Dual Antiplatelet Therapy with Clopidogrel and Aspirin Versus Aspirin Monotherapy in Patients Undergoing Coronary Artery Bypass Graft Surgery

Abstract

Background

The optimal antiplatelet therapy after coronary artery bypass grafting remains unclear. We evaluated the association of dual antiplatelet therapy (DAPT) with clopidogrel plus aspirin and clinical outcomes among patients undergoing coronary artery bypass grafting.

Methods and Results

A total of 18 069 consecutive patients who underwent primary isolated coronary artery bypass grafting between 2013 and 2017 were identified from a contemporary registry, and 10 854 (60.1%) received DAPT with clopidogrel plus aspirin as determined by claimed prescriptions after surgery. Cox regression models with inverse probability of treatment weighting were used to examine the associations between DAPT and outcomes. Patients who received DAPT, compared with those who received aspirin monotherapy, had a lower incidence of a composite of all‐cause death, myocardial infarction, stroke, or repeat revascularization at 6 months (2.9% versus 4.2%; inverse probability of treatment weighting–adjusted hazard ratio [HR], 0.65; 95% CI, 0.55–0.77; P<0.001) as well as death (HR, 0.61; 95% CI, 0.41–0.90), myocardial infarction (HR, 0.55; 95% CI, 0.40–0.74), and stroke (HR, 0.58; 95% CI, 0.46–0.74). The incidence of major bleeding did not differ significantly between the 2 groups (HR, 1.11; 95% CI, 0.69–1.78). Similar results were noted across multiple subgroups as well as when using different analytic methods.

Conclusions

Among patients undergoing coronary artery bypass grafting, DAPT with clopidogrel plus aspirin as secondary prevention was associated with reduced risk of major adverse cardiovascular and cerebrovascular events within 6 months as compared with aspirin monotherapy, and there was no significant increase in major bleeding.

Nonstandard Abbreviations and Acronyms

DAPT

dual antiplatelet therapy

IPTW

inverse probability of treatment weighting

MACCE

major adverse cardiac and cerebrovascular events

Clinical Perspective

What Is New?

• The present study is one of the largest analyses that primarily focused on the association of dual antiplatelet therapy and major outcomes in patients who underwent coronary artery bypass grafting.

• We found that post–coronary artery bypass grafting dual antiplatelet therapy with clopidogrel plus aspirin was associated with reduced risk of major adverse cardiovascular and cerebrovascular events within 6 months as compared with aspirin monotherapy, and there was no significant increase in major bleeding.

• The association of dual antiplatelet therapy with fewer clinical events was consistent across key clinical subgroups, including age, sex, clinical presentations, diabetes mellitus, and bypass techniques.

What Are the Clinical Implications?

• Dual antiplatelet therapy with clopidogrel plus aspirin could be a promising secondary prevention strategy that may improve the outcome of patients who had ​coronary artery bypass grafting, including those with acute coronary syndrome or stable angina, presence or absence of diabetes mellitus, or had on‐pump or off‐pump bypass.

Coronary artery bypass grafting (CABG) has been established as an effective treatment for patients with extensive coronary artery disease.¹ Aspirin is recommended as a fundamental secondary prevention medication for patients with CABG to maintain the benefits of revascularization and prevent major adverse cardiovascular events.² However, patients treated with CABG still have a notable risk of subsequent major ischemic cardiac and cerebrovascular events, which may exceed 10% in the first 6 to 12 months after the surgery.³ Reduced postoperative responsiveness to aspirin, platelet activation, and thrombosis results in systemic hypercoagulability and early graft failure. These have been identified as vital contributing factors in this context.⁴, ⁵, ⁶, ⁷

Dual antiplatelet therapy (DAPT) with aspirin plus a P2Y₁₂ receptor antagonist (eg, clopidogrel or ticagrelor) to enhance the antiplatelet effect⁸, ⁹ has been reported to slow down native coronary stenosis progression¹⁰ and increase graft patency¹¹, ¹², ¹³, ¹⁴, ¹⁵, ¹⁶ in patients who underwent CABG as well as prevent recurrent stroke in patients with ischemic cerebrovascular disease.¹⁷, ¹⁸ However, the question of whether the benefits associated with DAPT, especially the potential improvement in graft patency, translate into better clinical outcomes remains inadequately investigated with mixed results,¹⁶, ¹⁹, ²⁰, ²¹, ²², ²³ and DAPT may increase the risk of bleeding.⁹, ²³ Clinical practice guidelines recommend that DAPT may be considered for selected patients who are at high risk of recurrent ischemic events, including those who presented with acute coronary syndrome (ACS) or received coronary stent implantation within 1 year before CABG, or underwent off‐pump CABG.², ²⁴, ²⁵, ²⁶ However, they also acknowledge the need for more evidence in this area because these recommendations are merely based on expert consensus or underpowered secondary data (level C of evidence).², ²⁴, ²⁵, ²⁶ There is even less evidence available on the efficacy and safety of DAPT in specific patient populations, such as patients with stable angina who constitute the majority of CABG procedures and those who underwent on‐pump CABG.

Therefore, the present study examined the association between post‐CABG DAPT with clopidogrel plus aspirin and clinical outcomes in a large all‐comer clinical practice registry. We hypothesized that DAPT may be associated with reduced risk of major adverse cardiac and cerebrovascular events (MACCE) when compared with aspirin monotherapy.

Methods

Study Design and Data Source

The data, analytical methods, and study materials that support the findings of this study may be made available from the corresponding author on reasonable request. This retrospective cohort analysis used data from a contemporary registry of consecutive patients who underwent CABG at Fuwai Hospital (Beijing, China). Data on patient characteristics, procedures, and medications were extracted from the registry and supplemented with electronic medical records. All data were collected by experienced clinical researchers, and clinical definitions followed those of the Society of Thoracic Surgeons National Adult Cardiac Database (http://www.sts.org). The accuracy and completeness of these data were ensured through multiple procedures described previously.²⁷, ²⁸

Patients were followed up by routine outpatient visit or via telephone by trained cardiovascular research nurses as part of standard institutional procedures. At follow‐up, patients were asked to enumerate all of their current medications to the interviewer, including drug name, dose, and schedule.²⁷ If any adverse events were reported during the follow‐up process, patients were asked to provide related medical records for further confirmation. Details on aspirin and clopidogrel use were obtained by review of in‐hospital medication dispensing records, discharge summaries, and follow‐up. The institutional review board at Fuwai Hospital approved the use of clinical data for this study and waived the requirement of informed consent.

Study Population

All adult patients who underwent primary CABG between January 1, 2013 and December 31, 2017 were considered for the analysis. Patients were excluded if they had a concomitant cardiac surgery (eg, valve replacement or ventricular aneurysm resection), received simultaneous or staged hybrid coronary revascularization, were exposed to antiplatelet agents other than aspirin and clopidogrel (eg, ticagrelor or prasugrel) or treated with clopidogrel alone post‐CABG, required vitamin K antagonist therapy, or died before the initiation of any antiplatelet therapy. Patients with DAPT were defined as those who had at least 1 postoperative prescription for aspirin plus clopidogrel. The aspirin monotherapy cohort was defined as patients who received aspirin alone post‐CABG. Patients were considered exposed to the treatment (DAPT or aspirin monotherapy) through the end of follow‐up, analogous to an intention‐to‐treat design.

Clinical Management

Patients were managed in accordance with local practice guidelines, and all procedures were performed using standard bypass techniques (Data S1).²⁷, ²⁹ The choice between on‐pump and off‐pump CABG was at the discretion of the principal surgeon. Whenever possible, the internal thoracic artery was preferentially used for revascularization of the left anterior descending artery. The perioperative antiplatelet therapy was also left to the individual surgeon’s evaluation and decision, though clopidogrel should have been discontinued at least 5 days before surgery if clinically feasible. Local guidelines and regulatory authorities do not specify rules or restrictions for the selection of patients who receive post‐CABG DAPT administration. Routinely, aspirin was started within 24 hours (ideally within 6 hours) after CABG in a daily dose of 100 mg and recommended to continue indefinitely. For patients who received DAPT, 75 mg of clopidogrel was added to 100 mg of aspirin daily without a loading dose, preferably within 48 hours after CABG, but when clinical stability was ensured and chest tube output was <30 mL/hour for at least 2 hours.¹¹ The duration of DAPT was determined by the treating physician, with the treatment typically maintained for a minimum of 1 month. Additional secondary prevention therapies (eg, statins, β‐blockers, or renin‐angiotensin system blockade) were recommended for all patients, if indicated, following clinical guidelines.

Outcomes

The primary outcome was the first occurrence of MACCE, defined as a composite of all‐cause mortality, myocardial infarction, stroke, and repeat revascularization within 6 months after CABG. Secondary outcomes included individual components of the primary outcome and major bleeding, which was defined as a composite of in‐hospital reoperation because of bleeding and hospitalization for bleeding after discharge. All outcome measures were prespecified, rigorously verified, and adjudicated by independent clinicians. Detailed definitions of the outcomes are provided in Data S1.

Statistical Analysis

Detailed statistical methods are available in the Data S1. We used inverse probability of treatment weighting (IPTW) based on propensity scores to construct a weighted cohort of patients who differed with respect to postoperative antiplatelet therapy but were similar with respect to other measured characteristics.³⁰ A propensity score for the predicted probability of receiving DAPT in each patient was calculated from a nonparsimonious multivariable logistic regression model fitted with patient characteristics that may confound the relationship between antiplatelet therapy and clinical outcomes (ie, demographic characteristics, medical history, concurrent medication use, procedure‐related characteristics, and year of surgery; the full list of the 33 variables included in the propensity model is provided in the Data S1). The IPTW analysis was performed to estimate the average treatment effect, that is, the effect of treatment on the entire population eligible for isolated CABG.³⁰ Stabilized weights were used to reduce the variability in the IPTW models.³¹ Balance among covariates was assessed using standardized differences, and a difference of ≤10% was considered the ideal balance.³¹ To account for missing data (1.3% for preoperative hemoglobin and platelet count; <0.3% for height, weight, and preoperative creatinine), a single mean imputation stratified by study groups was used.

Time‐to‐event analyses for MACCE and all‐cause mortality were performed using weighted Cox proportional hazards models. Cardiovascular death, myocardial infarction, stroke, repeat revascularization, and major bleeding were analyzed in the weighted population, accounting for death (or noncardiovascular death) as a competing risk using the Fine and Gray method.³² Hazard ratios (HRs) and 95% CIs were estimated with a robust variance estimator to account for the weighted nature of the population. The proportional hazards assumption was confirmed by Schoenfeld residuals plots. Survival curves were constructed using the IPTW‐adjusted Kaplan‐Meier method and compared using the IPTW‐adjusted log‐rank test.³³ For each outcome analyzed, the follow‐up period began upon initiation of DAPT or aspirin monotherapy (time 0). Patients were censored on first occurrence of the event, death, loss to follow‐up, or reaching 180 days of follow‐up. Prespecified subgroup analyses were performed by refitting separate IPTW survival models for each subgroup and conducting tests for interaction.

We performed several sensitivity analyses to assess the robustness of our findings (Data S1). First, alternative analytic strategies (ie, propensity score matching, multivariable Cox regression, and doubly robust estimation combining the propensity score and outcome regression)³⁴, ³⁵ were used to compare outcomes between study groups. Second, the primary analyses were repeated after adjusting for other secondary prevention medications at discharge (ie, statin, β‐blocker, angiotensin‐converting enzyme inhibitor/angiotensin receptor blocker) or including principal surgeon as a random effect to confirm that the observed associations were not the result of differences in the use of postoperative medications or unobserved characteristics between surgeons, respectively. Third, analyses of study outcomes were performed after excluding those who were treated with P2Y₁₂ inhibitors within 5 days before CABG to ensure that treatment‐related outcome differences were not confounded by preoperative antiplatelet therapy.³⁶ Fourth, because a proportion of patients in the DAPT group started clopidogrel a few days after the initiation of aspirin rather than starting the 2 medications simultaneously, we repeated the primary analysis after exclusion of patients from the DAPT group who did not start clopidogrel and aspirin on the same day to minimize possible immortal time bias.³⁷, ³⁸ Finally, we calculated the E‐value to quantify the potential for unmeasured confounders to explain the effect of DAPT on estimated HRs.³⁹

All tests were 2‐tailed, with P values <0.05 indicating statistical significance. Data were analyzed using SAS 9.4 (SAS Institute, Cary, NC).

Results

Patient Characteristics

Between 2013 and 2017, there were 22 819 patients who met the study inclusion criteria. After the exclusion criteria were applied, 18 069 (79.2%) patients were included in the analysis; 10 854 (60.1%) received DAPT with clopidogrel plus aspirin after surgery, and 7215 (39.9%) received aspirin monotherapy (Figure S1). Table 1 summarizes selected baseline characteristics of study patients before propensity score weighting (a list of all characteristics is provided in Table S1). Patients who received DAPT, as compared with those who received aspirin monotherapy, were younger; had a higher prevalence of insulin‐dependent diabetes mellitus, peripheral artery disease, and previous percutaneous coronary intervention; and were more likely to be treated with intravenous nitrate or statins before surgery. Emergency surgery and use of cardiopulmonary bypass were more common in the DAPT group. The overall proportion of DAPT prescriptions increased over the 5‐year study period but varied substantially across surgeons (Figures S2 and S3). Details on the administration of antiplatelet regimens are provided in Table S2. For patients who received DAPT, 87.6% of them maintained the therapy at discharge and 54.5% at 6 months (calculated by dividing the total on‐treatment patients at the 2 time‐points by the corresponding surviving patients). Covariates were well balanced in the propensity‐weighted cohort, with all standardized differences <10% (Figure S4).

Table 1.

Selected Characteristics of Patients Before Inverse Probability Weighting

Characteristics	DAPT (N=10 854)	Aspirin Monotherapy (N=7215)	P Value
Age, y, mean (SD)	60.8 (8.6)	61.9 (8.5)	<0.01
Age ≥65 y, n (%)	3542 (32.6)	2651 (36.7)	<0.01
Female sex, n (%)	2441 (22.5)	1715 (23.8)	0.04
BMI*, kg/m2, mean (SD)	25.7 (3.0)	25.7 (3.0)	0.21
Medical history, n (%)
Smoking	6029 (55.5)	3944 (54.7)	0.24
Diabetes mellitus	4201 (38.7)	2782 (38.6)	0.84
Insulin‐treated diabetes mellitus	903 (8.3)	409 (5.7)	<0.01
Hypertension	7091 (65.3)	4700 (65.1)	0.79
Hyperlipidemia	7134 (65.7)	4553 (63.1)	<0.01
Peripheral artery disease	1051 (9.7)	454 (6.3)	<0.01
Previous myocardial infarction	2880 (26.5)	1852 (25.7)	0.20
Previous PCI	1485 (13.7)	708 (9.8)	<0.01
Previous CVA	1248 (11.5)	781 (10.8)	0.16
Clinical presentation, n (%)			0.06
Stable angina	4810 (44.3)	3051 (42.3)
Unstable angina	5341 (49.2)	3683 (51.0)
NSTEMI	310 (2.9)	215 (3.0)
STEMI	393 (3.6)	266 (3.7)
LVEF, n (%)			0.11
≥50%	10183 (93.8)	6713 (93.0)
40%–49%	528 (4.9)	379 (5.3)
30%– 39%	135 (1.2)	115 (1.6)
<30%	8 (0.1)	8 (0.1)
EuroSCORE†, n (%)			<0.01
0–2	7000 (64.5)	4453 (61.7)
3–5	3135 (28.9)	2194 (30.4)
≥6	719 (6.6)	568 (7.9)
Medication use before surgery, n (%)
Aspirin	4128 (38.0)	2868 (39.8)	0.02
Clopidogrel	3868 (35.6)	2511 (34.8)	0.25
Clopidogrel within 5 days	770 (7.1)	540 (7.5)	0.32
Intravenous nitrate	1533 (14.1)	817 (11.3)	<0.01
β‐blocker	9699 (89.4)	6346 (88.0)	0.01
Statin	9194 (84.7)	5622 (77.9)	<0.01
ACEI/ARB	4362 (40.2)	3022 (41.9)	0.02
Surgical procedure characteristics
Emergency surgery‡, n (%)	330 (3.0)	126 (1.7)	<0.01
On pump, no. (%)	5561 (51.2)	3272 (45.3)	<0.01
LIMA to LAD graft, n (%)	10067 (92.7)	6844 (94.9)	<0.01
No. of grafts, mean (SD)	3.3 (0.9)	3.4 (0.9)	<0.01
No. of arterial grafts	1.0 (0.4)	1.0 (0.2)	<0.01
No. of venous grafts	2.3 (0.9)	2.4 (0.9)	<0.01

ACEI indicates angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; BMI, body mass index; CVA, cerebrovascular accident; DAPT, dual antiplatelet therapy; EuroSCORE, European System for Cardiac Operative Risk Evaluation I; LAD, left anterior descending artery; LIMA, left internal mammary artery; LVEF, left ventricular ejection fraction; NSTEMI, non‐ST‐segment–elevation myocardial infarction; PCI, percutaneous coronary intervention; and STEMI, ST‐segment–elevation myocardial infarction.

^*Calculated as weight in kilograms divided by height in meters squared.

^† The European System for Cardiac Operative Risk Evaluation (EuroSCORE) is a risk model for predicting the risk of death after cardiac surgery; scores range from 0 to 100%, with higher scores indicating greater risk.

^‡ Operation before the beginning of the next working day after decision to operate.

Study Outcomes

Table 2 shows the incidences of the primary and secondary outcomes, and Figure 1 and Figure S5 show the weighted Kaplan–Meier curves of the study outcomes for the DAPT and aspirin monotherapy groups. The primary composite outcome occurred in 312 (2.9%) patients who received DAPT during the 6‐month follow‐up period as compared with 305 patients (4.2%) who received aspirin monotherapy (unadjusted HR, 0.67; 95% CI, 0.58–0.79; P<0.001). After adjustment using the IPTW approach, DAPT was associated with lower risks of MACCE (HR, 0.65; 95% CI, 0.55–0.77; P<0.001) and all‐cause mortality (0.6% versus 0.9%; HR, 0.61; 95% CI, 0.41–0.90; P=0.012) (Table 2). The DAPT group also had lower cumulative incidences of myocardial infarction (0.9% versus 1.3%; HR, 0.55; 95% CI, 0.40–0.74; P<0.001) and stroke (1.3% versus 2.3%; HR, 0.58; 95% CI, 0.46–0.74; P<0.001) in analyses accounting for the competing risk of death. The risk of major bleeding was similar for the DAPT and aspirin monotherapy groups (0.5% versus 0.4%), with an IPTW‐adjusted HR of 1.11 (95% CI, 0.69–1.78; P=0.67). In a post hoc analysis evaluating the net clinical benefit of MACCE offset by major bleeding, a composite of MACCE and major bleeding occurred in 355 (3.3%) patients with DAPT and 330 (4.6%) patients with aspirin monotherapy (HR, 0.68; 95% CI, 0.58–0.80; P<0.001). DAPT with clopidogrel would lead to 136 fewer MACCE per 10 000 patients at the expense of 8 additional major bleeding events, with a number needed to treat of 74 and a number needed to harm of 1380.

Table 2.

Association of DAPT Versus Aspirin Monotherapy With Outcomes Within 6 Months After Coronary Artery Bypass Graft Surgery

	DAPT (N=10 854)	Aspirin Monotherapy (N=7215)	Adjusted HR (95% CI)*	P Value
Primary outcome, n (%)
MACCE†	312 (2.9)	305 (4.2)	0.65 (0.55–0.77)	<0.001
Secondary outcomes, n (%)
All‐cause death	61 (0.6)	66 (0.9)	0.61 (0.41–0.90)	0.012
Cardiovascular death	44 (0.4)	49 (0.7)	0.57 (0.36–0.90)	0.015
MI	95 (0.9)	96 (1.3)	0.55 (0.40–0.74)	<0.001
Stroke‡	142 (1.3)	165 (2.3)	0.58 (0.46–0.74)	<0.001
Repeat revascularization	34 (0.3)	21 (0.3)	1.06 (0.61–1.86)	0.83
Cardiovascular death, MI, or ischemic stroke	263 (2.4)	282 (3.9)	0.59 (0.49–0.70)	<0.001
Cardiovascular death or MI	132 (1.2)	124 (1.7)	0.62 (0.48–0.81)	<0.001
Major bleeding	53 (0.5)	30 (0.4)	1.11 (0.69–1.78)	0.67
In‐hospital reoperation for bleeding	16 (0.1)	9 (0.1)	1.27 (0.55–2.94)	0.57
Hospitalization for bleeding	37 (0.3)	21 (0.3)	1.17 (0.66–2.08)	0.59
Net clinical benefit outcome, n (%)
MACCE, major bleeding	355 (3.3)	330 (4.6)	0.68 (0.58–0.80)	<0.001

DAPT indicates dual antiplatelet therapy; HR, hazard ratio; MACCE, major adverse cardiac and cerebrovascular events; and MI, myocardial infarction.

^*Estimated using inverse probability of treatment‐weighted Cox regression or Fine and Gray model.

^† A composite of all‐cause mortality, MI, stroke, and repeat revascularization.

^‡ Seven patients in the DAPT group and 2 patients in the aspirin‐alone group had hemorrhagic stroke.

Figure 1. Inverse probability of treatment weighting‐adjusted Kaplan‐Meier analysis for the primary outcome.

The at‐risk table shows the actual number of patients at risk. DAPT indicates dual antiplatelet therapy; and IPTW, inverse probability of treatment weighting.

Subgroup Analyses

The association of DAPT and primary outcome was consistent across predefined clinical subgroups defined by age (older or younger than 65 years), sex, clinical presentation (acute coronary syndrome [ACS] or stable angina), presence or absence of diabetes mellitus, presence or absence of hyperlipidemia, surgery risk (European System for Cardiac Operative Risk Evaluation I [EuroSCORE] ≥3 or ≤2), off‐pump or on‐pump bypass, and number of venous grafts (≥3 or ≤2) as well as in post hoc subsets defined by year of surgery (2013 to 2015, 2016 to 2017) (Figure 2). The difference between DAPT and aspirin monotherapy was less pronounced in the subgroup of patients who presented with ACS (HR, 0.75; 95% CI, 0.61–0.94) than it was in the subgroup of patients with stable angina (HR, 0.53; 95% CI, 0.40–0.69; P=0.04 for interaction), although DAPT was favored in both (Table 3 and Figure 2). Similarly, the association of DAPT and patient outcome tended to attenuate among patients with diabetes mellitus as opposed to among those without diabetes mellitus (Table 3, P=0.23 for interaction). No heterogeneity was noted among other subgroups (Figure 2).

Figure 2. Subgroup analyses for the primary outcome.

Separate propensity score models were fitted to predict the probability of DAPT for each subgroup, and hazard ratios were estimated using inverse probability of treatment‐weighted Cox proportional hazards models. DAPT indicates dual antiplatelet therapy; EuroSCORE, European System for Cardiac Operative Risk Evaluation I; and HR, hazard ratio.

Table 3.

Association of DAPT vs Aspirin Monotherapy With Outcomes in Selected Subsets of Patients

	DAPT (N=10 854)	Aspirin Monotherapy (N=7215)	Adjusted HR (95% CI)*	P Value
Clinical presentation, n (%)
Acute coronary syndrome	N=6044	N=4164
MACCE†	195 (3.2)	177 (4.3)	0.75 (0.61–0.94)	0.010
All‐cause death	42 (0.7)	39 (0.9)	0.85 (0.53–1.39)	0.52
Stroke	81 (1.3)	88 (2.1)	0.64 (0.47–0.89)	0.007
MACCE, major bleeding	219 (3.6)	195 (4.7)	0.77 (0.62–0.94)	0.011
Stable angina	N=4810	N=3051
MACCE†	117 (2.4)	128 (4.2)	0.53 (0.40–0.69)	<0.001
All‐cause death	19 (0.4)	27 (0.9)	0.42 (0.22–0.81)	0.010
Stroke	61 (1.3)	77 (2.5)	0.48 (0.34–0.69)	<0.001
MACCE, major bleeding	136 (2.8)	135 (4.4)	0.58 (0.45–0.75)	<0.001
Diabetes mellitus status, n (%)
Diabetes mellitus	N=4201	N=2782
MACCE†	139 (3.3)	120 (4.3)	0.76 (0.59–0.99)	0.040
All‐cause death	27 (0.6)	26 (0.9)	0.68 (0.37–1.24)	0.21
Stroke	79 (1.9)	77 (2.8)	0.71 (0.51–0.99)	0.044
MACCE, major bleeding	152 (3.6)	133 (4.8)	0.75 (0.59–0.96)	0.024
No diabetes mellitus	N=6653	N=4433
MACCE†	173 (2.6)	185 (4.2)	0.59 (0.47–0.73)	<0.001
All‐cause death	34 (0.5)	40 (0.9)	0.54 (0.32–0.90)	0.018
Stroke	63 (0.9)	88 (2.0)	0.45 (0.32–0.63)	<0.001
MACCE, major bleeding	203 (3.1)	197 (4.4)	0.65 (0.53–0.80)	<0.001
Cardiopulmonary bypass, n (%)
Off‐pump	N=5293	N=3943
MACCE†	141 (2.7)	158 (4.0)	0.67 (0.52–0.85)	0.001
All‐cause death	26 (0.5)	30 (0.8)	0.72 (0.39–1.32)	0.28
Stroke	58 (1.1)	90 (2.3)	0.51 (0.36–0.73)	<0.001
MACCE, major bleeding	165 (3.1)	168 (4.3)	0.74 (0.59–0.93)	0.010
On‐pump	N=5561	N=3272
MACCE†	171 (3.1)	147 (4.5)	0.66 (0.52–0.84)	<0.001
All‐cause death	35 (0.6)	36 (1.1)	0.52 (0.31–0.87)	0.014
Stroke	84 (1.5)	75 (2.3)	0.64 (0.46–0.88)	0.007
MACCE, major bleeding	190 (3.4)	162 (5.0)	0.66 (0.53–0.83)	<0.001

DAPT indicates dual antiplatelet therapy; HR, hazard ratio; and MACCE, major adverse cardiac and cerebrovascular events.

^*Separate propensity‐score models were fitted to predict the probability of DAPT for each subgroup, and HRs were estimated with the use of inverse probability of treatment‐weighted Cox regression or Fine and Gray model.

^† A composite of all‐cause mortality, myocardial infarction, stroke, and repeat revascularization.

Sensitivity Analyses

Propensity score matching created a well‐balanced cohort of 6635 patient pairs (Table S3). Findings from the propensity score–matched analyses were consistent with the primary IPTW‐adjusted analyses, demonstrating lower risk of MACCE with DAPT (HR, 0.68; 95% CI, 0.57–0.82; P<0.001) and comparable risk of major bleeding (HR, 1.14; 95% CI, 0.69–1.90; P=0.60) (Table S4). Both multivariable Cox regression and double robust estimation yielded nearly identical results with the original IPTW analysis (Tables S5).

In the primary propensity score–weighted cohort, the risk of MACCE remained lower in the DAPT group compared with the aspirin monotherapy group after further adjustment for other secondary prevention medications (HR, 0.67; 95% CI, 0.57–0.80; P<0.001), incorporation of principal surgeon as a random effect (HR, 0.65; 95% CI, 0.54–0.78; P<0.001), and exclusion of 1368 patients who were treated with P2Y₁₂ inhibitors within 5 days before CABG (HR, 0.67; 95% CI, 0.56–0.79; P<0.001) (Table S6). The exclusion of patients who started clopidogrel and aspirin on different days after CABG in the DAPT group (N=5484; analysis conducted to account for potential immortal time bias) did not alter the findings substantially (Table 4). The E‐value corresponding to the lower bound was 1.92 for MACCE (E‐value for point estimate, 2.45) and 1.00 for major bleeding (E‐value for point estimate, 1.46).

Table 4.

Association of DAPT vs Aspirin Monotherapy With Outcomes After Exclusion of Patients in the DAPT Group Who Started Clopidogrel and Aspirin on Different Days

	DAPT (N=5370)	Aspirin Monotherapy (N=7215)	Adjusted HR (95% CI)*	P Value
Primary outcome, n (%)
MACCE†	175 (3.3)	305 (4.2)	0.75 (0.61–0.92)	0.005
Secondary outcomes, n (%)
All‐cause death	21 (0.4)	66 (0.9)	0.48 (0.27–0.84)	0.010
Cardiovascular death	15 (0.3)	49 (0.7)	0.47 (0.24–0.90)	0.024
MI	70 (1.3)	96 (1.3)	0.88 (0.63–1.23)	0.46
Stroke	79 (1.5)	165 (2.3)	0.65 (0.48–0.86)	0.003
Repeat revascularization	19 (0.4)	21 (0.3)	1.01 (0.52–1.96)	0.97
Cardiovascular death, MI, or ischemic stroke	154 (2.9)	282 (3.9)	0.72 (0.58–0.89)	0.003
Cardiovascular death or MI	81 (1.5)	124 (1.7)	0.85 (0.63–1.15)	0.29
Major bleeding	31 (0.6)	30 (0.4)	1.47 (0.87–2.50)	0.15
In‐hospital reoperation for bleeding	11 (0.2)	9 (0.1)	1.59 (0.64–3.97)	0.32
Hospitalization for bleeding	20 (0.4)	21 (0.3)	1.46 (0.77–2.77)	0.25
Net clinical benefit outcome, n (%)
MACCE, major bleeding	201 (3.7)	330 (4.6)	0.81 (0.67–0.98)	0.028

DAPT indicates dual antiplatelet therapy; HR, hazard ratio; MACCE, major adverse cardiac and cerebrovascular events; and MI, myocardial infarction.

^*Estimated using inverse probability of treatment‐weighted Cox regression or Fine and Gray model.

^† A composite of all‐cause mortality, MI, stroke, and repeat revascularization.

Discussion

In this large cohort study comparing outcomes of DAPT with aspirin monotherapy in patients undergoing isolated CABG, the use of DAPT was associated with significantly lower risk of a composite outcome of death, myocardial infarction, stroke, or repeat revascularization as well as 3 individual components of the outcome—death, myocardial infarction, and stroke—at 6 months after the surgery. The association was apparent across clinically important subgroups, including patients with preoperative ACS or stable angina, those with or without diabetes mellitus, and those who underwent on‐pump or off‐pump bypass. There was no evidence of a higher risk of major bleeding among patients who received DAPT.

Inhibition of platelet activation and aggregation is crucial for maintaining graft patency and improving outcomes for patients with CABG, especially considering the continued popularity of saphenous vein grafts as the secondary bypass conduit supplement to the left internal mammary artery.¹, ⁶ DAPT with a P2Y₁₂ inhibitor added to aspirin is an appealing secondary prevention strategy for patients who had CABG because its potent synergistic antithrombotic effects might help to overcome resistance to aspirin and prevent thrombosis in grafts, native coronary arteries, and even cerebrovascular arteries.⁴, ⁸, ¹⁰, ¹⁸ A recent randomized controlled trial that compared the effect of ticagrelor plus aspirin versus aspirin alone on saphenous vein graft patency¹⁴ and 3 comprehensive meta‐analyses incorporating data from both randomized controlled trials and observational studies¹⁵, ¹⁶, ²³ found that adding a P2Y₁₂ inhibitor (clopidogrel or ticagrelor) to aspirin after CABG reduced the risk of saphenous vein graft failure. However, this benefit was not confirmed in another newly published randomized study on the same topic.⁴⁰

Limited data are available to determine whether DAPT improves patient outcomes.¹⁹, ²⁰, ²² In a meta‐analysis of 20 315 patients from 11 randomized controlled trials and 11 observational studies,²³ 7481 (37%) patients received postoperative DAPT (97% with clopidogrel), and DAPT was associated with lower cardiovascular mortality in the pooled observational sample (odds ratio, 0.67; 95% CI, 0.49–0.93). However, a subanalysis limited to patients from randomized controlled trials failed to demonstrate a reduction in mortality (odds ratio, 0.84; 95% CI, 0.56–1.26). Moreover, DAPT was associated with a 31% higher risk of bleeding in the overall study population. It is noteworthy that almost all of the studies involved were designed to evaluate the efficacy of DAPT on graft patency rather than its effects on clinical outcomes, were underpowered,¹⁵, ¹⁶, ²³ had substantially heterogeneous populations,¹⁴, ⁴⁰ and used different DAPT definitions, thereby precluding conclusive inference and resulting in inconsistent guideline recommendations², ²⁴, ²⁵ and variation in real‐world clinical practice patterns.⁴¹

In the present study, 60% of patients were prescribed DAPT after CABG, which is considerably higher than the proportions reported in previous studies (20%–50%).²³ This finding may reflect differences in institutional practice experiences¹¹ or the removal of financial barriers to clopidogrel because the drug was covered for patients in our study by local healthcare insurance. We noted substantial variation in DAPT prescription rates (26% to 95%) across the 74 surgeons involved in our study, which is consistent with a survey among Canadian cardiac surgeons.⁴¹ Only half of the patients in our DAPT group started therapy on postoperative day 0 or 1, and maintenance to DAPT rapidly declined to 54.5% by 6 months, similar to previous studies.²⁰, ²¹, ²², ⁴⁰, ⁴² These findings suggest potential hesitation by surgeons regarding routine use of DAPT in patients who had CABG.

To our knowledge, the present study is the largest to date to principally investigate the association between DAPT and clinical outcomes in a contemporary CABG population. Our findings suggest that post‐CABG DAPT with clopidogrel is an effective and safe secondary prevention regimen that can improve patient outcomes. We found a 39% reduction in mortality among patients treated with DAPT, an effect similar to 2 thorough meta‐analyses.²³, ⁴³ This improvement in survival could be the cumulative result of better graft patency, as previously demonstrated¹⁰, ¹⁵, ¹⁶, ²³ and supported by the lower incidences of myocardial infarction and cardiovascular death in the current analysis, and potential pleiotropic benefits of clopidogrel.⁴⁴ Additionally, for the first time, we observed a 50% lower risk of stroke at 6 months after CABG for patients taking DAPT with clopidogrel versus aspirin monotherapy. This finding is consistent with a trial of 5170 patients with recent transient ischemic attack or minor stroke that showed a combination of clopidogrel and aspirin was superior to aspirin alone in reducing the risk of stroke.¹⁷ Collectively, the findings indicate that post‐CABG DAPT contributed to a significantly lower risk of MACCE when compared with aspirin monotherapy.

In this study, the addition of clopidogrel to aspirin was associated with a risk of major bleeding that was comparable to aspirin monotherapy. This result was not unexpected given the conservative daily dose of DAPT (clopidogrel 75 mg plus aspirin 100 mg), relatively young and healthy patient population, and strict definitions of major bleeding. We developed a composite major bleeding outcome of roughly similar severity to the composite ischemic outcome to aid clinicians in evaluating the risk/benefit tradeoff of antiplatelet therapy, but omitting other minor bleeding events. Notably, previous studies indicated that DAPT was more likely to introduce minor bleeding rather than major bleeding requiring surgical intervention or intense hospitalization care,⁷, ¹⁴, ⁴⁰, ⁴⁵ and a lower‐than‐expected incidence of major bleeding under current definitions may have biased the result toward the null. In addition, the low number of major bleeding events precluded evaluation of the impact of DAPT in patients at higher risk for bleeding, such as those of older age, with diabetes mellitus, or with renal dysfunction.²⁵, ²⁶

DAPT may offer specific benefits in subsets of the CABG patient population who are in a prothrombotic state or have residual cardiovascular risk, such as those presenting with ACS or diabetes mellitus and those who underwent off‐pump bypass surgery. Current guidelines recommend post‐CABG DAPT in patients with ACS²⁵, ²⁶ based on data from secondary analysis of DAPT trials in the setting of non‐ST‐ or ST‐segment–elevation myocardial infarction, in which only a small proportion (≈10%) of participants underwent CABG at 20 to 100 days after the initiation of DAPT.¹⁹, ²³ The present study adds credibility to the recommendation in a large contemporary CABG cohort. However, unlike for ACS, scarce data exist on the effect of DAPT after surgical revascularization for stable angina, and it remains unclear whether the presenting symptoms result in clinical differences relevant to secondary antiplatelet therapy.¹¹, ¹³, ²³, ⁴⁶ We observed a greater benefit of DAPT in patients with stable angina, possibly because of these patients having fewer factors linked to low responsiveness to clopidogrel.⁴⁷ Our results also support the strategy of intensifying platelet inhibition with DAPT after off‐pump CABG and expanding the use of DAPT to those who underwent on‐pump bypass surgery. Given the continued debate on off‐pump versus on‐pump bypass grafting as well as the aforementioned response variability to clopidogrel,¹, ³, ⁴⁷ the underlying mechanisms explaining the variation in DAPT‐related benefits warrant further exploration.⁴⁴

In our study, most of the events of interest occurred within 30 days after CABG, which is similar to previous analyses¹⁵ and indicates that combined antiplatelet therapy may be most beneficial in the early post‐CABG phase when thrombosis plays a predominant role in graft failure.⁶ Use of DAPT for 1 to 3 or 6 months followed by conversion to monotherapy with P2Y₁₂ inhibitors or aspirin, which have been proven safe in patients with ACS and coronary stent implantation,⁴⁸ might also be a clinically feasible strategy to avoid potential excess bleeding in the CABG population. However, no specific study has examined the optimal treatment duration of DAPT after CABG, although an arbitrary 1‐year treatment has been suggested for ACS and off‐pump patients.², ²⁶ The optimal dose of DAPT and the appropriate postoperative initiation time require further investigation.

Several limitations of our analysis should be considered. First, as an observational study, the analyses are subject to selection bias, and residual unmeasured confounding may persist despite adjustment for a variety of known patient variables using propensity scores to approximate randomization. Second, our study was based on the experience of a single high‐volume center. The conclusions may be influenced by patient referral patterns and local medical management, and therefore may not generalize to the larger CABG population. Third, detailed data on discontinuation time and on‐treatment duration were unavailable for the present analyses, which were based on the intention‐to‐treat principle, and we were unable to conduct time‐dependent Cox regression analyses to account for the impact of the variation in DAPT exposure over time on the association of interest. Instead, we delineated the proportion of “on‐treatment” participants in the 2 groups at discharge and at 6 months. We observed a high rate of DAPT discontinuation, but the reasons were not documented in patient records. Nevertheless, noncompliance should favor a type II error, which cannot explain the findings of the present study. Fourth, the 6‐month event rates in our analysis were lower than those in previous studies.³, ²¹ One potential explanation is that our study population tended to be younger and had fewer coexisting conditions. However, we cannot exclude the possibility of underreporting, although this is unlikely because a standardized case report form was used for the adjudication of all events and careful study oversight of outcomes; we anticipate that any problems would have impacted both groups equally. Fifth, low clopidogrel responsiveness has been reported in up to 30% of patients,²⁵, ⁴⁷ but we were unable to consider this information in our study because of lack of data on the platelet function test and CYP2C19 gene polymorphisms. Finally, novel P2Y₁₂ inhibitors (eg, ticagrelor) were not widely prescribed for patients with CABG in our center during the study period; therefore, the results of this study cannot be translated to patients given these agents.

Conclusions

Among patients undergoing primary isolated CABG, the use of DAPT with clopidogrel plus aspirin compared with aspirin monotherapy was associated with significant reduction in the risks of MACCE, mortality, myocardial infarction, and stroke, without a significant increase in major bleeding. These findings suggest that DAPT with clopidogrel could be a promising secondary prevention strategy for CABG to improve patient outcomes. Future studies are needed to provide an optimal and personalized post‐CABG DAPT strategy.

Sources of Funding

This study was supported by grants from the Beijing Municipal Science and Technology Commission (D171100002917001). The funder had no role in study design, data collection, data analysis, data interpretation, or writing of the manuscript.

Disclosures

None.

Supporting information

Data S1

Table S1–S6

Figure S1–S5

(J Am Heart Assoc. 2021;10:e020413. DOI: 10.1161/JAHA.120.020413.)