The SORT OUT II trial showed that the first generation of drug-eluting stents, Cypher® and Taxus®, exhibit similar efficacy and safety when used for percutaneous coronary intervention in everyday practice, including in a high proportion of off-label situations.
The early drug-eluting stent trials launched to obtain US Food and Drug Administration approval of these devices were performed using highly selected, low-risk patients and simple coronary lesions.1,2 The trials showed that drug-eluting stents cut the restenosis rate by half compared with bare-metal stents, thereby reducing the rate of target lesion revascularizations and major adverse cardiac events. In clinical practice, the case mix of patients treated with percutaneous coronary intervention (PCI) is more heterogeneous and includes high-risk patients and complex lesions, such as patients with ST-segment elevation myocardial infarction (MI), long or ostial coronary lesions, and chronic total occlusions.3,4 Evidence regarding the efficacy and safety of drug-eluting stents in these “off-label” situations has been limited. Because these lesions and patients have increased adverse event rates with bare-metal stents, it has generally been thought that the benefits of drug-eluting stents could be transposed to all patients admitted for PCI. Off-label use of drug-eluting stents, therefore, quickly became predominant.5-7
The SORT OUT II (Danish Organization on Randomized Trials With Clinical Outcome) trial was designed to compare the first 2 available drug-eluting stents—the sirolimus-eluting Cypher® stent and the paclitaxel-eluting Taxus® stent—in unselected subjects, with the use of symptom-driven clinical end points. After subject inclusion in the trial was completed in early 2006, the global enthusiasm for drug-eluting stents was severely shaken by emerging reports of stent thrombosis, especially with drug-eluting stents. A “firestorm” was raised after data from the SCAAR (Swedish Coronary Angiography and Angioplasty Registry) trial suggested that there were increased rates of stent thrombosis and death with drug-eluting stents compared with bare-metal stents.8,9 Complete endothelial coverage of drug-eluting stents, which is likely to reduce the risk of stent thrombosis, is delayed by the antiproliferative drug; therefore, it has been suggested that patients with drug-eluting stents require prolonged and potentially indefinite dual antiplatelet therapy with clopidogrel and low-dose aspirin to avoid stent thrombosis. We provide an updated summary of results from the SORT OUT II trial, with a special focus on stent thrombosis and the effect of discontinuation of dual antiplatelet therapy after 1 year.10
Subjects and methods
The SORT OUT II trial was a national study performed in Denmark. During the randomization period, 11,766 subjects were treated with PCI, and 7556 received drug-eluting stents. Of these, 2098 (28%) were randomly assigned to receive either Cypher or Taxus stents by investigators in the 5 Danish university centers that are responsible for providing round-the-clock PCI to the entire country (5.4 million inhabitants). Inclusion criteria were use of drug-eluting stents at the operator’s discretion and provision of written informed consent. Subjects who did not live in Denmark or did not speak Danish were excluded. Dual antiplatelet therapy was recommended for 12 months, and aspirin therapy was recommended for life.
All subjects were followed until death, emigration, or 18 months after entry into the study. Events were detected in national registries, including the Denmark National Heart Registry, in which all coronary angiographies, PCIs, and bypass operations were registered. All events were adjudicated by an independent end point committee. The primary end point was a composite clinical end point of major cardiac adverse events, defined as cardiac death, acute MI, and target lesion or target vessel revascularization. Of note, the study protocol was distinguished from that of other stent trials by not scheduling a preplanned repeated coronary angiography within 1 year after PCI. Secondary end points were individual components of the composite end point, all-cause mortality, and stent thrombosis as defined by the Academic Research Consortium.11
Approximately 50% of subjects presented with acute coronary syndromes, 17% of subjects had ST-segment elevation MI, and 60% of coronary lesions were high-risk C-type off-label lesions.4 The Cypher and Taxus groups did not differ significantly in the rate of major adverse cardiac events (9.3% vs 11.2%, respectively; hazard ratio [HR] = 0.83; 95% confidence interval [CI], 0.63-1.08; = .16) or in any of the secondary end points. Specifically, there were no significant differences in the rate of stent thrombosis between the Cypher and Taxus stents (2.5% vs 2.8%, respectively; HR = 0.90; 95% CI,
0.54-1.51; = .70; Figure 1). The stent thrombosis rate was relatively constant after 1 month, regardless of the stent thrombosis definition,and there was no detectable rise inthe rate of stent thrombosis at the cessation of treatment with clopidogrel after 1 year (Figure 1). In addition, no increases in the rates of acute MI or cardiac death were observed after 1 year (Figure 2).
The SORT OUT II trial showed that in clinical practice with a substantial level of off-label situations, there were no significant differences between the Cypher and Taxus stents. Two recent meta-analyses of trials with head-to-head comparisons of these drug-eluting stents have shown comparable results, although one of them suggested that the risk of stent thrombosis was substantially lower (HR = 0.66) with the Cypher stent.12,13 This contention cannot be supported by the SORT OUT II results. In this regard, aside from our avoidance of preplanned repeated angiography, the definitions of stent thrombosis used in the earlier studies ranged from “definite” (stent thrombosis shown on angiography or at autopsy in the setting of acute coronary syndrome), to “probable” (acute MI in the stented coronary territory or unexplained death within 30 days after PCI), to “possible” (unexplained death after 30 days).11 The rate of definite stent thrombosis undoubtedly represents an underestimate of the true stent thrombosis rate, whereas the sum of definite and probable stent thrombosis is probably a more valid reflection of the reality, and the sum of definite, probable, and possible stent thrombosis represents an overestimate.
Figure 1 shows the SORT OUT II results for each definition of stent thrombosis. Clearly, there were no apparent changes in the slopes of any of the graphs at the time of cessation of dual antiplatelet therapy after 1 year. The highest estimate of the stent thrombosis rate, that is, the sum of definite, probable, and possible stent thrombosis, reached approximately 1% within the first month, and these cases of subacute stent thrombosis were likely to be primarily related to issues associated with the technique, for example, suboptimal deployment or edge dissections.14 Thereafter, the rate of stent thrombosis was constant at approximately 1 stent thrombosis per 1000 subjects per month across the whole observation period, with no additional signal at the time of cessation of dual antiplatelet therapy. We found no increases in the rates of major adverse cardiac events, acute MI, or death or acute MI after 1 year (Figure 2). The latest recommendation stating that “ideally dual antiplatelet therapy should be extended to 12 months” appears to be safe.15
The SORT OUT II trial showed that the first generation of drug-eluting stents, Cypher and Taxus, displayed similar efficacy and safety when used for PCI in everyday practice, including a high proportion of off-label situations.The stent thrombosis rate during the 18-month observation period was low, and the rate was not significantly increased by discontinuation of clopidogrel therapy after 1 year.
These results underline the rationale for the current recommendation of 1 year of dual antiplatelet therapy and lifelong aspirin therapy after implantation of drug-eluting stents. As indicated by the accompanying case report, however, temporary withdrawal of aspirin, clopidogrel, or both, may contribute to stent thrombosis. Discontinuation of dual antiplatelet therapy, for example, for elective noncardiac surgery, with secondarily increased risk of stent thrombosis, must therefore be carefully weighed against the risk of complications caused by increased perioperative bleeding. More studies are needed to determine the optimal strategy in these situations.
The authors have no relationship with any commercial entity that might represent a conflict of interest with the content of this article.