CDU Washington Editor
GAITHERSBURG, Maryland — Whatever the debates concerning the most appropriate lengths of trials and the most appropriate endpoints to be used in judging drug-eluting stents (DES) vs. bare metal stents (BMS), the most important endpoint is likely to be economic — specifically, what the healthcare market will bear and what reimbursement is willing to pay.
While DES suffered a marked sales slump in 2006, the apparent fall-out from a variety of new clinical reports concerning thrombosis as well as general media coverage, most sector analysts are predicting a rebound and continued dominance of DES over BMS in 2007 and beyond.
And the recent two-day consideration by the FDA of recently identified DES concerns would indicate that the continued demand for DES over BMS won't be impeded by any new actions by the agency, at least not for DES devices already approved. But an obvious issue to be watched closely is whether new DES ready to come online — and other device/drug combinations — will have to pass tougher tests proposed by the agency.
Overall, while the two-day meeting featured plenty of criticism of the manufacturers of the devices and the possible bias in their trials, the clinicians presenting at the meeting tended to provide a balance of pro and con in the debate. That sort of balance clearly served to support a status quo position by the agency, but also a wait-and-see attitude that suggests the door will be held open to additional longer-term evidence.
Healing not uniform
During the public comment session, Herman Gold, a professor of medicine and an interventional cardiologist at Massachusetts General Hospital (Boston), said that "although earlier pivotal studies indicate that these stents are safe, with longer follow-up, it has become clear that late stent thrombosis may impose significant risk for some patients." Gold said that "healing is not uniform within a DES, and strut coverage is not uniform, either."
The coating toward the middle of the stent is more subject to depletion, he said, and that a 37% exposure of the stent is associated with an increase in the risk of latent stent thrombosis (LST) of 11%. However, the most powerful predictor of thrombosis, Gold said, is lack of endothelial coverage on the inner lining of the blood vessel, which DES suppresses inadvertently while suppressing smooth muscle proliferation.
Alan Michelson, MD, a professor of medicine at the University of Massachusetts Memorial Hospital (Worcester, Massachusetts), is credited with having authored the definitive book on platelets. He informed the panel that LST "is associated with discontinuation of anti-platelet therapy in some patients, but not all," and that in some patients, pharmacological resistance is the culprit rather than discontinued therapy.
And he argued that "a clinically meaningful definition of aspirin resistance needs to be developed," but that the science behind this use of aspirin is not very robust due to lack of studies.
He did not recommend trying to test aspirin resistance with its use in an anti-platelet setting "other than [in] research trials, which I strongly encourage." He added that a similar examination of clopidogrel [Plavix] is also in order. Nonetheless, he recommended that clopidogrel doses be boosted from 75 mg to 150 mg a day in patients for whom sub-acute thrombosis could lead to disastrous outcomes.
Discernible LST increase
One of the questions the agency had tasked the panel to investigate was whether DES are associated with a discernible increase in the rate of LST compared to BMS. Christopher White, MD, of the Ochsner Clinic (New Orleans), said that where rates of thrombosis are concerned, "the number spoke for itself," but he said he was nonetheless uncertain that the rate of thrombosis is elevated. "I'm worried but not certain" about thrombosis, he said, adding, "I think death and MI are not" conspicuous in the DES group.
Sharon-Lise Normand, a professor of public healthcare policy and a biostatistician at the Harvard School of Public Health (Boston), said she didn't see "a strong signal of thrombosis."
Norman Kato, MD, of the Cardiac Care Medical Group (Encino, California), said that "I've been torn with some of what I see as an emotional issue." He said that the data seem to bounce around from a 1% to a half-percent frequency [of thrombosis], and that if he had to answer the question it would be "emotionally, yes, evidence-based, no."
After the first day of deliberations, Panel Chair William Maisel, MD, of Beth Israel Deaconess Hospital (Boston) said, "I think we have made essentially no change to current recommendations."
2nd-day focus on off-label use
The second day of the panel deliberation was both more interesting, and more complicated. It featured a variety of scorching remarks aimed at the companies making DES devices. But seeming to carry more weight were registry data and testimony by administrators at two locations in the Kaiser Permanente network.
Additionally, the panel's deliberations on the second-day of the proceedings were bound to be more controversial as it attempted to tackle the off-label uses of DES, up to 60% of overall use, according to an agency presenter. That use clearly involves more complex cases and more difficult, or difficult-to-assess, outcomes. Still, the data were too equivocal to support anything more drastic than that DES labels recommend dual antiplatelet therapy continue for a year, and more cautious recommendations to surgeons for DES use.
Ralph Brindis, MD, interventional cardiologist with Northern California Kaiser Permanente (San Francisco), said that his organization as a whole was "impressed with the marked reduction in clinical restenosis and improvement in quality of life" seen in patients implanted with DES rather than BMS.
Kaiser data for more than 4,200 patients over three years, ending Oct. 31, 2006, showed that more than 95% of those who underwent angioplasty chose to have a stent installed, and in that group 95% opted for DES.
David Magid, MD, director of the Colorado Permanente Clinical Research Unit (Denver), noted that data on patient risk characteristics showed that 74% of Kaiser patients who chose a DES "were receiving stents in the setting of non-FDA-approved indications."
Magid said that among the predictors of post-surgical infarction, diabetes mellitus correlated with the strongest hazard ratio at a rate of 3.2, with previous graft surgery coming in second at 2.6. Lesion length of greater than 30 mm exhibited a hazard ratio of 1.9, followed by congestive heart failure (CHF) at 1.8. As for the predictors for all-cause death, the hazard ratio for CHF led all groups at 3.8, and bifurcated lesions followed at 1.8.
Age and diabetes were the third and fourth most powerful predictors at 1.7 and 1.6.
Plavix protects early
On the use of clopidogrel beyond the recommended term, Magid said that the in-house nature of the organization's prescription drug benefit "allows us to carefully track" compliance. The data show that "we have seen a protective effect of Plavix in that three month period beyond the recommended period of time, and there is also a trend toward protection in the second three-month period, but there were no differences" after that.
"There is no evidence of any protective effect associated with clopidogrel use beyond" a six-month addition to the recommended duration of use for either type of stent among the Kaiser patient population, he said.
Sanjay Kaul, director of the vascular physiology and thrombosis research lab at Cedars Sinai Medical Center (Los Angeles), said that DES devices offer "a quantum leap in interventional cardiology," but that they predispose patients to thrombosis. And he posed one of the key questions: Do the clinical trials reflect real-world practice in defining occurrence of restenosis?
He also asked what the duration of risk might be for various outcomes and insisted that the FDA's least-burdensome pathway would be improved by insisting on trials that are larger, of longer duration and more inclusive of "real-world" populations.
"Reducing restenosis is desirable if we have ways to mitigate the risk of stent thrombosis," Kaul said, but "the optimal duration" of dual antiplatelet therapy is unknown, the data less than certain due to an information gap regarding compliance. He added that clopidogrel use for this is an off-label use of the drug.
Proof standards more stringent?
"With regard to the approval recommendations … we need larger and longer pre-approval randomized clinical trials" and a broader cross-section of patients and trial designs that emphasize clinical outcomes over statistical measures, " Kaul said. "I think we will all agree that the [resulting] approval process is likely to benefit from" a more stringent standard of proof "than the current standard of least burdensome pathway."
He commended the FDA for "taking the initiative" on setting up surveillance registries and insisted that "the crux of the debate here is the over-utilization of revascularization," adding that the DES vs. BMS debate can't be adequately argued until "the root cause" for the reported adverse events is found. Kaul also suggested that DES use is part of the trend toward defensive medicine and the "need [for] tort reform." He added: "All of this is going to take a lot of willpower."
Jeffrey Moses, MD, an interventional cardiologist at Columbia Presbyterian Physician Network (New York), said, "the negative characteristics [of DES] portrayed in the press have really frightened patients," thus compromising the interventionalist's ability to advise patients.
"Let's not lose perspective on what restenosis is all about," Moses said, arguing that abandoning off-label DES in favor of BMS would induce 300,000 clinical events a year and hundreds of thousands of bypass surgeries each year. The resulting "burden in the [U.S.] healthcare system would be $1 billion" a year, he said, citing Medicare data.
Remarking that the quality of the restenosis is as important as frequency, he said that the majority of restenosis associated with BMS was "diffuse" and that the recurrence rate ranges from 35% to 80%, resulting in major cardiac event rates of more than 50% over two years.
Real-world vs. non-real-world
"Unfortunately, all we're going to have for the foreseeable future are these nine randomized trials" for deciding which stent is better, Moses said.. As for the discussion of real-world and non-real-world data, he said the registry data provide four-year follow-up and clinical measures have improved as a result of widespread adoption of DES, including "profound reductions in restenosis."
The off-label data provided by the ARTS II follow-up of the Cypher stent indicated that after three years, the composite rate of death, stroke, heart attack and revascularization compared to coronary bypass was better for the stent vs. BMS, although the difference was apparently not statistically significant. However, significant differences were seen between the Cypher and BMS devices for this grouping of events as well as for the incidence of stroke and heart attack.
DES "are unquestionably the standard of care," Moses stated, noting that in intent-to-treat trials, "overall thrombosis rates are not very different and term-to-thrombosis is narrowed." He added that events secondary to thrombosis are eliminated from most studies, but that such an exclusion tilts the numbers toward BMS.
Panel chairman Maisel commented that the panel saw "significant uncertainties" and as a result, found it "difficult to make conclusive judgments regarding this issue." On the other hand, the panel felt comfortable in recommending that patients stay on dual antiplatelet therapy for at least a year, and suggested that stent labels inform doctors that high-risk patients do not seem to benefit as much as patients who are covered by the cleared indications.
Steve Nissen, MD, of the Cleveland Clinic (Cleveland, Ohio), said, "I want to warn people a little more than we do now."
On the other hand, Christopher White, MD, of the Ochsner Clinic (New Orleans), said he had not "heard anything today that will change my practice."