data for its premarket approval application for the Endeavor, a second-generation drug-eluting stent (DES).

The tsunami of information that Medtronic provided to the FDA’s circulatory system advisory panel on Wednesday included data on seven clinical trials — three of these randomized controlled trials (RCTs) — for more than 2,200 patients. And more than half of these, nearly 1,300, had been followed for over two years.

By the end of the hearing, any concerns about potentially iffy data from one of the trials were swamped by the sheer volume of data. That data showed that the Endeavor outperformed bare metal stents (BMS) on all efficacy endpoints and demonstrated comparable outcomes in trials going head-to-head with the Taxus DES, made by Boston Scientific (Natick, Massachusetts).

Literally at the end of the day, the panel voted unanimously to recommended FDA approval of the Endeavor, contingent on post-approval enrollment of 5,000 subjects for five years to check for very late stent thrombosis as a primary endpoint, and death from myocardial infarction as a secondary endpoint, all in a single-arm registry.

The panel also recommended that the label list the latest recommendations by the American College of Cardiology (Washington) and the American Heart Association (Dallas) regarding dual antiplatelet therapy.

But even with the wealth of RCT data, some attendees in attendance still had concerns.

During the morning open comment period, Bruce Ferguson, representing the Society of Thoracic Surgeons (STS; Chicago), said “[t]he use of composite endpoints in cardiovascular trials” often makes use of endpoints of least interest to patients.

Many doctors, he said, feel that RCTs “provide inadequate information to provide guidance for population subsets” that fall outside the trial population and that STS recommends “caution in the use of RCT data as the only criteria,” for evaluation of cardiovascular devices. He suggested that product labels describe the limitations of trials.

“The short cycle development of new technologies” creates problems because “new information is always coming” that may shed a different light on the benefits and hazards of cardiovascular devices. “The overall process of device evaluation is not patient-centered.” He said this made the case that coronary artery bypass graft patients should serve as the control group, and that while early mortality is of great interest to patients, “mortality is the least frequent event in composite outcomes.”

STS, he added, recommends “caution in relying on pivotal RCT data” for device approval, but he did not offer an alternate source of data.

Martin Leon, MD, the principal investigator of the Endeavor II trial and professor of medicine at the Columbia University Medical Center (New York), said that the Endeavor II trial showed that the device reduced target vessel failure from 15.1% to 7.9%, compared to the Driver BMS, also made by Medtronic. He said that the data showed a 55% drop in target vessel revascularization and a 61% reduction in target lesion revascularization. “We can conclude [from the data] there is a similar safety profile” and that “there were improved angiographic results.”

The Endeavor III trial used the Cypher, made by Cordis (Miami Lakes, Florida), as a control device and randomized 436 patients in a 3:1 ratio of study device to control device.

There was no thrombosis in either arm at 30 days, but Leon said that Endeavor III “did not meet its primary endpoints for lumen loss” at eight months. The Endeavor performed better on freedom from target-vessel failure up to about 220 days, but the gap closed after that time, and neither stent opened a statistically meaningful lead by day 720.

Leon said that Endeavor IV trial data, “presented for the first time today,” randomized more than 1,500 patients, with 773 patients getting the Endeavor and 775 the Taxus. At 30 days, the Endeavor arm accounted for three cases of thrombosis compared to only one for Taxus. But the Taxus was associated with higher rates of infarction, mostly of the non-Q wave type. Roughly half of the 17 non-Q wave infarctions among the control subjects were peri-procedural.

The primary efficacy endpoint result was a rate of target-vessel failure at nine months, which for the Endeavor was pegged at 6.8%, whereas the Taxus was tabulated at 7.4%. Event-free survival for target-vessel failure for Endeavor clocked in at a higher rate in the early going of Endeavor IV, but, again, the gap had almost completely closed by day 270.

One panelist asked what advantage Endeavor offered over Taxus.

Leon said, “I believe that this device has comparable efficacy, but has superior deliverability and safety,” perhaps a reference to the higher peri-procedural incidence of infarction for the Taxus.

For patients who seem somewhat high-risk, Leon said he would prefer the Endeavor. While he said that deliverability is “subjective,” he described Driver deliverability as superior, and the Endeavor compared well to it.

During the afternoon open public hearing, Bill Maisel, director of pacemaker services at Beth Israel Deaconess Hospital (Boston) and the former chair of the panel, said “[c]ontrary to popular belief, FDA got it right” with regard to approval of the Taxus and the Cypher.

Concerning follow-up studies, he said he didn’t want the current panel to “be doomed to repeat our mistakes.” Maisel said that due to inadequate follow-up studies, “patients and physicians were bombarded with conflicting headlines” regarding DES. He said that in the Swedish studies enrollees receiving DES devices “were sicker patients, and [study authors] supposedly adjusted for it,” but that their adjustments were not up to par.

“We need to study physician bias” to determine why they choose a given stent in a given patient, given co-morbidities and so on, Maisel said. He said that the original proposed surveillance for the other two DES devices “was reasonable,” but noted a significant benefit for longer follow-up.

As for control choice, he said that the use of BMS devices “is a little bit like comparing apples with oranges,” and that a better control group for post-approval studies would be a concurrent registry with patients who have other DES models, for up to 10,000 enrollees. He said that blinding of registries is unnecessary and an impediment to publication of valuable data.

Scott Ward, president of Medtronic’s vascular division, told Medical Device Daily saw the concerns expressed by panel members as genuinely important.

Regarding a more aggressive effort to communicate with doctors concerning off-label use of the Endeavor, assuming FDA approval, Ward said this isn’t an area “in which we have many options.” He said only that “it’s important for doctors to use these products in accordance with their best medical judgment.”

Ward said that Medtronic sees a lot of potential in the stent market. There are “still signs of unmet need in heart disease,” he said, and that the company has “a strong pipeline of products,” including one that may prove more effective for high-risk patients.