Clopidogrel outshines aspirin for secondary prevention in high-risk PCI patients

In the long-term maintenance period after percutaneous coronary intervention (PCI), monotherapy with clopidogrel appears to work better than aspirin at preventing major adverse cardiac and cerebrovascular events (MACCE) in Asian patients at high risk of recurrent ischaemic events, as shown in the open-label SMART-CHOICE 3 trial conducted in South Korea.

Over a median follow-up period of 2.3 years, the primary endpoint of a composite of death from any cause, myocardial infarction (MI), or stroke occurred with significantly less frequency in the clopidogrel arm than in the aspirin arm (4.4 percent vs 6.6 percent; hazard ratio [HR], 0.71, 95 percent confidence interval [CI], 0.54–0.93; p=0.013). A total of 45 patients needed to be treated to avert one such event. [Lancet 2025;doi:10.1016/S0140-6736(25)00449-0]

The difference was primarily driven by a reduction in the incidence of MI (1.0 percent vs 2.2 percent; HR, 0.54, 95 percent CI, 0.33–0.90) with clopidogrel vs aspirin, according to principal investigator Prof Joo-Yong Hahn from the Samsung Medical Center in Seoul, South Korea, who reported the findings at the ACC annual meeting.

As for bleeding outcomes, the number of any bleeding (BARC type 2, 3, or 5) and major bleeding (BARC type 3 or 5) events was similar between the clopidogrel and aspirin arms (HR, 0.97, 95 percent CI, 0.67–1.42 and HR, 1.00, 95 percent CI, 0.58–1.73, respectively).

Sensitivity analyses in the per-protocol population yielded consistent results for MACCE (HR, 0.70, 95 percent CI, 0.54–0.92) and major bleeding (HR, 0.95, 95 percent CI, 0.54–1.67), Hahn said.

Consistent effect across subgroups

Clopidogrel monotherapy demonstrated a consistent effect on the primary endpoint across all prespecified subgroups, except the one defined by MI history.

Hahn noted that the cumulative incidence of MACCE was lower with clopidogrel vs placebo among patients without previous MI (HR, 0.56, 95 percent CI, 0.39–0.81) but was not significantly different among those who had had MI (HR, 0.99, 95 percent CI, 0.67–1.49). This finding, he added, may be due to patients without previous MI potentially having a higher risk profile than those without previous MI or due to chance as a consequence of multiple comparisons.

An exploratory subgroup analysis of 731 patients allocated to the clopidogrel group who underwent CYP2C19 genotype testing for treatment effectiveness, the incidence of MACCE did not significantly differ between patients with rapid or normal metabolism vs intermediate or poor metabolism of clopidogrel.

“The SMART-CHOICE 3 trial was the first to demonstrate the benefits of clopidogrel monotherapy compared with aspirin monotherapy on a composite of hard endpoints in patients at a high risk of recurrent ischaemic events after PCI,” Hahn said.

“In general, the more potent antiplatelet therapy increases bleeding risk, but in our study, clopidogrel reduced the ischaemic endpoints … but without increased risk of bleeding [compared with aspirin], so it’s a very ideal result,” he noted.

SMART-CHOICE 3 included 5,506 patients (median age 65.0 years, 18.2 percent female) at high risk of recurrent ischaemic events (previous myocardial infarction at any time before enrolment, medication-treated diabetes, or complex coronary lesions) who completed a standard duration of dual antiplatelet therapy following PCI. A median of 17.5 months after PCI, these patients were randomly assigned to treatment with clopidogrel 75 mg (n=2,752) or aspirin 100 mg (n=2,754), administered orally once daily.

At baseline, 46.3 percent of patients had prior MI, 40.8 percent had diabetes, and 76.0 percent underwent PCI for complex coronary artery lesions.

Statistical concerns, population bias

Hahn acknowledged that the actual event rate was lower than expected, with only around half the number of the primary outcome events (290 vs 402), even though the final sample size was increased by 10 percent from the original sample size. This led to a relatively high risk of a type I error, although a permutation test indicated that the empirical type I error rate was low, he noted.

Additionally, since the trial focused on patients at high risk of recurrent ischaemic events, patients with high bleeding risk were under-represented, which might have contributed to the small number of overall bleeding events, according to Hahn.

Finally, the study population exclusively comprised Korean patients, who often have intermediate or poor clopidogrel metabolism, he noted. That being said, “clopidogrel might in fact be more effective in populations where reduced-function CYP2C19 alleles are less common.”

Robustness of findings questioned

In a linked editorial, Dr Xavier Rossello from the Universitat Illes Balears in Palma de Mallorca, Spain, and Dr Anna Meta Dyrvig Kristensen from the Copenhagen University Hospital—Bispebjerg and Frederiksberg in Copenhagen, Denmark, commented that while SMART-CHOICE 3 tackles the important topic of the choice of antithrombotic regimen in a field seemingly dominated by aspirin, the trial limitations leave the evidence gap only partially addressed.

“A manual calculation of the fragility index (FI) suggests a lack of robustness [in the trial’s findings],” Rossello and Kristensen wrote. They pointed out that the FI, which represents the minimum number of participant changes (ie, from not having the endpoint to having the endpoint) needed to render the results nonsignificant, is notably low.

The FI of seven for the primary composite endpoint and three for the secondary endpoint of MI of SMART-CHOICE 3 falls well below the median of 12 of landmark cardiovascular trials, they said.