Chicago, IL - Like the proverbial cat that just keeps coming back, the issue of early and late stent thrombosis following drug-eluting stent (DES) implantation continues to blemish the largely undisputed benefits of DESs over bare-metal stents. Now, in a new study appearing in the May 4, 2005 issue of the Journal of the American Medical Association, researchers say the incidence of "real-world" stent thrombosis within nine months of DES implantation is "substantially higher" than that reported in the clinical trials.[1]

"In a larger cohort of consecutive patients undergoing drug-eluting-stent implantation, we noted a nine-month cumulative stent thrombosis incidence of 1.3%, substantially higher than rates reported in major clinical trials (0.4% at one year for sirolimus and 0.6% at nine months for paclitaxel)," Dr Ioannis Iakovou (Centro Cuore Columbus, Milan, Italy) and colleagues write.

DES thrombosis was the talk of the 2005 American College of Cardiology meeting, after results from the head-to-head REALITY trial indicated a significant difference in 30-day stent thrombosis rates between the Cypher and the Taxus stents, 0.4% vs 1.8%, respectively (p=0.0196). Since then, a host of papers have tackled the stent thrombosis issue with DES. Dr Raul Moreno et al (Hospital Clinico San Carlos, Madrid, Spain), writing in the March 15, 2005 issue of the Journal of the American College of Cardiology, found no differences between bare-metal and sirolimus- and paclitaxel-eluting stents in their pooled analysis of 10 randomized trials. Stent thrombosis after DES implantation, they concluded, was related to stent length.[2] Likewise, Dr Andrew TL Ong and colleagues (Erasmus Medical Center, Rotterdam, the Netherlands), writing in the same issue, reported no difference between bare-metal and sirolimus- and paclitaxel-eluting stent thrombosis rates out to 30 days and found finger bifurcation stenting in AMI patients increasing stent thrombosis rates.[3] In yet another study, Dr Kenichi Fujii et al (Columbia University, New York, NY) report that stent underexpansion and significant residual reference segment stenosis are key predictors of stent thrombosis.[4]

Iakovou et al's latest addition to the literature examines stent thrombosis both within and beyond 30 days, using the rationale that endothelialization is likely delayed, potentially increasing the risk of thrombosis beyond this initial one-month period. In their prospective, observational study conducted at three hospitals in Germany and Italy, a total of 1062 patients, receiving 2272 paclitaxel- or sirolimus-eluting stents, were enrolled and then followed over nine months.

At follow-up, a total of 1.3% of patients experienced stent thrombosis, nine in the sirolimus group (0.8%) and 20 in the paclitaxel group (1.7%). Half of the thromboses occurred in the first 30 days, but the other half occurred after the first month; overall, 45% of the stent thromboses resulted in death. Independent predictors of stent thrombosis included premature discontinuation of antiplatelet therapy, renal failure, diabetes, and lower ejection fraction. Stent type was not an independent predictor of thrombosis, but the authors acknowledge that the nearly double thrombosis rate seen with the paclitaxel-eluting stent is "of concern" and consistent with findings from ISAR-DESIRE—which demonstrated better overall outcomes for the Cypher stent—as well as the thrombosis results fromREALITY. However, they conclude, "without larger numbers it is difficult to make firm conclusions; subsequent randomized trials will be critically important."

Commenting on the study for heartwire, Dr Adnan Kastrati (Deutsches Herzzentrum, Munich, Germany), primary investigator for ISAR-DESIRE, pointed out that, first of all, the stent thrombosis rate in the DES arm of the current study is not unlike the long-term follow-up studies of bare-metal stents, including the recently published study by Meier et al, which reported a stent thrombosis rate of 1.6%.[5] Second, Kastrati says, it is not clear from the published results how many patients received ticlopidine, "which needs several days before demonstrating its full antiplatelet effect," while patients treated with clopidogrel received only a 300-mg loading dose. "It is now known that 300-mg clopidogrel shows inferior efficacy compared with the 600-mg dose, which is becoming the standard loading dose in many centers," Kastrati said.

In an editorial accompanying the study, Drs Mauricio G Cohen and E Magnus Ohman (University of North Carolina, Chapel Hill) point out that, unlike the pivotal DES trials, Iakovou et al's study enrolled high numbers of patients with diabetes, multivessel disease, small reference-vessel diameters, and complex lesions.[6] More than anything else, they say, the new study should further reinforce the need for long-term aspirin and clopidogrel therapy, since premature discontinuation of these agents was the strongest predictor of thrombosis in the study by Iakovou and colleagues.

As Ohman emphasized to heartwire, "This paper, by a very well-respected interventional cardiology group with high volumes, confirms that this is an issue that we need to pay attention to. And it emphasizes that keeping antithrombotic therapy long-term is something that we seriously have to consider, because the ramifications of patients stopping this therapy prematurely, as this paper points out, are pretty devastating."

Asked whether he felt the different stent thromboses rates seen between the two DESs were important, Ohman warned against overinterpretation of the observational data, despite the fact that they appear to parallel the stent thromboses data reported in REALITY. Still, he said, "If we wanted to establish whether one stent has more stent thrombosis than another DES, we're going to need to do prospective, large, randomized trials. Observational series like this are important in many aspects, but they cannot tell us all the risk/benefit ratios and all the trade-off decisions of one strategy over another. The data are there; I'm cautious about making too much of them. One would want to see larger-scale studies, and I think some of these are under way, so this will come out in the future."

Likewise, Kastrati, whose ISAR-DESIRE data was the first to raise the possibility of different effects between DESs, emphasized that the trend toward a higher risk of stent thrombosis with the paclitaxel-eluting stent compared with the sirolimus-eluting stent "would have been concerning if the data were generated by a specifically designed study.

"Although the conventional characteristics were not statistically different between the two drug-eluting-stent groups, there are often occult factors that guide stent type selection, which may have introduced a bias in this analysis," Kastrati explained. "Fortunately, more than 3500 patients have been included in randomized studies of head-to-head comparison between the paclitaxel-eluting stent and sirolimus-eluting stent in various subsets of patients, and a meta-analysis of these studies would shed more light on this issue."