6-month versus 12-month or longer dual antiplatelet therapy after percutaneous coronary intervention in patients with acute coronary syndrome (SMART-DATE): a randomised, open-label, non-inferiority trial
Selected in The Lancet by D. Giacoppo
Joo-Yong Hahn, Young Bin Song, Ju-Hyeon Oh, Deok-Kyu Cho, Jin Bae Lee, Joon-Hyung Doh, Sang-Hyun Kim, Jin-Ok Jeong, Jang-Ho Bae, Byung-Ok Kim, Jang Hyun Cho, Il-Woo Suh, Doo-il Kim, Hoon-Ki Park, Jong-Seon Park, Woong Gil Choi, Wang Soo Lee, Jihoon Kim, Ki Hong Choi, Taek Kyu Park, Joo Myung Lee, Jeong Hoon Yang, Jin-Ho Choi, Seung-Hyuk Choi, Hyeon-Cheol Gwon
LinkRead the abstract
Why this study – the rationale/objective?
The optimal duration of dual antiplatelet therapy after drug-eluting stent implantation is still a matter of debate. Randomised clinical trials comparing different dual antiplatelet therapy regimens have mostly included patients presenting with stable coronary artery disease and relatively low cardiovascular risk profile. Current guidelines recommend dual antiplatelet therapy for 12 months or longer in patients with acute coronary syndrome mainly as a result of the benefits observed in the PCI-CURE trial, the only available dedicated prospective randomised study, though outdated and presenting several important limitations. Additional data are scant and mainly based on mixed results from observational investigations. Recently, the DAPT-STEMI and REDUCE trials have shown the non-inferiority of shorter dual antiplatelet therapy durations in patients with acute coronary syndrome, however size of samples was limited, number of events lower than expected, and some numerical trends arose concerns. Against this background, the SMART-DATE sought to inspect whether 6 and 12 months of dual antiplatelet therapy in patients who underwent percutaneous coronary intervention for acute coronary syndrome present comparable safety.
How was it executed – the methodology?
The SMART-DATE is a 1:1 randomised, open-label, non-inferiority trial conducted between September 2012 and December 2015 at 31 centres in South Korea.
A total of 2712 patients were assigned to 6-month (n=1357) or ≥12-month (n=1355) dual antiplatelet therapy with acetylsalicylic acid and a P2Y12 inhibitor. All patients received 300 mg of acetylsalicylic acid orally and 300-600 mg of clopidogrel followed by 75 mg daily or 180 mg of ticagrelor followed by 90 mg twice daily or 60 mg of prasugrel followed by 10 mg daily.
Web-based computer-generated randomisation was done at the time of the index procedure and was stratified by site, diabetes, clinical presentation (unstable angina, non-ST segment elevation myocardial infarction, or ST segment elevation myocardial infarction), and P2Y12 inhibitor (after that prasugrel and ticagrelor became available for clinical use). A second randomization process led to the 1:1:1 assignment of three types of drug-eluting stent: zotarolimus-eluting stent, everolimus-eluting stent, and biolimus-eluting stent.
Patients were eligible for inclusion if they had acute coronary syndrome and a ≥50% stenosis amenable for stenting in a native coronary vessel with 2.25-4.25 mm reference diameter. Key exclusion criteria included contraindication to heparin of study medications, active pathological bleeding, major bleeding ≤3 months, major surgery ≤2 months, history of bleeding diathesis or known coagulopathy, and planned elective surgical procedure ≥12 months.
The primary endpoint was a composite of all-cause death, myocardial infarction or stroke at 18 months. Secondary endpoints included the individual components of the primary endpoint, cardiac death, a composite of cardiac death or myocardial infarction, definite or probable stent thrombosis and BARC 2-5 bleeding, BARC 3-5 bleeding (major bleeding), and a net composite endpoint including the components of the primary endpoint or BARC 2-5 bleeding.
By assuming a 4.5% of all-cause death, myocardial infarction or stroke at 18 months and anticipating a 2.0% attrition in each group, a sample size of 2700 patients would guarantee with a 2.0% non-inferiority margin at least 80% power at a one-sided α of 0.05.
What is the main result?
The enrolled population had median age of 62 [IQR 54-71] years and 27.5% of diabetes. The median duration of dual antiplatelet therapy was 184 days (IQR 177–236) in the 6-month group and 531 days (IQR 378–540) in the ≥12-month group. Clopidogrel was used in 1082 (79.7%) of patients assigned to 6-month dual antiplatelet therapy and in 1109 (81.8%) of patients assigned to ≥12-month dual antiplatelet therapy. Follow-up for the primary endpoint was completed in 97.5% of patients.
At 18 months, the primary endpoint was comparable between the 6- and 12-month dual antiplatelet therapy groups (63 vs. 56 events; 4.7% vs. 4.2%; HR 1.13, 95% CI 0.79–1.62, p=0.51; pnon-inferiority=0.03). The treatment effects of 6-month dual antiplatelet therapy compared with ≥12-month dual antiplatelet therapy was consistent across several subgroups, including presence vs. absence of ST-segment elevation and type of P2Y12 inhibitor. However, a significant interaction between the treatment effects of the two regimens and treated vs. untreated left main coronary artery or left anterior descending (pinteraction=0.014).
By reviewing the individual components of the composite endpoint, although death (35 vs. 39 events; 2.6% vs. 2.9%; HR 0.90, 95% CI 0.57–1.42, p=0.90) and stroke (11 vs. 12 events; 0.8% vs. 0.9%; HR 0.92, 95% CI 0·41–2·08, p=0.84) were similar between the 6- and ≥12-month dual antiplatelet therapy groups, a significant excess in myocardial infarction (24 vs. 10 eventg; 1.8% vs. 0.8%; HR 2.41, 95% CI 1.15-5.05, p=0.02) was observed and interestingly Kaplan-Meier curves started to diverge after 6-months. The increased incidence of myocardial infarction seemed to be both target-vessel- and non-target-vessel-related. By applying a 6-month landmark time point, while the risk of myocardial infarction was similar within the 6 months following the procedure (HR 1.28, 95% CI 0.48-3.45, p=0.62), from 6 to 18 months it significantly increased in patients who interrupted dual antiplatelet therapy compared with those taking also the P2Y12 inhibitor for ≥12 months (HR 5.06, 95% CI 1.46-17.47, p=0.01). This result translated into a numerical excess in the primary composite endpoint from 6 to 18 months in the group of patients who interrupted dual antiplatelet therapy at 6 months compared with the group of patients who continued the treatment for ≥12 months (HR 1.69, 95% CI 0.97-2.94, p=0.07). With respect to myocardial infarction at 18 months, a significant interaction between the effects of the two regimens infarction and unstable angina or acute myocardial infarction was observed (piinteraction=0.03) and the significant risk increase seemed to be driven by the subgroup of patients presenting with acute myocardial infarction.
However, cardiac death (18 vs. 24 events; 1.4% vs. 1.8%; HR 0.75, 95% CI 0.41–1.38, p=0.36) and definite or probable stent thrombosis (15 vs. 10 events; 1·1% vs. 0·7%; HR 1.50, 95% CI 0.68-3.35, p=0.32) did not differ between group leaving partially unexplained the increase in myocardial infarction.
Although patients assigned to the prolonged dual antiplatelet therapy regimen tended to bleed more (35 vs. 51 events; 2.7% vs. 3.9%; HR 0·69, 95% CI 0.45–1.05, p=0.09), major bleeding (6 vs. 10 events; 0.5% vs. 0.8%; HR 0.60, 95% CI 0.22-1.65, p=0.33) did not differ between patients assigned to 6- and ≥12-month regimens. The net composite endpoint (96 vs. 99; 7.2% vs. 7.4%; HR 0.97, 95% CI 0.73-1.29, p=0.84) accounting for both ischemic and bleeding outcomes was comparable between groups.
Critical reading and the relevance for clinical practice
The results of this study impose several important considerations. Although in terms of all-cause death, myocardial infarction, or stroke the non-inferiority of 6-month dual antiplatelet therapy compared with ≥12-month duration was met and the net difference between regimens was quite comparable, the results of the SMART-DATE trial still cannot enable a shorter treatment in patients with acute coronary syndrome for several reasons. First, the non-inferiority margin might be too wide, implying limitations in the detection of the equivalence of two therapy durations for the primary composite endpoint. Second, there was a significant increase in the risk of myocardial infarction in patients assigned to 6-month dual antiplatelet therapy that at landmark analysis became evident following interruption of the P2Y12 inhibitor. However, this result did not reflect into higher incidences of cardiac death and stent thrombosis. Third, patients assigned to 6-month dual antiplatelet therapy had a non-significant reduction in bleeding events. However, this trend appears to be significantly mitigated after considering only major bleeding and death was similar between groups without detection of signals favouring one of the two regimens. In aggregate, ≥12-month dual antiplatelet therapy did not entail an increased risk of clinically-relevant or life-threatening bleeding events, while unanswered questions remain about the possible superior ischemic protection of prolonged dual antiplatelet therapy in patients with acute coronary syndrome.
Some limitations of the study need to be taken into account. First, the randomisation was performed at the time of index procedure and the study had open-label fashion leading thus to potential sources of bias. Second, the use of clopidogrel was predominant in both study groups compared with new-generation P2Y12 inhibitors. Third, the number of patients enrolled could not allow exploring differences in terms of individual endpoints and a larger population would have permitted a narrower non-inferiority margin for the primary endpoint. Fourth, protocol adherence significantly differed between 6 and ≥12-month groups (73.7% vs. 95.7%, p<0.001) and numbers at risk significantly decreased from 15 to 18 months at time-to-event analyses. Fifth, the population enrolled in the SMART-DATE had relatively low ischemic risk profile (hypertension 48.9%, dyslipidaemia 24.3%, prior myocardial infarction 2.0%, prior revascularization 4.3%, left ventricular ejection fraction 55.5±10.8%) and, although target lesion was more frequently placed in the left anterior descending (58.9%), 45.2% had multivessel disease and only 20.1% received multivessel intervention. Such characteristics impose some restrictions in the translatability of findings to the real world subset, which can be more complex. Finally, although large-scale and multicentre, the trial was not international, therefore ethnical influences or regional therapeutic standards cannot be completely neglected.