Dr Matthias E Pfisterer

Presenting results of the Basel Stent Cost-effectiveness Trial—Late Thrombotic Events (BASKET-LATE) during a late-breaking clinical-trials session here at the American College of Cardiology 2006 Scientific Sessions, Dr Matthias E Pfisterer (University Hospital Basel, Switzerland) acknowledged that his study was underpowered to detect major differences in thrombosis-related events. Still, he insisted, the indisputably increased thrombosis risk with drug-eluting stents warrants new strategies to prevent late thrombosis.

In the short term, Pfisterer told heartwire, longer dual antiplatelet therapy might be the best temporary solution. In the long term, new stents with bioabsorbable platforms or polymers may solve the late thrombosis problem. Testing for clopidogrel resistance may also help avert stent thrombosis related to drug resistance, he added.

Debate over the true risk of late thrombosis continues to dog drug-eluting stents. As Pfisterer noted in his presentation, delayed endothelialization following stent implantation is the proposed mechanism behind the widely acknowledged increased risk of late thrombosis with DES, although the actual real-world thrombosis rate is hotly disputed.

As Dr William O'Neill observed following Pfisterer's presentation, "The real question that clinicians are going to have is: are you really trading late restenosis, which is not a health hazard, for death or reinfarction? That's obviously a very important and worrisome trade-off."

The real question that clinicians are going to have is, are you really trading late restenosis, which is not a health hazard, for death or reinfarction? That's obviously a very important and worrisome trade-off.

As previously reported by heartwire, the original BASKET trial enrolled a relatively complex patient group and randomized them to a bare-metal stent, a paclitaxel-eluting stent, or a sirolimus-eluting stent. For BASKET-LATE, Pfisterer and colleagues followed the 746 patients who were MACE free at six months (when the main BASKET cost-effectiveness trial had concluded and clopidogrel was stopped) for an additional 12 months. The aim of BASKET-LATE was to define the incidence of clinical events related to stent thrombosis—and the timing of these events—in patients treated with either a bare-metal stent or a DES. For the purposes of the analysis, sirolimus- and paclitaxel-eluting stent-treated patients were combined in a single group, late stent thrombosis was defined as at least six months after PCI, and "thrombosis-related events" encompassed angiographically confirmed stent thrombosis, sudden cardiac death, or a target vessel MI.

Somewhat to the surprise of his audience here today, Pfisterer showed that while MACE rates were no different between bare-metal-stent or DES-treated patients, rates of nonfatal MI and cardiac death/nonfatal MI were actually significantly higher in DES-treated patients. By contrast, and as expected, target vessel revascularization was higher in the bare-metal-stent group, although when only restenosis-related TVR was included, this difference was not statistically significant.

When event rates for cardiac death/MI as well as rates of TVR were plotted over the full 18 months after stent implantation, the curves for TVR rates for DES vs bare-metal stents separated at approximately four months, in keeping with observations from the major DES trials (p=0.02). For cardiac death/MI, however, the curves for bare-metal stents and DES were more or less superimposable until approximately eight months, at which time event rates for bare-metal stents appear to plateau, while the rate for DES continue to climb (p=0.06). Timing of thrombotic events appeared to be evenly distributed across the 12 months after clopidogrel discontinuation.

While the stent thrombosis rate was not statistically different between the DES- and bare-metal-stent-treated patients, in absolute numbers, angiographically documented late stent thromboses were twice as high in DES-treated patients, while thrombosis-related clinical events were nearly three times as high. In multivariate analyses, the most important predictors of late cardiac death/MI were prior MI, initial need for GP IIb/III inhibitors (reflecting troponin-positive ACS patients, more complex lesions, or suboptimal revascularization) and DES use, Pfisterer said.

"With currently recommended antiplatelet therapies, one has to balance in real-world DES use the benefit of lower TVR rates with the cost of an increased rate of late cardiac death/MI," Pfisterer concluded. He estimated this cost/benefit to be 3.3 late death/MIs for every five TVRs avoided per 100 patients receiving a DES.

Asked why his findings appeared to depart so dramatically from pivotal trial results, Pfisterer emphasized that BASKET-LATE reflected a much higher-risk-patient cohort, of whom almost 60% were unstable-angina or STEMI patients.

Discussing the results following the presentation, Dr E Murat Tuzcu (Cleveland Clinic, OH) charged Pfisterer of overstating his findings.

"You have not shown a statistically significant difference in late stent thrombosis and thrombosis-related events [between the DES and bare-metal-stent groups], but you make sweeping conclusions in that regard. Moreover, some of the presentation and outcome of late events could be influenced by their definitions as 'thrombosis-related events,' which were not statistically different between the two groups. So do you really think you can make those kind of sweeping conclusions? This is strikingly different from what we have seen in other studies."

Tuzcu also questioned whether nonfatal myocardial infarction at 200 to 300 days after cessation of clopidogrel treatment could "confidently" be attributed to lack of dual antiplatelet therapy.

This is strikingly different from what we have seen in other studies.

In response, Pfisterer acknowledged that the major limitation of his study was that it was not powered to look for infrequent events. "In fact, if you wanted to do a study to look at these small differences, you would need a study with tens of thousands of patients, and such a study will most likely never be performed," Pfisterer said. "But we had the unique possibility to look at this phenomenon in our patient group, and in this sense, it is a valuable study. Yes, you can criticize the conclusions for being a little bit stretched, but it's what these data would suggest."

Commenting on the study for heartwire, Dr William Gray (Columbia University, New York, NY) said that the results were intriguing but pointed out that in many centers, physicians are already prescribing clopidogrel for one year.

"The data are interesting," Gray stated. "Late-stent-thrombosis rate is important, and if you looked retrospectively across all the trials with follow-up out to two years, the data are probably there, but it's probably not as profound as this. We're probably going to have to do something larger than this, and probably with longer-duration clopidogrel."

The clopidogrel requirement, however, raises a host of other issues, including clopidogrel resistance and drug cost. In some cases, patients may not be able to afford long-term clopidogrel therapy, and the potential increased risk of late thrombosis following drug-eluting-stent implantation may negate the benefits of drug-eluting stents if patients stop their clopidogrel early.

"The cost of clopidogrel is not insignificant," Gray acknowledged. "It is a problem for some patients who we treat and it can influence our clinical decision-making as to what kind of stents we implant. In some cases, a bare-metal stent may be the better option."