Background. Dual antiplatelet therapy reduces ischemic events after percutaneous coronary intervention (PCI) and in patients with acute coronary syndromes. The relationship between target vessel revascularization (TVR) and ischemic events in patients treated with aspirin and clopidogrel or aspirin alone from 1 month to 1 year after PCI has not been studied. Methods. Patients enrolled in the CREDO trial were treated with aspirin and clopidogrel or aspirin and placebo for up to 1 year. We compared the rates of TVR and ischemic events (cardiac death, myocardial infarction or stroke) in the two groups, and modeled the effect of clopidogrel treatment on ischemic events after adjusting for relevant parameters. Results. One month after PCI, 1,955 patients have remained asymptomatic. By 1 year, ischemic events occurred in 5.3% of placebo- and 3.1% of clopidogrel-treated patients; p = 0.02. The rate of TVR was 11.9% and 12.2%, respectively; p = 0.82. Only 7 patients (clopidogrel: 3 and placebo: 4) experienced TVR within 7 days of an ischemic event. After adjustment, long-term dual antiplatelet therapy was associated with a 48% reduction in events; p = 0.01. Patients who experienced TVR had a significantly higher rate of ischemic events than those without TVR, regardless of treatment assignment: 12.3% vs. 3.1%, respectively; p < 0.001. Conclusion. Thus, after successful PCI, prolonged dual antiplatelet therapy reduces ischemic events without affecting TVR. Overall, patients with TVR experienced an ischemic event much more often that was not related to the PCI vessel. This suggests that the benefit of antiplatelet therapy after coronary revascularization is indexed to the patient’s underlying atherothrombotic process, rather than the artery that underwent intervention. J INVASIVE CARDIOL 2007;19:287–290

       When clopidogrel is added to background aspirin therapy, it reduces the composite of cardiovascular death, myocardial infarction (MI) or stroke in patients with acute coronary syndromes (ACS)¹ or percutaneous revascularization.² The most consistent reduction is observed in MI and lasts for at least 1 year after initiation of therapy. Because dual antiplatelet therapy with aspirin and thienopyridines also reduces the incidence of acute and subacute stent thrombosis,^3,4 it may be construed that the reduction in adverse ischemic events is the result of prevention of the need for revascularization in these patients. Nevertheless, little has been reported about the need for (repeat) revascularization and the relationship, if any, between ischemic events and revascularization. Thus, we set out to explore the impact of dual antiplatelet therapy, compared with aspirin monotherapy, continued from 1 month to 1 year after percutaneous coronary intervention (PCI) on the rates of repeat revascularization in the Clopidogrel for the Reduction of Events During Observation (CREDO) trial. Furthermore, we planned to investigate the relationship between the occurrence of an ischemic event and the need for target vessel revascularization (TVR).

Methods

       The CREDO study was reported in detail, including the 1-year follow up results.² In brief, patients with objective evidence for symptomatic coronary artery disease (CAD) who were referred for elective or urgent PCI were randomized to clopidogrel 300 mg as a loading dose 3–24 hours prior to PCI or placebo, in addition to aspirin and other standard therapies. After PCI, all patients received aspirin and clopidogrel for 1 month, after which they were re-randomized to clopidogrel 75 mg daily or placebo for 11 additional months. This post hoc analysis focuses on the period between 1 and 12 months after the index PCI, when randomization to prolonged dual antiplatelet therapy or to aspirin monotherapy was in effect. The endpoints analyzed were all-cause death, MI, stroke and TVR, all adjudicated by an independent committee blinded to treatment assignment. Routine angiography was not recommended and all TVR procedures were clinically indicated. Acute Q-wave MI was defined as the presence of a new significant Q-wave with a duration of at least 0.04 seconds or a depth equal to one-fourth of the corresponding R-wave amplitude in 2 or more contiguous leads. Postdischarge non-Q-wave MI was defined as the elevation of CK or CK-MB isoenzyme of at least twice the upper limit of normal in 2 samples collected at different sampling times. Stroke was defined as a new focal neurologic deficit of vascular origin lasting at least 24 hours. It was further classified as an intracranial hemorrhage, ischemic infarction (if a computed tomographic or magnetic resonance imaging scan was available), or of uncertain cause. Statistical analyses were performed using SAS version 8.2 (Cary, North Carolina) for comparison of continuous or categorical variables, as well as a Cox proportional hazards model for the adjusted association between treatment assignment and events. A p-value < 0.05 was considered statistically significant.

Results

       There were 2,116 patients enrolled in the study and 1,955 qualified for this analysis, as they had not experienced an ischemic or TVR event in the first 28 days after PCI. Their baseline characteristics are depicted in Table 1.

Table 1

Table 1. Baseline characteristics of the study patients (data are presented as median with interquartile range or proportions, as appropriate).

Only 63% of clopidogrel- and 61% of placebo-treated patients completed the entire 1 year of therapy. Most of the discontinuations in both groups were due to patient choice or to adverse events (15.8% for clopidogrel vs. 13.0% for placebo; p = 0.09). The rate of major bleeding up to 1 year was 8.8% for clopidogrel and 6.7% for placebo; p = 0.07. Two-thirds of these events occurred in patients undergoing coronary artery bypass grafting surgery (CABG) (64 of 93 events for clopidogrel and 55 of 71 events for placebo). Between 29 days and 1 year, the cumulative rate of death, MI or stroke was 5.3% for placebo and 3.1% for clopidogrel; p = 0.02 (Table 2).

Table 2

Table 2. Ischemic events in the study population between day 29 and 1 year after percutaneous coronary intervention (p-value for time to event).

The rate of TVR was 11.9% and 12.2%, respectively; p = 0.82 (Figure 1).

Figure 1

Figure 1. Ischemic events and target vessel revascularization (TVR) in the study population between day 29 and 1 year after percutaneous coronary intervention.

Any revascularization procedure was performed in 18.1% and 18.9% of patients, respectively; p = 0.54. Patients with TVR — 109 in the placebo group and 113 in the clopidogrel group — were significantly more likely to experience an ischemic endpoint than those who did not undergo TVR: 12.3% vs. 3.1%, respectively; p < 0.001. Among patients with and without TVR, clopidogrel resulted in a 37–53% reduction in the rate of the primary ischemic endpoint; p < 0.001 (Figure 2),

Figure 2

Figure 2. Ischemic events in patients who did or did not experience target vessel revascularization during the 1-year follow up.

and there was no statistical interaction between TVR and treatment assignment (p = 0.50). We further explored the relationship between TVR and ischemic events. There were only 7 patients (3 treated with clopidogrel and 4 with placebo) who underwent TVR within 7 days of an ischemic event. Exclusion of these patients did not affect the results presented in Figure 2. Among patients who completed the 12 months of therapy, the incidence of death, MI or stroke was 3.8% for the placebo group (n = 23/605) and 2.3% for the clopidogrel group (n = 14/611); p = 0.13. The rates of death were 2.0% vs. 1.8%, respectively; p = 0.81, and MI occurred in 1.7% and 0.8% of patients, respectively; p = 0.19. There was no difference in TVR rates or in the rate of any revascularization procedure. The time to first ischemic endpoint was modeled in a multivariate analysis which included the following parameters: age, gender, race, cardiac risk factors, indication for PCI, prior MI, PCI or CABG, history of heart failure, baseline medications, renal function, hematological parameters and treatment assignment. Clopidogrel treatment was associated with a 48% reduction in time to first event (hazard ratio [HR] 0.52, 95% confidence interval [CI] 0.32–0.86; p = 0.01). The only other parameters independently associated with outcome were age > 70 years (HR 1.12, CI 1.06–1.17; p < 0.001), diabetes mellitus (HR 1.9, CI 1.7–3.2; p = 0.01), history of heart failure (HR 2.3, CI 1.3–4.2; p = 0.004), history of CABG (HR 2.3, CI 1.4–3.9; p = 0.004) and a white cell count in the second quartile (HR 0.29, CI 0.13–0.69; p = 0.005). A similar analysis was performed for TVR. Treatment with clopidogrel was not significantly associated with TVR (p = 0.65). Only increasing age (p = 0.05), prior angiography (p = 0.003) and declining renal function (p < 0.001) were predictive of TVR between 1 month and 1 year after randomization. After exclusion of patients who did not complete the full treatment assignment, clopidogrel was still associated with a 30% reduction in ischemic events (HR 0.70, 0.35–1.38) but this association was no longer statistically significant; p = 0.30.

Discussion

       The principal findings of this analysis are that, in patients who remain asymptomatic 4 weeks after their stent-based PCI, dual antiplatelet therapy with aspirin and clopidogrel for up to 1 year, compared with aspirin alone, significantly reduced the rate of ischemic events without affecting the rate of TVR. For every 1,000 patients treated, 22 ischemic events were prevented. Since the events prevented occurred beyond the first month (and in addition to the benefit observed in the first 28 days), and with more than a million coronary intervention procedures performed in the United States each year, the implication of this finding is quite important. Since dual antiplatelet therapy prevented ischemic events not related to the target vessel, these findings remain important in the era of drug-eluting stents and provide an additional impetus for prolonged therapy beyond the prevention of late stent thrombosis. Furthermore, there is an intriguing relationship between the occurrence of TVR and ischemic events, such that patients who underwent TVR were significantly more likely to also experience an ischemic event, regardless of treatment assignment. The timing of these events does suggest that they are not causally related, but rather that both MACE and TVR events are markers of a heightened propensity for future plaque rupture as well as restenosis. In the group of patients who underwent TVR, 85 such events were eliminated for every 1,000 patients treated because of the much higher rate of events. The reduction in ischemic events is attenuated by the increased risk of bleeding, which mostly occurred early and was related to surgical revascularization. The implications of these observations are critical in a number of ways. First, the data reinforce the concept that the lengthy treatment with dual antiplatelet therapy prevents future ischemic events rather than clinically indicated TVR. Indeed, data from the large multicenter Can Rapid risk stratification of Unstable angina patients Suppress ADverse outcomes with Early implementation of ACC/AHA Guidelines (CRUSADE) registry highlight this observation.⁵ While over 90% of patients treated with PCI were discharged on clopidogrel, fewer than 50% and 25% of those receiving medical therapy or CABG, respectively, were treated with clopidogrel, and thus did not benefit from its ability to prevent ischemic events. In the same report, Tricoci et al show that patients not receiving clopidogrel on discharge were also less likely to be treated with other guidelines-mandated therapies such as aspirin, beta-blockers, statins and risk factor modification, suggesting an overall lower quality of care. Second, the association between TVR and ischemic events is tantalizing. If TVR is primarily due to scar tissue formation in (bare-metal) stents, and ischemic events are related to plaque destabilization and thrombosis, and the two are not temporally related, it would mean that a proximate cause such as inflammation might be responsible for both. In our study, patients who underwent TVR had a 4-fold higher rate of ischemic events and a similar relative risk reduction with clopidogrel — about 40% — as those who did not need TVR. Thus, prolonged dual antiplatelet therapy attenuates the risk of ischemic events in those who also undergo TVR, without affecting the rate of TVR itself. The effects of clopidogrel on inflammatory markers, such as C-reactive protein (CRP) and CD40L have been studied in small cohorts, treated with or without PCI. As compared with aspirin alone, the addition of clopidogrel decreased the rate of platelet -monocyte and platelet-neutrophil aggregates,⁶ and suppressed the increase in CD40L after stenting⁷ without affecting levels of CRP.^8,9 It is notable that in a higher-risk group of patients treated with PCI in the setting of an acute coronary syndrome and a heightened inflammatory state, prolonged dual antiplatelet therapy compared with aspirin alone was associated with a 25% reduction in death or infarction from 1 month to 1 year of follow up (6% vs. 8%; p = 0.047), as well as an 18% reduction in the need for any revascularization (14.2% vs. 17.1%; p = 0.03).¹⁰ The association between white cell count and 1-year events supports the role of inflammation in the causation of ischemic events remote from the target vessel. In this report, there is no separation between TVR and other revascularization procedures, but it is likely, in view of our data, that the difference is in the need for revascularization related to plaque rupture, and not to restenosis.

Study limitations

       This is a post hoc analysis of the CREDO trial and thus suffers from the inherent limitations of such an approach. TVR was not a prespecified endpoint. Most importantly, a substantial minority of patients (nearly 40%) did not complete the prespecified duration of treatment, and we could not assess whether the events occurred while on treatment or not. In those completing the 1 year of treatment, similar trends were observed for a reduction in ischemic events, as in the whole cohort. It is possible that better compliance would have increased the difference between the two groups. Finally, we cannot exclude the possibility that the mechanisms responsible for TVR (such as vascular inflammation in the treated segment) were also responsible for a higher rate of ischemic events unrelated to the treated segment. Studying these events in patients treated with drugeluting stents and having a much lower rate of restenosis may clarify this issue.

Conclusions

       Nonetheless, we conclude that prolonged dual antiplatelet therapy is crucial in the prevention of ischemic events after PCI, beyond the benefit observed in the first 28 days. Our data reinforce the observations from registries and studies of ACS suggesting that the benefit is not related to prevention of TVR, but rather to attenuation of risk for future episodes of plaque destabilization. Since we cannot predict very well who will develop the need for TVR, and since these patients are at much higher risk for ischemic events as well, it is worthwhile to highlight the efficacy of prolonged dual antiplatelet therapy after PCI.