Are drug-eluting stents dangerous?

Avalanche of new research suggests millions at risk of thrombosis

By Owen Dyer

Potential complications of Coronary Stenting: Late Thrombosis in a Drug-Eluting Stent credit: Source: Curfman, GD. N Engl J Med 2007; published at www.nejm.org on Feb 12, 2007. Copyright 2007 Massachusetts Medical Society. All rights reserved

We may live in an age of astonishing medical discoveries, but lately it seems that new medicines and techniques are being discredited faster than they can be invented. Latest to be touted then doubted is the drug-eluting stent. If you ever felt that more research was needed on this technique, your wish has been granted and then some.

The New England Journal of Medicine has published five research articles and two commentaries on the subject over the past month, while another important review and advisory appeared in the journal Circulation. Last month the Canadian Medical Association Journal attacked the problem in some detail. And in December, the US Food and Drug Administration held a series of meetings on questions surrounding drug-eluting stents — meetings that some observers anticipated might even lead to a suspension of their licence.

That hasn't happened, and it looks increasingly unlikely. But the FDA remains nervous about drug-eluting stents, which have been implanted in about four million Americans. The bugbear that has everyone so concerned is late thrombosis, a potentially fatal complication.

A BAD VIBE
Needless to say, with so many studies coming out at once, including some that are themselves meta-analyses of multiple studies, there's something for everyone in the results. The manufacturers of the two drug-eluting stents currently on the market claimed to draw comfort from the research.

But the overall picture that emerged is rather negative. Not one of the five NEJM studies found a long-term survival benefit from drug-eluting stents when compared to bare metal stents. Several found that drug-eluting stents reduced the need for revascularisation of the target lesion, but in all cases this benefit was offset by increased rates of late thrombosis.

There's also concern that even these latest studies aren't following patients for long enough. Most of them stretched to three or four years. But while complications associated with bare metal stents — namely revascularisation — tend to occur early, thromboses in drug-eluting stents are often cropping up later in the study period. The obvious implication is that with longer follow-up, drug-eluting stents might prove significantly more dangerous.

MORE DRUGS PLEASE
The long-term persistence of clotting danger with drug-eluting stents has given rise to a parallel debate over how long patients should be given antiplatelet prophylaxis — usually thienopyridine (clopidogrel or ticlopidine) in conjunction with aspirin.

But the thienopyridine is often discontinued as early as three months after stent placement. There's a growing consensus that this is far too early even with bare metal stents, and especially with drug-eluting stents.

Surprisingly, the NEJM research articles barely address this question. But it is definitely causing plenty of head-scratching at the FDA. In January, the journal Circulation carried an advisory from the American Heart Association recommending 12 months antiplatelet therapy with all stents. The FDA has now endorsed this position in a commentary published in the NEJM. And they're clearly leaving the door open for even longer extensions.

That means that both drug-eluting stents on the market, the sirolimus-eluting Cypher stent and the paclitaxel-eluting Taxus stent, now have inappropriate labelling, since they recommend three and six months respectively.

MATH = TROUBLE
A few back-of-the-envelope calculations can show why the FDA is so obviously worried by the developing controversy over drug-eluting stents. Millions of patients already have them, and the rate of stent thrombosis suggested by research is at least 1%.

Each stent thrombosis is a grave event, with two-thirds of cases leading to myocardial infarction. Death rates from stent thrombosis are estimated at between 20 and 45% — let's assume 33% for the sake of argument. That means that even if no more are implanted, 40,000 Americans are going to have thromboses in their drug-eluting stents. Twenty-six thousand will have heart attacks. Thirteen thousand will die.

The really scary news is that these baseline assumptions come from trials of patients who met the FDA's approval criteria for drug-eluting stents: newly diagnosed coronary lesions, less than 30 mm long, in clinically stable patients without additional serious medical conditions.

By the FDA's own calculations, only 40% of patients treated with drug-eluting stents met these criteria. The rest are getting off-label treatment, typically for much more serious and complicated conditions.

PENNY WISE
So where does all of this leave Canada, which approved these stents a year earlier than the US? Surprisingly, much better off than might be feared.

We may have dodged the bullet thanks to the conservatism — what some might prefer to call the stinginess — of our public health system. While bare metal stents go for about $800 here, the drug-eluting stents cost two to four times as much. In the absence of clear data showing a survival benefit, Canadian hospitals were loath to shell out the extra money. So while 80% of stents used in the US today are drug-eluting, in Canada the proportion is about 30%.

Moreover, percutaneous coronary intervention is much rarer in Canada. While the US sees about 750,000 such procedures a year, here there are only about 35,000, about half as many per capita. So a Canadian cardiac patient is less than a quarter as likely as an American to be outfitted with a drug-eluting stent.

BACK TO THE FUTURE
A commentary in the Canadian Medical Association Journal in December noted that in cardiac medicine, "the interventional community is often accused of making decisions based on faith more than on fact".

It is indeed hard now to see the evidence that signposted the route from the treatment of 30 years ago to that of today. Angioplasty was adopted without a clear demonstration that it offered a survival benefit over drugs. Then stents were added without demonstrating they were better than balloon-only treatment. Then drug-eluting technology was thrown in without a clear scientific basis.

Indeed, a December article in NEJM appeared to bring us full circle. In 2,166 MI patients with total occlusion of the infarct-related artery, those randomized to stenting achieved no benefit over those given simple medical treatment, 1970s-style. The study questions whether reopening the artery is even a good idea.

QUESTIONABLE MOTIVES
The January 16 CMAJ provided what may be the most enlightening piece of research yet. Researchers from Holland and from Montreal's McGill University reviewed the quality, provenance and funding of many of the original cost-effectiveness analyses of these devices. Of 19 identified studies, 10 supported widespread use and nine were more cautious.

But only nine studies met criteria for "high quality", and only one of these recommended widespread use of drug-eluting stents, compared to nine of 10 lower quality studies. All seven of the studies that were sponsored by manufacturers recommended widespread use, compared to just three of 12 unsponsored ones.

All five studies that came from the US recommended widespread use. Of the 14 others, nine were against widespread use, including all four Canadian ones.