Concomitant Use of Single Antiplatelet Therapy With Edoxaban or Warfarin in Patients With Atrial Fibrillation: Analysis From the ENGAGE AF-TIMI48 Trial.

BACKGROUND: We studied the concomitant use of single antiplatelet therapy (SAPT) on the efficacy and safety of the anti-Xa agent edoxaban in patients with atrial fibrillation (AF). METHODS AND
RESULTS: ENGAGE AF-TIMI 48 was a randomized trial that compared 2 dose regimens of edoxaban with warfarin. We studied both the approved high-dose edoxaban regimen (HDER; 60 mg daily reduced by one half in patients with anticipated increased drug exposure), as well as a lower-dose edoxaban regimen (LDER; 30 mg daily, also reduced by one half in patients with anticipated increased drug regimen). SAPT (aspirin in 92.5%) was administered at the discretion of the treating physician. Cox proportional hazard regressions stratified by SAPT at 3 months with treatment as a covariate were performed. The 4912 patients who received SAPT were more frequently male, with histories of coronary artery disease and diabetes, and had higher CHADS2Vasc and HAS BLED scores than did the 14 977 patients not receiving SAPT. When compared to patients not receiving SAPT, those receiving SAPT had a higher incidence of major bleeding; (adjusted hazard ratio [HRadj]=1.46; 95% CI, 1.27-1.67, P<0.001). SAPT did not alter the relative efficacy of edoxaban compared to warfarin in preventing stroke or systemic embolic events (SEEs): edoxaban versus warfarin without SAPT, hazard ratio (HRadj for HDER)=0.94; (95% CI: 0.77-1.15) with SAPT, HRadj=0.70 (95% CI: 0.50-0.98), P interaction (Pint)=0.14. (HRadj for LDER versus warfarin without SAPT=1.19 (95% CI 0.99-1.43) With SAPT, 1.03 (95% CI, 0.76-1.39) Pint=0.42. Major bleeding was lower with edoxaban than warfarin both without SAPT, HRadj for HDER=0.80 (95% CI, 0.68-0.95), and with SAPT, HRadj=0.82 (95% CI, 0.65-1.03; Pint=0.91). For LDER without SAPT (HRadj=0.56 [95% CI 0.46-0.67]) and with SAPT (HRadj=0.51 [95% CI 0.39-0.66]).
CONCLUSIONS: Patients with AF who were selected by their physicians to receive SAPT in addition to an anticoagulant had a similar risk of stroke/SEE and higher rates of bleeding than those not receiving SAPT. Edoxaban exhibited similar relative efficacy and reduced bleeding compared to warfarin, with or without concomitant SAPT. CLINICAL TRIAL REGISTRATION: URL: http://www.clinicaltrials.gov/. Unique identifier: NCT00781391.

RCT Entities:   Population Interventions Outcomes

Introduction

As previously reported, in patients with nonvalvular atrial fibrillation (AF), the Effective aNticoaGulation with factor xA next GEneration in AF‐Thrombolysis In Myocardial Infarction 48 (ENGAGE AF‐TIMI 48) trial showed that the factor Xa inhibitor, edoxaban, was noninferior to warfarin in the prevention of stroke or systemic embolic event (stroke/SEE) and resulted in significantly lower rates of bleeding and cardiovascular death.1 Patients with nonvalvular AF are frequently elderly and have a high prevalence of chronic coronary artery disease (CAD).2, 3 Though oral anticoagulants are more effective than antiplatelet agents in preventing stroke/SEE in patients with AF, it is thought that the latter may be more protective in reducing vascular events in patients with CAD or at high risk of acute coronary events.4, 5 The choice of optimal antithrombotic management to prevent both thromboembolic and acute ischemic events in patients with AF and coexisting CAD is challenging given that combination therapy of anticoagulant and ‐antiplatelet agents is associated with an increased risk of bleeding and its efficacy is not clear.6, 7, 8, 9, 10 Here, we report on the effects of single antiplatelet therapy (SAPT) on the comparison of edoxaban with warfarin in patients with non‐valvular AF.

Methods and Results

Study Population and Treatments

ENGAGE AF‐TIMI 48 was a multinational, double‐blind randomized trial that compared the efficacy and safety of 2 dosing regimens of edoxaban with warfarin.1, 11 The trial was approved by all institutional review committees and subjects provided informed consent. Edoxaban was provided by the sponsor, Daiichi Sankyo (Parsippany, NY), who also funded the trial. Briefly, 21 105 patients with a history of documented AF and a CHADS2 score ≥2 were enrolled. Key exclusion criteria were severe renal dysfunction (creatinine clearance [CrCl] <30 mL/min), a high bleeding risk, receiving or anticipated to receive dual antiplatelet therapy, or a history of stroke, acute coronary syndrome, or coronary revascularization within 30 days of randomization. The trial studied 2 dose regimens of edoxaban. The higher dose edoxaban regimen (HDER) was approved by the US Food and Drug Administration and the European as well as Japanese Medicine Agencies. This dose (60 mg/day) was reduced to 30 mg/day if any of the following characteristics, which would be expected to increase drug exposure, were present at the time of randomization or occurred during the trial12: CrCl 30 to 50 mL/min; body weight ≤60 kg; or concomitant use of potent P‐glycoprotein inhibitors (verapamil, quinidine, or dronedarone). The lower‐dose regimen (LDER) was 30 mg/day and reduced to 15 mg/day for the same reasons. SAPT was administered as directed by the treating physician; aspirin ≤100 mg daily was strongly encouraged. If a clinical indication for dual antiplatelet therapy arose after randomization, the study drug was temporarily interrupted, but open‐label vitamin K antagonist (VKA) was permitted. Warfarin was well managed during the trial with median time in the therapeutic range (mTTR) of 68.4%.1

Patients with events (death, stroke, systemic embolic event [SEE], or major bleeding) occurring before the 3 months visit were excluded from the primary analysis because a sizeable percentage of patients (n=498; 7.46%) discontinued SAPT after they entered the ENGAGE AF‐TIMI 48 trial and were begun on anticoagulant therapy. Therefore, in our primary analysis we compared SAPT with no SAPT beginning 3 months after randomization. Patients with or without SAPT use at randomization were evaluated in a sensitivity analysis.

Endpoints

Endpoints were the same as those prespecified in the ENGAGE‐TIMI 48 trial.1, 11, 13 The primary efficacy endpoint was stroke/SEE and the primary safety endpoint was major bleeding as per the International Society on Thrombosis and Hemostasis (ISTH) criteria; the primary net clinical outcome was a composite of stroke/SEE, all‐cause death, or major bleeding. Cardiovascular death, myocardial infarction (MI), intracranial hemorrhage (ICH), life ‐threatenig bleeding, and major plus clinically relevant nonmajor bleeding were also analyzed.1 All end points were adjudicated by a blinded clinical endpoint committee.

Statistical Methods

Baseline characteristics across subgroups were compared using the chi‐square test for categorical variables and the Wilcoxon rank‐sum test for continuous variables. The Cox proportional hazard method was used to calculate the hazard ratio (HR) of edoxaban to warfarin. Comparison of outcomes in patients receiving and not receiving SAPT at 3 months was performed going forward, after adjustment for the following baseline characteristics: age, sex, geographic region; weight; CrCl; smoking; CAD; previous MI; previous coronary revascularization; dyslipidemia; diabetes; peripheral arterial disease; history of carotid arterial disease; type of AF; VKA naïve. Nonlinearity in continuous covariates was handled by cubic splines. Statistical analyses were performed in SAS software (version 9.2; SAS Institute, Inc., Cary, NC). All outcomes reported were annualized.

Results

At enrollment, 7036 subjects were randomized to warfarin, 7035 to the HDER and 7034 to the LDER, respectively (Figure 1). At randomization, 6678 of the 21 105 subjects (31.6%) were receiving SAPT and 7.46% of all subjects discontinued SAPT before the 3‐month visit in both the warfarin and edoxaban groups. The present analysis was carried out in 19 909 subjects, 4912 of whom (24.7%) were and 14 977 (75.3%) who were not receiving a SAPT at the 3‐month visit. After the 3 month visit, the percentage of patients receiving SAPT remained between 24% and 25% during the remainder of the trial, and the present analysis is based on these 4912 patients (Table 1). Of the 4912 patients on SAPT, 4525 (92.5%) were taking aspirin, (most [92%] of whom were taking ≤100 mg/day aspirin). The percentage of patients receiving aspirin at each time point were also similar (Table S1). The remainder received another antiplatelet agent, usually clopidogrel.

Figure 1

Study consort diagram. APT indicates antiplatelet therapy; HDER, high‐dose edoxaban registry; LDER, low dose edoxaban regimen; SEE, systemic embolic event.

Table 1

Prevalence of SAPT Use at Randomization and 6 Time Points Throughout the Study

Landmark Period	Total, n/N (%)	Warfarin, n/N (%)	Low‐Dose Edoxaban, n/N (%)	High‐Dose Edoxaban, n/N (%)
At baseline	6678/21 105 (31.6)	2253/7036 (32.0)	2179/7034 (31.0)	2246/7035 (31.9)
At 3 months	4912/19 909 (24.7)	1645/6643 (24.8)	1625/6671 (24.4)	1642/6595 (24.9)
At 6 months	4618/19 276 (24.0)	1551/6425 (24.1)	1527/6481 (23.6)	1540/6370 (24.2)
At 12 months	4567/18 794 (24.3)	1541/6250 (24.7)	1524/6301 (24.2)	1502/6243 (24.1)
At 18 months	4574/18 470 (24.8)	1518/6147 (24.7)	1525/6201 (24.6)	1531/6122 (25.0)
At 24 months	4447/18 095 (24.6)	1460/6002 (24.3)	1492/6090 (24.5)	1495/6003 (24.9)
At 30 months	3279/13 225 (24.8)	1089/4392 (24.8)	1127/4439 (25.4)	1063/4394 (24.2)

SAPT indicates single antiplatelet therapy.

Patients receiving SAPT were more frequently male, smokers, and more likely to have a history of CAD, previous MI, previous coronary revascularization, dyslipidemia, diabetes, and carotid and peripheral arterial disease (Table 2), previous coronary revascularization, paroxysmal AF, a CHADS2 score ≥4, a CHA2DS2‐Vasc score ≥4, a HAS‐BLED score ≥3, and to have been VKA naïve at the time of randomization. During the trial, the mTTR was lower in patients randomized to warfarin who were on SAPT (67.7%), compared to those who were not on SAPT (69.0%; P=0.002).

Table 2

Baseline Characteristics in Patients With and Without Antiplatelet Therapy at 3 Months (Including Low Edoxaban Group)

Variables	Not on SAPT (N=14 997)	On SAPT (N=4912)	P Value
Demographic
Age, y, median (IQR)	72.0 (64.0–77.0)	72.0 (64.0–78.0)	0.515
Age ≥75 y, n (%)	5907 (39.4)	1973 (40.2)	0.333
Male, n (%)	9039 (60.3)	3346 (68.1)	0.000
Region (%)
North America	2723 (18.2)	1662 (33.8)	0.000
Latin America	1996 (13.3)	496 (10.1)
Western Europe	2485 (16.6)	508 (10.3)
Eastern Europe	5634 (37.6)	1227 (25.0)
Asia	2159 (14.4)	1019 (20.7)
Clinical factors and medical history
Weight ≤60 kg, n (%)	1408 (9.4)	483 (9.8)	0.356
CrCl at randomization
Median (IQR), mL/min	71.1 (54.5–92.5)	70.0 (53.4–91.9)	0.013
≤50 mL/min, n (%)	2775 (18.5)	966 (19.7)	0.070
Current/former smoker, n (%)	5866 (39.1)	2287 (46.6)	0.000
Previous CAD, n (%)	4172 (27.8)	2403 (48.9)	0.000
Previous MI, n (%)	1395 (9.3)	869 (17.7)	0.000
Previous coronary revascularization, n (%)	1177 (7.8)	1274 (25.9)	0.000
Hypertension, n (%)	14 040 (93.6)	4606 (93.8)	0.705
Dyslipidemia, n (%)	7520 (50.1)	2979 (60.6)	0.000
Diabetes, n (%)	5190 (34.6)	1993 (40.6)	0.000
History of congestive heart failure, n (%)	8669 (57.8)	2762 (56.2)	0.053
Peripheral arterial disease, n (%)	511 (3.4)	278 (5.7)	0.000
Carotid arterial disease, n (%)	744 (5.0)	454 (9.2)	0.000
Previous stroke or TIA, n (%)	4216 (28.1)	1387 (28.2)	0.866
Type of AF
Paroxysmal, n (%)	3530 (23.5)	1510 (30.8)	0.000
Persistent, n (%)	3376 (22.5)	1211 (24.7)
Permanent, n (%)	8089 (53.9)	2189 (44.6)
CHADS2 score ≥4, n (%)	3243 (21.6)	1190 (24.2)	0.000
CHA2DS2‐Vasc score ≥4, n (%)	10 301 (68.7)	3694 (75.2)	0.000
HAS‐BLED score ≥3, n (%)	5253 (35.0)	3895 (79.3)	0.000
Medication
VKA naïve, n (%)	5633 (37.6)	2459 (50.1)	0.000
Dose reduced at randomization, n (%)	3669 (24.5)	1265 (25.8)	0.070

AF indicates atrial fibrillation; CAD, coronary artery disease; CrCl, creatinine clearance; IQR, interquartile range; MI, myocardial infarction; TIA, transient ischemic attack; VKA, vitamin K antagonist.

Outcomes in Patients With and Without SAPT

When the 3 arms (warfarin and the 2 edoxaban arms) were considered together, event rates for the primary efficacy endpoint (stroke/SEE) for those who received SAPT (1.71%/year) were similar to those in the non‐SAPT group (1.56%/year; P=0.26). Adjusted HR (HRadj; SAPT vs no SAPT) was 1.12 (95% CI, 0.95–1.32; P=0.19). Major bleeding occurred more frequently in patients who received SAPT (3.37%/year) than not (1.99%/year; P<0.001); HRadj=1.46 (95% CI, 1.27–1.67; P<0.001; Figure 2).

Figure 2

Outcomes in patients with and without antiplatelet therapy. Adj HR indicates adjusted hazard ratio; HR, hazard ratio; SAPT, single antiplatelet therapy; SEE, systemic embolic event.

In patients randomized to warfarin, those receiving SAPT had a numerically higher event rate for the primary endpoint (1.88%/year) compared to those in the non SAPT group (1.49%/year; P=0.08). In the HDER, the primary endpoint events were similar in the 2 groups (1.31%/year on SAPT vs 1.42%/year not on SAPT; P=0.61; Figure 3), and for the LDER they were 1.94 (on SAPT) and 1.78 (not on SAPT; P=0.49; Table 3).

Figure 3

Efficacy endpoints of high dose edoxaban strategy vs warfarin in patients with and without antiplatelet therapy. CV death indicates cardiovascular death; Edox, edoxaban; HR, hazard ratio; SEE, systemic embolic event; Warf, warfarin.

Table 3

Efficacy Endpoints of Low Dose Edoxaban Strategy

Outcome	Annualized Event Rate (%/year)		LDE vs WAR
	LDE	WAR	HR (95% CI)	P int
Stroke/SEE
No antiplatelet	1.78	1.49	1.19 (0.99–1.43)
Antiplatelet	1.94	1.88	1.03 (0.76–1.39)	0.42
Ischemic stroke
No antiplatelet	1.55	1.07	1.44 (1.17–1.78)
Antiplatelet	1.64	1.19	1.37 (0.96–1.96)	0.83
Hemorrhagic stroke
No antiplatelet	0.14	0.36	0.38 (0.22–0.65)
Antiplatelet	0.18	0.61	0.29 (0.13–0.64)	0.59
Myocardial infarction
No antiplatelet	0.76	0.59	1.29 (0.97–1.72)
Antiplatelet	0.89	0.94	0.95 (0.61–1.47)	0.24
Cardiovascular death
No antiplatelet	2.34	2.61	0.89 (0.77–1.04)
Antiplatelet	2.69	3.56	0.75 (0.59–0.95)	0.21

HR indicates adjusted hazard ratio; LDE, low dose edoxaban strategy; SEE, systemic embolic event; WAR, warfarin.

In the warfarin arm, cardiovascular death occurred more frequently in patients receiving SAPT (3.56%/year) than not receiving it (2.61%/year; P=0.001). In HDER, corresponding values were 2.94%/year versus 2.12%/year (P=0.002), respectively (Figure 3), whereas in LDER these values were 2.69% and 2.34% (P=0.19; Table 3).

In the warfarin arm the primary safety endpoint (ISTH major bleeding) occurred more frequently in patients receiving SAPT (4.38%/year) than not (2.54%/year; P<0.001). This was also the case for patients in the HDER (3.55% and 2.04%; P<0.001; Figure 4,) as well as in the LDER (2.23% and 1.41%, respectively; P=0.001; Table 4).

Figure 4

Bleeding endpoints and net clinical outcome of high dose edoxaban strategy vs warfarin in patients with and without antiplatelet therapy. Edox indicates edoxaban; HR, hazard ratio; SEE, systemic embolic event; Warf, warfarin.

Table 4

Bleeding Endpoint and Net Clinical Outcome of LDE

Safety	Annualized Event Rate (%/year)		LDE vs WAR
	LDE	WAR	HR (95% CI)	P int
Major bleeding
No antiplatelet	1.41	2.54	0.56 (0.46–0.67)
Antiplatelet	2.23	4.38	0.51 (0.39–0.66)	0.59
Fatal bleeding
No antiplatelet	0.11	0.24	0.47 (0.25–0.9)
Antiplatelet	0.13	0.56	0.23 (0.09–0.61)	0.22
Intracranial bleeding
No antiplatelet	0.23	0.57	0.40 (0.26–0.62)
Antiplatelet	0.21	1.18	0.18 (0.08–0.37)	0.07
Life‐threatening
No antiplatelet	0.34	0.64	0.52 (0.37–0.74)
Antiplatelet	0.31	1.14	0.28 (0.15–0.5)	0.07
Any bleeding
No antiplatelet	7.47	11.07	0.67 (0.62–0.74)
Antiplatelet	13.04	18.66	0.69 (0.61–0.79)	0.73
Net: death/stroke/SEE/major bleeding
No antiplatelet	5.81	6.50	0.89 (0.81–0.98)
Antiplatelet	6.91	9.47	0.72 (0.62–0.84)	0.02

HR indicates adjusted hazard ratio; LDE, low dose edoxaban strategy; SEE, systemic embolic event; WAR, warfarin.

Efficacy of Edoxaban Versus Warfarin Stratified by SAPT

HRsadj values for the primary endpoint (stroke/SEE) for HDER versus warfarin were 0.70 (95% CI, 0.50–0.98) in the SAPT group and 0.94 (95% CI, 0.77–1.15) in the non‐SAPT group (P int=0.14; Figure 3). Corresponding HRsadj values for the LDER versus warfarin comparisons were 1.03 (95% CI, 0.76–1.39) on SAPT and 1.19 (95% CI, 0.99–1.43) for not on SAPT, respectively (P int=0.42). Similar findings were noted for ischemic stroke (Table 3).

HRsadj values for HDER versus warfarin for cardiovascular death were similar for those in the SAPT group (HR=0.83; 95% CI, 0.66–1.05) and in the non‐SAPT group (HRadj=0.81; 95% CI, 0.69–0.94; P int=0.83; Figure 3). The corresponding comparisons of HRsadj in LDER versus warfarin were 0.75 (95% CI, 0.59–0.95) and 0.89 (95% CI, 0.77–1.04; P int=0.21.) (Table 3).

HRsadj values for safety of edoxaban versus warfarin stratified by SAPT for the primary safety endpoint (ISTH major bleeding) were 0.82 (95% CI, 0.65–1.03) in the SAPT group and 0.80 (95% CI, 0.68–0.95) in the non‐SAPT group (P int=0.91), whereas corresponding HRsadj value for the LDER versus warfarin were 0.51 (95% CI, 0.39–0.66) and 0.56 (95% CI, 0.46–0.67; P int=0.59). There were consistent reductions in bleeding (including ICH, life‐threatening bleeding, and fatal bleeding) with both edoxaban regimens compared to warfarin, with and without concomitant SAPT (Figure 4; Table 4).

The prespecified net clinical outcome, which consisted of both efficacy and safety endpoints, occurred significantly more frequently in patients in the SAPT group, in all 3 arms (warfarin and both edoxaban arms; Figure 4; Table 4). HRsadj value of HDER to warfarin were similar in the 2 groups as well, with HRsadj of 0.82 (95% CI, 0.71–0.95) and 0.89 (95% CI, 0.81–0.98) in the SAPT and non‐SAPT groups, respectively (P int=0.35; Figure 4). Corresponding values for HRsadj of LDER to warfarin were 0.72 (95% CI: 0.62–0.84) and 0.89 (95% CI, 0.81–0.98; P int=0.02; Table 4).

Sensitivity Analyses Stratified by SAPT at Randomization

The results of the sensitivity analysis for the comparison of outcomes of the edoxaban regimen compared to warfarin stratified by SAPT at randomization yielded similar results to those in the principal analysis, described above.

Findings With Aspirin

The results presented in Figures 3 and 4 and in Tables S2 and S3 for all patients receiving SAPT, were quite similar for the large subgroup (92.5%) of SAPT patients receiving aspirin (Tables S2 and S3).

Discussion

Current guidelines recommend that low‐dose aspirin (75–100 mg/day) and/or a P2Y12 antagonist may be given concurrently with an anticoagulant to prevent myocardial ischemic events and stroke in AF patients after coronary revascularization, a recent acute coronary syndrome or with high‐risk CAD. The associated increased risk of bleeding should be evaluated and efforts made to minimize it whenever possible.10, 14, 15, 16, 17, 18 Also, the American Heart Association/American College of Cardiology Foundation guideline for secondary prevention in patients with AF and coronary or other atherosclerotic vascular disease recommends treatment with warfarin and low‐dose aspirin (≤100 mg daily).14 However, for AF patients with stable coronary or peripheral arterial disease (ie, no acute events or revascularization for ≥12 months), oral anticoagulant therapy without antiplatelet therapy may be considered.10, 16 In clinical practice, a combination of anticoagulant and dual antiplatelet therapy (triple antithrombotic therapy) may be administered, preferably for short periods in patients with AF who are at very high risk of a platelet‐driven event, such as patients with a recent acute coronary syndrome or stent implantation.19 This increases the risk of serious bleeding and it should be reduced to double therapy (ie, an anticoagulant together with SAPT) whenever, or as soon as, possible.

The present report from the ENGAGE AF‐TIMI 48 trial provides data on the relative efficacy and safety of combination antithrombotic therapy. At the time of enrollment, approximately one‐third of patients were receiving SAPT, usually aspirin. SAPT was discontinued in one quarter of these patients after randomization to anticoagulant therapy. SAPT was prescribed by the treating physician more commonly in patients with established CAD, diabetes, dyslipidemia, peripheral arterial disease, and those who were therefore at higher risk of development of acute coronary syndromes than were patients without these comorbidities. SAPT administration was used most frequently in North America and less frequently elsewhere. This difference may be explained, at least in part, by the greater frequency of patients at high risk enrolled in North America.

We observed that the addition of SAPT to an anticoagulant (warfarin or edoxaban) was associated with a significantly greater risk of bleeding. However, the addition of SAPT did not modify the relative efficacy and safety of edoxaban as compared to warfarin. Notably, when compared to warfarin, both edoxaban regimens significantly reduced all forms of bleeding, including ICH and life‐threatening bleeding, both in patients who were as well as those who were not, receiving a SAPT.

The trade‐off between benefit and safety of adding SAPT to an anticoagulant in patients with both AF and CAD or others at risk of an acute coronary event is often challenging for clinicians.9, 10 A meta‐analysis of 10 randomized trials comparing the combination of an oral anticoagulant and aspirin with anticoagulant alone in patients with AF at risk of coronary events showed no reduction in arterial thromboembolic events in favor of the combination, but did show an increased risk of major bleeding.20 Lamberts et al. examined the efficacy and safety of adding aspirin to a VKA in AF patients with stable CAD in a nation‐wide Danish registry.8 Like the meta‐analysis, they found that the risk of coronary events with the combination was similar to that observed with VKA alone, whereas the risk of bleeding increased significantly when aspirin or clopidogrel were added to VKA. The WOEST (What is the Optimal antiplateElet and anticoagulant therapy in patients with oral anticoagulation and coronary StenTing) trial compared dual therapy (VKA plus clopidogrel) to triple therapy (VKA, clopidogrel and aspirin) in patients receiving oral anticoagulants undergoing PCI. Dual therapy was associated with a significant reduction in bleeding without an increase in rate of thrombotic events.21

Our results with edoxaban are generally consistent with earlier studies. The Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation (ARISTOTLE,) 22 and the Rivaroxaban Once‐daily, oral, direct Factor Xa inhibition Compared with vitamin K antagonism for prevention of stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF)23 trials showed that concomitant aspirin use did not alter the relative effects of apixaban and rivaroxaban on stroke/SEE and major bleeding compared to warfarin. In the Randomized Evaluation of Long‐term Anticoagulation Therapy (RE‐LY) trial, the addition of SAPT did not affect the safety and relative efficacy of dabigatran 110 mg BID when compared to warfarin. However, in contrast to our findings, the effect of dabigatran 150 mg BID on the reduction of stroke/SEE appeared to be attenuated among patients receiving an antiplatelet agent (HR, 0.80) in comparison to those who were not (HR=0.52, P int=0.058).24

Limitations

Patients with recent acute coronary syndrome, or recent stent implantation were excluded from this trial. Therefore, the results may be applicable only to patients in whom a single antiplatelet agent may be indicated. One of the limitations of this analysis is that it was based on SAPT at 3 months rather than at randomization, because a sizeable percentage of patients (25%) discontinued SAPT after they entered the ENGAGE AF trial. Therefore, the events occurring during the first 3 months post‐randomization were not included in the analysis reported herein. However, the sensitivity analysis which included all patients who entered the trial exhibited similar results. Administration of SAPT was not randomized, and although the analyses were adjusted for the baseline characteristics, such adjustments are never complete. In the future, randomized controlled trials on the outcomes of new oral anticoagulants with and without SAPT would be informative.

Conclusions

Patients with nonvalvular AF who were prescribed a single antiplatelet agent along with concomitant anticoagulant therapy had higher risks of bleeding than those who were prescribed only an anticoagulant. However, combination therapy did not alter the reduction in bleeding in both dose strategies of edoxaban compared with well‐managed warfarin. All forms of bleeding were highest in patients randomized to warfarin who were treated with a SAPT. Because of this finding, patients with AF who are deemed to require the addition of a SAPT should receive a Xa inhibitor for anticoagulation whenever possible.

Sources of Funding

The ENGAGE AF TIMI 48 trial was funded by Daiichi Sankyo Pharma Development, Edison, NJ).

Disclosures

Dr Xu, Dr Nordio and Ms Murphy have no disclosures. Drs Patel, Shi, and Mercuri are employees of Daiichi Sankyo, which funded this trial. Dr Ruff reports grant support through his institution from Daiichi‐Sankyo and has served as a consultant and received honoraria from Daiichi‐Sankyo, Boehringer Ingelheim, Bayer, and Portola; and grant support through his institution outside the submitted work from AstraZeneca, Eisai, Intarcia and GlaxoSmithKline (GSK). Dr Giugliano has served as a consultant and had received honoraria from Bristol‐Myers Squibb, Janssen, Daiichi‐Sankyo, Merck, and Sanofi; and grant support through his institution from Daiichi‐Sankyo, Merck, Johnson & Johnson, Sanofi, and AstraZeneca. Dr Antman reports receiving grant support through his institution from Daiichi‐Sankyo. Dr Braunwald reports grants (through the Brigham and Women's Hospital) and personal fees for lectures from Daiichi‐Sankyo. He has received grants from Duke University, AstraZeneca, Merck & Co, and GSK; uncompensated fees for consultancy from Merck & Co, personal fees for consultancies from Genyzme, Medicines Co, and Sanofi‐Aventis; uncompensated personal fees for lectures from Merck, and personal fees for lectures from Menarini International and Medscape.

Table S1. Prevalence of Aspirin Use at Randomization and 6 Time Points throughout the Study

Table S2. Efficacy Endpoints of Edoxaban vs Warfarin in Patients On or Not On Aspirin

Table S3. Bleeding Endpoints of Edoxaban vs Warfarin in Patients On or Not On Aspirin.

Click here for additional data file.